slurm submission log: 2024-05-29 11:13:55.587090
created following sbatch script: 

###############################

#!/bin/bash

#SBATCH --account=nlp
#SBATCH --cpus-per-task=16
#SBATCH --dependency=afterok:7667695
#SBATCH --gres=gpu:2
#SBATCH --job-name=tthrush-job-598030
#SBATCH --mem=100G
#SBATCH --nodelist=sphinx2
#SBATCH --open-mode=append
#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1/train_job_output.txt
#SBATCH --partition=sphinx
#SBATCH --time=14-0

# activate your desired anaconda environment
. /nlp/scr/tthrush/miniconda3/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection

# cd to working directory
cd .

# launch commands
srun --unbuffered run_as_child_processes 'torchrun --master_port 29528 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/lambada --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1 --output_hub_id pythia-70m_lambada --model_id EleutherAI/pythia-70m --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2 --per_device_train_batch_size 256 --seed 1 --num_train_epochs 14'

###############################

submission to slurm complete!


###############################
slurm submission output

Submitted batch job 7667696



###############################

slurm submission log: 2024-05-30 08:40:48.786099
created following sbatch script: 

###############################

#!/bin/bash

#SBATCH --account=nlp
#SBATCH --cpus-per-task=16
#SBATCH --gres=gpu:2
#SBATCH --job-name=tthrush-job-2044262
#SBATCH --mem=100G
#SBATCH --nodelist=sphinx2
#SBATCH --open-mode=append
#SBATCH --output=/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1/train_job_output.txt
#SBATCH --partition=sphinx
#SBATCH --time=14-0

# activate your desired anaconda environment
. /nlp/scr/tthrush/miniconda3/etc/profile.d/conda.sh ; conda activate pretraining-coreset-selection

# cd to working directory
cd .

# launch commands
srun --unbuffered run_as_child_processes 'torchrun --master_port 29510 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/lambada --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1 --output_hub_id pythia-70m_lambada --model_id EleutherAI/pythia-70m --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2 --per_device_train_batch_size 256 --seed 1 --num_train_epochs 1'

###############################

submission to slurm complete!


###############################
slurm submission output

Submitted batch job 7670605



###############################

###############################
start time: 2024-05-30 18:12:24.450817
machine: sphinx2
conda env: pretraining-coreset-selection
###############################
running following processes

	torchrun --master_port 29510 --nproc_per_node=2 train_llm.py --dataset_id /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/lambada --output_dir /juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1 --output_hub_id pythia-70m_lambada --model_id EleutherAI/pythia-70m --learning_rate 1e-3 --warmup_ratio=0.1 --gradient_accumulation_steps 2 --per_device_train_batch_size 256 --seed 1 --num_train_epochs 1


###############################
command outputs: 


[2024-05-30 18:12:31,541] torch.distributed.run: [WARNING] 
[2024-05-30 18:12:31,541] torch.distributed.run: [WARNING] *****************************************
[2024-05-30 18:12:31,541] torch.distributed.run: [WARNING] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. 
[2024-05-30 18:12:31,541] torch.distributed.run: [WARNING] *****************************************
05/30/2024 18:12:52 - INFO - __main__ - Script parameters ScriptArguments(seed=1, dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/lambada', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1', output_hub_id='pythia-70m_lambada', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1.0, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
05/30/2024 18:12:52 - INFO - __main__ - Script parameters ScriptArguments(seed=1, dataset_id='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/data/lambada', output_dir='/juice5/scr5/tthrush/pretraining-coreset-selection/llm_pretraining/paper_writeup_tests/ordinal_constrained/llms/pythia-70m_lambada_1', output_hub_id='pythia-70m_lambada', hf_hub_token=True, model_id='EleutherAI/pythia-70m', per_device_train_batch_size=256, num_train_epochs=1.0, learning_rate=0.001, gradient_accumulation_steps=2, from_scratch=True, warmup_ratio=0.1, adam_beta1=0.9, adam_beta2=0.95, adam_epsilon=1e-08, weight_decay=0.01, lr_scheduler_type='cosine', local_rank=0, resume_from_checkpoint=False, deepspeed=None, peft=False)
/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
  warnings.warn(
/nlp/scr/tthrush/miniconda3/envs/pretraining-coreset-selection/lib/python3.10/site-packages/huggingface_hub/file_download.py:1132: FutureWarning: `resume_download` is deprecated and will be removed in version 1.0.0. Downloads always resume when possible. If you want to force a new download, use `force_download=True`.
  warnings.warn(

  0%|          | 0/10708 [00:00<?, ?it/s][rank0]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())
[rank1]:[W reducer.cpp:1360] Warning: find_unused_parameters=True was specified in DDP constructor, but did not find any unused parameters in the forward pass. This flag results in an extra traversal of the autograd graph every iteration,  which can adversely affect performance. If your model indeed never has any unused parameters in the forward pass, consider turning this flag off. Note that this warning may be a false positive if your model has flow control causing later iterations to have unused parameters. (function operator())

  0%|          | 1/10708 [00:52<157:19:31, 52.90s/it]
  0%|          | 2/10708 [01:25<121:46:56, 40.95s/it]
  0%|          | 3/10708 [01:59<112:18:29, 37.77s/it]
  0%|          | 4/10708 [02:27<100:35:45, 33.83s/it]
  0%|          | 5/10708 [02:52<91:11:54, 30.68s/it] 
  0%|          | 6/10708 [03:16<84:19:41, 28.37s/it]
  0%|          | 7/10708 [03:38<78:30:24, 26.41s/it]
  0%|          | 8/10708 [04:02<76:10:14, 25.63s/it]
  0%|          | 9/10708 [04:22<71:03:55, 23.91s/it]
  0%|          | 10/10708 [04:43<68:34:27, 23.08s/it]
  0%|          | 11/10708 [05:00<62:45:52, 21.12s/it]
  0%|          | 12/10708 [05:19<60:37:12, 20.40s/it]
  0%|          | 13/10708 [05:39<60:22:41, 20.32s/it]
  0%|          | 14/10708 [06:00<61:18:33, 20.64s/it]
  0%|          | 15/10708 [06:22<61:58:57, 20.87s/it]
  0%|          | 16/10708 [06:38<57:25:50, 19.34s/it]
  0%|          | 17/10708 [06:53<54:20:05, 18.30s/it]
  0%|          | 18/10708 [07:06<48:49:47, 16.44s/it]
  0%|          | 19/10708 [07:18<45:34:54, 15.35s/it]
  0%|          | 20/10708 [07:32<43:53:32, 14.78s/it]
  0%|          | 21/10708 [07:44<41:48:13, 14.08s/it]
  0%|          | 22/10708 [07:57<40:26:58, 13.63s/it]
  0%|          | 23/10708 [08:08<38:24:10, 12.94s/it]
  0%|          | 24/10708 [08:19<36:06:41, 12.17s/it]
  0%|          | 25/10708 [08:29<34:36:32, 11.66s/it]
                                                     
{'loss': 10.6713, 'grad_norm': 1.3470221757888794, 'learning_rate': 2.3342670401493933e-05, 'epoch': 0.0}

  0%|          | 25/10708 [08:29<34:36:32, 11.66s/it]
  0%|          | 26/10708 [08:37<31:40:44, 10.68s/it]
  0%|          | 27/10708 [08:45<28:51:13,  9.73s/it]
  0%|          | 28/10708 [08:51<25:38:07,  8.64s/it]
  0%|          | 29/10708 [08:58<23:53:56,  8.06s/it]
  0%|          | 30/10708 [09:04<22:28:05,  7.58s/it]
  0%|          | 31/10708 [09:10<21:07:05,  7.12s/it]
  0%|          | 32/10708 [09:16<19:41:58,  6.64s/it]
  0%|          | 33/10708 [09:21<18:50:13,  6.35s/it]
  0%|          | 34/10708 [09:27<18:04:41,  6.10s/it]
  0%|          | 35/10708 [09:32<16:56:23,  5.71s/it]
  0%|          | 36/10708 [09:37<16:19:24,  5.51s/it]
  0%|          | 37/10708 [09:42<16:03:24,  5.42s/it]
  0%|          | 38/10708 [09:47<15:26:55,  5.21s/it]
  0%|          | 39/10708 [09:51<14:34:58,  4.92s/it]
  0%|          | 40/10708 [09:55<13:53:58,  4.69s/it]
  0%|          | 41/10708 [10:02<15:27:12,  5.22s/it]
  0%|          | 42/10708 [10:40<44:54:31, 15.16s/it]
  0%|          | 43/10708 [10:47<37:40:05, 12.71s/it]
  0%|          | 44/10708 [11:06<42:53:46, 14.48s/it]
  0%|          | 45/10708 [11:34<55:38:42, 18.79s/it]
  0%|          | 46/10708 [11:41<44:43:15, 15.10s/it]
  0%|          | 47/10708 [12:17<63:33:00, 21.46s/it]
  0%|          | 48/10708 [12:35<60:34:17, 20.46s/it]
  0%|          | 49/10708 [12:39<45:29:55, 15.37s/it]
  0%|          | 50/10708 [12:42<34:42:55, 11.73s/it]{'loss': 9.9592, 'grad_norm': 1.214390754699707, 'learning_rate': 4.6685340802987865e-05, 'epoch': 0.0}
                                                     

  0%|          | 50/10708 [12:42<34:42:55, 11.73s/it]
  0%|          | 51/10708 [12:45<27:11:39,  9.19s/it]
  0%|          | 52/10708 [12:49<22:08:04,  7.48s/it]
  0%|          | 53/10708 [12:52<18:08:49,  6.13s/it]
  1%|          | 54/10708 [12:55<15:39:06,  5.29s/it]
  1%|          | 55/10708 [12:58<13:45:57,  4.65s/it]
  1%|          | 56/10708 [13:01<12:20:30,  4.17s/it]
  1%|          | 57/10708 [13:05<11:40:14,  3.94s/it]
  1%|          | 58/10708 [13:08<10:59:16,  3.71s/it]
  1%|          | 59/10708 [13:10<9:59:50,  3.38s/it] 
  1%|          | 60/10708 [13:13<9:29:51,  3.21s/it]
  1%|          | 61/10708 [13:16<9:08:57,  3.09s/it]
  1%|          | 62/10708 [13:19<8:42:46,  2.95s/it]
  1%|          | 63/10708 [13:21<8:20:05,  2.82s/it]
  1%|          | 64/10708 [13:24<8:03:33,  2.73s/it]
  1%|          | 65/10708 [13:26<7:41:28,  2.60s/it]
  1%|          | 66/10708 [13:29<8:20:04,  2.82s/it]
  1%|          | 67/10708 [13:32<8:14:32,  2.79s/it]
  1%|          | 68/10708 [13:34<7:49:04,  2.65s/it]
  1%|          | 69/10708 [13:36<7:17:39,  2.47s/it]
  1%|          | 70/10708 [13:39<7:38:51,  2.59s/it]
  1%|          | 71/10708 [13:42<7:43:06,  2.61s/it]
  1%|          | 72/10708 [13:44<7:19:36,  2.48s/it]
  1%|          | 73/10708 [13:46<7:07:49,  2.41s/it]
  1%|          | 74/10708 [13:49<6:50:37,  2.32s/it]
  1%|          | 75/10708 [13:51<6:39:55,  2.26s/it]
                                                    
{'loss': 9.2468, 'grad_norm': 1.0271425247192383, 'learning_rate': 7.00280112044818e-05, 'epoch': 0.01}

  1%|          | 75/10708 [13:51<6:39:55,  2.26s/it]
  1%|          | 76/10708 [13:53<6:45:56,  2.29s/it]
  1%|          | 77/10708 [13:55<6:39:11,  2.25s/it]
  1%|          | 78/10708 [13:57<6:37:03,  2.24s/it]
  1%|          | 79/10708 [13:59<6:31:14,  2.21s/it]
  1%|          | 80/10708 [14:02<6:37:58,  2.25s/it]
  1%|          | 81/10708 [14:04<6:22:22,  2.16s/it]
  1%|          | 82/10708 [14:06<6:09:09,  2.08s/it]
  1%|          | 83/10708 [14:08<6:01:09,  2.04s/it]
  1%|          | 84/10708 [14:09<5:46:36,  1.96s/it]
  1%|          | 85/10708 [14:11<5:35:46,  1.90s/it]
  1%|          | 86/10708 [14:13<5:34:55,  1.89s/it]
  1%|          | 87/10708 [14:15<5:33:01,  1.88s/it]
  1%|          | 88/10708 [14:17<5:58:30,  2.03s/it]
  1%|          | 89/10708 [14:19<5:43:13,  1.94s/it]
  1%|          | 90/10708 [14:21<5:39:21,  1.92s/it]
  1%|          | 91/10708 [14:23<5:36:07,  1.90s/it]
  1%|          | 92/10708 [14:25<5:44:03,  1.94s/it]
  1%|          | 93/10708 [14:27<5:57:46,  2.02s/it]
  1%|          | 94/10708 [14:28<5:21:09,  1.82s/it]
  1%|          | 95/10708 [14:30<5:09:23,  1.75s/it]
  1%|          | 96/10708 [14:31<4:50:00,  1.64s/it]
  1%|          | 97/10708 [14:33<5:07:19,  1.74s/it]
  1%|          | 98/10708 [14:35<4:45:56,  1.62s/it]
  1%|          | 99/10708 [14:36<4:25:51,  1.50s/it]
  1%|          | 100/10708 [14:37<4:32:52,  1.54s/it]{'loss': 8.4707, 'grad_norm': 0.7098316550254822, 'learning_rate': 9.337068160597573e-05, 'epoch': 0.01}

                                                     

  1%|          | 100/10708 [14:37<4:32:52,  1.54s/it]
  1%|          | 101/10708 [14:39<4:21:56,  1.48s/it]
  1%|          | 102/10708 [14:40<4:15:29,  1.45s/it]
  1%|          | 103/10708 [14:42<4:22:08,  1.48s/it]
  1%|          | 104/10708 [14:44<4:39:59,  1.58s/it]
  1%|          | 105/10708 [14:45<4:42:55,  1.60s/it]
  1%|          | 106/10708 [14:47<4:41:26,  1.59s/it]
  1%|          | 107/10708 [14:48<4:37:35,  1.57s/it]
  1%|          | 108/10708 [14:50<4:37:50,  1.57s/it]
  1%|          | 109/10708 [14:51<4:34:54,  1.56s/it]
  1%|          | 110/10708 [14:53<4:28:11,  1.52s/it]
  1%|          | 111/10708 [14:54<4:29:49,  1.53s/it]
  1%|          | 112/10708 [14:56<4:34:03,  1.55s/it]
  1%|          | 113/10708 [14:58<4:33:53,  1.55s/it]
  1%|          | 114/10708 [14:59<4:32:30,  1.54s/it]
  1%|          | 115/10708 [15:00<4:27:47,  1.52s/it]
  1%|          | 116/10708 [15:02<4:23:38,  1.49s/it]
  1%|          | 117/10708 [15:03<4:26:53,  1.51s/it]
  1%|          | 118/10708 [15:05<4:14:01,  1.44s/it]
  1%|          | 119/10708 [15:06<4:01:55,  1.37s/it]
  1%|          | 120/10708 [15:07<3:58:57,  1.35s/it]
  1%|          | 121/10708 [15:09<4:00:33,  1.36s/it]
  1%|          | 122/10708 [15:10<3:57:18,  1.35s/it]
  1%|          | 123/10708 [15:11<3:44:24,  1.27s/it]
  1%|          | 124/10708 [15:17<7:24:58,  2.52s/it]
  1%|          | 125/10708 [15:19<7:42:14,  2.62s/it]{'loss': 7.8414, 'grad_norm': 0.4142545163631439, 'learning_rate': 0.00011671335200746966, 'epoch': 0.01}

                                                     

  1%|          | 125/10708 [15:19<7:42:14,  2.62s/it]
  1%|          | 126/10708 [15:21<6:27:31,  2.20s/it]
  1%|          | 127/10708 [15:22<5:31:20,  1.88s/it]
  1%|          | 128/10708 [15:31<12:06:33,  4.12s/it]
  1%|          | 129/10708 [15:32<9:42:45,  3.31s/it] 
  1%|          | 130/10708 [15:34<7:57:51,  2.71s/it]
  1%|          | 131/10708 [15:35<6:47:04,  2.31s/it]
  1%|          | 132/10708 [15:36<5:45:17,  1.96s/it]
  1%|          | 133/10708 [15:38<5:19:12,  1.81s/it]
  1%|▏         | 134/10708 [15:39<4:41:10,  1.60s/it]
  1%|▏         | 135/10708 [15:40<4:09:51,  1.42s/it]
  1%|▏         | 136/10708 [15:41<3:48:04,  1.29s/it]
  1%|▏         | 137/10708 [15:42<3:54:16,  1.33s/it]
  1%|▏         | 138/10708 [15:43<3:35:22,  1.22s/it]
  1%|▏         | 139/10708 [15:44<3:31:13,  1.20s/it]
  1%|▏         | 140/10708 [15:45<3:24:08,  1.16s/it]
  1%|▏         | 141/10708 [15:47<3:20:06,  1.14s/it]
  1%|▏         | 142/10708 [15:48<3:16:29,  1.12s/it]
  1%|▏         | 143/10708 [15:49<3:11:38,  1.09s/it]
  1%|▏         | 144/10708 [15:50<3:10:30,  1.08s/it]
  1%|▏         | 145/10708 [15:51<3:11:46,  1.09s/it]
  1%|▏         | 146/10708 [15:52<3:16:29,  1.12s/it]
  1%|▏         | 147/10708 [15:53<3:13:54,  1.10s/it]
  1%|▏         | 148/10708 [15:54<3:15:13,  1.11s/it]
  1%|▏         | 149/10708 [15:55<3:12:34,  1.09s/it]
  1%|▏         | 150/10708 [15:56<3:05:32,  1.05s/it]
                                                     
{'loss': 7.3608, 'grad_norm': 0.5246418118476868, 'learning_rate': 0.0001400560224089636, 'epoch': 0.01}

  1%|▏         | 150/10708 [15:56<3:05:32,  1.05s/it]
  1%|▏         | 151/10708 [15:57<3:07:11,  1.06s/it]
  1%|▏         | 152/10708 [15:58<3:02:04,  1.03s/it]
  1%|▏         | 153/10708 [15:59<3:00:45,  1.03s/it]
  1%|▏         | 154/10708 [16:00<3:02:06,  1.04s/it]
  1%|▏         | 155/10708 [16:01<3:08:43,  1.07s/it]
  1%|▏         | 156/10708 [16:03<3:07:41,  1.07s/it]
  1%|▏         | 157/10708 [16:04<3:07:03,  1.06s/it]
  1%|▏         | 158/10708 [16:05<3:06:27,  1.06s/it]
  1%|▏         | 159/10708 [16:06<2:58:16,  1.01s/it]
  1%|▏         | 160/10708 [16:06<2:50:57,  1.03it/s]
  2%|▏         | 161/10708 [16:07<2:51:54,  1.02it/s]
  2%|▏         | 162/10708 [16:09<2:59:34,  1.02s/it]
  2%|▏         | 163/10708 [16:10<2:58:39,  1.02s/it]
  2%|▏         | 164/10708 [16:11<2:57:27,  1.01s/it]
  2%|▏         | 165/10708 [16:11<2:44:09,  1.07it/s]
  2%|▏         | 166/10708 [16:12<2:33:53,  1.14it/s]
  2%|▏         | 167/10708 [16:13<2:41:21,  1.09it/s]
  2%|▏         | 168/10708 [16:14<2:55:50,  1.00s/it]
  2%|▏         | 169/10708 [16:15<2:41:23,  1.09it/s]
  2%|▏         | 170/10708 [16:16<2:47:16,  1.05it/s]
  2%|▏         | 171/10708 [16:17<2:47:14,  1.05it/s]
  2%|▏         | 172/10708 [16:18<2:44:28,  1.07it/s]
  2%|▏         | 173/10708 [16:19<2:34:35,  1.14it/s]
  2%|▏         | 174/10708 [16:20<2:37:00,  1.12it/s]
  2%|▏         | 175/10708 [16:20<2:38:03,  1.11it/s]
                                                     
{'loss': 6.9549, 'grad_norm': 0.4910980463027954, 'learning_rate': 0.00016339869281045753, 'epoch': 0.02}

  2%|▏         | 175/10708 [16:20<2:38:03,  1.11it/s]
  2%|▏         | 176/10708 [16:21<2:35:37,  1.13it/s]
  2%|▏         | 177/10708 [16:22<2:35:25,  1.13it/s]
  2%|▏         | 178/10708 [16:23<2:38:23,  1.11it/s]
  2%|▏         | 179/10708 [16:24<2:39:33,  1.10it/s]
  2%|▏         | 180/10708 [16:25<2:41:39,  1.09it/s]
  2%|▏         | 181/10708 [16:26<2:46:00,  1.06it/s]
  2%|▏         | 182/10708 [16:27<2:31:52,  1.16it/s]
  2%|▏         | 183/10708 [16:28<2:31:51,  1.16it/s]
  2%|▏         | 184/10708 [16:28<2:34:24,  1.14it/s]
  2%|▏         | 185/10708 [16:29<2:34:25,  1.14it/s]
  2%|▏         | 186/10708 [16:30<2:25:52,  1.20it/s]
  2%|▏         | 187/10708 [16:31<2:29:53,  1.17it/s]
  2%|▏         | 188/10708 [16:32<2:43:57,  1.07it/s]
  2%|▏         | 189/10708 [16:33<2:45:31,  1.06it/s]
  2%|▏         | 190/10708 [16:34<2:42:47,  1.08it/s]
  2%|▏         | 191/10708 [16:35<2:29:26,  1.17it/s]
  2%|▏         | 192/10708 [16:35<2:20:53,  1.24it/s]
  2%|▏         | 193/10708 [16:36<2:17:20,  1.28it/s]
  2%|▏         | 194/10708 [16:37<2:18:53,  1.26it/s]
  2%|▏         | 195/10708 [16:38<2:22:13,  1.23it/s]
  2%|▏         | 196/10708 [16:39<2:20:33,  1.25it/s]
  2%|▏         | 197/10708 [16:39<2:15:30,  1.29it/s]
  2%|▏         | 198/10708 [16:40<2:06:58,  1.38it/s]
  2%|▏         | 199/10708 [16:40<2:02:48,  1.43it/s]
  2%|▏         | 200/10708 [16:41<2:07:57,  1.37it/s]{'loss': 6.618, 'grad_norm': 0.3432289659976959, 'learning_rate': 0.00018674136321195146, 'epoch': 0.02}
                                                     

  2%|▏         | 200/10708 [16:41<2:07:57,  1.37it/s]
  2%|▏         | 201/10708 [16:42<2:26:49,  1.19it/s]
  2%|▏         | 202/10708 [16:43<2:27:13,  1.19it/s]
  2%|▏         | 203/10708 [16:44<2:27:24,  1.19it/s]
  2%|▏         | 204/10708 [16:45<2:14:52,  1.30it/s]
  2%|▏         | 205/10708 [16:45<2:14:37,  1.30it/s]
  2%|▏         | 206/10708 [16:46<2:18:51,  1.26it/s]
  2%|▏         | 207/10708 [16:47<2:12:49,  1.32it/s]
  2%|▏         | 208/10708 [16:48<2:07:20,  1.37it/s]
  2%|▏         | 209/10708 [17:37<44:20:37, 15.21s/it]
  2%|▏         | 210/10708 [18:37<83:51:32, 28.76s/it]
  2%|▏         | 211/10708 [19:10<87:27:00, 29.99s/it]
  2%|▏         | 212/10708 [20:14<116:59:46, 40.13s/it]
  2%|▏         | 213/10708 [21:28<147:04:07, 50.45s/it]
  2%|▏         | 214/10708 [22:37<163:33:22, 56.11s/it]
  2%|▏         | 215/10708 [23:17<149:15:07, 51.21s/it]
  2%|▏         | 216/10708 [24:56<191:15:09, 65.62s/it]
  2%|▏         | 217/10708 [25:38<170:07:50, 58.38s/it]
  2%|▏         | 218/10708 [26:39<172:18:11, 59.13s/it]
  2%|▏         | 219/10708 [28:12<201:53:44, 69.29s/it]
  2%|▏         | 220/10708 [28:53<177:11:27, 60.82s/it]
  2%|▏         | 221/10708 [30:07<188:40:19, 64.77s/it]
  2%|▏         | 222/10708 [30:28<150:55:01, 51.81s/it]
  2%|▏         | 223/10708 [31:08<140:02:54, 48.09s/it]
  2%|▏         | 224/10708 [31:54<138:42:35, 47.63s/it]
  2%|▏         | 225/10708 [32:57<151:29:35, 52.02s/it]{'loss': 6.3517, 'grad_norm': 0.48459678888320923, 'learning_rate': 0.0002100840336134454, 'epoch': 0.02}
                                                       

  2%|▏         | 225/10708 [32:57<151:29:35, 52.02s/it]
  2%|▏         | 226/10708 [33:15<122:13:35, 41.98s/it]
  2%|▏         | 227/10708 [33:56<121:32:34, 41.75s/it]
  2%|▏         | 228/10708 [34:21<106:37:10, 36.63s/it]
  2%|▏         | 229/10708 [35:08<115:12:50, 39.58s/it]
  2%|▏         | 230/10708 [35:38<106:47:40, 36.69s/it]
  2%|▏         | 231/10708 [36:04<98:09:48, 33.73s/it] 
  2%|▏         | 232/10708 [36:36<96:31:49, 33.17s/it]
  2%|▏         | 233/10708 [37:05<92:29:48, 31.79s/it]
  2%|▏         | 234/10708 [37:07<66:51:27, 22.98s/it]
  2%|▏         | 235/10708 [37:09<48:02:52, 16.52s/it]
  2%|▏         | 236/10708 [37:15<39:17:32, 13.51s/it]
  2%|▏         | 237/10708 [37:17<29:02:25,  9.98s/it]
  2%|▏         | 238/10708 [37:19<21:47:39,  7.49s/it]
  2%|▏         | 239/10708 [37:20<16:30:00,  5.67s/it]
  2%|▏         | 240/10708 [37:21<12:07:27,  4.17s/it]
  2%|▏         | 241/10708 [37:21<9:05:13,  3.13s/it] 
  2%|▏         | 242/10708 [37:22<6:57:14,  2.39s/it]
  2%|▏         | 243/10708 [37:24<6:31:25,  2.24s/it]
  2%|▏         | 244/10708 [37:26<6:25:53,  2.21s/it]
  2%|▏         | 245/10708 [37:28<6:17:55,  2.17s/it]
  2%|▏         | 246/10708 [37:30<6:01:04,  2.07s/it]
  2%|▏         | 247/10708 [37:32<5:32:30,  1.91s/it]
  2%|▏         | 248/10708 [37:32<4:21:59,  1.50s/it]
  2%|▏         | 249/10708 [37:33<3:40:50,  1.27s/it]
  2%|▏         | 250/10708 [37:33<3:05:42,  1.07s/it]
                                                     {'loss': 6.1397, 'grad_norm': 0.6109344363212585, 'learning_rate': 0.00023342670401493932, 'epoch': 0.02}


  2%|▏         | 250/10708 [37:33<3:05:42,  1.07s/it]
  2%|▏         | 251/10708 [37:34<2:52:45,  1.01it/s]
  2%|▏         | 252/10708 [37:35<2:32:15,  1.14it/s]
  2%|▏         | 253/10708 [37:36<2:27:16,  1.18it/s]
  2%|▏         | 254/10708 [37:36<2:23:42,  1.21it/s]
  2%|▏         | 255/10708 [37:37<2:10:56,  1.33it/s]
  2%|▏         | 256/10708 [37:38<2:26:48,  1.19it/s]
  2%|▏         | 257/10708 [37:39<2:19:45,  1.25it/s]
  2%|▏         | 258/10708 [37:39<2:12:21,  1.32it/s]
  2%|▏         | 259/10708 [37:40<2:04:00,  1.40it/s]
  2%|▏         | 260/10708 [37:41<2:05:28,  1.39it/s]
  2%|▏         | 261/10708 [37:41<2:07:42,  1.36it/s]
  2%|▏         | 262/10708 [37:42<2:00:10,  1.45it/s]
  2%|▏         | 263/10708 [37:43<2:01:01,  1.44it/s]
  2%|▏         | 264/10708 [37:43<2:00:43,  1.44it/s]
  2%|▏         | 265/10708 [37:44<2:00:16,  1.45it/s]
  2%|▏         | 266/10708 [37:45<2:08:05,  1.36it/s]
  2%|▏         | 267/10708 [37:46<2:06:40,  1.37it/s]
  3%|▎         | 268/10708 [37:46<2:05:50,  1.38it/s]
  3%|▎         | 269/10708 [37:47<2:02:12,  1.42it/s]
  3%|▎         | 270/10708 [37:48<2:04:41,  1.40it/s]
  3%|▎         | 271/10708 [37:48<1:59:36,  1.45it/s]
  3%|▎         | 272/10708 [37:49<1:57:06,  1.49it/s]
  3%|▎         | 273/10708 [37:50<1:58:35,  1.47it/s]
  3%|▎         | 274/10708 [37:50<1:58:40,  1.47it/s]
  3%|▎         | 275/10708 [37:51<2:00:49,  1.44it/s]
                                                     {'loss': 5.9724, 'grad_norm': 0.5953487157821655, 'learning_rate': 0.0002567693744164332, 'epoch': 0.03}

  3%|▎         | 275/10708 [37:51<2:00:49,  1.44it/s]
  3%|▎         | 276/10708 [37:52<1:57:19,  1.48it/s]
  3%|▎         | 277/10708 [37:53<1:58:31,  1.47it/s]
  3%|▎         | 278/10708 [37:53<1:58:38,  1.47it/s]
  3%|▎         | 279/10708 [37:54<2:03:25,  1.41it/s]
  3%|▎         | 280/10708 [37:55<1:59:10,  1.46it/s]
  3%|▎         | 281/10708 [37:55<1:58:22,  1.47it/s]
  3%|▎         | 282/10708 [37:56<2:00:55,  1.44it/s]
  3%|▎         | 283/10708 [37:57<1:55:19,  1.51it/s]
  3%|▎         | 284/10708 [37:57<1:52:30,  1.54it/s]
  3%|▎         | 285/10708 [37:58<1:49:52,  1.58it/s]
  3%|▎         | 286/10708 [37:58<1:50:06,  1.58it/s]
  3%|▎         | 287/10708 [37:59<1:50:15,  1.58it/s]
  3%|▎         | 288/10708 [38:00<1:52:29,  1.54it/s]
  3%|▎         | 289/10708 [38:01<1:58:06,  1.47it/s]
  3%|▎         | 290/10708 [38:01<1:59:21,  1.45it/s]
  3%|▎         | 291/10708 [38:02<2:01:52,  1.42it/s]
  3%|▎         | 292/10708 [38:03<1:55:20,  1.51it/s]
  3%|▎         | 293/10708 [38:03<1:52:02,  1.55it/s]
  3%|▎         | 294/10708 [38:04<1:51:25,  1.56it/s]
  3%|▎         | 295/10708 [38:04<1:53:48,  1.53it/s]
  3%|▎         | 296/10708 [38:05<1:51:56,  1.55it/s]
  3%|▎         | 297/10708 [38:06<2:02:33,  1.42it/s]
  3%|▎         | 298/10708 [38:07<1:56:18,  1.49it/s]
  3%|▎         | 299/10708 [38:07<1:57:17,  1.48it/s]
  3%|▎         | 300/10708 [38:08<1:55:35,  1.50it/s]
                                                     
{'loss': 5.8248, 'grad_norm': 0.6946554183959961, 'learning_rate': 0.0002801120448179272, 'epoch': 0.03}

  3%|▎         | 300/10708 [38:08<1:55:35,  1.50it/s]
  3%|▎         | 301/10708 [38:08<1:53:24,  1.53it/s]
  3%|▎         | 302/10708 [38:09<1:54:31,  1.51it/s]
  3%|▎         | 303/10708 [38:10<1:53:16,  1.53it/s]
  3%|▎         | 304/10708 [38:10<1:52:59,  1.53it/s]
  3%|▎         | 305/10708 [38:11<1:51:15,  1.56it/s]
  3%|▎         | 306/10708 [38:12<2:00:45,  1.44it/s]
  3%|▎         | 307/10708 [38:13<1:59:10,  1.45it/s]
  3%|▎         | 308/10708 [38:13<1:53:34,  1.53it/s]
  3%|▎         | 309/10708 [38:14<1:51:24,  1.56it/s]
  3%|▎         | 310/10708 [38:14<1:57:21,  1.48it/s]
  3%|▎         | 311/10708 [38:15<1:56:16,  1.49it/s]
  3%|▎         | 312/10708 [38:16<1:51:26,  1.55it/s]
  3%|▎         | 313/10708 [38:16<1:52:47,  1.54it/s]
  3%|▎         | 314/10708 [38:17<1:56:28,  1.49it/s]
  3%|▎         | 315/10708 [38:18<1:51:43,  1.55it/s]
  3%|▎         | 316/10708 [38:18<1:51:17,  1.56it/s]
  3%|▎         | 317/10708 [38:19<1:53:41,  1.52it/s]
  3%|▎         | 318/10708 [38:20<1:48:31,  1.60it/s]
  3%|▎         | 319/10708 [38:20<1:49:52,  1.58it/s]
  3%|▎         | 320/10708 [38:21<1:44:52,  1.65it/s]
  3%|▎         | 321/10708 [38:22<2:00:22,  1.44it/s]
  3%|▎         | 322/10708 [38:23<2:07:58,  1.35it/s]
  3%|▎         | 323/10708 [38:23<1:57:57,  1.47it/s]
  3%|▎         | 324/10708 [38:24<1:50:58,  1.56it/s]
  3%|▎         | 325/10708 [38:24<1:48:34,  1.59it/s]{'loss': 5.6989, 'grad_norm': 0.5192378163337708, 'learning_rate': 0.0003034547152194211, 'epoch': 0.03}
                                                     

  3%|▎         | 325/10708 [38:24<1:48:34,  1.59it/s]
  3%|▎         | 326/10708 [38:25<1:45:49,  1.63it/s]
  3%|▎         | 327/10708 [38:25<1:45:56,  1.63it/s]
  3%|▎         | 328/10708 [38:26<1:43:27,  1.67it/s]
  3%|▎         | 329/10708 [38:27<1:45:13,  1.64it/s]
  3%|▎         | 330/10708 [38:27<1:41:34,  1.70it/s]
  3%|▎         | 331/10708 [38:28<1:42:10,  1.69it/s]
  3%|▎         | 332/10708 [38:28<1:44:35,  1.65it/s]
  3%|▎         | 333/10708 [38:29<1:51:07,  1.56it/s]
  3%|▎         | 334/10708 [38:30<1:47:43,  1.61it/s]
  3%|▎         | 335/10708 [38:30<1:46:53,  1.62it/s]
  3%|▎         | 336/10708 [38:31<1:44:26,  1.66it/s]
  3%|▎         | 337/10708 [38:32<1:54:32,  1.51it/s]
  3%|▎         | 338/10708 [38:32<1:55:08,  1.50it/s]
  3%|▎         | 339/10708 [38:33<1:48:24,  1.59it/s]
  3%|▎         | 340/10708 [38:33<1:46:37,  1.62it/s]
  3%|▎         | 341/10708 [38:34<1:41:57,  1.69it/s]
  3%|▎         | 342/10708 [38:35<1:46:13,  1.63it/s]
  3%|▎         | 343/10708 [38:35<1:44:19,  1.66it/s]
  3%|▎         | 344/10708 [38:36<1:44:15,  1.66it/s]
  3%|▎         | 345/10708 [38:37<2:03:37,  1.40it/s]
  3%|▎         | 346/10708 [38:37<1:55:09,  1.50it/s]
  3%|▎         | 347/10708 [38:38<1:52:49,  1.53it/s]
  3%|▎         | 348/10708 [38:39<1:47:53,  1.60it/s]
  3%|▎         | 349/10708 [38:39<1:47:22,  1.61it/s]
  3%|▎         | 350/10708 [38:40<1:44:09,  1.66it/s]
                                                     {'loss': 5.5892, 'grad_norm': 0.5809213519096375, 'learning_rate': 0.00032679738562091506, 'epoch': 0.03}


  3%|▎         | 350/10708 [38:40<1:44:09,  1.66it/s]
  3%|▎         | 351/10708 [38:40<1:44:37,  1.65it/s]
  3%|▎         | 352/10708 [38:41<1:41:27,  1.70it/s]
  3%|▎         | 353/10708 [38:42<1:43:42,  1.66it/s]
  3%|▎         | 354/10708 [38:42<1:39:56,  1.73it/s]
  3%|▎         | 355/10708 [38:43<1:41:00,  1.71it/s]
  3%|▎         | 356/10708 [38:43<1:40:03,  1.72it/s]
  3%|▎         | 357/10708 [38:44<1:38:33,  1.75it/s]
  3%|▎         | 358/10708 [38:44<1:37:02,  1.78it/s]
  3%|▎         | 359/10708 [38:45<1:35:50,  1.80it/s]
  3%|▎         | 360/10708 [38:45<1:37:13,  1.77it/s]
  3%|▎         | 361/10708 [38:46<1:41:14,  1.70it/s]
  3%|▎         | 362/10708 [38:47<1:39:56,  1.73it/s]
  3%|▎         | 363/10708 [38:47<1:39:45,  1.73it/s]
  3%|▎         | 364/10708 [38:48<1:39:30,  1.73it/s]
  3%|▎         | 365/10708 [38:48<1:41:15,  1.70it/s]
  3%|▎         | 366/10708 [38:49<1:41:56,  1.69it/s]
  3%|▎         | 367/10708 [38:50<1:41:23,  1.70it/s]
  3%|▎         | 368/10708 [38:50<1:39:08,  1.74it/s]
  3%|▎         | 369/10708 [38:51<1:43:15,  1.67it/s]
  3%|▎         | 370/10708 [38:51<1:41:29,  1.70it/s]
  3%|▎         | 371/10708 [38:52<1:42:11,  1.69it/s]
  3%|▎         | 372/10708 [38:52<1:39:30,  1.73it/s]
  3%|▎         | 373/10708 [38:53<1:39:56,  1.72it/s]
  3%|▎         | 374/10708 [38:54<1:38:44,  1.74it/s]
  4%|▎         | 375/10708 [38:54<1:37:36,  1.76it/s]
                                                     
{'loss': 5.4919, 'grad_norm': 0.8862716555595398, 'learning_rate': 0.00035014005602240897, 'epoch': 0.04}

  4%|▎         | 375/10708 [38:54<1:37:36,  1.76it/s]
  4%|▎         | 376/10708 [38:55<1:40:08,  1.72it/s]
  4%|▎         | 377/10708 [38:55<1:39:19,  1.73it/s]
  4%|▎         | 378/10708 [38:56<1:38:44,  1.74it/s]
  4%|▎         | 379/10708 [38:57<1:39:21,  1.73it/s]
  4%|▎         | 380/10708 [38:57<1:36:49,  1.78it/s]
  4%|▎         | 381/10708 [38:58<1:36:48,  1.78it/s]
  4%|▎         | 382/10708 [38:58<1:36:23,  1.79it/s]
  4%|▎         | 383/10708 [38:59<1:35:16,  1.81it/s]
  4%|▎         | 384/10708 [38:59<1:38:12,  1.75it/s]
  4%|▎         | 385/10708 [39:00<1:38:05,  1.75it/s]
  4%|▎         | 386/10708 [39:00<1:39:47,  1.72it/s]
  4%|▎         | 387/10708 [39:01<1:36:52,  1.78it/s]
  4%|▎         | 388/10708 [39:02<1:36:18,  1.79it/s]
  4%|▎         | 389/10708 [39:02<1:37:27,  1.76it/s]
  4%|▎         | 390/10708 [39:03<1:37:56,  1.76it/s]
  4%|▎         | 391/10708 [39:03<1:41:03,  1.70it/s]
  4%|▎         | 392/10708 [39:04<1:39:20,  1.73it/s]
  4%|▎         | 393/10708 [39:04<1:36:58,  1.77it/s]
  4%|▎         | 394/10708 [39:05<1:38:13,  1.75it/s]
  4%|▎         | 395/10708 [39:06<1:36:47,  1.78it/s]
  4%|▎         | 396/10708 [39:06<1:35:15,  1.80it/s]
  4%|▎         | 397/10708 [39:07<1:37:34,  1.76it/s]
  4%|▎         | 398/10708 [39:07<1:37:07,  1.77it/s]
  4%|▎         | 399/10708 [39:08<1:39:25,  1.73it/s]
  4%|▎         | 400/10708 [39:08<1:37:05,  1.77it/s]{'loss': 5.4039, 'grad_norm': 0.534810483455658, 'learning_rate': 0.0003734827264239029, 'epoch': 0.04}
                                                     

  4%|▎         | 400/10708 [39:08<1:37:05,  1.77it/s]
  4%|▎         | 401/10708 [39:09<1:36:01,  1.79it/s]
  4%|▍         | 402/10708 [39:10<1:36:42,  1.78it/s]
  4%|▍         | 403/10708 [39:10<1:38:45,  1.74it/s]
  4%|▍         | 404/10708 [39:11<1:36:48,  1.77it/s]
  4%|▍         | 405/10708 [39:11<1:41:21,  1.69it/s]
  4%|▍         | 406/10708 [39:12<1:40:44,  1.70it/s]
  4%|▍         | 407/10708 [39:12<1:41:36,  1.69it/s]
  4%|▍         | 408/10708 [39:13<1:39:13,  1.73it/s]
  4%|▍         | 409/10708 [39:14<1:37:11,  1.77it/s]
  4%|▍         | 410/10708 [39:14<1:39:31,  1.72it/s]
  4%|▍         | 411/10708 [39:15<1:37:07,  1.77it/s]
  4%|▍         | 412/10708 [39:15<1:37:15,  1.76it/s]
  4%|▍         | 413/10708 [39:16<1:37:11,  1.77it/s]
  4%|▍         | 414/10708 [39:16<1:37:41,  1.76it/s]
  4%|▍         | 415/10708 [39:17<1:36:20,  1.78it/s]
  4%|▍         | 416/10708 [39:18<1:36:39,  1.77it/s]
  4%|▍         | 417/10708 [39:18<1:35:20,  1.80it/s]
  4%|▍         | 418/10708 [39:19<1:33:57,  1.83it/s]
  4%|▍         | 419/10708 [39:19<1:37:28,  1.76it/s]
  4%|▍         | 420/10708 [39:20<1:37:13,  1.76it/s]
  4%|▍         | 421/10708 [39:20<1:34:52,  1.81it/s]
  4%|▍         | 422/10708 [39:21<1:37:19,  1.76it/s]
  4%|▍         | 423/10708 [39:21<1:37:37,  1.76it/s]
  4%|▍         | 424/10708 [39:22<1:38:34,  1.74it/s]
  4%|▍         | 425/10708 [39:23<1:55:33,  1.48it/s]{'loss': 5.3117, 'grad_norm': 0.6840770840644836, 'learning_rate': 0.0003968253968253968, 'epoch': 0.04}

                                                     

  4%|▍         | 425/10708 [39:23<1:55:33,  1.48it/s]
  4%|▍         | 426/10708 [39:24<1:51:22,  1.54it/s]
  4%|▍         | 427/10708 [39:24<1:51:32,  1.54it/s]
  4%|▍         | 428/10708 [39:25<1:48:25,  1.58it/s]
  4%|▍         | 429/10708 [39:26<1:56:53,  1.47it/s]
  4%|▍         | 430/10708 [39:26<1:50:48,  1.55it/s]
  4%|▍         | 431/10708 [39:27<1:59:11,  1.44it/s]
  4%|▍         | 432/10708 [39:28<1:55:21,  1.48it/s]
  4%|▍         | 433/10708 [39:28<2:04:23,  1.38it/s]
  4%|▍         | 434/10708 [39:29<1:57:51,  1.45it/s]
  4%|▍         | 435/10708 [39:30<1:56:06,  1.47it/s]
  4%|▍         | 436/10708 [39:30<1:52:04,  1.53it/s]
  4%|▍         | 437/10708 [39:31<1:48:07,  1.58it/s]
  4%|▍         | 438/10708 [39:31<1:45:15,  1.63it/s]
  4%|▍         | 439/10708 [39:32<1:49:58,  1.56it/s]
  4%|▍         | 440/10708 [39:33<1:46:49,  1.60it/s]
  4%|▍         | 441/10708 [39:33<1:44:05,  1.64it/s]
  4%|▍         | 442/10708 [39:34<1:42:00,  1.68it/s]
  4%|▍         | 443/10708 [39:35<2:06:20,  1.35it/s]
  4%|▍         | 444/10708 [39:36<1:56:44,  1.47it/s]
  4%|▍         | 445/10708 [39:36<1:52:07,  1.53it/s]
  4%|▍         | 446/10708 [39:37<1:59:38,  1.43it/s]
  4%|▍         | 447/10708 [39:38<1:55:31,  1.48it/s]
  4%|▍         | 448/10708 [39:38<1:49:52,  1.56it/s]
  4%|▍         | 449/10708 [39:39<1:47:24,  1.59it/s]
  4%|▍         | 450/10708 [39:40<2:03:32,  1.38it/s]{'loss': 5.2472, 'grad_norm': 0.7636505961418152, 'learning_rate': 0.0004201680672268908, 'epoch': 0.04}
                                                     

  4%|▍         | 450/10708 [39:40<2:03:32,  1.38it/s]
  4%|▍         | 451/10708 [39:40<1:55:35,  1.48it/s]
  4%|▍         | 452/10708 [39:41<2:05:11,  1.37it/s]
  4%|▍         | 453/10708 [39:42<1:57:27,  1.46it/s]
  4%|▍         | 454/10708 [39:42<1:57:20,  1.46it/s]
  4%|▍         | 455/10708 [39:43<1:52:58,  1.51it/s]
  4%|▍         | 456/10708 [39:44<1:48:40,  1.57it/s]
  4%|▍         | 457/10708 [39:44<1:51:08,  1.54it/s]
  4%|▍         | 458/10708 [39:45<1:47:35,  1.59it/s]
  4%|▍         | 459/10708 [39:46<1:55:33,  1.48it/s]
  4%|▍         | 460/10708 [39:46<1:55:01,  1.48it/s]
  4%|▍         | 461/10708 [39:47<1:53:26,  1.51it/s]
  4%|▍         | 462/10708 [39:48<2:03:33,  1.38it/s]
  4%|▍         | 463/10708 [39:48<1:57:47,  1.45it/s]
  4%|▍         | 464/10708 [39:49<1:56:36,  1.46it/s]
  4%|▍         | 465/10708 [39:50<1:53:43,  1.50it/s]
  4%|▍         | 466/10708 [39:50<1:50:40,  1.54it/s]
  4%|▍         | 467/10708 [39:51<1:48:21,  1.58it/s]
  4%|▍         | 468/10708 [39:51<1:46:09,  1.61it/s]
  4%|▍         | 469/10708 [39:52<1:41:50,  1.68it/s]
  4%|▍         | 470/10708 [39:53<1:41:18,  1.68it/s]
  4%|▍         | 471/10708 [39:53<1:38:14,  1.74it/s]
  4%|▍         | 472/10708 [39:54<1:36:59,  1.76it/s]
  4%|▍         | 473/10708 [39:54<1:34:53,  1.80it/s]
  4%|▍         | 474/10708 [39:55<1:35:06,  1.79it/s]
  4%|▍         | 475/10708 [39:55<1:33:38,  1.82it/s]{'loss': 5.1726, 'grad_norm': 0.6458236575126648, 'learning_rate': 0.0004435107376283847, 'epoch': 0.04}
                                                     

  4%|▍         | 475/10708 [39:55<1:33:38,  1.82it/s]
  4%|▍         | 476/10708 [39:56<1:38:46,  1.73it/s]
  4%|▍         | 477/10708 [39:56<1:38:34,  1.73it/s]
  4%|▍         | 478/10708 [39:57<1:39:33,  1.71it/s]
  4%|▍         | 479/10708 [39:58<1:37:55,  1.74it/s]
  4%|▍         | 480/10708 [39:58<1:34:20,  1.81it/s]
  4%|▍         | 481/10708 [39:59<2:00:28,  1.41it/s]
  5%|▍         | 482/10708 [40:00<1:54:47,  1.48it/s]
  5%|▍         | 483/10708 [40:00<1:48:19,  1.57it/s]
  5%|▍         | 484/10708 [40:01<1:48:08,  1.58it/s]
  5%|▍         | 485/10708 [40:01<1:41:19,  1.68it/s]
  5%|▍         | 486/10708 [40:02<1:42:28,  1.66it/s]
  5%|▍         | 487/10708 [40:03<1:39:54,  1.71it/s]
  5%|▍         | 488/10708 [40:03<1:37:47,  1.74it/s]
  5%|▍         | 489/10708 [40:04<1:37:10,  1.75it/s]
  5%|▍         | 490/10708 [40:04<1:36:19,  1.77it/s]
  5%|▍         | 491/10708 [40:05<1:35:09,  1.79it/s]
  5%|▍         | 492/10708 [40:05<1:32:10,  1.85it/s]
  5%|▍         | 493/10708 [40:06<1:43:09,  1.65it/s]
  5%|▍         | 494/10708 [40:07<1:39:07,  1.72it/s]
  5%|▍         | 495/10708 [40:07<1:35:15,  1.79it/s]
  5%|▍         | 496/10708 [40:08<1:37:22,  1.75it/s]
  5%|▍         | 497/10708 [40:08<1:33:39,  1.82it/s]
  5%|▍         | 498/10708 [40:09<1:30:54,  1.87it/s]
  5%|▍         | 499/10708 [40:09<1:31:29,  1.86it/s]
  5%|▍         | 500/10708 [40:10<1:29:38,  1.90it/s]{'loss': 5.1191, 'grad_norm': 0.577984094619751, 'learning_rate': 0.00046685340802987864, 'epoch': 0.05}

                                                     

  5%|▍         | 500/10708 [40:10<1:29:38,  1.90it/s]
  5%|▍         | 501/10708 [40:10<1:28:37,  1.92it/s]
  5%|▍         | 502/10708 [40:11<1:29:24,  1.90it/s]
  5%|▍         | 503/10708 [40:11<1:29:33,  1.90it/s]
  5%|▍         | 504/10708 [40:12<1:32:58,  1.83it/s]
  5%|▍         | 505/10708 [40:13<1:43:07,  1.65it/s]
  5%|▍         | 506/10708 [40:13<1:37:44,  1.74it/s]
  5%|▍         | 507/10708 [40:14<1:35:45,  1.78it/s]
  5%|▍         | 508/10708 [40:14<1:32:30,  1.84it/s]
  5%|▍         | 509/10708 [40:15<1:30:11,  1.88it/s]
  5%|▍         | 510/10708 [40:15<1:30:02,  1.89it/s]
  5%|▍         | 511/10708 [40:16<1:30:23,  1.88it/s]
  5%|▍         | 512/10708 [40:16<1:30:11,  1.88it/s]
  5%|▍         | 513/10708 [40:17<1:28:50,  1.91it/s]
  5%|▍         | 514/10708 [40:17<1:29:44,  1.89it/s]
  5%|▍         | 515/10708 [40:18<1:29:33,  1.90it/s]
  5%|▍         | 516/10708 [40:18<1:29:30,  1.90it/s]
  5%|▍         | 517/10708 [40:19<1:28:28,  1.92it/s]
  5%|▍         | 518/10708 [40:19<1:28:46,  1.91it/s]
  5%|▍         | 519/10708 [40:20<1:28:50,  1.91it/s]
  5%|▍         | 520/10708 [40:21<1:29:28,  1.90it/s]
  5%|▍         | 521/10708 [40:21<1:30:08,  1.88it/s]
  5%|▍         | 522/10708 [40:22<1:28:30,  1.92it/s]
  5%|▍         | 523/10708 [40:22<1:27:33,  1.94it/s]
  5%|▍         | 524/10708 [40:23<1:29:50,  1.89it/s]
  5%|▍         | 525/10708 [40:23<1:28:20,  1.92it/s]
                                                     
{'loss': 5.0574, 'grad_norm': 0.4912331700325012, 'learning_rate': 0.0004901960784313725, 'epoch': 0.05}

  5%|▍         | 525/10708 [40:23<1:28:20,  1.92it/s]
  5%|▍         | 526/10708 [40:24<1:30:30,  1.87it/s]
  5%|▍         | 527/10708 [40:24<1:29:17,  1.90it/s]
  5%|▍         | 528/10708 [40:25<1:28:57,  1.91it/s]
  5%|▍         | 529/10708 [40:25<1:27:53,  1.93it/s]
  5%|▍         | 530/10708 [40:26<1:28:34,  1.92it/s]
  5%|▍         | 531/10708 [40:26<1:28:14,  1.92it/s]
  5%|▍         | 532/10708 [40:27<1:38:01,  1.73it/s]
  5%|▍         | 533/10708 [40:28<1:37:05,  1.75it/s]
  5%|▍         | 534/10708 [40:28<1:34:43,  1.79it/s]
  5%|▍         | 535/10708 [40:29<1:33:26,  1.81it/s]
  5%|▌         | 536/10708 [40:29<1:34:32,  1.79it/s]
  5%|▌         | 537/10708 [40:30<1:32:58,  1.82it/s]
  5%|▌         | 538/10708 [40:30<1:31:20,  1.86it/s]
  5%|▌         | 539/10708 [40:31<1:29:21,  1.90it/s]
  5%|▌         | 540/10708 [40:31<1:29:56,  1.88it/s]
  5%|▌         | 541/10708 [40:32<1:28:55,  1.91it/s]
  5%|▌         | 542/10708 [40:32<1:28:57,  1.90it/s]
  5%|▌         | 543/10708 [40:33<1:30:44,  1.87it/s]
  5%|▌         | 544/10708 [40:33<1:29:16,  1.90it/s]
  5%|▌         | 545/10708 [40:34<1:27:44,  1.93it/s]
  5%|▌         | 546/10708 [40:34<1:29:09,  1.90it/s]
  5%|▌         | 547/10708 [40:35<1:28:54,  1.90it/s]
  5%|▌         | 548/10708 [40:35<1:30:41,  1.87it/s]
  5%|▌         | 549/10708 [40:36<1:30:03,  1.88it/s]
  5%|▌         | 550/10708 [40:37<1:28:31,  1.91it/s]
                                                     
{'loss': 4.9976, 'grad_norm': 0.6723682284355164, 'learning_rate': 0.0005135387488328664, 'epoch': 0.05}

  5%|▌         | 550/10708 [40:37<1:28:31,  1.91it/s]
  5%|▌         | 551/10708 [40:37<1:30:22,  1.87it/s]
  5%|▌         | 552/10708 [40:38<1:29:28,  1.89it/s]
  5%|▌         | 553/10708 [40:38<1:30:50,  1.86it/s]
  5%|▌         | 554/10708 [40:39<1:28:57,  1.90it/s]
  5%|▌         | 555/10708 [40:39<1:30:24,  1.87it/s]
  5%|▌         | 556/10708 [40:40<1:29:09,  1.90it/s]
  5%|▌         | 557/10708 [40:40<1:28:40,  1.91it/s]
  5%|▌         | 558/10708 [40:41<1:28:44,  1.91it/s]
  5%|▌         | 559/10708 [40:41<1:27:12,  1.94it/s]
  5%|▌         | 560/10708 [40:42<1:26:47,  1.95it/s]
  5%|▌         | 561/10708 [40:42<1:26:24,  1.96it/s]
  5%|▌         | 562/10708 [40:43<1:28:11,  1.92it/s]
  5%|▌         | 563/10708 [40:43<1:28:20,  1.91it/s]
  5%|▌         | 564/10708 [40:44<1:29:59,  1.88it/s]
  5%|▌         | 565/10708 [40:44<1:28:27,  1.91it/s]
  5%|▌         | 566/10708 [40:45<1:28:28,  1.91it/s]
  5%|▌         | 567/10708 [40:45<1:27:43,  1.93it/s]
  5%|▌         | 568/10708 [40:46<1:28:45,  1.90it/s]
  5%|▌         | 569/10708 [40:46<1:29:27,  1.89it/s]
  5%|▌         | 570/10708 [40:47<1:27:50,  1.92it/s]
  5%|▌         | 571/10708 [40:48<1:39:15,  1.70it/s]
  5%|▌         | 572/10708 [40:48<1:36:14,  1.76it/s]
  5%|▌         | 573/10708 [40:49<1:33:21,  1.81it/s]
  5%|▌         | 574/10708 [40:49<1:30:27,  1.87it/s]
  5%|▌         | 575/10708 [40:50<1:29:32,  1.89it/s]{'loss': 4.9544, 'grad_norm': 0.5663116574287415, 'learning_rate': 0.0005368814192343605, 'epoch': 0.05}
                                                     

  5%|▌         | 575/10708 [40:50<1:29:32,  1.89it/s]
  5%|▌         | 576/10708 [40:50<1:29:36,  1.88it/s]
  5%|▌         | 577/10708 [40:51<1:28:07,  1.92it/s]
  5%|▌         | 578/10708 [40:51<1:27:54,  1.92it/s]
  5%|▌         | 579/10708 [40:52<1:27:48,  1.92it/s]
  5%|▌         | 580/10708 [40:52<1:28:30,  1.91it/s]
  5%|▌         | 581/10708 [40:53<1:39:28,  1.70it/s]
  5%|▌         | 582/10708 [40:54<1:36:51,  1.74it/s]
  5%|▌         | 583/10708 [40:54<1:33:26,  1.81it/s]
  5%|▌         | 584/10708 [40:55<1:30:42,  1.86it/s]
  5%|▌         | 585/10708 [40:55<1:31:10,  1.85it/s]
  5%|▌         | 586/10708 [40:56<1:29:03,  1.89it/s]
  5%|▌         | 587/10708 [40:56<1:28:32,  1.91it/s]
  5%|▌         | 588/10708 [40:57<1:27:12,  1.93it/s]
  6%|▌         | 589/10708 [40:57<1:26:43,  1.94it/s]
  6%|▌         | 590/10708 [40:58<1:26:37,  1.95it/s]
  6%|▌         | 591/10708 [40:58<1:26:24,  1.95it/s]
  6%|▌         | 592/10708 [40:59<1:26:19,  1.95it/s]
  6%|▌         | 593/10708 [40:59<1:26:45,  1.94it/s]
  6%|▌         | 594/10708 [41:00<1:25:52,  1.96it/s]
  6%|▌         | 595/10708 [41:00<1:26:48,  1.94it/s]
  6%|▌         | 596/10708 [41:01<1:26:44,  1.94it/s]
  6%|▌         | 597/10708 [41:01<1:26:27,  1.95it/s]
  6%|▌         | 598/10708 [41:02<1:28:35,  1.90it/s]
  6%|▌         | 599/10708 [41:02<1:27:08,  1.93it/s]
  6%|▌         | 600/10708 [41:03<1:27:08,  1.93it/s]
                                                     
{'loss': 4.9112, 'grad_norm': 0.49156028032302856, 'learning_rate': 0.0005602240896358544, 'epoch': 0.06}

  6%|▌         | 600/10708 [41:03<1:27:08,  1.93it/s]
  6%|▌         | 601/10708 [41:03<1:27:39,  1.92it/s]
  6%|▌         | 602/10708 [41:04<1:27:20,  1.93it/s]
  6%|▌         | 603/10708 [41:04<1:26:49,  1.94it/s]
  6%|▌         | 604/10708 [41:05<1:27:06,  1.93it/s]
  6%|▌         | 605/10708 [41:05<1:26:06,  1.96it/s]
  6%|▌         | 606/10708 [41:06<1:26:19,  1.95it/s]
  6%|▌         | 607/10708 [41:07<1:25:59,  1.96it/s]
  6%|▌         | 608/10708 [41:07<1:27:18,  1.93it/s]
  6%|▌         | 609/10708 [41:08<1:26:18,  1.95it/s]
  6%|▌         | 610/10708 [41:08<1:26:47,  1.94it/s]
  6%|▌         | 611/10708 [41:09<1:25:53,  1.96it/s]
  6%|▌         | 612/10708 [41:09<1:25:54,  1.96it/s]
  6%|▌         | 613/10708 [41:10<1:26:56,  1.94it/s]
  6%|▌         | 614/10708 [41:10<1:27:57,  1.91it/s]
  6%|▌         | 615/10708 [41:11<1:27:23,  1.92it/s]
  6%|▌         | 616/10708 [41:11<1:28:31,  1.90it/s]
  6%|▌         | 617/10708 [41:12<1:28:32,  1.90it/s]
  6%|▌         | 618/10708 [41:12<1:27:08,  1.93it/s]
  6%|▌         | 619/10708 [41:13<1:39:00,  1.70it/s]
  6%|▌         | 620/10708 [41:14<1:36:18,  1.75it/s]
  6%|▌         | 621/10708 [41:14<1:32:24,  1.82it/s]
  6%|▌         | 622/10708 [41:15<1:30:19,  1.86it/s]
  6%|▌         | 623/10708 [41:15<1:29:09,  1.89it/s]
  6%|▌         | 624/10708 [41:16<1:28:19,  1.90it/s]
  6%|▌         | 625/10708 [41:16<1:27:39,  1.92it/s]
                                                     {'loss': 4.8691, 'grad_norm': 0.5351319313049316, 'learning_rate': 0.0005835667600373484, 'epoch': 0.06}


  6%|▌         | 625/10708 [41:16<1:27:39,  1.92it/s]
  6%|▌         | 626/10708 [41:17<1:28:11,  1.91it/s]
  6%|▌         | 627/10708 [41:17<1:28:09,  1.91it/s]
  6%|▌         | 628/10708 [41:18<1:27:05,  1.93it/s]
  6%|▌         | 629/10708 [41:18<1:28:24,  1.90it/s]
  6%|▌         | 630/10708 [41:19<1:27:09,  1.93it/s]
  6%|▌         | 631/10708 [41:19<1:26:13,  1.95it/s]
  6%|▌         | 632/10708 [41:20<1:27:14,  1.92it/s]
  6%|▌         | 633/10708 [41:20<1:26:35,  1.94it/s]
  6%|▌         | 634/10708 [41:21<1:25:32,  1.96it/s]
  6%|▌         | 635/10708 [41:21<1:25:50,  1.96it/s]
  6%|▌         | 636/10708 [41:22<1:26:33,  1.94it/s]
  6%|▌         | 637/10708 [41:22<1:27:45,  1.91it/s]
  6%|▌         | 638/10708 [41:23<1:27:22,  1.92it/s]
  6%|▌         | 639/10708 [41:23<1:27:12,  1.92it/s]
  6%|▌         | 640/10708 [41:24<1:26:09,  1.95it/s]
  6%|▌         | 641/10708 [41:24<1:27:45,  1.91it/s]
  6%|▌         | 642/10708 [41:25<1:27:29,  1.92it/s]
  6%|▌         | 643/10708 [41:25<1:26:20,  1.94it/s]
  6%|▌         | 644/10708 [41:26<1:26:58,  1.93it/s]
  6%|▌         | 645/10708 [41:26<1:26:43,  1.93it/s]
  6%|▌         | 646/10708 [41:27<1:25:50,  1.95it/s]
  6%|▌         | 647/10708 [41:27<1:25:58,  1.95it/s]
  6%|▌         | 648/10708 [41:28<1:25:34,  1.96it/s]
  6%|▌         | 649/10708 [41:29<1:28:02,  1.90it/s]
  6%|▌         | 650/10708 [41:29<1:38:41,  1.70it/s]
                                                     
{'loss': 4.826, 'grad_norm': 0.49542924761772156, 'learning_rate': 0.0006069094304388422, 'epoch': 0.06}

  6%|▌         | 650/10708 [41:29<1:38:41,  1.70it/s]
  6%|▌         | 651/10708 [41:30<1:34:31,  1.77it/s]
  6%|▌         | 652/10708 [41:30<1:33:13,  1.80it/s]
  6%|▌         | 653/10708 [41:31<1:30:12,  1.86it/s]
  6%|▌         | 654/10708 [41:31<1:29:18,  1.88it/s]
  6%|▌         | 655/10708 [41:32<1:28:11,  1.90it/s]
  6%|▌         | 656/10708 [41:32<1:28:16,  1.90it/s]
  6%|▌         | 657/10708 [41:33<1:27:31,  1.91it/s]
  6%|▌         | 658/10708 [41:33<1:27:25,  1.92it/s]
  6%|▌         | 659/10708 [41:34<1:27:22,  1.92it/s]
  6%|▌         | 660/10708 [41:34<1:27:28,  1.91it/s]
  6%|▌         | 661/10708 [41:35<1:27:33,  1.91it/s]
  6%|▌         | 662/10708 [41:35<1:26:28,  1.94it/s]
  6%|▌         | 663/10708 [41:36<1:26:33,  1.93it/s]
  6%|▌         | 664/10708 [41:36<1:26:47,  1.93it/s]
  6%|▌         | 665/10708 [41:37<1:27:33,  1.91it/s]
  6%|▌         | 666/10708 [41:38<1:26:32,  1.93it/s]
  6%|▌         | 667/10708 [41:38<1:37:07,  1.72it/s]
  6%|▌         | 668/10708 [41:39<1:33:54,  1.78it/s]
  6%|▌         | 669/10708 [41:39<1:30:39,  1.85it/s]
  6%|▋         | 670/10708 [41:40<1:31:11,  1.83it/s]
  6%|▋         | 671/10708 [41:40<1:28:54,  1.88it/s]
  6%|▋         | 672/10708 [41:41<1:37:11,  1.72it/s]
  6%|▋         | 673/10708 [41:42<1:41:24,  1.65it/s]
  6%|▋         | 674/10708 [41:42<1:43:03,  1.62it/s]
  6%|▋         | 675/10708 [41:43<1:43:33,  1.61it/s]
                                                     
{'loss': 4.7974, 'grad_norm': 0.45134425163269043, 'learning_rate': 0.0006302521008403362, 'epoch': 0.06}

  6%|▋         | 675/10708 [41:43<1:43:33,  1.61it/s]
  6%|▋         | 676/10708 [41:43<1:39:49,  1.67it/s]
  6%|▋         | 677/10708 [41:44<1:35:35,  1.75it/s]
  6%|▋         | 678/10708 [41:45<1:32:06,  1.82it/s]
  6%|▋         | 679/10708 [41:45<1:30:32,  1.85it/s]
  6%|▋         | 680/10708 [41:46<1:29:35,  1.87it/s]
  6%|▋         | 681/10708 [41:46<1:29:11,  1.87it/s]
  6%|▋         | 682/10708 [41:47<1:27:44,  1.90it/s]
  6%|▋         | 683/10708 [41:47<1:26:36,  1.93it/s]
  6%|▋         | 684/10708 [41:48<1:26:26,  1.93it/s]
  6%|▋         | 685/10708 [41:48<1:25:26,  1.96it/s]
  6%|▋         | 686/10708 [41:49<1:24:58,  1.97it/s]
  6%|▋         | 687/10708 [41:49<1:25:46,  1.95it/s]
  6%|▋         | 688/10708 [41:50<1:26:22,  1.93it/s]
  6%|▋         | 689/10708 [41:50<1:26:13,  1.94it/s]
  6%|▋         | 690/10708 [41:51<1:25:19,  1.96it/s]
  6%|▋         | 691/10708 [41:51<1:26:02,  1.94it/s]
  6%|▋         | 692/10708 [41:52<1:25:30,  1.95it/s]
  6%|▋         | 693/10708 [41:52<1:25:41,  1.95it/s]
  6%|▋         | 694/10708 [41:53<1:25:29,  1.95it/s]
  6%|▋         | 695/10708 [41:53<1:25:10,  1.96it/s]
  6%|▋         | 696/10708 [41:54<1:36:10,  1.74it/s]
  7%|▋         | 697/10708 [41:54<1:32:21,  1.81it/s]
  7%|▋         | 698/10708 [41:55<1:30:15,  1.85it/s]
  7%|▋         | 699/10708 [41:55<1:28:09,  1.89it/s]
  7%|▋         | 700/10708 [41:56<1:28:09,  1.89it/s]
                                                     
{'loss': 4.7427, 'grad_norm': 0.49986767768859863, 'learning_rate': 0.0006535947712418301, 'epoch': 0.07}

  7%|▋         | 700/10708 [41:56<1:28:09,  1.89it/s]
  7%|▋         | 701/10708 [41:57<1:28:06,  1.89it/s]
  7%|▋         | 702/10708 [41:57<1:28:41,  1.88it/s]
  7%|▋         | 703/10708 [41:58<1:27:28,  1.91it/s]
  7%|▋         | 704/10708 [41:58<1:26:36,  1.93it/s]
  7%|▋         | 705/10708 [41:59<1:25:41,  1.95it/s]
  7%|▋         | 706/10708 [41:59<1:25:37,  1.95it/s]
  7%|▋         | 707/10708 [42:00<1:25:18,  1.95it/s]
  7%|▋         | 708/10708 [42:00<1:25:00,  1.96it/s]
  7%|▋         | 709/10708 [42:01<1:24:52,  1.96it/s]
  7%|▋         | 710/10708 [42:01<1:24:59,  1.96it/s]
  7%|▋         | 711/10708 [42:02<1:24:37,  1.97it/s]
  7%|▋         | 712/10708 [42:02<1:25:14,  1.95it/s]
  7%|▋         | 713/10708 [42:03<1:25:04,  1.96it/s]
  7%|▋         | 714/10708 [42:03<1:24:42,  1.97it/s]
  7%|▋         | 715/10708 [42:04<1:25:34,  1.95it/s]
  7%|▋         | 716/10708 [42:04<1:25:00,  1.96it/s]
  7%|▋         | 717/10708 [42:05<1:24:55,  1.96it/s]
  7%|▋         | 718/10708 [42:05<1:25:06,  1.96it/s]
  7%|▋         | 719/10708 [42:06<1:24:48,  1.96it/s]
  7%|▋         | 720/10708 [42:06<1:24:53,  1.96it/s]
  7%|▋         | 721/10708 [42:07<1:24:53,  1.96it/s]
  7%|▋         | 722/10708 [42:07<1:26:02,  1.93it/s]
  7%|▋         | 723/10708 [42:08<1:25:04,  1.96it/s]
  7%|▋         | 724/10708 [42:08<1:24:27,  1.97it/s]
  7%|▋         | 725/10708 [42:09<1:25:21,  1.95it/s]
                                                     
{'loss': 4.7221, 'grad_norm': 0.525814950466156, 'learning_rate': 0.0006769374416433241, 'epoch': 0.07}

  7%|▋         | 725/10708 [42:09<1:25:21,  1.95it/s]
  7%|▋         | 726/10708 [42:09<1:25:28,  1.95it/s]
  7%|▋         | 727/10708 [42:10<1:25:25,  1.95it/s]
  7%|▋         | 728/10708 [42:10<1:24:30,  1.97it/s]
  7%|▋         | 729/10708 [42:11<1:25:24,  1.95it/s]
  7%|▋         | 730/10708 [42:11<1:25:07,  1.95it/s]
  7%|▋         | 731/10708 [42:12<1:24:51,  1.96it/s]
  7%|▋         | 732/10708 [42:12<1:24:16,  1.97it/s]
  7%|▋         | 733/10708 [42:13<1:25:20,  1.95it/s]
  7%|▋         | 734/10708 [42:13<1:24:51,  1.96it/s]
  7%|▋         | 735/10708 [42:14<1:25:25,  1.95it/s]
  7%|▋         | 736/10708 [42:14<1:24:43,  1.96it/s]
  7%|▋         | 737/10708 [42:15<1:24:53,  1.96it/s]
  7%|▋         | 738/10708 [42:15<1:24:13,  1.97it/s]
  7%|▋         | 739/10708 [42:16<1:25:55,  1.93it/s]
  7%|▋         | 740/10708 [42:16<1:24:55,  1.96it/s]
  7%|▋         | 741/10708 [42:17<1:24:52,  1.96it/s]
  7%|▋         | 742/10708 [42:17<1:25:28,  1.94it/s]
  7%|▋         | 743/10708 [42:18<1:24:54,  1.96it/s]
  7%|▋         | 744/10708 [42:19<1:25:40,  1.94it/s]
  7%|▋         | 745/10708 [42:19<1:24:56,  1.95it/s]
  7%|▋         | 746/10708 [42:20<1:24:24,  1.97it/s]
  7%|▋         | 747/10708 [42:20<1:24:56,  1.95it/s]
  7%|▋         | 748/10708 [42:21<1:25:27,  1.94it/s]
  7%|▋         | 749/10708 [42:21<1:26:18,  1.92it/s]
  7%|▋         | 750/10708 [42:22<1:25:12,  1.95it/s]
                                                     
{'loss': 4.6868, 'grad_norm': 0.5079945921897888, 'learning_rate': 0.0007002801120448179, 'epoch': 0.07}

  7%|▋         | 750/10708 [42:22<1:25:12,  1.95it/s]
  7%|▋         | 751/10708 [42:22<1:24:26,  1.97it/s]
  7%|▋         | 752/10708 [42:23<1:23:45,  1.98it/s]
  7%|▋         | 753/10708 [42:23<1:23:25,  1.99it/s]
  7%|▋         | 754/10708 [42:24<1:23:03,  2.00it/s]
  7%|▋         | 755/10708 [42:24<1:23:01,  2.00it/s]
  7%|▋         | 756/10708 [42:25<1:22:47,  2.00it/s]
  7%|▋         | 757/10708 [42:25<1:22:36,  2.01it/s]
  7%|▋         | 758/10708 [42:26<1:22:39,  2.01it/s]
  7%|▋         | 759/10708 [42:26<1:22:41,  2.01it/s]
  7%|▋         | 760/10708 [42:27<1:22:42,  2.00it/s]
  7%|▋         | 761/10708 [42:27<1:22:41,  2.00it/s]
  7%|▋         | 762/10708 [42:28<1:22:35,  2.01it/s]
  7%|▋         | 763/10708 [42:28<1:23:06,  1.99it/s]
  7%|▋         | 764/10708 [42:29<1:23:16,  1.99it/s]
  7%|▋         | 765/10708 [42:29<1:22:55,  2.00it/s]
  7%|▋         | 766/10708 [42:30<1:23:00,  2.00it/s]
  7%|▋         | 767/10708 [42:30<1:23:35,  1.98it/s]
  7%|▋         | 768/10708 [42:31<1:23:42,  1.98it/s]
  7%|▋         | 769/10708 [42:31<1:23:22,  1.99it/s]
  7%|▋         | 770/10708 [42:32<1:24:29,  1.96it/s]
  7%|▋         | 771/10708 [42:32<1:24:14,  1.97it/s]
  7%|▋         | 772/10708 [42:33<1:24:05,  1.97it/s]
  7%|▋         | 773/10708 [42:33<1:24:24,  1.96it/s]
  7%|▋         | 774/10708 [42:34<1:23:54,  1.97it/s]
  7%|▋         | 775/10708 [42:34<1:24:16,  1.96it/s]
                                                     
{'loss': 4.6598, 'grad_norm': 0.45551303029060364, 'learning_rate': 0.0007236227824463119, 'epoch': 0.07}

  7%|▋         | 775/10708 [42:34<1:24:16,  1.96it/s]
  7%|▋         | 776/10708 [42:35<1:24:15,  1.96it/s]
  7%|▋         | 777/10708 [42:35<1:24:35,  1.96it/s]
  7%|▋         | 778/10708 [42:36<1:24:26,  1.96it/s]
  7%|▋         | 779/10708 [42:36<1:24:51,  1.95it/s]
  7%|▋         | 780/10708 [42:37<1:25:51,  1.93it/s]
  7%|▋         | 781/10708 [42:37<1:25:45,  1.93it/s]
  7%|▋         | 782/10708 [42:38<1:24:40,  1.95it/s]
  7%|▋         | 783/10708 [42:38<1:24:36,  1.96it/s]
  7%|▋         | 784/10708 [42:39<1:24:25,  1.96it/s]
  7%|▋         | 785/10708 [42:39<1:24:11,  1.96it/s]
  7%|▋         | 786/10708 [42:40<1:24:12,  1.96it/s]
  7%|▋         | 787/10708 [42:40<1:23:38,  1.98it/s]
  7%|▋         | 788/10708 [42:41<1:23:53,  1.97it/s]
  7%|▋         | 789/10708 [42:41<1:23:23,  1.98it/s]
  7%|▋         | 790/10708 [42:42<1:24:51,  1.95it/s]
  7%|▋         | 791/10708 [42:42<1:25:19,  1.94it/s]
  7%|▋         | 792/10708 [42:43<1:35:47,  1.73it/s]
  7%|▋         | 793/10708 [42:44<1:31:49,  1.80it/s]
  7%|▋         | 794/10708 [42:44<1:28:52,  1.86it/s]
  7%|▋         | 795/10708 [42:45<1:27:08,  1.90it/s]
  7%|▋         | 796/10708 [42:45<1:25:55,  1.92it/s]
  7%|▋         | 797/10708 [42:46<1:25:39,  1.93it/s]
  7%|▋         | 798/10708 [42:46<1:24:40,  1.95it/s]
  7%|▋         | 799/10708 [42:47<1:24:47,  1.95it/s]
  7%|▋         | 800/10708 [42:47<1:24:00,  1.97it/s]{'loss': 4.6436, 'grad_norm': 0.4544575810432434, 'learning_rate': 0.0007469654528478058, 'epoch': 0.07}
                                                     

  7%|▋         | 800/10708 [42:47<1:24:00,  1.97it/s]
  7%|▋         | 801/10708 [42:48<1:24:25,  1.96it/s]
  7%|▋         | 802/10708 [42:48<1:25:27,  1.93it/s]
  7%|▋         | 803/10708 [42:49<1:24:36,  1.95it/s]
  8%|▊         | 804/10708 [42:49<1:25:06,  1.94it/s]
  8%|▊         | 805/10708 [42:50<1:24:20,  1.96it/s]
  8%|▊         | 806/10708 [42:50<1:23:59,  1.96it/s]
  8%|▊         | 807/10708 [42:51<1:23:29,  1.98it/s]
  8%|▊         | 808/10708 [42:51<1:23:54,  1.97it/s]
  8%|▊         | 809/10708 [42:52<1:23:29,  1.98it/s]
  8%|▊         | 810/10708 [42:52<1:23:31,  1.98it/s]
  8%|▊         | 811/10708 [42:53<1:23:29,  1.98it/s]
  8%|▊         | 812/10708 [42:53<1:23:24,  1.98it/s]
  8%|▊         | 813/10708 [42:54<1:23:53,  1.97it/s]
  8%|▊         | 814/10708 [42:54<1:23:19,  1.98it/s]
  8%|▊         | 815/10708 [42:55<1:23:33,  1.97it/s]
  8%|▊         | 816/10708 [42:55<1:22:54,  1.99it/s]
  8%|▊         | 817/10708 [42:56<1:24:09,  1.96it/s]
  8%|▊         | 818/10708 [42:56<1:23:29,  1.97it/s]
  8%|▊         | 819/10708 [42:57<1:23:35,  1.97it/s]
  8%|▊         | 820/10708 [42:57<1:23:54,  1.96it/s]
  8%|▊         | 821/10708 [42:58<1:23:43,  1.97it/s]
  8%|▊         | 822/10708 [42:58<1:23:35,  1.97it/s]
  8%|▊         | 823/10708 [42:59<1:23:01,  1.98it/s]
  8%|▊         | 824/10708 [42:59<1:23:00,  1.98it/s]
  8%|▊         | 825/10708 [43:00<1:22:39,  1.99it/s]
                                                     
{'loss': 4.6065, 'grad_norm': 0.41252049803733826, 'learning_rate': 0.0007703081232492998, 'epoch': 0.08}

  8%|▊         | 825/10708 [43:00<1:22:39,  1.99it/s]
  8%|▊         | 826/10708 [43:00<1:22:29,  2.00it/s]
  8%|▊         | 827/10708 [43:01<1:23:03,  1.98it/s]
  8%|▊         | 828/10708 [43:01<1:22:47,  1.99it/s]
  8%|▊         | 829/10708 [43:02<1:22:24,  2.00it/s]
  8%|▊         | 830/10708 [43:02<1:24:17,  1.95it/s]
  8%|▊         | 831/10708 [43:03<1:23:29,  1.97it/s]
  8%|▊         | 832/10708 [43:03<1:23:15,  1.98it/s]
  8%|▊         | 833/10708 [43:04<1:23:11,  1.98it/s]
  8%|▊         | 834/10708 [43:04<1:22:53,  1.99it/s]
  8%|▊         | 835/10708 [43:05<1:22:57,  1.98it/s]
  8%|▊         | 836/10708 [43:05<1:22:46,  1.99it/s]
  8%|▊         | 837/10708 [43:06<1:23:06,  1.98it/s]
  8%|▊         | 838/10708 [43:06<1:22:56,  1.98it/s]
  8%|▊         | 839/10708 [43:07<1:23:32,  1.97it/s]
  8%|▊         | 840/10708 [43:07<1:23:32,  1.97it/s]
  8%|▊         | 841/10708 [43:08<1:24:03,  1.96it/s]
  8%|▊         | 842/10708 [43:08<1:23:19,  1.97it/s]
  8%|▊         | 843/10708 [43:09<1:23:19,  1.97it/s]
  8%|▊         | 844/10708 [43:09<1:23:58,  1.96it/s]
  8%|▊         | 845/10708 [43:10<1:23:30,  1.97it/s]
  8%|▊         | 846/10708 [43:10<1:23:30,  1.97it/s]
  8%|▊         | 847/10708 [43:11<1:22:54,  1.98it/s]
  8%|▊         | 848/10708 [43:11<1:22:29,  1.99it/s]
  8%|▊         | 849/10708 [43:12<1:22:08,  2.00it/s]
  8%|▊         | 850/10708 [43:12<1:22:08,  2.00it/s]
                                                     
{'loss': 4.5911, 'grad_norm': 0.40663063526153564, 'learning_rate': 0.0007936507936507937, 'epoch': 0.08}

  8%|▊         | 850/10708 [43:12<1:22:08,  2.00it/s]
  8%|▊         | 851/10708 [43:13<1:22:26,  1.99it/s]
  8%|▊         | 852/10708 [43:13<1:22:42,  1.99it/s]
  8%|▊         | 853/10708 [43:14<1:22:41,  1.99it/s]
  8%|▊         | 854/10708 [43:14<1:22:21,  1.99it/s]
  8%|▊         | 855/10708 [43:15<1:22:34,  1.99it/s]
  8%|▊         | 856/10708 [43:15<1:22:47,  1.98it/s]
  8%|▊         | 857/10708 [43:16<1:22:30,  1.99it/s]
  8%|▊         | 858/10708 [43:17<1:33:53,  1.75it/s]
  8%|▊         | 859/10708 [43:17<1:30:17,  1.82it/s]
  8%|▊         | 860/10708 [43:18<1:28:31,  1.85it/s]
  8%|▊         | 861/10708 [43:18<1:27:25,  1.88it/s]
  8%|▊         | 862/10708 [43:19<1:26:06,  1.91it/s]
  8%|▊         | 863/10708 [43:19<1:26:08,  1.90it/s]
  8%|▊         | 864/10708 [43:20<1:25:04,  1.93it/s]
  8%|▊         | 865/10708 [43:20<1:24:36,  1.94it/s]
  8%|▊         | 866/10708 [43:21<1:24:04,  1.95it/s]
  8%|▊         | 867/10708 [43:21<1:23:19,  1.97it/s]
  8%|▊         | 868/10708 [43:22<1:23:21,  1.97it/s]
  8%|▊         | 869/10708 [43:22<1:23:36,  1.96it/s]
  8%|▊         | 870/10708 [43:23<1:23:04,  1.97it/s]
  8%|▊         | 871/10708 [43:23<1:23:27,  1.96it/s]
  8%|▊         | 872/10708 [43:24<1:22:59,  1.98it/s]
  8%|▊         | 873/10708 [43:24<1:22:48,  1.98it/s]
  8%|▊         | 874/10708 [43:25<1:22:22,  1.99it/s]
  8%|▊         | 875/10708 [43:25<1:22:20,  1.99it/s]
                                                     
{'loss': 4.5558, 'grad_norm': 0.3500451445579529, 'learning_rate': 0.0008169934640522876, 'epoch': 0.08}

  8%|▊         | 875/10708 [43:25<1:22:20,  1.99it/s]
  8%|▊         | 876/10708 [43:26<1:22:08,  2.00it/s]
  8%|▊         | 877/10708 [43:26<1:22:16,  1.99it/s]
  8%|▊         | 878/10708 [43:27<1:22:38,  1.98it/s]
  8%|▊         | 879/10708 [43:27<1:22:55,  1.98it/s]
  8%|▊         | 880/10708 [43:28<1:23:09,  1.97it/s]
  8%|▊         | 881/10708 [43:28<1:23:27,  1.96it/s]
  8%|▊         | 882/10708 [43:29<1:23:03,  1.97it/s]
  8%|▊         | 883/10708 [43:29<1:22:44,  1.98it/s]
  8%|▊         | 884/10708 [43:30<1:22:46,  1.98it/s]
  8%|▊         | 885/10708 [43:30<1:23:01,  1.97it/s]
  8%|▊         | 886/10708 [43:31<1:22:32,  1.98it/s]
  8%|▊         | 887/10708 [43:31<1:22:25,  1.99it/s]
  8%|▊         | 888/10708 [43:32<1:22:57,  1.97it/s]
  8%|▊         | 889/10708 [43:32<1:22:56,  1.97it/s]
  8%|▊         | 890/10708 [43:33<1:23:18,  1.96it/s]
  8%|▊         | 891/10708 [43:33<1:23:34,  1.96it/s]
  8%|▊         | 892/10708 [43:34<1:24:03,  1.95it/s]
  8%|▊         | 893/10708 [43:34<1:23:45,  1.95it/s]
  8%|▊         | 894/10708 [43:35<1:23:20,  1.96it/s]
  8%|▊         | 895/10708 [43:35<1:22:36,  1.98it/s]
  8%|▊         | 896/10708 [43:36<1:22:32,  1.98it/s]
  8%|▊         | 897/10708 [43:36<1:22:21,  1.99it/s]
  8%|▊         | 898/10708 [43:37<1:22:45,  1.98it/s]
  8%|▊         | 899/10708 [43:37<1:23:06,  1.97it/s]
  8%|▊         | 900/10708 [43:38<1:22:58,  1.97it/s]
                                                     {'loss': 4.5391, 'grad_norm': 0.3365751802921295, 'learning_rate': 0.0008403361344537816, 'epoch': 0.08}

  8%|▊         | 900/10708 [43:38<1:22:58,  1.97it/s]
  8%|▊         | 901/10708 [43:39<1:24:11,  1.94it/s]
  8%|▊         | 902/10708 [43:39<1:34:56,  1.72it/s]
  8%|▊         | 903/10708 [43:40<1:43:25,  1.58it/s]
  8%|▊         | 904/10708 [43:41<1:38:02,  1.67it/s]
  8%|▊         | 905/10708 [43:41<1:33:30,  1.75it/s]
  8%|▊         | 906/10708 [43:42<1:31:10,  1.79it/s]
  8%|▊         | 907/10708 [43:42<1:28:22,  1.85it/s]
  8%|▊         | 908/10708 [43:43<1:39:58,  1.63it/s]
  8%|▊         | 909/10708 [43:43<1:35:20,  1.71it/s]
  8%|▊         | 910/10708 [43:44<1:32:41,  1.76it/s]
  9%|▊         | 911/10708 [43:45<1:40:04,  1.63it/s]
  9%|▊         | 912/10708 [43:45<1:35:03,  1.72it/s]
  9%|▊         | 913/10708 [43:46<1:41:36,  1.61it/s]
  9%|▊         | 914/10708 [43:46<1:36:03,  1.70it/s]
  9%|▊         | 915/10708 [43:47<1:33:01,  1.75it/s]
  9%|▊         | 916/10708 [43:47<1:30:26,  1.80it/s]
  9%|▊         | 917/10708 [43:48<1:28:23,  1.85it/s]
  9%|▊         | 918/10708 [43:48<1:27:55,  1.86it/s]
  9%|▊         | 919/10708 [43:49<1:49:02,  1.50it/s]
  9%|▊         | 920/10708 [43:50<1:42:55,  1.58it/s]
  9%|▊         | 921/10708 [43:50<1:37:25,  1.67it/s]
  9%|▊         | 922/10708 [43:51<1:33:23,  1.75it/s]
  9%|▊         | 923/10708 [43:51<1:30:03,  1.81it/s]
  9%|▊         | 924/10708 [43:52<1:27:35,  1.86it/s]
  9%|▊         | 925/10708 [43:52<1:26:09,  1.89it/s]
                                                     
{'loss': 4.5246, 'grad_norm': 0.34793150424957275, 'learning_rate': 0.0008636788048552755, 'epoch': 0.09}

  9%|▊         | 925/10708 [43:52<1:26:09,  1.89it/s]
  9%|▊         | 926/10708 [43:53<1:25:08,  1.91it/s]
  9%|▊         | 927/10708 [43:54<1:24:30,  1.93it/s]
  9%|▊         | 928/10708 [43:54<1:24:07,  1.94it/s]
  9%|▊         | 929/10708 [43:55<1:24:52,  1.92it/s]
  9%|▊         | 930/10708 [43:55<1:25:17,  1.91it/s]
  9%|▊         | 931/10708 [43:56<1:24:46,  1.92it/s]
  9%|▊         | 932/10708 [43:56<1:24:04,  1.94it/s]
  9%|▊         | 933/10708 [43:57<1:24:11,  1.94it/s]
  9%|▊         | 934/10708 [43:57<1:23:17,  1.96it/s]
  9%|▊         | 935/10708 [43:58<1:23:40,  1.95it/s]
  9%|▊         | 936/10708 [43:58<1:23:31,  1.95it/s]
  9%|▉         | 937/10708 [43:59<1:22:37,  1.97it/s]
  9%|▉         | 938/10708 [43:59<1:22:17,  1.98it/s]
  9%|▉         | 939/10708 [44:00<1:22:56,  1.96it/s]
  9%|▉         | 940/10708 [44:00<1:35:55,  1.70it/s]
  9%|▉         | 941/10708 [44:01<1:31:43,  1.77it/s]
  9%|▉         | 942/10708 [44:01<1:28:46,  1.83it/s]
  9%|▉         | 943/10708 [44:02<1:26:52,  1.87it/s]
  9%|▉         | 944/10708 [44:02<1:26:13,  1.89it/s]
  9%|▉         | 945/10708 [44:03<1:25:06,  1.91it/s]
  9%|▉         | 946/10708 [44:03<1:24:33,  1.92it/s]
  9%|▉         | 947/10708 [44:04<1:24:21,  1.93it/s]
  9%|▉         | 948/10708 [44:05<1:24:21,  1.93it/s]
  9%|▉         | 949/10708 [44:05<1:23:31,  1.95it/s]
  9%|▉         | 950/10708 [44:06<1:23:12,  1.95it/s]
                                                     
{'loss': 4.4864, 'grad_norm': 0.32668358087539673, 'learning_rate': 0.0008870214752567694, 'epoch': 0.09}

  9%|▉         | 950/10708 [44:06<1:23:12,  1.95it/s]
  9%|▉         | 951/10708 [44:06<1:23:08,  1.96it/s]
  9%|▉         | 952/10708 [44:07<1:22:59,  1.96it/s]
  9%|▉         | 953/10708 [44:07<1:22:59,  1.96it/s]
  9%|▉         | 954/10708 [44:08<1:22:49,  1.96it/s]
  9%|▉         | 955/10708 [44:08<1:33:50,  1.73it/s]
  9%|▉         | 956/10708 [44:09<1:30:11,  1.80it/s]
  9%|▉         | 957/10708 [44:09<1:28:07,  1.84it/s]
  9%|▉         | 958/10708 [44:10<1:27:00,  1.87it/s]
  9%|▉         | 959/10708 [44:10<1:25:35,  1.90it/s]
  9%|▉         | 960/10708 [44:11<1:24:37,  1.92it/s]
  9%|▉         | 961/10708 [44:11<1:23:40,  1.94it/s]
  9%|▉         | 962/10708 [44:12<1:23:39,  1.94it/s]
  9%|▉         | 963/10708 [44:12<1:22:38,  1.97it/s]
  9%|▉         | 964/10708 [44:13<1:22:43,  1.96it/s]
  9%|▉         | 965/10708 [44:13<1:22:22,  1.97it/s]
  9%|▉         | 966/10708 [44:14<1:22:03,  1.98it/s]
  9%|▉         | 967/10708 [44:14<1:22:41,  1.96it/s]
  9%|▉         | 968/10708 [44:15<1:22:28,  1.97it/s]
  9%|▉         | 969/10708 [44:15<1:22:42,  1.96it/s]
  9%|▉         | 970/10708 [44:16<1:22:36,  1.96it/s]
  9%|▉         | 971/10708 [44:16<1:22:38,  1.96it/s]
  9%|▉         | 972/10708 [44:17<1:22:14,  1.97it/s]
  9%|▉         | 973/10708 [44:17<1:21:59,  1.98it/s]
  9%|▉         | 974/10708 [44:18<1:21:37,  1.99it/s]
  9%|▉         | 975/10708 [44:18<1:22:09,  1.97it/s]{'loss': 4.4765, 'grad_norm': 0.39217931032180786, 'learning_rate': 0.0009103641456582633, 'epoch': 0.09}
                                                     

  9%|▉         | 975/10708 [44:18<1:22:09,  1.97it/s]
  9%|▉         | 976/10708 [44:19<1:22:27,  1.97it/s]
  9%|▉         | 977/10708 [44:20<1:32:50,  1.75it/s]
  9%|▉         | 978/10708 [44:20<1:29:31,  1.81it/s]
  9%|▉         | 979/10708 [44:21<1:27:23,  1.86it/s]
  9%|▉         | 980/10708 [44:21<1:25:20,  1.90it/s]
  9%|▉         | 981/10708 [44:22<1:25:03,  1.91it/s]
  9%|▉         | 982/10708 [44:22<1:24:23,  1.92it/s]
  9%|▉         | 983/10708 [44:23<1:23:42,  1.94it/s]
  9%|▉         | 984/10708 [44:23<1:23:20,  1.94it/s]
  9%|▉         | 985/10708 [44:24<1:23:00,  1.95it/s]
  9%|▉         | 986/10708 [44:24<1:22:52,  1.96it/s]
  9%|▉         | 987/10708 [44:25<1:22:24,  1.97it/s]
  9%|▉         | 988/10708 [44:25<1:21:56,  1.98it/s]
  9%|▉         | 989/10708 [44:26<1:21:34,  1.99it/s]
  9%|▉         | 990/10708 [44:26<1:21:10,  2.00it/s]
  9%|▉         | 991/10708 [44:27<1:23:38,  1.94it/s]
  9%|▉         | 992/10708 [44:27<1:23:13,  1.95it/s]
  9%|▉         | 993/10708 [44:28<1:22:25,  1.96it/s]
  9%|▉         | 994/10708 [44:28<1:22:33,  1.96it/s]
  9%|▉         | 995/10708 [44:29<1:33:01,  1.74it/s]
  9%|▉         | 996/10708 [44:30<1:30:01,  1.80it/s]
  9%|▉         | 997/10708 [44:30<1:39:19,  1.63it/s]
  9%|▉         | 998/10708 [44:31<1:34:44,  1.71it/s]
  9%|▉         | 999/10708 [44:31<1:30:47,  1.78it/s]
  9%|▉         | 1000/10708 [44:32<1:28:02,  1.84it/s]{'loss': 4.4455, 'grad_norm': 0.3237075209617615, 'learning_rate': 0.0009337068160597573, 'epoch': 0.09}

                                                      

  9%|▉         | 1000/10708 [44:32<1:28:02,  1.84it/s]
  9%|▉         | 1001/10708 [44:32<1:26:32,  1.87it/s]
  9%|▉         | 1002/10708 [44:33<1:25:42,  1.89it/s]
  9%|▉         | 1003/10708 [44:33<1:24:49,  1.91it/s]
  9%|▉         | 1004/10708 [44:34<1:23:26,  1.94it/s]
  9%|▉         | 1005/10708 [44:34<1:22:47,  1.95it/s]
  9%|▉         | 1006/10708 [44:35<1:23:06,  1.95it/s]
  9%|▉         | 1007/10708 [44:35<1:22:41,  1.96it/s]
  9%|▉         | 1008/10708 [44:36<1:22:16,  1.96it/s]
  9%|▉         | 1009/10708 [44:36<1:22:04,  1.97it/s]
  9%|▉         | 1010/10708 [44:37<1:22:17,  1.96it/s]
  9%|▉         | 1011/10708 [44:37<1:22:36,  1.96it/s]
  9%|▉         | 1012/10708 [44:38<1:21:56,  1.97it/s]
  9%|▉         | 1013/10708 [44:39<1:34:15,  1.71it/s]
  9%|▉         | 1014/10708 [44:39<1:41:14,  1.60it/s]
  9%|▉         | 1015/10708 [44:40<1:36:17,  1.68it/s]
  9%|▉         | 1016/10708 [44:40<1:32:09,  1.75it/s]
  9%|▉         | 1017/10708 [44:41<1:29:38,  1.80it/s]
 10%|▉         | 1018/10708 [44:41<1:26:50,  1.86it/s]
 10%|▉         | 1019/10708 [44:42<1:25:17,  1.89it/s]
 10%|▉         | 1020/10708 [44:42<1:23:47,  1.93it/s]
 10%|▉         | 1021/10708 [44:43<1:23:24,  1.94it/s]
 10%|▉         | 1022/10708 [44:43<1:22:26,  1.96it/s]
 10%|▉         | 1023/10708 [44:44<1:24:28,  1.91it/s]
 10%|▉         | 1024/10708 [44:45<1:24:53,  1.90it/s]
 10%|▉         | 1025/10708 [44:45<1:24:02,  1.92it/s]{'loss': 4.4435, 'grad_norm': 0.36041951179504395, 'learning_rate': 0.0009570494864612512, 'epoch': 0.1}
                                                      

 10%|▉         | 1025/10708 [44:45<1:24:02,  1.92it/s]
 10%|▉         | 1026/10708 [44:46<1:22:59,  1.94it/s]
 10%|▉         | 1027/10708 [44:46<1:23:24,  1.93it/s]
 10%|▉         | 1028/10708 [44:47<1:23:07,  1.94it/s]
 10%|▉         | 1029/10708 [44:47<1:22:15,  1.96it/s]
 10%|▉         | 1030/10708 [44:48<1:35:42,  1.69it/s]
 10%|▉         | 1031/10708 [44:48<1:30:59,  1.77it/s]
 10%|▉         | 1032/10708 [44:49<1:27:55,  1.83it/s]
 10%|▉         | 1033/10708 [44:49<1:25:59,  1.88it/s]
 10%|▉         | 1034/10708 [44:50<1:24:32,  1.91it/s]
 10%|▉         | 1035/10708 [44:50<1:23:39,  1.93it/s]
 10%|▉         | 1036/10708 [44:51<1:36:43,  1.67it/s]
 10%|▉         | 1037/10708 [44:52<1:31:39,  1.76it/s]
 10%|▉         | 1038/10708 [44:52<1:28:25,  1.82it/s]
 10%|▉         | 1039/10708 [44:53<1:26:41,  1.86it/s]
 10%|▉         | 1040/10708 [44:53<1:25:49,  1.88it/s]
 10%|▉         | 1041/10708 [44:54<1:25:16,  1.89it/s]
 10%|▉         | 1042/10708 [44:54<1:24:01,  1.92it/s]
 10%|▉         | 1043/10708 [44:55<1:22:38,  1.95it/s]
 10%|▉         | 1044/10708 [44:55<1:22:25,  1.95it/s]
 10%|▉         | 1045/10708 [44:56<1:21:54,  1.97it/s]
 10%|▉         | 1046/10708 [44:56<1:21:53,  1.97it/s]
 10%|▉         | 1047/10708 [44:57<1:22:24,  1.95it/s]
 10%|▉         | 1048/10708 [44:57<1:21:32,  1.97it/s]
 10%|▉         | 1049/10708 [44:58<1:21:42,  1.97it/s]
 10%|▉         | 1050/10708 [44:58<1:21:41,  1.97it/s]{'loss': 4.4147, 'grad_norm': 0.3646816313266754, 'learning_rate': 0.000980392156862745, 'epoch': 0.1}
                                                      

 10%|▉         | 1050/10708 [44:58<1:21:41,  1.97it/s]
 10%|▉         | 1051/10708 [44:59<1:21:45,  1.97it/s]
 10%|▉         | 1052/10708 [44:59<1:21:57,  1.96it/s]
 10%|▉         | 1053/10708 [45:00<1:21:27,  1.98it/s]
 10%|▉         | 1054/10708 [45:00<1:21:12,  1.98it/s]
 10%|▉         | 1055/10708 [45:01<1:21:07,  1.98it/s]
 10%|▉         | 1056/10708 [45:01<1:21:50,  1.97it/s]
 10%|▉         | 1057/10708 [45:02<1:21:36,  1.97it/s]
 10%|▉         | 1058/10708 [45:02<1:21:26,  1.97it/s]
 10%|▉         | 1059/10708 [45:03<1:21:58,  1.96it/s]
 10%|▉         | 1060/10708 [45:03<1:22:13,  1.96it/s]
 10%|▉         | 1061/10708 [45:04<1:21:30,  1.97it/s]
 10%|▉         | 1062/10708 [45:04<1:21:21,  1.98it/s]
 10%|▉         | 1063/10708 [45:05<1:20:45,  1.99it/s]
 10%|▉         | 1064/10708 [45:05<1:20:22,  2.00it/s]
 10%|▉         | 1065/10708 [45:06<1:20:51,  1.99it/s]
 10%|▉         | 1066/10708 [45:06<1:21:54,  1.96it/s]
 10%|▉         | 1067/10708 [45:07<1:21:35,  1.97it/s]
 10%|▉         | 1068/10708 [45:07<1:21:28,  1.97it/s]
 10%|▉         | 1069/10708 [45:08<1:22:52,  1.94it/s]
 10%|▉         | 1070/10708 [45:08<1:22:56,  1.94it/s]
 10%|█         | 1071/10708 [45:09<1:22:27,  1.95it/s]
 10%|█         | 1072/10708 [45:09<1:21:41,  1.97it/s]
 10%|█         | 1073/10708 [45:10<1:21:28,  1.97it/s]
 10%|█         | 1074/10708 [45:11<1:21:49,  1.96it/s]
 10%|█         | 1075/10708 [45:11<1:22:10,  1.95it/s]
                                                      
{'loss': 4.3984, 'grad_norm': 0.33610478043556213, 'learning_rate': 0.0009999995749148268, 'epoch': 0.1}

 10%|█         | 1075/10708 [45:11<1:22:10,  1.95it/s]
 10%|█         | 1076/10708 [45:12<1:22:02,  1.96it/s]
 10%|█         | 1077/10708 [45:12<1:21:47,  1.96it/s]
 10%|█         | 1078/10708 [45:13<1:21:25,  1.97it/s]
 10%|█         | 1079/10708 [45:13<1:20:53,  1.98it/s]
 10%|█         | 1080/10708 [45:14<1:32:25,  1.74it/s]
 10%|█         | 1081/10708 [45:14<1:29:33,  1.79it/s]
 10%|█         | 1082/10708 [45:15<1:26:52,  1.85it/s]
 10%|█         | 1083/10708 [45:15<1:25:08,  1.88it/s]
 10%|█         | 1084/10708 [45:16<1:23:23,  1.92it/s]
 10%|█         | 1085/10708 [45:16<1:22:59,  1.93it/s]
 10%|█         | 1086/10708 [45:17<1:22:30,  1.94it/s]
 10%|█         | 1087/10708 [45:17<1:22:13,  1.95it/s]
 10%|█         | 1088/10708 [45:18<1:22:02,  1.95it/s]
 10%|█         | 1089/10708 [45:19<1:32:55,  1.73it/s]
 10%|█         | 1090/10708 [45:19<1:30:40,  1.77it/s]
 10%|█         | 1091/10708 [45:20<1:27:54,  1.82it/s]
 10%|█         | 1092/10708 [45:20<1:35:10,  1.68it/s]
 10%|█         | 1093/10708 [45:21<1:31:10,  1.76it/s]
 10%|█         | 1094/10708 [45:21<1:27:54,  1.82it/s]
 10%|█         | 1095/10708 [45:22<1:25:32,  1.87it/s]
 10%|█         | 1096/10708 [45:22<1:24:03,  1.91it/s]
 10%|█         | 1097/10708 [45:23<1:23:09,  1.93it/s]
 10%|█         | 1098/10708 [45:23<1:22:09,  1.95it/s]
 10%|█         | 1099/10708 [45:24<1:21:23,  1.97it/s]
 10%|█         | 1100/10708 [45:24<1:21:40,  1.96it/s]
                                                      
{'loss': 4.3823, 'grad_norm': 0.3307268023490906, 'learning_rate': 0.0009999776566238286, 'epoch': 0.1}

 10%|█         | 1100/10708 [45:24<1:21:40,  1.96it/s]
 10%|█         | 1101/10708 [45:25<1:21:24,  1.97it/s]
 10%|█         | 1102/10708 [45:25<1:20:43,  1.98it/s]
 10%|█         | 1103/10708 [45:26<1:20:46,  1.98it/s]
 10%|█         | 1104/10708 [45:26<1:20:42,  1.98it/s]
 10%|█         | 1105/10708 [45:27<1:21:17,  1.97it/s]
 10%|█         | 1106/10708 [45:27<1:20:40,  1.98it/s]
 10%|█         | 1107/10708 [45:28<1:21:24,  1.97it/s]
 10%|█         | 1108/10708 [45:28<1:20:55,  1.98it/s]
 10%|█         | 1109/10708 [45:29<1:20:51,  1.98it/s]
 10%|█         | 1110/10708 [45:29<1:20:22,  1.99it/s]
 10%|█         | 1111/10708 [45:30<1:20:37,  1.98it/s]
 10%|█         | 1112/10708 [45:30<1:20:26,  1.99it/s]
 10%|█         | 1113/10708 [45:31<1:20:41,  1.98it/s]
 10%|█         | 1114/10708 [45:31<1:20:40,  1.98it/s]
 10%|█         | 1115/10708 [45:32<1:20:37,  1.98it/s]
 10%|█         | 1116/10708 [45:32<1:20:47,  1.98it/s]
 10%|█         | 1117/10708 [45:33<1:20:57,  1.97it/s]
 10%|█         | 1118/10708 [45:33<1:20:40,  1.98it/s]
 10%|█         | 1119/10708 [45:34<1:20:25,  1.99it/s]
 10%|█         | 1120/10708 [45:34<1:20:28,  1.99it/s]
 10%|█         | 1121/10708 [45:35<1:21:28,  1.96it/s]
 10%|█         | 1122/10708 [45:35<1:21:17,  1.97it/s]
 10%|█         | 1123/10708 [45:36<1:21:15,  1.97it/s]
 10%|█         | 1124/10708 [45:36<1:21:03,  1.97it/s]
 11%|█         | 1125/10708 [45:37<1:20:43,  1.98it/s]{'loss': 4.3688, 'grad_norm': 0.32420244812965393, 'learning_rate': 0.0009999225302168126, 'epoch': 0.11}
                                                      

 11%|█         | 1125/10708 [45:37<1:20:43,  1.98it/s]
 11%|█         | 1126/10708 [45:37<1:20:54,  1.97it/s]
 11%|█         | 1127/10708 [45:38<1:31:08,  1.75it/s]
 11%|█         | 1128/10708 [45:39<1:29:02,  1.79it/s]
 11%|█         | 1129/10708 [45:39<1:26:14,  1.85it/s]
 11%|█         | 1130/10708 [45:40<1:24:54,  1.88it/s]
 11%|█         | 1131/10708 [45:40<1:23:22,  1.91it/s]
 11%|█         | 1132/10708 [45:41<1:22:46,  1.93it/s]
 11%|█         | 1133/10708 [45:41<1:21:46,  1.95it/s]
 11%|█         | 1134/10708 [45:42<1:22:46,  1.93it/s]
 11%|█         | 1135/10708 [45:42<1:21:57,  1.95it/s]
 11%|█         | 1136/10708 [45:43<1:22:19,  1.94it/s]
 11%|█         | 1137/10708 [45:43<1:21:36,  1.95it/s]
 11%|█         | 1138/10708 [45:44<1:21:06,  1.97it/s]
 11%|█         | 1139/10708 [45:44<1:20:37,  1.98it/s]
 11%|█         | 1140/10708 [45:45<1:20:40,  1.98it/s]
 11%|█         | 1141/10708 [45:45<1:20:34,  1.98it/s]
 11%|█         | 1142/10708 [45:46<1:20:02,  1.99it/s]
 11%|█         | 1143/10708 [45:46<1:21:01,  1.97it/s]
 11%|█         | 1144/10708 [45:47<1:20:50,  1.97it/s]
 11%|█         | 1145/10708 [45:47<1:20:42,  1.97it/s]
 11%|█         | 1146/10708 [45:48<1:20:52,  1.97it/s]
 11%|█         | 1147/10708 [45:48<1:20:35,  1.98it/s]
 11%|█         | 1148/10708 [45:49<1:20:17,  1.98it/s]
 11%|█         | 1149/10708 [45:49<1:20:14,  1.99it/s]
 11%|█         | 1150/10708 [45:50<1:20:02,  1.99it/s]
                                                      {'loss': 4.3511, 'grad_norm': 0.269052654504776, 'learning_rate': 0.0009998341993552305, 'epoch': 0.11}


 11%|█         | 1150/10708 [45:50<1:20:02,  1.99it/s]
 11%|█         | 1151/10708 [45:50<1:20:12,  1.99it/s]
 11%|█         | 1152/10708 [45:51<1:20:22,  1.98it/s]
 11%|█         | 1153/10708 [45:51<1:20:27,  1.98it/s]
 11%|█         | 1154/10708 [45:52<1:20:06,  1.99it/s]
 11%|█         | 1155/10708 [45:52<1:19:47,  2.00it/s]
 11%|█         | 1156/10708 [45:53<1:19:56,  1.99it/s]
 11%|█         | 1157/10708 [45:53<1:19:37,  2.00it/s]
 11%|█         | 1158/10708 [45:54<1:20:06,  1.99it/s]
 11%|█         | 1159/10708 [45:54<1:20:06,  1.99it/s]
 11%|█         | 1160/10708 [45:55<1:20:19,  1.98it/s]
 11%|█         | 1161/10708 [45:55<1:20:22,  1.98it/s]
 11%|█         | 1162/10708 [45:56<1:20:04,  1.99it/s]
 11%|█         | 1163/10708 [45:56<1:20:28,  1.98it/s]
 11%|█         | 1164/10708 [45:57<1:20:20,  1.98it/s]
 11%|█         | 1165/10708 [45:57<1:20:11,  1.98it/s]
 11%|█         | 1166/10708 [45:58<1:31:04,  1.75it/s]
 11%|█         | 1167/10708 [45:59<1:27:36,  1.81it/s]
 11%|█         | 1168/10708 [45:59<1:25:29,  1.86it/s]
 11%|█         | 1169/10708 [46:00<1:23:47,  1.90it/s]
 11%|█         | 1170/10708 [46:00<1:22:31,  1.93it/s]
 11%|█         | 1171/10708 [46:01<1:21:27,  1.95it/s]
 11%|█         | 1172/10708 [46:01<1:21:11,  1.96it/s]
 11%|█         | 1173/10708 [46:02<1:20:44,  1.97it/s]
 11%|█         | 1174/10708 [46:02<1:20:25,  1.98it/s]
 11%|█         | 1175/10708 [46:03<1:21:13,  1.96it/s]{'loss': 4.3168, 'grad_norm': 0.27362313866615295, 'learning_rate': 0.0009997126699059474, 'epoch': 0.11}

                                                      

 11%|█         | 1175/10708 [46:03<1:21:13,  1.96it/s]
 11%|█         | 1176/10708 [46:03<1:21:04,  1.96it/s]
 11%|█         | 1177/10708 [46:04<1:20:25,  1.98it/s]
 11%|█         | 1178/10708 [46:04<1:20:25,  1.97it/s]
 11%|█         | 1179/10708 [46:05<1:19:54,  1.99it/s]
 11%|█         | 1180/10708 [46:05<1:20:04,  1.98it/s]
 11%|█         | 1181/10708 [46:06<1:20:00,  1.98it/s]
 11%|█         | 1182/10708 [46:06<1:19:56,  1.99it/s]
 11%|█         | 1183/10708 [46:07<1:20:23,  1.97it/s]
 11%|█         | 1184/10708 [46:07<1:20:13,  1.98it/s]
 11%|█         | 1185/10708 [46:08<1:19:38,  1.99it/s]
 11%|█         | 1186/10708 [46:08<1:19:36,  1.99it/s]
 11%|█         | 1187/10708 [46:09<1:19:45,  1.99it/s]
 11%|█         | 1188/10708 [46:09<1:20:02,  1.98it/s]
 11%|█         | 1189/10708 [46:10<1:19:30,  2.00it/s]
 11%|█         | 1190/10708 [46:10<1:19:54,  1.99it/s]
 11%|█         | 1191/10708 [46:11<1:19:55,  1.98it/s]
 11%|█         | 1192/10708 [46:11<1:19:28,  2.00it/s]
 11%|█         | 1193/10708 [46:12<1:20:38,  1.97it/s]
 11%|█         | 1194/10708 [46:12<1:20:13,  1.98it/s]
 11%|█         | 1195/10708 [46:13<1:20:39,  1.97it/s]
 11%|█         | 1196/10708 [46:13<1:20:57,  1.96it/s]
 11%|█         | 1197/10708 [46:14<1:20:05,  1.98it/s]
 11%|█         | 1198/10708 [46:14<1:19:56,  1.98it/s]
 11%|█         | 1199/10708 [46:15<1:19:41,  1.99it/s]
 11%|█         | 1200/10708 [46:15<1:20:13,  1.98it/s]
                                                      
{'loss': 4.2993, 'grad_norm': 0.2982044816017151, 'learning_rate': 0.0009995579499408522, 'epoch': 0.11}

 11%|█         | 1200/10708 [46:15<1:20:13,  1.98it/s]
 11%|█         | 1201/10708 [46:16<1:20:30,  1.97it/s]
 11%|█         | 1202/10708 [46:16<1:20:25,  1.97it/s]
 11%|█         | 1203/10708 [46:17<1:20:21,  1.97it/s]
 11%|█         | 1204/10708 [46:17<1:20:14,  1.97it/s]
 11%|█▏        | 1205/10708 [46:18<1:20:30,  1.97it/s]
 11%|█▏        | 1206/10708 [46:18<1:19:54,  1.98it/s]
 11%|█▏        | 1207/10708 [46:19<1:19:50,  1.98it/s]
 11%|█▏        | 1208/10708 [46:19<1:19:24,  1.99it/s]
 11%|█▏        | 1209/10708 [46:20<1:19:18,  2.00it/s]
 11%|█▏        | 1210/10708 [46:20<1:19:26,  1.99it/s]
 11%|█▏        | 1211/10708 [46:21<1:19:12,  2.00it/s]
 11%|█▏        | 1212/10708 [46:21<1:19:24,  1.99it/s]
 11%|█▏        | 1213/10708 [46:22<1:19:43,  1.98it/s]
 11%|█▏        | 1214/10708 [46:22<1:19:21,  1.99it/s]
 11%|█▏        | 1215/10708 [46:23<1:19:45,  1.98it/s]
 11%|█▏        | 1216/10708 [46:23<1:19:40,  1.99it/s]
 11%|█▏        | 1217/10708 [46:24<1:19:34,  1.99it/s]
 11%|█▏        | 1218/10708 [46:24<1:19:16,  2.00it/s]
 11%|█▏        | 1219/10708 [46:25<1:19:19,  1.99it/s]
 11%|█▏        | 1220/10708 [46:25<1:19:36,  1.99it/s]
 11%|█▏        | 1221/10708 [46:26<1:19:36,  1.99it/s]
 11%|█▏        | 1222/10708 [46:26<1:19:25,  1.99it/s]
 11%|█▏        | 1223/10708 [46:27<1:19:33,  1.99it/s]
 11%|█▏        | 1224/10708 [46:27<1:19:27,  1.99it/s]
 11%|█▏        | 1225/10708 [46:28<1:19:04,  2.00it/s]
                                                      
{'loss': 4.2955, 'grad_norm': 0.3212431073188782, 'learning_rate': 0.0009993700497363213, 'epoch': 0.11}

 11%|█▏        | 1225/10708 [46:28<1:19:04,  2.00it/s]
 11%|█▏        | 1226/10708 [46:28<1:18:59,  2.00it/s]
 11%|█▏        | 1227/10708 [46:29<1:18:47,  2.01it/s]
 11%|█▏        | 1228/10708 [46:30<1:28:59,  1.78it/s]
 11%|█▏        | 1229/10708 [46:30<1:26:11,  1.83it/s]
 11%|█▏        | 1230/10708 [46:31<1:24:54,  1.86it/s]
 11%|█▏        | 1231/10708 [46:31<1:23:10,  1.90it/s]
 12%|█▏        | 1232/10708 [46:32<1:21:45,  1.93it/s]
 12%|█▏        | 1233/10708 [46:32<1:20:45,  1.96it/s]
 12%|█▏        | 1234/10708 [46:33<1:19:53,  1.98it/s]
 12%|█▏        | 1235/10708 [46:33<1:19:29,  1.99it/s]
 12%|█▏        | 1236/10708 [46:34<1:19:09,  1.99it/s]
 12%|█▏        | 1237/10708 [46:34<1:19:13,  1.99it/s]
 12%|█▏        | 1238/10708 [46:35<1:19:02,  2.00it/s]
 12%|█▏        | 1239/10708 [46:35<1:18:49,  2.00it/s]
 12%|█▏        | 1240/10708 [46:36<1:18:32,  2.01it/s]
 12%|█▏        | 1241/10708 [46:36<1:18:27,  2.01it/s]
 12%|█▏        | 1242/10708 [46:37<1:18:36,  2.01it/s]
 12%|█▏        | 1243/10708 [46:37<1:18:47,  2.00it/s]
 12%|█▏        | 1244/10708 [46:38<1:19:01,  2.00it/s]
 12%|█▏        | 1245/10708 [46:38<1:18:46,  2.00it/s]
 12%|█▏        | 1246/10708 [46:39<1:18:25,  2.01it/s]
 12%|█▏        | 1247/10708 [46:39<1:18:22,  2.01it/s]
 12%|█▏        | 1248/10708 [46:40<1:18:21,  2.01it/s]
 12%|█▏        | 1249/10708 [46:40<1:18:15,  2.01it/s]
 12%|█▏        | 1250/10708 [46:41<1:18:35,  2.01it/s]
                                                      {'loss': 4.2716, 'grad_norm': 0.2938788831233978, 'learning_rate': 0.000999148981772536, 'epoch': 0.12}


 12%|█▏        | 1250/10708 [46:41<1:18:35,  2.01it/s]
 12%|█▏        | 1251/10708 [46:41<1:18:28,  2.01it/s]
 12%|█▏        | 1252/10708 [46:42<1:18:21,  2.01it/s]
 12%|█▏        | 1253/10708 [46:42<1:18:21,  2.01it/s]
 12%|█▏        | 1254/10708 [46:43<1:18:25,  2.01it/s]
 12%|█▏        | 1255/10708 [46:43<1:18:21,  2.01it/s]
 12%|█▏        | 1256/10708 [46:44<1:18:37,  2.00it/s]
 12%|█▏        | 1257/10708 [46:44<1:18:52,  2.00it/s]
 12%|█▏        | 1258/10708 [46:45<1:18:44,  2.00it/s]
 12%|█▏        | 1259/10708 [46:45<1:18:33,  2.00it/s]
 12%|█▏        | 1260/10708 [46:46<1:19:14,  1.99it/s]
 12%|█▏        | 1261/10708 [46:46<1:19:36,  1.98it/s]
 12%|█▏        | 1262/10708 [46:47<1:19:18,  1.99it/s]
 12%|█▏        | 1263/10708 [46:47<1:18:53,  2.00it/s]
 12%|█▏        | 1264/10708 [46:48<1:18:32,  2.00it/s]
 12%|█▏        | 1265/10708 [46:48<1:18:29,  2.00it/s]
 12%|█▏        | 1266/10708 [46:49<1:18:25,  2.01it/s]
 12%|█▏        | 1267/10708 [46:49<1:18:01,  2.02it/s]
 12%|█▏        | 1268/10708 [46:50<1:18:06,  2.01it/s]
 12%|█▏        | 1269/10708 [46:50<1:18:16,  2.01it/s]
 12%|█▏        | 1270/10708 [46:51<1:18:44,  2.00it/s]
 12%|█▏        | 1271/10708 [46:51<1:18:46,  2.00it/s]
 12%|█▏        | 1272/10708 [46:52<1:18:32,  2.00it/s]
 12%|█▏        | 1273/10708 [46:52<1:18:46,  2.00it/s]
 12%|█▏        | 1274/10708 [46:53<1:18:36,  2.00it/s]
 12%|█▏        | 1275/10708 [46:53<1:29:08,  1.76it/s]
                                                      
{'loss': 4.2507, 'grad_norm': 0.280631422996521, 'learning_rate': 0.0009988947607326533, 'epoch': 0.12}

 12%|█▏        | 1275/10708 [46:53<1:29:08,  1.76it/s]
 12%|█▏        | 1276/10708 [46:54<1:26:21,  1.82it/s]
 12%|█▏        | 1277/10708 [46:54<1:24:17,  1.86it/s]
 12%|█▏        | 1278/10708 [46:55<1:22:32,  1.90it/s]
 12%|█▏        | 1279/10708 [46:55<1:21:03,  1.94it/s]
 12%|█▏        | 1280/10708 [46:56<1:20:54,  1.94it/s]
 12%|█▏        | 1281/10708 [46:56<1:20:02,  1.96it/s]
 12%|█▏        | 1282/10708 [46:57<1:20:00,  1.96it/s]
 12%|█▏        | 1283/10708 [46:57<1:19:30,  1.98it/s]
 12%|█▏        | 1284/10708 [46:58<1:19:43,  1.97it/s]
 12%|█▏        | 1285/10708 [46:58<1:19:17,  1.98it/s]
 12%|█▏        | 1286/10708 [46:59<1:29:11,  1.76it/s]
 12%|█▏        | 1287/10708 [47:00<1:25:46,  1.83it/s]
 12%|█▏        | 1288/10708 [47:00<1:24:07,  1.87it/s]
 12%|█▏        | 1289/10708 [47:01<1:22:39,  1.90it/s]
 12%|█▏        | 1290/10708 [47:01<1:21:36,  1.92it/s]
 12%|█▏        | 1291/10708 [47:02<1:20:29,  1.95it/s]
 12%|█▏        | 1292/10708 [47:02<1:19:36,  1.97it/s]
 12%|█▏        | 1293/10708 [47:03<1:19:13,  1.98it/s]
 12%|█▏        | 1294/10708 [47:03<1:18:44,  1.99it/s]
 12%|█▏        | 1295/10708 [47:04<1:18:31,  2.00it/s]
 12%|█▏        | 1296/10708 [47:04<1:18:23,  2.00it/s]
 12%|█▏        | 1297/10708 [47:05<1:19:23,  1.98it/s]
 12%|█▏        | 1298/10708 [47:05<1:18:55,  1.99it/s]
 12%|█▏        | 1299/10708 [47:06<1:18:30,  2.00it/s]
 12%|█▏        | 1300/10708 [47:06<1:18:56,  1.99it/s]
                                                      
{'loss': 4.2445, 'grad_norm': 0.2629048228263855, 'learning_rate': 0.0009986074035018317, 'epoch': 0.12}

 12%|█▏        | 1300/10708 [47:06<1:18:56,  1.99it/s]
 12%|█▏        | 1301/10708 [47:07<1:18:49,  1.99it/s]
 12%|█▏        | 1302/10708 [47:07<1:18:49,  1.99it/s]
 12%|█▏        | 1303/10708 [47:08<1:18:39,  1.99it/s]
 12%|█▏        | 1304/10708 [47:08<1:18:46,  1.99it/s]
 12%|█▏        | 1305/10708 [47:09<1:18:33,  1.99it/s]
 12%|█▏        | 1306/10708 [47:09<1:18:47,  1.99it/s]
 12%|█▏        | 1307/10708 [47:10<1:18:24,  2.00it/s]
 12%|█▏        | 1308/10708 [47:10<1:18:16,  2.00it/s]
 12%|█▏        | 1309/10708 [47:11<1:18:00,  2.01it/s]
 12%|█▏        | 1310/10708 [47:11<1:17:48,  2.01it/s]
 12%|█▏        | 1311/10708 [47:12<1:17:44,  2.01it/s]
 12%|█▏        | 1312/10708 [47:12<1:17:54,  2.01it/s]
 12%|█▏        | 1313/10708 [47:13<1:18:05,  2.01it/s]
 12%|█▏        | 1314/10708 [47:13<1:18:38,  1.99it/s]
 12%|█▏        | 1315/10708 [47:14<1:18:15,  2.00it/s]
 12%|█▏        | 1316/10708 [47:14<1:18:20,  2.00it/s]
 12%|█▏        | 1317/10708 [47:15<1:18:46,  1.99it/s]
 12%|█▏        | 1318/10708 [47:15<1:18:52,  1.98it/s]
 12%|█▏        | 1319/10708 [47:16<1:18:27,  1.99it/s]
 12%|█▏        | 1320/10708 [47:16<1:18:14,  2.00it/s]
 12%|█▏        | 1321/10708 [47:17<1:18:37,  1.99it/s]
 12%|█▏        | 1322/10708 [47:17<1:19:05,  1.98it/s]
 12%|█▏        | 1323/10708 [47:18<1:18:26,  1.99it/s]
 12%|█▏        | 1324/10708 [47:18<1:18:14,  2.00it/s]
 12%|█▏        | 1325/10708 [47:19<1:18:16,  2.00it/s]
                                                      {'loss': 4.2248, 'grad_norm': 0.27730560302734375, 'learning_rate': 0.0009982869291661081, 'epoch': 0.12}


 12%|█▏        | 1325/10708 [47:19<1:18:16,  2.00it/s]
 12%|█▏        | 1326/10708 [47:19<1:18:24,  1.99it/s]
 12%|█▏        | 1327/10708 [47:20<1:18:29,  1.99it/s]
 12%|█▏        | 1328/10708 [47:20<1:18:31,  1.99it/s]
 12%|█▏        | 1329/10708 [47:21<1:18:18,  2.00it/s]
 12%|█▏        | 1330/10708 [47:21<1:18:05,  2.00it/s]
 12%|█▏        | 1331/10708 [47:22<1:17:59,  2.00it/s]
 12%|█▏        | 1332/10708 [47:22<1:18:27,  1.99it/s]
 12%|█▏        | 1333/10708 [47:23<1:18:54,  1.98it/s]
 12%|█▏        | 1334/10708 [47:23<1:18:15,  2.00it/s]
 12%|█▏        | 1335/10708 [47:24<1:18:11,  2.00it/s]
 12%|█▏        | 1336/10708 [47:24<1:17:57,  2.00it/s]
 12%|█▏        | 1337/10708 [47:25<1:17:46,  2.01it/s]
 12%|█▏        | 1338/10708 [47:25<1:27:58,  1.78it/s]
 13%|█▎        | 1339/10708 [47:26<1:25:08,  1.83it/s]
 13%|█▎        | 1340/10708 [47:26<1:23:13,  1.88it/s]
 13%|█▎        | 1341/10708 [47:27<1:21:26,  1.92it/s]
 13%|█▎        | 1342/10708 [47:27<1:20:20,  1.94it/s]
 13%|█▎        | 1343/10708 [47:28<1:19:54,  1.95it/s]
 13%|█▎        | 1344/10708 [47:28<1:19:03,  1.97it/s]
 13%|█▎        | 1345/10708 [47:29<1:18:28,  1.99it/s]
 13%|█▎        | 1346/10708 [47:29<1:18:16,  1.99it/s]
 13%|█▎        | 1347/10708 [47:30<1:18:15,  1.99it/s]
 13%|█▎        | 1348/10708 [47:30<1:28:08,  1.77it/s]
 13%|█▎        | 1349/10708 [47:31<1:25:07,  1.83it/s]
 13%|█▎        | 1350/10708 [47:31<1:22:37,  1.89it/s]
                                                      
{'loss': 4.2136, 'grad_norm': 0.28000473976135254, 'learning_rate': 0.0009979333590111321, 'epoch': 0.13}

 13%|█▎        | 1350/10708 [47:31<1:22:37,  1.89it/s]
 13%|█▎        | 1351/10708 [47:33<2:01:41,  1.28it/s]
 13%|█▎        | 1352/10708 [47:33<1:48:20,  1.44it/s]
 13%|█▎        | 1353/10708 [47:34<1:39:46,  1.56it/s]
 13%|█▎        | 1354/10708 [47:34<1:33:19,  1.67it/s]
 13%|█▎        | 1355/10708 [47:35<1:28:55,  1.75it/s]
 13%|█▎        | 1356/10708 [47:35<1:25:19,  1.83it/s]
 13%|█▎        | 1357/10708 [47:36<1:23:41,  1.86it/s]
 13%|█▎        | 1358/10708 [47:36<1:22:00,  1.90it/s]
 13%|█▎        | 1359/10708 [47:37<1:20:53,  1.93it/s]
 13%|█▎        | 1360/10708 [47:37<1:20:13,  1.94it/s]
 13%|█▎        | 1361/10708 [47:38<1:19:13,  1.97it/s]
 13%|█▎        | 1362/10708 [47:38<1:18:33,  1.98it/s]
 13%|█▎        | 1363/10708 [47:39<1:18:11,  1.99it/s]
 13%|█▎        | 1364/10708 [47:39<1:18:10,  1.99it/s]
 13%|█▎        | 1365/10708 [47:40<1:17:56,  2.00it/s]
 13%|█▎        | 1366/10708 [47:40<1:17:39,  2.00it/s]
 13%|█▎        | 1367/10708 [47:41<1:17:34,  2.01it/s]
 13%|█▎        | 1368/10708 [47:41<1:18:01,  2.00it/s]
 13%|█▎        | 1369/10708 [47:42<1:18:06,  1.99it/s]
 13%|█▎        | 1370/10708 [47:42<1:18:05,  1.99it/s]
 13%|█▎        | 1371/10708 [47:43<1:28:41,  1.75it/s]
 13%|█▎        | 1372/10708 [47:44<1:25:20,  1.82it/s]
 13%|█▎        | 1373/10708 [47:44<1:22:41,  1.88it/s]
 13%|█▎        | 1374/10708 [47:45<1:21:02,  1.92it/s]
 13%|█▎        | 1375/10708 [47:45<1:20:12,  1.94it/s]{'loss': 4.2038, 'grad_norm': 0.2727168798446655, 'learning_rate': 0.0009975467165207505, 'epoch': 0.13}
                                                      

 13%|█▎        | 1375/10708 [47:45<1:20:12,  1.94it/s]
 13%|█▎        | 1376/10708 [47:46<1:19:22,  1.96it/s]
 13%|█▎        | 1377/10708 [47:46<1:18:36,  1.98it/s]
 13%|█▎        | 1378/10708 [47:47<1:18:45,  1.97it/s]
 13%|█▎        | 1379/10708 [47:47<1:18:44,  1.97it/s]
 13%|█▎        | 1380/10708 [47:48<1:18:43,  1.97it/s]
 13%|█▎        | 1381/10708 [47:48<1:18:09,  1.99it/s]
 13%|█▎        | 1382/10708 [47:49<1:17:53,  2.00it/s]
 13%|█▎        | 1383/10708 [47:49<1:17:49,  2.00it/s]
 13%|█▎        | 1384/10708 [47:50<1:17:58,  1.99it/s]
 13%|█▎        | 1385/10708 [47:50<1:17:49,  2.00it/s]
 13%|█▎        | 1386/10708 [47:51<1:17:50,  2.00it/s]
 13%|█▎        | 1387/10708 [47:51<1:28:04,  1.76it/s]
 13%|█▎        | 1388/10708 [47:52<1:24:46,  1.83it/s]
 13%|█▎        | 1389/10708 [47:52<1:23:16,  1.87it/s]
 13%|█▎        | 1390/10708 [47:53<1:21:36,  1.90it/s]
 13%|█▎        | 1391/10708 [47:54<1:30:00,  1.73it/s]
 13%|█▎        | 1392/10708 [47:54<1:26:05,  1.80it/s]
 13%|█▎        | 1393/10708 [47:55<1:24:19,  1.84it/s]
 13%|█▎        | 1394/10708 [47:55<1:22:35,  1.88it/s]
 13%|█▎        | 1395/10708 [47:56<1:21:39,  1.90it/s]
 13%|█▎        | 1396/10708 [47:56<1:20:26,  1.93it/s]
 13%|█▎        | 1397/10708 [47:57<1:21:42,  1.90it/s]
 13%|█▎        | 1398/10708 [47:57<1:20:40,  1.92it/s]
 13%|█▎        | 1399/10708 [47:58<1:19:32,  1.95it/s]
 13%|█▎        | 1400/10708 [47:58<1:19:21,  1.95it/s]
                                                      
{'loss': 4.1847, 'grad_norm': 0.24278225004673004, 'learning_rate': 0.0009971270273754478, 'epoch': 0.13}

 13%|█▎        | 1400/10708 [47:58<1:19:21,  1.95it/s]
 13%|█▎        | 1401/10708 [47:59<1:18:58,  1.96it/s]
 13%|█▎        | 1402/10708 [47:59<1:18:21,  1.98it/s]
 13%|█▎        | 1403/10708 [48:00<1:18:07,  1.98it/s]
 13%|█▎        | 1404/10708 [48:00<1:17:57,  1.99it/s]
 13%|█▎        | 1405/10708 [48:01<1:17:53,  1.99it/s]
 13%|█▎        | 1406/10708 [48:01<1:17:30,  2.00it/s]
 13%|█▎        | 1407/10708 [48:02<1:17:21,  2.00it/s]
 13%|█▎        | 1408/10708 [48:02<1:17:57,  1.99it/s]
 13%|█▎        | 1409/10708 [48:03<1:17:40,  2.00it/s]
 13%|█▎        | 1410/10708 [48:03<1:18:30,  1.97it/s]
 13%|█▎        | 1411/10708 [48:04<1:18:40,  1.97it/s]
 13%|█▎        | 1412/10708 [48:04<1:18:18,  1.98it/s]
 13%|█▎        | 1413/10708 [48:05<1:18:17,  1.98it/s]
 13%|█▎        | 1414/10708 [48:05<1:18:16,  1.98it/s]
 13%|█▎        | 1415/10708 [48:06<1:18:10,  1.98it/s]
 13%|█▎        | 1416/10708 [48:06<1:17:52,  1.99it/s]
 13%|█▎        | 1417/10708 [48:07<1:17:37,  2.00it/s]
 13%|█▎        | 1418/10708 [48:07<1:17:11,  2.01it/s]
 13%|█▎        | 1419/10708 [48:08<1:17:06,  2.01it/s]
 13%|█▎        | 1420/10708 [48:08<1:17:05,  2.01it/s]
 13%|█▎        | 1421/10708 [48:09<1:17:17,  2.00it/s]
 13%|█▎        | 1422/10708 [48:09<1:17:01,  2.01it/s]
 13%|█▎        | 1423/10708 [48:10<1:16:55,  2.01it/s]
 13%|█▎        | 1424/10708 [48:10<1:16:55,  2.01it/s]
 13%|█▎        | 1425/10708 [48:11<1:17:01,  2.01it/s]{'loss': 4.1826, 'grad_norm': 0.23841816186904907, 'learning_rate': 0.0009966743194506418, 'epoch': 0.13}
                                                      

 13%|█▎        | 1425/10708 [48:11<1:17:01,  2.01it/s]
 13%|█▎        | 1426/10708 [48:11<1:17:52,  1.99it/s]
 13%|█▎        | 1427/10708 [48:12<1:18:08,  1.98it/s]
 13%|█▎        | 1428/10708 [48:12<1:18:04,  1.98it/s]
 13%|█▎        | 1429/10708 [48:13<1:18:32,  1.97it/s]
 13%|█▎        | 1430/10708 [48:13<1:19:38,  1.94it/s]
 13%|█▎        | 1431/10708 [48:14<1:19:02,  1.96it/s]
 13%|█▎        | 1432/10708 [48:14<1:18:09,  1.98it/s]
 13%|█▎        | 1433/10708 [48:15<1:17:46,  1.99it/s]
 13%|█▎        | 1434/10708 [48:15<1:18:03,  1.98it/s]
 13%|█▎        | 1435/10708 [48:16<1:17:51,  1.98it/s]
 13%|█▎        | 1436/10708 [48:16<1:17:41,  1.99it/s]
 13%|█▎        | 1437/10708 [48:17<1:17:18,  2.00it/s]
 13%|█▎        | 1438/10708 [48:17<1:17:07,  2.00it/s]
 13%|█▎        | 1439/10708 [48:18<1:17:28,  1.99it/s]
 13%|█▎        | 1440/10708 [48:18<1:17:00,  2.01it/s]
 13%|█▎        | 1441/10708 [48:19<1:17:07,  2.00it/s]
 13%|█▎        | 1442/10708 [48:19<1:17:52,  1.98it/s]
 13%|█▎        | 1443/10708 [48:20<1:18:03,  1.98it/s]
 13%|█▎        | 1444/10708 [48:20<1:17:31,  1.99it/s]
 13%|█▎        | 1445/10708 [48:21<1:17:58,  1.98it/s]
 14%|█▎        | 1446/10708 [48:21<1:17:30,  1.99it/s]
 14%|█▎        | 1447/10708 [48:22<1:17:08,  2.00it/s]
 14%|█▎        | 1448/10708 [48:22<1:17:15,  2.00it/s]
 14%|█▎        | 1449/10708 [48:23<1:17:13,  2.00it/s]
 14%|█▎        | 1450/10708 [48:23<1:16:43,  2.01it/s]
                                                      {'loss': 4.1685, 'grad_norm': 0.24806207418441772, 'learning_rate': 0.0009961886228148306, 'epoch': 0.14}


 14%|█▎        | 1450/10708 [48:23<1:16:43,  2.01it/s]
 14%|█▎        | 1451/10708 [48:24<1:17:05,  2.00it/s]
 14%|█▎        | 1452/10708 [48:24<1:16:45,  2.01it/s]
 14%|█▎        | 1453/10708 [48:25<1:16:47,  2.01it/s]
 14%|█▎        | 1454/10708 [48:25<1:16:37,  2.01it/s]
 14%|█▎        | 1455/10708 [48:26<1:16:27,  2.02it/s]
 14%|█▎        | 1456/10708 [48:26<1:16:40,  2.01it/s]
 14%|█▎        | 1457/10708 [48:27<1:16:27,  2.02it/s]
 14%|█▎        | 1458/10708 [48:27<1:16:18,  2.02it/s]
 14%|█▎        | 1459/10708 [48:28<1:16:27,  2.02it/s]
 14%|█▎        | 1460/10708 [48:28<1:16:22,  2.02it/s]
 14%|█▎        | 1461/10708 [48:29<1:16:30,  2.01it/s]
 14%|█▎        | 1462/10708 [48:29<1:16:35,  2.01it/s]
 14%|█▎        | 1463/10708 [48:30<1:17:01,  2.00it/s]
 14%|█▎        | 1464/10708 [48:30<1:17:08,  2.00it/s]
 14%|█▎        | 1465/10708 [48:31<1:27:05,  1.77it/s]
 14%|█▎        | 1466/10708 [48:31<1:23:53,  1.84it/s]
 14%|█▎        | 1467/10708 [48:32<1:22:03,  1.88it/s]
 14%|█▎        | 1468/10708 [48:32<1:20:18,  1.92it/s]
 14%|█▎        | 1469/10708 [48:33<1:19:15,  1.94it/s]
 14%|█▎        | 1470/10708 [48:33<1:18:16,  1.97it/s]
 14%|█▎        | 1471/10708 [48:34<1:18:01,  1.97it/s]
 14%|█▎        | 1472/10708 [48:34<1:17:22,  1.99it/s]
 14%|█▍        | 1473/10708 [48:35<1:27:19,  1.76it/s]
 14%|█▍        | 1474/10708 [48:36<1:23:54,  1.83it/s]
 14%|█▍        | 1475/10708 [48:36<1:21:37,  1.89it/s]
                                                      
{'loss': 4.1558, 'grad_norm': 0.24821878969669342, 'learning_rate': 0.000995669969727596, 'epoch': 0.14}

 14%|█▍        | 1475/10708 [48:36<1:21:37,  1.89it/s]
 14%|█▍        | 1476/10708 [48:37<1:20:11,  1.92it/s]
 14%|█▍        | 1477/10708 [48:37<1:18:55,  1.95it/s]
 14%|█▍        | 1478/10708 [48:38<1:18:05,  1.97it/s]
 14%|█▍        | 1479/10708 [48:38<1:17:24,  1.99it/s]
 14%|█▍        | 1480/10708 [48:39<1:17:17,  1.99it/s]
 14%|█▍        | 1481/10708 [48:39<1:17:02,  2.00it/s]
 14%|█▍        | 1482/10708 [48:40<1:16:48,  2.00it/s]
 14%|█▍        | 1483/10708 [48:40<1:16:31,  2.01it/s]
 14%|█▍        | 1484/10708 [48:41<1:16:27,  2.01it/s]
 14%|█▍        | 1485/10708 [48:41<1:16:23,  2.01it/s]
 14%|█▍        | 1486/10708 [48:42<1:16:43,  2.00it/s]
 14%|█▍        | 1487/10708 [48:42<1:17:24,  1.99it/s]
 14%|█▍        | 1488/10708 [48:43<1:17:13,  1.99it/s]
 14%|█▍        | 1489/10708 [48:43<1:16:57,  2.00it/s]
 14%|█▍        | 1490/10708 [48:44<1:16:47,  2.00it/s]
 14%|█▍        | 1491/10708 [48:44<1:16:24,  2.01it/s]
 14%|█▍        | 1492/10708 [48:45<1:16:20,  2.01it/s]
 14%|█▍        | 1493/10708 [48:45<1:16:09,  2.02it/s]
 14%|█▍        | 1494/10708 [48:46<1:16:28,  2.01it/s]
 14%|█▍        | 1495/10708 [48:46<1:16:25,  2.01it/s]
 14%|█▍        | 1496/10708 [48:47<1:16:31,  2.01it/s]
 14%|█▍        | 1497/10708 [48:47<1:16:22,  2.01it/s]
 14%|█▍        | 1498/10708 [48:48<1:16:15,  2.01it/s]
 14%|█▍        | 1499/10708 [48:48<1:16:08,  2.02it/s]
 14%|█▍        | 1500/10708 [48:49<1:16:03,  2.02it/s]{'loss': 4.1462, 'grad_norm': 0.2577230930328369, 'learning_rate': 0.0009951183946374615, 'epoch': 0.14}
                                                      

 14%|█▍        | 1500/10708 [48:49<1:16:03,  2.02it/s]
 14%|█▍        | 1501/10708 [48:49<1:16:32,  2.00it/s]
 14%|█▍        | 1502/10708 [48:50<1:16:36,  2.00it/s]
 14%|█▍        | 1503/10708 [48:50<1:16:33,  2.00it/s]
 14%|█▍        | 1504/10708 [48:51<1:16:20,  2.01it/s]
 14%|█▍        | 1505/10708 [48:51<1:16:18,  2.01it/s]
 14%|█▍        | 1506/10708 [48:52<1:16:19,  2.01it/s]
 14%|█▍        | 1507/10708 [48:52<1:16:07,  2.01it/s]
 14%|█▍        | 1508/10708 [48:53<1:16:24,  2.01it/s]
 14%|█▍        | 1509/10708 [48:53<1:16:15,  2.01it/s]
 14%|█▍        | 1510/10708 [48:54<1:16:26,  2.01it/s]
 14%|█▍        | 1511/10708 [48:54<1:16:32,  2.00it/s]
 14%|█▍        | 1512/10708 [48:55<1:16:28,  2.00it/s]
 14%|█▍        | 1513/10708 [48:55<1:16:17,  2.01it/s]
 14%|█▍        | 1514/10708 [48:56<1:16:16,  2.01it/s]
 14%|█▍        | 1515/10708 [48:56<1:16:07,  2.01it/s]
 14%|█▍        | 1516/10708 [48:57<1:16:01,  2.01it/s]
 14%|█▍        | 1517/10708 [48:57<1:16:16,  2.01it/s]
 14%|█▍        | 1518/10708 [48:57<1:16:03,  2.01it/s]
 14%|█▍        | 1519/10708 [48:58<1:16:00,  2.01it/s]
 14%|█▍        | 1520/10708 [48:58<1:16:27,  2.00it/s]
 14%|█▍        | 1521/10708 [48:59<1:16:07,  2.01it/s]
 14%|█▍        | 1522/10708 [48:59<1:16:23,  2.00it/s]
 14%|█▍        | 1523/10708 [49:00<1:16:32,  2.00it/s]
 14%|█▍        | 1524/10708 [49:00<1:16:20,  2.01it/s]
 14%|█▍        | 1525/10708 [49:01<1:16:12,  2.01it/s]
                                                      {'loss': 4.1378, 'grad_norm': 0.2557726800441742, 'learning_rate': 0.0009945339341796032, 'epoch': 0.14}


 14%|█▍        | 1525/10708 [49:01<1:16:12,  2.01it/s]
 14%|█▍        | 1526/10708 [49:01<1:16:11,  2.01it/s]
 14%|█▍        | 1527/10708 [49:02<1:15:58,  2.01it/s]
 14%|█▍        | 1528/10708 [49:02<1:16:48,  1.99it/s]
 14%|█▍        | 1529/10708 [49:03<1:16:27,  2.00it/s]
 14%|█▍        | 1530/10708 [49:03<1:16:18,  2.00it/s]
 14%|█▍        | 1531/10708 [49:04<1:16:33,  2.00it/s]
 14%|█▍        | 1532/10708 [49:04<1:16:28,  2.00it/s]
 14%|█▍        | 1533/10708 [49:05<1:16:23,  2.00it/s]
 14%|█▍        | 1534/10708 [49:05<1:16:14,  2.01it/s]
 14%|█▍        | 1535/10708 [49:06<1:16:08,  2.01it/s]
 14%|█▍        | 1536/10708 [49:06<1:16:11,  2.01it/s]
 14%|█▍        | 1537/10708 [49:07<1:16:20,  2.00it/s]
 14%|█▍        | 1538/10708 [49:07<1:15:58,  2.01it/s]
 14%|█▍        | 1539/10708 [49:08<1:15:57,  2.01it/s]
 14%|█▍        | 1540/10708 [49:08<1:15:51,  2.01it/s]
 14%|█▍        | 1541/10708 [49:09<1:15:53,  2.01it/s]
 14%|█▍        | 1542/10708 [49:09<1:15:53,  2.01it/s]
 14%|█▍        | 1543/10708 [49:10<1:16:22,  2.00it/s]
 14%|█▍        | 1544/10708 [49:10<1:16:39,  1.99it/s]
 14%|█▍        | 1545/10708 [49:11<1:16:19,  2.00it/s]
 14%|█▍        | 1546/10708 [49:11<1:16:13,  2.00it/s]
 14%|█▍        | 1547/10708 [49:12<1:15:58,  2.01it/s]
 14%|█▍        | 1548/10708 [49:12<1:15:40,  2.02it/s]
 14%|█▍        | 1549/10708 [49:13<1:15:42,  2.02it/s]
 14%|█▍        | 1550/10708 [49:13<1:15:38,  2.02it/s]{'loss': 4.1201, 'grad_norm': 0.27535998821258545, 'learning_rate': 0.0009939166271734172, 'epoch': 0.14}
                                                      

 14%|█▍        | 1550/10708 [49:13<1:15:38,  2.02it/s]
 14%|█▍        | 1551/10708 [49:14<1:15:34,  2.02it/s]
 14%|█▍        | 1552/10708 [49:14<1:15:32,  2.02it/s]
 15%|█▍        | 1553/10708 [49:15<1:15:28,  2.02it/s]
 15%|█▍        | 1554/10708 [49:15<1:15:33,  2.02it/s]
 15%|█▍        | 1555/10708 [49:16<1:15:33,  2.02it/s]
 15%|█▍        | 1556/10708 [49:16<1:15:31,  2.02it/s]
 15%|█▍        | 1557/10708 [49:17<1:16:33,  1.99it/s]
 15%|█▍        | 1558/10708 [49:17<1:17:20,  1.97it/s]
 15%|█▍        | 1559/10708 [49:18<1:17:09,  1.98it/s]
 15%|█▍        | 1560/10708 [49:18<1:16:36,  1.99it/s]
 15%|█▍        | 1561/10708 [49:19<1:16:18,  2.00it/s]
 15%|█▍        | 1562/10708 [49:19<1:16:22,  2.00it/s]
 15%|█▍        | 1563/10708 [49:20<1:26:47,  1.76it/s]
 15%|█▍        | 1564/10708 [49:21<1:23:35,  1.82it/s]
 15%|█▍        | 1565/10708 [49:21<1:21:13,  1.88it/s]
 15%|█▍        | 1566/10708 [49:22<1:19:24,  1.92it/s]
 15%|█▍        | 1567/10708 [49:22<1:18:07,  1.95it/s]
 15%|█▍        | 1568/10708 [49:23<1:17:17,  1.97it/s]
 15%|█▍        | 1569/10708 [49:23<1:16:52,  1.98it/s]
 15%|█▍        | 1570/10708 [49:24<1:17:11,  1.97it/s]
 15%|█▍        | 1571/10708 [49:24<1:16:37,  1.99it/s]
 15%|█▍        | 1572/10708 [49:25<1:16:12,  2.00it/s]
 15%|█▍        | 1573/10708 [49:25<1:15:58,  2.00it/s]
 15%|█▍        | 1574/10708 [49:26<1:15:43,  2.01it/s]
 15%|█▍        | 1575/10708 [49:26<1:15:49,  2.01it/s]
                                                      
{'loss': 4.1304, 'grad_norm': 0.2331741601228714, 'learning_rate': 0.000993266514619941, 'epoch': 0.15}

 15%|█▍        | 1575/10708 [49:26<1:15:49,  2.01it/s]
 15%|█▍        | 1576/10708 [49:27<1:15:32,  2.01it/s]
 15%|█▍        | 1577/10708 [49:27<1:15:36,  2.01it/s]
 15%|█▍        | 1578/10708 [49:28<1:15:51,  2.01it/s]
 15%|█▍        | 1579/10708 [49:28<1:16:05,  2.00it/s]
 15%|█▍        | 1580/10708 [49:29<1:15:55,  2.00it/s]
 15%|█▍        | 1581/10708 [49:29<1:15:52,  2.00it/s]
 15%|█▍        | 1582/10708 [49:30<1:15:58,  2.00it/s]
 15%|█▍        | 1583/10708 [49:30<1:16:12,  2.00it/s]
 15%|█▍        | 1584/10708 [49:31<1:15:59,  2.00it/s]
 15%|█▍        | 1585/10708 [49:31<1:15:47,  2.01it/s]
 15%|█▍        | 1586/10708 [49:32<1:15:54,  2.00it/s]
 15%|█▍        | 1587/10708 [49:32<1:15:48,  2.01it/s]
 15%|█▍        | 1588/10708 [49:33<1:15:42,  2.01it/s]
 15%|█▍        | 1589/10708 [49:33<1:15:52,  2.00it/s]
 15%|█▍        | 1590/10708 [49:34<1:16:02,  2.00it/s]
 15%|█▍        | 1591/10708 [49:34<1:15:37,  2.01it/s]
 15%|█▍        | 1592/10708 [49:35<1:15:33,  2.01it/s]
 15%|█▍        | 1593/10708 [49:35<1:15:42,  2.01it/s]
 15%|█▍        | 1594/10708 [49:36<1:15:51,  2.00it/s]
 15%|█▍        | 1595/10708 [49:36<1:15:31,  2.01it/s]
 15%|█▍        | 1596/10708 [49:37<1:15:27,  2.01it/s]
 15%|█▍        | 1597/10708 [49:37<1:15:19,  2.02it/s]
 15%|█▍        | 1598/10708 [49:38<1:25:31,  1.78it/s]
 15%|█▍        | 1599/10708 [49:38<1:22:18,  1.84it/s]
 15%|█▍        | 1600/10708 [49:39<1:20:46,  1.88it/s]{'loss': 4.117, 'grad_norm': 0.23349474370479584, 'learning_rate': 0.0009925836396991307, 'epoch': 0.15}
                                                      

 15%|█▍        | 1600/10708 [49:39<1:20:46,  1.88it/s]
 15%|█▍        | 1601/10708 [49:39<1:19:11,  1.92it/s]
 15%|█▍        | 1602/10708 [49:40<1:18:07,  1.94it/s]
 15%|█▍        | 1603/10708 [49:40<1:17:09,  1.97it/s]
 15%|█▍        | 1604/10708 [49:41<1:25:51,  1.77it/s]
 15%|█▍        | 1605/10708 [49:42<1:23:17,  1.82it/s]
 15%|█▍        | 1606/10708 [49:42<1:21:17,  1.87it/s]
 15%|█▌        | 1607/10708 [49:43<1:19:32,  1.91it/s]
 15%|█▌        | 1608/10708 [49:43<1:18:18,  1.94it/s]
 15%|█▌        | 1609/10708 [49:44<1:17:46,  1.95it/s]
 15%|█▌        | 1610/10708 [49:44<1:17:04,  1.97it/s]
 15%|█▌        | 1611/10708 [49:45<1:16:29,  1.98it/s]
 15%|█▌        | 1612/10708 [49:45<1:16:04,  1.99it/s]
 15%|█▌        | 1613/10708 [49:46<1:15:50,  2.00it/s]
 15%|█▌        | 1614/10708 [49:46<1:15:53,  2.00it/s]
 15%|█▌        | 1615/10708 [49:47<1:15:30,  2.01it/s]
 15%|█▌        | 1616/10708 [49:47<1:15:18,  2.01it/s]
 15%|█▌        | 1617/10708 [49:47<1:15:06,  2.02it/s]
 15%|█▌        | 1618/10708 [49:48<1:14:50,  2.02it/s]
 15%|█▌        | 1619/10708 [49:48<1:15:01,  2.02it/s]
 15%|█▌        | 1620/10708 [49:49<1:14:58,  2.02it/s]
 15%|█▌        | 1621/10708 [49:49<1:15:11,  2.01it/s]
 15%|█▌        | 1622/10708 [49:50<1:15:06,  2.02it/s]
 15%|█▌        | 1623/10708 [49:50<1:15:01,  2.02it/s]
 15%|█▌        | 1624/10708 [49:51<1:14:43,  2.03it/s]
 15%|█▌        | 1625/10708 [49:51<1:15:00,  2.02it/s]
                                                      {'loss': 4.0941, 'grad_norm': 0.24619771540164948, 'learning_rate': 0.000991868047766992, 'epoch': 0.15}


 15%|█▌        | 1625/10708 [49:51<1:15:00,  2.02it/s]
 15%|█▌        | 1626/10708 [49:52<1:15:04,  2.02it/s]
 15%|█▌        | 1627/10708 [49:52<1:14:58,  2.02it/s]
 15%|█▌        | 1628/10708 [49:53<1:16:19,  1.98it/s]
 15%|█▌        | 1629/10708 [49:53<1:16:25,  1.98it/s]
 15%|█▌        | 1630/10708 [49:54<1:16:09,  1.99it/s]
 15%|█▌        | 1631/10708 [49:54<1:15:34,  2.00it/s]
 15%|█▌        | 1632/10708 [49:55<1:25:07,  1.78it/s]
 15%|█▌        | 1633/10708 [49:56<1:22:18,  1.84it/s]
 15%|█▌        | 1634/10708 [49:56<1:20:21,  1.88it/s]
 15%|█▌        | 1635/10708 [49:57<1:18:35,  1.92it/s]
 15%|█▌        | 1636/10708 [49:57<1:17:39,  1.95it/s]
 15%|█▌        | 1637/10708 [49:58<1:17:19,  1.96it/s]
 15%|█▌        | 1638/10708 [49:58<1:16:36,  1.97it/s]
 15%|█▌        | 1639/10708 [49:59<1:16:04,  1.99it/s]
 15%|█▌        | 1640/10708 [49:59<1:15:39,  2.00it/s]
 15%|█▌        | 1641/10708 [50:00<1:15:26,  2.00it/s]
 15%|█▌        | 1642/10708 [50:00<1:24:54,  1.78it/s]
 15%|█▌        | 1643/10708 [50:01<1:22:11,  1.84it/s]
 15%|█▌        | 1644/10708 [50:01<1:19:52,  1.89it/s]
 15%|█▌        | 1645/10708 [50:02<1:18:29,  1.92it/s]
 15%|█▌        | 1646/10708 [50:02<1:17:44,  1.94it/s]
 15%|█▌        | 1647/10708 [50:03<1:16:56,  1.96it/s]
 15%|█▌        | 1648/10708 [50:03<1:16:03,  1.99it/s]
 15%|█▌        | 1649/10708 [50:04<1:15:32,  2.00it/s]
 15%|█▌        | 1650/10708 [50:04<1:15:36,  2.00it/s]
                                                      
{'loss': 4.0929, 'grad_norm': 0.2356419861316681, 'learning_rate': 0.0009911197863525687, 'epoch': 0.15}

 15%|█▌        | 1650/10708 [50:04<1:15:36,  2.00it/s]
 15%|█▌        | 1651/10708 [50:05<1:15:18,  2.00it/s]
 15%|█▌        | 1652/10708 [50:05<1:15:02,  2.01it/s]
 15%|█▌        | 1653/10708 [50:06<1:15:07,  2.01it/s]
 15%|█▌        | 1654/10708 [50:06<1:15:02,  2.01it/s]
 15%|█▌        | 1655/10708 [50:07<1:14:48,  2.02it/s]
 15%|█▌        | 1656/10708 [50:07<1:14:48,  2.02it/s]
 15%|█▌        | 1657/10708 [50:08<1:14:37,  2.02it/s]
 15%|█▌        | 1658/10708 [50:08<1:14:38,  2.02it/s]
 15%|█▌        | 1659/10708 [50:09<1:14:39,  2.02it/s]
 16%|█▌        | 1660/10708 [50:09<1:15:01,  2.01it/s]
 16%|█▌        | 1661/10708 [50:10<1:15:18,  2.00it/s]
 16%|█▌        | 1662/10708 [50:10<1:15:05,  2.01it/s]
 16%|█▌        | 1663/10708 [50:11<1:15:03,  2.01it/s]
 16%|█▌        | 1664/10708 [50:11<1:14:51,  2.01it/s]
 16%|█▌        | 1665/10708 [50:12<1:14:42,  2.02it/s]
 16%|█▌        | 1666/10708 [50:12<1:14:36,  2.02it/s]
 16%|█▌        | 1667/10708 [50:13<1:14:26,  2.02it/s]
 16%|█▌        | 1668/10708 [50:13<1:15:15,  2.00it/s]
 16%|█▌        | 1669/10708 [50:14<1:15:03,  2.01it/s]
 16%|█▌        | 1670/10708 [50:14<1:14:47,  2.01it/s]
 16%|█▌        | 1671/10708 [50:15<1:15:07,  2.00it/s]
 16%|█▌        | 1672/10708 [50:15<1:15:10,  2.00it/s]
 16%|█▌        | 1673/10708 [50:16<1:15:13,  2.00it/s]
 16%|█▌        | 1674/10708 [50:16<1:15:09,  2.00it/s]
 16%|█▌        | 1675/10708 [50:17<1:14:55,  2.01it/s]{'loss': 4.0871, 'grad_norm': 0.24520528316497803, 'learning_rate': 0.0009903389051547854, 'epoch': 0.16}

                                                      

 16%|█▌        | 1675/10708 [50:17<1:14:55,  2.01it/s]
 16%|█▌        | 1676/10708 [50:17<1:14:54,  2.01it/s]
 16%|█▌        | 1677/10708 [50:18<1:15:14,  2.00it/s]
 16%|█▌        | 1678/10708 [50:18<1:14:52,  2.01it/s]
 16%|█▌        | 1679/10708 [50:19<1:14:41,  2.01it/s]
 16%|█▌        | 1680/10708 [50:19<1:14:39,  2.02it/s]
 16%|█▌        | 1681/10708 [50:20<1:14:31,  2.02it/s]
 16%|█▌        | 1682/10708 [50:20<1:14:30,  2.02it/s]
 16%|█▌        | 1683/10708 [50:21<1:14:25,  2.02it/s]
 16%|█▌        | 1684/10708 [50:21<1:14:20,  2.02it/s]
 16%|█▌        | 1685/10708 [50:22<1:14:17,  2.02it/s]
 16%|█▌        | 1686/10708 [50:22<1:14:16,  2.02it/s]
 16%|█▌        | 1687/10708 [50:23<1:14:18,  2.02it/s]
 16%|█▌        | 1688/10708 [50:23<1:14:17,  2.02it/s]
 16%|█▌        | 1689/10708 [50:24<1:14:40,  2.01it/s]
 16%|█▌        | 1690/10708 [50:24<1:15:14,  2.00it/s]
 16%|█▌        | 1691/10708 [50:25<1:15:01,  2.00it/s]
 16%|█▌        | 1692/10708 [50:25<1:14:43,  2.01it/s]
 16%|█▌        | 1693/10708 [50:26<1:14:40,  2.01it/s]
 16%|█▌        | 1694/10708 [50:26<1:14:53,  2.01it/s]
 16%|█▌        | 1695/10708 [50:27<1:14:40,  2.01it/s]
 16%|█▌        | 1696/10708 [50:27<1:14:32,  2.02it/s]
 16%|█▌        | 1697/10708 [50:28<1:14:30,  2.02it/s]
 16%|█▌        | 1698/10708 [50:28<1:14:45,  2.01it/s]
 16%|█▌        | 1699/10708 [50:29<1:14:32,  2.01it/s]
 16%|█▌        | 1700/10708 [50:29<1:15:12,  2.00it/s]
                                                      
{'loss': 4.0869, 'grad_norm': 0.24152620136737823, 'learning_rate': 0.0009895254560391468, 'epoch': 0.16}

 16%|█▌        | 1700/10708 [50:29<1:15:12,  2.00it/s]
 16%|█▌        | 1701/10708 [50:30<1:14:59,  2.00it/s]
 16%|█▌        | 1702/10708 [50:30<1:15:21,  1.99it/s]
 16%|█▌        | 1703/10708 [50:31<1:15:15,  1.99it/s]
 16%|█▌        | 1704/10708 [50:31<1:15:01,  2.00it/s]
 16%|█▌        | 1705/10708 [50:32<1:14:42,  2.01it/s]
 16%|█▌        | 1706/10708 [50:32<1:14:45,  2.01it/s]
 16%|█▌        | 1707/10708 [50:33<1:14:34,  2.01it/s]
 16%|█▌        | 1708/10708 [50:33<1:14:32,  2.01it/s]
 16%|█▌        | 1709/10708 [50:34<1:14:20,  2.02it/s]
 16%|█▌        | 1710/10708 [50:34<1:14:37,  2.01it/s]
 16%|█▌        | 1711/10708 [50:35<1:14:38,  2.01it/s]
 16%|█▌        | 1712/10708 [50:35<1:14:20,  2.02it/s]
 16%|█▌        | 1713/10708 [50:36<1:14:23,  2.02it/s]
 16%|█▌        | 1714/10708 [50:36<1:14:33,  2.01it/s]
 16%|█▌        | 1715/10708 [50:37<1:14:45,  2.00it/s]
 16%|█▌        | 1716/10708 [50:37<1:14:44,  2.01it/s]
 16%|█▌        | 1717/10708 [50:38<1:14:27,  2.01it/s]
 16%|█▌        | 1718/10708 [50:38<1:14:35,  2.01it/s]
 16%|█▌        | 1719/10708 [50:39<1:14:28,  2.01it/s]
 16%|█▌        | 1720/10708 [50:39<1:14:21,  2.01it/s]
 16%|█▌        | 1721/10708 [50:40<1:14:10,  2.02it/s]
 16%|█▌        | 1722/10708 [50:40<1:14:13,  2.02it/s]
 16%|█▌        | 1723/10708 [50:41<1:14:04,  2.02it/s]
 16%|█▌        | 1724/10708 [50:41<1:15:21,  1.99it/s]
 16%|█▌        | 1725/10708 [50:42<1:15:16,  1.99it/s]
                                                      
{'loss': 4.0746, 'grad_norm': 0.22092857956886292, 'learning_rate': 0.0009886794930342921, 'epoch': 0.16}

 16%|█▌        | 1725/10708 [50:42<1:15:16,  1.99it/s]
 16%|█▌        | 1726/10708 [50:42<1:15:06,  1.99it/s]
 16%|█▌        | 1727/10708 [50:43<1:15:14,  1.99it/s]
 16%|█▌        | 1728/10708 [50:43<1:15:27,  1.98it/s]
 16%|█▌        | 1729/10708 [50:44<1:14:54,  2.00it/s]
 16%|█▌        | 1730/10708 [50:44<1:14:39,  2.00it/s]
 16%|█▌        | 1731/10708 [50:45<1:14:21,  2.01it/s]
 16%|█▌        | 1732/10708 [50:45<1:14:12,  2.02it/s]
 16%|█▌        | 1733/10708 [50:46<1:14:05,  2.02it/s]
 16%|█▌        | 1734/10708 [50:46<1:13:57,  2.02it/s]
 16%|█▌        | 1735/10708 [50:47<1:14:03,  2.02it/s]
 16%|█▌        | 1736/10708 [50:47<1:14:46,  2.00it/s]
 16%|█▌        | 1737/10708 [50:48<1:14:32,  2.01it/s]
 16%|█▌        | 1738/10708 [50:48<1:26:45,  1.72it/s]
 16%|█▌        | 1739/10708 [50:49<1:23:06,  1.80it/s]
 16%|█▌        | 1740/10708 [50:49<1:20:17,  1.86it/s]
 16%|█▋        | 1741/10708 [50:50<1:18:28,  1.90it/s]
 16%|█▋        | 1742/10708 [50:50<1:17:06,  1.94it/s]
 16%|█▋        | 1743/10708 [50:51<1:16:03,  1.96it/s]
 16%|█▋        | 1744/10708 [50:51<1:15:33,  1.98it/s]
 16%|█▋        | 1745/10708 [50:52<1:27:14,  1.71it/s]
 16%|█▋        | 1746/10708 [50:53<1:23:16,  1.79it/s]
 16%|█▋        | 1747/10708 [50:53<1:20:18,  1.86it/s]
 16%|█▋        | 1748/10708 [50:54<1:18:21,  1.91it/s]
 16%|█▋        | 1749/10708 [50:54<1:17:10,  1.93it/s]
 16%|█▋        | 1750/10708 [50:55<1:16:13,  1.96it/s]
                                                      
{'loss': 4.0607, 'grad_norm': 0.24467609822750092, 'learning_rate': 0.0009878010723284078, 'epoch': 0.16}

 16%|█▋        | 1750/10708 [50:55<1:16:13,  1.96it/s]
 16%|█▋        | 1751/10708 [50:55<1:16:15,  1.96it/s]
 16%|█▋        | 1752/10708 [50:56<1:15:33,  1.98it/s]
 16%|█▋        | 1753/10708 [50:56<1:15:07,  1.99it/s]
 16%|█▋        | 1754/10708 [50:57<1:14:42,  2.00it/s]
 16%|█▋        | 1755/10708 [50:57<1:14:34,  2.00it/s]
 16%|█▋        | 1756/10708 [50:58<1:14:22,  2.01it/s]
 16%|█▋        | 1757/10708 [50:58<1:14:04,  2.01it/s]
 16%|█▋        | 1758/10708 [50:59<1:14:02,  2.01it/s]
 16%|█▋        | 1759/10708 [50:59<1:13:49,  2.02it/s]
 16%|█▋        | 1760/10708 [51:00<1:13:42,  2.02it/s]
 16%|█▋        | 1761/10708 [51:00<1:13:39,  2.02it/s]
 16%|█▋        | 1762/10708 [51:01<1:13:52,  2.02it/s]
 16%|█▋        | 1763/10708 [51:01<1:13:41,  2.02it/s]
 16%|█▋        | 1764/10708 [51:02<1:14:09,  2.01it/s]
 16%|█▋        | 1765/10708 [51:02<1:14:00,  2.01it/s]
 16%|█▋        | 1766/10708 [51:03<1:14:05,  2.01it/s]
 17%|█▋        | 1767/10708 [51:03<1:14:06,  2.01it/s]
 17%|█▋        | 1768/10708 [51:04<1:14:11,  2.01it/s]
 17%|█▋        | 1769/10708 [51:04<1:14:31,  2.00it/s]
 17%|█▋        | 1770/10708 [51:05<1:14:08,  2.01it/s]
 17%|█▋        | 1771/10708 [51:05<1:14:03,  2.01it/s]
 17%|█▋        | 1772/10708 [51:06<1:14:28,  2.00it/s]
 17%|█▋        | 1773/10708 [51:06<1:14:16,  2.00it/s]
 17%|█▋        | 1774/10708 [51:07<1:14:04,  2.01it/s]
 17%|█▋        | 1775/10708 [51:07<1:13:49,  2.02it/s]
                                                      
{'loss': 4.0651, 'grad_norm': 0.24454286694526672, 'learning_rate': 0.000986890252265494, 'epoch': 0.17}

 17%|█▋        | 1775/10708 [51:07<1:13:49,  2.02it/s]
 17%|█▋        | 1776/10708 [51:08<1:14:23,  2.00it/s]
 17%|█▋        | 1777/10708 [51:08<1:14:10,  2.01it/s]
 17%|█▋        | 1778/10708 [51:09<1:14:12,  2.01it/s]
 17%|█▋        | 1779/10708 [51:09<1:13:59,  2.01it/s]
 17%|█▋        | 1780/10708 [51:10<1:14:05,  2.01it/s]
 17%|█▋        | 1781/10708 [51:10<1:13:47,  2.02it/s]
 17%|█▋        | 1782/10708 [51:11<1:13:43,  2.02it/s]
 17%|█▋        | 1783/10708 [51:11<1:13:38,  2.02it/s]
 17%|█▋        | 1784/10708 [51:12<1:13:39,  2.02it/s]
 17%|█▋        | 1785/10708 [51:12<1:13:42,  2.02it/s]
 17%|█▋        | 1786/10708 [51:13<1:13:55,  2.01it/s]
 17%|█▋        | 1787/10708 [51:13<1:13:41,  2.02it/s]
 17%|█▋        | 1788/10708 [51:13<1:13:42,  2.02it/s]
 17%|█▋        | 1789/10708 [51:14<1:13:32,  2.02it/s]
 17%|█▋        | 1790/10708 [51:14<1:13:31,  2.02it/s]
 17%|█▋        | 1791/10708 [51:15<1:13:29,  2.02it/s]
 17%|█▋        | 1792/10708 [51:15<1:13:41,  2.02it/s]
 17%|█▋        | 1793/10708 [51:16<1:13:52,  2.01it/s]
 17%|█▋        | 1794/10708 [51:16<1:13:48,  2.01it/s]
 17%|█▋        | 1795/10708 [51:17<1:13:38,  2.02it/s]
 17%|█▋        | 1796/10708 [51:17<1:13:50,  2.01it/s]
 17%|█▋        | 1797/10708 [51:18<1:13:44,  2.01it/s]
 17%|█▋        | 1798/10708 [51:18<1:13:36,  2.02it/s]
 17%|█▋        | 1799/10708 [51:19<1:13:34,  2.02it/s]
 17%|█▋        | 1800/10708 [51:19<1:13:44,  2.01it/s]{'loss': 4.051, 'grad_norm': 0.22839736938476562, 'learning_rate': 0.0009859470933414913, 'epoch': 0.17}

                                                      

 17%|█▋        | 1800/10708 [51:19<1:13:44,  2.01it/s]
 17%|█▋        | 1801/10708 [51:20<1:13:40,  2.01it/s]
 17%|█▋        | 1802/10708 [51:20<1:13:43,  2.01it/s]
 17%|█▋        | 1803/10708 [51:21<1:13:40,  2.01it/s]
 17%|█▋        | 1804/10708 [51:21<1:13:38,  2.02it/s]
 17%|█▋        | 1805/10708 [51:22<1:13:43,  2.01it/s]
 17%|█▋        | 1806/10708 [51:22<1:13:35,  2.02it/s]
 17%|█▋        | 1807/10708 [51:23<1:13:54,  2.01it/s]
 17%|█▋        | 1808/10708 [51:23<1:13:33,  2.02it/s]
 17%|█▋        | 1809/10708 [51:24<1:13:33,  2.02it/s]
 17%|█▋        | 1810/10708 [51:24<1:13:29,  2.02it/s]
 17%|█▋        | 1811/10708 [51:25<1:13:26,  2.02it/s]
 17%|█▋        | 1812/10708 [51:25<1:13:18,  2.02it/s]
 17%|█▋        | 1813/10708 [51:26<1:13:24,  2.02it/s]
 17%|█▋        | 1814/10708 [51:26<1:13:16,  2.02it/s]
 17%|█▋        | 1815/10708 [51:27<1:13:24,  2.02it/s]
 17%|█▋        | 1816/10708 [51:27<1:13:25,  2.02it/s]
 17%|█▋        | 1817/10708 [51:28<1:13:18,  2.02it/s]
 17%|█▋        | 1818/10708 [51:28<1:13:42,  2.01it/s]
 17%|█▋        | 1819/10708 [51:29<1:13:25,  2.02it/s]
 17%|█▋        | 1820/10708 [51:29<1:13:23,  2.02it/s]
 17%|█▋        | 1821/10708 [51:30<1:13:20,  2.02it/s]
 17%|█▋        | 1822/10708 [51:30<1:13:17,  2.02it/s]
 17%|█▋        | 1823/10708 [51:31<1:13:21,  2.02it/s]
 17%|█▋        | 1824/10708 [51:31<1:13:38,  2.01it/s]
 17%|█▋        | 1825/10708 [51:32<1:13:23,  2.02it/s]
                                                      
{'loss': 4.0429, 'grad_norm': 0.22138071060180664, 'learning_rate': 0.0009849716582002606, 'epoch': 0.17}

 17%|█▋        | 1825/10708 [51:32<1:13:23,  2.02it/s]
 17%|█▋        | 1826/10708 [51:32<1:13:36,  2.01it/s]
 17%|█▋        | 1827/10708 [51:33<1:13:30,  2.01it/s]
 17%|█▋        | 1828/10708 [51:33<1:13:28,  2.01it/s]
 17%|█▋        | 1829/10708 [51:34<1:13:28,  2.01it/s]
 17%|█▋        | 1830/10708 [51:34<1:13:17,  2.02it/s]
 17%|█▋        | 1831/10708 [51:35<1:13:21,  2.02it/s]
 17%|█▋        | 1832/10708 [51:35<1:13:13,  2.02it/s]
 17%|█▋        | 1833/10708 [51:36<1:13:27,  2.01it/s]
 17%|█▋        | 1834/10708 [51:36<1:13:26,  2.01it/s]
 17%|█▋        | 1835/10708 [51:37<1:13:16,  2.02it/s]
 17%|█▋        | 1836/10708 [51:37<1:13:12,  2.02it/s]
 17%|█▋        | 1837/10708 [51:38<1:13:09,  2.02it/s]
 17%|█▋        | 1838/10708 [51:38<1:13:12,  2.02it/s]
 17%|█▋        | 1839/10708 [51:39<1:13:16,  2.02it/s]
 17%|█▋        | 1840/10708 [51:39<1:13:14,  2.02it/s]
 17%|█▋        | 1841/10708 [51:40<1:13:05,  2.02it/s]
 17%|█▋        | 1842/10708 [51:40<1:13:26,  2.01it/s]
 17%|█▋        | 1843/10708 [51:41<1:13:20,  2.01it/s]
 17%|█▋        | 1844/10708 [51:41<1:13:16,  2.02it/s]
 17%|█▋        | 1845/10708 [51:42<1:13:32,  2.01it/s]
 17%|█▋        | 1846/10708 [51:42<1:23:19,  1.77it/s]
 17%|█▋        | 1847/10708 [51:43<1:20:09,  1.84it/s]
 17%|█▋        | 1848/10708 [51:43<1:17:49,  1.90it/s]
 17%|█▋        | 1849/10708 [51:44<1:16:39,  1.93it/s]
 17%|█▋        | 1850/10708 [51:44<1:15:36,  1.95it/s]{'loss': 4.0442, 'grad_norm': 0.2243327647447586, 'learning_rate': 0.0009839640116294241, 'epoch': 0.17}
                                                      

 17%|█▋        | 1850/10708 [51:44<1:15:36,  1.95it/s]
 17%|█▋        | 1851/10708 [51:45<1:14:58,  1.97it/s]
 17%|█▋        | 1852/10708 [51:45<1:14:28,  1.98it/s]
 17%|█▋        | 1853/10708 [51:46<1:14:09,  1.99it/s]
 17%|█▋        | 1854/10708 [51:46<1:14:01,  1.99it/s]
 17%|█▋        | 1855/10708 [51:47<1:13:45,  2.00it/s]
 17%|█▋        | 1856/10708 [51:47<1:13:39,  2.00it/s]
 17%|█▋        | 1857/10708 [51:48<1:13:27,  2.01it/s]
 17%|█▋        | 1858/10708 [51:48<1:13:13,  2.01it/s]
 17%|█▋        | 1859/10708 [51:49<1:13:22,  2.01it/s]
 17%|█▋        | 1860/10708 [51:49<1:13:28,  2.01it/s]
 17%|█▋        | 1861/10708 [51:50<1:13:38,  2.00it/s]
 17%|█▋        | 1862/10708 [51:50<1:13:23,  2.01it/s]
 17%|█▋        | 1863/10708 [51:51<1:13:19,  2.01it/s]
 17%|█▋        | 1864/10708 [51:51<1:13:15,  2.01it/s]
 17%|█▋        | 1865/10708 [51:52<1:13:01,  2.02it/s]
 17%|█▋        | 1866/10708 [51:52<1:13:20,  2.01it/s]
 17%|█▋        | 1867/10708 [51:53<1:13:15,  2.01it/s]
 17%|█▋        | 1868/10708 [51:53<1:12:56,  2.02it/s]
 17%|█▋        | 1869/10708 [51:54<1:12:55,  2.02it/s]
 17%|█▋        | 1870/10708 [51:54<1:12:53,  2.02it/s]
 17%|█▋        | 1871/10708 [51:55<1:12:53,  2.02it/s]
 17%|█▋        | 1872/10708 [51:55<1:12:47,  2.02it/s]
 17%|█▋        | 1873/10708 [51:56<1:12:42,  2.03it/s]
 18%|█▊        | 1874/10708 [51:56<1:12:58,  2.02it/s]
 18%|█▊        | 1875/10708 [51:57<1:12:58,  2.02it/s]
                                                      
{'loss': 4.0255, 'grad_norm': 0.23586711287498474, 'learning_rate': 0.0009829242205560606, 'epoch': 0.18}

 18%|█▊        | 1875/10708 [51:57<1:12:58,  2.02it/s]
 18%|█▊        | 1876/10708 [51:57<1:13:17,  2.01it/s]
 18%|█▊        | 1877/10708 [51:58<1:13:11,  2.01it/s]
 18%|█▊        | 1878/10708 [51:59<1:22:41,  1.78it/s]
 18%|█▊        | 1879/10708 [51:59<1:19:46,  1.84it/s]
 18%|█▊        | 1880/10708 [52:00<1:17:28,  1.90it/s]
 18%|█▊        | 1881/10708 [52:00<1:15:57,  1.94it/s]
 18%|█▊        | 1882/10708 [52:01<1:14:58,  1.96it/s]
 18%|█▊        | 1883/10708 [52:01<1:14:16,  1.98it/s]
 18%|█▊        | 1884/10708 [52:02<1:13:47,  1.99it/s]
 18%|█▊        | 1885/10708 [52:02<1:13:30,  2.00it/s]
 18%|█▊        | 1886/10708 [52:03<1:13:26,  2.00it/s]
 18%|█▊        | 1887/10708 [52:03<1:13:28,  2.00it/s]
 18%|█▊        | 1888/10708 [52:04<1:13:44,  1.99it/s]
 18%|█▊        | 1889/10708 [52:04<1:13:20,  2.00it/s]
 18%|█▊        | 1890/10708 [52:05<1:13:02,  2.01it/s]
 18%|█▊        | 1891/10708 [52:05<1:12:54,  2.02it/s]
 18%|█▊        | 1892/10708 [52:06<1:12:52,  2.02it/s]
 18%|█▊        | 1893/10708 [52:06<1:12:57,  2.01it/s]
 18%|█▊        | 1894/10708 [52:07<1:22:15,  1.79it/s]
 18%|█▊        | 1895/10708 [52:07<1:19:44,  1.84it/s]
 18%|█▊        | 1896/10708 [52:08<1:17:41,  1.89it/s]
 18%|█▊        | 1897/10708 [52:08<1:16:12,  1.93it/s]
 18%|█▊        | 1898/10708 [52:09<1:15:24,  1.95it/s]
 18%|█▊        | 1899/10708 [52:09<1:14:39,  1.97it/s]
 18%|█▊        | 1900/10708 [52:10<1:14:00,  1.98it/s]
                                                      
{'loss': 4.0334, 'grad_norm': 0.22638219594955444, 'learning_rate': 0.0009818523540422617, 'epoch': 0.18}

 18%|█▊        | 1900/10708 [52:10<1:14:00,  1.98it/s]
 18%|█▊        | 1901/10708 [52:10<1:13:37,  1.99it/s]
 18%|█▊        | 1902/10708 [52:11<1:13:13,  2.00it/s]
 18%|█▊        | 1903/10708 [52:11<1:13:09,  2.01it/s]
 18%|█▊        | 1904/10708 [52:12<1:13:29,  2.00it/s]
 18%|█▊        | 1905/10708 [52:12<1:13:19,  2.00it/s]
 18%|█▊        | 1906/10708 [52:13<1:13:17,  2.00it/s]
 18%|█▊        | 1907/10708 [52:13<1:13:17,  2.00it/s]
 18%|█▊        | 1908/10708 [52:14<1:13:02,  2.01it/s]
 18%|█▊        | 1909/10708 [52:14<1:12:50,  2.01it/s]
 18%|█▊        | 1910/10708 [52:15<1:12:47,  2.01it/s]
 18%|█▊        | 1911/10708 [52:15<1:12:35,  2.02it/s]
 18%|█▊        | 1912/10708 [52:16<1:12:30,  2.02it/s]
 18%|█▊        | 1913/10708 [52:16<1:12:19,  2.03it/s]
 18%|█▊        | 1914/10708 [52:17<1:12:26,  2.02it/s]
 18%|█▊        | 1915/10708 [52:17<1:12:20,  2.03it/s]
 18%|█▊        | 1916/10708 [52:18<1:12:23,  2.02it/s]
 18%|█▊        | 1917/10708 [52:18<1:12:29,  2.02it/s]
 18%|█▊        | 1918/10708 [52:19<1:12:24,  2.02it/s]
 18%|█▊        | 1919/10708 [52:19<1:12:23,  2.02it/s]
 18%|█▊        | 1920/10708 [52:20<1:12:39,  2.02it/s]
 18%|█▊        | 1921/10708 [52:20<1:12:27,  2.02it/s]
 18%|█▊        | 1922/10708 [52:21<1:12:47,  2.01it/s]
 18%|█▊        | 1923/10708 [52:21<1:12:52,  2.01it/s]
 18%|█▊        | 1924/10708 [52:22<1:12:49,  2.01it/s]
 18%|█▊        | 1925/10708 [52:22<1:12:45,  2.01it/s]
                                                      
{'loss': 4.0235, 'grad_norm': 0.21523308753967285, 'learning_rate': 0.0009807484832805439, 'epoch': 0.18}

 18%|█▊        | 1925/10708 [52:22<1:12:45,  2.01it/s]
 18%|█▊        | 1926/10708 [52:23<1:12:44,  2.01it/s]
 18%|█▊        | 1927/10708 [52:23<1:12:56,  2.01it/s]
 18%|█▊        | 1928/10708 [52:24<1:12:46,  2.01it/s]
 18%|█▊        | 1929/10708 [52:24<1:12:36,  2.01it/s]
 18%|█▊        | 1930/10708 [52:25<1:12:31,  2.02it/s]
 18%|█▊        | 1931/10708 [52:25<1:12:19,  2.02it/s]
 18%|█▊        | 1932/10708 [52:26<1:12:23,  2.02it/s]
 18%|█▊        | 1933/10708 [52:26<1:12:21,  2.02it/s]
 18%|█▊        | 1934/10708 [52:27<1:12:24,  2.02it/s]
 18%|█▊        | 1935/10708 [52:27<1:12:27,  2.02it/s]
 18%|█▊        | 1936/10708 [52:28<1:12:50,  2.01it/s]
 18%|█▊        | 1937/10708 [52:28<1:12:48,  2.01it/s]
 18%|█▊        | 1938/10708 [52:29<1:12:43,  2.01it/s]
 18%|█▊        | 1939/10708 [52:29<1:12:38,  2.01it/s]
 18%|█▊        | 1940/10708 [52:30<1:12:47,  2.01it/s]
 18%|█▊        | 1941/10708 [52:30<1:12:36,  2.01it/s]
 18%|█▊        | 1942/10708 [52:31<1:12:35,  2.01it/s]
 18%|█▊        | 1943/10708 [52:31<1:12:25,  2.02it/s]
 18%|█▊        | 1944/10708 [52:32<1:12:15,  2.02it/s]
 18%|█▊        | 1945/10708 [52:32<1:12:23,  2.02it/s]
 18%|█▊        | 1946/10708 [52:33<1:12:23,  2.02it/s]
 18%|█▊        | 1947/10708 [52:33<1:22:10,  1.78it/s]
 18%|█▊        | 1948/10708 [52:34<1:19:19,  1.84it/s]
 18%|█▊        | 1949/10708 [52:34<1:17:06,  1.89it/s]
 18%|█▊        | 1950/10708 [52:35<1:15:37,  1.93it/s]
                                                      
{'loss': 4.0175, 'grad_norm': 0.2215900719165802, 'learning_rate': 0.00097961268158912, 'epoch': 0.18}

 18%|█▊        | 1950/10708 [52:35<1:15:37,  1.93it/s]
 18%|█▊        | 1951/10708 [52:35<1:14:31,  1.96it/s]
 18%|█▊        | 1952/10708 [52:36<1:13:54,  1.97it/s]
 18%|█▊        | 1953/10708 [52:36<1:13:22,  1.99it/s]
 18%|█▊        | 1954/10708 [52:37<1:12:57,  2.00it/s]
 18%|█▊        | 1955/10708 [52:37<1:13:06,  2.00it/s]
 18%|█▊        | 1956/10708 [52:38<1:13:11,  1.99it/s]
 18%|█▊        | 1957/10708 [52:38<1:12:46,  2.00it/s]
 18%|█▊        | 1958/10708 [52:39<1:12:31,  2.01it/s]
 18%|█▊        | 1959/10708 [52:39<1:12:22,  2.01it/s]
 18%|█▊        | 1960/10708 [52:40<1:12:10,  2.02it/s]
 18%|█▊        | 1961/10708 [52:40<1:12:18,  2.02it/s]
 18%|█▊        | 1962/10708 [52:41<1:12:07,  2.02it/s]
 18%|█▊        | 1963/10708 [52:41<1:12:01,  2.02it/s]
 18%|█▊        | 1964/10708 [52:42<1:11:59,  2.02it/s]
 18%|█▊        | 1965/10708 [52:42<1:12:09,  2.02it/s]
 18%|█▊        | 1966/10708 [52:43<1:12:08,  2.02it/s]
 18%|█▊        | 1967/10708 [52:43<1:12:13,  2.02it/s]
 18%|█▊        | 1968/10708 [52:44<1:12:14,  2.02it/s]
 18%|█▊        | 1969/10708 [52:44<1:12:09,  2.02it/s]
 18%|█▊        | 1970/10708 [52:45<1:12:14,  2.02it/s]
 18%|█▊        | 1971/10708 [52:45<1:12:04,  2.02it/s]
 18%|█▊        | 1972/10708 [52:46<1:12:21,  2.01it/s]
 18%|█▊        | 1973/10708 [52:46<1:12:18,  2.01it/s]
 18%|█▊        | 1974/10708 [52:47<1:12:13,  2.02it/s]
 18%|█▊        | 1975/10708 [52:47<1:12:19,  2.01it/s]
                                                      
{'loss': 4.0149, 'grad_norm': 0.23637443780899048, 'learning_rate': 0.0009784450244070297, 'epoch': 0.18}

 18%|█▊        | 1975/10708 [52:47<1:12:19,  2.01it/s]
 18%|█▊        | 1976/10708 [52:48<1:12:17,  2.01it/s]
 18%|█▊        | 1977/10708 [52:48<1:12:19,  2.01it/s]
 18%|█▊        | 1978/10708 [52:49<1:12:23,  2.01it/s]
 18%|█▊        | 1979/10708 [52:49<1:13:10,  1.99it/s]
 18%|█▊        | 1980/10708 [52:50<1:13:07,  1.99it/s]
 19%|█▊        | 1981/10708 [52:50<1:12:43,  2.00it/s]
 19%|█▊        | 1982/10708 [52:51<1:12:30,  2.01it/s]
 19%|█▊        | 1983/10708 [52:51<1:12:18,  2.01it/s]
 19%|█▊        | 1984/10708 [52:52<1:12:12,  2.01it/s]
 19%|█▊        | 1985/10708 [52:52<1:12:12,  2.01it/s]
 19%|█▊        | 1986/10708 [52:53<1:12:08,  2.02it/s]
 19%|█▊        | 1987/10708 [52:53<1:11:56,  2.02it/s]
 19%|█▊        | 1988/10708 [52:54<1:12:07,  2.01it/s]
 19%|█▊        | 1989/10708 [52:54<1:11:57,  2.02it/s]
 19%|█▊        | 1990/10708 [52:55<1:11:59,  2.02it/s]
 19%|█▊        | 1991/10708 [52:55<1:11:52,  2.02it/s]
 19%|█▊        | 1992/10708 [52:56<1:11:48,  2.02it/s]
 19%|█▊        | 1993/10708 [52:56<1:12:09,  2.01it/s]
 19%|█▊        | 1994/10708 [52:57<1:12:08,  2.01it/s]
 19%|█▊        | 1995/10708 [52:57<1:12:12,  2.01it/s]
 19%|█▊        | 1996/10708 [52:58<1:12:09,  2.01it/s]
 19%|█▊        | 1997/10708 [52:58<1:12:06,  2.01it/s]
 19%|█▊        | 1998/10708 [52:59<1:11:57,  2.02it/s]
 19%|█▊        | 1999/10708 [52:59<1:11:57,  2.02it/s]
 19%|█▊        | 2000/10708 [53:00<1:11:49,  2.02it/s]{'loss': 4.0127, 'grad_norm': 0.22611330449581146, 'learning_rate': 0.0009772455892891295, 'epoch': 0.19}

                                                      

 19%|█▊        | 2000/10708 [53:00<1:11:49,  2.02it/s]
 19%|█▊        | 2001/10708 [53:00<1:12:00,  2.02it/s]
 19%|█▊        | 2002/10708 [53:01<1:11:50,  2.02it/s]
 19%|█▊        | 2003/10708 [53:01<1:11:49,  2.02it/s]
 19%|█▊        | 2004/10708 [53:02<1:11:38,  2.03it/s]
 19%|█▊        | 2005/10708 [53:02<1:20:54,  1.79it/s]
 19%|█▊        | 2006/10708 [53:03<1:18:12,  1.85it/s]
 19%|█▊        | 2007/10708 [53:03<1:16:11,  1.90it/s]
 19%|█▉        | 2008/10708 [53:04<1:14:44,  1.94it/s]
 19%|█▉        | 2009/10708 [53:04<1:13:43,  1.97it/s]
 19%|█▉        | 2010/10708 [53:05<1:12:56,  1.99it/s]
 19%|█▉        | 2011/10708 [53:05<1:12:30,  2.00it/s]
 19%|█▉        | 2012/10708 [53:06<1:12:22,  2.00it/s]
 19%|█▉        | 2013/10708 [53:06<1:12:01,  2.01it/s]
 19%|█▉        | 2014/10708 [53:07<1:11:57,  2.01it/s]
 19%|█▉        | 2015/10708 [53:07<1:11:40,  2.02it/s]
 19%|█▉        | 2016/10708 [53:08<1:11:33,  2.02it/s]
 19%|█▉        | 2017/10708 [53:08<1:11:30,  2.03it/s]
 19%|█▉        | 2018/10708 [53:09<1:11:22,  2.03it/s]
 19%|█▉        | 2019/10708 [53:09<1:11:24,  2.03it/s]
 19%|█▉        | 2020/10708 [53:10<1:11:19,  2.03it/s]
 19%|█▉        | 2021/10708 [53:10<1:11:27,  2.03it/s]
 19%|█▉        | 2022/10708 [53:11<1:11:25,  2.03it/s]
 19%|█▉        | 2023/10708 [53:11<1:11:27,  2.03it/s]
 19%|█▉        | 2024/10708 [53:12<1:11:29,  2.02it/s]
 19%|█▉        | 2025/10708 [53:12<1:11:16,  2.03it/s]{'loss': 4.0009, 'grad_norm': 0.2276071310043335, 'learning_rate': 0.00097601445590094, 'epoch': 0.19}

                                                      

 19%|█▉        | 2025/10708 [53:12<1:11:16,  2.03it/s]
 19%|█▉        | 2026/10708 [53:13<1:11:53,  2.01it/s]
 19%|█▉        | 2027/10708 [53:13<1:11:44,  2.02it/s]
 19%|█▉        | 2028/10708 [53:14<1:11:36,  2.02it/s]
 19%|█▉        | 2029/10708 [53:14<1:11:50,  2.01it/s]
 19%|█▉        | 2030/10708 [53:15<1:11:41,  2.02it/s]
 19%|█▉        | 2031/10708 [53:15<1:11:38,  2.02it/s]
 19%|█▉        | 2032/10708 [53:16<1:11:30,  2.02it/s]
 19%|█▉        | 2033/10708 [53:16<1:11:29,  2.02it/s]
 19%|█▉        | 2034/10708 [53:17<1:11:23,  2.02it/s]
 19%|█▉        | 2035/10708 [53:17<1:11:24,  2.02it/s]
 19%|█▉        | 2036/10708 [53:18<1:11:17,  2.03it/s]
 19%|█▉        | 2037/10708 [53:18<1:11:10,  2.03it/s]
 19%|█▉        | 2038/10708 [53:19<1:11:18,  2.03it/s]
 19%|█▉        | 2039/10708 [53:19<1:11:18,  2.03it/s]
 19%|█▉        | 2040/10708 [53:20<1:11:23,  2.02it/s]
 19%|█▉        | 2041/10708 [53:20<1:11:22,  2.02it/s]
 19%|█▉        | 2042/10708 [53:21<1:11:29,  2.02it/s]
 19%|█▉        | 2043/10708 [53:21<1:11:21,  2.02it/s]
 19%|█▉        | 2044/10708 [53:22<1:11:21,  2.02it/s]
 19%|█▉        | 2045/10708 [53:22<1:11:19,  2.02it/s]
 19%|█▉        | 2046/10708 [53:22<1:11:39,  2.01it/s]
 19%|█▉        | 2047/10708 [53:23<1:11:43,  2.01it/s]
 19%|█▉        | 2048/10708 [53:23<1:11:31,  2.02it/s]
 19%|█▉        | 2049/10708 [53:24<1:11:29,  2.02it/s]
 19%|█▉        | 2050/10708 [53:24<1:11:23,  2.02it/s]
                                                      
{'loss': 3.9911, 'grad_norm': 0.22013770043849945, 'learning_rate': 0.0009747517060133565, 'epoch': 0.19}

 19%|█▉        | 2050/10708 [53:24<1:11:23,  2.02it/s]
 19%|█▉        | 2051/10708 [53:25<1:11:34,  2.02it/s]
 19%|█▉        | 2052/10708 [53:25<1:11:32,  2.02it/s]
 19%|█▉        | 2053/10708 [53:26<1:11:23,  2.02it/s]
 19%|█▉        | 2054/10708 [53:26<1:11:10,  2.03it/s]
 19%|█▉        | 2055/10708 [53:27<1:11:13,  2.02it/s]
 19%|█▉        | 2056/10708 [53:28<1:29:38,  1.61it/s]
 19%|█▉        | 2057/10708 [53:28<1:24:05,  1.71it/s]
 19%|█▉        | 2058/10708 [53:29<1:20:10,  1.80it/s]
 19%|█▉        | 2059/10708 [53:29<1:17:36,  1.86it/s]
 19%|█▉        | 2060/10708 [53:30<1:15:40,  1.90it/s]
 19%|█▉        | 2061/10708 [53:30<1:14:52,  1.92it/s]
 19%|█▉        | 2062/10708 [53:31<1:13:45,  1.95it/s]
 19%|█▉        | 2063/10708 [53:31<1:13:07,  1.97it/s]
 19%|█▉        | 2064/10708 [53:32<1:12:38,  1.98it/s]
 19%|█▉        | 2065/10708 [53:32<1:12:33,  1.99it/s]
 19%|█▉        | 2066/10708 [53:33<1:12:20,  1.99it/s]
 19%|█▉        | 2067/10708 [53:33<1:12:24,  1.99it/s]
 19%|█▉        | 2068/10708 [53:34<1:12:09,  2.00it/s]
 19%|█▉        | 2069/10708 [53:34<1:11:50,  2.00it/s]
 19%|█▉        | 2070/10708 [53:35<1:11:47,  2.01it/s]
 19%|█▉        | 2071/10708 [53:35<1:11:44,  2.01it/s]
 19%|█▉        | 2072/10708 [53:36<1:11:43,  2.01it/s]
 19%|█▉        | 2073/10708 [53:36<1:11:31,  2.01it/s]
 19%|█▉        | 2074/10708 [53:37<1:11:22,  2.02it/s]
 19%|█▉        | 2075/10708 [53:37<1:11:20,  2.02it/s]{'loss': 3.9859, 'grad_norm': 0.2191363275051117, 'learning_rate': 0.000973457423497216, 'epoch': 0.19}

                                                      

 19%|█▉        | 2075/10708 [53:37<1:11:20,  2.02it/s]
 19%|█▉        | 2076/10708 [53:38<1:11:23,  2.02it/s]
 19%|█▉        | 2077/10708 [53:38<1:11:33,  2.01it/s]
 19%|█▉        | 2078/10708 [53:39<1:11:18,  2.02it/s]
 19%|█▉        | 2079/10708 [53:39<1:11:20,  2.02it/s]
 19%|█▉        | 2080/10708 [53:40<1:11:07,  2.02it/s]
 19%|█▉        | 2081/10708 [53:40<1:11:03,  2.02it/s]
 19%|█▉        | 2082/10708 [53:41<1:10:59,  2.03it/s]
 19%|█▉        | 2083/10708 [53:41<1:10:53,  2.03it/s]
 19%|█▉        | 2084/10708 [53:42<1:10:56,  2.03it/s]
 19%|█▉        | 2085/10708 [53:42<1:11:10,  2.02it/s]
 19%|█▉        | 2086/10708 [53:43<1:11:02,  2.02it/s]
 19%|█▉        | 2087/10708 [53:43<1:10:56,  2.03it/s]
 19%|█▉        | 2088/10708 [53:44<1:10:53,  2.03it/s]
 20%|█▉        | 2089/10708 [53:44<1:11:07,  2.02it/s]
 20%|█▉        | 2090/10708 [53:45<1:10:58,  2.02it/s]
 20%|█▉        | 2091/10708 [53:45<1:10:53,  2.03it/s]
 20%|█▉        | 2092/10708 [53:46<1:10:52,  2.03it/s]
 20%|█▉        | 2093/10708 [53:46<1:10:50,  2.03it/s]
 20%|█▉        | 2094/10708 [53:47<1:11:00,  2.02it/s]
 20%|█▉        | 2095/10708 [53:47<1:10:56,  2.02it/s]
 20%|█▉        | 2096/10708 [53:48<1:10:51,  2.03it/s]
 20%|█▉        | 2097/10708 [53:48<1:11:07,  2.02it/s]
 20%|█▉        | 2098/10708 [53:49<1:11:00,  2.02it/s]
 20%|█▉        | 2099/10708 [53:49<1:10:52,  2.02it/s]
 20%|█▉        | 2100/10708 [53:50<1:11:00,  2.02it/s]
                                                      
{'loss': 3.9818, 'grad_norm': 0.22106432914733887, 'learning_rate': 0.0009721316943177283, 'epoch': 0.2}

 20%|█▉        | 2100/10708 [53:50<1:11:00,  2.02it/s]
 20%|█▉        | 2101/10708 [53:50<1:11:11,  2.02it/s]
 20%|█▉        | 2102/10708 [53:51<1:11:05,  2.02it/s]
 20%|█▉        | 2103/10708 [53:51<1:11:04,  2.02it/s]
 20%|█▉        | 2104/10708 [53:52<1:11:02,  2.02it/s]
 20%|█▉        | 2105/10708 [53:52<1:10:57,  2.02it/s]
 20%|█▉        | 2106/10708 [53:53<1:10:53,  2.02it/s]
 20%|█▉        | 2107/10708 [53:53<1:10:44,  2.03it/s]
 20%|█▉        | 2108/10708 [53:54<1:10:42,  2.03it/s]
 20%|█▉        | 2109/10708 [53:54<1:10:58,  2.02it/s]
 20%|█▉        | 2110/10708 [53:55<1:11:15,  2.01it/s]
 20%|█▉        | 2111/10708 [53:55<1:11:09,  2.01it/s]
 20%|█▉        | 2112/10708 [53:56<1:11:26,  2.01it/s]
 20%|█▉        | 2113/10708 [53:56<1:11:17,  2.01it/s]
 20%|█▉        | 2114/10708 [53:57<1:11:02,  2.02it/s]
 20%|█▉        | 2115/10708 [53:57<1:11:04,  2.02it/s]
 20%|█▉        | 2116/10708 [53:58<1:11:01,  2.02it/s]
 20%|█▉        | 2117/10708 [53:58<1:11:10,  2.01it/s]
 20%|█▉        | 2118/10708 [53:59<1:11:01,  2.02it/s]
 20%|█▉        | 2119/10708 [53:59<1:11:00,  2.02it/s]
 20%|█▉        | 2120/10708 [54:00<1:10:57,  2.02it/s]
 20%|█▉        | 2121/10708 [54:00<1:10:50,  2.02it/s]
 20%|█▉        | 2122/10708 [54:01<1:10:54,  2.02it/s]
 20%|█▉        | 2123/10708 [54:01<1:10:55,  2.02it/s]
 20%|█▉        | 2124/10708 [54:02<1:10:48,  2.02it/s]
 20%|█▉        | 2125/10708 [54:02<1:10:49,  2.02it/s]
                                                      
{'loss': 3.985, 'grad_norm': 0.22493590414524078, 'learning_rate': 0.0009707746065287648, 'epoch': 0.2}

 20%|█▉        | 2125/10708 [54:02<1:10:49,  2.02it/s]
 20%|█▉        | 2126/10708 [54:03<1:10:50,  2.02it/s]
 20%|█▉        | 2127/10708 [54:03<1:11:00,  2.01it/s]
 20%|█▉        | 2128/10708 [54:04<1:10:50,  2.02it/s]
 20%|█▉        | 2129/10708 [54:04<1:10:46,  2.02it/s]
 20%|█▉        | 2130/10708 [54:05<1:10:46,  2.02it/s]
 20%|█▉        | 2131/10708 [54:05<1:10:44,  2.02it/s]
 20%|█▉        | 2132/10708 [54:06<1:10:40,  2.02it/s]
 20%|█▉        | 2133/10708 [54:06<1:10:30,  2.03it/s]
 20%|█▉        | 2134/10708 [54:07<1:10:32,  2.03it/s]
 20%|█▉        | 2135/10708 [54:07<1:10:32,  2.03it/s]
 20%|█▉        | 2136/10708 [54:08<1:10:53,  2.02it/s]
 20%|█▉        | 2137/10708 [54:08<1:10:45,  2.02it/s]
 20%|█▉        | 2138/10708 [54:09<1:10:39,  2.02it/s]
 20%|█▉        | 2139/10708 [54:09<1:10:32,  2.02it/s]
 20%|█▉        | 2140/10708 [54:09<1:10:35,  2.02it/s]
 20%|█▉        | 2141/10708 [54:10<1:10:38,  2.02it/s]
 20%|██        | 2142/10708 [54:10<1:10:32,  2.02it/s]
 20%|██        | 2143/10708 [54:11<1:10:34,  2.02it/s]
 20%|██        | 2144/10708 [54:11<1:10:30,  2.02it/s]
 20%|██        | 2145/10708 [54:12<1:10:30,  2.02it/s]
 20%|██        | 2146/10708 [54:12<1:10:32,  2.02it/s]
 20%|██        | 2147/10708 [54:13<1:10:36,  2.02it/s]
 20%|██        | 2148/10708 [54:13<1:10:33,  2.02it/s]
 20%|██        | 2149/10708 [54:14<1:10:29,  2.02it/s]
 20%|██        | 2150/10708 [54:14<1:10:23,  2.03it/s]
                                                      
{'loss': 3.9674, 'grad_norm': 0.21516136825084686, 'learning_rate': 0.0009693862502670111, 'epoch': 0.2}

 20%|██        | 2150/10708 [54:14<1:10:23,  2.03it/s]
 20%|██        | 2151/10708 [54:15<1:10:24,  2.03it/s]
 20%|██        | 2152/10708 [54:15<1:10:29,  2.02it/s]
 20%|██        | 2153/10708 [54:16<1:10:24,  2.02it/s]
 20%|██        | 2154/10708 [54:16<1:10:28,  2.02it/s]
 20%|██        | 2155/10708 [54:17<1:10:20,  2.03it/s]
 20%|██        | 2156/10708 [54:17<1:10:25,  2.02it/s]
 20%|██        | 2157/10708 [54:18<1:10:20,  2.03it/s]
 20%|██        | 2158/10708 [54:18<1:10:16,  2.03it/s]
 20%|██        | 2159/10708 [54:19<1:10:16,  2.03it/s]
 20%|██        | 2160/10708 [54:19<1:10:17,  2.03it/s]
 20%|██        | 2161/10708 [54:20<1:10:17,  2.03it/s]
 20%|██        | 2162/10708 [54:20<1:10:15,  2.03it/s]
 20%|██        | 2163/10708 [54:21<1:10:16,  2.03it/s]
 20%|██        | 2164/10708 [54:21<1:10:10,  2.03it/s]
 20%|██        | 2165/10708 [54:22<1:10:06,  2.03it/s]
 20%|██        | 2166/10708 [54:22<1:10:11,  2.03it/s]
 20%|██        | 2167/10708 [54:23<1:10:21,  2.02it/s]
 20%|██        | 2168/10708 [54:23<1:10:20,  2.02it/s]
 20%|██        | 2169/10708 [54:24<1:10:15,  2.03it/s]
 20%|██        | 2170/10708 [54:24<1:10:17,  2.02it/s]
 20%|██        | 2171/10708 [54:25<1:10:11,  2.03it/s]
 20%|██        | 2172/10708 [54:25<1:10:12,  2.03it/s]
 20%|██        | 2173/10708 [54:26<1:10:15,  2.02it/s]
 20%|██        | 2174/10708 [54:26<1:10:16,  2.02it/s]
 20%|██        | 2175/10708 [54:27<1:10:16,  2.02it/s]
                                                      
{'loss': 3.964, 'grad_norm': 0.2376551479101181, 'learning_rate': 0.0009679667177459793, 'epoch': 0.2}

 20%|██        | 2175/10708 [54:27<1:10:16,  2.02it/s]
 20%|██        | 2176/10708 [54:27<1:10:30,  2.02it/s]
 20%|██        | 2177/10708 [54:28<1:10:20,  2.02it/s]
 20%|██        | 2178/10708 [54:28<1:10:13,  2.02it/s]
 20%|██        | 2179/10708 [54:29<1:10:14,  2.02it/s]
 20%|██        | 2180/10708 [54:29<1:10:15,  2.02it/s]
 20%|██        | 2181/10708 [54:30<1:10:21,  2.02it/s]
 20%|██        | 2182/10708 [54:30<1:10:17,  2.02it/s]
 20%|██        | 2183/10708 [54:31<1:10:23,  2.02it/s]
 20%|██        | 2184/10708 [54:31<1:10:09,  2.03it/s]
 20%|██        | 2185/10708 [54:32<1:10:03,  2.03it/s]
 20%|██        | 2186/10708 [54:32<1:10:08,  2.02it/s]
 20%|██        | 2187/10708 [54:33<1:10:06,  2.03it/s]
 20%|██        | 2188/10708 [54:33<1:10:09,  2.02it/s]
 20%|██        | 2189/10708 [54:34<1:10:06,  2.03it/s]
 20%|██        | 2190/10708 [54:34<1:10:02,  2.03it/s]
 20%|██        | 2191/10708 [54:35<1:10:02,  2.03it/s]
 20%|██        | 2192/10708 [54:35<1:10:01,  2.03it/s]
 20%|██        | 2193/10708 [54:36<1:09:57,  2.03it/s]
 20%|██        | 2194/10708 [54:36<1:09:55,  2.03it/s]
 20%|██        | 2195/10708 [54:37<1:10:00,  2.03it/s]
 21%|██        | 2196/10708 [54:37<1:10:11,  2.02it/s]
 21%|██        | 2197/10708 [54:38<1:10:16,  2.02it/s]
 21%|██        | 2198/10708 [54:38<1:10:12,  2.02it/s]
 21%|██        | 2199/10708 [54:39<1:10:05,  2.02it/s]
 21%|██        | 2200/10708 [54:39<1:10:01,  2.03it/s]
                                                      
{'loss': 3.9718, 'grad_norm': 0.22017095983028412, 'learning_rate': 0.0009665161032498842, 'epoch': 0.21}

 21%|██        | 2200/10708 [54:39<1:10:01,  2.03it/s]
 21%|██        | 2201/10708 [54:40<1:10:07,  2.02it/s]
 21%|██        | 2202/10708 [54:40<1:10:00,  2.03it/s]
 21%|██        | 2203/10708 [54:41<1:10:06,  2.02it/s]
 21%|██        | 2204/10708 [54:41<1:10:11,  2.02it/s]
 21%|██        | 2205/10708 [54:42<1:10:04,  2.02it/s]
 21%|██        | 2206/10708 [54:42<1:10:24,  2.01it/s]
 21%|██        | 2207/10708 [54:43<1:10:24,  2.01it/s]
 21%|██        | 2208/10708 [54:43<1:10:18,  2.01it/s]
 21%|██        | 2209/10708 [54:44<1:10:14,  2.02it/s]
 21%|██        | 2210/10708 [54:44<1:10:07,  2.02it/s]
 21%|██        | 2211/10708 [54:45<1:10:04,  2.02it/s]
 21%|██        | 2212/10708 [54:45<1:09:58,  2.02it/s]
 21%|██        | 2213/10708 [54:46<1:09:52,  2.03it/s]
 21%|██        | 2214/10708 [54:46<1:10:05,  2.02it/s]
 21%|██        | 2215/10708 [54:47<1:10:07,  2.02it/s]
 21%|██        | 2216/10708 [54:47<1:10:03,  2.02it/s]
 21%|██        | 2217/10708 [54:48<1:10:21,  2.01it/s]
 21%|██        | 2218/10708 [54:48<1:10:09,  2.02it/s]
 21%|██        | 2219/10708 [54:49<1:10:02,  2.02it/s]
 21%|██        | 2220/10708 [54:49<1:09:54,  2.02it/s]
 21%|██        | 2221/10708 [54:50<1:09:59,  2.02it/s]
 21%|██        | 2222/10708 [54:50<1:09:55,  2.02it/s]
 21%|██        | 2223/10708 [54:51<1:09:52,  2.02it/s]
 21%|██        | 2224/10708 [54:51<1:09:49,  2.02it/s]
 21%|██        | 2225/10708 [54:52<1:09:47,  2.03it/s]
                                                      
{'loss': 3.9606, 'grad_norm': 0.2246270328760147, 'learning_rate': 0.0009650345031273802, 'epoch': 0.21}

 21%|██        | 2225/10708 [54:52<1:09:47,  2.03it/s]
 21%|██        | 2226/10708 [54:52<1:09:48,  2.02it/s]
 21%|██        | 2227/10708 [54:52<1:09:54,  2.02it/s]
 21%|██        | 2228/10708 [54:53<1:09:49,  2.02it/s]
 21%|██        | 2229/10708 [54:53<1:09:53,  2.02it/s]
 21%|██        | 2230/10708 [54:54<1:09:53,  2.02it/s]
 21%|██        | 2231/10708 [54:54<1:09:52,  2.02it/s]
 21%|██        | 2232/10708 [54:55<1:09:49,  2.02it/s]
 21%|██        | 2233/10708 [54:55<1:09:43,  2.03it/s]
 21%|██        | 2234/10708 [54:56<1:09:51,  2.02it/s]
 21%|██        | 2235/10708 [54:56<1:09:40,  2.03it/s]
 21%|██        | 2236/10708 [54:57<1:09:44,  2.02it/s]
 21%|██        | 2237/10708 [54:57<1:09:45,  2.02it/s]
 21%|██        | 2238/10708 [54:58<1:09:44,  2.02it/s]
 21%|██        | 2239/10708 [54:58<1:09:40,  2.03it/s]
 21%|██        | 2240/10708 [54:59<1:09:34,  2.03it/s]
 21%|██        | 2241/10708 [54:59<1:09:34,  2.03it/s]
 21%|██        | 2242/10708 [55:00<1:09:28,  2.03it/s]
 21%|██        | 2243/10708 [55:01<1:18:29,  1.80it/s]
 21%|██        | 2244/10708 [55:01<1:15:44,  1.86it/s]
 21%|██        | 2245/10708 [55:02<1:13:53,  1.91it/s]
 21%|██        | 2246/10708 [55:02<1:12:33,  1.94it/s]
 21%|██        | 2247/10708 [55:03<1:11:42,  1.97it/s]
 21%|██        | 2248/10708 [55:03<1:11:05,  1.98it/s]
 21%|██        | 2249/10708 [55:04<1:10:38,  2.00it/s]
 21%|██        | 2250/10708 [55:04<1:10:27,  2.00it/s]
                                                      
{'loss': 3.9538, 'grad_norm': 0.21303685009479523, 'learning_rate': 0.0009635220157851621, 'epoch': 0.21}

 21%|██        | 2250/10708 [55:04<1:10:27,  2.00it/s]
 21%|██        | 2251/10708 [55:05<1:10:16,  2.01it/s]
 21%|██        | 2252/10708 [55:05<1:10:00,  2.01it/s]
 21%|██        | 2253/10708 [55:06<1:09:48,  2.02it/s]
 21%|██        | 2254/10708 [55:06<1:09:40,  2.02it/s]
 21%|██        | 2255/10708 [55:07<1:09:32,  2.03it/s]
 21%|██        | 2256/10708 [55:07<1:09:29,  2.03it/s]
 21%|██        | 2257/10708 [55:08<1:09:34,  2.02it/s]
 21%|██        | 2258/10708 [55:08<1:09:29,  2.03it/s]
 21%|██        | 2259/10708 [55:08<1:09:38,  2.02it/s]
 21%|██        | 2260/10708 [55:09<1:09:37,  2.02it/s]
 21%|██        | 2261/10708 [55:09<1:09:32,  2.02it/s]
 21%|██        | 2262/10708 [55:10<1:09:29,  2.03it/s]
 21%|██        | 2263/10708 [55:10<1:09:25,  2.03it/s]
 21%|██        | 2264/10708 [55:11<1:09:29,  2.03it/s]
 21%|██        | 2265/10708 [55:11<1:09:31,  2.02it/s]
 21%|██        | 2266/10708 [55:12<1:09:22,  2.03it/s]
 21%|██        | 2267/10708 [55:12<1:09:22,  2.03it/s]
 21%|██        | 2268/10708 [55:13<1:09:25,  2.03it/s]
 21%|██        | 2269/10708 [55:13<1:09:19,  2.03it/s]
 21%|██        | 2270/10708 [55:14<1:09:25,  2.03it/s]
 21%|██        | 2271/10708 [55:14<1:09:28,  2.02it/s]
 21%|██        | 2272/10708 [55:15<1:09:48,  2.01it/s]
 21%|██        | 2273/10708 [55:15<1:09:40,  2.02it/s]
 21%|██        | 2274/10708 [55:16<1:09:40,  2.02it/s]
 21%|██        | 2275/10708 [55:16<1:09:29,  2.02it/s]
                                                      
{'loss': 3.9509, 'grad_norm': 0.23261603713035583, 'learning_rate': 0.0009619787416814296, 'epoch': 0.21}

 21%|██        | 2275/10708 [55:16<1:09:29,  2.02it/s]
 21%|██▏       | 2276/10708 [55:17<1:09:29,  2.02it/s]
 21%|██▏       | 2277/10708 [55:17<1:09:20,  2.03it/s]
 21%|██▏       | 2278/10708 [55:18<1:09:16,  2.03it/s]
 21%|██▏       | 2279/10708 [55:18<1:09:18,  2.03it/s]
 21%|██▏       | 2280/10708 [55:19<1:09:19,  2.03it/s]
 21%|██▏       | 2281/10708 [55:19<1:09:18,  2.03it/s]
 21%|██▏       | 2282/10708 [55:20<1:09:12,  2.03it/s]
 21%|██▏       | 2283/10708 [55:20<1:09:13,  2.03it/s]
 21%|██▏       | 2284/10708 [55:21<1:09:26,  2.02it/s]
 21%|██▏       | 2285/10708 [55:21<1:09:31,  2.02it/s]
 21%|██▏       | 2286/10708 [55:22<1:09:30,  2.02it/s]
 21%|██▏       | 2287/10708 [55:22<1:09:34,  2.02it/s]
 21%|██▏       | 2288/10708 [55:23<1:09:28,  2.02it/s]
 21%|██▏       | 2289/10708 [55:23<1:09:31,  2.02it/s]
 21%|██▏       | 2290/10708 [55:24<1:09:22,  2.02it/s]
 21%|██▏       | 2291/10708 [55:24<1:09:25,  2.02it/s]
 21%|██▏       | 2292/10708 [55:25<1:09:18,  2.02it/s]
 21%|██▏       | 2293/10708 [55:25<1:09:11,  2.03it/s]
 21%|██▏       | 2294/10708 [55:26<1:09:14,  2.03it/s]
 21%|██▏       | 2295/10708 [55:26<1:09:13,  2.03it/s]
 21%|██▏       | 2296/10708 [55:27<1:09:11,  2.03it/s]
 21%|██▏       | 2297/10708 [55:27<1:09:15,  2.02it/s]
 21%|██▏       | 2298/10708 [55:28<1:09:09,  2.03it/s]
 21%|██▏       | 2299/10708 [55:28<1:09:09,  2.03it/s]
 21%|██▏       | 2300/10708 [55:29<1:09:06,  2.03it/s]
                                                      
{'loss': 3.9436, 'grad_norm': 0.22946855425834656, 'learning_rate': 0.000960404783319214, 'epoch': 0.21}

 21%|██▏       | 2300/10708 [55:29<1:09:06,  2.03it/s]
 21%|██▏       | 2301/10708 [55:29<1:09:21,  2.02it/s]
 21%|██▏       | 2302/10708 [55:30<1:09:19,  2.02it/s]
 22%|██▏       | 2303/10708 [55:30<1:09:14,  2.02it/s]
 22%|██▏       | 2304/10708 [55:31<1:09:17,  2.02it/s]
 22%|██▏       | 2305/10708 [55:31<1:09:13,  2.02it/s]
 22%|██▏       | 2306/10708 [55:32<1:09:01,  2.03it/s]
 22%|██▏       | 2307/10708 [55:32<1:08:59,  2.03it/s]
 22%|██▏       | 2308/10708 [55:33<1:08:59,  2.03it/s]
 22%|██▏       | 2309/10708 [55:33<1:08:52,  2.03it/s]
 22%|██▏       | 2310/10708 [55:34<1:08:55,  2.03it/s]
 22%|██▏       | 2311/10708 [55:34<1:08:59,  2.03it/s]
 22%|██▏       | 2312/10708 [55:35<1:09:04,  2.03it/s]
 22%|██▏       | 2313/10708 [55:35<1:09:10,  2.02it/s]
 22%|██▏       | 2314/10708 [55:36<1:09:07,  2.02it/s]
 22%|██▏       | 2315/10708 [55:36<1:09:04,  2.02it/s]
 22%|██▏       | 2316/10708 [55:37<1:09:11,  2.02it/s]
 22%|██▏       | 2317/10708 [55:37<1:09:09,  2.02it/s]
 22%|██▏       | 2318/10708 [55:38<1:09:14,  2.02it/s]
 22%|██▏       | 2319/10708 [55:38<1:09:09,  2.02it/s]
 22%|██▏       | 2320/10708 [55:39<1:09:13,  2.02it/s]
 22%|██▏       | 2321/10708 [55:39<1:09:12,  2.02it/s]
 22%|██▏       | 2322/10708 [55:40<1:09:03,  2.02it/s]
 22%|██▏       | 2323/10708 [55:40<1:09:01,  2.02it/s]
 22%|██▏       | 2324/10708 [55:41<1:08:54,  2.03it/s]
 22%|██▏       | 2325/10708 [55:41<1:08:59,  2.02it/s]
                                                      {'loss': 3.9419, 'grad_norm': 0.21622878313064575, 'learning_rate': 0.0009588002452395707, 'epoch': 0.22}


 22%|██▏       | 2325/10708 [55:41<1:08:59,  2.02it/s]
 22%|██▏       | 2326/10708 [55:42<1:09:00,  2.02it/s]
 22%|██▏       | 2327/10708 [55:42<1:08:57,  2.03it/s]
 22%|██▏       | 2328/10708 [55:43<1:08:56,  2.03it/s]
 22%|██▏       | 2329/10708 [55:43<1:08:53,  2.03it/s]
 22%|██▏       | 2330/10708 [55:44<1:08:54,  2.03it/s]
 22%|██▏       | 2331/10708 [55:44<1:08:52,  2.03it/s]
 22%|██▏       | 2332/10708 [55:45<1:08:53,  2.03it/s]
 22%|██▏       | 2333/10708 [55:45<1:08:53,  2.03it/s]
 22%|██▏       | 2334/10708 [55:46<1:08:54,  2.03it/s]
 22%|██▏       | 2335/10708 [55:46<1:09:02,  2.02it/s]
 22%|██▏       | 2336/10708 [55:47<1:08:57,  2.02it/s]
 22%|██▏       | 2337/10708 [55:47<1:08:48,  2.03it/s]
 22%|██▏       | 2338/10708 [55:48<1:08:52,  2.03it/s]
 22%|██▏       | 2339/10708 [55:48<1:08:52,  2.03it/s]
 22%|██▏       | 2340/10708 [55:49<1:08:53,  2.02it/s]
 22%|██▏       | 2341/10708 [55:49<1:08:49,  2.03it/s]
 22%|██▏       | 2342/10708 [55:49<1:08:44,  2.03it/s]
 22%|██▏       | 2343/10708 [55:50<1:08:49,  2.03it/s]
 22%|██▏       | 2344/10708 [55:50<1:08:43,  2.03it/s]
 22%|██▏       | 2345/10708 [55:51<1:08:43,  2.03it/s]
 22%|██▏       | 2346/10708 [55:51<1:09:10,  2.01it/s]
 22%|██▏       | 2347/10708 [55:52<1:09:01,  2.02it/s]
 22%|██▏       | 2348/10708 [55:52<1:08:53,  2.02it/s]
 22%|██▏       | 2349/10708 [55:53<1:08:55,  2.02it/s]
 22%|██▏       | 2350/10708 [55:53<1:08:53,  2.02it/s]
                                                      {'loss': 3.936, 'grad_norm': 0.22584141790866852, 'learning_rate': 0.0009571652340146357, 'epoch': 0.22}


 22%|██▏       | 2350/10708 [55:53<1:08:53,  2.02it/s]
 22%|██▏       | 2351/10708 [55:54<1:08:56,  2.02it/s]
 22%|██▏       | 2352/10708 [55:54<1:08:51,  2.02it/s]
 22%|██▏       | 2353/10708 [55:55<1:08:46,  2.02it/s]
 22%|██▏       | 2354/10708 [55:55<1:08:44,  2.03it/s]
 22%|██▏       | 2355/10708 [55:56<1:08:38,  2.03it/s]
 22%|██▏       | 2356/10708 [55:56<1:08:35,  2.03it/s]
 22%|██▏       | 2357/10708 [55:57<1:08:32,  2.03it/s]
 22%|██▏       | 2358/10708 [55:57<1:08:30,  2.03it/s]
 22%|██▏       | 2359/10708 [55:58<1:08:41,  2.03it/s]
 22%|██▏       | 2360/10708 [55:58<1:08:37,  2.03it/s]
 22%|██▏       | 2361/10708 [55:59<1:08:46,  2.02it/s]
 22%|██▏       | 2362/10708 [55:59<1:08:46,  2.02it/s]
 22%|██▏       | 2363/10708 [56:00<1:08:46,  2.02it/s]
 22%|██▏       | 2364/10708 [56:00<1:08:47,  2.02it/s]
 22%|██▏       | 2365/10708 [56:01<1:08:46,  2.02it/s]
 22%|██▏       | 2366/10708 [56:01<1:08:43,  2.02it/s]
 22%|██▏       | 2367/10708 [56:02<1:08:41,  2.02it/s]
 22%|██▏       | 2368/10708 [56:02<1:08:33,  2.03it/s]
 22%|██▏       | 2369/10708 [56:03<1:08:33,  2.03it/s]
 22%|██▏       | 2370/10708 [56:03<1:08:32,  2.03it/s]
 22%|██▏       | 2371/10708 [56:04<1:08:25,  2.03it/s]
 22%|██▏       | 2372/10708 [56:04<1:08:22,  2.03it/s]
 22%|██▏       | 2373/10708 [56:05<1:08:22,  2.03it/s]
 22%|██▏       | 2374/10708 [56:05<1:08:28,  2.03it/s]
 22%|██▏       | 2375/10708 [56:06<1:08:29,  2.03it/s]
                                                      
{'loss': 3.9378, 'grad_norm': 0.21406668424606323, 'learning_rate': 0.0009554998582405466, 'epoch': 0.22}

 22%|██▏       | 2375/10708 [56:06<1:08:29,  2.03it/s]
 22%|██▏       | 2376/10708 [56:06<1:08:34,  2.02it/s]
 22%|██▏       | 2377/10708 [56:07<1:08:38,  2.02it/s]
 22%|██▏       | 2378/10708 [56:07<1:08:45,  2.02it/s]
 22%|██▏       | 2379/10708 [56:08<1:08:42,  2.02it/s]
 22%|██▏       | 2380/10708 [56:08<1:08:35,  2.02it/s]
 22%|██▏       | 2381/10708 [56:09<1:08:31,  2.03it/s]
 22%|██▏       | 2382/10708 [56:09<1:08:24,  2.03it/s]
 22%|██▏       | 2383/10708 [56:10<1:08:38,  2.02it/s]
 22%|██▏       | 2384/10708 [56:10<1:08:33,  2.02it/s]
 22%|██▏       | 2385/10708 [56:11<1:08:30,  2.02it/s]
 22%|██▏       | 2386/10708 [56:11<1:08:26,  2.03it/s]
 22%|██▏       | 2387/10708 [56:12<1:08:21,  2.03it/s]
 22%|██▏       | 2388/10708 [56:12<1:08:16,  2.03it/s]
 22%|██▏       | 2389/10708 [56:13<1:08:24,  2.03it/s]
 22%|██▏       | 2390/10708 [56:13<1:08:29,  2.02it/s]
 22%|██▏       | 2391/10708 [56:14<1:08:25,  2.03it/s]
 22%|██▏       | 2392/10708 [56:14<1:08:21,  2.03it/s]
 22%|██▏       | 2393/10708 [56:15<1:08:16,  2.03it/s]
 22%|██▏       | 2394/10708 [56:15<1:08:17,  2.03it/s]
 22%|██▏       | 2395/10708 [56:16<1:08:19,  2.03it/s]
 22%|██▏       | 2396/10708 [56:16<1:08:21,  2.03it/s]
 22%|██▏       | 2397/10708 [56:17<1:08:18,  2.03it/s]
 22%|██▏       | 2398/10708 [56:17<1:08:37,  2.02it/s]
 22%|██▏       | 2399/10708 [56:18<1:08:33,  2.02it/s]
 22%|██▏       | 2400/10708 [56:18<1:08:26,  2.02it/s]
                                                      {'loss': 3.9296, 'grad_norm': 0.21196231245994568, 'learning_rate': 0.00095380422853023, 'epoch': 0.22}


 22%|██▏       | 2400/10708 [56:18<1:08:26,  2.02it/s]
 22%|██▏       | 2401/10708 [56:19<1:08:25,  2.02it/s]
 22%|██▏       | 2402/10708 [56:19<1:08:16,  2.03it/s]
 22%|██▏       | 2403/10708 [56:20<1:08:23,  2.02it/s]
 22%|██▏       | 2404/10708 [56:20<1:08:17,  2.03it/s]
 22%|██▏       | 2405/10708 [56:21<1:08:14,  2.03it/s]
 22%|██▏       | 2406/10708 [56:21<1:08:16,  2.03it/s]
 22%|██▏       | 2407/10708 [56:22<1:08:08,  2.03it/s]
 22%|██▏       | 2408/10708 [56:22<1:08:05,  2.03it/s]
 22%|██▏       | 2409/10708 [56:23<1:08:06,  2.03it/s]
 23%|██▎       | 2410/10708 [56:23<1:08:03,  2.03it/s]
 23%|██▎       | 2411/10708 [56:24<1:08:14,  2.03it/s]
 23%|██▎       | 2412/10708 [56:24<1:08:11,  2.03it/s]
 23%|██▎       | 2413/10708 [56:25<1:08:16,  2.03it/s]
 23%|██▎       | 2414/10708 [56:25<1:08:09,  2.03it/s]
 23%|██▎       | 2415/10708 [56:26<1:08:00,  2.03it/s]
 23%|██▎       | 2416/10708 [56:26<1:08:02,  2.03it/s]
 23%|██▎       | 2417/10708 [56:27<1:08:00,  2.03it/s]
 23%|██▎       | 2418/10708 [56:27<1:07:54,  2.03it/s]
 23%|██▎       | 2419/10708 [56:27<1:08:06,  2.03it/s]
 23%|██▎       | 2420/10708 [56:28<1:08:00,  2.03it/s]
 23%|██▎       | 2421/10708 [56:28<1:08:10,  2.03it/s]
 23%|██▎       | 2422/10708 [56:29<1:08:02,  2.03it/s]
 23%|██▎       | 2423/10708 [56:29<1:07:55,  2.03it/s]
 23%|██▎       | 2424/10708 [56:30<1:07:56,  2.03it/s]
 23%|██▎       | 2425/10708 [56:30<1:07:54,  2.03it/s]{'loss': 3.9279, 'grad_norm': 0.2121303677558899, 'learning_rate': 0.0009520784575060553, 'epoch': 0.23}
                                                      

 23%|██▎       | 2425/10708 [56:30<1:07:54,  2.03it/s]
 23%|██▎       | 2426/10708 [56:31<1:08:02,  2.03it/s]
 23%|██▎       | 2427/10708 [56:31<1:08:01,  2.03it/s]
 23%|██▎       | 2428/10708 [56:32<1:07:57,  2.03it/s]
 23%|██▎       | 2429/10708 [56:32<1:07:58,  2.03it/s]
 23%|██▎       | 2430/10708 [56:33<1:07:57,  2.03it/s]
 23%|██▎       | 2431/10708 [56:33<1:08:05,  2.03it/s]
 23%|██▎       | 2432/10708 [56:34<1:08:01,  2.03it/s]
 23%|██▎       | 2433/10708 [56:34<1:07:58,  2.03it/s]
 23%|██▎       | 2434/10708 [56:35<1:07:59,  2.03it/s]
 23%|██▎       | 2435/10708 [56:35<1:07:56,  2.03it/s]
 23%|██▎       | 2436/10708 [56:36<1:08:11,  2.02it/s]
 23%|██▎       | 2437/10708 [56:36<1:08:07,  2.02it/s]
 23%|██▎       | 2438/10708 [56:37<1:08:08,  2.02it/s]
 23%|██▎       | 2439/10708 [56:37<1:07:59,  2.03it/s]
 23%|██▎       | 2440/10708 [56:38<1:08:01,  2.03it/s]
 23%|██▎       | 2441/10708 [56:38<1:07:56,  2.03it/s]
 23%|██▎       | 2442/10708 [56:39<1:07:47,  2.03it/s]
 23%|██▎       | 2443/10708 [56:39<1:07:51,  2.03it/s]
 23%|██▎       | 2444/10708 [56:40<1:07:47,  2.03it/s]
 23%|██▎       | 2445/10708 [56:40<1:07:42,  2.03it/s]
 23%|██▎       | 2446/10708 [56:41<1:07:45,  2.03it/s]
 23%|██▎       | 2447/10708 [56:42<1:18:57,  1.74it/s]
 23%|██▎       | 2448/10708 [56:42<1:15:35,  1.82it/s]
 23%|██▎       | 2449/10708 [56:43<1:13:12,  1.88it/s]
 23%|██▎       | 2450/10708 [56:43<1:11:45,  1.92it/s]{'loss': 3.9252, 'grad_norm': 0.21555061638355255, 'learning_rate': 0.0009503226597923529, 'epoch': 0.23}

                                                      

 23%|██▎       | 2450/10708 [56:43<1:11:45,  1.92it/s]
 23%|██▎       | 2451/10708 [56:44<1:10:42,  1.95it/s]
 23%|██▎       | 2452/10708 [56:44<1:09:48,  1.97it/s]
 23%|██▎       | 2453/10708 [56:45<1:09:10,  1.99it/s]
 23%|██▎       | 2454/10708 [56:45<1:08:41,  2.00it/s]
 23%|██▎       | 2455/10708 [56:46<1:08:21,  2.01it/s]
 23%|██▎       | 2456/10708 [56:46<1:08:10,  2.02it/s]
 23%|██▎       | 2457/10708 [56:46<1:08:01,  2.02it/s]
 23%|██▎       | 2458/10708 [56:47<1:07:55,  2.02it/s]
 23%|██▎       | 2459/10708 [56:48<1:18:53,  1.74it/s]
 23%|██▎       | 2460/10708 [56:48<1:15:31,  1.82it/s]
 23%|██▎       | 2461/10708 [56:49<1:13:09,  1.88it/s]
 23%|██▎       | 2462/10708 [56:49<1:11:43,  1.92it/s]
 23%|██▎       | 2463/10708 [56:50<1:10:30,  1.95it/s]
 23%|██▎       | 2464/10708 [56:50<1:09:46,  1.97it/s]
 23%|██▎       | 2465/10708 [56:51<1:09:02,  1.99it/s]
 23%|██▎       | 2466/10708 [56:51<1:08:47,  2.00it/s]
 23%|██▎       | 2467/10708 [56:52<1:08:27,  2.01it/s]
 23%|██▎       | 2468/10708 [56:52<1:08:04,  2.02it/s]
 23%|██▎       | 2469/10708 [56:53<1:07:54,  2.02it/s]
 23%|██▎       | 2470/10708 [56:53<1:07:53,  2.02it/s]
 23%|██▎       | 2471/10708 [56:54<1:07:47,  2.03it/s]
 23%|██▎       | 2472/10708 [56:54<1:07:46,  2.03it/s]
 23%|██▎       | 2473/10708 [56:55<1:07:47,  2.02it/s]
 23%|██▎       | 2474/10708 [56:55<1:07:50,  2.02it/s]
 23%|██▎       | 2475/10708 [56:56<1:07:41,  2.03it/s]
                                                      
{'loss': 3.9189, 'grad_norm': 0.21377475559711456, 'learning_rate': 0.0009485369520078029, 'epoch': 0.23}

 23%|██▎       | 2475/10708 [56:56<1:07:41,  2.03it/s]
 23%|██▎       | 2476/10708 [56:56<1:07:46,  2.02it/s]
 23%|██▎       | 2477/10708 [56:57<1:07:40,  2.03it/s]
 23%|██▎       | 2478/10708 [56:57<1:07:30,  2.03it/s]
 23%|██▎       | 2479/10708 [56:58<1:07:31,  2.03it/s]
 23%|██▎       | 2480/10708 [56:58<1:07:33,  2.03it/s]
 23%|██▎       | 2481/10708 [56:59<1:07:27,  2.03it/s]
 23%|██▎       | 2482/10708 [56:59<1:07:32,  2.03it/s]
 23%|██▎       | 2483/10708 [57:00<1:07:30,  2.03it/s]
 23%|██▎       | 2484/10708 [57:00<1:07:29,  2.03it/s]
 23%|██▎       | 2485/10708 [57:01<1:07:26,  2.03it/s]
 23%|██▎       | 2486/10708 [57:01<1:07:26,  2.03it/s]
 23%|██▎       | 2487/10708 [57:02<1:07:33,  2.03it/s]
 23%|██▎       | 2488/10708 [57:02<1:07:25,  2.03it/s]
 23%|██▎       | 2489/10708 [57:03<1:07:25,  2.03it/s]
 23%|██▎       | 2490/10708 [57:03<1:07:24,  2.03it/s]
 23%|██▎       | 2491/10708 [57:04<1:07:25,  2.03it/s]
 23%|██▎       | 2492/10708 [57:04<1:07:32,  2.03it/s]
 23%|██▎       | 2493/10708 [57:05<1:07:32,  2.03it/s]
 23%|██▎       | 2494/10708 [57:05<1:07:31,  2.03it/s]
 23%|██▎       | 2495/10708 [57:05<1:07:25,  2.03it/s]
 23%|██▎       | 2496/10708 [57:06<1:07:22,  2.03it/s]
 23%|██▎       | 2497/10708 [57:06<1:07:26,  2.03it/s]
 23%|██▎       | 2498/10708 [57:07<1:07:19,  2.03it/s]
 23%|██▎       | 2499/10708 [57:07<1:07:16,  2.03it/s]
 23%|██▎       | 2500/10708 [57:08<1:07:14,  2.03it/s]{'loss': 3.9213, 'grad_norm': 0.2201596349477768, 'learning_rate': 0.000946721452757688, 'epoch': 0.23}
                                                      

 23%|██▎       | 2500/10708 [57:08<1:07:14,  2.03it/s]
 23%|██▎       | 2501/10708 [57:08<1:07:29,  2.03it/s]
 23%|██▎       | 2502/10708 [57:09<1:07:26,  2.03it/s]
 23%|██▎       | 2503/10708 [57:09<1:07:19,  2.03it/s]
 23%|██▎       | 2504/10708 [57:10<1:07:24,  2.03it/s]
 23%|██▎       | 2505/10708 [57:10<1:07:21,  2.03it/s]
 23%|██▎       | 2506/10708 [57:11<1:07:19,  2.03it/s]
 23%|██▎       | 2507/10708 [57:11<1:07:23,  2.03it/s]
 23%|██▎       | 2508/10708 [57:12<1:07:23,  2.03it/s]
 23%|██▎       | 2509/10708 [57:12<1:07:35,  2.02it/s]
 23%|██▎       | 2510/10708 [57:13<1:07:55,  2.01it/s]
 23%|██▎       | 2511/10708 [57:13<1:07:51,  2.01it/s]
 23%|██▎       | 2512/10708 [57:14<1:07:44,  2.02it/s]
 23%|██▎       | 2513/10708 [57:14<1:07:31,  2.02it/s]
 23%|██▎       | 2514/10708 [57:15<1:07:26,  2.02it/s]
 23%|██▎       | 2515/10708 [57:15<1:07:25,  2.03it/s]
 23%|██▎       | 2516/10708 [57:16<1:07:16,  2.03it/s]
 24%|██▎       | 2517/10708 [57:17<1:15:29,  1.81it/s]
 24%|██▎       | 2518/10708 [57:17<1:13:00,  1.87it/s]
 24%|██▎       | 2519/10708 [57:18<1:11:08,  1.92it/s]
 24%|██▎       | 2520/10708 [57:18<1:10:03,  1.95it/s]
 24%|██▎       | 2521/10708 [57:19<1:09:09,  1.97it/s]
 24%|██▎       | 2522/10708 [57:19<1:08:28,  1.99it/s]
 24%|██▎       | 2523/10708 [57:20<1:08:13,  2.00it/s]
 24%|██▎       | 2524/10708 [57:20<1:07:50,  2.01it/s]
 24%|██▎       | 2525/10708 [57:20<1:07:35,  2.02it/s]{'loss': 3.9138, 'grad_norm': 0.2335921972990036, 'learning_rate': 0.0009448762826260158, 'epoch': 0.24}
                                                      

 24%|██▎       | 2525/10708 [57:20<1:07:35,  2.02it/s]
 24%|██▎       | 2526/10708 [57:21<1:07:32,  2.02it/s]
 24%|██▎       | 2527/10708 [57:21<1:07:26,  2.02it/s]
 24%|██▎       | 2528/10708 [57:22<1:07:21,  2.02it/s]
 24%|██▎       | 2529/10708 [57:22<1:07:11,  2.03it/s]
 24%|██▎       | 2530/10708 [57:23<1:07:08,  2.03it/s]
 24%|██▎       | 2531/10708 [57:23<1:07:04,  2.03it/s]
 24%|██▎       | 2532/10708 [57:24<1:07:00,  2.03it/s]
 24%|██▎       | 2533/10708 [57:24<1:07:04,  2.03it/s]
 24%|██▎       | 2534/10708 [57:25<1:07:09,  2.03it/s]
 24%|██▎       | 2535/10708 [57:25<1:07:07,  2.03it/s]
 24%|██▎       | 2536/10708 [57:26<1:07:14,  2.03it/s]
 24%|██▎       | 2537/10708 [57:26<1:07:03,  2.03it/s]
 24%|██▎       | 2538/10708 [57:27<1:07:06,  2.03it/s]
 24%|██▎       | 2539/10708 [57:27<1:07:00,  2.03it/s]
 24%|██▎       | 2540/10708 [57:28<1:06:57,  2.03it/s]
 24%|██▎       | 2541/10708 [57:28<1:07:00,  2.03it/s]
 24%|██▎       | 2542/10708 [57:29<1:06:56,  2.03it/s]
 24%|██▎       | 2543/10708 [57:29<1:06:52,  2.04it/s]
 24%|██▍       | 2544/10708 [57:30<1:07:02,  2.03it/s]
 24%|██▍       | 2545/10708 [57:30<1:06:59,  2.03it/s]
 24%|██▍       | 2546/10708 [57:31<1:06:58,  2.03it/s]
 24%|██▍       | 2547/10708 [57:31<1:07:00,  2.03it/s]
 24%|██▍       | 2548/10708 [57:32<1:06:52,  2.03it/s]
 24%|██▍       | 2549/10708 [57:32<1:06:54,  2.03it/s]
 24%|██▍       | 2550/10708 [57:33<1:06:53,  2.03it/s]{'loss': 3.9164, 'grad_norm': 0.22024033963680267, 'learning_rate': 0.0009430015641675107, 'epoch': 0.24}
                                                      

 24%|██▍       | 2550/10708 [57:33<1:06:53,  2.03it/s]
 24%|██▍       | 2551/10708 [57:33<1:06:56,  2.03it/s]
 24%|██▍       | 2552/10708 [57:34<1:06:56,  2.03it/s]
 24%|██▍       | 2553/10708 [57:34<1:06:55,  2.03it/s]
 24%|██▍       | 2554/10708 [57:35<1:06:56,  2.03it/s]
 24%|██▍       | 2555/10708 [57:35<1:06:53,  2.03it/s]
 24%|██▍       | 2556/10708 [57:36<1:06:52,  2.03it/s]
 24%|██▍       | 2557/10708 [57:36<1:06:52,  2.03it/s]
 24%|██▍       | 2558/10708 [57:37<1:06:52,  2.03it/s]
 24%|██▍       | 2559/10708 [57:37<1:06:56,  2.03it/s]
 24%|██▍       | 2560/10708 [57:38<1:06:52,  2.03it/s]
 24%|██▍       | 2561/10708 [57:38<1:06:55,  2.03it/s]
 24%|██▍       | 2562/10708 [57:39<1:06:50,  2.03it/s]
 24%|██▍       | 2563/10708 [57:39<1:06:52,  2.03it/s]
 24%|██▍       | 2564/10708 [57:40<1:06:47,  2.03it/s]
 24%|██▍       | 2565/10708 [57:40<1:06:54,  2.03it/s]
 24%|██▍       | 2566/10708 [57:41<1:06:52,  2.03it/s]
 24%|██▍       | 2567/10708 [57:41<1:06:49,  2.03it/s]
 24%|██▍       | 2568/10708 [57:42<1:06:46,  2.03it/s]
 24%|██▍       | 2569/10708 [57:42<1:06:41,  2.03it/s]
 24%|██▍       | 2570/10708 [57:43<1:06:56,  2.03it/s]
 24%|██▍       | 2571/10708 [57:43<1:07:03,  2.02it/s]
 24%|██▍       | 2572/10708 [57:44<1:06:55,  2.03it/s]
 24%|██▍       | 2573/10708 [57:44<1:06:52,  2.03it/s]
 24%|██▍       | 2574/10708 [57:45<1:06:46,  2.03it/s]
 24%|██▍       | 2575/10708 [57:45<1:06:45,  2.03it/s]{'loss': 3.9007, 'grad_norm': 0.22088299691677094, 'learning_rate': 0.0009410974218994732, 'epoch': 0.24}
                                                      

 24%|██▍       | 2575/10708 [57:45<1:06:45,  2.03it/s]
 24%|██▍       | 2576/10708 [57:46<1:06:51,  2.03it/s]
 24%|██▍       | 2577/10708 [57:46<1:06:45,  2.03it/s]
 24%|██▍       | 2578/10708 [57:47<1:06:47,  2.03it/s]
 24%|██▍       | 2579/10708 [57:47<1:06:39,  2.03it/s]
 24%|██▍       | 2580/10708 [57:48<1:06:38,  2.03it/s]
 24%|██▍       | 2581/10708 [57:48<1:06:39,  2.03it/s]
 24%|██▍       | 2582/10708 [57:49<1:06:47,  2.03it/s]
 24%|██▍       | 2583/10708 [57:49<1:06:53,  2.02it/s]
 24%|██▍       | 2584/10708 [57:50<1:07:12,  2.01it/s]
 24%|██▍       | 2585/10708 [57:50<1:07:00,  2.02it/s]
 24%|██▍       | 2586/10708 [57:51<1:06:58,  2.02it/s]
 24%|██▍       | 2587/10708 [57:51<1:06:52,  2.02it/s]
 24%|██▍       | 2588/10708 [57:52<1:06:45,  2.03it/s]
 24%|██▍       | 2589/10708 [57:52<1:06:48,  2.03it/s]
 24%|██▍       | 2590/10708 [57:53<1:06:44,  2.03it/s]
 24%|██▍       | 2591/10708 [57:53<1:06:40,  2.03it/s]
 24%|██▍       | 2592/10708 [57:54<1:06:47,  2.03it/s]
 24%|██▍       | 2593/10708 [57:54<1:06:38,  2.03it/s]
 24%|██▍       | 2594/10708 [57:54<1:06:34,  2.03it/s]
 24%|██▍       | 2595/10708 [57:55<1:06:38,  2.03it/s]
 24%|██▍       | 2596/10708 [57:55<1:06:35,  2.03it/s]
 24%|██▍       | 2597/10708 [57:56<1:06:39,  2.03it/s]
 24%|██▍       | 2598/10708 [57:56<1:06:40,  2.03it/s]
 24%|██▍       | 2599/10708 [57:57<1:06:30,  2.03it/s]
 24%|██▍       | 2600/10708 [57:57<1:06:26,  2.03it/s]
                                                      
{'loss': 3.9088, 'grad_norm': 0.23119039833545685, 'learning_rate': 0.0009391639822935095, 'epoch': 0.24}

 24%|██▍       | 2600/10708 [57:57<1:06:26,  2.03it/s]
 24%|██▍       | 2601/10708 [57:58<1:06:29,  2.03it/s]
 24%|██▍       | 2602/10708 [57:58<1:06:26,  2.03it/s]
 24%|██▍       | 2603/10708 [57:59<1:06:33,  2.03it/s]
 24%|██▍       | 2604/10708 [57:59<1:06:33,  2.03it/s]
 24%|██▍       | 2605/10708 [58:00<1:06:35,  2.03it/s]
 24%|██▍       | 2606/10708 [58:00<1:06:35,  2.03it/s]
 24%|██▍       | 2607/10708 [58:01<1:06:38,  2.03it/s]
 24%|██▍       | 2608/10708 [58:01<1:06:34,  2.03it/s]
 24%|██▍       | 2609/10708 [58:02<1:06:34,  2.03it/s]
 24%|██▍       | 2610/10708 [58:02<1:06:27,  2.03it/s]
 24%|██▍       | 2611/10708 [58:03<1:06:27,  2.03it/s]
 24%|██▍       | 2612/10708 [58:03<1:06:32,  2.03it/s]
 24%|██▍       | 2613/10708 [58:04<1:06:26,  2.03it/s]
 24%|██▍       | 2614/10708 [58:04<1:06:26,  2.03it/s]
 24%|██▍       | 2615/10708 [58:05<1:06:26,  2.03it/s]
 24%|██▍       | 2616/10708 [58:05<1:06:34,  2.03it/s]
 24%|██▍       | 2617/10708 [58:06<1:06:29,  2.03it/s]
 24%|██▍       | 2618/10708 [58:06<1:06:30,  2.03it/s]
 24%|██▍       | 2619/10708 [58:07<1:06:25,  2.03it/s]
 24%|██▍       | 2620/10708 [58:07<1:06:25,  2.03it/s]
 24%|██▍       | 2621/10708 [58:08<1:06:25,  2.03it/s]
 24%|██▍       | 2622/10708 [58:08<1:06:21,  2.03it/s]
 24%|██▍       | 2623/10708 [58:09<1:06:21,  2.03it/s]
 25%|██▍       | 2624/10708 [58:09<1:06:20,  2.03it/s]
 25%|██▍       | 2625/10708 [58:10<1:06:17,  2.03it/s]{'loss': 3.9025, 'grad_norm': 0.22815874218940735, 'learning_rate': 0.0009372013737671314, 'epoch': 0.25}
                                                      

 25%|██▍       | 2625/10708 [58:10<1:06:17,  2.03it/s]
 25%|██▍       | 2626/10708 [58:10<1:06:22,  2.03it/s]
 25%|██▍       | 2627/10708 [58:11<1:06:16,  2.03it/s]
 25%|██▍       | 2628/10708 [58:11<1:06:21,  2.03it/s]
 25%|██▍       | 2629/10708 [58:12<1:06:21,  2.03it/s]
 25%|██▍       | 2630/10708 [58:12<1:06:16,  2.03it/s]
 25%|██▍       | 2631/10708 [58:13<1:06:21,  2.03it/s]
 25%|██▍       | 2632/10708 [58:13<1:06:28,  2.02it/s]
 25%|██▍       | 2633/10708 [58:14<1:06:25,  2.03it/s]
 25%|██▍       | 2634/10708 [58:14<1:06:26,  2.03it/s]
 25%|██▍       | 2635/10708 [58:15<1:06:26,  2.03it/s]
 25%|██▍       | 2636/10708 [58:15<1:06:28,  2.02it/s]
 25%|██▍       | 2637/10708 [58:16<1:06:23,  2.03it/s]
 25%|██▍       | 2638/10708 [58:16<1:06:19,  2.03it/s]
 25%|██▍       | 2639/10708 [58:17<1:06:13,  2.03it/s]
 25%|██▍       | 2640/10708 [58:17<1:06:14,  2.03it/s]
 25%|██▍       | 2641/10708 [58:18<1:06:09,  2.03it/s]
 25%|██▍       | 2642/10708 [58:18<1:06:10,  2.03it/s]
 25%|██▍       | 2643/10708 [58:19<1:06:11,  2.03it/s]
 25%|██▍       | 2644/10708 [58:19<1:06:06,  2.03it/s]
 25%|██▍       | 2645/10708 [58:20<1:06:08,  2.03it/s]
 25%|██▍       | 2646/10708 [58:20<1:06:08,  2.03it/s]
 25%|██▍       | 2647/10708 [58:21<1:06:07,  2.03it/s]
 25%|██▍       | 2648/10708 [58:21<1:06:10,  2.03it/s]
 25%|██▍       | 2649/10708 [58:22<1:06:04,  2.03it/s]
 25%|██▍       | 2650/10708 [58:22<1:06:02,  2.03it/s]{'loss': 3.8991, 'grad_norm': 0.22534961998462677, 'learning_rate': 0.000935209726675228, 'epoch': 0.25}

                                                      

 25%|██▍       | 2650/10708 [58:22<1:06:02,  2.03it/s]
 25%|██▍       | 2651/10708 [58:23<1:06:09,  2.03it/s]
 25%|██▍       | 2652/10708 [58:23<1:06:05,  2.03it/s]
 25%|██▍       | 2653/10708 [58:24<1:06:13,  2.03it/s]
 25%|██▍       | 2654/10708 [58:24<1:06:11,  2.03it/s]
 25%|██▍       | 2655/10708 [58:25<1:06:05,  2.03it/s]
 25%|██▍       | 2656/10708 [58:25<1:06:11,  2.03it/s]
 25%|██▍       | 2657/10708 [58:26<1:06:09,  2.03it/s]
 25%|██▍       | 2658/10708 [58:26<1:06:13,  2.03it/s]
 25%|██▍       | 2659/10708 [58:27<1:06:13,  2.03it/s]
 25%|██▍       | 2660/10708 [58:27<1:06:17,  2.02it/s]
 25%|██▍       | 2661/10708 [58:28<1:06:29,  2.02it/s]
 25%|██▍       | 2662/10708 [58:28<1:06:19,  2.02it/s]
 25%|██▍       | 2663/10708 [58:28<1:06:18,  2.02it/s]
 25%|██▍       | 2664/10708 [58:29<1:06:18,  2.02it/s]
 25%|██▍       | 2665/10708 [58:29<1:06:07,  2.03it/s]
 25%|██▍       | 2666/10708 [58:30<1:06:02,  2.03it/s]
 25%|██▍       | 2667/10708 [58:30<1:06:04,  2.03it/s]
 25%|██▍       | 2668/10708 [58:31<1:05:57,  2.03it/s]
 25%|██▍       | 2669/10708 [58:31<1:06:00,  2.03it/s]
 25%|██▍       | 2670/10708 [58:32<1:05:53,  2.03it/s]
 25%|██▍       | 2671/10708 [58:32<1:05:48,  2.04it/s]
 25%|██▍       | 2672/10708 [58:33<1:05:48,  2.04it/s]
 25%|██▍       | 2673/10708 [58:33<1:05:55,  2.03it/s]
 25%|██▍       | 2674/10708 [58:34<1:05:49,  2.03it/s]
 25%|██▍       | 2675/10708 [58:34<1:05:52,  2.03it/s]
                                                      
{'loss': 3.8885, 'grad_norm': 0.2086808830499649, 'learning_rate': 0.0009331891733014053, 'epoch': 0.25}

 25%|██▍       | 2675/10708 [58:34<1:05:52,  2.03it/s]
 25%|██▍       | 2676/10708 [58:35<1:05:59,  2.03it/s]
 25%|██▌       | 2677/10708 [58:35<1:06:00,  2.03it/s]
 25%|██▌       | 2678/10708 [58:36<1:05:57,  2.03it/s]
 25%|██▌       | 2679/10708 [58:36<1:05:50,  2.03it/s]
 25%|██▌       | 2680/10708 [58:37<1:05:54,  2.03it/s]
 25%|██▌       | 2681/10708 [58:37<1:05:47,  2.03it/s]
 25%|██▌       | 2682/10708 [58:38<1:05:51,  2.03it/s]
 25%|██▌       | 2683/10708 [58:38<1:05:58,  2.03it/s]
 25%|██▌       | 2684/10708 [58:39<1:05:53,  2.03it/s]
 25%|██▌       | 2685/10708 [58:39<1:05:51,  2.03it/s]
 25%|██▌       | 2686/10708 [58:40<1:05:54,  2.03it/s]
 25%|██▌       | 2687/10708 [58:40<1:05:47,  2.03it/s]
 25%|██▌       | 2688/10708 [58:41<1:05:46,  2.03it/s]
 25%|██▌       | 2689/10708 [58:41<1:05:46,  2.03it/s]
 25%|██▌       | 2690/10708 [58:42<1:05:41,  2.03it/s]
 25%|██▌       | 2691/10708 [58:42<1:05:42,  2.03it/s]
 25%|██▌       | 2692/10708 [58:43<1:05:49,  2.03it/s]
 25%|██▌       | 2693/10708 [58:43<1:05:52,  2.03it/s]
 25%|██▌       | 2694/10708 [58:44<1:05:52,  2.03it/s]
 25%|██▌       | 2695/10708 [58:44<1:05:44,  2.03it/s]
 25%|██▌       | 2696/10708 [58:45<1:05:45,  2.03it/s]
 25%|██▌       | 2697/10708 [58:45<1:05:46,  2.03it/s]
 25%|██▌       | 2698/10708 [58:46<1:05:49,  2.03it/s]
 25%|██▌       | 2699/10708 [58:46<1:05:51,  2.03it/s]
 25%|██▌       | 2700/10708 [58:47<1:05:43,  2.03it/s]{'loss': 3.8907, 'grad_norm': 0.2118133306503296, 'learning_rate': 0.0009311398478492026, 'epoch': 0.25}

                                                      

 25%|██▌       | 2700/10708 [58:47<1:05:43,  2.03it/s]
 25%|██▌       | 2701/10708 [58:47<1:05:48,  2.03it/s]
 25%|██▌       | 2702/10708 [58:48<1:05:50,  2.03it/s]
 25%|██▌       | 2703/10708 [58:48<1:05:51,  2.03it/s]
 25%|██▌       | 2704/10708 [58:49<1:05:57,  2.02it/s]
 25%|██▌       | 2705/10708 [58:49<1:05:55,  2.02it/s]
 25%|██▌       | 2706/10708 [58:50<1:05:50,  2.03it/s]
 25%|██▌       | 2707/10708 [58:50<1:05:47,  2.03it/s]
 25%|██▌       | 2708/10708 [58:51<1:05:46,  2.03it/s]
 25%|██▌       | 2709/10708 [58:51<1:05:40,  2.03it/s]
 25%|██▌       | 2710/10708 [58:52<1:05:39,  2.03it/s]
 25%|██▌       | 2711/10708 [58:52<1:05:38,  2.03it/s]
 25%|██▌       | 2712/10708 [58:53<1:05:43,  2.03it/s]
 25%|██▌       | 2713/10708 [58:53<1:05:45,  2.03it/s]
 25%|██▌       | 2714/10708 [58:54<1:05:45,  2.03it/s]
 25%|██▌       | 2715/10708 [58:54<1:05:41,  2.03it/s]
 25%|██▌       | 2716/10708 [58:55<1:05:43,  2.03it/s]
 25%|██▌       | 2717/10708 [58:55<1:05:38,  2.03it/s]
 25%|██▌       | 2718/10708 [58:56<1:05:40,  2.03it/s]
 25%|██▌       | 2719/10708 [58:56<1:05:40,  2.03it/s]
 25%|██▌       | 2720/10708 [58:57<1:05:36,  2.03it/s]
 25%|██▌       | 2721/10708 [58:57<1:05:38,  2.03it/s]
 25%|██▌       | 2722/10708 [58:58<1:05:34,  2.03it/s]
 25%|██▌       | 2723/10708 [58:58<1:05:34,  2.03it/s]
 25%|██▌       | 2724/10708 [58:59<1:05:35,  2.03it/s]
 25%|██▌       | 2725/10708 [58:59<1:05:30,  2.03it/s]{'loss': 3.8804, 'grad_norm': 0.2219672054052353, 'learning_rate': 0.0009290618864331769, 'epoch': 0.25}

                                                      

 25%|██▌       | 2725/10708 [58:59<1:05:30,  2.03it/s]
 25%|██▌       | 2726/10708 [59:00<1:05:40,  2.03it/s]
 25%|██▌       | 2727/10708 [59:00<1:05:32,  2.03it/s]
 25%|██▌       | 2728/10708 [59:01<1:05:30,  2.03it/s]
 25%|██▌       | 2729/10708 [59:01<1:05:29,  2.03it/s]
 25%|██▌       | 2730/10708 [59:02<1:05:25,  2.03it/s]
 26%|██▌       | 2731/10708 [59:02<1:05:26,  2.03it/s]
 26%|██▌       | 2732/10708 [59:02<1:05:27,  2.03it/s]
 26%|██▌       | 2733/10708 [59:03<1:05:31,  2.03it/s]
 26%|██▌       | 2734/10708 [59:03<1:05:46,  2.02it/s]
 26%|██▌       | 2735/10708 [59:04<1:05:40,  2.02it/s]
 26%|██▌       | 2736/10708 [59:04<1:05:35,  2.03it/s]
 26%|██▌       | 2737/10708 [59:05<1:05:35,  2.03it/s]
 26%|██▌       | 2738/10708 [59:05<1:05:30,  2.03it/s]
 26%|██▌       | 2739/10708 [59:06<1:06:13,  2.01it/s]
 26%|██▌       | 2740/10708 [59:07<1:08:45,  1.93it/s]
 26%|██▌       | 2741/10708 [59:07<1:08:57,  1.93it/s]
 26%|██▌       | 2742/10708 [59:08<1:07:47,  1.96it/s]
 26%|██▌       | 2743/10708 [59:08<1:07:04,  1.98it/s]
 26%|██▌       | 2744/10708 [59:09<1:06:37,  1.99it/s]
 26%|██▌       | 2745/10708 [59:09<1:06:10,  2.01it/s]
 26%|██▌       | 2746/10708 [59:10<1:05:56,  2.01it/s]
 26%|██▌       | 2747/10708 [59:10<1:07:46,  1.96it/s]
 26%|██▌       | 2748/10708 [59:11<1:06:58,  1.98it/s]
 26%|██▌       | 2749/10708 [59:11<1:06:21,  2.00it/s]
 26%|██▌       | 2750/10708 [59:12<1:06:10,  2.00it/s]
                                                      {'loss': 3.885, 'grad_norm': 0.22504831850528717, 'learning_rate': 0.0009269554270698634, 'epoch': 0.26}


 26%|██▌       | 2750/10708 [59:12<1:06:10,  2.00it/s]
 26%|██▌       | 2751/10708 [59:12<1:07:57,  1.95it/s]
 26%|██▌       | 2752/10708 [59:13<1:07:05,  1.98it/s]
 26%|██▌       | 2753/10708 [59:13<1:06:46,  1.99it/s]
 26%|██▌       | 2754/10708 [59:14<1:06:35,  1.99it/s]
 26%|██▌       | 2755/10708 [59:14<1:06:20,  2.00it/s]
 26%|██▌       | 2756/10708 [59:15<1:05:56,  2.01it/s]
 26%|██▌       | 2757/10708 [59:15<1:05:44,  2.02it/s]
 26%|██▌       | 2758/10708 [59:16<1:05:35,  2.02it/s]
 26%|██▌       | 2759/10708 [59:16<1:05:31,  2.02it/s]
 26%|██▌       | 2760/10708 [59:17<1:06:20,  2.00it/s]
 26%|██▌       | 2761/10708 [59:17<1:06:00,  2.01it/s]
 26%|██▌       | 2762/10708 [59:18<1:06:41,  1.99it/s]
 26%|██▌       | 2763/10708 [59:18<1:07:25,  1.96it/s]
 26%|██▌       | 2764/10708 [59:19<1:09:13,  1.91it/s]
 26%|██▌       | 2765/10708 [59:19<1:07:59,  1.95it/s]
 26%|██▌       | 2766/10708 [59:20<1:07:02,  1.97it/s]
 26%|██▌       | 2767/10708 [59:20<1:06:28,  1.99it/s]
 26%|██▌       | 2768/10708 [59:21<1:07:00,  1.97it/s]
 26%|██▌       | 2769/10708 [59:21<1:08:15,  1.94it/s]
 26%|██▌       | 2770/10708 [59:22<1:07:17,  1.97it/s]
 26%|██▌       | 2771/10708 [59:22<1:07:39,  1.95it/s]
 26%|██▌       | 2772/10708 [59:23<1:06:53,  1.98it/s]
 26%|██▌       | 2773/10708 [59:23<1:06:21,  1.99it/s]
 26%|██▌       | 2774/10708 [59:24<1:09:50,  1.89it/s]
 26%|██▌       | 2775/10708 [59:24<1:08:22,  1.93it/s]
                                                      
{'loss': 3.873, 'grad_norm': 0.19740962982177734, 'learning_rate': 0.0009248206096686082, 'epoch': 0.26}

 26%|██▌       | 2775/10708 [59:24<1:08:22,  1.93it/s]
 26%|██▌       | 2776/10708 [59:25<1:07:31,  1.96it/s]
 26%|██▌       | 2777/10708 [59:25<1:06:49,  1.98it/s]
 26%|██▌       | 2778/10708 [59:26<1:07:15,  1.97it/s]
 26%|██▌       | 2779/10708 [59:26<1:06:33,  1.99it/s]
 26%|██▌       | 2780/10708 [59:27<1:06:03,  2.00it/s]
 26%|██▌       | 2781/10708 [59:27<1:05:45,  2.01it/s]
 26%|██▌       | 2782/10708 [59:28<1:05:32,  2.02it/s]
 26%|██▌       | 2783/10708 [59:28<1:05:25,  2.02it/s]
 26%|██▌       | 2784/10708 [59:29<1:07:24,  1.96it/s]
 26%|██▌       | 2785/10708 [59:29<1:06:39,  1.98it/s]
 26%|██▌       | 2786/10708 [59:30<1:06:10,  2.00it/s]
 26%|██▌       | 2787/10708 [59:30<1:05:47,  2.01it/s]
 26%|██▌       | 2788/10708 [59:31<1:08:31,  1.93it/s]
 26%|██▌       | 2789/10708 [59:31<1:07:24,  1.96it/s]
 26%|██▌       | 2790/10708 [59:32<1:08:43,  1.92it/s]
 26%|██▌       | 2791/10708 [59:32<1:07:34,  1.95it/s]
 26%|██▌       | 2792/10708 [59:33<1:06:47,  1.98it/s]
 26%|██▌       | 2793/10708 [59:33<1:06:13,  1.99it/s]
 26%|██▌       | 2794/10708 [59:34<1:05:50,  2.00it/s]
 26%|██▌       | 2795/10708 [59:34<1:05:33,  2.01it/s]
 26%|██▌       | 2796/10708 [59:35<1:06:24,  1.99it/s]
 26%|██▌       | 2797/10708 [59:35<1:05:54,  2.00it/s]
 26%|██▌       | 2798/10708 [59:36<1:05:34,  2.01it/s]
 26%|██▌       | 2799/10708 [59:36<1:06:21,  1.99it/s]
 26%|██▌       | 2800/10708 [59:37<1:08:45,  1.92it/s]
                                                      
{'loss': 3.8832, 'grad_norm': 0.2392578274011612, 'learning_rate': 0.000922657576022275, 'epoch': 0.26}

 26%|██▌       | 2800/10708 [59:37<1:08:45,  1.92it/s]
 26%|██▌       | 2801/10708 [59:37<1:07:35,  1.95it/s]
 26%|██▌       | 2802/10708 [59:38<1:06:46,  1.97it/s]
 26%|██▌       | 2803/10708 [59:38<1:07:18,  1.96it/s]
 26%|██▌       | 2804/10708 [59:39<1:07:29,  1.95it/s]
 26%|██▌       | 2805/10708 [59:39<1:07:42,  1.95it/s]
 26%|██▌       | 2806/10708 [59:40<1:10:05,  1.88it/s]
 26%|██▌       | 2807/10708 [59:41<1:09:23,  1.90it/s]
 26%|██▌       | 2808/10708 [59:41<1:10:07,  1.88it/s]
 26%|██▌       | 2809/10708 [59:42<1:10:43,  1.86it/s]
 26%|██▌       | 2810/10708 [59:42<1:09:56,  1.88it/s]
 26%|██▋       | 2811/10708 [59:43<1:08:30,  1.92it/s]
 26%|██▋       | 2812/10708 [59:43<1:08:34,  1.92it/s]
 26%|██▋       | 2813/10708 [59:44<1:07:26,  1.95it/s]
 26%|██▋       | 2814/10708 [59:44<1:06:35,  1.98it/s]
 26%|██▋       | 2815/10708 [59:45<1:06:00,  1.99it/s]
 26%|██▋       | 2816/10708 [59:45<1:05:36,  2.00it/s]
 26%|██▋       | 2817/10708 [59:46<1:05:23,  2.01it/s]
 26%|██▋       | 2818/10708 [59:46<1:05:17,  2.01it/s]
 26%|██▋       | 2819/10708 [59:47<1:05:06,  2.02it/s]
 26%|██▋       | 2820/10708 [59:47<1:05:07,  2.02it/s]
 26%|██▋       | 2821/10708 [59:48<1:05:05,  2.02it/s]
 26%|██▋       | 2822/10708 [59:48<1:05:02,  2.02it/s]
 26%|██▋       | 2823/10708 [59:49<1:04:54,  2.02it/s]
 26%|██▋       | 2824/10708 [59:49<1:04:54,  2.02it/s]
 26%|██▋       | 2825/10708 [59:50<1:04:52,  2.03it/s]{'loss': 3.8782, 'grad_norm': 0.23474058508872986, 'learning_rate': 0.0009204664697978289, 'epoch': 0.26}

                                                      

 26%|██▋       | 2825/10708 [59:50<1:04:52,  2.03it/s]
 26%|██▋       | 2826/10708 [59:50<1:04:50,  2.03it/s]
 26%|██▋       | 2827/10708 [59:51<1:04:48,  2.03it/s]
 26%|██▋       | 2828/10708 [59:51<1:04:50,  2.03it/s]
 26%|██▋       | 2829/10708 [59:52<1:04:44,  2.03it/s]
 26%|██▋       | 2830/10708 [59:52<1:04:45,  2.03it/s]
 26%|██▋       | 2831/10708 [59:52<1:04:39,  2.03it/s]
 26%|██▋       | 2832/10708 [59:53<1:04:38,  2.03it/s]
 26%|██▋       | 2833/10708 [59:53<1:04:40,  2.03it/s]
 26%|██▋       | 2834/10708 [59:54<1:04:42,  2.03it/s]
 26%|██▋       | 2835/10708 [59:54<1:04:40,  2.03it/s]
 26%|██▋       | 2836/10708 [59:55<1:04:35,  2.03it/s]
 26%|██▋       | 2837/10708 [59:55<1:04:36,  2.03it/s]
 27%|██▋       | 2838/10708 [59:56<1:04:34,  2.03it/s]
 27%|██▋       | 2839/10708 [59:56<1:04:31,  2.03it/s]
 27%|██▋       | 2840/10708 [59:57<1:04:33,  2.03it/s]
 27%|██▋       | 2841/10708 [59:57<1:04:31,  2.03it/s]
 27%|██▋       | 2842/10708 [59:58<1:04:35,  2.03it/s]
 27%|██▋       | 2843/10708 [59:58<1:04:34,  2.03it/s]
 27%|██▋       | 2844/10708 [59:59<1:04:31,  2.03it/s]
 27%|██▋       | 2845/10708 [59:59<1:04:34,  2.03it/s]
 27%|██▋       | 2846/10708 [1:00:00<1:04:29,  2.03it/s]
 27%|██▋       | 2847/10708 [1:00:00<1:04:28,  2.03it/s]
 27%|██▋       | 2848/10708 [1:00:01<1:04:34,  2.03it/s]
 27%|██▋       | 2849/10708 [1:00:01<1:04:31,  2.03it/s]
 27%|██▋       | 2850/10708 [1:00:02<1:04:33,  2.03it/s]{'loss': 3.8725, 'grad_norm': 0.20619118213653564, 'learning_rate': 0.0009182474365267923, 'epoch': 0.27}
                                                        

 27%|██▋       | 2850/10708 [1:00:02<1:04:33,  2.03it/s]
 27%|██▋       | 2851/10708 [1:00:02<1:04:35,  2.03it/s]
 27%|██▋       | 2852/10708 [1:00:03<1:04:33,  2.03it/s]
 27%|██▋       | 2853/10708 [1:00:03<1:04:33,  2.03it/s]
 27%|██▋       | 2854/10708 [1:00:04<1:04:38,  2.02it/s]
 27%|██▋       | 2855/10708 [1:00:04<1:04:39,  2.02it/s]
 27%|██▋       | 2856/10708 [1:00:05<1:04:33,  2.03it/s]
 27%|██▋       | 2857/10708 [1:00:05<1:04:31,  2.03it/s]
 27%|██▋       | 2858/10708 [1:00:06<1:04:29,  2.03it/s]
 27%|██▋       | 2859/10708 [1:00:06<1:04:29,  2.03it/s]
 27%|██▋       | 2860/10708 [1:00:07<1:04:28,  2.03it/s]
 27%|██▋       | 2861/10708 [1:00:07<1:04:23,  2.03it/s]
 27%|██▋       | 2862/10708 [1:00:08<1:04:25,  2.03it/s]
 27%|██▋       | 2863/10708 [1:00:08<1:04:36,  2.02it/s]
 27%|██▋       | 2864/10708 [1:00:09<1:04:30,  2.03it/s]
 27%|██▋       | 2865/10708 [1:00:09<1:04:27,  2.03it/s]
 27%|██▋       | 2866/10708 [1:00:10<1:04:25,  2.03it/s]
 27%|██▋       | 2867/10708 [1:00:10<1:04:22,  2.03it/s]
 27%|██▋       | 2868/10708 [1:00:11<1:04:21,  2.03it/s]
 27%|██▋       | 2869/10708 [1:00:11<1:04:23,  2.03it/s]
 27%|██▋       | 2870/10708 [1:00:12<1:04:22,  2.03it/s]
 27%|██▋       | 2871/10708 [1:00:12<1:04:24,  2.03it/s]
 27%|██▋       | 2872/10708 [1:00:13<1:04:29,  2.03it/s]
 27%|██▋       | 2873/10708 [1:00:13<1:04:33,  2.02it/s]
 27%|██▋       | 2874/10708 [1:00:14<1:04:28,  2.03it/s]
 27%|██▋       | 2875/10708 [1:00:14<1:04:30,  2.02it/s]{'loss': 3.8578, 'grad_norm': 0.2226843684911728, 'learning_rate': 0.00091600062359558, 'epoch': 0.27}

                                                        

 27%|██▋       | 2875/10708 [1:00:14<1:04:30,  2.02it/s]
 27%|██▋       | 2876/10708 [1:00:15<1:04:58,  2.01it/s]
 27%|██▋       | 2877/10708 [1:00:15<1:04:47,  2.01it/s]
 27%|██▋       | 2878/10708 [1:00:16<1:04:38,  2.02it/s]
 27%|██▋       | 2879/10708 [1:00:16<1:04:31,  2.02it/s]
 27%|██▋       | 2880/10708 [1:00:17<1:04:26,  2.02it/s]
 27%|██▋       | 2881/10708 [1:00:17<1:04:28,  2.02it/s]
 27%|██▋       | 2882/10708 [1:00:18<1:04:24,  2.03it/s]
 27%|██▋       | 2883/10708 [1:00:18<1:04:25,  2.02it/s]
 27%|██▋       | 2884/10708 [1:00:19<1:04:24,  2.02it/s]
 27%|██▋       | 2885/10708 [1:00:19<1:04:23,  2.02it/s]
 27%|██▋       | 2886/10708 [1:00:20<1:04:23,  2.02it/s]
 27%|██▋       | 2887/10708 [1:00:20<1:04:19,  2.03it/s]
 27%|██▋       | 2888/10708 [1:00:21<1:04:17,  2.03it/s]
 27%|██▋       | 2889/10708 [1:00:21<1:04:14,  2.03it/s]
 27%|██▋       | 2890/10708 [1:00:22<1:04:07,  2.03it/s]
 27%|██▋       | 2891/10708 [1:00:22<1:04:02,  2.03it/s]
 27%|██▋       | 2892/10708 [1:00:23<1:04:08,  2.03it/s]
 27%|██▋       | 2893/10708 [1:00:23<1:04:04,  2.03it/s]
 27%|██▋       | 2894/10708 [1:00:24<1:04:07,  2.03it/s]
 27%|██▋       | 2895/10708 [1:00:24<1:04:08,  2.03it/s]
 27%|██▋       | 2896/10708 [1:00:25<1:04:10,  2.03it/s]
 27%|██▋       | 2897/10708 [1:00:25<1:04:07,  2.03it/s]
 27%|██▋       | 2898/10708 [1:00:26<1:04:03,  2.03it/s]
 27%|██▋       | 2899/10708 [1:00:26<1:04:02,  2.03it/s]
 27%|██▋       | 2900/10708 [1:00:27<1:04:04,  2.03it/s]
                                                        
{'loss': 3.8648, 'grad_norm': 0.23569121956825256, 'learning_rate': 0.0009137261802357099, 'epoch': 0.27}

 27%|██▋       | 2900/10708 [1:00:27<1:04:04,  2.03it/s]
 27%|██▋       | 2901/10708 [1:00:27<1:04:12,  2.03it/s]
 27%|██▋       | 2902/10708 [1:00:28<1:04:13,  2.03it/s]
 27%|██▋       | 2903/10708 [1:00:28<1:04:11,  2.03it/s]
 27%|██▋       | 2904/10708 [1:00:28<1:04:08,  2.03it/s]
 27%|██▋       | 2905/10708 [1:00:29<1:04:02,  2.03it/s]
 27%|██▋       | 2906/10708 [1:00:29<1:03:56,  2.03it/s]
 27%|██▋       | 2907/10708 [1:00:30<1:03:58,  2.03it/s]
 27%|██▋       | 2908/10708 [1:00:30<1:03:53,  2.03it/s]
 27%|██▋       | 2909/10708 [1:00:31<1:04:04,  2.03it/s]
 27%|██▋       | 2910/10708 [1:00:31<1:04:03,  2.03it/s]
 27%|██▋       | 2911/10708 [1:00:32<1:04:08,  2.03it/s]
 27%|██▋       | 2912/10708 [1:00:32<1:04:09,  2.03it/s]
 27%|██▋       | 2913/10708 [1:00:33<1:04:05,  2.03it/s]
 27%|██▋       | 2914/10708 [1:00:33<1:04:01,  2.03it/s]
 27%|██▋       | 2915/10708 [1:00:34<1:03:56,  2.03it/s]
 27%|██▋       | 2916/10708 [1:00:34<1:04:01,  2.03it/s]
 27%|██▋       | 2917/10708 [1:00:35<1:04:06,  2.03it/s]
 27%|██▋       | 2918/10708 [1:00:35<1:03:58,  2.03it/s]
 27%|██▋       | 2919/10708 [1:00:36<1:04:05,  2.03it/s]
 27%|██▋       | 2920/10708 [1:00:36<1:03:53,  2.03it/s]
 27%|██▋       | 2921/10708 [1:00:37<1:03:57,  2.03it/s]
 27%|██▋       | 2922/10708 [1:00:37<1:03:57,  2.03it/s]
 27%|██▋       | 2923/10708 [1:00:38<1:03:52,  2.03it/s]
 27%|██▋       | 2924/10708 [1:00:38<1:03:52,  2.03it/s]
 27%|██▋       | 2925/10708 [1:00:39<1:03:49,  2.03it/s]
                                                        {'loss': 3.8624, 'grad_norm': 0.21057862043380737, 'learning_rate': 0.0009114242575138905, 'epoch': 0.27}


 27%|██▋       | 2925/10708 [1:00:39<1:03:49,  2.03it/s]
 27%|██▋       | 2926/10708 [1:00:39<1:03:50,  2.03it/s]
 27%|██▋       | 2927/10708 [1:00:40<1:03:49,  2.03it/s]
 27%|██▋       | 2928/10708 [1:00:40<1:03:47,  2.03it/s]
 27%|██▋       | 2929/10708 [1:00:41<1:03:47,  2.03it/s]
 27%|██▋       | 2930/10708 [1:00:41<1:03:46,  2.03it/s]
 27%|██▋       | 2931/10708 [1:00:42<1:03:44,  2.03it/s]
 27%|██▋       | 2932/10708 [1:00:42<1:03:46,  2.03it/s]
 27%|██▋       | 2933/10708 [1:00:43<1:03:54,  2.03it/s]
 27%|██▋       | 2934/10708 [1:00:43<1:04:06,  2.02it/s]
 27%|██▋       | 2935/10708 [1:00:44<1:04:00,  2.02it/s]
 27%|██▋       | 2936/10708 [1:00:44<1:03:58,  2.02it/s]
 27%|██▋       | 2937/10708 [1:00:45<1:03:52,  2.03it/s]
 27%|██▋       | 2938/10708 [1:00:45<1:03:51,  2.03it/s]
 27%|██▋       | 2939/10708 [1:00:46<1:12:04,  1.80it/s]
 27%|██▋       | 2940/10708 [1:00:46<1:09:32,  1.86it/s]
 27%|██▋       | 2941/10708 [1:00:47<1:07:45,  1.91it/s]
 27%|██▋       | 2942/10708 [1:00:47<1:06:39,  1.94it/s]
 27%|██▋       | 2943/10708 [1:00:48<1:05:49,  1.97it/s]
 27%|██▋       | 2944/10708 [1:00:48<1:05:20,  1.98it/s]
 28%|██▊       | 2945/10708 [1:00:49<1:04:52,  1.99it/s]
 28%|██▊       | 2946/10708 [1:00:49<1:04:27,  2.01it/s]
 28%|██▊       | 2947/10708 [1:00:50<1:04:08,  2.02it/s]
 28%|██▊       | 2948/10708 [1:00:50<1:03:58,  2.02it/s]
 28%|██▊       | 2949/10708 [1:00:51<1:03:48,  2.03it/s]
 28%|██▊       | 2950/10708 [1:00:51<1:03:41,  2.03it/s]{'loss': 3.8641, 'grad_norm': 0.2016034573316574, 'learning_rate': 0.0009090950083219877, 'epoch': 0.28}
                                                        

 28%|██▊       | 2950/10708 [1:00:51<1:03:41,  2.03it/s]
 28%|██▊       | 2951/10708 [1:00:52<1:03:46,  2.03it/s]
 28%|██▊       | 2952/10708 [1:00:52<1:03:44,  2.03it/s]
 28%|██▊       | 2953/10708 [1:00:53<1:03:46,  2.03it/s]
 28%|██▊       | 2954/10708 [1:00:53<1:03:45,  2.03it/s]
 28%|██▊       | 2955/10708 [1:00:54<1:03:51,  2.02it/s]
 28%|██▊       | 2956/10708 [1:00:54<1:03:50,  2.02it/s]
 28%|██▊       | 2957/10708 [1:00:55<1:03:54,  2.02it/s]
 28%|██▊       | 2958/10708 [1:00:55<1:03:53,  2.02it/s]
 28%|██▊       | 2959/10708 [1:00:56<1:03:49,  2.02it/s]
 28%|██▊       | 2960/10708 [1:00:56<1:03:47,  2.02it/s]
 28%|██▊       | 2961/10708 [1:00:57<1:03:39,  2.03it/s]
 28%|██▊       | 2962/10708 [1:00:57<1:03:45,  2.02it/s]
 28%|██▊       | 2963/10708 [1:00:58<1:03:38,  2.03it/s]
 28%|██▊       | 2964/10708 [1:00:58<1:03:37,  2.03it/s]
 28%|██▊       | 2965/10708 [1:00:59<1:03:37,  2.03it/s]
 28%|██▊       | 2966/10708 [1:00:59<1:03:29,  2.03it/s]
 28%|██▊       | 2967/10708 [1:01:00<1:03:28,  2.03it/s]
 28%|██▊       | 2968/10708 [1:01:00<1:03:31,  2.03it/s]
 28%|██▊       | 2969/10708 [1:01:01<1:03:31,  2.03it/s]
 28%|██▊       | 2970/10708 [1:01:01<1:03:34,  2.03it/s]
 28%|██▊       | 2971/10708 [1:01:02<1:03:41,  2.02it/s]
 28%|██▊       | 2972/10708 [1:01:02<1:03:38,  2.03it/s]
 28%|██▊       | 2973/10708 [1:01:03<1:03:34,  2.03it/s]
 28%|██▊       | 2974/10708 [1:01:03<1:03:31,  2.03it/s]
 28%|██▊       | 2975/10708 [1:01:04<1:03:32,  2.03it/s]
                                                        
{'loss': 3.8538, 'grad_norm': 0.20680154860019684, 'learning_rate': 0.0009067385873668695, 'epoch': 0.28}

 28%|██▊       | 2975/10708 [1:01:04<1:03:32,  2.03it/s]
 28%|██▊       | 2976/10708 [1:01:04<1:03:41,  2.02it/s]
 28%|██▊       | 2977/10708 [1:01:05<1:03:34,  2.03it/s]
 28%|██▊       | 2978/10708 [1:01:05<1:03:32,  2.03it/s]
 28%|██▊       | 2979/10708 [1:01:06<1:03:31,  2.03it/s]
 28%|██▊       | 2980/10708 [1:01:06<1:03:34,  2.03it/s]
 28%|██▊       | 2981/10708 [1:01:07<1:03:27,  2.03it/s]
 28%|██▊       | 2982/10708 [1:01:07<1:03:27,  2.03it/s]
 28%|██▊       | 2983/10708 [1:01:08<1:03:22,  2.03it/s]
 28%|██▊       | 2984/10708 [1:01:08<1:03:26,  2.03it/s]
 28%|██▊       | 2985/10708 [1:01:09<1:03:28,  2.03it/s]
 28%|██▊       | 2986/10708 [1:01:09<1:03:26,  2.03it/s]
 28%|██▊       | 2987/10708 [1:01:10<1:03:24,  2.03it/s]
 28%|██▊       | 2988/10708 [1:01:10<1:03:23,  2.03it/s]
 28%|██▊       | 2989/10708 [1:01:11<1:03:24,  2.03it/s]
 28%|██▊       | 2990/10708 [1:01:11<1:03:24,  2.03it/s]
 28%|██▊       | 2991/10708 [1:01:12<1:03:17,  2.03it/s]
 28%|██▊       | 2992/10708 [1:01:12<1:03:23,  2.03it/s]
 28%|██▊       | 2993/10708 [1:01:13<1:03:24,  2.03it/s]
 28%|██▊       | 2994/10708 [1:01:13<1:03:34,  2.02it/s]
 28%|██▊       | 2995/10708 [1:01:14<1:03:32,  2.02it/s]
 28%|██▊       | 2996/10708 [1:01:14<1:03:27,  2.03it/s]
 28%|██▊       | 2997/10708 [1:01:15<1:03:25,  2.03it/s]
 28%|██▊       | 2998/10708 [1:01:15<1:03:16,  2.03it/s]
 28%|██▊       | 2999/10708 [1:01:16<1:03:48,  2.01it/s]
 28%|██▊       | 3000/10708 [1:01:16<1:03:34,  2.02it/s]{'loss': 3.8552, 'grad_norm': 0.21460427343845367, 'learning_rate': 0.000904355151160131, 'epoch': 0.28}
                                                        

 28%|██▊       | 3000/10708 [1:01:16<1:03:34,  2.02it/s]
 28%|██▊       | 3001/10708 [1:01:17<1:03:33,  2.02it/s]
 28%|██▊       | 3002/10708 [1:01:17<1:03:26,  2.02it/s]
 28%|██▊       | 3003/10708 [1:01:18<1:04:33,  1.99it/s]
 28%|██▊       | 3004/10708 [1:01:18<1:04:49,  1.98it/s]
 28%|██▊       | 3005/10708 [1:01:19<1:04:25,  1.99it/s]
 28%|██▊       | 3006/10708 [1:01:19<1:05:05,  1.97it/s]
 28%|██▊       | 3007/10708 [1:01:20<1:04:25,  1.99it/s]
 28%|██▊       | 3008/10708 [1:01:20<1:04:31,  1.99it/s]
 28%|██▊       | 3009/10708 [1:01:21<1:04:07,  2.00it/s]
 28%|██▊       | 3010/10708 [1:01:21<1:04:31,  1.99it/s]
 28%|██▊       | 3011/10708 [1:01:22<1:04:05,  2.00it/s]
 28%|██▊       | 3012/10708 [1:01:22<1:03:48,  2.01it/s]
 28%|██▊       | 3013/10708 [1:01:23<1:03:40,  2.01it/s]
 28%|██▊       | 3014/10708 [1:01:23<1:04:02,  2.00it/s]
 28%|██▊       | 3015/10708 [1:01:24<1:03:39,  2.01it/s]
 28%|██▊       | 3016/10708 [1:01:24<1:05:50,  1.95it/s]
 28%|██▊       | 3017/10708 [1:01:25<1:04:59,  1.97it/s]
 28%|██▊       | 3018/10708 [1:01:25<1:04:20,  1.99it/s]
 28%|██▊       | 3019/10708 [1:01:26<1:04:54,  1.97it/s]
 28%|██▊       | 3020/10708 [1:01:26<1:05:24,  1.96it/s]
 28%|██▊       | 3021/10708 [1:01:27<1:05:34,  1.95it/s]
 28%|██▊       | 3022/10708 [1:01:27<1:08:02,  1.88it/s]
 28%|██▊       | 3023/10708 [1:01:28<1:08:37,  1.87it/s]
 28%|██▊       | 3024/10708 [1:01:28<1:06:53,  1.91it/s]
 28%|██▊       | 3025/10708 [1:01:29<1:06:53,  1.91it/s]
                                                        
{'loss': 3.8524, 'grad_norm': 0.20757624506950378, 'learning_rate': 0.0009019448580076985, 'epoch': 0.28}

 28%|██▊       | 3025/10708 [1:01:29<1:06:53,  1.91it/s]
 28%|██▊       | 3026/10708 [1:01:29<1:06:43,  1.92it/s]
 28%|██▊       | 3027/10708 [1:01:30<1:05:39,  1.95it/s]
 28%|██▊       | 3028/10708 [1:01:30<1:06:06,  1.94it/s]
 28%|██▊       | 3029/10708 [1:01:31<1:06:17,  1.93it/s]
 28%|██▊       | 3030/10708 [1:01:31<1:05:13,  1.96it/s]
 28%|██▊       | 3031/10708 [1:01:32<1:06:38,  1.92it/s]
 28%|██▊       | 3032/10708 [1:01:32<1:06:36,  1.92it/s]
 28%|██▊       | 3033/10708 [1:01:33<1:05:27,  1.95it/s]
 28%|██▊       | 3034/10708 [1:01:33<1:04:41,  1.98it/s]
 28%|██▊       | 3035/10708 [1:01:34<1:06:01,  1.94it/s]
 28%|██▊       | 3036/10708 [1:01:34<1:07:17,  1.90it/s]
 28%|██▊       | 3037/10708 [1:01:35<1:06:55,  1.91it/s]
 28%|██▊       | 3038/10708 [1:01:35<1:05:44,  1.94it/s]
 28%|██▊       | 3039/10708 [1:01:36<1:05:53,  1.94it/s]
 28%|██▊       | 3040/10708 [1:01:37<1:06:02,  1.93it/s]
 28%|██▊       | 3041/10708 [1:01:37<1:05:07,  1.96it/s]
 28%|██▊       | 3042/10708 [1:01:37<1:04:31,  1.98it/s]
 28%|██▊       | 3043/10708 [1:01:38<1:04:01,  2.00it/s]
 28%|██▊       | 3044/10708 [1:01:38<1:03:41,  2.01it/s]
 28%|██▊       | 3045/10708 [1:01:39<1:03:26,  2.01it/s]
 28%|██▊       | 3046/10708 [1:01:39<1:03:17,  2.02it/s]
 28%|██▊       | 3047/10708 [1:01:40<1:03:12,  2.02it/s]
 28%|██▊       | 3048/10708 [1:01:40<1:02:58,  2.03it/s]
 28%|██▊       | 3049/10708 [1:01:41<1:02:53,  2.03it/s]
 28%|██▊       | 3050/10708 [1:01:41<1:02:55,  2.03it/s]
                                                        
{'loss': 3.8427, 'grad_norm': 0.22704316675662994, 'learning_rate': 0.0008995078679993151, 'epoch': 0.28}

 28%|██▊       | 3050/10708 [1:01:41<1:02:55,  2.03it/s]
 28%|██▊       | 3051/10708 [1:01:42<1:03:21,  2.01it/s]
 29%|██▊       | 3052/10708 [1:01:42<1:04:11,  1.99it/s]
 29%|██▊       | 3053/10708 [1:01:43<1:03:51,  2.00it/s]
 29%|██▊       | 3054/10708 [1:01:43<1:03:38,  2.00it/s]
 29%|██▊       | 3055/10708 [1:01:44<1:03:25,  2.01it/s]
 29%|██▊       | 3056/10708 [1:01:44<1:04:17,  1.98it/s]
 29%|██▊       | 3057/10708 [1:01:45<1:03:49,  2.00it/s]
 29%|██▊       | 3058/10708 [1:01:45<1:03:28,  2.01it/s]
 29%|██▊       | 3059/10708 [1:01:46<1:03:14,  2.02it/s]
 29%|██▊       | 3060/10708 [1:01:46<1:03:05,  2.02it/s]
 29%|██▊       | 3061/10708 [1:01:47<1:03:04,  2.02it/s]
 29%|██▊       | 3062/10708 [1:01:47<1:03:58,  1.99it/s]
 29%|██▊       | 3063/10708 [1:01:48<1:04:40,  1.97it/s]
 29%|██▊       | 3064/10708 [1:01:48<1:05:05,  1.96it/s]
 29%|██▊       | 3065/10708 [1:01:49<1:04:24,  1.98it/s]
 29%|██▊       | 3066/10708 [1:01:49<1:03:53,  1.99it/s]
 29%|██▊       | 3067/10708 [1:01:50<1:04:32,  1.97it/s]
 29%|██▊       | 3068/10708 [1:01:51<1:04:57,  1.96it/s]
 29%|██▊       | 3069/10708 [1:01:51<1:04:15,  1.98it/s]
 29%|██▊       | 3070/10708 [1:01:51<1:03:52,  1.99it/s]
 29%|██▊       | 3071/10708 [1:01:52<1:04:34,  1.97it/s]
 29%|██▊       | 3072/10708 [1:01:53<1:04:02,  1.99it/s]
 29%|██▊       | 3073/10708 [1:01:53<1:04:40,  1.97it/s]
 29%|██▊       | 3074/10708 [1:01:54<1:04:01,  1.99it/s]
 29%|██▊       | 3075/10708 [1:01:54<1:04:37,  1.97it/s]
                                                        
{'loss': 3.8502, 'grad_norm': 0.211347758769989, 'learning_rate': 0.0008970443429979078, 'epoch': 0.29}

 29%|██▊       | 3075/10708 [1:01:54<1:04:37,  1.97it/s]
 29%|██▊       | 3076/10708 [1:01:55<1:04:05,  1.98it/s]
 29%|██▊       | 3077/10708 [1:01:55<1:03:42,  2.00it/s]
 29%|██▊       | 3078/10708 [1:01:56<1:03:20,  2.01it/s]
 29%|██▉       | 3079/10708 [1:01:56<1:04:02,  1.99it/s]
 29%|██▉       | 3080/10708 [1:01:57<1:03:36,  2.00it/s]
 29%|██▉       | 3081/10708 [1:01:57<1:03:14,  2.01it/s]
 29%|██▉       | 3082/10708 [1:01:58<1:07:21,  1.89it/s]
 29%|██▉       | 3083/10708 [1:01:58<1:05:50,  1.93it/s]
 29%|██▉       | 3084/10708 [1:01:59<1:04:51,  1.96it/s]
 29%|██▉       | 3085/10708 [1:01:59<1:04:05,  1.98it/s]
 29%|██▉       | 3086/10708 [1:02:00<1:03:40,  2.00it/s]
 29%|██▉       | 3087/10708 [1:02:00<1:03:22,  2.00it/s]
 29%|██▉       | 3088/10708 [1:02:01<1:03:01,  2.01it/s]
 29%|██▉       | 3089/10708 [1:02:01<1:03:00,  2.02it/s]
 29%|██▉       | 3090/10708 [1:02:02<1:02:50,  2.02it/s]
 29%|██▉       | 3091/10708 [1:02:02<1:02:38,  2.03it/s]
 29%|██▉       | 3092/10708 [1:02:03<1:02:38,  2.03it/s]
 29%|██▉       | 3093/10708 [1:02:03<1:02:35,  2.03it/s]
 29%|██▉       | 3094/10708 [1:02:04<1:02:36,  2.03it/s]
 29%|██▉       | 3095/10708 [1:02:04<1:02:34,  2.03it/s]
 29%|██▉       | 3096/10708 [1:02:05<1:02:30,  2.03it/s]
 29%|██▉       | 3097/10708 [1:02:05<1:02:32,  2.03it/s]
 29%|██▉       | 3098/10708 [1:02:06<1:02:26,  2.03it/s]
 29%|██▉       | 3099/10708 [1:02:06<1:03:23,  2.00it/s]
 29%|██▉       | 3100/10708 [1:02:07<1:03:02,  2.01it/s]
                                                        {'loss': 3.8417, 'grad_norm': 0.2125086933374405, 'learning_rate': 0.0008945544466288367, 'epoch': 0.29}

 29%|██▉       | 3100/10708 [1:02:07<1:03:02,  2.01it/s]
 29%|██▉       | 3101/10708 [1:02:07<1:03:58,  1.98it/s]
 29%|██▉       | 3102/10708 [1:02:08<1:04:45,  1.96it/s]
 29%|██▉       | 3103/10708 [1:02:08<1:04:06,  1.98it/s]
 29%|██▉       | 3104/10708 [1:02:09<1:03:35,  1.99it/s]
 29%|██▉       | 3105/10708 [1:02:09<1:03:08,  2.01it/s]
 29%|██▉       | 3106/10708 [1:02:10<1:02:53,  2.01it/s]
 29%|██▉       | 3107/10708 [1:02:10<1:02:46,  2.02it/s]
 29%|██▉       | 3108/10708 [1:02:11<1:02:39,  2.02it/s]
 29%|██▉       | 3109/10708 [1:02:11<1:02:38,  2.02it/s]
 29%|██▉       | 3110/10708 [1:02:11<1:02:29,  2.03it/s]
 29%|██▉       | 3111/10708 [1:02:12<1:02:27,  2.03it/s]
 29%|██▉       | 3112/10708 [1:02:12<1:02:34,  2.02it/s]
 29%|██▉       | 3113/10708 [1:02:13<1:02:39,  2.02it/s]
 29%|██▉       | 3114/10708 [1:02:14<1:03:34,  1.99it/s]
 29%|██▉       | 3115/10708 [1:02:14<1:06:10,  1.91it/s]
 29%|██▉       | 3116/10708 [1:02:15<1:04:58,  1.95it/s]
 29%|██▉       | 3117/10708 [1:02:15<1:04:09,  1.97it/s]
 29%|██▉       | 3118/10708 [1:02:16<1:03:37,  1.99it/s]
 29%|██▉       | 3119/10708 [1:02:16<1:04:04,  1.97it/s]
 29%|██▉       | 3120/10708 [1:02:17<1:06:34,  1.90it/s]
 29%|██▉       | 3121/10708 [1:02:17<1:05:16,  1.94it/s]
 29%|██▉       | 3122/10708 [1:02:18<1:05:23,  1.93it/s]
 29%|██▉       | 3123/10708 [1:02:18<1:04:24,  1.96it/s]
 29%|██▉       | 3124/10708 [1:02:19<1:03:49,  1.98it/s]
 29%|██▉       | 3125/10708 [1:02:19<1:03:19,  2.00it/s]
                                                        {'loss': 3.8373, 'grad_norm': 0.21936264634132385, 'learning_rate': 0.0008920383442690269, 'epoch': 0.29}


 29%|██▉       | 3125/10708 [1:02:19<1:03:19,  2.00it/s]
 29%|██▉       | 3126/10708 [1:02:20<1:03:04,  2.00it/s]
 29%|██▉       | 3127/10708 [1:02:20<1:02:47,  2.01it/s]
 29%|██▉       | 3128/10708 [1:02:21<1:02:33,  2.02it/s]
 29%|██▉       | 3129/10708 [1:02:21<1:03:20,  1.99it/s]
 29%|██▉       | 3130/10708 [1:02:22<1:03:04,  2.00it/s]
 29%|██▉       | 3131/10708 [1:02:22<1:02:45,  2.01it/s]
 29%|██▉       | 3132/10708 [1:02:23<1:03:33,  1.99it/s]
 29%|██▉       | 3133/10708 [1:02:23<1:04:07,  1.97it/s]
 29%|██▉       | 3134/10708 [1:02:24<1:03:31,  1.99it/s]
 29%|██▉       | 3135/10708 [1:02:24<1:04:01,  1.97it/s]
 29%|██▉       | 3136/10708 [1:02:25<1:03:27,  1.99it/s]
 29%|██▉       | 3137/10708 [1:02:25<1:03:03,  2.00it/s]
 29%|██▉       | 3138/10708 [1:02:26<1:02:41,  2.01it/s]
 29%|██▉       | 3139/10708 [1:02:26<1:02:36,  2.02it/s]
 29%|██▉       | 3140/10708 [1:02:27<1:02:28,  2.02it/s]
 29%|██▉       | 3141/10708 [1:02:27<1:02:19,  2.02it/s]
 29%|██▉       | 3142/10708 [1:02:28<1:02:11,  2.03it/s]
 29%|██▉       | 3143/10708 [1:02:28<1:02:08,  2.03it/s]
 29%|██▉       | 3144/10708 [1:02:29<1:02:09,  2.03it/s]
 29%|██▉       | 3145/10708 [1:02:29<1:02:04,  2.03it/s]
 29%|██▉       | 3146/10708 [1:02:30<1:02:03,  2.03it/s]
 29%|██▉       | 3147/10708 [1:02:30<1:02:03,  2.03it/s]
 29%|██▉       | 3148/10708 [1:02:31<1:04:00,  1.97it/s]
 29%|██▉       | 3149/10708 [1:02:31<1:03:20,  1.99it/s]
 29%|██▉       | 3150/10708 [1:02:32<1:03:54,  1.97it/s]
                                                        {'loss': 3.8362, 'grad_norm': 0.23142272233963013, 'learning_rate': 0.0008894962030359843, 'epoch': 0.29}

 29%|██▉       | 3150/10708 [1:02:32<1:03:54,  1.97it/s]
 29%|██▉       | 3151/10708 [1:02:32<1:05:38,  1.92it/s]
 29%|██▉       | 3152/10708 [1:02:33<1:04:37,  1.95it/s]
 29%|██▉       | 3153/10708 [1:02:33<1:04:52,  1.94it/s]
 29%|██▉       | 3154/10708 [1:02:34<1:03:59,  1.97it/s]
 29%|██▉       | 3155/10708 [1:02:34<1:14:43,  1.68it/s]
 29%|██▉       | 3156/10708 [1:02:35<1:11:45,  1.75it/s]
 29%|██▉       | 3157/10708 [1:02:35<1:08:52,  1.83it/s]
 29%|██▉       | 3158/10708 [1:02:36<1:06:40,  1.89it/s]
 30%|██▉       | 3159/10708 [1:02:36<1:05:14,  1.93it/s]
 30%|██▉       | 3160/10708 [1:02:37<1:06:09,  1.90it/s]
 30%|██▉       | 3161/10708 [1:02:37<1:04:52,  1.94it/s]
 30%|██▉       | 3162/10708 [1:02:38<1:03:53,  1.97it/s]
 30%|██▉       | 3163/10708 [1:02:39<1:04:24,  1.95it/s]
 30%|██▉       | 3164/10708 [1:02:39<1:03:38,  1.98it/s]
 30%|██▉       | 3165/10708 [1:02:39<1:03:04,  1.99it/s]
 30%|██▉       | 3166/10708 [1:02:40<1:03:38,  1.98it/s]
 30%|██▉       | 3167/10708 [1:02:40<1:03:05,  1.99it/s]
 30%|██▉       | 3168/10708 [1:02:41<1:12:54,  1.72it/s]
 30%|██▉       | 3169/10708 [1:02:42<1:10:38,  1.78it/s]
 30%|██▉       | 3170/10708 [1:02:42<1:08:02,  1.85it/s]
 30%|██▉       | 3171/10708 [1:02:43<1:06:10,  1.90it/s]
 30%|██▉       | 3172/10708 [1:02:43<1:06:00,  1.90it/s]
 30%|██▉       | 3173/10708 [1:02:44<1:05:42,  1.91it/s]
 30%|██▉       | 3174/10708 [1:02:44<1:05:35,  1.91it/s]
 30%|██▉       | 3175/10708 [1:02:45<1:05:27,  1.92it/s]
                                                        
{'loss': 3.8316, 'grad_norm': 0.211162269115448, 'learning_rate': 0.0008869281917766966, 'epoch': 0.3}

 30%|██▉       | 3175/10708 [1:02:45<1:05:27,  1.92it/s]
 30%|██▉       | 3176/10708 [1:02:45<1:04:26,  1.95it/s]
 30%|██▉       | 3177/10708 [1:02:46<1:03:37,  1.97it/s]
 30%|██▉       | 3178/10708 [1:02:46<1:03:10,  1.99it/s]
 30%|██▉       | 3179/10708 [1:02:47<1:02:46,  2.00it/s]
 30%|██▉       | 3180/10708 [1:02:47<1:02:21,  2.01it/s]
 30%|██▉       | 3181/10708 [1:02:48<1:03:12,  1.98it/s]
 30%|██▉       | 3182/10708 [1:02:48<1:02:47,  2.00it/s]
 30%|██▉       | 3183/10708 [1:02:49<1:02:32,  2.01it/s]
 30%|██▉       | 3184/10708 [1:02:49<1:02:12,  2.02it/s]
 30%|██▉       | 3185/10708 [1:02:50<1:01:58,  2.02it/s]
 30%|██▉       | 3186/10708 [1:02:50<1:01:56,  2.02it/s]
 30%|██▉       | 3187/10708 [1:02:51<1:01:52,  2.03it/s]
 30%|██▉       | 3188/10708 [1:02:51<1:01:58,  2.02it/s]
 30%|██▉       | 3189/10708 [1:02:52<1:01:52,  2.03it/s]
 30%|██▉       | 3190/10708 [1:02:52<1:01:52,  2.02it/s]
 30%|██▉       | 3191/10708 [1:02:53<1:01:49,  2.03it/s]
 30%|██▉       | 3192/10708 [1:02:53<1:01:43,  2.03it/s]
 30%|██▉       | 3193/10708 [1:02:54<1:01:40,  2.03it/s]
 30%|██▉       | 3194/10708 [1:02:54<1:01:35,  2.03it/s]
 30%|██▉       | 3195/10708 [1:02:55<1:01:36,  2.03it/s]
 30%|██▉       | 3196/10708 [1:02:55<1:01:40,  2.03it/s]
 30%|██▉       | 3197/10708 [1:02:56<1:01:35,  2.03it/s]
 30%|██▉       | 3198/10708 [1:02:56<1:03:41,  1.97it/s]
 30%|██▉       | 3199/10708 [1:02:57<1:03:58,  1.96it/s]
 30%|██▉       | 3200/10708 [1:02:57<1:04:10,  1.95it/s]
                                                        
{'loss': 3.8293, 'grad_norm': 0.21196815371513367, 'learning_rate': 0.000884334481056417, 'epoch': 0.3}

 30%|██▉       | 3200/10708 [1:02:57<1:04:10,  1.95it/s]
 30%|██▉       | 3201/10708 [1:02:58<1:04:15,  1.95it/s]
 30%|██▉       | 3202/10708 [1:02:58<1:03:22,  1.97it/s]
 30%|██▉       | 3203/10708 [1:02:59<1:03:52,  1.96it/s]
 30%|██▉       | 3204/10708 [1:02:59<1:03:07,  1.98it/s]
 30%|██▉       | 3205/10708 [1:03:00<1:02:40,  2.00it/s]
 30%|██▉       | 3206/10708 [1:03:00<1:03:20,  1.97it/s]
 30%|██▉       | 3207/10708 [1:03:01<1:02:49,  1.99it/s]
 30%|██▉       | 3208/10708 [1:03:01<1:02:25,  2.00it/s]
 30%|██▉       | 3209/10708 [1:03:02<1:02:03,  2.01it/s]
 30%|██▉       | 3210/10708 [1:03:02<1:02:50,  1.99it/s]
 30%|██▉       | 3211/10708 [1:03:03<1:02:25,  2.00it/s]
 30%|██▉       | 3212/10708 [1:03:03<1:02:07,  2.01it/s]
 30%|███       | 3213/10708 [1:03:04<1:02:53,  1.99it/s]
 30%|███       | 3214/10708 [1:03:04<1:03:19,  1.97it/s]
 30%|███       | 3215/10708 [1:03:05<1:03:48,  1.96it/s]
 30%|███       | 3216/10708 [1:03:05<1:04:02,  1.95it/s]
 30%|███       | 3217/10708 [1:03:06<1:04:12,  1.94it/s]
 30%|███       | 3218/10708 [1:03:06<1:03:17,  1.97it/s]
 30%|███       | 3219/10708 [1:03:07<1:02:38,  1.99it/s]
 30%|███       | 3220/10708 [1:03:07<1:02:17,  2.00it/s]
 30%|███       | 3221/10708 [1:03:08<1:02:01,  2.01it/s]
 30%|███       | 3222/10708 [1:03:08<1:01:44,  2.02it/s]
 30%|███       | 3223/10708 [1:03:09<1:01:44,  2.02it/s]
 30%|███       | 3224/10708 [1:03:09<1:02:33,  1.99it/s]
 30%|███       | 3225/10708 [1:03:10<1:02:11,  2.01it/s]{'loss': 3.8297, 'grad_norm': 0.2390924096107483, 'learning_rate': 0.000881715243147337, 'epoch': 0.3}

                                                        

 30%|███       | 3225/10708 [1:03:10<1:02:11,  2.01it/s]
 30%|███       | 3226/10708 [1:03:10<1:02:01,  2.01it/s]
 30%|███       | 3227/10708 [1:03:11<1:02:38,  1.99it/s]
 30%|███       | 3228/10708 [1:03:11<1:03:11,  1.97it/s]
 30%|███       | 3229/10708 [1:03:12<1:03:35,  1.96it/s]
 30%|███       | 3230/10708 [1:03:12<1:04:55,  1.92it/s]
 30%|███       | 3231/10708 [1:03:13<1:04:53,  1.92it/s]
 30%|███       | 3232/10708 [1:03:13<1:04:45,  1.92it/s]
 30%|███       | 3233/10708 [1:03:14<1:03:47,  1.95it/s]
 30%|███       | 3234/10708 [1:03:15<1:05:04,  1.91it/s]
 30%|███       | 3235/10708 [1:03:15<1:03:53,  1.95it/s]
 30%|███       | 3236/10708 [1:03:15<1:03:10,  1.97it/s]
 30%|███       | 3237/10708 [1:03:16<1:03:24,  1.96it/s]
 30%|███       | 3238/10708 [1:03:17<1:02:45,  1.98it/s]
 30%|███       | 3239/10708 [1:03:17<1:03:12,  1.97it/s]
 30%|███       | 3240/10708 [1:03:18<1:02:32,  1.99it/s]
 30%|███       | 3241/10708 [1:03:18<1:04:11,  1.94it/s]
 30%|███       | 3242/10708 [1:03:19<1:05:13,  1.91it/s]
 30%|███       | 3243/10708 [1:03:19<1:03:59,  1.94it/s]
 30%|███       | 3244/10708 [1:03:20<1:03:06,  1.97it/s]
 30%|███       | 3245/10708 [1:03:20<1:02:29,  1.99it/s]
 30%|███       | 3246/10708 [1:03:21<1:03:09,  1.97it/s]
 30%|███       | 3247/10708 [1:03:21<1:02:42,  1.98it/s]
 30%|███       | 3248/10708 [1:03:22<1:04:22,  1.93it/s]
 30%|███       | 3249/10708 [1:03:22<1:03:24,  1.96it/s]
 30%|███       | 3250/10708 [1:03:23<1:02:46,  1.98it/s]
                                                        {'loss': 3.8338, 'grad_norm': 0.20314745604991913, 'learning_rate': 0.0008790706520171435, 'epoch': 0.3}

 30%|███       | 3250/10708 [1:03:23<1:02:46,  1.98it/s]
 30%|███       | 3251/10708 [1:03:23<1:02:22,  1.99it/s]
 30%|███       | 3252/10708 [1:03:24<1:02:01,  2.00it/s]
 30%|███       | 3253/10708 [1:03:24<1:01:48,  2.01it/s]
 30%|███       | 3254/10708 [1:03:25<1:01:35,  2.02it/s]
 30%|███       | 3255/10708 [1:03:25<1:01:30,  2.02it/s]
 30%|███       | 3256/10708 [1:03:26<1:01:21,  2.02it/s]
 30%|███       | 3257/10708 [1:03:26<1:01:14,  2.03it/s]
 30%|███       | 3258/10708 [1:03:27<1:01:09,  2.03it/s]
 30%|███       | 3259/10708 [1:03:27<1:02:08,  2.00it/s]
 30%|███       | 3260/10708 [1:03:28<1:02:49,  1.98it/s]
 30%|███       | 3261/10708 [1:03:28<1:02:15,  1.99it/s]
 30%|███       | 3262/10708 [1:03:29<1:01:55,  2.00it/s]
 30%|███       | 3263/10708 [1:03:29<1:01:45,  2.01it/s]
 30%|███       | 3264/10708 [1:03:30<1:01:30,  2.02it/s]
 30%|███       | 3265/10708 [1:03:30<1:02:19,  1.99it/s]
 31%|███       | 3266/10708 [1:03:31<1:02:47,  1.98it/s]
 31%|███       | 3267/10708 [1:03:31<1:02:11,  1.99it/s]
 31%|███       | 3268/10708 [1:03:32<1:04:46,  1.91it/s]
 31%|███       | 3269/10708 [1:03:32<1:03:34,  1.95it/s]
 31%|███       | 3270/10708 [1:03:33<1:02:46,  1.97it/s]
 31%|███       | 3271/10708 [1:03:33<1:02:19,  1.99it/s]
 31%|███       | 3272/10708 [1:03:34<1:01:55,  2.00it/s]
 31%|███       | 3273/10708 [1:03:34<1:01:41,  2.01it/s]
 31%|███       | 3274/10708 [1:03:35<1:01:25,  2.02it/s]
 31%|███       | 3275/10708 [1:03:35<1:01:13,  2.02it/s]
                                                        {'loss': 3.8274, 'grad_norm': 0.20323984324932098, 'learning_rate': 0.0008764008833174637, 'epoch': 0.31}


 31%|███       | 3275/10708 [1:03:35<1:01:13,  2.02it/s]
 31%|███       | 3276/10708 [1:03:36<1:01:14,  2.02it/s]
 31%|███       | 3277/10708 [1:03:36<1:01:06,  2.03it/s]
 31%|███       | 3278/10708 [1:03:37<1:01:05,  2.03it/s]
 31%|███       | 3279/10708 [1:03:37<1:01:58,  2.00it/s]
 31%|███       | 3280/10708 [1:03:38<1:02:38,  1.98it/s]
 31%|███       | 3281/10708 [1:03:38<1:02:08,  1.99it/s]
 31%|███       | 3282/10708 [1:03:39<1:01:47,  2.00it/s]
 31%|███       | 3283/10708 [1:03:39<1:01:28,  2.01it/s]
 31%|███       | 3284/10708 [1:03:40<1:01:18,  2.02it/s]
 31%|███       | 3285/10708 [1:03:40<1:01:08,  2.02it/s]
 31%|███       | 3286/10708 [1:03:41<1:01:00,  2.03it/s]
 31%|███       | 3287/10708 [1:03:41<1:00:56,  2.03it/s]
 31%|███       | 3288/10708 [1:03:42<1:00:54,  2.03it/s]
 31%|███       | 3289/10708 [1:03:42<1:05:02,  1.90it/s]
 31%|███       | 3290/10708 [1:03:43<1:03:50,  1.94it/s]
 31%|███       | 3291/10708 [1:03:43<1:04:59,  1.90it/s]
 31%|███       | 3292/10708 [1:03:44<1:04:43,  1.91it/s]
 31%|███       | 3293/10708 [1:03:44<1:03:31,  1.95it/s]
 31%|███       | 3294/10708 [1:03:45<1:02:40,  1.97it/s]
 31%|███       | 3295/10708 [1:03:45<1:02:06,  1.99it/s]
 31%|███       | 3296/10708 [1:03:46<1:01:41,  2.00it/s]
 31%|███       | 3297/10708 [1:03:46<1:01:24,  2.01it/s]
 31%|███       | 3298/10708 [1:03:47<1:02:03,  1.99it/s]
 31%|███       | 3299/10708 [1:03:47<1:01:41,  2.00it/s]
 31%|███       | 3300/10708 [1:03:48<1:01:22,  2.01it/s]
                                                        {'loss': 3.8288, 'grad_norm': 0.2019776850938797, 'learning_rate': 0.0008737061143721993, 'epoch': 0.31}

 31%|███       | 3300/10708 [1:03:48<1:01:22,  2.01it/s]
 31%|███       | 3301/10708 [1:03:48<1:01:15,  2.02it/s]
 31%|███       | 3302/10708 [1:03:49<1:01:03,  2.02it/s]
 31%|███       | 3303/10708 [1:03:49<1:02:57,  1.96it/s]
 31%|███       | 3304/10708 [1:03:50<1:02:15,  1.98it/s]
 31%|███       | 3305/10708 [1:03:50<1:02:41,  1.97it/s]
 31%|███       | 3306/10708 [1:03:51<1:02:01,  1.99it/s]
 31%|███       | 3307/10708 [1:03:51<1:01:39,  2.00it/s]
 31%|███       | 3308/10708 [1:03:52<1:01:24,  2.01it/s]
 31%|███       | 3309/10708 [1:03:52<1:01:11,  2.02it/s]
 31%|███       | 3310/10708 [1:03:53<1:01:03,  2.02it/s]
 31%|███       | 3311/10708 [1:03:53<1:00:56,  2.02it/s]
 31%|███       | 3312/10708 [1:03:54<1:00:51,  2.03it/s]
 31%|███       | 3313/10708 [1:03:54<1:00:52,  2.02it/s]
 31%|███       | 3314/10708 [1:03:55<1:00:51,  2.03it/s]
 31%|███       | 3315/10708 [1:03:55<1:00:47,  2.03it/s]
 31%|███       | 3316/10708 [1:03:56<1:00:43,  2.03it/s]
 31%|███       | 3317/10708 [1:03:56<1:00:37,  2.03it/s]
 31%|███       | 3318/10708 [1:03:57<1:00:40,  2.03it/s]
 31%|███       | 3319/10708 [1:03:57<1:00:35,  2.03it/s]
 31%|███       | 3320/10708 [1:03:58<1:00:34,  2.03it/s]
 31%|███       | 3321/10708 [1:03:58<1:00:32,  2.03it/s]
 31%|███       | 3322/10708 [1:03:59<1:00:36,  2.03it/s]
 31%|███       | 3323/10708 [1:03:59<1:00:38,  2.03it/s]
 31%|███       | 3324/10708 [1:04:00<1:00:38,  2.03it/s]
 31%|███       | 3325/10708 [1:04:00<1:00:42,  2.03it/s]{'loss': 3.8213, 'grad_norm': 0.20708830654621124, 'learning_rate': 0.0008709865241657479, 'epoch': 0.31}

                                                        

 31%|███       | 3325/10708 [1:04:00<1:00:42,  2.03it/s]
 31%|███       | 3326/10708 [1:04:01<1:00:42,  2.03it/s]
 31%|███       | 3327/10708 [1:04:01<1:00:45,  2.02it/s]
 31%|███       | 3328/10708 [1:04:02<1:00:40,  2.03it/s]
 31%|███       | 3329/10708 [1:04:02<1:00:35,  2.03it/s]
 31%|███       | 3330/10708 [1:04:03<1:00:38,  2.03it/s]
 31%|███       | 3331/10708 [1:04:03<1:00:36,  2.03it/s]
 31%|███       | 3332/10708 [1:04:04<1:00:36,  2.03it/s]
 31%|███       | 3333/10708 [1:04:04<1:00:28,  2.03it/s]
 31%|███       | 3334/10708 [1:04:05<1:00:26,  2.03it/s]
 31%|███       | 3335/10708 [1:04:05<1:00:26,  2.03it/s]
 31%|███       | 3336/10708 [1:04:05<1:00:25,  2.03it/s]
 31%|███       | 3337/10708 [1:04:06<1:00:24,  2.03it/s]
 31%|███       | 3338/10708 [1:04:06<1:00:26,  2.03it/s]
 31%|███       | 3339/10708 [1:04:07<1:00:22,  2.03it/s]
 31%|███       | 3340/10708 [1:04:07<1:00:27,  2.03it/s]
 31%|███       | 3341/10708 [1:04:08<1:00:27,  2.03it/s]
 31%|███       | 3342/10708 [1:04:08<1:00:25,  2.03it/s]
 31%|███       | 3343/10708 [1:04:09<1:00:22,  2.03it/s]
 31%|███       | 3344/10708 [1:04:09<1:00:23,  2.03it/s]
 31%|███       | 3345/10708 [1:04:10<1:00:25,  2.03it/s]
 31%|███       | 3346/10708 [1:04:10<1:00:29,  2.03it/s]
 31%|███▏      | 3347/10708 [1:04:11<1:00:25,  2.03it/s]
 31%|███▏      | 3348/10708 [1:04:11<1:00:22,  2.03it/s]
 31%|███▏      | 3349/10708 [1:04:12<1:00:20,  2.03it/s]
 31%|███▏      | 3350/10708 [1:04:12<1:00:21,  2.03it/s]{'loss': 3.8208, 'grad_norm': 0.20793431997299194, 'learning_rate': 0.0008682422933311158, 'epoch': 0.31}

                                                        

 31%|███▏      | 3350/10708 [1:04:12<1:00:21,  2.03it/s]
 31%|███▏      | 3351/10708 [1:04:13<1:00:33,  2.02it/s]
 31%|███▏      | 3352/10708 [1:04:13<1:00:32,  2.03it/s]
 31%|███▏      | 3353/10708 [1:04:14<1:00:26,  2.03it/s]
 31%|███▏      | 3354/10708 [1:04:14<1:00:28,  2.03it/s]
 31%|███▏      | 3355/10708 [1:04:15<1:00:23,  2.03it/s]
 31%|███▏      | 3356/10708 [1:04:15<1:00:21,  2.03it/s]
 31%|███▏      | 3357/10708 [1:04:16<1:00:21,  2.03it/s]
 31%|███▏      | 3358/10708 [1:04:16<1:00:15,  2.03it/s]
 31%|███▏      | 3359/10708 [1:04:17<1:00:08,  2.04it/s]
 31%|███▏      | 3360/10708 [1:04:17<1:00:17,  2.03it/s]
 31%|███▏      | 3361/10708 [1:04:18<1:00:12,  2.03it/s]
 31%|███▏      | 3362/10708 [1:04:18<1:00:21,  2.03it/s]
 31%|███▏      | 3363/10708 [1:04:19<1:00:16,  2.03it/s]
 31%|███▏      | 3364/10708 [1:04:19<1:00:09,  2.03it/s]
 31%|███▏      | 3365/10708 [1:04:20<1:00:13,  2.03it/s]
 31%|███▏      | 3366/10708 [1:04:20<1:00:13,  2.03it/s]
 31%|███▏      | 3367/10708 [1:04:21<1:00:09,  2.03it/s]
 31%|███▏      | 3368/10708 [1:04:21<1:00:10,  2.03it/s]
 31%|███▏      | 3369/10708 [1:04:22<1:00:06,  2.03it/s]
 31%|███▏      | 3370/10708 [1:04:22<1:00:19,  2.03it/s]
 31%|███▏      | 3371/10708 [1:04:23<1:00:14,  2.03it/s]
 31%|███▏      | 3372/10708 [1:04:23<1:00:10,  2.03it/s]
 31%|███▏      | 3373/10708 [1:04:24<1:00:09,  2.03it/s]
 32%|███▏      | 3374/10708 [1:04:24<1:00:08,  2.03it/s]
 32%|███▏      | 3375/10708 [1:04:25<1:00:09,  2.03it/s]
                                                        
{'loss': 3.8176, 'grad_norm': 0.2111804038286209, 'learning_rate': 0.0008654736041379197, 'epoch': 0.32}

 32%|███▏      | 3375/10708 [1:04:25<1:00:09,  2.03it/s]
 32%|███▏      | 3376/10708 [1:04:25<1:00:10,  2.03it/s]
 32%|███▏      | 3377/10708 [1:04:26<1:00:08,  2.03it/s]
 32%|███▏      | 3378/10708 [1:04:26<1:00:13,  2.03it/s]
 32%|███▏      | 3379/10708 [1:04:27<1:00:08,  2.03it/s]
 32%|███▏      | 3380/10708 [1:04:27<1:00:13,  2.03it/s]
 32%|███▏      | 3381/10708 [1:04:28<1:00:07,  2.03it/s]
 32%|███▏      | 3382/10708 [1:04:28<1:00:05,  2.03it/s]
 32%|███▏      | 3383/10708 [1:04:29<1:00:07,  2.03it/s]
 32%|███▏      | 3384/10708 [1:04:29<1:00:01,  2.03it/s]
 32%|███▏      | 3385/10708 [1:04:30<59:58,  2.03it/s]  
 32%|███▏      | 3386/10708 [1:04:30<59:58,  2.03it/s]
 32%|███▏      | 3387/10708 [1:04:31<1:00:03,  2.03it/s]
 32%|███▏      | 3388/10708 [1:04:31<1:00:01,  2.03it/s]
 32%|███▏      | 3389/10708 [1:04:32<1:00:07,  2.03it/s]
 32%|███▏      | 3390/10708 [1:04:32<1:00:06,  2.03it/s]
 32%|███▏      | 3391/10708 [1:04:33<1:00:09,  2.03it/s]
 32%|███▏      | 3392/10708 [1:04:33<1:00:04,  2.03it/s]
 32%|███▏      | 3393/10708 [1:04:34<1:00:04,  2.03it/s]
 32%|███▏      | 3394/10708 [1:04:34<1:00:02,  2.03it/s]
 32%|███▏      | 3395/10708 [1:04:35<1:00:00,  2.03it/s]
 32%|███▏      | 3396/10708 [1:04:35<1:00:07,  2.03it/s]
 32%|███▏      | 3397/10708 [1:04:36<1:00:11,  2.02it/s]
 32%|███▏      | 3398/10708 [1:04:36<1:00:12,  2.02it/s]
 32%|███▏      | 3399/10708 [1:04:37<1:00:11,  2.02it/s]
 32%|███▏      | 3400/10708 [1:04:37<1:00:08,  2.03it/s]
                                                        
{'loss': 3.8153, 'grad_norm': 0.21064741909503937, 'learning_rate': 0.0008626806404802811, 'epoch': 0.32}

 32%|███▏      | 3400/10708 [1:04:37<1:00:08,  2.03it/s]
 32%|███▏      | 3401/10708 [1:04:38<1:00:10,  2.02it/s]
 32%|███▏      | 3402/10708 [1:04:38<1:00:11,  2.02it/s]
 32%|███▏      | 3403/10708 [1:04:38<1:00:00,  2.03it/s]
 32%|███▏      | 3404/10708 [1:04:39<1:00:01,  2.03it/s]
 32%|███▏      | 3405/10708 [1:04:39<1:00:00,  2.03it/s]
 32%|███▏      | 3406/10708 [1:04:40<1:00:00,  2.03it/s]
 32%|███▏      | 3407/10708 [1:04:40<59:56,  2.03it/s]  
 32%|███▏      | 3408/10708 [1:04:41<59:56,  2.03it/s]
 32%|███▏      | 3409/10708 [1:04:41<59:54,  2.03it/s]
 32%|███▏      | 3410/10708 [1:04:42<59:52,  2.03it/s]
 32%|███▏      | 3411/10708 [1:04:42<59:49,  2.03it/s]
 32%|███▏      | 3412/10708 [1:04:43<1:00:01,  2.03it/s]
 32%|███▏      | 3413/10708 [1:04:43<1:00:04,  2.02it/s]
 32%|███▏      | 3414/10708 [1:04:44<59:59,  2.03it/s]  
 32%|███▏      | 3415/10708 [1:04:44<59:58,  2.03it/s]
 32%|███▏      | 3416/10708 [1:04:45<59:57,  2.03it/s]
 32%|███▏      | 3417/10708 [1:04:45<59:54,  2.03it/s]
 32%|███▏      | 3418/10708 [1:04:46<59:50,  2.03it/s]
 32%|███▏      | 3419/10708 [1:04:46<59:49,  2.03it/s]
 32%|███▏      | 3420/10708 [1:04:47<59:46,  2.03it/s]
 32%|███▏      | 3421/10708 [1:04:47<59:45,  2.03it/s]
 32%|███▏      | 3422/10708 [1:04:48<59:47,  2.03it/s]
 32%|███▏      | 3423/10708 [1:04:48<59:47,  2.03it/s]
 32%|███▏      | 3424/10708 [1:04:49<59:45,  2.03it/s]
 32%|███▏      | 3425/10708 [1:04:49<59:48,  2.03it/s]
                                                      
{'loss': 3.8083, 'grad_norm': 0.19689825177192688, 'learning_rate': 0.0008598635878646122, 'epoch': 0.32}

 32%|███▏      | 3425/10708 [1:04:49<59:48,  2.03it/s]
 32%|███▏      | 3426/10708 [1:04:50<59:47,  2.03it/s]
 32%|███▏      | 3427/10708 [1:04:50<59:46,  2.03it/s]
 32%|███▏      | 3428/10708 [1:04:51<59:46,  2.03it/s]
 32%|███▏      | 3429/10708 [1:04:51<59:38,  2.03it/s]
 32%|███▏      | 3430/10708 [1:04:52<59:41,  2.03it/s]
 32%|███▏      | 3431/10708 [1:04:52<59:50,  2.03it/s]
 32%|███▏      | 3432/10708 [1:04:53<59:46,  2.03it/s]
 32%|███▏      | 3433/10708 [1:04:53<59:46,  2.03it/s]
 32%|███▏      | 3434/10708 [1:04:54<59:38,  2.03it/s]
 32%|███▏      | 3435/10708 [1:04:54<59:39,  2.03it/s]
 32%|███▏      | 3436/10708 [1:04:55<59:36,  2.03it/s]
 32%|███▏      | 3437/10708 [1:04:55<59:34,  2.03it/s]
 32%|███▏      | 3438/10708 [1:04:56<59:40,  2.03it/s]
 32%|███▏      | 3439/10708 [1:04:56<59:37,  2.03it/s]
 32%|███▏      | 3440/10708 [1:04:57<59:36,  2.03it/s]
 32%|███▏      | 3441/10708 [1:04:57<59:40,  2.03it/s]
 32%|███▏      | 3442/10708 [1:04:58<59:37,  2.03it/s]
 32%|███▏      | 3443/10708 [1:04:58<59:40,  2.03it/s]
 32%|███▏      | 3444/10708 [1:04:59<59:36,  2.03it/s]
 32%|███▏      | 3445/10708 [1:04:59<59:33,  2.03it/s]
 32%|███▏      | 3446/10708 [1:05:00<59:33,  2.03it/s]
 32%|███▏      | 3447/10708 [1:05:00<59:29,  2.03it/s]
 32%|███▏      | 3448/10708 [1:05:01<59:31,  2.03it/s]
 32%|███▏      | 3449/10708 [1:05:01<59:39,  2.03it/s]
 32%|███▏      | 3450/10708 [1:05:02<59:35,  2.03it/s]{'loss': 3.8139, 'grad_norm': 0.20398055016994476, 'learning_rate': 0.0008570226333972944, 'epoch': 0.32}
                                                      

 32%|███▏      | 3450/10708 [1:05:02<59:35,  2.03it/s]
 32%|███▏      | 3451/10708 [1:05:02<59:39,  2.03it/s]
 32%|███▏      | 3452/10708 [1:05:03<59:34,  2.03it/s]
 32%|███▏      | 3453/10708 [1:05:03<59:38,  2.03it/s]
 32%|███▏      | 3454/10708 [1:05:04<59:35,  2.03it/s]
 32%|███▏      | 3455/10708 [1:05:04<59:38,  2.03it/s]
 32%|███▏      | 3456/10708 [1:05:05<59:46,  2.02it/s]
 32%|███▏      | 3457/10708 [1:05:05<59:38,  2.03it/s]
 32%|███▏      | 3458/10708 [1:05:06<59:37,  2.03it/s]
 32%|███▏      | 3459/10708 [1:05:06<59:34,  2.03it/s]
 32%|███▏      | 3460/10708 [1:05:07<59:33,  2.03it/s]
 32%|███▏      | 3461/10708 [1:05:07<59:32,  2.03it/s]
 32%|███▏      | 3462/10708 [1:05:08<59:26,  2.03it/s]
 32%|███▏      | 3463/10708 [1:05:08<59:27,  2.03it/s]
 32%|███▏      | 3464/10708 [1:05:09<59:23,  2.03it/s]
 32%|███▏      | 3465/10708 [1:05:09<59:22,  2.03it/s]
 32%|███▏      | 3466/10708 [1:05:10<59:26,  2.03it/s]
 32%|███▏      | 3467/10708 [1:05:10<59:26,  2.03it/s]
 32%|███▏      | 3468/10708 [1:05:11<59:24,  2.03it/s]
 32%|███▏      | 3469/10708 [1:05:11<59:27,  2.03it/s]
 32%|███▏      | 3470/10708 [1:05:11<59:27,  2.03it/s]
 32%|███▏      | 3471/10708 [1:05:12<59:29,  2.03it/s]
 32%|███▏      | 3472/10708 [1:05:12<59:28,  2.03it/s]
 32%|███▏      | 3473/10708 [1:05:13<59:33,  2.02it/s]
 32%|███▏      | 3474/10708 [1:05:13<59:33,  2.02it/s]
 32%|███▏      | 3475/10708 [1:05:14<59:37,  2.02it/s]{'loss': 3.8028, 'grad_norm': 0.20424214005470276, 'learning_rate': 0.0008541579657722507, 'epoch': 0.32}
                                                      

 32%|███▏      | 3475/10708 [1:05:14<59:37,  2.02it/s]
 32%|███▏      | 3476/10708 [1:05:14<59:37,  2.02it/s]
 32%|███▏      | 3477/10708 [1:05:15<59:32,  2.02it/s]
 32%|███▏      | 3478/10708 [1:05:15<59:27,  2.03it/s]
 32%|███▏      | 3479/10708 [1:05:16<59:27,  2.03it/s]
 32%|███▏      | 3480/10708 [1:05:16<59:25,  2.03it/s]
 33%|███▎      | 3481/10708 [1:05:17<59:20,  2.03it/s]
 33%|███▎      | 3482/10708 [1:05:17<59:21,  2.03it/s]
 33%|███▎      | 3483/10708 [1:05:18<59:24,  2.03it/s]
 33%|███▎      | 3484/10708 [1:05:18<59:19,  2.03it/s]
 33%|███▎      | 3485/10708 [1:05:19<59:17,  2.03it/s]
 33%|███▎      | 3486/10708 [1:05:19<59:13,  2.03it/s]
 33%|███▎      | 3487/10708 [1:05:20<59:27,  2.02it/s]
 33%|███▎      | 3488/10708 [1:05:20<59:22,  2.03it/s]
 33%|███▎      | 3489/10708 [1:05:21<59:24,  2.02it/s]
 33%|███▎      | 3490/10708 [1:05:21<59:20,  2.03it/s]
 33%|███▎      | 3491/10708 [1:05:22<59:17,  2.03it/s]
 33%|███▎      | 3492/10708 [1:05:22<59:16,  2.03it/s]
 33%|███▎      | 3493/10708 [1:05:23<59:16,  2.03it/s]
 33%|███▎      | 3494/10708 [1:05:23<59:10,  2.03it/s]
 33%|███▎      | 3495/10708 [1:05:24<59:09,  2.03it/s]
 33%|███▎      | 3496/10708 [1:05:24<59:09,  2.03it/s]
 33%|███▎      | 3497/10708 [1:05:25<59:10,  2.03it/s]
 33%|███▎      | 3498/10708 [1:05:25<59:12,  2.03it/s]
 33%|███▎      | 3499/10708 [1:05:26<59:10,  2.03it/s]
 33%|███▎      | 3500/10708 [1:05:26<59:15,  2.03it/s]{'loss': 3.8112, 'grad_norm': 0.21291854977607727, 'learning_rate': 0.0008512697752584134, 'epoch': 0.33}
                                                      

 33%|███▎      | 3500/10708 [1:05:26<59:15,  2.03it/s]
 33%|███▎      | 3501/10708 [1:05:27<59:16,  2.03it/s]
 33%|███▎      | 3502/10708 [1:05:27<59:19,  2.02it/s]
 33%|███▎      | 3503/10708 [1:05:28<59:15,  2.03it/s]
 33%|███▎      | 3504/10708 [1:05:28<59:07,  2.03it/s]
 33%|███▎      | 3505/10708 [1:05:29<59:08,  2.03it/s]
 33%|███▎      | 3506/10708 [1:05:29<59:06,  2.03it/s]
 33%|███▎      | 3507/10708 [1:05:30<59:06,  2.03it/s]
 33%|███▎      | 3508/10708 [1:05:30<59:09,  2.03it/s]
 33%|███▎      | 3509/10708 [1:05:31<59:07,  2.03it/s]
 33%|███▎      | 3510/10708 [1:05:31<59:08,  2.03it/s]
 33%|███▎      | 3511/10708 [1:05:32<59:04,  2.03it/s]
 33%|███▎      | 3512/10708 [1:05:32<59:01,  2.03it/s]
 33%|███▎      | 3513/10708 [1:05:33<59:06,  2.03it/s]
 33%|███▎      | 3514/10708 [1:05:33<59:02,  2.03it/s]
 33%|███▎      | 3515/10708 [1:05:34<59:03,  2.03it/s]
 33%|███▎      | 3516/10708 [1:05:34<59:06,  2.03it/s]
 33%|███▎      | 3517/10708 [1:05:35<59:00,  2.03it/s]
 33%|███▎      | 3518/10708 [1:05:35<59:04,  2.03it/s]
 33%|███▎      | 3519/10708 [1:05:36<59:02,  2.03it/s]
 33%|███▎      | 3520/10708 [1:05:36<59:03,  2.03it/s]
 33%|███▎      | 3521/10708 [1:05:37<58:58,  2.03it/s]
 33%|███▎      | 3522/10708 [1:05:37<58:54,  2.03it/s]
 33%|███▎      | 3523/10708 [1:05:38<58:55,  2.03it/s]
 33%|███▎      | 3524/10708 [1:05:38<58:57,  2.03it/s]
 33%|███▎      | 3525/10708 [1:05:39<59:00,  2.03it/s]
                                                      
{'loss': 3.802, 'grad_norm': 0.2138051688671112, 'learning_rate': 0.0008483582536870862, 'epoch': 0.33}

 33%|███▎      | 3525/10708 [1:05:39<59:00,  2.03it/s]
 33%|███▎      | 3526/10708 [1:05:39<59:01,  2.03it/s]
 33%|███▎      | 3527/10708 [1:05:40<59:07,  2.02it/s]
 33%|███▎      | 3528/10708 [1:05:40<59:02,  2.03it/s]
 33%|███▎      | 3529/10708 [1:05:41<59:00,  2.03it/s]
 33%|███▎      | 3530/10708 [1:05:41<58:57,  2.03it/s]
 33%|███▎      | 3531/10708 [1:05:42<58:56,  2.03it/s]
 33%|███▎      | 3532/10708 [1:05:42<59:00,  2.03it/s]
 33%|███▎      | 3533/10708 [1:05:43<59:03,  2.02it/s]
 33%|███▎      | 3534/10708 [1:05:43<59:02,  2.03it/s]
 33%|███▎      | 3535/10708 [1:05:44<59:03,  2.02it/s]
 33%|███▎      | 3536/10708 [1:05:44<58:58,  2.03it/s]
 33%|███▎      | 3537/10708 [1:05:45<58:53,  2.03it/s]
 33%|███▎      | 3538/10708 [1:05:45<58:51,  2.03it/s]
 33%|███▎      | 3539/10708 [1:05:46<58:53,  2.03it/s]
 33%|███▎      | 3540/10708 [1:05:46<58:47,  2.03it/s]
 33%|███▎      | 3541/10708 [1:05:46<58:45,  2.03it/s]
 33%|███▎      | 3542/10708 [1:05:47<58:47,  2.03it/s]
 33%|███▎      | 3543/10708 [1:05:47<58:44,  2.03it/s]
 33%|███▎      | 3544/10708 [1:05:48<58:45,  2.03it/s]
 33%|███▎      | 3545/10708 [1:05:48<58:50,  2.03it/s]
 33%|███▎      | 3546/10708 [1:05:49<58:48,  2.03it/s]
 33%|███▎      | 3547/10708 [1:05:49<58:45,  2.03it/s]
 33%|███▎      | 3548/10708 [1:05:50<58:43,  2.03it/s]
 33%|███▎      | 3549/10708 [1:05:50<58:45,  2.03it/s]
 33%|███▎      | 3550/10708 [1:05:51<58:48,  2.03it/s]{'loss': 3.7983, 'grad_norm': 0.21436575055122375, 'learning_rate': 0.0008454235944392029, 'epoch': 0.33}
                                                      

 33%|███▎      | 3550/10708 [1:05:51<58:48,  2.03it/s]
 33%|███▎      | 3551/10708 [1:05:51<59:03,  2.02it/s]
 33%|███▎      | 3552/10708 [1:05:52<58:55,  2.02it/s]
 33%|███▎      | 3553/10708 [1:05:52<58:50,  2.03it/s]
 33%|███▎      | 3554/10708 [1:05:53<1:06:37,  1.79it/s]
 33%|███▎      | 3555/10708 [1:05:54<1:04:10,  1.86it/s]
 33%|███▎      | 3556/10708 [1:05:54<1:02:31,  1.91it/s]
 33%|███▎      | 3557/10708 [1:05:55<1:01:22,  1.94it/s]
 33%|███▎      | 3558/10708 [1:05:55<1:00:30,  1.97it/s]
 33%|███▎      | 3559/10708 [1:05:56<1:00:03,  1.98it/s]
 33%|███▎      | 3560/10708 [1:05:56<59:37,  2.00it/s]  
 33%|███▎      | 3561/10708 [1:05:57<59:19,  2.01it/s]
 33%|███▎      | 3562/10708 [1:05:57<59:12,  2.01it/s]
 33%|███▎      | 3563/10708 [1:05:58<58:55,  2.02it/s]
 33%|███▎      | 3564/10708 [1:05:58<58:49,  2.02it/s]
 33%|███▎      | 3565/10708 [1:05:59<58:42,  2.03it/s]
 33%|███▎      | 3566/10708 [1:05:59<58:37,  2.03it/s]
 33%|███▎      | 3567/10708 [1:06:00<58:36,  2.03it/s]
 33%|███▎      | 3568/10708 [1:06:00<58:37,  2.03it/s]
 33%|███▎      | 3569/10708 [1:06:01<58:38,  2.03it/s]
 33%|███▎      | 3570/10708 [1:06:01<58:34,  2.03it/s]
 33%|███▎      | 3571/10708 [1:06:01<58:31,  2.03it/s]
 33%|███▎      | 3572/10708 [1:06:02<58:36,  2.03it/s]
 33%|███▎      | 3573/10708 [1:06:02<58:35,  2.03it/s]
 33%|███▎      | 3574/10708 [1:06:03<58:31,  2.03it/s]
 33%|███▎      | 3575/10708 [1:06:03<58:30,  2.03it/s]{'loss': 3.7967, 'grad_norm': 0.20647336542606354, 'learning_rate': 0.0008424659924324833, 'epoch': 0.33}

                                                      

 33%|███▎      | 3575/10708 [1:06:03<58:30,  2.03it/s]
 33%|███▎      | 3576/10708 [1:06:04<58:29,  2.03it/s]
 33%|███▎      | 3577/10708 [1:06:04<58:25,  2.03it/s]
 33%|███▎      | 3578/10708 [1:06:05<58:28,  2.03it/s]
 33%|███▎      | 3579/10708 [1:06:05<58:34,  2.03it/s]
 33%|███▎      | 3580/10708 [1:06:06<58:34,  2.03it/s]
 33%|███▎      | 3581/10708 [1:06:06<58:31,  2.03it/s]
 33%|███▎      | 3582/10708 [1:06:07<58:37,  2.03it/s]
 33%|███▎      | 3583/10708 [1:06:07<58:32,  2.03it/s]
 33%|███▎      | 3584/10708 [1:06:08<58:28,  2.03it/s]
 33%|███▎      | 3585/10708 [1:06:08<58:27,  2.03it/s]
 33%|███▎      | 3586/10708 [1:06:09<58:23,  2.03it/s]
 33%|███▎      | 3587/10708 [1:06:09<58:22,  2.03it/s]
 34%|███▎      | 3588/10708 [1:06:10<58:22,  2.03it/s]
 34%|███▎      | 3589/10708 [1:06:10<58:25,  2.03it/s]
 34%|███▎      | 3590/10708 [1:06:11<58:23,  2.03it/s]
 34%|███▎      | 3591/10708 [1:06:11<58:25,  2.03it/s]
 34%|███▎      | 3592/10708 [1:06:12<58:24,  2.03it/s]
 34%|███▎      | 3593/10708 [1:06:12<58:24,  2.03it/s]
 34%|███▎      | 3594/10708 [1:06:13<58:28,  2.03it/s]
 34%|███▎      | 3595/10708 [1:06:13<58:34,  2.02it/s]
 34%|███▎      | 3596/10708 [1:06:14<58:25,  2.03it/s]
 34%|███▎      | 3597/10708 [1:06:14<58:26,  2.03it/s]
 34%|███▎      | 3598/10708 [1:06:15<58:21,  2.03it/s]
 34%|███▎      | 3599/10708 [1:06:15<58:22,  2.03it/s]
 34%|███▎      | 3600/10708 [1:06:16<58:21,  2.03it/s]
                                                      
{'loss': 3.7953, 'grad_norm': 0.21377171576023102, 'learning_rate': 0.0008394856441084869, 'epoch': 0.34}

 34%|███▎      | 3600/10708 [1:06:16<58:21,  2.03it/s]
 34%|███▎      | 3601/10708 [1:06:16<58:22,  2.03it/s]
 34%|███▎      | 3602/10708 [1:06:17<58:28,  2.03it/s]
 34%|███▎      | 3603/10708 [1:06:17<58:24,  2.03it/s]
 34%|███▎      | 3604/10708 [1:06:18<58:20,  2.03it/s]
 34%|███▎      | 3605/10708 [1:06:18<58:18,  2.03it/s]
 34%|███▎      | 3606/10708 [1:06:19<58:20,  2.03it/s]
 34%|███▎      | 3607/10708 [1:06:19<58:19,  2.03it/s]
 34%|███▎      | 3608/10708 [1:06:20<58:16,  2.03it/s]
 34%|███▎      | 3609/10708 [1:06:20<58:16,  2.03it/s]
 34%|███▎      | 3610/10708 [1:06:21<58:20,  2.03it/s]
 34%|███▎      | 3611/10708 [1:06:21<58:19,  2.03it/s]
 34%|███▎      | 3612/10708 [1:06:22<58:19,  2.03it/s]
 34%|███▎      | 3613/10708 [1:06:22<58:15,  2.03it/s]
 34%|███▍      | 3614/10708 [1:06:23<58:16,  2.03it/s]
 34%|███▍      | 3615/10708 [1:06:23<58:13,  2.03it/s]
 34%|███▍      | 3616/10708 [1:06:24<58:07,  2.03it/s]
 34%|███▍      | 3617/10708 [1:06:24<58:10,  2.03it/s]
 34%|███▍      | 3618/10708 [1:06:25<58:09,  2.03it/s]
 34%|███▍      | 3619/10708 [1:06:25<58:10,  2.03it/s]
 34%|███▍      | 3620/10708 [1:06:26<58:21,  2.02it/s]
 34%|███▍      | 3621/10708 [1:06:26<58:24,  2.02it/s]
 34%|███▍      | 3622/10708 [1:06:27<58:21,  2.02it/s]
 34%|███▍      | 3623/10708 [1:06:27<58:25,  2.02it/s]
 34%|███▍      | 3624/10708 [1:06:28<58:19,  2.02it/s]
 34%|███▍      | 3625/10708 [1:06:28<58:17,  2.03it/s]{'loss': 3.7922, 'grad_norm': 0.20145882666110992, 'learning_rate': 0.0008364827474195651, 'epoch': 0.34}

                                                      

 34%|███▍      | 3625/10708 [1:06:28<58:17,  2.03it/s]
 34%|███▍      | 3626/10708 [1:06:29<58:23,  2.02it/s]
 34%|███▍      | 3627/10708 [1:06:29<58:18,  2.02it/s]
 34%|███▍      | 3628/10708 [1:06:30<58:11,  2.03it/s]
 34%|███▍      | 3629/10708 [1:06:30<58:10,  2.03it/s]
 34%|███▍      | 3630/10708 [1:06:31<58:08,  2.03it/s]
 34%|███▍      | 3631/10708 [1:06:31<58:06,  2.03it/s]
 34%|███▍      | 3632/10708 [1:06:32<58:07,  2.03it/s]
 34%|███▍      | 3633/10708 [1:06:32<58:07,  2.03it/s]
 34%|███▍      | 3634/10708 [1:06:33<58:07,  2.03it/s]
 34%|███▍      | 3635/10708 [1:06:33<58:04,  2.03it/s]
 34%|███▍      | 3636/10708 [1:06:34<58:02,  2.03it/s]
 34%|███▍      | 3637/10708 [1:06:34<58:02,  2.03it/s]
 34%|███▍      | 3638/10708 [1:06:35<57:55,  2.03it/s]
 34%|███▍      | 3639/10708 [1:06:35<57:58,  2.03it/s]
 34%|███▍      | 3640/10708 [1:06:35<58:01,  2.03it/s]
 34%|███▍      | 3641/10708 [1:06:36<57:58,  2.03it/s]
 34%|███▍      | 3642/10708 [1:06:36<58:02,  2.03it/s]
 34%|███▍      | 3643/10708 [1:06:37<57:58,  2.03it/s]
 34%|███▍      | 3644/10708 [1:06:37<57:56,  2.03it/s]
 34%|███▍      | 3645/10708 [1:06:38<57:55,  2.03it/s]
 34%|███▍      | 3646/10708 [1:06:38<57:56,  2.03it/s]
 34%|███▍      | 3647/10708 [1:06:39<57:55,  2.03it/s]
 34%|███▍      | 3648/10708 [1:06:39<57:53,  2.03it/s]
 34%|███▍      | 3649/10708 [1:06:40<58:00,  2.03it/s]
 34%|███▍      | 3650/10708 [1:06:40<58:04,  2.03it/s]
                                                      
{'loss': 3.7903, 'grad_norm': 0.204352468252182, 'learning_rate': 0.0008334575018157144, 'epoch': 0.34}

 34%|███▍      | 3650/10708 [1:06:40<58:04,  2.03it/s]
 34%|███▍      | 3651/10708 [1:06:41<58:03,  2.03it/s]
 34%|███▍      | 3652/10708 [1:06:41<58:01,  2.03it/s]
 34%|███▍      | 3653/10708 [1:06:42<58:00,  2.03it/s]
 34%|███▍      | 3654/10708 [1:06:42<57:54,  2.03it/s]
 34%|███▍      | 3655/10708 [1:06:43<58:00,  2.03it/s]
 34%|███▍      | 3656/10708 [1:06:43<57:58,  2.03it/s]
 34%|███▍      | 3657/10708 [1:06:44<57:59,  2.03it/s]
 34%|███▍      | 3658/10708 [1:06:44<57:54,  2.03it/s]
 34%|███▍      | 3659/10708 [1:06:45<57:57,  2.03it/s]
 34%|███▍      | 3660/10708 [1:06:45<57:55,  2.03it/s]
 34%|███▍      | 3661/10708 [1:06:46<57:51,  2.03it/s]
 34%|███▍      | 3662/10708 [1:06:46<57:48,  2.03it/s]
 34%|███▍      | 3663/10708 [1:06:47<57:53,  2.03it/s]
 34%|███▍      | 3664/10708 [1:06:47<57:50,  2.03it/s]
 34%|███▍      | 3665/10708 [1:06:48<57:50,  2.03it/s]
 34%|███▍      | 3666/10708 [1:06:48<57:44,  2.03it/s]
 34%|███▍      | 3667/10708 [1:06:49<57:46,  2.03it/s]
 34%|███▍      | 3668/10708 [1:06:49<57:42,  2.03it/s]
 34%|███▍      | 3669/10708 [1:06:50<57:46,  2.03it/s]
 34%|███▍      | 3670/10708 [1:06:50<57:43,  2.03it/s]
 34%|███▍      | 3671/10708 [1:06:51<57:43,  2.03it/s]
 34%|███▍      | 3672/10708 [1:06:51<57:39,  2.03it/s]
 34%|███▍      | 3673/10708 [1:06:52<57:36,  2.04it/s]
 34%|███▍      | 3674/10708 [1:06:52<57:39,  2.03it/s]
 34%|███▍      | 3675/10708 [1:06:53<57:36,  2.03it/s]
                                                      
{'loss': 3.7895, 'grad_norm': 0.19522659480571747, 'learning_rate': 0.0008304101082313276, 'epoch': 0.34}

 34%|███▍      | 3675/10708 [1:06:53<57:36,  2.03it/s]
 34%|███▍      | 3676/10708 [1:06:53<57:47,  2.03it/s]
 34%|███▍      | 3677/10708 [1:06:54<57:43,  2.03it/s]
 34%|███▍      | 3678/10708 [1:06:54<57:39,  2.03it/s]
 34%|███▍      | 3679/10708 [1:06:55<57:45,  2.03it/s]
 34%|███▍      | 3680/10708 [1:06:55<57:45,  2.03it/s]
 34%|███▍      | 3681/10708 [1:06:56<57:47,  2.03it/s]
 34%|███▍      | 3682/10708 [1:06:56<57:46,  2.03it/s]
 34%|███▍      | 3683/10708 [1:06:57<57:44,  2.03it/s]
 34%|███▍      | 3684/10708 [1:06:57<57:45,  2.03it/s]
 34%|███▍      | 3685/10708 [1:06:58<57:42,  2.03it/s]
 34%|███▍      | 3686/10708 [1:06:58<57:37,  2.03it/s]
 34%|███▍      | 3687/10708 [1:06:59<57:35,  2.03it/s]
 34%|███▍      | 3688/10708 [1:06:59<57:38,  2.03it/s]
 34%|███▍      | 3689/10708 [1:07:00<57:38,  2.03it/s]
 34%|███▍      | 3690/10708 [1:07:00<57:37,  2.03it/s]
 34%|███▍      | 3691/10708 [1:07:01<57:35,  2.03it/s]
 34%|███▍      | 3692/10708 [1:07:01<57:39,  2.03it/s]
 34%|███▍      | 3693/10708 [1:07:02<57:35,  2.03it/s]
 34%|███▍      | 3694/10708 [1:07:02<57:35,  2.03it/s]
 35%|███▍      | 3695/10708 [1:07:03<57:48,  2.02it/s]
 35%|███▍      | 3696/10708 [1:07:03<57:43,  2.02it/s]
 35%|███▍      | 3697/10708 [1:07:04<57:40,  2.03it/s]
 35%|███▍      | 3698/10708 [1:07:04<57:33,  2.03it/s]
 35%|███▍      | 3699/10708 [1:07:05<57:36,  2.03it/s]
 35%|███▍      | 3700/10708 [1:07:05<57:33,  2.03it/s]{'loss': 3.7911, 'grad_norm': 0.19612690806388855, 'learning_rate': 0.000827340769071849, 'epoch': 0.35}
                                                      

 35%|███▍      | 3700/10708 [1:07:05<57:33,  2.03it/s]
 35%|███▍      | 3701/10708 [1:07:06<57:34,  2.03it/s]
 35%|███▍      | 3702/10708 [1:07:06<57:32,  2.03it/s]
 35%|███▍      | 3703/10708 [1:07:07<57:36,  2.03it/s]
 35%|███▍      | 3704/10708 [1:07:07<57:29,  2.03it/s]
 35%|███▍      | 3705/10708 [1:07:08<57:33,  2.03it/s]
 35%|███▍      | 3706/10708 [1:07:08<57:28,  2.03it/s]
 35%|███▍      | 3707/10708 [1:07:09<57:26,  2.03it/s]
 35%|███▍      | 3708/10708 [1:07:09<57:20,  2.03it/s]
 35%|███▍      | 3709/10708 [1:07:09<57:24,  2.03it/s]
 35%|███▍      | 3710/10708 [1:07:10<57:26,  2.03it/s]
 35%|███▍      | 3711/10708 [1:07:10<57:28,  2.03it/s]
 35%|███▍      | 3712/10708 [1:07:11<57:32,  2.03it/s]
 35%|███▍      | 3713/10708 [1:07:11<57:38,  2.02it/s]
 35%|███▍      | 3714/10708 [1:07:12<57:32,  2.03it/s]
 35%|███▍      | 3715/10708 [1:07:12<57:32,  2.03it/s]
 35%|███▍      | 3716/10708 [1:07:13<57:38,  2.02it/s]
 35%|███▍      | 3717/10708 [1:07:13<57:33,  2.02it/s]
 35%|███▍      | 3718/10708 [1:07:14<57:28,  2.03it/s]
 35%|███▍      | 3719/10708 [1:07:14<57:27,  2.03it/s]
 35%|███▍      | 3720/10708 [1:07:15<57:26,  2.03it/s]
 35%|███▍      | 3721/10708 [1:07:15<57:22,  2.03it/s]
 35%|███▍      | 3722/10708 [1:07:16<57:14,  2.03it/s]
 35%|███▍      | 3723/10708 [1:07:16<57:18,  2.03it/s]
 35%|███▍      | 3724/10708 [1:07:17<57:18,  2.03it/s]
 35%|███▍      | 3725/10708 [1:07:17<57:22,  2.03it/s]{'loss': 3.7854, 'grad_norm': 0.20849332213401794, 'learning_rate': 0.0008242496882003305, 'epoch': 0.35}
                                                      

 35%|███▍      | 3725/10708 [1:07:17<57:22,  2.03it/s]
 35%|███▍      | 3726/10708 [1:07:18<57:29,  2.02it/s]
 35%|███▍      | 3727/10708 [1:07:18<57:21,  2.03it/s]
 35%|███▍      | 3728/10708 [1:07:19<57:24,  2.03it/s]
 35%|███▍      | 3729/10708 [1:07:19<57:20,  2.03it/s]
 35%|███▍      | 3730/10708 [1:07:20<57:16,  2.03it/s]
 35%|███▍      | 3731/10708 [1:07:20<57:21,  2.03it/s]
 35%|███▍      | 3732/10708 [1:07:21<57:23,  2.03it/s]
 35%|███▍      | 3733/10708 [1:07:21<57:22,  2.03it/s]
 35%|███▍      | 3734/10708 [1:07:22<57:16,  2.03it/s]
 35%|███▍      | 3735/10708 [1:07:22<57:22,  2.03it/s]
 35%|███▍      | 3736/10708 [1:07:23<57:18,  2.03it/s]
 35%|███▍      | 3737/10708 [1:07:23<57:08,  2.03it/s]
 35%|███▍      | 3738/10708 [1:07:24<57:11,  2.03it/s]
 35%|███▍      | 3739/10708 [1:07:24<57:10,  2.03it/s]
 35%|███▍      | 3740/10708 [1:07:25<57:06,  2.03it/s]
 35%|███▍      | 3741/10708 [1:07:25<57:14,  2.03it/s]
 35%|███▍      | 3742/10708 [1:07:26<57:09,  2.03it/s]
 35%|███▍      | 3743/10708 [1:07:26<57:09,  2.03it/s]
 35%|███▍      | 3744/10708 [1:07:27<57:08,  2.03it/s]
 35%|███▍      | 3745/10708 [1:07:27<57:04,  2.03it/s]
 35%|███▍      | 3746/10708 [1:07:28<57:02,  2.03it/s]
 35%|███▍      | 3747/10708 [1:07:28<56:59,  2.04it/s]
 35%|███▌      | 3748/10708 [1:07:29<57:01,  2.03it/s]
 35%|███▌      | 3749/10708 [1:07:29<57:02,  2.03it/s]
 35%|███▌      | 3750/10708 [1:07:30<57:01,  2.03it/s]{'loss': 3.7866, 'grad_norm': 0.20268572866916656, 'learning_rate': 0.000821137070923891, 'epoch': 0.35}

                                                      

 35%|███▌      | 3750/10708 [1:07:30<57:01,  2.03it/s]
 35%|███▌      | 3751/10708 [1:07:30<57:06,  2.03it/s]
 35%|███▌      | 3752/10708 [1:07:31<57:09,  2.03it/s]
 35%|███▌      | 3753/10708 [1:07:31<57:03,  2.03it/s]
 35%|███▌      | 3754/10708 [1:07:32<57:06,  2.03it/s]
 35%|███▌      | 3755/10708 [1:07:32<57:03,  2.03it/s]
 35%|███▌      | 3756/10708 [1:07:33<57:00,  2.03it/s]
 35%|███▌      | 3757/10708 [1:07:33<56:58,  2.03it/s]
 35%|███▌      | 3758/10708 [1:07:34<56:59,  2.03it/s]
 35%|███▌      | 3759/10708 [1:07:34<56:55,  2.03it/s]
 35%|███▌      | 3760/10708 [1:07:35<56:58,  2.03it/s]
 35%|███▌      | 3761/10708 [1:07:35<56:55,  2.03it/s]
 35%|███▌      | 3762/10708 [1:07:36<57:00,  2.03it/s]
 35%|███▌      | 3763/10708 [1:07:36<57:04,  2.03it/s]
 35%|███▌      | 3764/10708 [1:07:37<57:00,  2.03it/s]
 35%|███▌      | 3765/10708 [1:07:37<56:58,  2.03it/s]
 35%|███▌      | 3766/10708 [1:07:38<56:57,  2.03it/s]
 35%|███▌      | 3767/10708 [1:07:38<56:54,  2.03it/s]
 35%|███▌      | 3768/10708 [1:07:39<56:58,  2.03it/s]
 35%|███▌      | 3769/10708 [1:07:39<56:57,  2.03it/s]
 35%|███▌      | 3770/10708 [1:07:40<56:58,  2.03it/s]
 35%|███▌      | 3771/10708 [1:07:40<56:57,  2.03it/s]
 35%|███▌      | 3772/10708 [1:07:41<57:02,  2.03it/s]
 35%|███▌      | 3773/10708 [1:07:41<57:01,  2.03it/s]
 35%|███▌      | 3774/10708 [1:07:42<56:59,  2.03it/s]
 35%|███▌      | 3775/10708 [1:07:42<56:56,  2.03it/s]{'loss': 3.778, 'grad_norm': 0.20329724252223969, 'learning_rate': 0.00081800312398008, 'epoch': 0.35}

                                                      

 35%|███▌      | 3775/10708 [1:07:42<56:56,  2.03it/s]
 35%|███▌      | 3776/10708 [1:07:43<56:58,  2.03it/s]
 35%|███▌      | 3777/10708 [1:07:43<56:58,  2.03it/s]
 35%|███▌      | 3778/10708 [1:07:43<57:02,  2.03it/s]
 35%|███▌      | 3779/10708 [1:07:44<56:59,  2.03it/s]
 35%|███▌      | 3780/10708 [1:07:44<56:56,  2.03it/s]
 35%|███▌      | 3781/10708 [1:07:45<56:52,  2.03it/s]
 35%|███▌      | 3782/10708 [1:07:45<56:51,  2.03it/s]
 35%|███▌      | 3783/10708 [1:07:46<56:50,  2.03it/s]
 35%|███▌      | 3784/10708 [1:07:46<56:47,  2.03it/s]
 35%|███▌      | 3785/10708 [1:07:47<56:46,  2.03it/s]
 35%|███▌      | 3786/10708 [1:07:47<56:45,  2.03it/s]
 35%|███▌      | 3787/10708 [1:07:48<56:45,  2.03it/s]
 35%|███▌      | 3788/10708 [1:07:48<56:42,  2.03it/s]
 35%|███▌      | 3789/10708 [1:07:49<56:46,  2.03it/s]
 35%|███▌      | 3790/10708 [1:07:49<56:59,  2.02it/s]
 35%|███▌      | 3791/10708 [1:07:50<56:53,  2.03it/s]
 35%|███▌      | 3792/10708 [1:07:50<56:50,  2.03it/s]
 35%|███▌      | 3793/10708 [1:07:51<56:51,  2.03it/s]
 35%|███▌      | 3794/10708 [1:07:51<56:47,  2.03it/s]
 35%|███▌      | 3795/10708 [1:07:52<56:43,  2.03it/s]
 35%|███▌      | 3796/10708 [1:07:52<56:44,  2.03it/s]
 35%|███▌      | 3797/10708 [1:07:53<56:46,  2.03it/s]
 35%|███▌      | 3798/10708 [1:07:53<56:38,  2.03it/s]
 35%|███▌      | 3799/10708 [1:07:54<56:39,  2.03it/s]
 35%|███▌      | 3800/10708 [1:07:54<56:39,  2.03it/s]
                                                      
{'loss': 3.7755, 'grad_norm': 0.21012020111083984, 'learning_rate': 0.0008148480555231468, 'epoch': 0.35}

 35%|███▌      | 3800/10708 [1:07:54<56:39,  2.03it/s]
 35%|███▌      | 3801/10708 [1:07:55<56:44,  2.03it/s]
 36%|███▌      | 3802/10708 [1:07:55<56:42,  2.03it/s]
 36%|███▌      | 3803/10708 [1:07:56<56:38,  2.03it/s]
 36%|███▌      | 3804/10708 [1:07:56<56:46,  2.03it/s]
 36%|███▌      | 3805/10708 [1:07:57<56:41,  2.03it/s]
 36%|███▌      | 3806/10708 [1:07:57<56:42,  2.03it/s]
 36%|███▌      | 3807/10708 [1:07:58<56:39,  2.03it/s]
 36%|███▌      | 3808/10708 [1:07:58<56:35,  2.03it/s]
 36%|███▌      | 3809/10708 [1:07:59<56:36,  2.03it/s]
 36%|███▌      | 3810/10708 [1:07:59<56:36,  2.03it/s]
 36%|███▌      | 3811/10708 [1:08:00<56:33,  2.03it/s]
 36%|███▌      | 3812/10708 [1:08:00<56:34,  2.03it/s]
 36%|███▌      | 3813/10708 [1:08:01<56:37,  2.03it/s]
 36%|███▌      | 3814/10708 [1:08:01<56:29,  2.03it/s]
 36%|███▌      | 3815/10708 [1:08:02<56:26,  2.04it/s]
 36%|███▌      | 3816/10708 [1:08:02<56:25,  2.04it/s]
 36%|███▌      | 3817/10708 [1:08:03<56:27,  2.03it/s]
 36%|███▌      | 3818/10708 [1:08:03<56:26,  2.03it/s]
 36%|███▌      | 3819/10708 [1:08:04<56:27,  2.03it/s]
 36%|███▌      | 3820/10708 [1:08:04<56:24,  2.03it/s]
 36%|███▌      | 3821/10708 [1:08:05<56:29,  2.03it/s]
 36%|███▌      | 3822/10708 [1:08:05<56:27,  2.03it/s]
 36%|███▌      | 3823/10708 [1:08:06<56:27,  2.03it/s]
 36%|███▌      | 3824/10708 [1:08:06<56:31,  2.03it/s]
 36%|███▌      | 3825/10708 [1:08:07<56:25,  2.03it/s]{'loss': 3.7812, 'grad_norm': 0.21310366690158844, 'learning_rate': 0.0008116720751102145, 'epoch': 0.36}

                                                      

 36%|███▌      | 3825/10708 [1:08:07<56:25,  2.03it/s]
 36%|███▌      | 3826/10708 [1:08:07<56:32,  2.03it/s]
 36%|███▌      | 3827/10708 [1:08:08<56:32,  2.03it/s]
 36%|███▌      | 3828/10708 [1:08:08<56:25,  2.03it/s]
 36%|███▌      | 3829/10708 [1:08:09<56:29,  2.03it/s]
 36%|███▌      | 3830/10708 [1:08:09<56:27,  2.03it/s]
 36%|███▌      | 3831/10708 [1:08:10<56:23,  2.03it/s]
 36%|███▌      | 3832/10708 [1:08:10<56:21,  2.03it/s]
 36%|███▌      | 3833/10708 [1:08:11<56:23,  2.03it/s]
 36%|███▌      | 3834/10708 [1:08:11<56:25,  2.03it/s]
 36%|███▌      | 3835/10708 [1:08:12<56:22,  2.03it/s]
 36%|███▌      | 3836/10708 [1:08:12<56:18,  2.03it/s]
 36%|███▌      | 3837/10708 [1:08:13<56:17,  2.03it/s]
 36%|███▌      | 3838/10708 [1:08:13<56:24,  2.03it/s]
 36%|███▌      | 3839/10708 [1:08:14<56:32,  2.02it/s]
 36%|███▌      | 3840/10708 [1:08:14<56:32,  2.02it/s]
 36%|███▌      | 3841/10708 [1:08:15<56:28,  2.03it/s]
 36%|███▌      | 3842/10708 [1:08:15<56:24,  2.03it/s]
 36%|███▌      | 3843/10708 [1:08:16<56:22,  2.03it/s]
 36%|███▌      | 3844/10708 [1:08:16<56:23,  2.03it/s]
 36%|███▌      | 3845/10708 [1:08:16<56:27,  2.03it/s]
 36%|███▌      | 3846/10708 [1:08:17<56:26,  2.03it/s]
 36%|███▌      | 3847/10708 [1:08:17<56:22,  2.03it/s]
 36%|███▌      | 3848/10708 [1:08:18<56:18,  2.03it/s]
 36%|███▌      | 3849/10708 [1:08:18<56:17,  2.03it/s]
 36%|███▌      | 3850/10708 [1:08:19<56:14,  2.03it/s]{'loss': 3.7763, 'grad_norm': 0.24404746294021606, 'learning_rate': 0.0008084753936873617, 'epoch': 0.36}

                                                      

 36%|███▌      | 3850/10708 [1:08:19<56:14,  2.03it/s]
 36%|███▌      | 3851/10708 [1:08:19<56:18,  2.03it/s]
 36%|███▌      | 3852/10708 [1:08:20<56:13,  2.03it/s]
 36%|███▌      | 3853/10708 [1:08:20<56:19,  2.03it/s]
 36%|███▌      | 3854/10708 [1:08:21<56:17,  2.03it/s]
 36%|███▌      | 3855/10708 [1:08:21<56:14,  2.03it/s]
 36%|███▌      | 3856/10708 [1:08:22<56:09,  2.03it/s]
 36%|███▌      | 3857/10708 [1:08:22<56:10,  2.03it/s]
 36%|███▌      | 3858/10708 [1:08:23<56:03,  2.04it/s]
 36%|███▌      | 3859/10708 [1:08:23<56:05,  2.03it/s]
 36%|███▌      | 3860/10708 [1:08:24<56:08,  2.03it/s]
 36%|███▌      | 3861/10708 [1:08:24<56:10,  2.03it/s]
 36%|███▌      | 3862/10708 [1:08:25<56:09,  2.03it/s]
 36%|███▌      | 3863/10708 [1:08:25<56:08,  2.03it/s]
 36%|███▌      | 3864/10708 [1:08:26<1:05:27,  1.74it/s]
 36%|███▌      | 3865/10708 [1:08:27<1:02:35,  1.82it/s]
 36%|███▌      | 3866/10708 [1:08:27<1:00:36,  1.88it/s]
 36%|███▌      | 3867/10708 [1:08:28<59:21,  1.92it/s]  
 36%|███▌      | 3868/10708 [1:08:28<58:20,  1.95it/s]
 36%|███▌      | 3869/10708 [1:08:29<57:42,  1.97it/s]
 36%|███▌      | 3870/10708 [1:08:29<57:14,  1.99it/s]
 36%|███▌      | 3871/10708 [1:08:30<56:51,  2.00it/s]
 36%|███▌      | 3872/10708 [1:08:30<56:37,  2.01it/s]
 36%|███▌      | 3873/10708 [1:08:31<56:24,  2.02it/s]
 36%|███▌      | 3874/10708 [1:08:31<56:18,  2.02it/s]
 36%|███▌      | 3875/10708 [1:08:32<56:11,  2.03it/s]
                                                      
{'loss': 3.7766, 'grad_norm': 0.20735499262809753, 'learning_rate': 0.0008052582235756115, 'epoch': 0.36}

 36%|███▌      | 3875/10708 [1:08:32<56:11,  2.03it/s]
 36%|███▌      | 3876/10708 [1:08:32<56:10,  2.03it/s]
 36%|███▌      | 3877/10708 [1:08:33<56:08,  2.03it/s]
 36%|███▌      | 3878/10708 [1:08:33<56:07,  2.03it/s]
 36%|███▌      | 3879/10708 [1:08:33<55:58,  2.03it/s]
 36%|███▌      | 3880/10708 [1:08:34<56:01,  2.03it/s]
 36%|███▌      | 3881/10708 [1:08:34<55:56,  2.03it/s]
 36%|███▋      | 3882/10708 [1:08:35<1:05:49,  1.73it/s]
 36%|███▋      | 3883/10708 [1:08:36<1:02:57,  1.81it/s]
 36%|███▋      | 3884/10708 [1:08:36<1:01:04,  1.86it/s]
 36%|███▋      | 3885/10708 [1:08:37<59:29,  1.91it/s]  
 36%|███▋      | 3886/10708 [1:08:37<58:19,  1.95it/s]
 36%|███▋      | 3887/10708 [1:08:38<57:43,  1.97it/s]
 36%|███▋      | 3888/10708 [1:08:38<57:07,  1.99it/s]
 36%|███▋      | 3889/10708 [1:08:39<56:45,  2.00it/s]
 36%|███▋      | 3890/10708 [1:08:39<56:30,  2.01it/s]
 36%|███▋      | 3891/10708 [1:08:40<56:18,  2.02it/s]
 36%|███▋      | 3892/10708 [1:08:40<56:08,  2.02it/s]
 36%|███▋      | 3893/10708 [1:08:41<56:08,  2.02it/s]
 36%|███▋      | 3894/10708 [1:08:41<56:04,  2.03it/s]
 36%|███▋      | 3895/10708 [1:08:42<55:57,  2.03it/s]
 36%|███▋      | 3896/10708 [1:08:42<55:57,  2.03it/s]
 36%|███▋      | 3897/10708 [1:08:43<55:55,  2.03it/s]
 36%|███▋      | 3898/10708 [1:08:43<56:04,  2.02it/s]
 36%|███▋      | 3899/10708 [1:08:44<55:56,  2.03it/s]
 36%|███▋      | 3900/10708 [1:08:44<55:51,  2.03it/s]
                                                      
{'loss': 3.77, 'grad_norm': 0.21825850009918213, 'learning_rate': 0.0008020207784568292, 'epoch': 0.36}

 36%|███▋      | 3900/10708 [1:08:44<55:51,  2.03it/s]
 36%|███▋      | 3901/10708 [1:08:45<55:55,  2.03it/s]
 36%|███▋      | 3902/10708 [1:08:45<55:51,  2.03it/s]
 36%|███▋      | 3903/10708 [1:08:46<55:51,  2.03it/s]
 36%|███▋      | 3904/10708 [1:08:46<55:51,  2.03it/s]
 36%|███▋      | 3905/10708 [1:08:47<55:46,  2.03it/s]
 36%|███▋      | 3906/10708 [1:08:47<55:49,  2.03it/s]
 36%|███▋      | 3907/10708 [1:08:48<55:51,  2.03it/s]
 36%|███▋      | 3908/10708 [1:08:48<55:48,  2.03it/s]
 37%|███▋      | 3909/10708 [1:08:49<55:59,  2.02it/s]
 37%|███▋      | 3910/10708 [1:08:49<55:54,  2.03it/s]
 37%|███▋      | 3911/10708 [1:08:50<55:51,  2.03it/s]
 37%|███▋      | 3912/10708 [1:08:50<55:46,  2.03it/s]
 37%|███▋      | 3913/10708 [1:08:51<55:46,  2.03it/s]
 37%|███▋      | 3914/10708 [1:08:51<55:43,  2.03it/s]
 37%|███▋      | 3915/10708 [1:08:52<55:41,  2.03it/s]
 37%|███▋      | 3916/10708 [1:08:52<55:38,  2.03it/s]
 37%|███▋      | 3917/10708 [1:08:52<55:36,  2.04it/s]
 37%|███▋      | 3918/10708 [1:08:53<55:37,  2.03it/s]
 37%|███▋      | 3919/10708 [1:08:53<55:46,  2.03it/s]
 37%|███▋      | 3920/10708 [1:08:54<55:42,  2.03it/s]
 37%|███▋      | 3921/10708 [1:08:54<55:44,  2.03it/s]
 37%|███▋      | 3922/10708 [1:08:55<55:46,  2.03it/s]
 37%|███▋      | 3923/10708 [1:08:55<55:42,  2.03it/s]
 37%|███▋      | 3924/10708 [1:08:56<55:36,  2.03it/s]
 37%|███▋      | 3925/10708 [1:08:56<55:39,  2.03it/s]
                                                      
{'loss': 3.7693, 'grad_norm': 0.20076662302017212, 'learning_rate': 0.0007987632733595299, 'epoch': 0.37}

 37%|███▋      | 3925/10708 [1:08:56<55:39,  2.03it/s]
 37%|███▋      | 3926/10708 [1:08:57<55:42,  2.03it/s]
 37%|███▋      | 3927/10708 [1:08:57<55:39,  2.03it/s]
 37%|███▋      | 3928/10708 [1:08:58<55:41,  2.03it/s]
 37%|███▋      | 3929/10708 [1:08:58<55:40,  2.03it/s]
 37%|███▋      | 3930/10708 [1:08:59<55:39,  2.03it/s]
 37%|███▋      | 3931/10708 [1:08:59<55:38,  2.03it/s]
 37%|███▋      | 3932/10708 [1:09:00<55:38,  2.03it/s]
 37%|███▋      | 3933/10708 [1:09:00<55:43,  2.03it/s]
 37%|███▋      | 3934/10708 [1:09:01<55:41,  2.03it/s]
 37%|███▋      | 3935/10708 [1:09:01<55:38,  2.03it/s]
 37%|███▋      | 3936/10708 [1:09:02<55:33,  2.03it/s]
 37%|███▋      | 3937/10708 [1:09:02<55:29,  2.03it/s]
 37%|███▋      | 3938/10708 [1:09:03<55:29,  2.03it/s]
 37%|███▋      | 3939/10708 [1:09:03<55:27,  2.03it/s]
 37%|███▋      | 3940/10708 [1:09:04<55:30,  2.03it/s]
 37%|███▋      | 3941/10708 [1:09:04<55:31,  2.03it/s]
 37%|███▋      | 3942/10708 [1:09:05<55:27,  2.03it/s]
 37%|███▋      | 3943/10708 [1:09:05<55:26,  2.03it/s]
 37%|███▋      | 3944/10708 [1:09:06<55:25,  2.03it/s]
 37%|███▋      | 3945/10708 [1:09:06<55:22,  2.04it/s]
 37%|███▋      | 3946/10708 [1:09:07<55:22,  2.04it/s]
 37%|███▋      | 3947/10708 [1:09:07<55:23,  2.03it/s]
 37%|███▋      | 3948/10708 [1:09:08<55:26,  2.03it/s]
 37%|███▋      | 3949/10708 [1:09:08<55:22,  2.03it/s]
 37%|███▋      | 3950/10708 [1:09:09<55:24,  2.03it/s]
                                                      
{'loss': 3.764, 'grad_norm': 0.21502697467803955, 'learning_rate': 0.0007954859246445964, 'epoch': 0.37}

 37%|███▋      | 3950/10708 [1:09:09<55:24,  2.03it/s]
 37%|███▋      | 3951/10708 [1:09:09<55:26,  2.03it/s]
 37%|███▋      | 3952/10708 [1:09:10<55:24,  2.03it/s]
 37%|███▋      | 3953/10708 [1:09:10<55:26,  2.03it/s]
 37%|███▋      | 3954/10708 [1:09:11<55:25,  2.03it/s]
 37%|███▋      | 3955/10708 [1:09:11<55:22,  2.03it/s]
 37%|███▋      | 3956/10708 [1:09:12<55:23,  2.03it/s]
 37%|███▋      | 3957/10708 [1:09:12<55:19,  2.03it/s]
 37%|███▋      | 3958/10708 [1:09:13<55:22,  2.03it/s]
 37%|███▋      | 3959/10708 [1:09:13<55:27,  2.03it/s]
 37%|███▋      | 3960/10708 [1:09:14<55:23,  2.03it/s]
 37%|███▋      | 3961/10708 [1:09:14<55:19,  2.03it/s]
 37%|███▋      | 3962/10708 [1:09:15<55:17,  2.03it/s]
 37%|███▋      | 3963/10708 [1:09:15<55:20,  2.03it/s]
 37%|███▋      | 3964/10708 [1:09:16<55:15,  2.03it/s]
 37%|███▋      | 3965/10708 [1:09:16<55:20,  2.03it/s]
 37%|███▋      | 3966/10708 [1:09:17<55:23,  2.03it/s]
 37%|███▋      | 3967/10708 [1:09:17<55:23,  2.03it/s]
 37%|███▋      | 3968/10708 [1:09:18<55:21,  2.03it/s]
 37%|███▋      | 3969/10708 [1:09:18<55:15,  2.03it/s]
 37%|███▋      | 3970/10708 [1:09:19<55:18,  2.03it/s]
 37%|███▋      | 3971/10708 [1:09:19<55:19,  2.03it/s]
 37%|███▋      | 3972/10708 [1:09:20<55:18,  2.03it/s]
 37%|███▋      | 3973/10708 [1:09:20<55:19,  2.03it/s]
 37%|███▋      | 3974/10708 [1:09:21<55:16,  2.03it/s]
 37%|███▋      | 3975/10708 [1:09:21<55:13,  2.03it/s]
                                                      {'loss': 3.7705, 'grad_norm': 0.20843757688999176, 'learning_rate': 0.0007921889499909092, 'epoch': 0.37}


 37%|███▋      | 3975/10708 [1:09:21<55:13,  2.03it/s]
 37%|███▋      | 3976/10708 [1:09:22<55:17,  2.03it/s]
 37%|███▋      | 3977/10708 [1:09:22<55:13,  2.03it/s]
 37%|███▋      | 3978/10708 [1:09:23<55:19,  2.03it/s]
 37%|███▋      | 3979/10708 [1:09:23<55:17,  2.03it/s]
 37%|███▋      | 3980/10708 [1:09:24<55:17,  2.03it/s]
 37%|███▋      | 3981/10708 [1:09:24<55:10,  2.03it/s]
 37%|███▋      | 3982/10708 [1:09:25<55:12,  2.03it/s]
 37%|███▋      | 3983/10708 [1:09:25<55:09,  2.03it/s]
 37%|███▋      | 3984/10708 [1:09:25<55:11,  2.03it/s]
 37%|███▋      | 3985/10708 [1:09:26<55:10,  2.03it/s]
 37%|███▋      | 3986/10708 [1:09:26<55:06,  2.03it/s]
 37%|███▋      | 3987/10708 [1:09:27<55:06,  2.03it/s]
 37%|███▋      | 3988/10708 [1:09:27<55:07,  2.03it/s]
 37%|███▋      | 3989/10708 [1:09:28<55:04,  2.03it/s]
 37%|███▋      | 3990/10708 [1:09:28<55:08,  2.03it/s]
 37%|███▋      | 3991/10708 [1:09:29<55:04,  2.03it/s]
 37%|███▋      | 3992/10708 [1:09:29<55:03,  2.03it/s]
 37%|███▋      | 3993/10708 [1:09:30<55:04,  2.03it/s]
 37%|███▋      | 3994/10708 [1:09:30<55:05,  2.03it/s]
 37%|███▋      | 3995/10708 [1:09:31<55:01,  2.03it/s]
 37%|███▋      | 3996/10708 [1:09:31<55:03,  2.03it/s]
 37%|███▋      | 3997/10708 [1:09:32<55:04,  2.03it/s]
 37%|███▋      | 3998/10708 [1:09:32<55:07,  2.03it/s]
 37%|███▋      | 3999/10708 [1:09:33<55:12,  2.03it/s]
 37%|███▋      | 4000/10708 [1:09:33<55:13,  2.02it/s]{'loss': 3.7682, 'grad_norm': 0.20502804219722748, 'learning_rate': 0.0007888725683808871, 'epoch': 0.37}

                                                      

 37%|███▋      | 4000/10708 [1:09:33<55:13,  2.02it/s]
 37%|███▋      | 4001/10708 [1:09:34<55:16,  2.02it/s]
 37%|███▋      | 4002/10708 [1:09:34<55:16,  2.02it/s]
 37%|███▋      | 4003/10708 [1:09:35<55:10,  2.03it/s]
 37%|███▋      | 4004/10708 [1:09:35<55:11,  2.02it/s]
 37%|███▋      | 4005/10708 [1:09:36<55:07,  2.03it/s]
 37%|███▋      | 4006/10708 [1:09:36<55:10,  2.02it/s]
 37%|███▋      | 4007/10708 [1:09:37<55:08,  2.03it/s]
 37%|███▋      | 4008/10708 [1:09:37<55:01,  2.03it/s]
 37%|███▋      | 4009/10708 [1:09:38<55:07,  2.03it/s]
 37%|███▋      | 4010/10708 [1:09:38<55:06,  2.03it/s]
 37%|███▋      | 4011/10708 [1:09:39<55:03,  2.03it/s]
 37%|███▋      | 4012/10708 [1:09:39<55:02,  2.03it/s]
 37%|███▋      | 4013/10708 [1:09:40<55:01,  2.03it/s]
 37%|███▋      | 4014/10708 [1:09:40<55:04,  2.03it/s]
 37%|███▋      | 4015/10708 [1:09:41<54:59,  2.03it/s]
 38%|███▊      | 4016/10708 [1:09:41<54:56,  2.03it/s]
 38%|███▊      | 4017/10708 [1:09:42<54:53,  2.03it/s]
 38%|███▊      | 4018/10708 [1:09:42<54:56,  2.03it/s]
 38%|███▊      | 4019/10708 [1:09:43<54:56,  2.03it/s]
 38%|███▊      | 4020/10708 [1:09:43<55:09,  2.02it/s]
 38%|███▊      | 4021/10708 [1:09:44<54:59,  2.03it/s]
 38%|███▊      | 4022/10708 [1:09:44<54:54,  2.03it/s]
 38%|███▊      | 4023/10708 [1:09:45<54:58,  2.03it/s]
 38%|███▊      | 4024/10708 [1:09:45<54:57,  2.03it/s]
 38%|███▊      | 4025/10708 [1:09:46<54:54,  2.03it/s]
                                                      
{'loss': 3.7539, 'grad_norm': 0.22189170122146606, 'learning_rate': 0.0007855370000859438, 'epoch': 0.38}

 38%|███▊      | 4025/10708 [1:09:46<54:54,  2.03it/s]
 38%|███▊      | 4026/10708 [1:09:46<55:03,  2.02it/s]
 38%|███▊      | 4027/10708 [1:09:47<55:03,  2.02it/s]
 38%|███▊      | 4028/10708 [1:09:47<55:02,  2.02it/s]
 38%|███▊      | 4029/10708 [1:09:48<55:02,  2.02it/s]
 38%|███▊      | 4030/10708 [1:09:48<54:54,  2.03it/s]
 38%|███▊      | 4031/10708 [1:09:49<54:56,  2.03it/s]
 38%|███▊      | 4032/10708 [1:09:49<54:52,  2.03it/s]
 38%|███▊      | 4033/10708 [1:09:50<54:48,  2.03it/s]
 38%|███▊      | 4034/10708 [1:09:50<54:45,  2.03it/s]
 38%|███▊      | 4035/10708 [1:09:51<54:42,  2.03it/s]
 38%|███▊      | 4036/10708 [1:09:51<54:43,  2.03it/s]
 38%|███▊      | 4037/10708 [1:09:52<54:41,  2.03it/s]
 38%|███▊      | 4038/10708 [1:09:52<54:42,  2.03it/s]
 38%|███▊      | 4039/10708 [1:09:53<54:52,  2.03it/s]
 38%|███▊      | 4040/10708 [1:09:53<54:47,  2.03it/s]
 38%|███▊      | 4041/10708 [1:09:54<54:41,  2.03it/s]
 38%|███▊      | 4042/10708 [1:09:54<54:42,  2.03it/s]
 38%|███▊      | 4043/10708 [1:09:55<54:48,  2.03it/s]
 38%|███▊      | 4044/10708 [1:09:55<54:46,  2.03it/s]
 38%|███▊      | 4045/10708 [1:09:56<54:42,  2.03it/s]
 38%|███▊      | 4046/10708 [1:09:56<54:39,  2.03it/s]
 38%|███▊      | 4047/10708 [1:09:57<54:35,  2.03it/s]
 38%|███▊      | 4048/10708 [1:09:57<54:34,  2.03it/s]
 38%|███▊      | 4049/10708 [1:09:58<54:32,  2.03it/s]
 38%|███▊      | 4050/10708 [1:09:58<54:33,  2.03it/s]
                                                      {'loss': 3.7622, 'grad_norm': 0.18867304921150208, 'learning_rate': 0.0007821824666518578, 'epoch': 0.38}

 38%|███▊      | 4050/10708 [1:09:58<54:33,  2.03it/s]
 38%|███▊      | 4051/10708 [1:09:59<54:43,  2.03it/s]
 38%|███▊      | 4052/10708 [1:09:59<54:42,  2.03it/s]
 38%|███▊      | 4053/10708 [1:10:00<54:46,  2.03it/s]
 38%|███▊      | 4054/10708 [1:10:00<54:43,  2.03it/s]
 38%|███▊      | 4055/10708 [1:10:00<54:40,  2.03it/s]
 38%|███▊      | 4056/10708 [1:10:01<54:41,  2.03it/s]
 38%|███▊      | 4057/10708 [1:10:01<54:49,  2.02it/s]
 38%|███▊      | 4058/10708 [1:10:02<54:44,  2.02it/s]
 38%|███▊      | 4059/10708 [1:10:02<54:42,  2.03it/s]
 38%|███▊      | 4060/10708 [1:10:03<54:41,  2.03it/s]
 38%|███▊      | 4061/10708 [1:10:03<54:36,  2.03it/s]
 38%|███▊      | 4062/10708 [1:10:04<54:32,  2.03it/s]
 38%|███▊      | 4063/10708 [1:10:04<54:33,  2.03it/s]
 38%|███▊      | 4064/10708 [1:10:05<54:29,  2.03it/s]
 38%|███▊      | 4065/10708 [1:10:05<54:27,  2.03it/s]
 38%|███▊      | 4066/10708 [1:10:06<54:29,  2.03it/s]
 38%|███▊      | 4067/10708 [1:10:06<54:36,  2.03it/s]
 38%|███▊      | 4068/10708 [1:10:07<54:39,  2.02it/s]
 38%|███▊      | 4069/10708 [1:10:07<54:38,  2.03it/s]
 38%|███▊      | 4070/10708 [1:10:08<54:34,  2.03it/s]
 38%|███▊      | 4071/10708 [1:10:08<54:34,  2.03it/s]
 38%|███▊      | 4072/10708 [1:10:09<54:32,  2.03it/s]
 38%|███▊      | 4073/10708 [1:10:09<54:29,  2.03it/s]
 38%|███▊      | 4074/10708 [1:10:10<54:34,  2.03it/s]
 38%|███▊      | 4075/10708 [1:10:10<54:28,  2.03it/s]{'loss': 3.7562, 'grad_norm': 0.20043686032295227, 'learning_rate': 0.000778809190884056, 'epoch': 0.38}

                                                      

 38%|███▊      | 4075/10708 [1:10:10<54:28,  2.03it/s]
 38%|███▊      | 4076/10708 [1:10:11<54:33,  2.03it/s]
 38%|███▊      | 4077/10708 [1:10:11<54:28,  2.03it/s]
 38%|███▊      | 4078/10708 [1:10:12<54:25,  2.03it/s]
 38%|███▊      | 4079/10708 [1:10:12<54:21,  2.03it/s]
 38%|███▊      | 4080/10708 [1:10:13<54:25,  2.03it/s]
 38%|███▊      | 4081/10708 [1:10:13<54:29,  2.03it/s]
 38%|███▊      | 4082/10708 [1:10:14<54:26,  2.03it/s]
 38%|███▊      | 4083/10708 [1:10:14<54:26,  2.03it/s]
 38%|███▊      | 4084/10708 [1:10:15<54:20,  2.03it/s]
 38%|███▊      | 4085/10708 [1:10:15<54:15,  2.03it/s]
 38%|███▊      | 4086/10708 [1:10:16<54:18,  2.03it/s]
 38%|███▊      | 4087/10708 [1:10:16<54:16,  2.03it/s]
 38%|███▊      | 4088/10708 [1:10:17<54:15,  2.03it/s]
 38%|███▊      | 4089/10708 [1:10:17<54:16,  2.03it/s]
 38%|███▊      | 4090/10708 [1:10:18<54:15,  2.03it/s]
 38%|███▊      | 4091/10708 [1:10:18<54:12,  2.03it/s]
 38%|███▊      | 4092/10708 [1:10:19<54:11,  2.03it/s]
 38%|███▊      | 4093/10708 [1:10:19<54:23,  2.03it/s]
 38%|███▊      | 4094/10708 [1:10:20<54:17,  2.03it/s]
 38%|███▊      | 4095/10708 [1:10:20<54:15,  2.03it/s]
 38%|███▊      | 4096/10708 [1:10:21<54:16,  2.03it/s]
 38%|███▊      | 4097/10708 [1:10:21<54:17,  2.03it/s]
 38%|███▊      | 4098/10708 [1:10:22<54:17,  2.03it/s]
 38%|███▊      | 4099/10708 [1:10:22<54:16,  2.03it/s]
 38%|███▊      | 4100/10708 [1:10:23<54:17,  2.03it/s]
                                                      
{'loss': 3.7579, 'grad_norm': 0.20425733923912048, 'learning_rate': 0.0007754173968328167, 'epoch': 0.38}

 38%|███▊      | 4100/10708 [1:10:23<54:17,  2.03it/s]
 38%|███▊      | 4101/10708 [1:10:23<54:19,  2.03it/s]
 38%|███▊      | 4102/10708 [1:10:24<54:19,  2.03it/s]
 38%|███▊      | 4103/10708 [1:10:24<54:18,  2.03it/s]
 38%|███▊      | 4104/10708 [1:10:25<54:11,  2.03it/s]
 38%|███▊      | 4105/10708 [1:10:25<54:09,  2.03it/s]
 38%|███▊      | 4106/10708 [1:10:26<54:08,  2.03it/s]
 38%|███▊      | 4107/10708 [1:10:26<54:05,  2.03it/s]
 38%|███▊      | 4108/10708 [1:10:27<54:06,  2.03it/s]
 38%|███▊      | 4109/10708 [1:10:27<54:07,  2.03it/s]
 38%|███▊      | 4110/10708 [1:10:28<54:03,  2.03it/s]
 38%|███▊      | 4111/10708 [1:10:28<54:10,  2.03it/s]
 38%|███▊      | 4112/10708 [1:10:29<54:05,  2.03it/s]
 38%|███▊      | 4113/10708 [1:10:29<54:05,  2.03it/s]
 38%|███▊      | 4114/10708 [1:10:30<54:09,  2.03it/s]
 38%|███▊      | 4115/10708 [1:10:30<54:04,  2.03it/s]
 38%|███▊      | 4116/10708 [1:10:31<54:08,  2.03it/s]
 38%|███▊      | 4117/10708 [1:10:31<54:04,  2.03it/s]
 38%|███▊      | 4118/10708 [1:10:32<54:08,  2.03it/s]
 38%|███▊      | 4119/10708 [1:10:32<54:04,  2.03it/s]
 38%|███▊      | 4120/10708 [1:10:33<54:01,  2.03it/s]
 38%|███▊      | 4121/10708 [1:10:33<54:05,  2.03it/s]
 38%|███▊      | 4122/10708 [1:10:33<54:01,  2.03it/s]
 39%|███▊      | 4123/10708 [1:10:34<54:00,  2.03it/s]
 39%|███▊      | 4124/10708 [1:10:34<54:00,  2.03it/s]
 39%|███▊      | 4125/10708 [1:10:35<53:56,  2.03it/s]{'loss': 3.7517, 'grad_norm': 0.20143786072731018, 'learning_rate': 0.0007720073097783878, 'epoch': 0.39}

                                                      

 39%|███▊      | 4125/10708 [1:10:35<53:56,  2.03it/s]
 39%|███▊      | 4126/10708 [1:10:35<54:02,  2.03it/s]
 39%|███▊      | 4127/10708 [1:10:36<53:59,  2.03it/s]
 39%|███▊      | 4128/10708 [1:10:36<53:54,  2.03it/s]
 39%|███▊      | 4129/10708 [1:10:37<53:58,  2.03it/s]
 39%|███▊      | 4130/10708 [1:10:37<54:01,  2.03it/s]
 39%|███▊      | 4131/10708 [1:10:38<54:03,  2.03it/s]
 39%|███▊      | 4132/10708 [1:10:38<53:59,  2.03it/s]
 39%|███▊      | 4133/10708 [1:10:39<54:03,  2.03it/s]
 39%|███▊      | 4134/10708 [1:10:39<53:58,  2.03it/s]
 39%|███▊      | 4135/10708 [1:10:40<53:54,  2.03it/s]
 39%|███▊      | 4136/10708 [1:10:40<53:58,  2.03it/s]
 39%|███▊      | 4137/10708 [1:10:41<53:57,  2.03it/s]
 39%|███▊      | 4138/10708 [1:10:41<54:04,  2.02it/s]
 39%|███▊      | 4139/10708 [1:10:42<54:13,  2.02it/s]
 39%|███▊      | 4140/10708 [1:10:42<54:07,  2.02it/s]
 39%|███▊      | 4141/10708 [1:10:43<54:16,  2.02it/s]
 39%|███▊      | 4142/10708 [1:10:43<54:15,  2.02it/s]
 39%|███▊      | 4143/10708 [1:10:44<54:07,  2.02it/s]
 39%|███▊      | 4144/10708 [1:10:44<54:04,  2.02it/s]
 39%|███▊      | 4145/10708 [1:10:45<53:58,  2.03it/s]
 39%|███▊      | 4146/10708 [1:10:45<54:02,  2.02it/s]
 39%|███▊      | 4147/10708 [1:10:46<53:59,  2.03it/s]
 39%|███▊      | 4148/10708 [1:10:46<53:54,  2.03it/s]
 39%|███▊      | 4149/10708 [1:10:47<53:51,  2.03it/s]
 39%|███▉      | 4150/10708 [1:10:47<53:52,  2.03it/s]
                                                      
{'loss': 3.7498, 'grad_norm': 0.22743378579616547, 'learning_rate': 0.0007685791562160235, 'epoch': 0.39}

 39%|███▉      | 4150/10708 [1:10:47<53:52,  2.03it/s]
 39%|███▉      | 4151/10708 [1:10:48<53:58,  2.02it/s]
 39%|███▉      | 4152/10708 [1:10:48<53:53,  2.03it/s]
 39%|███▉      | 4153/10708 [1:10:49<53:56,  2.03it/s]
 39%|███▉      | 4154/10708 [1:10:49<53:53,  2.03it/s]
 39%|███▉      | 4155/10708 [1:10:50<53:53,  2.03it/s]
 39%|███▉      | 4156/10708 [1:10:50<53:54,  2.03it/s]
 39%|███▉      | 4157/10708 [1:10:51<53:53,  2.03it/s]
 39%|███▉      | 4158/10708 [1:10:51<53:50,  2.03it/s]
 39%|███▉      | 4159/10708 [1:10:52<53:47,  2.03it/s]
 39%|███▉      | 4160/10708 [1:10:52<53:51,  2.03it/s]
 39%|███▉      | 4161/10708 [1:10:53<53:50,  2.03it/s]
 39%|███▉      | 4162/10708 [1:10:53<53:51,  2.03it/s]
 39%|███▉      | 4163/10708 [1:10:54<53:53,  2.02it/s]
 39%|███▉      | 4164/10708 [1:10:54<53:45,  2.03it/s]
 39%|███▉      | 4165/10708 [1:10:55<53:45,  2.03it/s]
 39%|███▉      | 4166/10708 [1:10:55<53:50,  2.02it/s]
 39%|███▉      | 4167/10708 [1:10:56<53:48,  2.03it/s]
 39%|███▉      | 4168/10708 [1:10:56<53:48,  2.03it/s]
 39%|███▉      | 4169/10708 [1:10:57<53:41,  2.03it/s]
 39%|███▉      | 4170/10708 [1:10:57<53:41,  2.03it/s]
 39%|███▉      | 4171/10708 [1:10:58<53:37,  2.03it/s]
 39%|███▉      | 4172/10708 [1:10:58<53:31,  2.04it/s]
 39%|███▉      | 4173/10708 [1:10:59<53:36,  2.03it/s]
 39%|███▉      | 4174/10708 [1:10:59<53:36,  2.03it/s]
 39%|███▉      | 4175/10708 [1:11:00<53:35,  2.03it/s]{'loss': 3.7493, 'grad_norm': 0.20892256498336792, 'learning_rate': 0.0007651331638409417, 'epoch': 0.39}

                                                      

 39%|███▉      | 4175/10708 [1:11:00<53:35,  2.03it/s]
 39%|███▉      | 4176/10708 [1:11:00<53:44,  2.03it/s]
 39%|███▉      | 4177/10708 [1:11:01<53:40,  2.03it/s]
 39%|███▉      | 4178/10708 [1:11:01<53:37,  2.03it/s]
 39%|███▉      | 4179/10708 [1:11:02<53:34,  2.03it/s]
 39%|███▉      | 4180/10708 [1:11:02<53:39,  2.03it/s]
 39%|███▉      | 4181/10708 [1:11:03<53:41,  2.03it/s]
 39%|███▉      | 4182/10708 [1:11:03<53:40,  2.03it/s]
 39%|███▉      | 4183/10708 [1:11:04<53:37,  2.03it/s]
 39%|███▉      | 4184/10708 [1:11:04<53:35,  2.03it/s]
 39%|███▉      | 4185/10708 [1:11:05<53:32,  2.03it/s]
 39%|███▉      | 4186/10708 [1:11:05<53:34,  2.03it/s]
 39%|███▉      | 4187/10708 [1:11:06<53:33,  2.03it/s]
 39%|███▉      | 4188/10708 [1:11:06<53:30,  2.03it/s]
 39%|███▉      | 4189/10708 [1:11:07<53:33,  2.03it/s]
 39%|███▉      | 4190/10708 [1:11:07<53:29,  2.03it/s]
 39%|███▉      | 4191/10708 [1:11:08<53:28,  2.03it/s]
 39%|███▉      | 4192/10708 [1:11:08<53:26,  2.03it/s]
 39%|███▉      | 4193/10708 [1:11:09<53:27,  2.03it/s]
 39%|███▉      | 4194/10708 [1:11:09<53:27,  2.03it/s]
 39%|███▉      | 4195/10708 [1:11:09<53:28,  2.03it/s]
 39%|███▉      | 4196/10708 [1:11:10<53:28,  2.03it/s]
 39%|███▉      | 4197/10708 [1:11:10<53:25,  2.03it/s]
 39%|███▉      | 4198/10708 [1:11:11<53:21,  2.03it/s]
 39%|███▉      | 4199/10708 [1:11:11<53:20,  2.03it/s]
 39%|███▉      | 4200/10708 [1:11:12<53:26,  2.03it/s]{'loss': 3.75, 'grad_norm': 0.2255452275276184, 'learning_rate': 0.000761669561533199, 'epoch': 0.39}

                                                      

 39%|███▉      | 4200/10708 [1:11:12<53:26,  2.03it/s]
 39%|███▉      | 4201/10708 [1:11:12<53:35,  2.02it/s]
 39%|███▉      | 4202/10708 [1:11:13<53:33,  2.02it/s]
 39%|███▉      | 4203/10708 [1:11:13<53:34,  2.02it/s]
 39%|███▉      | 4204/10708 [1:11:14<53:30,  2.03it/s]
 39%|███▉      | 4205/10708 [1:11:14<53:23,  2.03it/s]
 39%|███▉      | 4206/10708 [1:11:15<53:24,  2.03it/s]
 39%|███▉      | 4207/10708 [1:11:15<53:19,  2.03it/s]
 39%|███▉      | 4208/10708 [1:11:16<53:19,  2.03it/s]
 39%|███▉      | 4209/10708 [1:11:16<53:24,  2.03it/s]
 39%|███▉      | 4210/10708 [1:11:17<53:26,  2.03it/s]
 39%|███▉      | 4211/10708 [1:11:17<53:29,  2.02it/s]
 39%|███▉      | 4212/10708 [1:11:18<53:21,  2.03it/s]
 39%|███▉      | 4213/10708 [1:11:18<53:19,  2.03it/s]
 39%|███▉      | 4214/10708 [1:11:19<53:18,  2.03it/s]
 39%|███▉      | 4215/10708 [1:11:19<53:21,  2.03it/s]
 39%|███▉      | 4216/10708 [1:11:20<53:19,  2.03it/s]
 39%|███▉      | 4217/10708 [1:11:20<53:15,  2.03it/s]
 39%|███▉      | 4218/10708 [1:11:21<53:12,  2.03it/s]
 39%|███▉      | 4219/10708 [1:11:21<53:12,  2.03it/s]
 39%|███▉      | 4220/10708 [1:11:22<53:16,  2.03it/s]
 39%|███▉      | 4221/10708 [1:11:22<53:16,  2.03it/s]
 39%|███▉      | 4222/10708 [1:11:23<53:12,  2.03it/s]
 39%|███▉      | 4223/10708 [1:11:23<53:13,  2.03it/s]
 39%|███▉      | 4224/10708 [1:11:24<53:12,  2.03it/s]
 39%|███▉      | 4225/10708 [1:11:24<53:18,  2.03it/s]
                                                      {'loss': 3.7459, 'grad_norm': 0.2000829577445984, 'learning_rate': 0.0007581885793424901, 'epoch': 0.39}


 39%|███▉      | 4225/10708 [1:11:24<53:18,  2.03it/s]
 39%|███▉      | 4226/10708 [1:11:25<53:20,  2.03it/s]
 39%|███▉      | 4227/10708 [1:11:25<53:12,  2.03it/s]
 39%|███▉      | 4228/10708 [1:11:26<53:09,  2.03it/s]
 39%|███▉      | 4229/10708 [1:11:26<53:15,  2.03it/s]
 40%|███▉      | 4230/10708 [1:11:27<53:17,  2.03it/s]
 40%|███▉      | 4231/10708 [1:11:27<53:16,  2.03it/s]
 40%|███▉      | 4232/10708 [1:11:28<53:12,  2.03it/s]
 40%|███▉      | 4233/10708 [1:11:28<53:09,  2.03it/s]
 40%|███▉      | 4234/10708 [1:11:29<53:08,  2.03it/s]
 40%|███▉      | 4235/10708 [1:11:29<53:06,  2.03it/s]
 40%|███▉      | 4236/10708 [1:11:30<53:05,  2.03it/s]
 40%|███▉      | 4237/10708 [1:11:30<53:05,  2.03it/s]
 40%|███▉      | 4238/10708 [1:11:31<53:02,  2.03it/s]
 40%|███▉      | 4239/10708 [1:11:31<53:00,  2.03it/s]
 40%|███▉      | 4240/10708 [1:11:32<53:05,  2.03it/s]
 40%|███▉      | 4241/10708 [1:11:32<53:04,  2.03it/s]
 40%|███▉      | 4242/10708 [1:11:33<53:03,  2.03it/s]
 40%|███▉      | 4243/10708 [1:11:33<53:01,  2.03it/s]
 40%|███▉      | 4244/10708 [1:11:34<53:01,  2.03it/s]
 40%|███▉      | 4245/10708 [1:11:34<53:05,  2.03it/s]
 40%|███▉      | 4246/10708 [1:11:35<52:59,  2.03it/s]
 40%|███▉      | 4247/10708 [1:11:35<52:57,  2.03it/s]
 40%|███▉      | 4248/10708 [1:11:36<53:02,  2.03it/s]
 40%|███▉      | 4249/10708 [1:11:36<53:03,  2.03it/s]
 40%|███▉      | 4250/10708 [1:11:37<53:05,  2.03it/s]{'loss': 3.7429, 'grad_norm': 0.21603931486606598, 'learning_rate': 0.0007546904484728679, 'epoch': 0.4}

                                                      

 40%|███▉      | 4250/10708 [1:11:37<53:05,  2.03it/s]
 40%|███▉      | 4251/10708 [1:11:37<53:08,  2.03it/s]
 40%|███▉      | 4252/10708 [1:11:38<53:08,  2.02it/s]
 40%|███▉      | 4253/10708 [1:11:38<53:00,  2.03it/s]
 40%|███▉      | 4254/10708 [1:11:39<52:56,  2.03it/s]
 40%|███▉      | 4255/10708 [1:11:39<52:56,  2.03it/s]
 40%|███▉      | 4256/10708 [1:11:40<53:09,  2.02it/s]
 40%|███▉      | 4257/10708 [1:11:40<53:03,  2.03it/s]
 40%|███▉      | 4258/10708 [1:11:41<52:58,  2.03it/s]
 40%|███▉      | 4259/10708 [1:11:41<52:55,  2.03it/s]
 40%|███▉      | 4260/10708 [1:11:42<53:03,  2.03it/s]
 40%|███▉      | 4261/10708 [1:11:42<53:05,  2.02it/s]
 40%|███▉      | 4262/10708 [1:11:43<53:01,  2.03it/s]
 40%|███▉      | 4263/10708 [1:11:43<53:11,  2.02it/s]
 40%|███▉      | 4264/10708 [1:11:44<53:05,  2.02it/s]
 40%|███▉      | 4265/10708 [1:11:44<53:08,  2.02it/s]
 40%|███▉      | 4266/10708 [1:11:44<53:05,  2.02it/s]
 40%|███▉      | 4267/10708 [1:11:45<52:58,  2.03it/s]
 40%|███▉      | 4268/10708 [1:11:45<52:56,  2.03it/s]
 40%|███▉      | 4269/10708 [1:11:46<52:55,  2.03it/s]
 40%|███▉      | 4270/10708 [1:11:46<52:53,  2.03it/s]
 40%|███▉      | 4271/10708 [1:11:47<52:48,  2.03it/s]
 40%|███▉      | 4272/10708 [1:11:47<52:45,  2.03it/s]
 40%|███▉      | 4273/10708 [1:11:48<52:49,  2.03it/s]
 40%|███▉      | 4274/10708 [1:11:48<52:47,  2.03it/s]
 40%|███▉      | 4275/10708 [1:11:49<52:39,  2.04it/s]
                                                      
{'loss': 3.7398, 'grad_norm': 0.21388515830039978, 'learning_rate': 0.0007511754012673866, 'epoch': 0.4}

 40%|███▉      | 4275/10708 [1:11:49<52:39,  2.04it/s]
 40%|███▉      | 4276/10708 [1:11:49<52:49,  2.03it/s]
 40%|███▉      | 4277/10708 [1:11:50<52:45,  2.03it/s]
 40%|███▉      | 4278/10708 [1:11:50<52:48,  2.03it/s]
 40%|███▉      | 4279/10708 [1:11:51<52:42,  2.03it/s]
 40%|███▉      | 4280/10708 [1:11:51<52:54,  2.02it/s]
 40%|███▉      | 4281/10708 [1:11:52<52:52,  2.03it/s]
 40%|███▉      | 4282/10708 [1:11:52<52:48,  2.03it/s]
 40%|███▉      | 4283/10708 [1:11:53<52:51,  2.03it/s]
 40%|████      | 4284/10708 [1:11:53<52:49,  2.03it/s]
 40%|████      | 4285/10708 [1:11:54<52:53,  2.02it/s]
 40%|████      | 4286/10708 [1:11:54<52:48,  2.03it/s]
 40%|████      | 4287/10708 [1:11:55<52:46,  2.03it/s]
 40%|████      | 4288/10708 [1:11:55<52:44,  2.03it/s]
 40%|████      | 4289/10708 [1:11:56<52:41,  2.03it/s]
 40%|████      | 4290/10708 [1:11:56<52:41,  2.03it/s]
 40%|████      | 4291/10708 [1:11:57<52:42,  2.03it/s]
 40%|████      | 4292/10708 [1:11:57<52:49,  2.02it/s]
 40%|████      | 4293/10708 [1:11:58<52:42,  2.03it/s]
 40%|████      | 4294/10708 [1:11:58<52:39,  2.03it/s]
 40%|████      | 4295/10708 [1:11:59<52:44,  2.03it/s]
 40%|████      | 4296/10708 [1:11:59<52:42,  2.03it/s]
 40%|████      | 4297/10708 [1:12:00<52:41,  2.03it/s]
 40%|████      | 4298/10708 [1:12:00<52:37,  2.03it/s]
 40%|████      | 4299/10708 [1:12:01<52:38,  2.03it/s]
 40%|████      | 4300/10708 [1:12:01<52:45,  2.02it/s]
                                                      {'loss': 3.7391, 'grad_norm': 0.2087078094482422, 'learning_rate': 0.0007476436711926697, 'epoch': 0.4}


 40%|████      | 4300/10708 [1:12:01<52:45,  2.02it/s]
 40%|████      | 4301/10708 [1:12:02<52:54,  2.02it/s]
 40%|████      | 4302/10708 [1:12:02<52:42,  2.03it/s]
 40%|████      | 4303/10708 [1:12:03<52:38,  2.03it/s]
 40%|████      | 4304/10708 [1:12:03<52:37,  2.03it/s]
 40%|████      | 4305/10708 [1:12:04<52:33,  2.03it/s]
 40%|████      | 4306/10708 [1:12:04<52:34,  2.03it/s]
 40%|████      | 4307/10708 [1:12:05<52:37,  2.03it/s]
 40%|████      | 4308/10708 [1:12:05<52:32,  2.03it/s]
 40%|████      | 4309/10708 [1:12:06<52:31,  2.03it/s]
 40%|████      | 4310/10708 [1:12:06<52:36,  2.03it/s]
 40%|████      | 4311/10708 [1:12:07<52:32,  2.03it/s]
 40%|████      | 4312/10708 [1:12:07<52:32,  2.03it/s]
 40%|████      | 4313/10708 [1:12:08<52:40,  2.02it/s]
 40%|████      | 4314/10708 [1:12:08<52:39,  2.02it/s]
 40%|████      | 4315/10708 [1:12:09<52:34,  2.03it/s]
 40%|████      | 4316/10708 [1:12:09<52:38,  2.02it/s]
 40%|████      | 4317/10708 [1:12:10<52:33,  2.03it/s]
 40%|████      | 4318/10708 [1:12:10<52:32,  2.03it/s]
 40%|████      | 4319/10708 [1:12:11<52:34,  2.03it/s]
 40%|████      | 4320/10708 [1:12:11<52:34,  2.02it/s]
 40%|████      | 4321/10708 [1:12:12<52:30,  2.03it/s]
 40%|████      | 4322/10708 [1:12:12<52:28,  2.03it/s]
 40%|████      | 4323/10708 [1:12:13<52:35,  2.02it/s]
 40%|████      | 4324/10708 [1:12:13<52:38,  2.02it/s]
 40%|████      | 4325/10708 [1:12:14<52:29,  2.03it/s]{'loss': 3.7412, 'grad_norm': 0.20429947972297668, 'learning_rate': 0.0007440954928234037, 'epoch': 0.4}

                                                      

 40%|████      | 4325/10708 [1:12:14<52:29,  2.03it/s]
 40%|████      | 4326/10708 [1:12:14<52:32,  2.02it/s]
 40%|████      | 4327/10708 [1:12:15<52:25,  2.03it/s]
 40%|████      | 4328/10708 [1:12:15<52:29,  2.03it/s]
 40%|████      | 4329/10708 [1:12:16<52:29,  2.03it/s]
 40%|████      | 4330/10708 [1:12:16<52:24,  2.03it/s]
 40%|████      | 4331/10708 [1:12:17<52:27,  2.03it/s]
 40%|████      | 4332/10708 [1:12:17<52:24,  2.03it/s]
 40%|████      | 4333/10708 [1:12:18<52:27,  2.03it/s]
 40%|████      | 4334/10708 [1:12:18<52:22,  2.03it/s]
 40%|████      | 4335/10708 [1:12:19<52:16,  2.03it/s]
 40%|████      | 4336/10708 [1:12:19<52:18,  2.03it/s]
 41%|████      | 4337/10708 [1:12:19<52:11,  2.03it/s]
 41%|████      | 4338/10708 [1:12:20<52:08,  2.04it/s]
 41%|████      | 4339/10708 [1:12:20<52:10,  2.03it/s]
 41%|████      | 4340/10708 [1:12:21<52:10,  2.03it/s]
 41%|████      | 4341/10708 [1:12:21<52:14,  2.03it/s]
 41%|████      | 4342/10708 [1:12:22<52:13,  2.03it/s]
 41%|████      | 4343/10708 [1:12:22<52:11,  2.03it/s]
 41%|████      | 4344/10708 [1:12:23<52:11,  2.03it/s]
 41%|████      | 4345/10708 [1:12:23<52:10,  2.03it/s]
 41%|████      | 4346/10708 [1:12:24<52:06,  2.03it/s]
 41%|████      | 4347/10708 [1:12:24<52:11,  2.03it/s]
 41%|████      | 4348/10708 [1:12:25<52:10,  2.03it/s]
 41%|████      | 4349/10708 [1:12:25<52:06,  2.03it/s]
 41%|████      | 4350/10708 [1:12:26<52:10,  2.03it/s]
                                                      {'loss': 3.7396, 'grad_norm': 0.19755712151527405, 'learning_rate': 0.0007405311018267577, 'epoch': 0.41}

 41%|████      | 4350/10708 [1:12:26<52:10,  2.03it/s]
 41%|████      | 4351/10708 [1:12:26<52:13,  2.03it/s]
 41%|████      | 4352/10708 [1:12:27<52:11,  2.03it/s]
 41%|████      | 4353/10708 [1:12:27<52:11,  2.03it/s]
 41%|████      | 4354/10708 [1:12:28<52:08,  2.03it/s]
 41%|████      | 4355/10708 [1:12:28<52:05,  2.03it/s]
 41%|████      | 4356/10708 [1:12:29<52:08,  2.03it/s]
 41%|████      | 4357/10708 [1:12:29<52:07,  2.03it/s]
 41%|████      | 4358/10708 [1:12:30<52:07,  2.03it/s]
 41%|████      | 4359/10708 [1:12:30<52:05,  2.03it/s]
 41%|████      | 4360/10708 [1:12:31<52:06,  2.03it/s]
 41%|████      | 4361/10708 [1:12:31<52:14,  2.02it/s]
 41%|████      | 4362/10708 [1:12:32<52:12,  2.03it/s]
 41%|████      | 4363/10708 [1:12:32<52:07,  2.03it/s]
 41%|████      | 4364/10708 [1:12:33<52:07,  2.03it/s]
 41%|████      | 4365/10708 [1:12:33<52:07,  2.03it/s]
 41%|████      | 4366/10708 [1:12:34<52:02,  2.03it/s]
 41%|████      | 4367/10708 [1:12:34<52:04,  2.03it/s]
 41%|████      | 4368/10708 [1:12:35<52:04,  2.03it/s]
 41%|████      | 4369/10708 [1:12:35<51:58,  2.03it/s]
 41%|████      | 4370/10708 [1:12:36<52:02,  2.03it/s]
 41%|████      | 4371/10708 [1:12:36<52:01,  2.03it/s]
 41%|████      | 4372/10708 [1:12:37<51:59,  2.03it/s]
 41%|████      | 4373/10708 [1:12:37<51:57,  2.03it/s]
 41%|████      | 4374/10708 [1:12:38<51:53,  2.03it/s]
 41%|████      | 4375/10708 [1:12:38<51:56,  2.03it/s]
                                                      
{'loss': 3.7321, 'grad_norm': 0.20076099038124084, 'learning_rate': 0.0007369507349467307, 'epoch': 0.41}

 41%|████      | 4375/10708 [1:12:38<51:56,  2.03it/s]
 41%|████      | 4376/10708 [1:12:39<52:02,  2.03it/s]
 41%|████      | 4377/10708 [1:12:39<51:59,  2.03it/s]
 41%|████      | 4378/10708 [1:12:40<51:56,  2.03it/s]
 41%|████      | 4379/10708 [1:12:40<51:56,  2.03it/s]
 41%|████      | 4380/10708 [1:12:41<51:59,  2.03it/s]
 41%|████      | 4381/10708 [1:12:41<51:57,  2.03it/s]
 41%|████      | 4382/10708 [1:12:42<51:52,  2.03it/s]
 41%|████      | 4383/10708 [1:12:42<51:58,  2.03it/s]
 41%|████      | 4384/10708 [1:12:43<51:58,  2.03it/s]
 41%|████      | 4385/10708 [1:12:43<51:58,  2.03it/s]
 41%|████      | 4386/10708 [1:12:44<51:56,  2.03it/s]
 41%|████      | 4387/10708 [1:12:44<51:53,  2.03it/s]
 41%|████      | 4388/10708 [1:12:45<51:54,  2.03it/s]
 41%|████      | 4389/10708 [1:12:45<51:49,  2.03it/s]
 41%|████      | 4390/10708 [1:12:46<51:53,  2.03it/s]
 41%|████      | 4391/10708 [1:12:46<51:54,  2.03it/s]
 41%|████      | 4392/10708 [1:12:47<51:50,  2.03it/s]
 41%|████      | 4393/10708 [1:12:47<51:54,  2.03it/s]
 41%|████      | 4394/10708 [1:12:48<51:54,  2.03it/s]
 41%|████      | 4395/10708 [1:12:48<51:48,  2.03it/s]
 41%|████      | 4396/10708 [1:12:49<51:49,  2.03it/s]
 41%|████      | 4397/10708 [1:12:49<51:43,  2.03it/s]
 41%|████      | 4398/10708 [1:12:50<58:25,  1.80it/s]
 41%|████      | 4399/10708 [1:12:50<56:18,  1.87it/s]
 41%|████      | 4400/10708 [1:12:51<54:58,  1.91it/s]{'loss': 3.7325, 'grad_norm': 0.20862436294555664, 'learning_rate': 0.0007333546299884268, 'epoch': 0.41}

                                                      

 41%|████      | 4400/10708 [1:12:51<54:58,  1.91it/s]
 41%|████      | 4401/10708 [1:12:51<54:03,  1.94it/s]
 41%|████      | 4402/10708 [1:12:52<53:22,  1.97it/s]
 41%|████      | 4403/10708 [1:12:52<52:55,  1.99it/s]
 41%|████      | 4404/10708 [1:12:53<52:30,  2.00it/s]
 41%|████      | 4405/10708 [1:12:53<52:16,  2.01it/s]
 41%|████      | 4406/10708 [1:12:54<52:05,  2.02it/s]
 41%|████      | 4407/10708 [1:12:54<51:57,  2.02it/s]
 41%|████      | 4408/10708 [1:12:55<51:54,  2.02it/s]
 41%|████      | 4409/10708 [1:12:55<51:47,  2.03it/s]
 41%|████      | 4410/10708 [1:12:56<51:51,  2.02it/s]
 41%|████      | 4411/10708 [1:12:56<51:48,  2.03it/s]
 41%|████      | 4412/10708 [1:12:57<52:15,  2.01it/s]
 41%|████      | 4413/10708 [1:12:57<52:02,  2.02it/s]
 41%|████      | 4414/10708 [1:12:58<51:57,  2.02it/s]
 41%|████      | 4415/10708 [1:12:58<51:47,  2.03it/s]
 41%|████      | 4416/10708 [1:12:59<51:47,  2.02it/s]
 41%|████      | 4417/10708 [1:12:59<51:39,  2.03it/s]
 41%|████▏     | 4418/10708 [1:13:00<51:33,  2.03it/s]
 41%|████▏     | 4419/10708 [1:13:00<51:36,  2.03it/s]
 41%|████▏     | 4420/10708 [1:13:01<51:37,  2.03it/s]
 41%|████▏     | 4421/10708 [1:13:01<51:35,  2.03it/s]
 41%|████▏     | 4422/10708 [1:13:02<51:35,  2.03it/s]
 41%|████▏     | 4423/10708 [1:13:02<51:39,  2.03it/s]
 41%|████▏     | 4424/10708 [1:13:03<51:41,  2.03it/s]
 41%|████▏     | 4425/10708 [1:13:03<51:36,  2.03it/s]{'loss': 3.7319, 'grad_norm': 0.20518772304058075, 'learning_rate': 0.0007297430258022609, 'epoch': 0.41}

                                                      

 41%|████▏     | 4425/10708 [1:13:03<51:36,  2.03it/s]
 41%|████▏     | 4426/10708 [1:13:04<51:37,  2.03it/s]
 41%|████▏     | 4427/10708 [1:13:04<51:35,  2.03it/s]
 41%|████▏     | 4428/10708 [1:13:05<51:34,  2.03it/s]
 41%|████▏     | 4429/10708 [1:13:05<51:37,  2.03it/s]
 41%|████▏     | 4430/10708 [1:13:06<51:35,  2.03it/s]
 41%|████▏     | 4431/10708 [1:13:06<51:32,  2.03it/s]
 41%|████▏     | 4432/10708 [1:13:07<51:28,  2.03it/s]
 41%|████▏     | 4433/10708 [1:13:07<51:26,  2.03it/s]
 41%|████▏     | 4434/10708 [1:13:07<51:31,  2.03it/s]
 41%|████▏     | 4435/10708 [1:13:08<51:45,  2.02it/s]
 41%|████▏     | 4436/10708 [1:13:08<51:39,  2.02it/s]
 41%|████▏     | 4437/10708 [1:13:09<51:29,  2.03it/s]
 41%|████▏     | 4438/10708 [1:13:09<51:33,  2.03it/s]
 41%|████▏     | 4439/10708 [1:13:10<51:30,  2.03it/s]
 41%|████▏     | 4440/10708 [1:13:10<51:25,  2.03it/s]
 41%|████▏     | 4441/10708 [1:13:11<51:41,  2.02it/s]
 41%|████▏     | 4442/10708 [1:13:11<51:34,  2.02it/s]
 41%|████▏     | 4443/10708 [1:13:12<51:29,  2.03it/s]
 42%|████▏     | 4444/10708 [1:13:12<51:28,  2.03it/s]
 42%|████▏     | 4445/10708 [1:13:13<51:27,  2.03it/s]
 42%|████▏     | 4446/10708 [1:13:13<51:31,  2.03it/s]
 42%|████▏     | 4447/10708 [1:13:14<51:24,  2.03it/s]
 42%|████▏     | 4448/10708 [1:13:14<51:22,  2.03it/s]
 42%|████▏     | 4449/10708 [1:13:15<51:21,  2.03it/s]
 42%|████▏     | 4450/10708 [1:13:15<51:21,  2.03it/s]{'loss': 3.7324, 'grad_norm': 0.19908738136291504, 'learning_rate': 0.0007261161622680937, 'epoch': 0.42}

                                                      

 42%|████▏     | 4450/10708 [1:13:15<51:21,  2.03it/s]
 42%|████▏     | 4451/10708 [1:13:16<51:28,  2.03it/s]
 42%|████▏     | 4452/10708 [1:13:16<51:25,  2.03it/s]
 42%|████▏     | 4453/10708 [1:13:17<51:29,  2.02it/s]
 42%|████▏     | 4454/10708 [1:13:17<51:29,  2.02it/s]
 42%|████▏     | 4455/10708 [1:13:18<51:22,  2.03it/s]
 42%|████▏     | 4456/10708 [1:13:18<51:23,  2.03it/s]
 42%|████▏     | 4457/10708 [1:13:19<51:20,  2.03it/s]
 42%|████▏     | 4458/10708 [1:13:19<51:27,  2.02it/s]
 42%|████▏     | 4459/10708 [1:13:20<51:25,  2.02it/s]
 42%|████▏     | 4460/10708 [1:13:20<51:21,  2.03it/s]
 42%|████▏     | 4461/10708 [1:13:21<51:19,  2.03it/s]
 42%|████▏     | 4462/10708 [1:13:21<51:16,  2.03it/s]
 42%|████▏     | 4463/10708 [1:13:22<51:20,  2.03it/s]
 42%|████▏     | 4464/10708 [1:13:22<51:19,  2.03it/s]
 42%|████▏     | 4465/10708 [1:13:23<51:21,  2.03it/s]
 42%|████▏     | 4466/10708 [1:13:23<51:20,  2.03it/s]
 42%|████▏     | 4467/10708 [1:13:24<51:17,  2.03it/s]
 42%|████▏     | 4468/10708 [1:13:24<51:14,  2.03it/s]
 42%|████▏     | 4469/10708 [1:13:25<51:10,  2.03it/s]
 42%|████▏     | 4470/10708 [1:13:25<51:08,  2.03it/s]
 42%|████▏     | 4471/10708 [1:13:26<51:15,  2.03it/s]
 42%|████▏     | 4472/10708 [1:13:26<51:11,  2.03it/s]
 42%|████▏     | 4473/10708 [1:13:27<51:09,  2.03it/s]
 42%|████▏     | 4474/10708 [1:13:27<51:07,  2.03it/s]
 42%|████▏     | 4475/10708 [1:13:28<51:04,  2.03it/s]
                                                      {'loss': 3.7344, 'grad_norm': 0.19804130494594574, 'learning_rate': 0.0007224742802793004, 'epoch': 0.42}


 42%|████▏     | 4475/10708 [1:13:28<51:04,  2.03it/s]
 42%|████▏     | 4476/10708 [1:13:28<51:14,  2.03it/s]
 42%|████▏     | 4477/10708 [1:13:29<51:09,  2.03it/s]
 42%|████▏     | 4478/10708 [1:13:29<51:08,  2.03it/s]
 42%|████▏     | 4479/10708 [1:13:30<51:06,  2.03it/s]
 42%|████▏     | 4480/10708 [1:13:30<51:04,  2.03it/s]
 42%|████▏     | 4481/10708 [1:13:31<51:10,  2.03it/s]
 42%|████▏     | 4482/10708 [1:13:31<51:05,  2.03it/s]
 42%|████▏     | 4483/10708 [1:13:32<51:03,  2.03it/s]
 42%|████▏     | 4484/10708 [1:13:32<51:08,  2.03it/s]
 42%|████▏     | 4485/10708 [1:13:33<51:01,  2.03it/s]
 42%|████▏     | 4486/10708 [1:13:33<51:06,  2.03it/s]
 42%|████▏     | 4487/10708 [1:13:34<51:04,  2.03it/s]
 42%|████▏     | 4488/10708 [1:13:34<50:59,  2.03it/s]
 42%|████▏     | 4489/10708 [1:13:35<50:59,  2.03it/s]
 42%|████▏     | 4490/10708 [1:13:35<51:01,  2.03it/s]
 42%|████▏     | 4491/10708 [1:13:36<50:56,  2.03it/s]
 42%|████▏     | 4492/10708 [1:13:36<50:55,  2.03it/s]
 42%|████▏     | 4493/10708 [1:13:37<51:02,  2.03it/s]
 42%|████▏     | 4494/10708 [1:13:37<51:19,  2.02it/s]
 42%|████▏     | 4495/10708 [1:13:38<51:11,  2.02it/s]
 42%|████▏     | 4496/10708 [1:13:38<51:08,  2.02it/s]
 42%|████▏     | 4497/10708 [1:13:39<51:10,  2.02it/s]
 42%|████▏     | 4498/10708 [1:13:39<51:04,  2.03it/s]
 42%|████▏     | 4499/10708 [1:13:40<50:59,  2.03it/s]
 42%|████▏     | 4500/10708 [1:13:40<51:00,  2.03it/s]{'loss': 3.7269, 'grad_norm': 0.2237486094236374, 'learning_rate': 0.0007188176217267694, 'epoch': 0.42}

                                                      

 42%|████▏     | 4500/10708 [1:13:40<51:00,  2.03it/s]
 42%|████▏     | 4501/10708 [1:13:41<51:03,  2.03it/s]
 42%|████▏     | 4502/10708 [1:13:41<50:59,  2.03it/s]
 42%|████▏     | 4503/10708 [1:13:42<50:57,  2.03it/s]
 42%|████▏     | 4504/10708 [1:13:42<50:56,  2.03it/s]
 42%|████▏     | 4505/10708 [1:13:43<50:53,  2.03it/s]
 42%|████▏     | 4506/10708 [1:13:43<51:01,  2.03it/s]
 42%|████▏     | 4507/10708 [1:13:43<50:58,  2.03it/s]
 42%|████▏     | 4508/10708 [1:13:44<51:01,  2.02it/s]
 42%|████▏     | 4509/10708 [1:13:44<50:53,  2.03it/s]
 42%|████▏     | 4510/10708 [1:13:45<50:57,  2.03it/s]
 42%|████▏     | 4511/10708 [1:13:45<50:55,  2.03it/s]
 42%|████▏     | 4512/10708 [1:13:46<50:48,  2.03it/s]
 42%|████▏     | 4513/10708 [1:13:46<50:49,  2.03it/s]
 42%|████▏     | 4514/10708 [1:13:47<50:50,  2.03it/s]
 42%|████▏     | 4515/10708 [1:13:47<50:48,  2.03it/s]
 42%|████▏     | 4516/10708 [1:13:48<50:50,  2.03it/s]
 42%|████▏     | 4517/10708 [1:13:48<50:45,  2.03it/s]
 42%|████▏     | 4518/10708 [1:13:49<50:47,  2.03it/s]
 42%|████▏     | 4519/10708 [1:13:49<50:45,  2.03it/s]
 42%|████▏     | 4520/10708 [1:13:50<50:51,  2.03it/s]
 42%|████▏     | 4521/10708 [1:13:50<50:49,  2.03it/s]
 42%|████▏     | 4522/10708 [1:13:51<50:47,  2.03it/s]
 42%|████▏     | 4523/10708 [1:13:51<50:46,  2.03it/s]
 42%|████▏     | 4524/10708 [1:13:52<50:46,  2.03it/s]
 42%|████▏     | 4525/10708 [1:13:52<50:42,  2.03it/s]{'loss': 3.7178, 'grad_norm': 0.20719149708747864, 'learning_rate': 0.0007151464294828369, 'epoch': 0.42}
                                                      

 42%|████▏     | 4525/10708 [1:13:52<50:42,  2.03it/s]
 42%|████▏     | 4526/10708 [1:13:53<50:50,  2.03it/s]
 42%|████▏     | 4527/10708 [1:13:53<50:47,  2.03it/s]
 42%|████▏     | 4528/10708 [1:13:54<50:47,  2.03it/s]
 42%|████▏     | 4529/10708 [1:13:54<50:44,  2.03it/s]
 42%|████▏     | 4530/10708 [1:13:55<50:40,  2.03it/s]
 42%|████▏     | 4531/10708 [1:13:55<50:40,  2.03it/s]
 42%|████▏     | 4532/10708 [1:13:56<50:39,  2.03it/s]
 42%|████▏     | 4533/10708 [1:13:56<50:43,  2.03it/s]
 42%|████▏     | 4534/10708 [1:13:57<50:39,  2.03it/s]
 42%|████▏     | 4535/10708 [1:13:57<50:38,  2.03it/s]
 42%|████▏     | 4536/10708 [1:13:58<50:37,  2.03it/s]
 42%|████▏     | 4537/10708 [1:13:58<50:40,  2.03it/s]
 42%|████▏     | 4538/10708 [1:13:59<50:41,  2.03it/s]
 42%|████▏     | 4539/10708 [1:13:59<50:39,  2.03it/s]
 42%|████▏     | 4540/10708 [1:14:00<50:38,  2.03it/s]
 42%|████▏     | 4541/10708 [1:14:00<50:37,  2.03it/s]
 42%|████▏     | 4542/10708 [1:14:01<50:37,  2.03it/s]
 42%|████▏     | 4543/10708 [1:14:01<50:36,  2.03it/s]
 42%|████▏     | 4544/10708 [1:14:02<50:35,  2.03it/s]
 42%|████▏     | 4545/10708 [1:14:02<50:35,  2.03it/s]
 42%|████▏     | 4546/10708 [1:14:03<50:32,  2.03it/s]
 42%|████▏     | 4547/10708 [1:14:03<50:32,  2.03it/s]
 42%|████▏     | 4548/10708 [1:14:04<50:29,  2.03it/s]
 42%|████▏     | 4549/10708 [1:14:04<50:31,  2.03it/s]
 42%|████▏     | 4550/10708 [1:14:05<50:31,  2.03it/s]
                                                      {'loss': 3.7275, 'grad_norm': 0.20492514967918396, 'learning_rate': 0.0007114609473851548, 'epoch': 0.42}


 42%|████▏     | 4550/10708 [1:14:05<50:31,  2.03it/s]
 43%|████▎     | 4551/10708 [1:14:05<50:32,  2.03it/s]
 43%|████▎     | 4552/10708 [1:14:06<50:33,  2.03it/s]
 43%|████▎     | 4553/10708 [1:14:06<50:27,  2.03it/s]
 43%|████▎     | 4554/10708 [1:14:07<50:25,  2.03it/s]
 43%|████▎     | 4555/10708 [1:14:07<50:28,  2.03it/s]
 43%|████▎     | 4556/10708 [1:14:08<50:25,  2.03it/s]
 43%|████▎     | 4557/10708 [1:14:08<50:22,  2.03it/s]
 43%|████▎     | 4558/10708 [1:14:09<50:22,  2.03it/s]
 43%|████▎     | 4559/10708 [1:14:09<50:24,  2.03it/s]
 43%|████▎     | 4560/10708 [1:14:10<50:25,  2.03it/s]
 43%|████▎     | 4561/10708 [1:14:10<50:28,  2.03it/s]
 43%|████▎     | 4562/10708 [1:14:11<50:27,  2.03it/s]
 43%|████▎     | 4563/10708 [1:14:11<50:23,  2.03it/s]
 43%|████▎     | 4564/10708 [1:14:12<50:32,  2.03it/s]
 43%|████▎     | 4565/10708 [1:14:12<50:30,  2.03it/s]
 43%|████▎     | 4566/10708 [1:14:13<50:26,  2.03it/s]
 43%|████▎     | 4567/10708 [1:14:13<50:31,  2.03it/s]
 43%|████▎     | 4568/10708 [1:14:14<50:28,  2.03it/s]
 43%|████▎     | 4569/10708 [1:14:14<50:28,  2.03it/s]
 43%|████▎     | 4570/10708 [1:14:15<50:30,  2.03it/s]
 43%|████▎     | 4571/10708 [1:14:15<50:25,  2.03it/s]
 43%|████▎     | 4572/10708 [1:14:16<58:55,  1.74it/s]
 43%|████▎     | 4573/10708 [1:14:16<56:24,  1.81it/s]
 43%|████▎     | 4574/10708 [1:14:17<54:32,  1.87it/s]
 43%|████▎     | 4575/10708 [1:14:17<53:11,  1.92it/s]
                                                      
{'loss': 3.726, 'grad_norm': 0.2284766137599945, 'learning_rate': 0.0007077614202204965, 'epoch': 0.43}

 43%|████▎     | 4575/10708 [1:14:17<53:11,  1.92it/s]
 43%|████▎     | 4576/10708 [1:14:18<52:22,  1.95it/s]
 43%|████▎     | 4577/10708 [1:14:18<51:42,  1.98it/s]
 43%|████▎     | 4578/10708 [1:14:19<51:24,  1.99it/s]
 43%|████▎     | 4579/10708 [1:14:19<51:01,  2.00it/s]
 43%|████▎     | 4580/10708 [1:14:20<50:44,  2.01it/s]
 43%|████▎     | 4581/10708 [1:14:20<50:36,  2.02it/s]
 43%|████▎     | 4582/10708 [1:14:21<50:28,  2.02it/s]
 43%|████▎     | 4583/10708 [1:14:21<50:20,  2.03it/s]
 43%|████▎     | 4584/10708 [1:14:22<50:21,  2.03it/s]
 43%|████▎     | 4585/10708 [1:14:22<50:30,  2.02it/s]
 43%|████▎     | 4586/10708 [1:14:23<50:38,  2.01it/s]
 43%|████▎     | 4587/10708 [1:14:23<50:27,  2.02it/s]
 43%|████▎     | 4588/10708 [1:14:24<50:26,  2.02it/s]
 43%|████▎     | 4589/10708 [1:14:24<50:19,  2.03it/s]
 43%|████▎     | 4590/10708 [1:14:25<50:15,  2.03it/s]
 43%|████▎     | 4591/10708 [1:14:25<58:24,  1.75it/s]
 43%|████▎     | 4592/10708 [1:14:26<55:52,  1.82it/s]
 43%|████▎     | 4593/10708 [1:14:26<54:07,  1.88it/s]
 43%|████▎     | 4594/10708 [1:14:27<52:58,  1.92it/s]
 43%|████▎     | 4595/10708 [1:14:27<52:02,  1.96it/s]
 43%|████▎     | 4596/10708 [1:14:28<51:27,  1.98it/s]
 43%|████▎     | 4597/10708 [1:14:28<51:00,  2.00it/s]
 43%|████▎     | 4598/10708 [1:14:29<50:48,  2.00it/s]
 43%|████▎     | 4599/10708 [1:14:29<50:36,  2.01it/s]
 43%|████▎     | 4600/10708 [1:14:30<50:30,  2.02it/s]
                                                      
{'loss': 3.7211, 'grad_norm': 0.2078220695257187, 'learning_rate': 0.0007040480937084966, 'epoch': 0.43}

 43%|████▎     | 4600/10708 [1:14:30<50:30,  2.02it/s]
 43%|████▎     | 4601/10708 [1:14:30<50:29,  2.02it/s]
 43%|████▎     | 4602/10708 [1:14:31<50:21,  2.02it/s]
 43%|████▎     | 4603/10708 [1:14:31<50:17,  2.02it/s]
 43%|████▎     | 4604/10708 [1:14:32<50:13,  2.03it/s]
 43%|████▎     | 4605/10708 [1:14:32<50:05,  2.03it/s]
 43%|████▎     | 4606/10708 [1:14:33<50:02,  2.03it/s]
 43%|████▎     | 4607/10708 [1:14:33<50:05,  2.03it/s]
 43%|████▎     | 4608/10708 [1:14:34<50:06,  2.03it/s]
 43%|████▎     | 4609/10708 [1:14:34<50:08,  2.03it/s]
 43%|████▎     | 4610/10708 [1:14:35<50:06,  2.03it/s]
 43%|████▎     | 4611/10708 [1:14:35<50:04,  2.03it/s]
 43%|████▎     | 4612/10708 [1:14:36<50:00,  2.03it/s]
 43%|████▎     | 4613/10708 [1:14:36<49:58,  2.03it/s]
 43%|████▎     | 4614/10708 [1:14:37<49:58,  2.03it/s]
 43%|████▎     | 4615/10708 [1:14:37<49:59,  2.03it/s]
 43%|████▎     | 4616/10708 [1:14:38<49:58,  2.03it/s]
 43%|████▎     | 4617/10708 [1:14:38<50:02,  2.03it/s]
 43%|████▎     | 4618/10708 [1:14:39<50:01,  2.03it/s]
 43%|████▎     | 4619/10708 [1:14:39<49:56,  2.03it/s]
 43%|████▎     | 4620/10708 [1:14:40<49:58,  2.03it/s]
 43%|████▎     | 4621/10708 [1:14:40<49:56,  2.03it/s]
 43%|████▎     | 4622/10708 [1:14:41<49:54,  2.03it/s]
 43%|████▎     | 4623/10708 [1:14:41<49:55,  2.03it/s]
 43%|████▎     | 4624/10708 [1:14:42<49:55,  2.03it/s]
 43%|████▎     | 4625/10708 [1:14:42<49:52,  2.03it/s]
                                                      {'loss': 3.7265, 'grad_norm': 0.20977582037448883, 'learning_rate': 0.0007003212144853318, 'epoch': 0.43}


 43%|████▎     | 4625/10708 [1:14:42<49:52,  2.03it/s]
 43%|████▎     | 4626/10708 [1:14:43<49:59,  2.03it/s]
 43%|████▎     | 4627/10708 [1:14:43<49:59,  2.03it/s]
 43%|████▎     | 4628/10708 [1:14:44<49:54,  2.03it/s]
 43%|████▎     | 4629/10708 [1:14:44<49:55,  2.03it/s]
 43%|████▎     | 4630/10708 [1:14:45<49:50,  2.03it/s]
 43%|████▎     | 4631/10708 [1:14:45<49:50,  2.03it/s]
 43%|████▎     | 4632/10708 [1:14:46<49:54,  2.03it/s]
 43%|████▎     | 4633/10708 [1:14:46<49:51,  2.03it/s]
 43%|████▎     | 4634/10708 [1:14:47<49:48,  2.03it/s]
 43%|████▎     | 4635/10708 [1:14:47<49:54,  2.03it/s]
 43%|████▎     | 4636/10708 [1:14:48<49:47,  2.03it/s]
 43%|████▎     | 4637/10708 [1:14:48<49:46,  2.03it/s]
 43%|████▎     | 4638/10708 [1:14:49<49:45,  2.03it/s]
 43%|████▎     | 4639/10708 [1:14:49<49:45,  2.03it/s]
 43%|████▎     | 4640/10708 [1:14:50<49:42,  2.03it/s]
 43%|████▎     | 4641/10708 [1:14:50<49:48,  2.03it/s]
 43%|████▎     | 4642/10708 [1:14:51<49:48,  2.03it/s]
 43%|████▎     | 4643/10708 [1:14:51<49:42,  2.03it/s]
 43%|████▎     | 4644/10708 [1:14:52<49:45,  2.03it/s]
 43%|████▎     | 4645/10708 [1:14:52<49:42,  2.03it/s]
 43%|████▎     | 4646/10708 [1:14:52<49:41,  2.03it/s]
 43%|████▎     | 4647/10708 [1:14:53<49:43,  2.03it/s]
 43%|████▎     | 4648/10708 [1:14:53<49:38,  2.03it/s]
 43%|████▎     | 4649/10708 [1:14:54<49:42,  2.03it/s]
 43%|████▎     | 4650/10708 [1:14:54<49:41,  2.03it/s]
                                                      
{'loss': 3.7142, 'grad_norm': 0.19301597774028778, 'learning_rate': 0.000696581030087339, 'epoch': 0.43}

 43%|████▎     | 4650/10708 [1:14:54<49:41,  2.03it/s]
 43%|████▎     | 4651/10708 [1:14:55<49:44,  2.03it/s]
 43%|████▎     | 4652/10708 [1:14:55<49:42,  2.03it/s]
 43%|████▎     | 4653/10708 [1:14:56<49:40,  2.03it/s]
 43%|████▎     | 4654/10708 [1:14:56<49:42,  2.03it/s]
 43%|████▎     | 4655/10708 [1:14:57<49:39,  2.03it/s]
 43%|████▎     | 4656/10708 [1:14:57<49:41,  2.03it/s]
 43%|████▎     | 4657/10708 [1:14:58<49:42,  2.03it/s]
 44%|████▎     | 4658/10708 [1:14:58<49:39,  2.03it/s]
 44%|████▎     | 4659/10708 [1:14:59<49:40,  2.03it/s]
 44%|████▎     | 4660/10708 [1:14:59<49:37,  2.03it/s]
 44%|████▎     | 4661/10708 [1:15:00<49:37,  2.03it/s]
 44%|████▎     | 4662/10708 [1:15:00<49:35,  2.03it/s]
 44%|████▎     | 4663/10708 [1:15:01<49:32,  2.03it/s]
 44%|████▎     | 4664/10708 [1:15:01<49:38,  2.03it/s]
 44%|████▎     | 4665/10708 [1:15:02<49:37,  2.03it/s]
 44%|████▎     | 4666/10708 [1:15:02<49:32,  2.03it/s]
 44%|████▎     | 4667/10708 [1:15:03<49:39,  2.03it/s]
 44%|████▎     | 4668/10708 [1:15:03<49:39,  2.03it/s]
 44%|████▎     | 4669/10708 [1:15:04<49:38,  2.03it/s]
 44%|████▎     | 4670/10708 [1:15:04<49:34,  2.03it/s]
 44%|████▎     | 4671/10708 [1:15:05<49:35,  2.03it/s]
 44%|████▎     | 4672/10708 [1:15:05<49:32,  2.03it/s]
 44%|████▎     | 4673/10708 [1:15:06<49:25,  2.04it/s]
 44%|████▎     | 4674/10708 [1:15:06<49:30,  2.03it/s]
 44%|████▎     | 4675/10708 [1:15:07<49:27,  2.03it/s]{'loss': 3.712, 'grad_norm': 0.20493794977664948, 'learning_rate': 0.0006928277889345741, 'epoch': 0.44}
                                                      

 44%|████▎     | 4675/10708 [1:15:07<49:27,  2.03it/s]
 44%|████▎     | 4676/10708 [1:15:07<49:31,  2.03it/s]
 44%|████▎     | 4677/10708 [1:15:08<49:28,  2.03it/s]
 44%|████▎     | 4678/10708 [1:15:08<49:27,  2.03it/s]
 44%|████▎     | 4679/10708 [1:15:09<49:26,  2.03it/s]
 44%|████▎     | 4680/10708 [1:15:09<49:26,  2.03it/s]
 44%|████▎     | 4681/10708 [1:15:10<49:24,  2.03it/s]
 44%|████▎     | 4682/10708 [1:15:10<49:25,  2.03it/s]
 44%|████▎     | 4683/10708 [1:15:11<49:24,  2.03it/s]
 44%|████▎     | 4684/10708 [1:15:11<49:23,  2.03it/s]
 44%|████▍     | 4685/10708 [1:15:12<49:24,  2.03it/s]
 44%|████▍     | 4686/10708 [1:15:12<49:20,  2.03it/s]
 44%|████▍     | 4687/10708 [1:15:13<49:18,  2.03it/s]
 44%|████▍     | 4688/10708 [1:15:13<49:28,  2.03it/s]
 44%|████▍     | 4689/10708 [1:15:14<49:23,  2.03it/s]
 44%|████▍     | 4690/10708 [1:15:14<49:18,  2.03it/s]
 44%|████▍     | 4691/10708 [1:15:15<49:19,  2.03it/s]
 44%|████▍     | 4692/10708 [1:15:15<49:17,  2.03it/s]
 44%|████▍     | 4693/10708 [1:15:16<49:16,  2.03it/s]
 44%|████▍     | 4694/10708 [1:15:16<49:19,  2.03it/s]
 44%|████▍     | 4695/10708 [1:15:17<49:18,  2.03it/s]
 44%|████▍     | 4696/10708 [1:15:17<49:14,  2.03it/s]
 44%|████▍     | 4697/10708 [1:15:18<49:16,  2.03it/s]
 44%|████▍     | 4698/10708 [1:15:18<49:13,  2.03it/s]
 44%|████▍     | 4699/10708 [1:15:19<49:12,  2.04it/s]
 44%|████▍     | 4700/10708 [1:15:19<49:15,  2.03it/s]{'loss': 3.7164, 'grad_norm': 0.21682792901992798, 'learning_rate': 0.0006890617403143123, 'epoch': 0.44}
                                                      

 44%|████▍     | 4700/10708 [1:15:19<49:15,  2.03it/s]
 44%|████▍     | 4701/10708 [1:15:20<49:18,  2.03it/s]
 44%|████▍     | 4702/10708 [1:15:20<49:19,  2.03it/s]
 44%|████▍     | 4703/10708 [1:15:21<49:16,  2.03it/s]
 44%|████▍     | 4704/10708 [1:15:21<49:17,  2.03it/s]
 44%|████▍     | 4705/10708 [1:15:22<49:17,  2.03it/s]
 44%|████▍     | 4706/10708 [1:15:22<49:14,  2.03it/s]
 44%|████▍     | 4707/10708 [1:15:23<49:14,  2.03it/s]
 44%|████▍     | 4708/10708 [1:15:23<49:12,  2.03it/s]
 44%|████▍     | 4709/10708 [1:15:24<49:16,  2.03it/s]
 44%|████▍     | 4710/10708 [1:15:24<49:15,  2.03it/s]
 44%|████▍     | 4711/10708 [1:15:24<49:16,  2.03it/s]
 44%|████▍     | 4712/10708 [1:15:25<49:13,  2.03it/s]
 44%|████▍     | 4713/10708 [1:15:25<49:11,  2.03it/s]
 44%|████▍     | 4714/10708 [1:15:26<49:09,  2.03it/s]
 44%|████▍     | 4715/10708 [1:15:26<49:06,  2.03it/s]
 44%|████▍     | 4716/10708 [1:15:27<49:06,  2.03it/s]
 44%|████▍     | 4717/10708 [1:15:27<49:08,  2.03it/s]
 44%|████▍     | 4718/10708 [1:15:28<49:06,  2.03it/s]
 44%|████▍     | 4719/10708 [1:15:28<49:04,  2.03it/s]
 44%|████▍     | 4720/10708 [1:15:29<49:03,  2.03it/s]
 44%|████▍     | 4721/10708 [1:15:29<49:01,  2.04it/s]
 44%|████▍     | 4722/10708 [1:15:30<49:01,  2.03it/s]
 44%|████▍     | 4723/10708 [1:15:30<49:03,  2.03it/s]
 44%|████▍     | 4724/10708 [1:15:31<49:05,  2.03it/s]
 44%|████▍     | 4725/10708 [1:15:31<49:05,  2.03it/s]
                                                      {'loss': 3.7219, 'grad_norm': 0.21389877796173096, 'learning_rate': 0.0006852831343644903, 'epoch': 0.44}


 44%|████▍     | 4725/10708 [1:15:31<49:05,  2.03it/s]
 44%|████▍     | 4726/10708 [1:15:32<49:10,  2.03it/s]
 44%|████▍     | 4727/10708 [1:15:32<49:07,  2.03it/s]
 44%|████▍     | 4728/10708 [1:15:33<49:06,  2.03it/s]
 44%|████▍     | 4729/10708 [1:15:33<49:02,  2.03it/s]
 44%|████▍     | 4730/10708 [1:15:34<49:04,  2.03it/s]
 44%|████▍     | 4731/10708 [1:15:34<49:03,  2.03it/s]
 44%|████▍     | 4732/10708 [1:15:35<49:03,  2.03it/s]
 44%|████▍     | 4733/10708 [1:15:35<49:05,  2.03it/s]
 44%|████▍     | 4734/10708 [1:15:36<48:59,  2.03it/s]
 44%|████▍     | 4735/10708 [1:15:36<48:58,  2.03it/s]
 44%|████▍     | 4736/10708 [1:15:37<49:00,  2.03it/s]
 44%|████▍     | 4737/10708 [1:15:37<49:01,  2.03it/s]
 44%|████▍     | 4738/10708 [1:15:38<48:58,  2.03it/s]
 44%|████▍     | 4739/10708 [1:15:38<48:55,  2.03it/s]
 44%|████▍     | 4740/10708 [1:15:39<48:54,  2.03it/s]
 44%|████▍     | 4741/10708 [1:15:39<48:56,  2.03it/s]
 44%|████▍     | 4742/10708 [1:15:40<48:51,  2.04it/s]
 44%|████▍     | 4743/10708 [1:15:40<48:55,  2.03it/s]
 44%|████▍     | 4744/10708 [1:15:41<48:59,  2.03it/s]
 44%|████▍     | 4745/10708 [1:15:41<48:56,  2.03it/s]
 44%|████▍     | 4746/10708 [1:15:42<48:56,  2.03it/s]
 44%|████▍     | 4747/10708 [1:15:42<48:53,  2.03it/s]
 44%|████▍     | 4748/10708 [1:15:43<49:06,  2.02it/s]
 44%|████▍     | 4749/10708 [1:15:43<49:03,  2.02it/s]
 44%|████▍     | 4750/10708 [1:15:44<49:01,  2.03it/s]{'loss': 3.709, 'grad_norm': 0.19242490828037262, 'learning_rate': 0.0006814922220570927, 'epoch': 0.44}

                                                      

 44%|████▍     | 4750/10708 [1:15:44<49:01,  2.03it/s]
 44%|████▍     | 4751/10708 [1:15:44<49:02,  2.02it/s]
 44%|████▍     | 4752/10708 [1:15:45<48:59,  2.03it/s]
 44%|████▍     | 4753/10708 [1:15:45<48:56,  2.03it/s]
 44%|████▍     | 4754/10708 [1:15:46<48:51,  2.03it/s]
 44%|████▍     | 4755/10708 [1:15:46<48:51,  2.03it/s]
 44%|████▍     | 4756/10708 [1:15:47<48:48,  2.03it/s]
 44%|████▍     | 4757/10708 [1:15:47<48:52,  2.03it/s]
 44%|████▍     | 4758/10708 [1:15:48<48:50,  2.03it/s]
 44%|████▍     | 4759/10708 [1:15:48<48:50,  2.03it/s]
 44%|████▍     | 4760/10708 [1:15:49<48:47,  2.03it/s]
 44%|████▍     | 4761/10708 [1:15:49<48:48,  2.03it/s]
 44%|████▍     | 4762/10708 [1:15:50<48:48,  2.03it/s]
 44%|████▍     | 4763/10708 [1:15:50<48:44,  2.03it/s]
 44%|████▍     | 4764/10708 [1:15:51<48:45,  2.03it/s]
 44%|████▍     | 4765/10708 [1:15:51<48:47,  2.03it/s]
 45%|████▍     | 4766/10708 [1:15:52<48:44,  2.03it/s]
 45%|████▍     | 4767/10708 [1:15:52<48:50,  2.03it/s]
 45%|████▍     | 4768/10708 [1:15:53<48:44,  2.03it/s]
 45%|████▍     | 4769/10708 [1:15:53<48:39,  2.03it/s]
 45%|████▍     | 4770/10708 [1:15:54<48:39,  2.03it/s]
 45%|████▍     | 4771/10708 [1:15:54<48:42,  2.03it/s]
 45%|████▍     | 4772/10708 [1:15:55<48:37,  2.03it/s]
 45%|████▍     | 4773/10708 [1:15:55<48:36,  2.03it/s]
 45%|████▍     | 4774/10708 [1:15:56<48:37,  2.03it/s]
 45%|████▍     | 4775/10708 [1:15:56<48:38,  2.03it/s]
                                                      
{'loss': 3.7143, 'grad_norm': 0.2177303582429886, 'learning_rate': 0.000677689255181483, 'epoch': 0.45}

 45%|████▍     | 4775/10708 [1:15:56<48:38,  2.03it/s]
 45%|████▍     | 4776/10708 [1:15:56<48:43,  2.03it/s]
 45%|████▍     | 4777/10708 [1:15:57<48:42,  2.03it/s]
 45%|████▍     | 4778/10708 [1:15:57<48:47,  2.03it/s]
 45%|████▍     | 4779/10708 [1:15:58<48:44,  2.03it/s]
 45%|████▍     | 4780/10708 [1:15:58<48:45,  2.03it/s]
 45%|████▍     | 4781/10708 [1:15:59<48:44,  2.03it/s]
 45%|████▍     | 4782/10708 [1:15:59<48:40,  2.03it/s]
 45%|████▍     | 4783/10708 [1:16:00<48:40,  2.03it/s]
 45%|████▍     | 4784/10708 [1:16:00<48:37,  2.03it/s]
 45%|████▍     | 4785/10708 [1:16:01<48:34,  2.03it/s]
 45%|████▍     | 4786/10708 [1:16:01<48:34,  2.03it/s]
 45%|████▍     | 4787/10708 [1:16:02<48:36,  2.03it/s]
 45%|████▍     | 4788/10708 [1:16:02<48:34,  2.03it/s]
 45%|████▍     | 4789/10708 [1:16:03<48:32,  2.03it/s]
 45%|████▍     | 4790/10708 [1:16:03<48:31,  2.03it/s]
 45%|████▍     | 4791/10708 [1:16:04<48:33,  2.03it/s]
 45%|████▍     | 4792/10708 [1:16:04<48:36,  2.03it/s]
 45%|████▍     | 4793/10708 [1:16:05<48:31,  2.03it/s]
 45%|████▍     | 4794/10708 [1:16:05<48:37,  2.03it/s]
 45%|████▍     | 4795/10708 [1:16:06<48:34,  2.03it/s]
 45%|████▍     | 4796/10708 [1:16:06<48:38,  2.03it/s]
 45%|████▍     | 4797/10708 [1:16:07<48:33,  2.03it/s]
 45%|████▍     | 4798/10708 [1:16:07<48:28,  2.03it/s]
 45%|████▍     | 4799/10708 [1:16:08<48:29,  2.03it/s]
 45%|████▍     | 4800/10708 [1:16:08<48:24,  2.03it/s]
                                                      
{'loss': 3.7052, 'grad_norm': 0.1967005878686905, 'learning_rate': 0.0006738744863276787, 'epoch': 0.45}

 45%|████▍     | 4800/10708 [1:16:08<48:24,  2.03it/s]
 45%|████▍     | 4801/10708 [1:16:09<48:32,  2.03it/s]
 45%|████▍     | 4802/10708 [1:16:09<48:30,  2.03it/s]
 45%|████▍     | 4803/10708 [1:16:10<48:29,  2.03it/s]
 45%|████▍     | 4804/10708 [1:16:10<48:28,  2.03it/s]
 45%|████▍     | 4805/10708 [1:16:11<48:24,  2.03it/s]
 45%|████▍     | 4806/10708 [1:16:11<48:21,  2.03it/s]
 45%|████▍     | 4807/10708 [1:16:12<48:24,  2.03it/s]
 45%|████▍     | 4808/10708 [1:16:12<48:22,  2.03it/s]
 45%|████▍     | 4809/10708 [1:16:13<48:28,  2.03it/s]
 45%|████▍     | 4810/10708 [1:16:13<48:32,  2.03it/s]
 45%|████▍     | 4811/10708 [1:16:14<48:31,  2.03it/s]
 45%|████▍     | 4812/10708 [1:16:14<48:27,  2.03it/s]
 45%|████▍     | 4813/10708 [1:16:15<48:23,  2.03it/s]
 45%|████▍     | 4814/10708 [1:16:15<48:24,  2.03it/s]
 45%|████▍     | 4815/10708 [1:16:16<48:24,  2.03it/s]
 45%|████▍     | 4816/10708 [1:16:16<48:25,  2.03it/s]
 45%|████▍     | 4817/10708 [1:16:17<48:25,  2.03it/s]
 45%|████▍     | 4818/10708 [1:16:17<48:20,  2.03it/s]
 45%|████▌     | 4819/10708 [1:16:18<48:21,  2.03it/s]
 45%|████▌     | 4820/10708 [1:16:18<48:16,  2.03it/s]
 45%|████▌     | 4821/10708 [1:16:19<48:15,  2.03it/s]
 45%|████▌     | 4822/10708 [1:16:19<48:15,  2.03it/s]
 45%|████▌     | 4823/10708 [1:16:20<48:16,  2.03it/s]
 45%|████▌     | 4824/10708 [1:16:20<48:18,  2.03it/s]
 45%|████▌     | 4825/10708 [1:16:21<48:16,  2.03it/s]{'loss': 3.7102, 'grad_norm': 0.2087836116552353, 'learning_rate': 0.0006700481688695759, 'epoch': 0.45}
                                                      

 45%|████▌     | 4825/10708 [1:16:21<48:16,  2.03it/s]
 45%|████▌     | 4826/10708 [1:16:21<48:16,  2.03it/s]
 45%|████▌     | 4827/10708 [1:16:22<48:14,  2.03it/s]
 45%|████▌     | 4828/10708 [1:16:22<48:12,  2.03it/s]
 45%|████▌     | 4829/10708 [1:16:23<48:12,  2.03it/s]
 45%|████▌     | 4830/10708 [1:16:23<48:11,  2.03it/s]
 45%|████▌     | 4831/10708 [1:16:24<48:11,  2.03it/s]
 45%|████▌     | 4832/10708 [1:16:24<48:10,  2.03it/s]
 45%|████▌     | 4833/10708 [1:16:25<48:10,  2.03it/s]
 45%|████▌     | 4834/10708 [1:16:25<48:12,  2.03it/s]
 45%|████▌     | 4835/10708 [1:16:26<48:09,  2.03it/s]
 45%|████▌     | 4836/10708 [1:16:26<48:10,  2.03it/s]
 45%|████▌     | 4837/10708 [1:16:27<48:06,  2.03it/s]
 45%|████▌     | 4838/10708 [1:16:27<48:14,  2.03it/s]
 45%|████▌     | 4839/10708 [1:16:28<48:11,  2.03it/s]
 45%|████▌     | 4840/10708 [1:16:28<48:18,  2.02it/s]
 45%|████▌     | 4841/10708 [1:16:29<48:15,  2.03it/s]
 45%|████▌     | 4842/10708 [1:16:29<48:13,  2.03it/s]
 45%|████▌     | 4843/10708 [1:16:29<48:09,  2.03it/s]
 45%|████▌     | 4844/10708 [1:16:30<48:05,  2.03it/s]
 45%|████▌     | 4845/10708 [1:16:30<48:09,  2.03it/s]
 45%|████▌     | 4846/10708 [1:16:31<48:09,  2.03it/s]
 45%|████▌     | 4847/10708 [1:16:31<48:04,  2.03it/s]
 45%|████▌     | 4848/10708 [1:16:32<48:06,  2.03it/s]
 45%|████▌     | 4849/10708 [1:16:32<48:07,  2.03it/s]
 45%|████▌     | 4850/10708 [1:16:33<48:05,  2.03it/s]{'loss': 3.6991, 'grad_norm': 0.20195670425891876, 'learning_rate': 0.000666210556948119, 'epoch': 0.45}
                                                      

 45%|████▌     | 4850/10708 [1:16:33<48:05,  2.03it/s]
 45%|████▌     | 4851/10708 [1:16:33<48:08,  2.03it/s]
 45%|████▌     | 4852/10708 [1:16:34<48:03,  2.03it/s]
 45%|████▌     | 4853/10708 [1:16:34<48:03,  2.03it/s]
 45%|████▌     | 4854/10708 [1:16:35<48:00,  2.03it/s]
 45%|████▌     | 4855/10708 [1:16:35<48:02,  2.03it/s]
 45%|████▌     | 4856/10708 [1:16:36<48:01,  2.03it/s]
 45%|████▌     | 4857/10708 [1:16:36<48:04,  2.03it/s]
 45%|████▌     | 4858/10708 [1:16:37<48:01,  2.03it/s]
 45%|████▌     | 4859/10708 [1:16:37<47:57,  2.03it/s]
 45%|████▌     | 4860/10708 [1:16:38<47:54,  2.03it/s]
 45%|████▌     | 4861/10708 [1:16:38<47:56,  2.03it/s]
 45%|████▌     | 4862/10708 [1:16:39<48:00,  2.03it/s]
 45%|████▌     | 4863/10708 [1:16:39<48:00,  2.03it/s]
 45%|████▌     | 4864/10708 [1:16:40<48:00,  2.03it/s]
 45%|████▌     | 4865/10708 [1:16:40<48:03,  2.03it/s]
 45%|████▌     | 4866/10708 [1:16:41<47:59,  2.03it/s]
 45%|████▌     | 4867/10708 [1:16:41<47:57,  2.03it/s]
 45%|████▌     | 4868/10708 [1:16:42<47:53,  2.03it/s]
 45%|████▌     | 4869/10708 [1:16:42<47:49,  2.03it/s]
 45%|████▌     | 4870/10708 [1:16:43<47:59,  2.03it/s]
 45%|████▌     | 4871/10708 [1:16:43<47:57,  2.03it/s]
 45%|████▌     | 4872/10708 [1:16:44<48:01,  2.03it/s]
 46%|████▌     | 4873/10708 [1:16:44<47:55,  2.03it/s]
 46%|████▌     | 4874/10708 [1:16:45<47:48,  2.03it/s]
 46%|████▌     | 4875/10708 [1:16:45<47:51,  2.03it/s]
                                                      
{'loss': 3.7026, 'grad_norm': 0.20650863647460938, 'learning_rate': 0.0006623619054544226, 'epoch': 0.46}

 46%|████▌     | 4875/10708 [1:16:45<47:51,  2.03it/s]
 46%|████▌     | 4876/10708 [1:16:46<47:52,  2.03it/s]
 46%|████▌     | 4877/10708 [1:16:46<47:56,  2.03it/s]
 46%|████▌     | 4878/10708 [1:16:47<47:49,  2.03it/s]
 46%|████▌     | 4879/10708 [1:16:47<47:47,  2.03it/s]
 46%|████▌     | 4880/10708 [1:16:48<47:51,  2.03it/s]
 46%|████▌     | 4881/10708 [1:16:48<47:46,  2.03it/s]
 46%|████▌     | 4882/10708 [1:16:49<47:48,  2.03it/s]
 46%|████▌     | 4883/10708 [1:16:49<47:47,  2.03it/s]
 46%|████▌     | 4884/10708 [1:16:50<47:45,  2.03it/s]
 46%|████▌     | 4885/10708 [1:16:50<47:50,  2.03it/s]
 46%|████▌     | 4886/10708 [1:16:51<47:46,  2.03it/s]
 46%|████▌     | 4887/10708 [1:16:51<47:41,  2.03it/s]
 46%|████▌     | 4888/10708 [1:16:52<47:47,  2.03it/s]
 46%|████▌     | 4889/10708 [1:16:52<47:42,  2.03it/s]
 46%|████▌     | 4890/10708 [1:16:53<47:39,  2.03it/s]
 46%|████▌     | 4891/10708 [1:16:53<47:40,  2.03it/s]
 46%|████▌     | 4892/10708 [1:16:54<47:37,  2.04it/s]
 46%|████▌     | 4893/10708 [1:16:54<47:39,  2.03it/s]
 46%|████▌     | 4894/10708 [1:16:55<47:38,  2.03it/s]
 46%|████▌     | 4895/10708 [1:16:55<47:40,  2.03it/s]
 46%|████▌     | 4896/10708 [1:16:56<47:45,  2.03it/s]
 46%|████▌     | 4897/10708 [1:16:56<47:41,  2.03it/s]
 46%|████▌     | 4898/10708 [1:16:57<47:36,  2.03it/s]
 46%|████▌     | 4899/10708 [1:16:57<47:35,  2.03it/s]
 46%|████▌     | 4900/10708 [1:16:58<47:34,  2.03it/s]{'loss': 3.6962, 'grad_norm': 0.20323608815670013, 'learning_rate': 0.00065850247001284, 'epoch': 0.46}

                                                      

 46%|████▌     | 4900/10708 [1:16:58<47:34,  2.03it/s]
 46%|████▌     | 4901/10708 [1:16:58<47:39,  2.03it/s]
 46%|████▌     | 4902/10708 [1:16:59<47:41,  2.03it/s]
 46%|████▌     | 4903/10708 [1:16:59<47:41,  2.03it/s]
 46%|████▌     | 4904/10708 [1:17:00<47:36,  2.03it/s]
 46%|████▌     | 4905/10708 [1:17:00<47:34,  2.03it/s]
 46%|████▌     | 4906/10708 [1:17:01<47:37,  2.03it/s]
 46%|████▌     | 4907/10708 [1:17:01<47:35,  2.03it/s]
 46%|████▌     | 4908/10708 [1:17:01<47:30,  2.03it/s]
 46%|████▌     | 4909/10708 [1:17:02<47:32,  2.03it/s]
 46%|████▌     | 4910/10708 [1:17:02<47:33,  2.03it/s]
 46%|████▌     | 4911/10708 [1:17:03<47:31,  2.03it/s]
 46%|████▌     | 4912/10708 [1:17:03<47:31,  2.03it/s]
 46%|████▌     | 4913/10708 [1:17:04<47:30,  2.03it/s]
 46%|████▌     | 4914/10708 [1:17:04<47:29,  2.03it/s]
 46%|████▌     | 4915/10708 [1:17:05<47:30,  2.03it/s]
 46%|████▌     | 4916/10708 [1:17:05<47:35,  2.03it/s]
 46%|████▌     | 4917/10708 [1:17:06<47:32,  2.03it/s]
 46%|████▌     | 4918/10708 [1:17:06<47:29,  2.03it/s]
 46%|████▌     | 4919/10708 [1:17:07<47:27,  2.03it/s]
 46%|████▌     | 4920/10708 [1:17:07<47:26,  2.03it/s]
 46%|████▌     | 4921/10708 [1:17:08<47:28,  2.03it/s]
 46%|████▌     | 4922/10708 [1:17:08<47:30,  2.03it/s]
 46%|████▌     | 4923/10708 [1:17:09<47:27,  2.03it/s]
 46%|████▌     | 4924/10708 [1:17:09<47:28,  2.03it/s]
 46%|████▌     | 4925/10708 [1:17:10<47:28,  2.03it/s]
                                                      
{'loss': 3.7027, 'grad_norm': 0.19085539877414703, 'learning_rate': 0.0006546325069639862, 'epoch': 0.46}

 46%|████▌     | 4925/10708 [1:17:10<47:28,  2.03it/s]
 46%|████▌     | 4926/10708 [1:17:10<47:32,  2.03it/s]
 46%|████▌     | 4927/10708 [1:17:11<47:33,  2.03it/s]
 46%|████▌     | 4928/10708 [1:17:11<47:28,  2.03it/s]
 46%|████▌     | 4929/10708 [1:17:12<47:28,  2.03it/s]
 46%|████▌     | 4930/10708 [1:17:12<47:22,  2.03it/s]
 46%|████▌     | 4931/10708 [1:17:13<47:27,  2.03it/s]
 46%|████▌     | 4932/10708 [1:17:13<47:26,  2.03it/s]
 46%|████▌     | 4933/10708 [1:17:14<47:18,  2.03it/s]
 46%|████▌     | 4934/10708 [1:17:14<47:20,  2.03it/s]
 46%|████▌     | 4935/10708 [1:17:15<47:18,  2.03it/s]
 46%|████▌     | 4936/10708 [1:17:15<47:14,  2.04it/s]
 46%|████▌     | 4937/10708 [1:17:16<47:15,  2.04it/s]
 46%|████▌     | 4938/10708 [1:17:16<47:16,  2.03it/s]
 46%|████▌     | 4939/10708 [1:17:17<47:19,  2.03it/s]
 46%|████▌     | 4940/10708 [1:17:17<47:26,  2.03it/s]
 46%|████▌     | 4941/10708 [1:17:18<47:22,  2.03it/s]
 46%|████▌     | 4942/10708 [1:17:18<47:23,  2.03it/s]
 46%|████▌     | 4943/10708 [1:17:19<47:19,  2.03it/s]
 46%|████▌     | 4944/10708 [1:17:19<47:17,  2.03it/s]
 46%|████▌     | 4945/10708 [1:17:20<47:17,  2.03it/s]
 46%|████▌     | 4946/10708 [1:17:20<47:16,  2.03it/s]
 46%|████▌     | 4947/10708 [1:17:21<47:14,  2.03it/s]
 46%|████▌     | 4948/10708 [1:17:21<47:14,  2.03it/s]
 46%|████▌     | 4949/10708 [1:17:22<47:15,  2.03it/s]
 46%|████▌     | 4950/10708 [1:17:22<47:14,  2.03it/s]
                                                      {'loss': 3.6981, 'grad_norm': 0.20606784522533417, 'learning_rate': 0.0006507522733477114, 'epoch': 0.46}


 46%|████▌     | 4950/10708 [1:17:22<47:14,  2.03it/s]
 46%|████▌     | 4951/10708 [1:17:23<47:20,  2.03it/s]
 46%|████▌     | 4952/10708 [1:17:23<47:21,  2.03it/s]
 46%|████▋     | 4953/10708 [1:17:24<47:19,  2.03it/s]
 46%|████▋     | 4954/10708 [1:17:24<47:18,  2.03it/s]
 46%|████▋     | 4955/10708 [1:17:25<47:15,  2.03it/s]
 46%|████▋     | 4956/10708 [1:17:25<47:18,  2.03it/s]
 46%|████▋     | 4957/10708 [1:17:26<47:16,  2.03it/s]
 46%|████▋     | 4958/10708 [1:17:26<47:14,  2.03it/s]
 46%|████▋     | 4959/10708 [1:17:27<47:15,  2.03it/s]
 46%|████▋     | 4960/10708 [1:17:27<47:09,  2.03it/s]
 46%|████▋     | 4961/10708 [1:17:28<47:08,  2.03it/s]
 46%|████▋     | 4962/10708 [1:17:28<47:09,  2.03it/s]
 46%|████▋     | 4963/10708 [1:17:29<47:08,  2.03it/s]
 46%|████▋     | 4964/10708 [1:17:29<47:02,  2.03it/s]
 46%|████▋     | 4965/10708 [1:17:30<47:03,  2.03it/s]
 46%|████▋     | 4966/10708 [1:17:30<47:05,  2.03it/s]
 46%|████▋     | 4967/10708 [1:17:31<47:02,  2.03it/s]
 46%|████▋     | 4968/10708 [1:17:31<47:07,  2.03it/s]
 46%|████▋     | 4969/10708 [1:17:32<47:03,  2.03it/s]
 46%|████▋     | 4970/10708 [1:17:32<47:03,  2.03it/s]
 46%|████▋     | 4971/10708 [1:17:33<47:00,  2.03it/s]
 46%|████▋     | 4972/10708 [1:17:33<46:59,  2.03it/s]
 46%|████▋     | 4973/10708 [1:17:34<47:01,  2.03it/s]
 46%|████▋     | 4974/10708 [1:17:34<46:58,  2.03it/s]
 46%|████▋     | 4975/10708 [1:17:34<46:56,  2.04it/s]
                                                      {'loss': 3.6949, 'grad_norm': 0.2154226005077362, 'learning_rate': 0.0006468620268860283, 'epoch': 0.46}


 46%|████▋     | 4975/10708 [1:17:34<46:56,  2.04it/s]
 46%|████▋     | 4976/10708 [1:17:35<47:03,  2.03it/s]
 46%|████▋     | 4977/10708 [1:17:35<47:00,  2.03it/s]
 46%|████▋     | 4978/10708 [1:17:36<46:57,  2.03it/s]
 46%|████▋     | 4979/10708 [1:17:36<46:58,  2.03it/s]
 47%|████▋     | 4980/10708 [1:17:37<46:56,  2.03it/s]
 47%|████▋     | 4981/10708 [1:17:37<46:57,  2.03it/s]
 47%|████▋     | 4982/10708 [1:17:38<46:58,  2.03it/s]
 47%|████▋     | 4983/10708 [1:17:38<46:56,  2.03it/s]
 47%|████▋     | 4984/10708 [1:17:39<46:54,  2.03it/s]
 47%|████▋     | 4985/10708 [1:17:39<46:58,  2.03it/s]
 47%|████▋     | 4986/10708 [1:17:40<46:58,  2.03it/s]
 47%|████▋     | 4987/10708 [1:17:40<46:55,  2.03it/s]
 47%|████▋     | 4988/10708 [1:17:41<46:58,  2.03it/s]
 47%|████▋     | 4989/10708 [1:17:41<46:54,  2.03it/s]
 47%|████▋     | 4990/10708 [1:17:42<47:05,  2.02it/s]
 47%|████▋     | 4991/10708 [1:17:42<47:01,  2.03it/s]
 47%|████▋     | 4992/10708 [1:17:43<47:09,  2.02it/s]
 47%|████▋     | 4993/10708 [1:17:43<47:13,  2.02it/s]
 47%|████▋     | 4994/10708 [1:17:44<47:07,  2.02it/s]
 47%|████▋     | 4995/10708 [1:17:44<47:01,  2.02it/s]
 47%|████▋     | 4996/10708 [1:17:45<46:56,  2.03it/s]
 47%|████▋     | 4997/10708 [1:17:45<46:54,  2.03it/s]
 47%|████▋     | 4998/10708 [1:17:46<46:55,  2.03it/s]
 47%|████▋     | 4999/10708 [1:17:46<46:50,  2.03it/s]
 47%|████▋     | 5000/10708 [1:17:47<46:52,  2.03it/s]
                                                      {'loss': 3.6975, 'grad_norm': 0.2237989604473114, 'learning_rate': 0.0006429620259659954, 'epoch': 0.47}

 47%|████▋     | 5000/10708 [1:17:47<46:52,  2.03it/s]
 47%|████▋     | 5001/10708 [1:17:47<46:54,  2.03it/s]
 47%|████▋     | 5002/10708 [1:17:48<46:52,  2.03it/s]
 47%|████▋     | 5003/10708 [1:17:48<46:53,  2.03it/s]
 47%|████▋     | 5004/10708 [1:17:49<46:49,  2.03it/s]
 47%|████▋     | 5005/10708 [1:17:49<46:54,  2.03it/s]
 47%|████▋     | 5006/10708 [1:17:50<46:48,  2.03it/s]
 47%|████▋     | 5007/10708 [1:17:50<46:47,  2.03it/s]
 47%|████▋     | 5008/10708 [1:17:51<46:46,  2.03it/s]
 47%|████▋     | 5009/10708 [1:17:51<46:44,  2.03it/s]
 47%|████▋     | 5010/10708 [1:17:52<46:42,  2.03it/s]
 47%|████▋     | 5011/10708 [1:17:52<46:43,  2.03it/s]
 47%|████▋     | 5012/10708 [1:17:53<46:42,  2.03it/s]
 47%|████▋     | 5013/10708 [1:17:53<46:42,  2.03it/s]
 47%|████▋     | 5014/10708 [1:17:54<46:39,  2.03it/s]
 47%|████▋     | 5015/10708 [1:17:54<46:38,  2.03it/s]
 47%|████▋     | 5016/10708 [1:17:55<46:38,  2.03it/s]
 47%|████▋     | 5017/10708 [1:17:55<46:37,  2.03it/s]
 47%|████▋     | 5018/10708 [1:17:56<46:33,  2.04it/s]
 47%|████▋     | 5019/10708 [1:17:56<46:34,  2.04it/s]
 47%|████▋     | 5020/10708 [1:17:57<46:35,  2.03it/s]
 47%|████▋     | 5021/10708 [1:17:57<46:33,  2.04it/s]
 47%|████▋     | 5022/10708 [1:17:58<46:42,  2.03it/s]
 47%|████▋     | 5023/10708 [1:17:58<46:40,  2.03it/s]
 47%|████▋     | 5024/10708 [1:17:59<46:45,  2.03it/s]
 47%|████▋     | 5025/10708 [1:17:59<46:43,  2.03it/s]{'loss': 3.7074, 'grad_norm': 0.20733600854873657, 'learning_rate': 0.0006390525296225539, 'epoch': 0.47}

                                                      

 47%|████▋     | 5025/10708 [1:17:59<46:43,  2.03it/s]
 47%|████▋     | 5026/10708 [1:18:00<46:50,  2.02it/s]
 47%|████▋     | 5027/10708 [1:18:00<46:42,  2.03it/s]
 47%|████▋     | 5028/10708 [1:18:01<46:40,  2.03it/s]
 47%|████▋     | 5029/10708 [1:18:01<46:38,  2.03it/s]
 47%|████▋     | 5030/10708 [1:18:02<46:35,  2.03it/s]
 47%|████▋     | 5031/10708 [1:18:02<46:36,  2.03it/s]
 47%|████▋     | 5032/10708 [1:18:03<46:34,  2.03it/s]
 47%|████▋     | 5033/10708 [1:18:03<46:30,  2.03it/s]
 47%|████▋     | 5034/10708 [1:18:04<46:39,  2.03it/s]
 47%|████▋     | 5035/10708 [1:18:04<46:35,  2.03it/s]
 47%|████▋     | 5036/10708 [1:18:05<46:35,  2.03it/s]
 47%|████▋     | 5037/10708 [1:18:05<46:35,  2.03it/s]
 47%|████▋     | 5038/10708 [1:18:06<46:30,  2.03it/s]
 47%|████▋     | 5039/10708 [1:18:06<46:26,  2.03it/s]
 47%|████▋     | 5040/10708 [1:18:07<46:28,  2.03it/s]
 47%|████▋     | 5041/10708 [1:18:07<46:29,  2.03it/s]
 47%|████▋     | 5042/10708 [1:18:07<46:31,  2.03it/s]
 47%|████▋     | 5043/10708 [1:18:08<46:31,  2.03it/s]
 47%|████▋     | 5044/10708 [1:18:08<46:28,  2.03it/s]
 47%|████▋     | 5045/10708 [1:18:09<46:26,  2.03it/s]
 47%|████▋     | 5046/10708 [1:18:09<46:27,  2.03it/s]
 47%|████▋     | 5047/10708 [1:18:10<46:24,  2.03it/s]
 47%|████▋     | 5048/10708 [1:18:10<46:23,  2.03it/s]
 47%|████▋     | 5049/10708 [1:18:11<46:23,  2.03it/s]
 47%|████▋     | 5050/10708 [1:18:11<46:24,  2.03it/s]{'loss': 3.6924, 'grad_norm': 0.19802828133106232, 'learning_rate': 0.0006351337975213238, 'epoch': 0.47}

                                                      

 47%|████▋     | 5050/10708 [1:18:11<46:24,  2.03it/s]
 47%|████▋     | 5051/10708 [1:18:12<46:26,  2.03it/s]
 47%|████▋     | 5052/10708 [1:18:12<46:30,  2.03it/s]
 47%|████▋     | 5053/10708 [1:18:13<46:31,  2.03it/s]
 47%|████▋     | 5054/10708 [1:18:13<46:32,  2.02it/s]
 47%|████▋     | 5055/10708 [1:18:14<46:30,  2.03it/s]
 47%|████▋     | 5056/10708 [1:18:14<46:24,  2.03it/s]
 47%|████▋     | 5057/10708 [1:18:15<46:24,  2.03it/s]
 47%|████▋     | 5058/10708 [1:18:15<46:21,  2.03it/s]
 47%|████▋     | 5059/10708 [1:18:16<46:19,  2.03it/s]
 47%|████▋     | 5060/10708 [1:18:16<46:18,  2.03it/s]
 47%|████▋     | 5061/10708 [1:18:17<46:21,  2.03it/s]
 47%|████▋     | 5062/10708 [1:18:17<46:19,  2.03it/s]
 47%|████▋     | 5063/10708 [1:18:18<46:18,  2.03it/s]
 47%|████▋     | 5064/10708 [1:18:18<46:14,  2.03it/s]
 47%|████▋     | 5065/10708 [1:18:19<46:16,  2.03it/s]
 47%|████▋     | 5066/10708 [1:18:19<46:14,  2.03it/s]
 47%|████▋     | 5067/10708 [1:18:20<46:13,  2.03it/s]
 47%|████▋     | 5068/10708 [1:18:20<46:15,  2.03it/s]
 47%|████▋     | 5069/10708 [1:18:21<46:13,  2.03it/s]
 47%|████▋     | 5070/10708 [1:18:21<46:09,  2.04it/s]
 47%|████▋     | 5071/10708 [1:18:22<46:12,  2.03it/s]
 47%|████▋     | 5072/10708 [1:18:22<46:12,  2.03it/s]
 47%|████▋     | 5073/10708 [1:18:23<46:14,  2.03it/s]
 47%|████▋     | 5074/10708 [1:18:23<46:12,  2.03it/s]
 47%|████▋     | 5075/10708 [1:18:24<46:12,  2.03it/s]
                                                      {'loss': 3.6924, 'grad_norm': 0.200722336769104, 'learning_rate': 0.0006312060899413565, 'epoch': 0.47}

 47%|████▋     | 5075/10708 [1:18:24<46:12,  2.03it/s]
 47%|████▋     | 5076/10708 [1:18:24<46:18,  2.03it/s]
 47%|████▋     | 5077/10708 [1:18:25<46:14,  2.03it/s]
 47%|████▋     | 5078/10708 [1:18:25<46:12,  2.03it/s]
 47%|████▋     | 5079/10708 [1:18:26<46:09,  2.03it/s]
 47%|████▋     | 5080/10708 [1:18:26<46:07,  2.03it/s]
 47%|████▋     | 5081/10708 [1:18:27<46:10,  2.03it/s]
 47%|████▋     | 5082/10708 [1:18:27<46:08,  2.03it/s]
 47%|████▋     | 5083/10708 [1:18:28<46:10,  2.03it/s]
 47%|████▋     | 5084/10708 [1:18:28<46:10,  2.03it/s]
 47%|████▋     | 5085/10708 [1:18:29<46:08,  2.03it/s]
 47%|████▋     | 5086/10708 [1:18:29<46:05,  2.03it/s]
 48%|████▊     | 5087/10708 [1:18:30<46:05,  2.03it/s]
 48%|████▊     | 5088/10708 [1:18:30<46:06,  2.03it/s]
 48%|████▊     | 5089/10708 [1:18:31<46:04,  2.03it/s]
 48%|████▊     | 5090/10708 [1:18:31<46:04,  2.03it/s]
 48%|████▊     | 5091/10708 [1:18:32<46:04,  2.03it/s]
 48%|████▊     | 5092/10708 [1:18:32<46:08,  2.03it/s]
 48%|████▊     | 5093/10708 [1:18:33<46:05,  2.03it/s]
 48%|████▊     | 5094/10708 [1:18:33<46:07,  2.03it/s]
 48%|████▊     | 5095/10708 [1:18:34<46:07,  2.03it/s]
 48%|████▊     | 5096/10708 [1:18:34<46:11,  2.03it/s]
 48%|████▊     | 5097/10708 [1:18:35<46:09,  2.03it/s]
 48%|████▊     | 5098/10708 [1:18:35<46:04,  2.03it/s]
 48%|████▊     | 5099/10708 [1:18:36<46:01,  2.03it/s]
 48%|████▊     | 5100/10708 [1:18:36<45:58,  2.03it/s]{'loss': 3.691, 'grad_norm': 0.19512537121772766, 'learning_rate': 0.0006272696677578482, 'epoch': 0.48}

                                                      

 48%|████▊     | 5100/10708 [1:18:36<45:58,  2.03it/s]
 48%|████▊     | 5101/10708 [1:18:37<46:03,  2.03it/s]
 48%|████▊     | 5102/10708 [1:18:37<46:01,  2.03it/s]
 48%|████▊     | 5103/10708 [1:18:38<46:02,  2.03it/s]
 48%|████▊     | 5104/10708 [1:18:38<46:00,  2.03it/s]
 48%|████▊     | 5105/10708 [1:18:39<46:01,  2.03it/s]
 48%|████▊     | 5106/10708 [1:18:39<45:59,  2.03it/s]
 48%|████▊     | 5107/10708 [1:18:39<46:01,  2.03it/s]
 48%|████▊     | 5108/10708 [1:18:40<45:56,  2.03it/s]
 48%|████▊     | 5109/10708 [1:18:40<45:57,  2.03it/s]
 48%|████▊     | 5110/10708 [1:18:41<45:57,  2.03it/s]
 48%|████▊     | 5111/10708 [1:18:41<45:55,  2.03it/s]
 48%|████▊     | 5112/10708 [1:18:42<45:56,  2.03it/s]
 48%|████▊     | 5113/10708 [1:18:42<45:54,  2.03it/s]
 48%|████▊     | 5114/10708 [1:18:43<46:03,  2.02it/s]
 48%|████▊     | 5115/10708 [1:18:43<46:00,  2.03it/s]
 48%|████▊     | 5116/10708 [1:18:44<45:58,  2.03it/s]
 48%|████▊     | 5117/10708 [1:18:44<45:56,  2.03it/s]
 48%|████▊     | 5118/10708 [1:18:45<45:50,  2.03it/s]
 48%|████▊     | 5119/10708 [1:18:45<45:49,  2.03it/s]
 48%|████▊     | 5120/10708 [1:18:46<45:48,  2.03it/s]
 48%|████▊     | 5121/10708 [1:18:46<45:46,  2.03it/s]
 48%|████▊     | 5122/10708 [1:18:47<45:50,  2.03it/s]
 48%|████▊     | 5123/10708 [1:18:47<45:51,  2.03it/s]
 48%|████▊     | 5124/10708 [1:18:48<45:53,  2.03it/s]
 48%|████▊     | 5125/10708 [1:18:48<45:50,  2.03it/s]
                                                      
{'loss': 3.6863, 'grad_norm': 0.20694175362586975, 'learning_rate': 0.0006233247924248109, 'epoch': 0.48}

 48%|████▊     | 5125/10708 [1:18:48<45:50,  2.03it/s]
 48%|████▊     | 5126/10708 [1:18:49<45:49,  2.03it/s]
 48%|████▊     | 5127/10708 [1:18:49<45:47,  2.03it/s]
 48%|████▊     | 5128/10708 [1:18:50<45:47,  2.03it/s]
 48%|████▊     | 5129/10708 [1:18:50<45:46,  2.03it/s]
 48%|████▊     | 5130/10708 [1:18:51<45:48,  2.03it/s]
 48%|████▊     | 5131/10708 [1:18:51<45:46,  2.03it/s]
 48%|████▊     | 5132/10708 [1:18:52<45:44,  2.03it/s]
 48%|████▊     | 5133/10708 [1:18:52<45:46,  2.03it/s]
 48%|████▊     | 5134/10708 [1:18:53<45:42,  2.03it/s]
 48%|████▊     | 5135/10708 [1:18:53<45:47,  2.03it/s]
 48%|████▊     | 5136/10708 [1:18:54<45:42,  2.03it/s]
 48%|████▊     | 5137/10708 [1:18:54<45:38,  2.03it/s]
 48%|████▊     | 5138/10708 [1:18:55<45:38,  2.03it/s]
 48%|████▊     | 5139/10708 [1:18:55<45:38,  2.03it/s]
 48%|████▊     | 5140/10708 [1:18:56<45:35,  2.04it/s]
 48%|████▊     | 5141/10708 [1:18:56<45:36,  2.03it/s]
 48%|████▊     | 5142/10708 [1:18:57<45:36,  2.03it/s]
 48%|████▊     | 5143/10708 [1:18:57<45:36,  2.03it/s]
 48%|████▊     | 5144/10708 [1:18:58<45:37,  2.03it/s]
 48%|████▊     | 5145/10708 [1:18:58<45:38,  2.03it/s]
 48%|████▊     | 5146/10708 [1:18:59<45:34,  2.03it/s]
 48%|████▊     | 5147/10708 [1:18:59<45:37,  2.03it/s]
 48%|████▊     | 5148/10708 [1:19:00<45:33,  2.03it/s]
 48%|████▊     | 5149/10708 [1:19:00<45:32,  2.03it/s]
 48%|████▊     | 5150/10708 [1:19:01<45:34,  2.03it/s]{'loss': 3.6928, 'grad_norm': 0.2077898532152176, 'learning_rate': 0.000619371725957709, 'epoch': 0.48}

                                                      

 48%|████▊     | 5150/10708 [1:19:01<45:34,  2.03it/s]
 48%|████▊     | 5151/10708 [1:19:01<45:38,  2.03it/s]
 48%|████▊     | 5152/10708 [1:19:02<45:37,  2.03it/s]
 48%|████▊     | 5153/10708 [1:19:02<45:34,  2.03it/s]
 48%|████▊     | 5154/10708 [1:19:03<45:37,  2.03it/s]
 48%|████▊     | 5155/10708 [1:19:03<45:38,  2.03it/s]
 48%|████▊     | 5156/10708 [1:19:04<45:33,  2.03it/s]
 48%|████▊     | 5157/10708 [1:19:04<45:32,  2.03it/s]
 48%|████▊     | 5158/10708 [1:19:05<45:30,  2.03it/s]
 48%|████▊     | 5159/10708 [1:19:05<45:27,  2.03it/s]
 48%|████▊     | 5160/10708 [1:19:06<45:26,  2.03it/s]
 48%|████▊     | 5161/10708 [1:19:06<45:32,  2.03it/s]
 48%|████▊     | 5162/10708 [1:19:07<45:27,  2.03it/s]
 48%|████▊     | 5163/10708 [1:19:07<45:31,  2.03it/s]
 48%|████▊     | 5164/10708 [1:19:08<45:29,  2.03it/s]
 48%|████▊     | 5165/10708 [1:19:08<45:30,  2.03it/s]
 48%|████▊     | 5166/10708 [1:19:09<45:27,  2.03it/s]
 48%|████▊     | 5167/10708 [1:19:09<45:23,  2.03it/s]
 48%|████▊     | 5168/10708 [1:19:10<45:21,  2.04it/s]
 48%|████▊     | 5169/10708 [1:19:10<45:26,  2.03it/s]
 48%|████▊     | 5170/10708 [1:19:11<45:28,  2.03it/s]
 48%|████▊     | 5171/10708 [1:19:11<45:29,  2.03it/s]
 48%|████▊     | 5172/10708 [1:19:11<45:26,  2.03it/s]
 48%|████▊     | 5173/10708 [1:19:12<45:24,  2.03it/s]
 48%|████▊     | 5174/10708 [1:19:12<45:22,  2.03it/s]
 48%|████▊     | 5175/10708 [1:19:13<45:29,  2.03it/s]
                                                      
{'loss': 3.6805, 'grad_norm': 0.21598412096500397, 'learning_rate': 0.0006154107309160549, 'epoch': 0.48}

 48%|████▊     | 5175/10708 [1:19:13<45:29,  2.03it/s]
 48%|████▊     | 5176/10708 [1:19:13<45:33,  2.02it/s]
 48%|████▊     | 5177/10708 [1:19:14<45:34,  2.02it/s]
 48%|████▊     | 5178/10708 [1:19:14<45:30,  2.03it/s]
 48%|████▊     | 5179/10708 [1:19:15<45:24,  2.03it/s]
 48%|████▊     | 5180/10708 [1:19:15<45:24,  2.03it/s]
 48%|████▊     | 5181/10708 [1:19:16<45:20,  2.03it/s]
 48%|████▊     | 5182/10708 [1:19:16<45:19,  2.03it/s]
 48%|████▊     | 5183/10708 [1:19:17<45:20,  2.03it/s]
 48%|████▊     | 5184/10708 [1:19:17<45:19,  2.03it/s]
 48%|████▊     | 5185/10708 [1:19:18<45:24,  2.03it/s]
 48%|████▊     | 5186/10708 [1:19:18<45:22,  2.03it/s]
 48%|████▊     | 5187/10708 [1:19:19<45:21,  2.03it/s]
 48%|████▊     | 5188/10708 [1:19:19<45:20,  2.03it/s]
 48%|████▊     | 5189/10708 [1:19:20<45:18,  2.03it/s]
 48%|████▊     | 5190/10708 [1:19:20<45:15,  2.03it/s]
 48%|████▊     | 5191/10708 [1:19:21<45:15,  2.03it/s]
 48%|████▊     | 5192/10708 [1:19:21<45:16,  2.03it/s]
 48%|████▊     | 5193/10708 [1:19:22<45:15,  2.03it/s]
 49%|████▊     | 5194/10708 [1:19:22<45:13,  2.03it/s]
 49%|████▊     | 5195/10708 [1:19:23<45:11,  2.03it/s]
 49%|████▊     | 5196/10708 [1:19:23<45:08,  2.04it/s]
 49%|████▊     | 5197/10708 [1:19:24<45:10,  2.03it/s]
 49%|████▊     | 5198/10708 [1:19:24<45:08,  2.03it/s]
 49%|████▊     | 5199/10708 [1:19:25<45:05,  2.04it/s]
 49%|████▊     | 5200/10708 [1:19:25<45:08,  2.03it/s]{'loss': 3.6767, 'grad_norm': 0.2086770385503769, 'learning_rate': 0.0006114420703859714, 'epoch': 0.49}

                                                      

 49%|████▊     | 5200/10708 [1:19:25<45:08,  2.03it/s]
 49%|████▊     | 5201/10708 [1:19:26<45:11,  2.03it/s]
 49%|████▊     | 5202/10708 [1:19:26<45:21,  2.02it/s]
 49%|████▊     | 5203/10708 [1:19:27<45:19,  2.02it/s]
 49%|████▊     | 5204/10708 [1:19:27<45:18,  2.02it/s]
 49%|████▊     | 5205/10708 [1:19:28<45:16,  2.03it/s]
 49%|████▊     | 5206/10708 [1:19:28<45:12,  2.03it/s]
 49%|████▊     | 5207/10708 [1:19:29<45:09,  2.03it/s]
 49%|████▊     | 5208/10708 [1:19:29<45:11,  2.03it/s]
 49%|████▊     | 5209/10708 [1:19:30<45:09,  2.03it/s]
 49%|████▊     | 5210/10708 [1:19:30<45:07,  2.03it/s]
 49%|████▊     | 5211/10708 [1:19:31<45:03,  2.03it/s]
 49%|████▊     | 5212/10708 [1:19:31<45:04,  2.03it/s]
 49%|████▊     | 5213/10708 [1:19:32<45:02,  2.03it/s]
 49%|████▊     | 5214/10708 [1:19:32<45:01,  2.03it/s]
 49%|████▊     | 5215/10708 [1:19:33<45:05,  2.03it/s]
 49%|████▊     | 5216/10708 [1:19:33<45:00,  2.03it/s]
 49%|████▊     | 5217/10708 [1:19:34<44:57,  2.04it/s]
 49%|████▊     | 5218/10708 [1:19:34<44:58,  2.03it/s]
 49%|████▊     | 5219/10708 [1:19:35<44:57,  2.04it/s]
 49%|████▊     | 5220/10708 [1:19:35<44:55,  2.04it/s]
 49%|████▉     | 5221/10708 [1:19:36<45:00,  2.03it/s]
 49%|████▉     | 5222/10708 [1:19:36<44:58,  2.03it/s]
 49%|████▉     | 5223/10708 [1:19:37<45:00,  2.03it/s]
 49%|████▉     | 5224/10708 [1:19:37<44:57,  2.03it/s]
 49%|████▉     | 5225/10708 [1:19:38<44:55,  2.03it/s]{'loss': 3.6771, 'grad_norm': 0.21891659498214722, 'learning_rate': 0.0006074660079627157, 'epoch': 0.49}

                                                      

 49%|████▉     | 5225/10708 [1:19:38<44:55,  2.03it/s]
 49%|████▉     | 5226/10708 [1:19:38<45:02,  2.03it/s]
 49%|████▉     | 5227/10708 [1:19:39<44:59,  2.03it/s]
 49%|████▉     | 5228/10708 [1:19:39<44:58,  2.03it/s]
 49%|████▉     | 5229/10708 [1:19:40<44:59,  2.03it/s]
 49%|████▉     | 5230/10708 [1:19:40<44:53,  2.03it/s]
 49%|████▉     | 5231/10708 [1:19:41<44:58,  2.03it/s]
 49%|████▉     | 5232/10708 [1:19:41<44:55,  2.03it/s]
 49%|████▉     | 5233/10708 [1:19:42<44:54,  2.03it/s]
 49%|████▉     | 5234/10708 [1:19:42<44:53,  2.03it/s]
 49%|████▉     | 5235/10708 [1:19:43<44:53,  2.03it/s]
 49%|████▉     | 5236/10708 [1:19:43<44:59,  2.03it/s]
 49%|████▉     | 5237/10708 [1:19:44<44:58,  2.03it/s]
 49%|████▉     | 5238/10708 [1:19:44<44:58,  2.03it/s]
 49%|████▉     | 5239/10708 [1:19:44<44:54,  2.03it/s]
 49%|████▉     | 5240/10708 [1:19:45<44:55,  2.03it/s]
 49%|████▉     | 5241/10708 [1:19:45<44:55,  2.03it/s]
 49%|████▉     | 5242/10708 [1:19:46<44:50,  2.03it/s]
 49%|████▉     | 5243/10708 [1:19:46<44:48,  2.03it/s]
 49%|████▉     | 5244/10708 [1:19:47<44:50,  2.03it/s]
 49%|████▉     | 5245/10708 [1:19:47<44:49,  2.03it/s]
 49%|████▉     | 5246/10708 [1:19:48<44:51,  2.03it/s]
 49%|████▉     | 5247/10708 [1:19:48<44:47,  2.03it/s]
 49%|████▉     | 5248/10708 [1:19:49<44:47,  2.03it/s]
 49%|████▉     | 5249/10708 [1:19:49<44:46,  2.03it/s]
 49%|████▉     | 5250/10708 [1:19:50<44:45,  2.03it/s]{'loss': 3.6807, 'grad_norm': 0.19686418771743774, 'learning_rate': 0.0006034828077331741, 'epoch': 0.49}

                                                      

 49%|████▉     | 5250/10708 [1:19:50<44:45,  2.03it/s]
 49%|████▉     | 5251/10708 [1:19:50<44:50,  2.03it/s]
 49%|████▉     | 5252/10708 [1:19:51<44:47,  2.03it/s]
 49%|████▉     | 5253/10708 [1:19:51<44:47,  2.03it/s]
 49%|████▉     | 5254/10708 [1:19:52<44:52,  2.03it/s]
 49%|████▉     | 5255/10708 [1:19:52<44:46,  2.03it/s]
 49%|████▉     | 5256/10708 [1:19:53<44:48,  2.03it/s]
 49%|████▉     | 5257/10708 [1:19:53<44:46,  2.03it/s]
 49%|████▉     | 5258/10708 [1:19:54<44:39,  2.03it/s]
 49%|████▉     | 5259/10708 [1:19:54<44:37,  2.04it/s]
 49%|████▉     | 5260/10708 [1:19:55<44:38,  2.03it/s]
 49%|████▉     | 5261/10708 [1:19:55<44:38,  2.03it/s]
 49%|████▉     | 5262/10708 [1:19:56<44:38,  2.03it/s]
 49%|████▉     | 5263/10708 [1:19:56<44:41,  2.03it/s]
 49%|████▉     | 5264/10708 [1:19:57<44:37,  2.03it/s]
 49%|████▉     | 5265/10708 [1:19:57<44:38,  2.03it/s]
 49%|████▉     | 5266/10708 [1:19:58<44:38,  2.03it/s]
 49%|████▉     | 5267/10708 [1:19:58<44:32,  2.04it/s]
 49%|████▉     | 5268/10708 [1:19:59<44:34,  2.03it/s]
 49%|████▉     | 5269/10708 [1:19:59<44:33,  2.03it/s]
 49%|████▉     | 5270/10708 [1:20:00<44:32,  2.03it/s]
 49%|████▉     | 5271/10708 [1:20:00<44:35,  2.03it/s]
 49%|████▉     | 5272/10708 [1:20:01<44:32,  2.03it/s]
 49%|████▉     | 5273/10708 [1:20:01<44:36,  2.03it/s]
 49%|████▉     | 5274/10708 [1:20:02<44:41,  2.03it/s]
 49%|████▉     | 5275/10708 [1:20:02<44:41,  2.03it/s]
                                                      
{'loss': 3.6729, 'grad_norm': 0.2121880054473877, 'learning_rate': 0.000599492734258319, 'epoch': 0.49}

 49%|████▉     | 5275/10708 [1:20:02<44:41,  2.03it/s]
 49%|████▉     | 5276/10708 [1:20:03<44:42,  2.02it/s]
 49%|████▉     | 5277/10708 [1:20:03<44:35,  2.03it/s]
 49%|████▉     | 5278/10708 [1:20:04<44:34,  2.03it/s]
 49%|████▉     | 5279/10708 [1:20:04<44:31,  2.03it/s]
 49%|████▉     | 5280/10708 [1:20:05<44:31,  2.03it/s]
 49%|████▉     | 5281/10708 [1:20:05<52:09,  1.73it/s]
 49%|████▉     | 5282/10708 [1:20:06<49:51,  1.81it/s]
 49%|████▉     | 5283/10708 [1:20:06<48:15,  1.87it/s]
 49%|████▉     | 5284/10708 [1:20:07<47:06,  1.92it/s]
 49%|████▉     | 5285/10708 [1:20:07<46:17,  1.95it/s]
 49%|████▉     | 5286/10708 [1:20:08<45:43,  1.98it/s]
 49%|████▉     | 5287/10708 [1:20:08<45:23,  1.99it/s]
 49%|████▉     | 5288/10708 [1:20:09<45:06,  2.00it/s]
 49%|████▉     | 5289/10708 [1:20:09<44:58,  2.01it/s]
 49%|████▉     | 5290/10708 [1:20:10<44:48,  2.01it/s]
 49%|████▉     | 5291/10708 [1:20:10<44:40,  2.02it/s]
 49%|████▉     | 5292/10708 [1:20:11<44:35,  2.02it/s]
 49%|████▉     | 5293/10708 [1:20:11<44:31,  2.03it/s]
 49%|████▉     | 5294/10708 [1:20:12<44:27,  2.03it/s]
 49%|████▉     | 5295/10708 [1:20:12<44:24,  2.03it/s]
 49%|████▉     | 5296/10708 [1:20:13<44:30,  2.03it/s]
 49%|████▉     | 5297/10708 [1:20:13<44:33,  2.02it/s]
 49%|████▉     | 5298/10708 [1:20:14<44:29,  2.03it/s]
 49%|████▉     | 5299/10708 [1:20:14<44:27,  2.03it/s]
 49%|████▉     | 5300/10708 [1:20:15<44:26,  2.03it/s]{'loss': 3.6767, 'grad_norm': 0.2113901525735855, 'learning_rate': 0.0005954960525556397, 'epoch': 0.49}
                                                      

 49%|████▉     | 5300/10708 [1:20:15<44:26,  2.03it/s]
 50%|████▉     | 5301/10708 [1:20:15<44:37,  2.02it/s]
 50%|████▉     | 5302/10708 [1:20:16<44:29,  2.02it/s]
 50%|████▉     | 5303/10708 [1:20:16<44:28,  2.03it/s]
 50%|████▉     | 5304/10708 [1:20:17<44:26,  2.03it/s]
 50%|████▉     | 5305/10708 [1:20:18<51:37,  1.74it/s]
 50%|████▉     | 5306/10708 [1:20:18<49:27,  1.82it/s]
 50%|████▉     | 5307/10708 [1:20:19<47:53,  1.88it/s]
 50%|████▉     | 5308/10708 [1:20:19<46:44,  1.93it/s]
 50%|████▉     | 5309/10708 [1:20:20<46:01,  1.95it/s]
 50%|████▉     | 5310/10708 [1:20:20<45:31,  1.98it/s]
 50%|████▉     | 5311/10708 [1:20:20<45:04,  2.00it/s]
 50%|████▉     | 5312/10708 [1:20:21<44:50,  2.01it/s]
 50%|████▉     | 5313/10708 [1:20:21<44:38,  2.01it/s]
 50%|████▉     | 5314/10708 [1:20:22<44:28,  2.02it/s]
 50%|████▉     | 5315/10708 [1:20:22<44:19,  2.03it/s]
 50%|████▉     | 5316/10708 [1:20:23<44:23,  2.02it/s]
 50%|████▉     | 5317/10708 [1:20:23<44:19,  2.03it/s]
 50%|████▉     | 5318/10708 [1:20:24<44:16,  2.03it/s]
 50%|████▉     | 5319/10708 [1:20:24<44:14,  2.03it/s]
 50%|████▉     | 5320/10708 [1:20:25<44:13,  2.03it/s]
 50%|████▉     | 5321/10708 [1:20:25<44:14,  2.03it/s]
 50%|████▉     | 5322/10708 [1:20:26<44:15,  2.03it/s]
 50%|████▉     | 5323/10708 [1:20:26<44:14,  2.03it/s]
 50%|████▉     | 5324/10708 [1:20:27<44:15,  2.03it/s]
 50%|████▉     | 5325/10708 [1:20:27<44:13,  2.03it/s]
                                                      
{'loss': 3.6796, 'grad_norm': 0.19681377708911896, 'learning_rate': 0.0005914930280815372, 'epoch': 0.5}

 50%|████▉     | 5325/10708 [1:20:27<44:13,  2.03it/s]
 50%|████▉     | 5326/10708 [1:20:28<44:12,  2.03it/s]
 50%|████▉     | 5327/10708 [1:20:28<44:08,  2.03it/s]
 50%|████▉     | 5328/10708 [1:20:29<44:06,  2.03it/s]
 50%|████▉     | 5329/10708 [1:20:29<44:07,  2.03it/s]
 50%|████▉     | 5330/10708 [1:20:30<44:05,  2.03it/s]
 50%|████▉     | 5331/10708 [1:20:30<44:06,  2.03it/s]
 50%|████▉     | 5332/10708 [1:20:31<44:07,  2.03it/s]
 50%|████▉     | 5333/10708 [1:20:31<44:04,  2.03it/s]
 50%|████▉     | 5334/10708 [1:20:32<44:02,  2.03it/s]
 50%|████▉     | 5335/10708 [1:20:32<44:01,  2.03it/s]
 50%|████▉     | 5336/10708 [1:20:33<44:04,  2.03it/s]
 50%|████▉     | 5337/10708 [1:20:33<44:04,  2.03it/s]
 50%|████▉     | 5338/10708 [1:20:34<44:11,  2.03it/s]
 50%|████▉     | 5339/10708 [1:20:34<44:07,  2.03it/s]
 50%|████▉     | 5340/10708 [1:20:35<44:01,  2.03it/s]
 50%|████▉     | 5341/10708 [1:20:35<44:01,  2.03it/s]
 50%|████▉     | 5342/10708 [1:20:36<43:58,  2.03it/s]
 50%|████▉     | 5343/10708 [1:20:36<43:58,  2.03it/s]
 50%|████▉     | 5344/10708 [1:20:37<44:00,  2.03it/s]
 50%|████▉     | 5345/10708 [1:20:37<43:57,  2.03it/s]
 50%|████▉     | 5346/10708 [1:20:38<43:57,  2.03it/s]
 50%|████▉     | 5347/10708 [1:20:38<43:57,  2.03it/s]
 50%|████▉     | 5348/10708 [1:20:39<43:52,  2.04it/s]
 50%|████▉     | 5349/10708 [1:20:39<43:52,  2.04it/s]
 50%|████▉     | 5350/10708 [1:20:40<43:56,  2.03it/s]{'loss': 3.6702, 'grad_norm': 0.19571775197982788, 'learning_rate': 0.0005874839267136956, 'epoch': 0.5}

                                                      

 50%|████▉     | 5350/10708 [1:20:40<43:56,  2.03it/s]
 50%|████▉     | 5351/10708 [1:20:40<44:05,  2.02it/s]
 50%|████▉     | 5352/10708 [1:20:41<44:00,  2.03it/s]
 50%|████▉     | 5353/10708 [1:20:41<43:59,  2.03it/s]
 50%|█████     | 5354/10708 [1:20:42<43:59,  2.03it/s]
 50%|█████     | 5355/10708 [1:20:42<43:53,  2.03it/s]
 50%|█████     | 5356/10708 [1:20:43<43:53,  2.03it/s]
 50%|█████     | 5357/10708 [1:20:43<44:03,  2.02it/s]
 50%|█████     | 5358/10708 [1:20:44<44:06,  2.02it/s]
 50%|█████     | 5359/10708 [1:20:44<43:58,  2.03it/s]
 50%|█████     | 5360/10708 [1:20:45<43:50,  2.03it/s]
 50%|█████     | 5361/10708 [1:20:45<43:51,  2.03it/s]
 50%|█████     | 5362/10708 [1:20:46<43:49,  2.03it/s]
 50%|█████     | 5363/10708 [1:20:46<43:46,  2.04it/s]
 50%|█████     | 5364/10708 [1:20:47<43:50,  2.03it/s]
 50%|█████     | 5365/10708 [1:20:47<43:45,  2.04it/s]
 50%|█████     | 5366/10708 [1:20:48<43:43,  2.04it/s]
 50%|█████     | 5367/10708 [1:20:48<43:49,  2.03it/s]
 50%|█████     | 5368/10708 [1:20:49<43:46,  2.03it/s]
 50%|█████     | 5369/10708 [1:20:49<43:47,  2.03it/s]
 50%|█████     | 5370/10708 [1:20:50<43:45,  2.03it/s]
 50%|█████     | 5371/10708 [1:20:50<43:42,  2.04it/s]
 50%|█████     | 5372/10708 [1:20:51<43:40,  2.04it/s]
 50%|█████     | 5373/10708 [1:20:51<43:42,  2.03it/s]
 50%|█████     | 5374/10708 [1:20:52<43:38,  2.04it/s]
 50%|█████     | 5375/10708 [1:20:52<43:39,  2.04it/s]
                                                      
{'loss': 3.6727, 'grad_norm': 0.2151510864496231, 'learning_rate': 0.0005834690147334202, 'epoch': 0.5}

 50%|█████     | 5375/10708 [1:20:52<43:39,  2.04it/s]
 50%|█████     | 5376/10708 [1:20:52<43:44,  2.03it/s]
 50%|█████     | 5377/10708 [1:20:53<43:40,  2.03it/s]
 50%|█████     | 5378/10708 [1:20:53<43:46,  2.03it/s]
 50%|█████     | 5379/10708 [1:20:54<43:41,  2.03it/s]
 50%|█████     | 5380/10708 [1:20:54<43:40,  2.03it/s]
 50%|█████     | 5381/10708 [1:20:55<43:41,  2.03it/s]
 50%|█████     | 5382/10708 [1:20:55<43:41,  2.03it/s]
 50%|█████     | 5383/10708 [1:20:56<43:40,  2.03it/s]
 50%|█████     | 5384/10708 [1:20:56<43:42,  2.03it/s]
 50%|█████     | 5385/10708 [1:20:57<43:42,  2.03it/s]
 50%|█████     | 5386/10708 [1:20:57<43:43,  2.03it/s]
 50%|█████     | 5387/10708 [1:20:58<43:37,  2.03it/s]
 50%|█████     | 5388/10708 [1:20:58<43:37,  2.03it/s]
 50%|█████     | 5389/10708 [1:20:59<43:40,  2.03it/s]
 50%|█████     | 5390/10708 [1:20:59<43:34,  2.03it/s]
 50%|█████     | 5391/10708 [1:21:00<43:34,  2.03it/s]
 50%|█████     | 5392/10708 [1:21:00<43:37,  2.03it/s]
 50%|█████     | 5393/10708 [1:21:01<43:34,  2.03it/s]
 50%|█████     | 5394/10708 [1:21:01<43:32,  2.03it/s]
 50%|█████     | 5395/10708 [1:21:02<43:32,  2.03it/s]
 50%|█████     | 5396/10708 [1:21:02<43:31,  2.03it/s]
 50%|█████     | 5397/10708 [1:21:03<43:31,  2.03it/s]
 50%|█████     | 5398/10708 [1:21:03<43:33,  2.03it/s]
 50%|█████     | 5399/10708 [1:21:04<43:26,  2.04it/s]
 50%|█████     | 5400/10708 [1:21:04<43:27,  2.04it/s]
                                                      
{'loss': 3.6734, 'grad_norm': 0.21585804224014282, 'learning_rate': 0.0005794485588079538, 'epoch': 0.5}

 50%|█████     | 5400/10708 [1:21:04<43:27,  2.04it/s]
 50%|█████     | 5401/10708 [1:21:05<43:32,  2.03it/s]
 50%|█████     | 5402/10708 [1:21:05<43:28,  2.03it/s]
 50%|█████     | 5403/10708 [1:21:06<43:27,  2.03it/s]
 50%|█████     | 5404/10708 [1:21:06<43:26,  2.03it/s]
 50%|█████     | 5405/10708 [1:21:07<43:25,  2.04it/s]
 50%|█████     | 5406/10708 [1:21:07<43:26,  2.03it/s]
 50%|█████     | 5407/10708 [1:21:08<43:26,  2.03it/s]
 51%|█████     | 5408/10708 [1:21:08<43:24,  2.03it/s]
 51%|█████     | 5409/10708 [1:21:09<43:25,  2.03it/s]
 51%|█████     | 5410/10708 [1:21:09<43:25,  2.03it/s]
 51%|█████     | 5411/10708 [1:21:10<43:27,  2.03it/s]
 51%|█████     | 5412/10708 [1:21:10<43:29,  2.03it/s]
 51%|█████     | 5413/10708 [1:21:11<43:29,  2.03it/s]
 51%|█████     | 5414/10708 [1:21:11<43:25,  2.03it/s]
 51%|█████     | 5415/10708 [1:21:12<43:24,  2.03it/s]
 51%|█████     | 5416/10708 [1:21:12<43:22,  2.03it/s]
 51%|█████     | 5417/10708 [1:21:13<43:22,  2.03it/s]
 51%|█████     | 5418/10708 [1:21:13<43:25,  2.03it/s]
 51%|█████     | 5419/10708 [1:21:14<43:22,  2.03it/s]
 51%|█████     | 5420/10708 [1:21:14<43:25,  2.03it/s]
 51%|█████     | 5421/10708 [1:21:15<43:22,  2.03it/s]
 51%|█████     | 5422/10708 [1:21:15<43:29,  2.03it/s]
 51%|█████     | 5423/10708 [1:21:16<43:25,  2.03it/s]
 51%|█████     | 5424/10708 [1:21:16<43:21,  2.03it/s]
 51%|█████     | 5425/10708 [1:21:17<43:25,  2.03it/s]
                                                      
{'loss': 3.6696, 'grad_norm': 0.20642201602458954, 'learning_rate': 0.0005754228259727626, 'epoch': 0.51}

 51%|█████     | 5425/10708 [1:21:17<43:25,  2.03it/s]
 51%|█████     | 5426/10708 [1:21:17<43:28,  2.03it/s]
 51%|█████     | 5427/10708 [1:21:18<43:25,  2.03it/s]
 51%|█████     | 5428/10708 [1:21:18<43:21,  2.03it/s]
 51%|█████     | 5429/10708 [1:21:19<43:15,  2.03it/s]
 51%|█████     | 5430/10708 [1:21:19<43:18,  2.03it/s]
 51%|█████     | 5431/10708 [1:21:20<43:13,  2.03it/s]
 51%|█████     | 5432/10708 [1:21:20<43:13,  2.03it/s]
 51%|█████     | 5433/10708 [1:21:21<43:14,  2.03it/s]
 51%|█████     | 5434/10708 [1:21:21<43:12,  2.03it/s]
 51%|█████     | 5435/10708 [1:21:22<43:14,  2.03it/s]
 51%|█████     | 5436/10708 [1:21:22<43:14,  2.03it/s]
 51%|█████     | 5437/10708 [1:21:23<44:41,  1.97it/s]
 51%|█████     | 5438/10708 [1:21:23<44:14,  1.99it/s]
 51%|█████     | 5439/10708 [1:21:24<43:58,  2.00it/s]
 51%|█████     | 5440/10708 [1:21:24<43:42,  2.01it/s]
 51%|█████     | 5441/10708 [1:21:25<44:09,  1.99it/s]
 51%|█████     | 5442/10708 [1:21:25<44:28,  1.97it/s]
 51%|█████     | 5443/10708 [1:21:26<44:04,  1.99it/s]
 51%|█████     | 5444/10708 [1:21:26<43:48,  2.00it/s]
 51%|█████     | 5445/10708 [1:21:27<43:37,  2.01it/s]
 51%|█████     | 5446/10708 [1:21:27<44:17,  1.98it/s]
 51%|█████     | 5447/10708 [1:21:28<43:52,  2.00it/s]
 51%|█████     | 5448/10708 [1:21:28<43:35,  2.01it/s]
 51%|█████     | 5449/10708 [1:21:29<43:24,  2.02it/s]
 51%|█████     | 5450/10708 [1:21:29<43:19,  2.02it/s]
                                                      
{'loss': 3.6606, 'grad_norm': 0.206051766872406, 'learning_rate': 0.0005713920836138015, 'epoch': 0.51}

 51%|█████     | 5450/10708 [1:21:29<43:19,  2.02it/s]
 51%|█████     | 5451/10708 [1:21:30<43:15,  2.03it/s]
 51%|█████     | 5452/10708 [1:21:30<43:08,  2.03it/s]
 51%|█████     | 5453/10708 [1:21:31<43:06,  2.03it/s]
 51%|█████     | 5454/10708 [1:21:31<43:06,  2.03it/s]
 51%|█████     | 5455/10708 [1:21:31<43:07,  2.03it/s]
 51%|█████     | 5456/10708 [1:21:32<43:48,  2.00it/s]
 51%|█████     | 5457/10708 [1:21:33<43:32,  2.01it/s]
 51%|█████     | 5458/10708 [1:21:33<43:19,  2.02it/s]
 51%|█████     | 5459/10708 [1:21:33<43:17,  2.02it/s]
 51%|█████     | 5460/10708 [1:21:34<43:10,  2.03it/s]
 51%|█████     | 5461/10708 [1:21:34<43:04,  2.03it/s]
 51%|█████     | 5462/10708 [1:21:35<43:03,  2.03it/s]
 51%|█████     | 5463/10708 [1:21:35<44:18,  1.97it/s]
 51%|█████     | 5464/10708 [1:21:36<43:57,  1.99it/s]
 51%|█████     | 5465/10708 [1:21:36<43:38,  2.00it/s]
 51%|█████     | 5466/10708 [1:21:37<43:25,  2.01it/s]
 51%|█████     | 5467/10708 [1:21:37<43:16,  2.02it/s]
 51%|█████     | 5468/10708 [1:21:38<43:08,  2.02it/s]
 51%|█████     | 5469/10708 [1:21:38<43:06,  2.03it/s]
 51%|█████     | 5470/10708 [1:21:39<43:04,  2.03it/s]
 51%|█████     | 5471/10708 [1:21:39<42:59,  2.03it/s]
 51%|█████     | 5472/10708 [1:21:40<42:53,  2.03it/s]
 51%|█████     | 5473/10708 [1:21:40<42:54,  2.03it/s]
 51%|█████     | 5474/10708 [1:21:41<42:54,  2.03it/s]
 51%|█████     | 5475/10708 [1:21:41<42:53,  2.03it/s]{'loss': 3.6672, 'grad_norm': 0.22587287425994873, 'learning_rate': 0.0005673565994497538, 'epoch': 0.51}
                                                      

 51%|█████     | 5475/10708 [1:21:41<42:53,  2.03it/s]
 51%|█████     | 5476/10708 [1:21:42<43:01,  2.03it/s]
 51%|█████     | 5477/10708 [1:21:42<42:57,  2.03it/s]
 51%|█████     | 5478/10708 [1:21:43<43:01,  2.03it/s]
 51%|█████     | 5479/10708 [1:21:43<43:05,  2.02it/s]
 51%|█████     | 5480/10708 [1:21:44<43:00,  2.03it/s]
 51%|█████     | 5481/10708 [1:21:44<42:56,  2.03it/s]
 51%|█████     | 5482/10708 [1:21:45<42:55,  2.03it/s]
 51%|█████     | 5483/10708 [1:21:45<42:52,  2.03it/s]
 51%|█████     | 5484/10708 [1:21:46<42:51,  2.03it/s]
 51%|█████     | 5485/10708 [1:21:46<42:52,  2.03it/s]
 51%|█████     | 5486/10708 [1:21:47<42:52,  2.03it/s]
 51%|█████     | 5487/10708 [1:21:47<42:51,  2.03it/s]
 51%|█████▏    | 5488/10708 [1:21:48<42:47,  2.03it/s]
 51%|█████▏    | 5489/10708 [1:21:48<42:47,  2.03it/s]
 51%|█████▏    | 5490/10708 [1:21:49<42:50,  2.03it/s]
 51%|█████▏    | 5491/10708 [1:21:49<42:50,  2.03it/s]
 51%|█████▏    | 5492/10708 [1:21:50<42:51,  2.03it/s]
 51%|█████▏    | 5493/10708 [1:21:50<42:48,  2.03it/s]
 51%|█████▏    | 5494/10708 [1:21:51<42:46,  2.03it/s]
 51%|█████▏    | 5495/10708 [1:21:51<42:47,  2.03it/s]
 51%|█████▏    | 5496/10708 [1:21:52<42:46,  2.03it/s]
 51%|█████▏    | 5497/10708 [1:21:52<42:44,  2.03it/s]
 51%|█████▏    | 5498/10708 [1:21:53<42:42,  2.03it/s]
 51%|█████▏    | 5499/10708 [1:21:53<42:40,  2.03it/s]
 51%|█████▏    | 5500/10708 [1:21:54<42:44,  2.03it/s]
                                                      
{'loss': 3.6682, 'grad_norm': 0.19558091461658478, 'learning_rate': 0.0005633166415142494, 'epoch': 0.51}

 51%|█████▏    | 5500/10708 [1:21:54<42:44,  2.03it/s]
 51%|█████▏    | 5501/10708 [1:21:54<42:42,  2.03it/s]
 51%|█████▏    | 5502/10708 [1:21:55<42:42,  2.03it/s]
 51%|█████▏    | 5503/10708 [1:21:55<42:42,  2.03it/s]
 51%|█████▏    | 5504/10708 [1:21:56<42:39,  2.03it/s]
 51%|█████▏    | 5505/10708 [1:21:56<42:39,  2.03it/s]
 51%|█████▏    | 5506/10708 [1:21:57<42:38,  2.03it/s]
 51%|█████▏    | 5507/10708 [1:21:57<42:40,  2.03it/s]
 51%|█████▏    | 5508/10708 [1:21:58<42:38,  2.03it/s]
 51%|█████▏    | 5509/10708 [1:21:58<42:37,  2.03it/s]
 51%|█████▏    | 5510/10708 [1:21:59<42:34,  2.03it/s]
 51%|█████▏    | 5511/10708 [1:21:59<42:35,  2.03it/s]
 51%|█████▏    | 5512/10708 [1:22:00<42:35,  2.03it/s]
 51%|█████▏    | 5513/10708 [1:22:00<42:35,  2.03it/s]
 51%|█████▏    | 5514/10708 [1:22:01<42:35,  2.03it/s]
 52%|█████▏    | 5515/10708 [1:22:01<42:32,  2.03it/s]
 52%|█████▏    | 5516/10708 [1:22:02<42:30,  2.04it/s]
 52%|█████▏    | 5517/10708 [1:22:02<42:32,  2.03it/s]
 52%|█████▏    | 5518/10708 [1:22:03<42:33,  2.03it/s]
 52%|█████▏    | 5519/10708 [1:22:03<42:34,  2.03it/s]
 52%|█████▏    | 5520/10708 [1:22:04<42:34,  2.03it/s]
 52%|█████▏    | 5521/10708 [1:22:04<42:31,  2.03it/s]
 52%|█████▏    | 5522/10708 [1:22:05<42:30,  2.03it/s]
 52%|█████▏    | 5523/10708 [1:22:05<42:28,  2.03it/s]
 52%|█████▏    | 5524/10708 [1:22:06<42:28,  2.03it/s]
 52%|█████▏    | 5525/10708 [1:22:06<42:28,  2.03it/s]{'loss': 3.6685, 'grad_norm': 0.2045350819826126, 'learning_rate': 0.0005592724781380634, 'epoch': 0.52}

                                                      

 52%|█████▏    | 5525/10708 [1:22:06<42:28,  2.03it/s]
 52%|█████▏    | 5526/10708 [1:22:07<42:28,  2.03it/s]
 52%|█████▏    | 5527/10708 [1:22:07<42:27,  2.03it/s]
 52%|█████▏    | 5528/10708 [1:22:07<42:27,  2.03it/s]
 52%|█████▏    | 5529/10708 [1:22:08<42:28,  2.03it/s]
 52%|█████▏    | 5530/10708 [1:22:08<42:33,  2.03it/s]
 52%|█████▏    | 5531/10708 [1:22:09<42:29,  2.03it/s]
 52%|█████▏    | 5532/10708 [1:22:09<42:28,  2.03it/s]
 52%|█████▏    | 5533/10708 [1:22:10<42:27,  2.03it/s]
 52%|█████▏    | 5534/10708 [1:22:10<42:24,  2.03it/s]
 52%|█████▏    | 5535/10708 [1:22:11<42:25,  2.03it/s]
 52%|█████▏    | 5536/10708 [1:22:11<42:26,  2.03it/s]
 52%|█████▏    | 5537/10708 [1:22:12<42:24,  2.03it/s]
 52%|█████▏    | 5538/10708 [1:22:12<42:22,  2.03it/s]
 52%|█████▏    | 5539/10708 [1:22:13<42:29,  2.03it/s]
 52%|█████▏    | 5540/10708 [1:22:13<42:32,  2.02it/s]
 52%|█████▏    | 5541/10708 [1:22:14<42:28,  2.03it/s]
 52%|█████▏    | 5542/10708 [1:22:14<42:27,  2.03it/s]
 52%|█████▏    | 5543/10708 [1:22:15<42:23,  2.03it/s]
 52%|█████▏    | 5544/10708 [1:22:15<42:25,  2.03it/s]
 52%|█████▏    | 5545/10708 [1:22:16<42:23,  2.03it/s]
 52%|█████▏    | 5546/10708 [1:22:16<42:23,  2.03it/s]
 52%|█████▏    | 5547/10708 [1:22:17<42:20,  2.03it/s]
 52%|█████▏    | 5548/10708 [1:22:17<42:20,  2.03it/s]
 52%|█████▏    | 5549/10708 [1:22:18<42:20,  2.03it/s]
 52%|█████▏    | 5550/10708 [1:22:18<42:17,  2.03it/s]
                                                      
{'loss': 3.6705, 'grad_norm': 0.2120000720024109, 'learning_rate': 0.0005552243779312915, 'epoch': 0.52}

 52%|█████▏    | 5550/10708 [1:22:18<42:17,  2.03it/s]
 52%|█████▏    | 5551/10708 [1:22:19<42:22,  2.03it/s]
 52%|█████▏    | 5552/10708 [1:22:19<42:16,  2.03it/s]
 52%|█████▏    | 5553/10708 [1:22:20<42:15,  2.03it/s]
 52%|█████▏    | 5554/10708 [1:22:20<42:13,  2.03it/s]
 52%|█████▏    | 5555/10708 [1:22:21<42:12,  2.03it/s]
 52%|█████▏    | 5556/10708 [1:22:21<42:15,  2.03it/s]
 52%|█████▏    | 5557/10708 [1:22:22<42:15,  2.03it/s]
 52%|█████▏    | 5558/10708 [1:22:22<42:13,  2.03it/s]
 52%|█████▏    | 5559/10708 [1:22:23<42:13,  2.03it/s]
 52%|█████▏    | 5560/10708 [1:22:23<42:12,  2.03it/s]
 52%|█████▏    | 5561/10708 [1:22:24<42:16,  2.03it/s]
 52%|█████▏    | 5562/10708 [1:22:24<42:12,  2.03it/s]
 52%|█████▏    | 5563/10708 [1:22:25<42:13,  2.03it/s]
 52%|█████▏    | 5564/10708 [1:22:25<42:12,  2.03it/s]
 52%|█████▏    | 5565/10708 [1:22:26<42:09,  2.03it/s]
 52%|█████▏    | 5566/10708 [1:22:26<42:11,  2.03it/s]
 52%|█████▏    | 5567/10708 [1:22:27<42:16,  2.03it/s]
 52%|█████▏    | 5568/10708 [1:22:27<42:12,  2.03it/s]
 52%|█████▏    | 5569/10708 [1:22:28<42:09,  2.03it/s]
 52%|█████▏    | 5570/10708 [1:22:28<42:08,  2.03it/s]
 52%|█████▏    | 5571/10708 [1:22:29<42:05,  2.03it/s]
 52%|█████▏    | 5572/10708 [1:22:29<42:03,  2.04it/s]
 52%|█████▏    | 5573/10708 [1:22:30<42:04,  2.03it/s]
 52%|█████▏    | 5574/10708 [1:22:30<42:06,  2.03it/s]
 52%|█████▏    | 5575/10708 [1:22:31<42:03,  2.03it/s]{'loss': 3.6622, 'grad_norm': 0.20891928672790527, 'learning_rate': 0.0005511726097655117, 'epoch': 0.52}
                                                      

 52%|█████▏    | 5575/10708 [1:22:31<42:03,  2.03it/s]
 52%|█████▏    | 5576/10708 [1:22:31<42:07,  2.03it/s]
 52%|█████▏    | 5577/10708 [1:22:32<42:04,  2.03it/s]
 52%|█████▏    | 5578/10708 [1:22:32<41:57,  2.04it/s]
 52%|█████▏    | 5579/10708 [1:22:33<41:57,  2.04it/s]
 52%|█████▏    | 5580/10708 [1:22:33<41:58,  2.04it/s]
 52%|█████▏    | 5581/10708 [1:22:34<41:56,  2.04it/s]
 52%|█████▏    | 5582/10708 [1:22:34<41:56,  2.04it/s]
 52%|█████▏    | 5583/10708 [1:22:35<41:55,  2.04it/s]
 52%|█████▏    | 5584/10708 [1:22:35<41:59,  2.03it/s]
 52%|█████▏    | 5585/10708 [1:22:36<41:54,  2.04it/s]
 52%|█████▏    | 5586/10708 [1:22:36<41:57,  2.03it/s]
 52%|█████▏    | 5587/10708 [1:22:37<42:00,  2.03it/s]
 52%|█████▏    | 5588/10708 [1:22:37<42:01,  2.03it/s]
 52%|█████▏    | 5589/10708 [1:22:38<42:03,  2.03it/s]
 52%|█████▏    | 5590/10708 [1:22:38<41:59,  2.03it/s]
 52%|█████▏    | 5591/10708 [1:22:38<42:00,  2.03it/s]
 52%|█████▏    | 5592/10708 [1:22:39<41:56,  2.03it/s]
 52%|█████▏    | 5593/10708 [1:22:39<41:58,  2.03it/s]
 52%|█████▏    | 5594/10708 [1:22:40<41:59,  2.03it/s]
 52%|█████▏    | 5595/10708 [1:22:40<41:57,  2.03it/s]
 52%|█████▏    | 5596/10708 [1:22:41<41:55,  2.03it/s]
 52%|█████▏    | 5597/10708 [1:22:41<41:53,  2.03it/s]
 52%|█████▏    | 5598/10708 [1:22:42<41:57,  2.03it/s]
 52%|█████▏    | 5599/10708 [1:22:42<41:58,  2.03it/s]
 52%|█████▏    | 5600/10708 [1:22:43<41:57,  2.03it/s]
                                                      
{'loss': 3.6596, 'grad_norm': 0.19663967192173004, 'learning_rate': 0.0005471174427559245, 'epoch': 0.52}

 52%|█████▏    | 5600/10708 [1:22:43<41:57,  2.03it/s]
 52%|█████▏    | 5601/10708 [1:22:43<42:02,  2.02it/s]
 52%|█████▏    | 5602/10708 [1:22:44<41:56,  2.03it/s]
 52%|█████▏    | 5603/10708 [1:22:44<41:58,  2.03it/s]
 52%|█████▏    | 5604/10708 [1:22:45<41:55,  2.03it/s]
 52%|█████▏    | 5605/10708 [1:22:45<41:54,  2.03it/s]
 52%|█████▏    | 5606/10708 [1:22:46<41:56,  2.03it/s]
 52%|█████▏    | 5607/10708 [1:22:46<41:52,  2.03it/s]
 52%|█████▏    | 5608/10708 [1:22:47<41:55,  2.03it/s]
 52%|█████▏    | 5609/10708 [1:22:47<41:52,  2.03it/s]
 52%|█████▏    | 5610/10708 [1:22:48<41:47,  2.03it/s]
 52%|█████▏    | 5611/10708 [1:22:48<41:46,  2.03it/s]
 52%|█████▏    | 5612/10708 [1:22:49<41:44,  2.03it/s]
 52%|█████▏    | 5613/10708 [1:22:49<41:42,  2.04it/s]
 52%|█████▏    | 5614/10708 [1:22:50<41:44,  2.03it/s]
 52%|█████▏    | 5615/10708 [1:22:50<41:45,  2.03it/s]
 52%|█████▏    | 5616/10708 [1:22:51<41:45,  2.03it/s]
 52%|█████▏    | 5617/10708 [1:22:51<41:46,  2.03it/s]
 52%|█████▏    | 5618/10708 [1:22:52<41:44,  2.03it/s]
 52%|█████▏    | 5619/10708 [1:22:52<41:46,  2.03it/s]
 52%|█████▏    | 5620/10708 [1:22:53<41:41,  2.03it/s]
 52%|█████▏    | 5621/10708 [1:22:53<41:41,  2.03it/s]
 53%|█████▎    | 5622/10708 [1:22:54<41:40,  2.03it/s]
 53%|█████▎    | 5623/10708 [1:22:54<41:41,  2.03it/s]
 53%|█████▎    | 5624/10708 [1:22:55<41:40,  2.03it/s]
 53%|█████▎    | 5625/10708 [1:22:55<41:40,  2.03it/s]
                                                      
{'loss': 3.6633, 'grad_norm': 0.2095218300819397, 'learning_rate': 0.0005430591462434792, 'epoch': 0.53}

 53%|█████▎    | 5625/10708 [1:22:55<41:40,  2.03it/s]
 53%|█████▎    | 5626/10708 [1:22:56<41:43,  2.03it/s]
 53%|█████▎    | 5627/10708 [1:22:56<41:41,  2.03it/s]
 53%|█████▎    | 5628/10708 [1:22:57<41:42,  2.03it/s]
 53%|█████▎    | 5629/10708 [1:22:57<41:39,  2.03it/s]
 53%|█████▎    | 5630/10708 [1:22:58<41:38,  2.03it/s]
 53%|█████▎    | 5631/10708 [1:22:58<41:37,  2.03it/s]
 53%|█████▎    | 5632/10708 [1:22:59<41:35,  2.03it/s]
 53%|█████▎    | 5633/10708 [1:22:59<41:36,  2.03it/s]
 53%|█████▎    | 5634/10708 [1:23:00<41:35,  2.03it/s]
 53%|█████▎    | 5635/10708 [1:23:00<41:38,  2.03it/s]
 53%|█████▎    | 5636/10708 [1:23:01<41:36,  2.03it/s]
 53%|█████▎    | 5637/10708 [1:23:01<41:35,  2.03it/s]
 53%|█████▎    | 5638/10708 [1:23:02<41:34,  2.03it/s]
 53%|█████▎    | 5639/10708 [1:23:02<41:35,  2.03it/s]
 53%|█████▎    | 5640/10708 [1:23:03<41:35,  2.03it/s]
 53%|█████▎    | 5641/10708 [1:23:03<41:36,  2.03it/s]
 53%|█████▎    | 5642/10708 [1:23:04<41:34,  2.03it/s]
 53%|█████▎    | 5643/10708 [1:23:04<41:35,  2.03it/s]
 53%|█████▎    | 5644/10708 [1:23:05<41:31,  2.03it/s]
 53%|█████▎    | 5645/10708 [1:23:05<41:30,  2.03it/s]
 53%|█████▎    | 5646/10708 [1:23:06<41:33,  2.03it/s]
 53%|█████▎    | 5647/10708 [1:23:06<41:30,  2.03it/s]
 53%|█████▎    | 5648/10708 [1:23:07<41:28,  2.03it/s]
 53%|█████▎    | 5649/10708 [1:23:07<41:29,  2.03it/s]
 53%|█████▎    | 5650/10708 [1:23:08<41:26,  2.03it/s]{'loss': 3.6583, 'grad_norm': 0.23184968531131744, 'learning_rate': 0.0005389979897769835, 'epoch': 0.53}
                                                      

 53%|█████▎    | 5650/10708 [1:23:08<41:26,  2.03it/s]
 53%|█████▎    | 5651/10708 [1:23:08<41:29,  2.03it/s]
 53%|█████▎    | 5652/10708 [1:23:09<41:28,  2.03it/s]
 53%|█████▎    | 5653/10708 [1:23:09<41:28,  2.03it/s]
 53%|█████▎    | 5654/10708 [1:23:10<41:29,  2.03it/s]
 53%|█████▎    | 5655/10708 [1:23:10<41:28,  2.03it/s]
 53%|█████▎    | 5656/10708 [1:23:10<41:22,  2.04it/s]
 53%|█████▎    | 5657/10708 [1:23:11<41:27,  2.03it/s]
 53%|█████▎    | 5658/10708 [1:23:11<41:26,  2.03it/s]
 53%|█████▎    | 5659/10708 [1:23:12<41:28,  2.03it/s]
 53%|█████▎    | 5660/10708 [1:23:12<41:23,  2.03it/s]
 53%|█████▎    | 5661/10708 [1:23:13<41:36,  2.02it/s]
 53%|█████▎    | 5662/10708 [1:23:13<41:36,  2.02it/s]
 53%|█████▎    | 5663/10708 [1:23:14<41:35,  2.02it/s]
 53%|█████▎    | 5664/10708 [1:23:14<41:27,  2.03it/s]
 53%|█████▎    | 5665/10708 [1:23:15<41:28,  2.03it/s]
 53%|█████▎    | 5666/10708 [1:23:15<41:23,  2.03it/s]
 53%|█████▎    | 5667/10708 [1:23:16<41:21,  2.03it/s]
 53%|█████▎    | 5668/10708 [1:23:16<41:19,  2.03it/s]
 53%|█████▎    | 5669/10708 [1:23:17<41:18,  2.03it/s]
 53%|█████▎    | 5670/10708 [1:23:17<41:18,  2.03it/s]
 53%|█████▎    | 5671/10708 [1:23:18<41:16,  2.03it/s]
 53%|█████▎    | 5672/10708 [1:23:18<41:17,  2.03it/s]
 53%|█████▎    | 5673/10708 [1:23:19<41:15,  2.03it/s]
 53%|█████▎    | 5674/10708 [1:23:19<41:16,  2.03it/s]
 53%|█████▎    | 5675/10708 [1:23:20<41:16,  2.03it/s]{'loss': 3.6602, 'grad_norm': 0.2061760127544403, 'learning_rate': 0.0005349342430952015, 'epoch': 0.53}

                                                      

 53%|█████▎    | 5675/10708 [1:23:20<41:16,  2.03it/s]
 53%|█████▎    | 5676/10708 [1:23:20<41:18,  2.03it/s]
 53%|█████▎    | 5677/10708 [1:23:21<41:19,  2.03it/s]
 53%|█████▎    | 5678/10708 [1:23:21<41:19,  2.03it/s]
 53%|█████▎    | 5679/10708 [1:23:22<41:14,  2.03it/s]
 53%|█████▎    | 5680/10708 [1:23:22<41:13,  2.03it/s]
 53%|█████▎    | 5681/10708 [1:23:23<41:16,  2.03it/s]
 53%|█████▎    | 5682/10708 [1:23:23<41:14,  2.03it/s]
 53%|█████▎    | 5683/10708 [1:23:24<41:12,  2.03it/s]
 53%|█████▎    | 5684/10708 [1:23:24<41:12,  2.03it/s]
 53%|█████▎    | 5685/10708 [1:23:25<41:09,  2.03it/s]
 53%|█████▎    | 5686/10708 [1:23:25<41:08,  2.03it/s]
 53%|█████▎    | 5687/10708 [1:23:26<41:09,  2.03it/s]
 53%|█████▎    | 5688/10708 [1:23:26<41:07,  2.03it/s]
 53%|█████▎    | 5689/10708 [1:23:27<41:09,  2.03it/s]
 53%|█████▎    | 5690/10708 [1:23:27<41:07,  2.03it/s]
 53%|█████▎    | 5691/10708 [1:23:28<41:03,  2.04it/s]
 53%|█████▎    | 5692/10708 [1:23:28<41:02,  2.04it/s]
 53%|█████▎    | 5693/10708 [1:23:29<41:03,  2.04it/s]
 53%|█████▎    | 5694/10708 [1:23:29<41:04,  2.03it/s]
 53%|█████▎    | 5695/10708 [1:23:30<41:08,  2.03it/s]
 53%|█████▎    | 5696/10708 [1:23:30<41:04,  2.03it/s]
 53%|█████▎    | 5697/10708 [1:23:31<41:02,  2.04it/s]
 53%|█████▎    | 5698/10708 [1:23:31<41:03,  2.03it/s]
 53%|█████▎    | 5699/10708 [1:23:32<41:01,  2.03it/s]
 53%|█████▎    | 5700/10708 [1:23:32<40:58,  2.04it/s]{'loss': 3.6605, 'grad_norm': 0.20227791368961334, 'learning_rate': 0.0005308681761089373, 'epoch': 0.53}

                                                      

 53%|█████▎    | 5700/10708 [1:23:32<40:58,  2.04it/s]
 53%|█████▎    | 5701/10708 [1:23:33<41:03,  2.03it/s]
 53%|█████▎    | 5702/10708 [1:23:33<41:00,  2.03it/s]
 53%|█████▎    | 5703/10708 [1:23:34<41:00,  2.03it/s]
 53%|█████▎    | 5704/10708 [1:23:34<41:04,  2.03it/s]
 53%|█████▎    | 5705/10708 [1:23:35<41:01,  2.03it/s]
 53%|█████▎    | 5706/10708 [1:23:35<41:02,  2.03it/s]
 53%|█████▎    | 5707/10708 [1:23:36<41:04,  2.03it/s]
 53%|█████▎    | 5708/10708 [1:23:36<41:02,  2.03it/s]
 53%|█████▎    | 5709/10708 [1:23:37<41:03,  2.03it/s]
 53%|█████▎    | 5710/10708 [1:23:37<41:01,  2.03it/s]
 53%|█████▎    | 5711/10708 [1:23:38<41:01,  2.03it/s]
 53%|█████▎    | 5712/10708 [1:23:38<40:59,  2.03it/s]
 53%|█████▎    | 5713/10708 [1:23:39<40:55,  2.03it/s]
 53%|█████▎    | 5714/10708 [1:23:39<40:58,  2.03it/s]
 53%|█████▎    | 5715/10708 [1:23:40<40:56,  2.03it/s]
 53%|█████▎    | 5716/10708 [1:23:40<40:55,  2.03it/s]
 53%|█████▎    | 5717/10708 [1:23:41<40:56,  2.03it/s]
 53%|█████▎    | 5718/10708 [1:23:41<40:55,  2.03it/s]
 53%|█████▎    | 5719/10708 [1:23:41<40:53,  2.03it/s]
 53%|█████▎    | 5720/10708 [1:23:42<40:53,  2.03it/s]
 53%|█████▎    | 5721/10708 [1:23:42<40:56,  2.03it/s]
 53%|█████▎    | 5722/10708 [1:23:43<40:58,  2.03it/s]
 53%|█████▎    | 5723/10708 [1:23:43<41:00,  2.03it/s]
 53%|█████▎    | 5724/10708 [1:23:44<40:57,  2.03it/s]
 53%|█████▎    | 5725/10708 [1:23:44<40:54,  2.03it/s]{'loss': 3.6522, 'grad_norm': 0.2060396671295166, 'learning_rate': 0.0005268000588831076, 'epoch': 0.53}
                                                      

 53%|█████▎    | 5725/10708 [1:23:44<40:54,  2.03it/s]
 53%|█████▎    | 5726/10708 [1:23:45<40:55,  2.03it/s]
 53%|█████▎    | 5727/10708 [1:23:45<40:53,  2.03it/s]
 53%|█████▎    | 5728/10708 [1:23:46<40:50,  2.03it/s]
 54%|█████▎    | 5729/10708 [1:23:46<40:55,  2.03it/s]
 54%|█████▎    | 5730/10708 [1:23:47<40:49,  2.03it/s]
 54%|█████▎    | 5731/10708 [1:23:47<40:45,  2.03it/s]
 54%|█████▎    | 5732/10708 [1:23:48<40:43,  2.04it/s]
 54%|█████▎    | 5733/10708 [1:23:48<40:43,  2.04it/s]
 54%|█████▎    | 5734/10708 [1:23:49<40:44,  2.04it/s]
 54%|█████▎    | 5735/10708 [1:23:49<40:41,  2.04it/s]
 54%|█████▎    | 5736/10708 [1:23:50<40:46,  2.03it/s]
 54%|█████▎    | 5737/10708 [1:23:50<40:42,  2.03it/s]
 54%|█████▎    | 5738/10708 [1:23:51<40:42,  2.03it/s]
 54%|█████▎    | 5739/10708 [1:23:51<40:45,  2.03it/s]
 54%|█████▎    | 5740/10708 [1:23:52<40:42,  2.03it/s]
 54%|█████▎    | 5741/10708 [1:23:52<40:44,  2.03it/s]
 54%|█████▎    | 5742/10708 [1:23:53<40:46,  2.03it/s]
 54%|█████▎    | 5743/10708 [1:23:53<40:43,  2.03it/s]
 54%|█████▎    | 5744/10708 [1:23:54<40:45,  2.03it/s]
 54%|█████▎    | 5745/10708 [1:23:54<40:45,  2.03it/s]
 54%|█████▎    | 5746/10708 [1:23:55<40:48,  2.03it/s]
 54%|█████▎    | 5747/10708 [1:23:55<40:44,  2.03it/s]
 54%|█████▎    | 5748/10708 [1:23:56<40:45,  2.03it/s]
 54%|█████▎    | 5749/10708 [1:23:56<40:44,  2.03it/s]
 54%|█████▎    | 5750/10708 [1:23:57<40:41,  2.03it/s]
                                                      
{'loss': 3.658, 'grad_norm': 0.2006624937057495, 'learning_rate': 0.0005227301616188044, 'epoch': 0.54}

 54%|█████▎    | 5750/10708 [1:23:57<40:41,  2.03it/s]
 54%|█████▎    | 5751/10708 [1:23:57<40:47,  2.03it/s]
 54%|█████▎    | 5752/10708 [1:23:58<40:40,  2.03it/s]
 54%|█████▎    | 5753/10708 [1:23:58<40:36,  2.03it/s]
 54%|█████▎    | 5754/10708 [1:23:59<40:37,  2.03it/s]
 54%|█████▎    | 5755/10708 [1:23:59<40:37,  2.03it/s]
 54%|█████▍    | 5756/10708 [1:24:00<40:33,  2.04it/s]
 54%|█████▍    | 5757/10708 [1:24:00<40:34,  2.03it/s]
 54%|█████▍    | 5758/10708 [1:24:01<40:33,  2.03it/s]
 54%|█████▍    | 5759/10708 [1:24:01<40:36,  2.03it/s]
 54%|█████▍    | 5760/10708 [1:24:02<40:35,  2.03it/s]
 54%|█████▍    | 5761/10708 [1:24:02<40:34,  2.03it/s]
 54%|█████▍    | 5762/10708 [1:24:03<40:32,  2.03it/s]
 54%|█████▍    | 5763/10708 [1:24:03<40:33,  2.03it/s]
 54%|█████▍    | 5764/10708 [1:24:04<40:30,  2.03it/s]
 54%|█████▍    | 5765/10708 [1:24:04<40:31,  2.03it/s]
 54%|█████▍    | 5766/10708 [1:24:05<40:31,  2.03it/s]
 54%|█████▍    | 5767/10708 [1:24:05<40:27,  2.04it/s]
 54%|█████▍    | 5768/10708 [1:24:06<40:31,  2.03it/s]
 54%|█████▍    | 5769/10708 [1:24:06<40:27,  2.03it/s]
 54%|█████▍    | 5770/10708 [1:24:07<40:28,  2.03it/s]
 54%|█████▍    | 5771/10708 [1:24:07<40:27,  2.03it/s]
 54%|█████▍    | 5772/10708 [1:24:08<40:24,  2.04it/s]
 54%|█████▍    | 5773/10708 [1:24:08<40:22,  2.04it/s]
 54%|█████▍    | 5774/10708 [1:24:09<40:24,  2.04it/s]
 54%|█████▍    | 5775/10708 [1:24:09<40:23,  2.04it/s]
                                                      
{'loss': 3.6575, 'grad_norm': 0.20077486336231232, 'learning_rate': 0.0005186587546353484, 'epoch': 0.54}

 54%|█████▍    | 5775/10708 [1:24:09<40:23,  2.04it/s]
 54%|█████▍    | 5776/10708 [1:24:10<40:26,  2.03it/s]
 54%|█████▍    | 5777/10708 [1:24:10<40:27,  2.03it/s]
 54%|█████▍    | 5778/10708 [1:24:11<40:27,  2.03it/s]
 54%|█████▍    | 5779/10708 [1:24:11<40:29,  2.03it/s]
 54%|█████▍    | 5780/10708 [1:24:12<40:26,  2.03it/s]
 54%|█████▍    | 5781/10708 [1:24:12<40:22,  2.03it/s]
 54%|█████▍    | 5782/10708 [1:24:12<40:23,  2.03it/s]
 54%|█████▍    | 5783/10708 [1:24:13<40:26,  2.03it/s]
 54%|█████▍    | 5784/10708 [1:24:13<40:27,  2.03it/s]
 54%|█████▍    | 5785/10708 [1:24:14<40:25,  2.03it/s]
 54%|█████▍    | 5786/10708 [1:24:14<40:24,  2.03it/s]
 54%|█████▍    | 5787/10708 [1:24:15<40:29,  2.03it/s]
 54%|█████▍    | 5788/10708 [1:24:15<40:25,  2.03it/s]
 54%|█████▍    | 5789/10708 [1:24:16<40:25,  2.03it/s]
 54%|█████▍    | 5790/10708 [1:24:16<40:26,  2.03it/s]
 54%|█████▍    | 5791/10708 [1:24:17<40:23,  2.03it/s]
 54%|█████▍    | 5792/10708 [1:24:17<40:20,  2.03it/s]
 54%|█████▍    | 5793/10708 [1:24:18<40:19,  2.03it/s]
 54%|█████▍    | 5794/10708 [1:24:18<40:19,  2.03it/s]
 54%|█████▍    | 5795/10708 [1:24:19<40:18,  2.03it/s]
 54%|█████▍    | 5796/10708 [1:24:19<40:18,  2.03it/s]
 54%|█████▍    | 5797/10708 [1:24:20<40:17,  2.03it/s]
 54%|█████▍    | 5798/10708 [1:24:20<40:14,  2.03it/s]
 54%|█████▍    | 5799/10708 [1:24:21<40:10,  2.04it/s]
 54%|█████▍    | 5800/10708 [1:24:21<40:13,  2.03it/s]
                                                      
{'loss': 3.6569, 'grad_norm': 0.215371236205101, 'learning_rate': 0.0005145861083523346, 'epoch': 0.54}

 54%|█████▍    | 5800/10708 [1:24:21<40:13,  2.03it/s]
 54%|█████▍    | 5801/10708 [1:24:22<40:26,  2.02it/s]
 54%|█████▍    | 5802/10708 [1:24:22<40:23,  2.02it/s]
 54%|█████▍    | 5803/10708 [1:24:23<40:17,  2.03it/s]
 54%|█████▍    | 5804/10708 [1:24:23<40:15,  2.03it/s]
 54%|█████▍    | 5805/10708 [1:24:24<40:15,  2.03it/s]
 54%|█████▍    | 5806/10708 [1:24:24<40:11,  2.03it/s]
 54%|█████▍    | 5807/10708 [1:24:25<40:11,  2.03it/s]
 54%|█████▍    | 5808/10708 [1:24:25<40:13,  2.03it/s]
 54%|█████▍    | 5809/10708 [1:24:26<40:11,  2.03it/s]
 54%|█████▍    | 5810/10708 [1:24:26<40:13,  2.03it/s]
 54%|█████▍    | 5811/10708 [1:24:27<40:12,  2.03it/s]
 54%|█████▍    | 5812/10708 [1:24:27<40:08,  2.03it/s]
 54%|█████▍    | 5813/10708 [1:24:28<40:10,  2.03it/s]
 54%|█████▍    | 5814/10708 [1:24:28<40:07,  2.03it/s]
 54%|█████▍    | 5815/10708 [1:24:29<40:12,  2.03it/s]
 54%|█████▍    | 5816/10708 [1:24:29<40:07,  2.03it/s]
 54%|█████▍    | 5817/10708 [1:24:30<40:08,  2.03it/s]
 54%|█████▍    | 5818/10708 [1:24:30<40:07,  2.03it/s]
 54%|█████▍    | 5819/10708 [1:24:31<40:04,  2.03it/s]
 54%|█████▍    | 5820/10708 [1:24:31<40:07,  2.03it/s]
 54%|█████▍    | 5821/10708 [1:24:32<40:05,  2.03it/s]
 54%|█████▍    | 5822/10708 [1:24:32<40:05,  2.03it/s]
 54%|█████▍    | 5823/10708 [1:24:33<40:03,  2.03it/s]
 54%|█████▍    | 5824/10708 [1:24:33<40:00,  2.03it/s]
 54%|█████▍    | 5825/10708 [1:24:34<40:04,  2.03it/s]{'loss': 3.6592, 'grad_norm': 0.19794149696826935, 'learning_rate': 0.0005105124932716718, 'epoch': 0.54}
                                                      

 54%|█████▍    | 5825/10708 [1:24:34<40:04,  2.03it/s]
 54%|█████▍    | 5826/10708 [1:24:34<40:09,  2.03it/s]
 54%|█████▍    | 5827/10708 [1:24:35<40:08,  2.03it/s]
 54%|█████▍    | 5828/10708 [1:24:35<40:05,  2.03it/s]
 54%|█████▍    | 5829/10708 [1:24:36<40:02,  2.03it/s]
 54%|█████▍    | 5830/10708 [1:24:36<40:03,  2.03it/s]
 54%|█████▍    | 5831/10708 [1:24:37<40:07,  2.03it/s]
 54%|█████▍    | 5832/10708 [1:24:37<40:04,  2.03it/s]
 54%|█████▍    | 5833/10708 [1:24:38<39:58,  2.03it/s]
 54%|█████▍    | 5834/10708 [1:24:38<39:57,  2.03it/s]
 54%|█████▍    | 5835/10708 [1:24:39<39:56,  2.03it/s]
 55%|█████▍    | 5836/10708 [1:24:39<39:54,  2.03it/s]
 55%|█████▍    | 5837/10708 [1:24:40<39:53,  2.04it/s]
 55%|█████▍    | 5838/10708 [1:24:40<39:53,  2.03it/s]
 55%|█████▍    | 5839/10708 [1:24:41<39:52,  2.04it/s]
 55%|█████▍    | 5840/10708 [1:24:41<39:51,  2.04it/s]
 55%|█████▍    | 5841/10708 [1:24:42<39:51,  2.04it/s]
 55%|█████▍    | 5842/10708 [1:24:42<39:48,  2.04it/s]
 55%|█████▍    | 5843/10708 [1:24:43<39:47,  2.04it/s]
 55%|█████▍    | 5844/10708 [1:24:43<39:52,  2.03it/s]
 55%|█████▍    | 5845/10708 [1:24:44<39:51,  2.03it/s]
 55%|█████▍    | 5846/10708 [1:24:44<39:52,  2.03it/s]
 55%|█████▍    | 5847/10708 [1:24:45<39:51,  2.03it/s]
 55%|█████▍    | 5848/10708 [1:24:45<39:49,  2.03it/s]
 55%|█████▍    | 5849/10708 [1:24:45<39:50,  2.03it/s]
 55%|█████▍    | 5850/10708 [1:24:46<39:46,  2.04it/s]
                                                      
{'loss': 3.6384, 'grad_norm': 0.2266710251569748, 'learning_rate': 0.0005064381799596149, 'epoch': 0.55}

 55%|█████▍    | 5850/10708 [1:24:46<39:46,  2.04it/s]
 55%|█████▍    | 5851/10708 [1:24:46<39:52,  2.03it/s]
 55%|█████▍    | 5852/10708 [1:24:47<39:51,  2.03it/s]
 55%|█████▍    | 5853/10708 [1:24:47<39:47,  2.03it/s]
 55%|█████▍    | 5854/10708 [1:24:48<39:46,  2.03it/s]
 55%|█████▍    | 5855/10708 [1:24:48<39:47,  2.03it/s]
 55%|█████▍    | 5856/10708 [1:24:49<39:45,  2.03it/s]
 55%|█████▍    | 5857/10708 [1:24:49<39:47,  2.03it/s]
 55%|█████▍    | 5858/10708 [1:24:50<39:49,  2.03it/s]
 55%|█████▍    | 5859/10708 [1:24:50<39:48,  2.03it/s]
 55%|█████▍    | 5860/10708 [1:24:51<39:48,  2.03it/s]
 55%|█████▍    | 5861/10708 [1:24:51<39:45,  2.03it/s]
 55%|█████▍    | 5862/10708 [1:24:52<39:43,  2.03it/s]
 55%|█████▍    | 5863/10708 [1:24:52<39:42,  2.03it/s]
 55%|█████▍    | 5864/10708 [1:24:53<39:44,  2.03it/s]
 55%|█████▍    | 5865/10708 [1:24:53<39:47,  2.03it/s]
 55%|█████▍    | 5866/10708 [1:24:54<39:43,  2.03it/s]
 55%|█████▍    | 5867/10708 [1:24:54<39:43,  2.03it/s]
 55%|█████▍    | 5868/10708 [1:24:55<39:42,  2.03it/s]
 55%|█████▍    | 5869/10708 [1:24:55<39:39,  2.03it/s]
 55%|█████▍    | 5870/10708 [1:24:56<39:38,  2.03it/s]
 55%|█████▍    | 5871/10708 [1:24:56<39:37,  2.03it/s]
 55%|█████▍    | 5872/10708 [1:24:57<39:35,  2.04it/s]
 55%|█████▍    | 5873/10708 [1:24:57<39:37,  2.03it/s]
 55%|█████▍    | 5874/10708 [1:24:58<39:36,  2.03it/s]
 55%|█████▍    | 5875/10708 [1:24:58<39:36,  2.03it/s]{'loss': 3.6416, 'grad_norm': 0.21231143176555634, 'learning_rate': 0.000502363439028795, 'epoch': 0.55}
                                                      

 55%|█████▍    | 5875/10708 [1:24:58<39:36,  2.03it/s]
 55%|█████▍    | 5876/10708 [1:24:59<39:42,  2.03it/s]
 55%|█████▍    | 5877/10708 [1:24:59<39:40,  2.03it/s]
 55%|█████▍    | 5878/10708 [1:25:00<39:37,  2.03it/s]
 55%|█████▍    | 5879/10708 [1:25:00<39:36,  2.03it/s]
 55%|█████▍    | 5880/10708 [1:25:01<39:34,  2.03it/s]
 55%|█████▍    | 5881/10708 [1:25:01<39:35,  2.03it/s]
 55%|█████▍    | 5882/10708 [1:25:02<39:33,  2.03it/s]
 55%|█████▍    | 5883/10708 [1:25:02<39:31,  2.03it/s]
 55%|█████▍    | 5884/10708 [1:25:03<39:29,  2.04it/s]
 55%|█████▍    | 5885/10708 [1:25:03<39:30,  2.03it/s]
 55%|█████▍    | 5886/10708 [1:25:04<39:30,  2.03it/s]
 55%|█████▍    | 5887/10708 [1:25:04<39:26,  2.04it/s]
 55%|█████▍    | 5888/10708 [1:25:05<39:29,  2.03it/s]
 55%|█████▍    | 5889/10708 [1:25:05<39:29,  2.03it/s]
 55%|█████▌    | 5890/10708 [1:25:06<39:28,  2.03it/s]
 55%|█████▌    | 5891/10708 [1:25:06<39:30,  2.03it/s]
 55%|█████▌    | 5892/10708 [1:25:07<39:27,  2.03it/s]
 55%|█████▌    | 5893/10708 [1:25:07<39:27,  2.03it/s]
 55%|█████▌    | 5894/10708 [1:25:08<39:27,  2.03it/s]
 55%|█████▌    | 5895/10708 [1:25:08<39:27,  2.03it/s]
 55%|█████▌    | 5896/10708 [1:25:09<39:27,  2.03it/s]
 55%|█████▌    | 5897/10708 [1:25:09<39:23,  2.04it/s]
 55%|█████▌    | 5898/10708 [1:25:10<39:26,  2.03it/s]
 55%|█████▌    | 5899/10708 [1:25:10<39:26,  2.03it/s]
 55%|█████▌    | 5900/10708 [1:25:11<39:26,  2.03it/s]
                                                      
{'loss': 3.6497, 'grad_norm': 0.20754210650920868, 'learning_rate': 0.0004982885411202452, 'epoch': 0.55}

 55%|█████▌    | 5900/10708 [1:25:11<39:26,  2.03it/s]
 55%|█████▌    | 5901/10708 [1:25:11<39:30,  2.03it/s]
 55%|█████▌    | 5902/10708 [1:25:12<39:29,  2.03it/s]
 55%|█████▌    | 5903/10708 [1:25:12<39:31,  2.03it/s]
 55%|█████▌    | 5904/10708 [1:25:13<39:27,  2.03it/s]
 55%|█████▌    | 5905/10708 [1:25:13<39:33,  2.02it/s]
 55%|█████▌    | 5906/10708 [1:25:14<39:27,  2.03it/s]
 55%|█████▌    | 5907/10708 [1:25:14<39:28,  2.03it/s]
 55%|█████▌    | 5908/10708 [1:25:15<39:27,  2.03it/s]
 55%|█████▌    | 5909/10708 [1:25:15<39:27,  2.03it/s]
 55%|█████▌    | 5910/10708 [1:25:16<39:27,  2.03it/s]
 55%|█████▌    | 5911/10708 [1:25:16<39:24,  2.03it/s]
 55%|█████▌    | 5912/10708 [1:25:16<39:25,  2.03it/s]
 55%|█████▌    | 5913/10708 [1:25:17<39:23,  2.03it/s]
 55%|█████▌    | 5914/10708 [1:25:17<39:29,  2.02it/s]
 55%|█████▌    | 5915/10708 [1:25:18<39:26,  2.03it/s]
 55%|█████▌    | 5916/10708 [1:25:18<39:22,  2.03it/s]
 55%|█████▌    | 5917/10708 [1:25:19<39:21,  2.03it/s]
 55%|█████▌    | 5918/10708 [1:25:19<39:19,  2.03it/s]
 55%|█████▌    | 5919/10708 [1:25:20<39:22,  2.03it/s]
 55%|█████▌    | 5920/10708 [1:25:20<39:22,  2.03it/s]
 55%|█████▌    | 5921/10708 [1:25:21<39:16,  2.03it/s]
 55%|█████▌    | 5922/10708 [1:25:21<39:15,  2.03it/s]
 55%|█████▌    | 5923/10708 [1:25:22<39:18,  2.03it/s]
 55%|█████▌    | 5924/10708 [1:25:22<39:13,  2.03it/s]
 55%|█████▌    | 5925/10708 [1:25:23<39:13,  2.03it/s]{'loss': 3.6411, 'grad_norm': 0.21027398109436035, 'learning_rate': 0.0004942137568854255, 'epoch': 0.55}

                                                      

 55%|█████▌    | 5925/10708 [1:25:23<39:13,  2.03it/s]
 55%|█████▌    | 5926/10708 [1:25:23<39:16,  2.03it/s]
 55%|█████▌    | 5927/10708 [1:25:24<39:15,  2.03it/s]
 55%|█████▌    | 5928/10708 [1:25:24<39:12,  2.03it/s]
 55%|█████▌    | 5929/10708 [1:25:25<39:11,  2.03it/s]
 55%|█████▌    | 5930/10708 [1:25:25<39:15,  2.03it/s]
 55%|█████▌    | 5931/10708 [1:25:26<39:14,  2.03it/s]
 55%|█████▌    | 5932/10708 [1:25:26<39:13,  2.03it/s]
 55%|█████▌    | 5933/10708 [1:25:27<39:12,  2.03it/s]
 55%|█████▌    | 5934/10708 [1:25:27<39:09,  2.03it/s]
 55%|█████▌    | 5935/10708 [1:25:28<39:09,  2.03it/s]
 55%|█████▌    | 5936/10708 [1:25:28<39:08,  2.03it/s]
 55%|█████▌    | 5937/10708 [1:25:29<39:06,  2.03it/s]
 55%|█████▌    | 5938/10708 [1:25:29<39:05,  2.03it/s]
 55%|█████▌    | 5939/10708 [1:25:30<39:04,  2.03it/s]
 55%|█████▌    | 5940/10708 [1:25:30<39:03,  2.03it/s]
 55%|█████▌    | 5941/10708 [1:25:31<39:03,  2.03it/s]
 55%|█████▌    | 5942/10708 [1:25:31<39:03,  2.03it/s]
 56%|█████▌    | 5943/10708 [1:25:32<39:02,  2.03it/s]
 56%|█████▌    | 5944/10708 [1:25:32<39:06,  2.03it/s]
 56%|█████▌    | 5945/10708 [1:25:33<39:02,  2.03it/s]
 56%|█████▌    | 5946/10708 [1:25:33<39:04,  2.03it/s]
 56%|█████▌    | 5947/10708 [1:25:34<39:03,  2.03it/s]
 56%|█████▌    | 5948/10708 [1:25:34<39:02,  2.03it/s]
 56%|█████▌    | 5949/10708 [1:25:35<39:03,  2.03it/s]
 56%|█████▌    | 5950/10708 [1:25:35<39:05,  2.03it/s]
                                                      
{'loss': 3.6449, 'grad_norm': 0.20810039341449738, 'learning_rate': 0.0004901393569682447, 'epoch': 0.56}

 56%|█████▌    | 5950/10708 [1:25:35<39:05,  2.03it/s]
 56%|█████▌    | 5951/10708 [1:25:36<39:06,  2.03it/s]
 56%|█████▌    | 5952/10708 [1:25:36<39:03,  2.03it/s]
 56%|█████▌    | 5953/10708 [1:25:37<39:01,  2.03it/s]
 56%|█████▌    | 5954/10708 [1:25:37<39:01,  2.03it/s]
 56%|█████▌    | 5955/10708 [1:25:38<39:03,  2.03it/s]
 56%|█████▌    | 5956/10708 [1:25:38<39:00,  2.03it/s]
 56%|█████▌    | 5957/10708 [1:25:39<39:01,  2.03it/s]
 56%|█████▌    | 5958/10708 [1:25:39<39:01,  2.03it/s]
 56%|█████▌    | 5959/10708 [1:25:40<38:57,  2.03it/s]
 56%|█████▌    | 5960/10708 [1:25:40<38:59,  2.03it/s]
 56%|█████▌    | 5961/10708 [1:25:41<38:56,  2.03it/s]
 56%|█████▌    | 5962/10708 [1:25:41<38:54,  2.03it/s]
 56%|█████▌    | 5963/10708 [1:25:42<38:54,  2.03it/s]
 56%|█████▌    | 5964/10708 [1:25:42<38:52,  2.03it/s]
 56%|█████▌    | 5965/10708 [1:25:43<38:58,  2.03it/s]
 56%|█████▌    | 5966/10708 [1:25:43<38:59,  2.03it/s]
 56%|█████▌    | 5967/10708 [1:25:44<38:57,  2.03it/s]
 56%|█████▌    | 5968/10708 [1:25:44<38:54,  2.03it/s]
 56%|█████▌    | 5969/10708 [1:25:45<38:52,  2.03it/s]
 56%|█████▌    | 5970/10708 [1:25:45<38:55,  2.03it/s]
 56%|█████▌    | 5971/10708 [1:25:46<38:51,  2.03it/s]
 56%|█████▌    | 5972/10708 [1:25:46<38:51,  2.03it/s]
 56%|█████▌    | 5973/10708 [1:25:47<38:50,  2.03it/s]
 56%|█████▌    | 5974/10708 [1:25:47<38:50,  2.03it/s]
 56%|█████▌    | 5975/10708 [1:25:48<38:49,  2.03it/s]{'loss': 3.6349, 'grad_norm': 0.19367223978042603, 'learning_rate': 0.0004860656119870867, 'epoch': 0.56}
                                                      

 56%|█████▌    | 5975/10708 [1:25:48<38:49,  2.03it/s]
 56%|█████▌    | 5976/10708 [1:25:48<39:12,  2.01it/s]
 56%|█████▌    | 5977/10708 [1:25:49<39:02,  2.02it/s]
 56%|█████▌    | 5978/10708 [1:25:49<39:00,  2.02it/s]
 56%|█████▌    | 5979/10708 [1:25:50<38:55,  2.02it/s]
 56%|█████▌    | 5980/10708 [1:25:50<38:51,  2.03it/s]
 56%|█████▌    | 5981/10708 [1:25:50<38:49,  2.03it/s]
 56%|█████▌    | 5982/10708 [1:25:51<38:47,  2.03it/s]
 56%|█████▌    | 5983/10708 [1:25:51<38:44,  2.03it/s]
 56%|█████▌    | 5984/10708 [1:25:52<38:43,  2.03it/s]
 56%|█████▌    | 5985/10708 [1:25:52<38:47,  2.03it/s]
 56%|█████▌    | 5986/10708 [1:25:53<38:44,  2.03it/s]
 56%|█████▌    | 5987/10708 [1:25:53<38:48,  2.03it/s]
 56%|█████▌    | 5988/10708 [1:25:54<38:45,  2.03it/s]
 56%|█████▌    | 5989/10708 [1:25:55<45:23,  1.73it/s]
 56%|█████▌    | 5990/10708 [1:25:55<43:22,  1.81it/s]
 56%|█████▌    | 5991/10708 [1:25:56<41:56,  1.87it/s]
 56%|█████▌    | 5992/10708 [1:25:56<41:00,  1.92it/s]
 56%|█████▌    | 5993/10708 [1:25:57<40:21,  1.95it/s]
 56%|█████▌    | 5994/10708 [1:25:57<39:47,  1.97it/s]
 56%|█████▌    | 5995/10708 [1:25:58<39:26,  1.99it/s]
 56%|█████▌    | 5996/10708 [1:25:58<39:12,  2.00it/s]
 56%|█████▌    | 5997/10708 [1:25:59<39:07,  2.01it/s]
 56%|█████▌    | 5998/10708 [1:25:59<38:56,  2.02it/s]
 56%|█████▌    | 5999/10708 [1:26:00<38:53,  2.02it/s]
 56%|█████▌    | 6000/10708 [1:26:00<38:45,  2.02it/s]{'loss': 3.6474, 'grad_norm': 0.22761598229408264, 'learning_rate': 0.00048199279251683405, 'epoch': 0.56}

                                                      

 56%|█████▌    | 6000/10708 [1:26:00<38:45,  2.02it/s]
 56%|█████▌    | 6001/10708 [1:26:01<38:46,  2.02it/s]
 56%|█████▌    | 6002/10708 [1:26:01<38:44,  2.02it/s]
 56%|█████▌    | 6003/10708 [1:26:02<38:41,  2.03it/s]
 56%|█████▌    | 6004/10708 [1:26:02<38:38,  2.03it/s]
 56%|█████▌    | 6005/10708 [1:26:03<38:38,  2.03it/s]
 56%|█████▌    | 6006/10708 [1:26:03<38:37,  2.03it/s]
 56%|█████▌    | 6007/10708 [1:26:04<38:34,  2.03it/s]
 56%|█████▌    | 6008/10708 [1:26:04<38:35,  2.03it/s]
 56%|█████▌    | 6009/10708 [1:26:05<38:33,  2.03it/s]
 56%|█████▌    | 6010/10708 [1:26:05<38:32,  2.03it/s]
 56%|█████▌    | 6011/10708 [1:26:06<38:33,  2.03it/s]
 56%|█████▌    | 6012/10708 [1:26:06<38:31,  2.03it/s]
 56%|█████▌    | 6013/10708 [1:26:07<38:32,  2.03it/s]
 56%|█████▌    | 6014/10708 [1:26:07<44:49,  1.75it/s]
 56%|█████▌    | 6015/10708 [1:26:08<42:54,  1.82it/s]
 56%|█████▌    | 6016/10708 [1:26:08<41:31,  1.88it/s]
 56%|█████▌    | 6017/10708 [1:26:09<40:38,  1.92it/s]
 56%|█████▌    | 6018/10708 [1:26:09<40:01,  1.95it/s]
 56%|█████▌    | 6019/10708 [1:26:10<39:30,  1.98it/s]
 56%|█████▌    | 6020/10708 [1:26:10<39:09,  2.00it/s]
 56%|█████▌    | 6021/10708 [1:26:11<38:56,  2.01it/s]
 56%|█████▌    | 6022/10708 [1:26:11<38:43,  2.02it/s]
 56%|█████▌    | 6023/10708 [1:26:12<38:35,  2.02it/s]
 56%|█████▋    | 6024/10708 [1:26:12<38:33,  2.03it/s]
 56%|█████▋    | 6025/10708 [1:26:13<38:32,  2.03it/s]
                                                      
{'loss': 3.6367, 'grad_norm': 0.2126239389181137, 'learning_rate': 0.00047792116907089863, 'epoch': 0.56}

 56%|█████▋    | 6025/10708 [1:26:13<38:32,  2.03it/s]
 56%|█████▋    | 6026/10708 [1:26:13<38:39,  2.02it/s]
 56%|█████▋    | 6027/10708 [1:26:14<38:33,  2.02it/s]
 56%|█████▋    | 6028/10708 [1:26:14<38:30,  2.03it/s]
 56%|█████▋    | 6029/10708 [1:26:15<38:27,  2.03it/s]
 56%|█████▋    | 6030/10708 [1:26:15<38:22,  2.03it/s]
 56%|█████▋    | 6031/10708 [1:26:16<38:21,  2.03it/s]
 56%|█████▋    | 6032/10708 [1:26:16<38:18,  2.03it/s]
 56%|█████▋    | 6033/10708 [1:26:17<38:16,  2.04it/s]
 56%|█████▋    | 6034/10708 [1:26:17<38:16,  2.04it/s]
 56%|█████▋    | 6035/10708 [1:26:18<38:15,  2.04it/s]
 56%|█████▋    | 6036/10708 [1:26:18<38:13,  2.04it/s]
 56%|█████▋    | 6037/10708 [1:26:19<38:14,  2.04it/s]
 56%|█████▋    | 6038/10708 [1:26:19<38:14,  2.04it/s]
 56%|█████▋    | 6039/10708 [1:26:20<38:12,  2.04it/s]
 56%|█████▋    | 6040/10708 [1:26:20<38:14,  2.03it/s]
 56%|█████▋    | 6041/10708 [1:26:21<38:15,  2.03it/s]
 56%|█████▋    | 6042/10708 [1:26:21<38:10,  2.04it/s]
 56%|█████▋    | 6043/10708 [1:26:22<38:11,  2.04it/s]
 56%|█████▋    | 6044/10708 [1:26:22<38:11,  2.04it/s]
 56%|█████▋    | 6045/10708 [1:26:23<38:09,  2.04it/s]
 56%|█████▋    | 6046/10708 [1:26:23<38:12,  2.03it/s]
 56%|█████▋    | 6047/10708 [1:26:24<38:11,  2.03it/s]
 56%|█████▋    | 6048/10708 [1:26:24<38:08,  2.04it/s]
 56%|█████▋    | 6049/10708 [1:26:24<38:11,  2.03it/s]
 56%|█████▋    | 6050/10708 [1:26:25<38:12,  2.03it/s]{'loss': 3.6352, 'grad_norm': 0.21229113638401031, 'learning_rate': 0.00047385101208325233, 'epoch': 0.56}
                                                      

 56%|█████▋    | 6050/10708 [1:26:25<38:12,  2.03it/s]
 57%|█████▋    | 6051/10708 [1:26:25<38:20,  2.02it/s]
 57%|█████▋    | 6052/10708 [1:26:26<38:14,  2.03it/s]
 57%|█████▋    | 6053/10708 [1:26:26<38:13,  2.03it/s]
 57%|█████▋    | 6054/10708 [1:26:27<38:15,  2.03it/s]
 57%|█████▋    | 6055/10708 [1:26:27<38:12,  2.03it/s]
 57%|█████▋    | 6056/10708 [1:26:28<38:16,  2.03it/s]
 57%|█████▋    | 6057/10708 [1:26:28<38:13,  2.03it/s]
 57%|█████▋    | 6058/10708 [1:26:29<38:11,  2.03it/s]
 57%|█████▋    | 6059/10708 [1:26:29<38:09,  2.03it/s]
 57%|█████▋    | 6060/10708 [1:26:30<38:06,  2.03it/s]
 57%|█████▋    | 6061/10708 [1:26:30<38:07,  2.03it/s]
 57%|█████▋    | 6062/10708 [1:26:31<38:03,  2.03it/s]
 57%|█████▋    | 6063/10708 [1:26:31<38:02,  2.03it/s]
 57%|█████▋    | 6064/10708 [1:26:32<38:01,  2.04it/s]
 57%|█████▋    | 6065/10708 [1:26:32<38:02,  2.03it/s]
 57%|█████▋    | 6066/10708 [1:26:33<38:03,  2.03it/s]
 57%|█████▋    | 6067/10708 [1:26:33<38:07,  2.03it/s]
 57%|█████▋    | 6068/10708 [1:26:34<38:03,  2.03it/s]
 57%|█████▋    | 6069/10708 [1:26:34<38:03,  2.03it/s]
 57%|█████▋    | 6070/10708 [1:26:35<38:02,  2.03it/s]
 57%|█████▋    | 6071/10708 [1:26:35<37:57,  2.04it/s]
 57%|█████▋    | 6072/10708 [1:26:36<37:56,  2.04it/s]
 57%|█████▋    | 6073/10708 [1:26:36<38:00,  2.03it/s]
 57%|█████▋    | 6074/10708 [1:26:37<37:57,  2.04it/s]
 57%|█████▋    | 6075/10708 [1:26:37<38:00,  2.03it/s]
                                                      
{'loss': 3.6412, 'grad_norm': 0.19763733446598053, 'learning_rate': 0.00046978259189046695, 'epoch': 0.57}

 57%|█████▋    | 6075/10708 [1:26:37<38:00,  2.03it/s]
 57%|█████▋    | 6076/10708 [1:26:38<38:00,  2.03it/s]
 57%|█████▋    | 6077/10708 [1:26:38<38:00,  2.03it/s]
 57%|█████▋    | 6078/10708 [1:26:39<38:00,  2.03it/s]
 57%|█████▋    | 6079/10708 [1:26:39<37:57,  2.03it/s]
 57%|█████▋    | 6080/10708 [1:26:40<37:55,  2.03it/s]
 57%|█████▋    | 6081/10708 [1:26:40<37:55,  2.03it/s]
 57%|█████▋    | 6082/10708 [1:26:41<37:54,  2.03it/s]
 57%|█████▋    | 6083/10708 [1:26:41<37:51,  2.04it/s]
 57%|█████▋    | 6084/10708 [1:26:42<37:54,  2.03it/s]
 57%|█████▋    | 6085/10708 [1:26:42<37:52,  2.03it/s]
 57%|█████▋    | 6086/10708 [1:26:43<37:54,  2.03it/s]
 57%|█████▋    | 6087/10708 [1:26:43<38:00,  2.03it/s]
 57%|█████▋    | 6088/10708 [1:26:44<37:56,  2.03it/s]
 57%|█████▋    | 6089/10708 [1:26:44<37:55,  2.03it/s]
 57%|█████▋    | 6090/10708 [1:26:45<37:52,  2.03it/s]
 57%|█████▋    | 6091/10708 [1:26:45<37:50,  2.03it/s]
 57%|█████▋    | 6092/10708 [1:26:46<37:51,  2.03it/s]
 57%|█████▋    | 6093/10708 [1:26:46<37:50,  2.03it/s]
 57%|█████▋    | 6094/10708 [1:26:47<37:49,  2.03it/s]
 57%|█████▋    | 6095/10708 [1:26:47<37:52,  2.03it/s]
 57%|█████▋    | 6096/10708 [1:26:48<37:52,  2.03it/s]
 57%|█████▋    | 6097/10708 [1:26:48<37:52,  2.03it/s]
 57%|█████▋    | 6098/10708 [1:26:49<37:50,  2.03it/s]
 57%|█████▋    | 6099/10708 [1:26:49<37:47,  2.03it/s]
 57%|█████▋    | 6100/10708 [1:26:50<37:49,  2.03it/s]{'loss': 3.6401, 'grad_norm': 0.22421593964099884, 'learning_rate': 0.0004657161787137569, 'epoch': 0.57}
                                                      

 57%|█████▋    | 6100/10708 [1:26:50<37:49,  2.03it/s]
 57%|█████▋    | 6101/10708 [1:26:50<37:52,  2.03it/s]
 57%|█████▋    | 6102/10708 [1:26:51<37:50,  2.03it/s]
 57%|█████▋    | 6103/10708 [1:26:51<37:49,  2.03it/s]
 57%|█████▋    | 6104/10708 [1:26:52<37:52,  2.03it/s]
 57%|█████▋    | 6105/10708 [1:26:52<37:46,  2.03it/s]
 57%|█████▋    | 6106/10708 [1:26:53<37:42,  2.03it/s]
 57%|█████▋    | 6107/10708 [1:26:53<37:45,  2.03it/s]
 57%|█████▋    | 6108/10708 [1:26:54<37:47,  2.03it/s]
 57%|█████▋    | 6109/10708 [1:26:54<37:46,  2.03it/s]
 57%|█████▋    | 6110/10708 [1:26:55<37:43,  2.03it/s]
 57%|█████▋    | 6111/10708 [1:26:55<37:40,  2.03it/s]
 57%|█████▋    | 6112/10708 [1:26:56<37:42,  2.03it/s]
 57%|█████▋    | 6113/10708 [1:26:56<37:41,  2.03it/s]
 57%|█████▋    | 6114/10708 [1:26:56<37:40,  2.03it/s]
 57%|█████▋    | 6115/10708 [1:26:57<37:41,  2.03it/s]
 57%|█████▋    | 6116/10708 [1:26:57<37:37,  2.03it/s]
 57%|█████▋    | 6117/10708 [1:26:58<37:40,  2.03it/s]
 57%|█████▋    | 6118/10708 [1:26:58<37:39,  2.03it/s]
 57%|█████▋    | 6119/10708 [1:26:59<37:37,  2.03it/s]
 57%|█████▋    | 6120/10708 [1:26:59<37:35,  2.03it/s]
 57%|█████▋    | 6121/10708 [1:27:00<37:35,  2.03it/s]
 57%|█████▋    | 6122/10708 [1:27:00<37:33,  2.04it/s]
 57%|█████▋    | 6123/10708 [1:27:01<37:32,  2.04it/s]
 57%|█████▋    | 6124/10708 [1:27:01<37:33,  2.03it/s]
 57%|█████▋    | 6125/10708 [1:27:02<37:31,  2.04it/s]{'loss': 3.6298, 'grad_norm': 0.2113819718360901, 'learning_rate': 0.000461652042641033, 'epoch': 0.57}

                                                      

 57%|█████▋    | 6125/10708 [1:27:02<37:31,  2.04it/s]
 57%|█████▋    | 6126/10708 [1:27:02<37:33,  2.03it/s]
 57%|█████▋    | 6127/10708 [1:27:03<37:37,  2.03it/s]
 57%|█████▋    | 6128/10708 [1:27:03<37:33,  2.03it/s]
 57%|█████▋    | 6129/10708 [1:27:04<37:36,  2.03it/s]
 57%|█████▋    | 6130/10708 [1:27:04<37:36,  2.03it/s]
 57%|█████▋    | 6131/10708 [1:27:05<37:34,  2.03it/s]
 57%|█████▋    | 6132/10708 [1:27:05<37:33,  2.03it/s]
 57%|█████▋    | 6133/10708 [1:27:06<37:32,  2.03it/s]
 57%|█████▋    | 6134/10708 [1:27:06<37:30,  2.03it/s]
 57%|█████▋    | 6135/10708 [1:27:07<37:26,  2.04it/s]
 57%|█████▋    | 6136/10708 [1:27:07<37:26,  2.03it/s]
 57%|█████▋    | 6137/10708 [1:27:08<37:29,  2.03it/s]
 57%|█████▋    | 6138/10708 [1:27:08<37:27,  2.03it/s]
 57%|█████▋    | 6139/10708 [1:27:09<37:27,  2.03it/s]
 57%|█████▋    | 6140/10708 [1:27:09<37:29,  2.03it/s]
 57%|█████▋    | 6141/10708 [1:27:10<37:29,  2.03it/s]
 57%|█████▋    | 6142/10708 [1:27:10<37:31,  2.03it/s]
 57%|█████▋    | 6143/10708 [1:27:11<37:27,  2.03it/s]
 57%|█████▋    | 6144/10708 [1:27:11<37:23,  2.03it/s]
 57%|█████▋    | 6145/10708 [1:27:12<37:23,  2.03it/s]
 57%|█████▋    | 6146/10708 [1:27:12<37:24,  2.03it/s]
 57%|█████▋    | 6147/10708 [1:27:13<37:28,  2.03it/s]
 57%|█████▋    | 6148/10708 [1:27:13<37:33,  2.02it/s]
 57%|█████▋    | 6149/10708 [1:27:14<37:29,  2.03it/s]
 57%|█████▋    | 6150/10708 [1:27:14<37:27,  2.03it/s]
                                                      
{'loss': 3.6324, 'grad_norm': 0.2047315537929535, 'learning_rate': 0.000457590453608962, 'epoch': 0.57}

 57%|█████▋    | 6150/10708 [1:27:14<37:27,  2.03it/s]
 57%|█████▋    | 6151/10708 [1:27:15<37:25,  2.03it/s]
 57%|█████▋    | 6152/10708 [1:27:15<37:24,  2.03it/s]
 57%|█████▋    | 6153/10708 [1:27:16<37:22,  2.03it/s]
 57%|█████▋    | 6154/10708 [1:27:16<37:22,  2.03it/s]
 57%|█████▋    | 6155/10708 [1:27:17<37:21,  2.03it/s]
 57%|█████▋    | 6156/10708 [1:27:17<37:17,  2.03it/s]
 57%|█████▋    | 6157/10708 [1:27:18<37:15,  2.04it/s]
 58%|█████▊    | 6158/10708 [1:27:18<37:17,  2.03it/s]
 58%|█████▊    | 6159/10708 [1:27:19<37:15,  2.04it/s]
 58%|█████▊    | 6160/10708 [1:27:19<37:12,  2.04it/s]
 58%|█████▊    | 6161/10708 [1:27:20<37:15,  2.03it/s]
 58%|█████▊    | 6162/10708 [1:27:20<37:15,  2.03it/s]
 58%|█████▊    | 6163/10708 [1:27:21<37:13,  2.04it/s]
 58%|█████▊    | 6164/10708 [1:27:21<37:12,  2.04it/s]
 58%|█████▊    | 6165/10708 [1:27:22<37:12,  2.04it/s]
 58%|█████▊    | 6166/10708 [1:27:22<37:08,  2.04it/s]
 58%|█████▊    | 6167/10708 [1:27:23<37:11,  2.04it/s]
 58%|█████▊    | 6168/10708 [1:27:23<37:09,  2.04it/s]
 58%|█████▊    | 6169/10708 [1:27:24<37:06,  2.04it/s]
 58%|█████▊    | 6170/10708 [1:27:24<37:10,  2.03it/s]
 58%|█████▊    | 6171/10708 [1:27:25<37:07,  2.04it/s]
 58%|█████▊    | 6172/10708 [1:27:25<37:07,  2.04it/s]
 58%|█████▊    | 6173/10708 [1:27:26<37:06,  2.04it/s]
 58%|█████▊    | 6174/10708 [1:27:26<37:08,  2.03it/s]
 58%|█████▊    | 6175/10708 [1:27:27<37:06,  2.04it/s]
                                                      
{'loss': 3.629, 'grad_norm': 0.20039784908294678, 'learning_rate': 0.0004535316813850389, 'epoch': 0.58}

 58%|█████▊    | 6175/10708 [1:27:27<37:06,  2.04it/s]
 58%|█████▊    | 6176/10708 [1:27:27<37:11,  2.03it/s]
 58%|█████▊    | 6177/10708 [1:27:27<37:09,  2.03it/s]
 58%|█████▊    | 6178/10708 [1:27:28<37:05,  2.04it/s]
 58%|█████▊    | 6179/10708 [1:27:28<37:04,  2.04it/s]
 58%|█████▊    | 6180/10708 [1:27:29<37:04,  2.04it/s]
 58%|█████▊    | 6181/10708 [1:27:29<37:00,  2.04it/s]
 58%|█████▊    | 6182/10708 [1:27:30<37:05,  2.03it/s]
 58%|█████▊    | 6183/10708 [1:27:30<37:06,  2.03it/s]
 58%|█████▊    | 6184/10708 [1:27:31<37:07,  2.03it/s]
 58%|█████▊    | 6185/10708 [1:27:31<37:06,  2.03it/s]
 58%|█████▊    | 6186/10708 [1:27:32<37:05,  2.03it/s]
 58%|█████▊    | 6187/10708 [1:27:32<37:04,  2.03it/s]
 58%|█████▊    | 6188/10708 [1:27:33<37:03,  2.03it/s]
 58%|█████▊    | 6189/10708 [1:27:33<37:04,  2.03it/s]
 58%|█████▊    | 6190/10708 [1:27:34<37:06,  2.03it/s]
 58%|█████▊    | 6191/10708 [1:27:34<37:03,  2.03it/s]
 58%|█████▊    | 6192/10708 [1:27:35<37:00,  2.03it/s]
 58%|█████▊    | 6193/10708 [1:27:35<37:01,  2.03it/s]
 58%|█████▊    | 6194/10708 [1:27:36<37:00,  2.03it/s]
 58%|█████▊    | 6195/10708 [1:27:36<36:57,  2.04it/s]
 58%|█████▊    | 6196/10708 [1:27:37<37:00,  2.03it/s]
 58%|█████▊    | 6197/10708 [1:27:37<36:57,  2.03it/s]
 58%|█████▊    | 6198/10708 [1:27:38<36:55,  2.04it/s]
 58%|█████▊    | 6199/10708 [1:27:38<36:57,  2.03it/s]
 58%|█████▊    | 6200/10708 [1:27:39<36:58,  2.03it/s]
                                                      
{'loss': 3.6359, 'grad_norm': 0.19807069003582, 'learning_rate': 0.00044947599554966784, 'epoch': 0.58}

 58%|█████▊    | 6200/10708 [1:27:39<36:58,  2.03it/s]
 58%|█████▊    | 6201/10708 [1:27:39<36:59,  2.03it/s]
 58%|█████▊    | 6202/10708 [1:27:40<37:01,  2.03it/s]
 58%|█████▊    | 6203/10708 [1:27:40<36:59,  2.03it/s]
 58%|█████▊    | 6204/10708 [1:27:41<37:02,  2.03it/s]
 58%|█████▊    | 6205/10708 [1:27:41<37:00,  2.03it/s]
 58%|█████▊    | 6206/10708 [1:27:42<36:57,  2.03it/s]
 58%|█████▊    | 6207/10708 [1:27:42<36:55,  2.03it/s]
 58%|█████▊    | 6208/10708 [1:27:43<36:57,  2.03it/s]
 58%|█████▊    | 6209/10708 [1:27:43<37:01,  2.02it/s]
 58%|█████▊    | 6210/10708 [1:27:44<36:56,  2.03it/s]
 58%|█████▊    | 6211/10708 [1:27:44<36:53,  2.03it/s]
 58%|█████▊    | 6212/10708 [1:27:45<36:53,  2.03it/s]
 58%|█████▊    | 6213/10708 [1:27:45<36:52,  2.03it/s]
 58%|█████▊    | 6214/10708 [1:27:46<36:56,  2.03it/s]
 58%|█████▊    | 6215/10708 [1:27:46<36:51,  2.03it/s]
 58%|█████▊    | 6216/10708 [1:27:47<36:49,  2.03it/s]
 58%|█████▊    | 6217/10708 [1:27:47<36:47,  2.03it/s]
 58%|█████▊    | 6218/10708 [1:27:48<36:45,  2.04it/s]
 58%|█████▊    | 6219/10708 [1:27:48<36:48,  2.03it/s]
 58%|█████▊    | 6220/10708 [1:27:49<36:47,  2.03it/s]
 58%|█████▊    | 6221/10708 [1:27:49<36:50,  2.03it/s]
 58%|█████▊    | 6222/10708 [1:27:50<36:49,  2.03it/s]
 58%|█████▊    | 6223/10708 [1:27:50<36:49,  2.03it/s]
 58%|█████▊    | 6224/10708 [1:27:51<36:49,  2.03it/s]
 58%|█████▊    | 6225/10708 [1:27:51<36:48,  2.03it/s]
                                                      {'loss': 3.6328, 'grad_norm': 0.19344808161258698, 'learning_rate': 0.00044542366547825775, 'epoch': 0.58}

 58%|█████▊    | 6225/10708 [1:27:51<36:48,  2.03it/s]
 58%|█████▊    | 6226/10708 [1:27:52<36:49,  2.03it/s]
 58%|█████▊    | 6227/10708 [1:27:52<36:46,  2.03it/s]
 58%|█████▊    | 6228/10708 [1:27:53<36:44,  2.03it/s]
 58%|█████▊    | 6229/10708 [1:27:53<36:46,  2.03it/s]
 58%|█████▊    | 6230/10708 [1:27:54<36:43,  2.03it/s]
 58%|█████▊    | 6231/10708 [1:27:54<36:40,  2.03it/s]
 58%|█████▊    | 6232/10708 [1:27:55<36:42,  2.03it/s]
 58%|█████▊    | 6233/10708 [1:27:55<36:39,  2.03it/s]
 58%|█████▊    | 6234/10708 [1:27:56<36:40,  2.03it/s]
 58%|█████▊    | 6235/10708 [1:27:56<36:40,  2.03it/s]
 58%|█████▊    | 6236/10708 [1:27:57<36:39,  2.03it/s]
 58%|█████▊    | 6237/10708 [1:27:57<36:37,  2.03it/s]
 58%|█████▊    | 6238/10708 [1:27:58<36:37,  2.03it/s]
 58%|█████▊    | 6239/10708 [1:27:58<36:36,  2.03it/s]
 58%|█████▊    | 6240/10708 [1:27:58<36:36,  2.03it/s]
 58%|█████▊    | 6241/10708 [1:27:59<36:36,  2.03it/s]
 58%|█████▊    | 6242/10708 [1:27:59<36:35,  2.03it/s]
 58%|█████▊    | 6243/10708 [1:28:00<36:35,  2.03it/s]
 58%|█████▊    | 6244/10708 [1:28:00<36:34,  2.03it/s]
 58%|█████▊    | 6245/10708 [1:28:01<36:30,  2.04it/s]
 58%|█████▊    | 6246/10708 [1:28:01<36:33,  2.03it/s]
 58%|█████▊    | 6247/10708 [1:28:02<36:32,  2.03it/s]
 58%|█████▊    | 6248/10708 [1:28:02<36:32,  2.03it/s]
 58%|█████▊    | 6249/10708 [1:28:03<36:32,  2.03it/s]
 58%|█████▊    | 6250/10708 [1:28:03<36:31,  2.03it/s]{'loss': 3.62, 'grad_norm': 0.19992946088314056, 'learning_rate': 0.0004413749603233309, 'epoch': 0.58}

                                                      

 58%|█████▊    | 6250/10708 [1:28:03<36:31,  2.03it/s]
 58%|█████▊    | 6251/10708 [1:28:04<36:34,  2.03it/s]
 58%|█████▊    | 6252/10708 [1:28:04<36:32,  2.03it/s]
 58%|█████▊    | 6253/10708 [1:28:05<36:35,  2.03it/s]
 58%|█████▊    | 6254/10708 [1:28:05<36:33,  2.03it/s]
 58%|█████▊    | 6255/10708 [1:28:06<36:32,  2.03it/s]
 58%|█████▊    | 6256/10708 [1:28:06<36:29,  2.03it/s]
 58%|█████▊    | 6257/10708 [1:28:07<36:28,  2.03it/s]
 58%|█████▊    | 6258/10708 [1:28:07<36:29,  2.03it/s]
 58%|█████▊    | 6259/10708 [1:28:08<36:28,  2.03it/s]
 58%|█████▊    | 6260/10708 [1:28:08<36:28,  2.03it/s]
 58%|█████▊    | 6261/10708 [1:28:09<36:27,  2.03it/s]
 58%|█████▊    | 6262/10708 [1:28:09<36:25,  2.03it/s]
 58%|█████▊    | 6263/10708 [1:28:10<36:25,  2.03it/s]
 58%|█████▊    | 6264/10708 [1:28:10<36:23,  2.03it/s]
 59%|█████▊    | 6265/10708 [1:28:11<36:25,  2.03it/s]
 59%|█████▊    | 6266/10708 [1:28:11<36:23,  2.03it/s]
 59%|█████▊    | 6267/10708 [1:28:12<36:24,  2.03it/s]
 59%|█████▊    | 6268/10708 [1:28:12<36:24,  2.03it/s]
 59%|█████▊    | 6269/10708 [1:28:13<36:25,  2.03it/s]
 59%|█████▊    | 6270/10708 [1:28:13<36:30,  2.03it/s]
 59%|█████▊    | 6271/10708 [1:28:14<36:30,  2.03it/s]
 59%|█████▊    | 6272/10708 [1:28:14<36:28,  2.03it/s]
 59%|█████▊    | 6273/10708 [1:28:15<36:24,  2.03it/s]
 59%|█████▊    | 6274/10708 [1:28:15<36:21,  2.03it/s]
 59%|█████▊    | 6275/10708 [1:28:16<36:22,  2.03it/s]
                                                      
{'loss': 3.6212, 'grad_norm': 0.19084656238555908, 'learning_rate': 0.0004373301489966447, 'epoch': 0.59}

 59%|█████▊    | 6275/10708 [1:28:16<36:22,  2.03it/s]
 59%|█████▊    | 6276/10708 [1:28:16<36:24,  2.03it/s]
 59%|█████▊    | 6277/10708 [1:28:17<36:24,  2.03it/s]
 59%|█████▊    | 6278/10708 [1:28:17<36:25,  2.03it/s]
 59%|█████▊    | 6279/10708 [1:28:18<36:22,  2.03it/s]
 59%|█████▊    | 6280/10708 [1:28:18<36:21,  2.03it/s]
 59%|█████▊    | 6281/10708 [1:28:19<36:20,  2.03it/s]
 59%|█████▊    | 6282/10708 [1:28:19<36:19,  2.03it/s]
 59%|█████▊    | 6283/10708 [1:28:20<36:21,  2.03it/s]
 59%|█████▊    | 6284/10708 [1:28:20<36:19,  2.03it/s]
 59%|█████▊    | 6285/10708 [1:28:21<36:15,  2.03it/s]
 59%|█████▊    | 6286/10708 [1:28:21<36:12,  2.04it/s]
 59%|█████▊    | 6287/10708 [1:28:22<36:14,  2.03it/s]
 59%|█████▊    | 6288/10708 [1:28:22<36:11,  2.04it/s]
 59%|█████▊    | 6289/10708 [1:28:23<36:09,  2.04it/s]
 59%|█████▊    | 6290/10708 [1:28:23<36:12,  2.03it/s]
 59%|█████▉    | 6291/10708 [1:28:24<36:10,  2.03it/s]
 59%|█████▉    | 6292/10708 [1:28:24<36:08,  2.04it/s]
 59%|█████▉    | 6293/10708 [1:28:25<36:11,  2.03it/s]
 59%|█████▉    | 6294/10708 [1:28:25<36:08,  2.04it/s]
 59%|█████▉    | 6295/10708 [1:28:26<36:05,  2.04it/s]
 59%|█████▉    | 6296/10708 [1:28:26<36:07,  2.04it/s]
 59%|█████▉    | 6297/10708 [1:28:27<36:07,  2.04it/s]
 59%|█████▉    | 6298/10708 [1:28:27<36:07,  2.03it/s]
 59%|█████▉    | 6299/10708 [1:28:28<36:08,  2.03it/s]
 59%|█████▉    | 6300/10708 [1:28:28<36:06,  2.03it/s]
                                                      
{'loss': 3.6303, 'grad_norm': 0.20419462025165558, 'learning_rate': 0.00043328950015133225, 'epoch': 0.59}

 59%|█████▉    | 6300/10708 [1:28:28<36:06,  2.03it/s]
 59%|█████▉    | 6301/10708 [1:28:29<36:10,  2.03it/s]
 59%|█████▉    | 6302/10708 [1:28:29<36:07,  2.03it/s]
 59%|█████▉    | 6303/10708 [1:28:29<36:07,  2.03it/s]
 59%|█████▉    | 6304/10708 [1:28:30<36:08,  2.03it/s]
 59%|█████▉    | 6305/10708 [1:28:30<36:06,  2.03it/s]
 59%|█████▉    | 6306/10708 [1:28:31<36:06,  2.03it/s]
 59%|█████▉    | 6307/10708 [1:28:31<36:06,  2.03it/s]
 59%|█████▉    | 6308/10708 [1:28:32<36:03,  2.03it/s]
 59%|█████▉    | 6309/10708 [1:28:32<36:01,  2.04it/s]
 59%|█████▉    | 6310/10708 [1:28:33<36:02,  2.03it/s]
 59%|█████▉    | 6311/10708 [1:28:33<36:03,  2.03it/s]
 59%|█████▉    | 6312/10708 [1:28:34<36:03,  2.03it/s]
 59%|█████▉    | 6313/10708 [1:28:34<36:03,  2.03it/s]
 59%|█████▉    | 6314/10708 [1:28:35<36:04,  2.03it/s]
 59%|█████▉    | 6315/10708 [1:28:35<36:04,  2.03it/s]
 59%|█████▉    | 6316/10708 [1:28:36<36:02,  2.03it/s]
 59%|█████▉    | 6317/10708 [1:28:36<36:04,  2.03it/s]
 59%|█████▉    | 6318/10708 [1:28:37<36:03,  2.03it/s]
 59%|█████▉    | 6319/10708 [1:28:37<36:03,  2.03it/s]
 59%|█████▉    | 6320/10708 [1:28:38<36:06,  2.03it/s]
 59%|█████▉    | 6321/10708 [1:28:38<36:00,  2.03it/s]
 59%|█████▉    | 6322/10708 [1:28:39<36:00,  2.03it/s]
 59%|█████▉    | 6323/10708 [1:28:39<35:58,  2.03it/s]
 59%|█████▉    | 6324/10708 [1:28:40<35:58,  2.03it/s]
 59%|█████▉    | 6325/10708 [1:28:40<35:58,  2.03it/s]
                                                      {'loss': 3.6212, 'grad_norm': 0.21039293706417084, 'learning_rate': 0.0004292532821640575, 'epoch': 0.59}

 59%|█████▉    | 6325/10708 [1:28:40<35:58,  2.03it/s]
 59%|█████▉    | 6326/10708 [1:28:41<35:58,  2.03it/s]
 59%|█████▉    | 6327/10708 [1:28:41<35:55,  2.03it/s]
 59%|█████▉    | 6328/10708 [1:28:42<35:56,  2.03it/s]
 59%|█████▉    | 6329/10708 [1:28:42<35:52,  2.03it/s]
 59%|█████▉    | 6330/10708 [1:28:43<35:52,  2.03it/s]
 59%|█████▉    | 6331/10708 [1:28:43<35:57,  2.03it/s]
 59%|█████▉    | 6332/10708 [1:28:44<35:55,  2.03it/s]
 59%|█████▉    | 6333/10708 [1:28:44<35:53,  2.03it/s]
 59%|█████▉    | 6334/10708 [1:28:45<35:54,  2.03it/s]
 59%|█████▉    | 6335/10708 [1:28:45<35:52,  2.03it/s]
 59%|█████▉    | 6336/10708 [1:28:46<35:53,  2.03it/s]
 59%|█████▉    | 6337/10708 [1:28:46<35:50,  2.03it/s]
 59%|█████▉    | 6338/10708 [1:28:47<35:51,  2.03it/s]
 59%|█████▉    | 6339/10708 [1:28:47<35:51,  2.03it/s]
 59%|█████▉    | 6340/10708 [1:28:48<35:49,  2.03it/s]
 59%|█████▉    | 6341/10708 [1:28:48<35:53,  2.03it/s]
 59%|█████▉    | 6342/10708 [1:28:49<35:50,  2.03it/s]
 59%|█████▉    | 6343/10708 [1:28:49<35:48,  2.03it/s]
 59%|█████▉    | 6344/10708 [1:28:50<35:50,  2.03it/s]
 59%|█████▉    | 6345/10708 [1:28:50<35:50,  2.03it/s]
 59%|█████▉    | 6346/10708 [1:28:51<35:48,  2.03it/s]
 59%|█████▉    | 6347/10708 [1:28:51<35:47,  2.03it/s]
 59%|█████▉    | 6348/10708 [1:28:52<35:43,  2.03it/s]
 59%|█████▉    | 6349/10708 [1:28:52<35:45,  2.03it/s]
 59%|█████▉    | 6350/10708 [1:28:53<35:43,  2.03it/s]{'loss': 3.6217, 'grad_norm': 0.1961202770471573, 'learning_rate': 0.00042522176311719095, 'epoch': 0.59}
                                                      

 59%|█████▉    | 6350/10708 [1:28:53<35:43,  2.03it/s]
 59%|█████▉    | 6351/10708 [1:28:53<35:42,  2.03it/s]
 59%|█████▉    | 6352/10708 [1:28:54<35:49,  2.03it/s]
 59%|█████▉    | 6353/10708 [1:28:54<35:44,  2.03it/s]
 59%|█████▉    | 6354/10708 [1:28:55<35:45,  2.03it/s]
 59%|█████▉    | 6355/10708 [1:28:55<35:43,  2.03it/s]
 59%|█████▉    | 6356/10708 [1:28:56<35:42,  2.03it/s]
 59%|█████▉    | 6357/10708 [1:28:56<35:42,  2.03it/s]
 59%|█████▉    | 6358/10708 [1:28:57<35:40,  2.03it/s]
 59%|█████▉    | 6359/10708 [1:28:57<35:37,  2.03it/s]
 59%|█████▉    | 6360/10708 [1:28:58<35:39,  2.03it/s]
 59%|█████▉    | 6361/10708 [1:28:58<35:38,  2.03it/s]
 59%|█████▉    | 6362/10708 [1:28:59<35:37,  2.03it/s]
 59%|█████▉    | 6363/10708 [1:28:59<35:37,  2.03it/s]
 59%|█████▉    | 6364/10708 [1:29:00<35:38,  2.03it/s]
 59%|█████▉    | 6365/10708 [1:29:00<35:39,  2.03it/s]
 59%|█████▉    | 6366/10708 [1:29:01<35:38,  2.03it/s]
 59%|█████▉    | 6367/10708 [1:29:01<35:36,  2.03it/s]
 59%|█████▉    | 6368/10708 [1:29:01<35:36,  2.03it/s]
 59%|█████▉    | 6369/10708 [1:29:02<35:36,  2.03it/s]
 59%|█████▉    | 6370/10708 [1:29:02<35:34,  2.03it/s]
 59%|█████▉    | 6371/10708 [1:29:03<35:38,  2.03it/s]
 60%|█████▉    | 6372/10708 [1:29:03<35:35,  2.03it/s]
 60%|█████▉    | 6373/10708 [1:29:04<35:34,  2.03it/s]
 60%|█████▉    | 6374/10708 [1:29:04<35:34,  2.03it/s]
 60%|█████▉    | 6375/10708 [1:29:05<35:30,  2.03it/s]
                                                      
{'loss': 3.6182, 'grad_norm': 0.20512153208255768, 'learning_rate': 0.0004211952107810024, 'epoch': 0.6}

 60%|█████▉    | 6375/10708 [1:29:05<35:30,  2.03it/s]
 60%|█████▉    | 6376/10708 [1:29:05<35:35,  2.03it/s]
 60%|█████▉    | 6377/10708 [1:29:06<35:32,  2.03it/s]
 60%|█████▉    | 6378/10708 [1:29:06<35:30,  2.03it/s]
 60%|█████▉    | 6379/10708 [1:29:07<35:32,  2.03it/s]
 60%|█████▉    | 6380/10708 [1:29:07<35:29,  2.03it/s]
 60%|█████▉    | 6381/10708 [1:29:08<35:30,  2.03it/s]
 60%|█████▉    | 6382/10708 [1:29:08<35:29,  2.03it/s]
 60%|█████▉    | 6383/10708 [1:29:09<35:27,  2.03it/s]
 60%|█████▉    | 6384/10708 [1:29:09<35:26,  2.03it/s]
 60%|█████▉    | 6385/10708 [1:29:10<35:25,  2.03it/s]
 60%|█████▉    | 6386/10708 [1:29:10<35:24,  2.03it/s]
 60%|█████▉    | 6387/10708 [1:29:11<35:23,  2.03it/s]
 60%|█████▉    | 6388/10708 [1:29:11<35:25,  2.03it/s]
 60%|█████▉    | 6389/10708 [1:29:12<35:26,  2.03it/s]
 60%|█████▉    | 6390/10708 [1:29:12<35:26,  2.03it/s]
 60%|█████▉    | 6391/10708 [1:29:13<35:30,  2.03it/s]
 60%|█████▉    | 6392/10708 [1:29:13<35:32,  2.02it/s]
 60%|█████▉    | 6393/10708 [1:29:14<35:27,  2.03it/s]
 60%|█████▉    | 6394/10708 [1:29:14<35:28,  2.03it/s]
 60%|█████▉    | 6395/10708 [1:29:15<35:24,  2.03it/s]
 60%|█████▉    | 6396/10708 [1:29:15<35:21,  2.03it/s]
 60%|█████▉    | 6397/10708 [1:29:16<35:20,  2.03it/s]
 60%|█████▉    | 6398/10708 [1:29:16<35:19,  2.03it/s]
 60%|█████▉    | 6399/10708 [1:29:17<35:19,  2.03it/s]
 60%|█████▉    | 6400/10708 [1:29:17<35:18,  2.03it/s]
                                                      
{'loss': 3.6165, 'grad_norm': 0.2101997584104538, 'learning_rate': 0.00041717389259587743, 'epoch': 0.6}

 60%|█████▉    | 6400/10708 [1:29:17<35:18,  2.03it/s]
 60%|█████▉    | 6401/10708 [1:29:18<35:20,  2.03it/s]
 60%|█████▉    | 6402/10708 [1:29:18<35:19,  2.03it/s]
 60%|█████▉    | 6403/10708 [1:29:19<35:22,  2.03it/s]
 60%|█████▉    | 6404/10708 [1:29:19<35:22,  2.03it/s]
 60%|█████▉    | 6405/10708 [1:29:20<35:20,  2.03it/s]
 60%|█████▉    | 6406/10708 [1:29:20<35:19,  2.03it/s]
 60%|█████▉    | 6407/10708 [1:29:21<35:19,  2.03it/s]
 60%|█████▉    | 6408/10708 [1:29:21<35:17,  2.03it/s]
 60%|█████▉    | 6409/10708 [1:29:22<35:17,  2.03it/s]
 60%|█████▉    | 6410/10708 [1:29:22<35:16,  2.03it/s]
 60%|█████▉    | 6411/10708 [1:29:23<35:13,  2.03it/s]
 60%|█████▉    | 6412/10708 [1:29:23<35:12,  2.03it/s]
 60%|█████▉    | 6413/10708 [1:29:24<35:15,  2.03it/s]
 60%|█████▉    | 6414/10708 [1:29:24<35:14,  2.03it/s]
 60%|█████▉    | 6415/10708 [1:29:25<35:12,  2.03it/s]
 60%|█████▉    | 6416/10708 [1:29:25<35:13,  2.03it/s]
 60%|█████▉    | 6417/10708 [1:29:26<35:12,  2.03it/s]
 60%|█████▉    | 6418/10708 [1:29:26<35:11,  2.03it/s]
 60%|█████▉    | 6419/10708 [1:29:27<35:10,  2.03it/s]
 60%|█████▉    | 6420/10708 [1:29:27<35:08,  2.03it/s]
 60%|█████▉    | 6421/10708 [1:29:28<35:11,  2.03it/s]
 60%|█████▉    | 6422/10708 [1:29:28<35:08,  2.03it/s]
 60%|█████▉    | 6423/10708 [1:29:29<35:09,  2.03it/s]
 60%|█████▉    | 6424/10708 [1:29:29<35:10,  2.03it/s]
 60%|██████    | 6425/10708 [1:29:30<35:07,  2.03it/s]
                                                      
{'loss': 3.6245, 'grad_norm': 0.1960935890674591, 'learning_rate': 0.0004131580756545531, 'epoch': 0.6}

 60%|██████    | 6425/10708 [1:29:30<35:07,  2.03it/s]
 60%|██████    | 6426/10708 [1:29:30<35:10,  2.03it/s]
 60%|██████    | 6427/10708 [1:29:31<35:08,  2.03it/s]
 60%|██████    | 6428/10708 [1:29:31<35:08,  2.03it/s]
 60%|██████    | 6429/10708 [1:29:32<35:08,  2.03it/s]
 60%|██████    | 6430/10708 [1:29:32<35:07,  2.03it/s]
 60%|██████    | 6431/10708 [1:29:33<35:07,  2.03it/s]
 60%|██████    | 6432/10708 [1:29:33<35:07,  2.03it/s]
 60%|██████    | 6433/10708 [1:29:33<35:06,  2.03it/s]
 60%|██████    | 6434/10708 [1:29:34<35:08,  2.03it/s]
 60%|██████    | 6435/10708 [1:29:34<35:04,  2.03it/s]
 60%|██████    | 6436/10708 [1:29:35<35:03,  2.03it/s]
 60%|██████    | 6437/10708 [1:29:35<35:02,  2.03it/s]
 60%|██████    | 6438/10708 [1:29:36<35:01,  2.03it/s]
 60%|██████    | 6439/10708 [1:29:36<35:01,  2.03it/s]
 60%|██████    | 6440/10708 [1:29:37<35:02,  2.03it/s]
 60%|██████    | 6441/10708 [1:29:37<35:01,  2.03it/s]
 60%|██████    | 6442/10708 [1:29:38<35:03,  2.03it/s]
 60%|██████    | 6443/10708 [1:29:38<35:00,  2.03it/s]
 60%|██████    | 6444/10708 [1:29:39<34:59,  2.03it/s]
 60%|██████    | 6445/10708 [1:29:39<34:57,  2.03it/s]
 60%|██████    | 6446/10708 [1:29:40<34:55,  2.03it/s]
 60%|██████    | 6447/10708 [1:29:40<34:53,  2.03it/s]
 60%|██████    | 6448/10708 [1:29:41<34:56,  2.03it/s]
 60%|██████    | 6449/10708 [1:29:41<34:52,  2.04it/s]
 60%|██████    | 6450/10708 [1:29:42<34:50,  2.04it/s]{'loss': 3.6196, 'grad_norm': 0.21287453174591064, 'learning_rate': 0.00040914802668437863, 'epoch': 0.6}

                                                      

 60%|██████    | 6450/10708 [1:29:42<34:50,  2.04it/s]
 60%|██████    | 6451/10708 [1:29:42<34:55,  2.03it/s]
 60%|██████    | 6452/10708 [1:29:43<34:55,  2.03it/s]
 60%|██████    | 6453/10708 [1:29:43<34:59,  2.03it/s]
 60%|██████    | 6454/10708 [1:29:44<34:57,  2.03it/s]
 60%|██████    | 6455/10708 [1:29:44<34:56,  2.03it/s]
 60%|██████    | 6456/10708 [1:29:45<34:54,  2.03it/s]
 60%|██████    | 6457/10708 [1:29:45<34:49,  2.03it/s]
 60%|██████    | 6458/10708 [1:29:46<34:51,  2.03it/s]
 60%|██████    | 6459/10708 [1:29:46<34:50,  2.03it/s]
 60%|██████    | 6460/10708 [1:29:47<34:50,  2.03it/s]
 60%|██████    | 6461/10708 [1:29:47<34:50,  2.03it/s]
 60%|██████    | 6462/10708 [1:29:48<35:03,  2.02it/s]
 60%|██████    | 6463/10708 [1:29:48<34:58,  2.02it/s]
 60%|██████    | 6464/10708 [1:29:49<34:55,  2.03it/s]
 60%|██████    | 6465/10708 [1:29:49<34:54,  2.03it/s]
 60%|██████    | 6466/10708 [1:29:50<34:49,  2.03it/s]
 60%|██████    | 6467/10708 [1:29:50<34:49,  2.03it/s]
 60%|██████    | 6468/10708 [1:29:51<34:50,  2.03it/s]
 60%|██████    | 6469/10708 [1:29:51<34:45,  2.03it/s]
 60%|██████    | 6470/10708 [1:29:52<34:45,  2.03it/s]
 60%|██████    | 6471/10708 [1:29:52<34:46,  2.03it/s]
 60%|██████    | 6472/10708 [1:29:53<34:42,  2.03it/s]
 60%|██████    | 6473/10708 [1:29:53<34:46,  2.03it/s]
 60%|██████    | 6474/10708 [1:29:54<34:44,  2.03it/s]
 60%|██████    | 6475/10708 [1:29:54<34:45,  2.03it/s]
                                                      
{'loss': 3.6113, 'grad_norm': 0.19377054274082184, 'learning_rate': 0.00040514401202959853, 'epoch': 0.6}

 60%|██████    | 6475/10708 [1:29:54<34:45,  2.03it/s]
 60%|██████    | 6476/10708 [1:29:55<34:45,  2.03it/s]
 60%|██████    | 6477/10708 [1:29:55<34:44,  2.03it/s]
 60%|██████    | 6478/10708 [1:29:56<34:45,  2.03it/s]
 61%|██████    | 6479/10708 [1:29:56<34:41,  2.03it/s]
 61%|██████    | 6480/10708 [1:29:57<34:40,  2.03it/s]
 61%|██████    | 6481/10708 [1:29:57<34:40,  2.03it/s]
 61%|██████    | 6482/10708 [1:29:58<34:38,  2.03it/s]
 61%|██████    | 6483/10708 [1:29:58<34:39,  2.03it/s]
 61%|██████    | 6484/10708 [1:29:59<34:38,  2.03it/s]
 61%|██████    | 6485/10708 [1:29:59<34:36,  2.03it/s]
 61%|██████    | 6486/10708 [1:30:00<34:34,  2.04it/s]
 61%|██████    | 6487/10708 [1:30:00<34:34,  2.03it/s]
 61%|██████    | 6488/10708 [1:30:01<34:33,  2.04it/s]
 61%|██████    | 6489/10708 [1:30:01<34:31,  2.04it/s]
 61%|██████    | 6490/10708 [1:30:02<34:33,  2.03it/s]
 61%|██████    | 6491/10708 [1:30:02<34:32,  2.03it/s]
 61%|██████    | 6492/10708 [1:30:03<34:32,  2.03it/s]
 61%|██████    | 6493/10708 [1:30:03<34:33,  2.03it/s]
 61%|██████    | 6494/10708 [1:30:04<34:31,  2.03it/s]
 61%|██████    | 6495/10708 [1:30:04<34:31,  2.03it/s]
 61%|██████    | 6496/10708 [1:30:05<34:31,  2.03it/s]
 61%|██████    | 6497/10708 [1:30:05<34:28,  2.04it/s]
 61%|██████    | 6498/10708 [1:30:05<34:27,  2.04it/s]
 61%|██████    | 6499/10708 [1:30:06<34:29,  2.03it/s]
 61%|██████    | 6500/10708 [1:30:06<34:28,  2.03it/s]{'loss': 3.6213, 'grad_norm': 0.19991187751293182, 'learning_rate': 0.0004011462976336637, 'epoch': 0.61}
                                                      

 61%|██████    | 6500/10708 [1:30:06<34:28,  2.03it/s]
 61%|██████    | 6501/10708 [1:30:07<34:33,  2.03it/s]
 61%|██████    | 6502/10708 [1:30:07<34:29,  2.03it/s]
 61%|██████    | 6503/10708 [1:30:08<34:28,  2.03it/s]
 61%|██████    | 6504/10708 [1:30:08<34:29,  2.03it/s]
 61%|██████    | 6505/10708 [1:30:09<34:29,  2.03it/s]
 61%|██████    | 6506/10708 [1:30:09<34:27,  2.03it/s]
 61%|██████    | 6507/10708 [1:30:10<34:26,  2.03it/s]
 61%|██████    | 6508/10708 [1:30:10<34:26,  2.03it/s]
 61%|██████    | 6509/10708 [1:30:11<34:24,  2.03it/s]
 61%|██████    | 6510/10708 [1:30:11<34:27,  2.03it/s]
 61%|██████    | 6511/10708 [1:30:12<34:26,  2.03it/s]
 61%|██████    | 6512/10708 [1:30:12<34:27,  2.03it/s]
 61%|██████    | 6513/10708 [1:30:13<34:31,  2.03it/s]
 61%|██████    | 6514/10708 [1:30:13<34:34,  2.02it/s]
 61%|██████    | 6515/10708 [1:30:14<34:32,  2.02it/s]
 61%|██████    | 6516/10708 [1:30:14<34:30,  2.02it/s]
 61%|██████    | 6517/10708 [1:30:15<34:25,  2.03it/s]
 61%|██████    | 6518/10708 [1:30:15<34:23,  2.03it/s]
 61%|██████    | 6519/10708 [1:30:16<34:20,  2.03it/s]
 61%|██████    | 6520/10708 [1:30:16<34:20,  2.03it/s]
 61%|██████    | 6521/10708 [1:30:17<34:20,  2.03it/s]
 61%|██████    | 6522/10708 [1:30:17<34:20,  2.03it/s]
 61%|██████    | 6523/10708 [1:30:18<34:20,  2.03it/s]
 61%|██████    | 6524/10708 [1:30:18<34:20,  2.03it/s]
 61%|██████    | 6525/10708 [1:30:19<34:21,  2.03it/s]
                                                      
{'loss': 3.621, 'grad_norm': 0.2028183490037918, 'learning_rate': 0.00039715514902156636, 'epoch': 0.61}

 61%|██████    | 6525/10708 [1:30:19<34:21,  2.03it/s]
 61%|██████    | 6526/10708 [1:30:19<34:21,  2.03it/s]
 61%|██████    | 6527/10708 [1:30:20<34:21,  2.03it/s]
 61%|██████    | 6528/10708 [1:30:20<34:19,  2.03it/s]
 61%|██████    | 6529/10708 [1:30:21<34:17,  2.03it/s]
 61%|██████    | 6530/10708 [1:30:21<34:19,  2.03it/s]
 61%|██████    | 6531/10708 [1:30:22<34:18,  2.03it/s]
 61%|██████    | 6532/10708 [1:30:22<34:19,  2.03it/s]
 61%|██████    | 6533/10708 [1:30:23<34:17,  2.03it/s]
 61%|██████    | 6534/10708 [1:30:23<34:17,  2.03it/s]
 61%|██████    | 6535/10708 [1:30:24<34:15,  2.03it/s]
 61%|██████    | 6536/10708 [1:30:24<34:13,  2.03it/s]
 61%|██████    | 6537/10708 [1:30:25<34:14,  2.03it/s]
 61%|██████    | 6538/10708 [1:30:25<34:14,  2.03it/s]
 61%|██████    | 6539/10708 [1:30:26<34:09,  2.03it/s]
 61%|██████    | 6540/10708 [1:30:26<34:10,  2.03it/s]
 61%|██████    | 6541/10708 [1:30:27<34:13,  2.03it/s]
 61%|██████    | 6542/10708 [1:30:27<34:09,  2.03it/s]
 61%|██████    | 6543/10708 [1:30:28<34:10,  2.03it/s]
 61%|██████    | 6544/10708 [1:30:28<34:09,  2.03it/s]
 61%|██████    | 6545/10708 [1:30:29<34:09,  2.03it/s]
 61%|██████    | 6546/10708 [1:30:29<34:08,  2.03it/s]
 61%|██████    | 6547/10708 [1:30:30<34:07,  2.03it/s]
 61%|██████    | 6548/10708 [1:30:30<34:07,  2.03it/s]
 61%|██████    | 6549/10708 [1:30:31<34:08,  2.03it/s]
 61%|██████    | 6550/10708 [1:30:31<34:09,  2.03it/s]{'loss': 3.6225, 'grad_norm': 0.21238887310028076, 'learning_rate': 0.00039317083128220494, 'epoch': 0.61}

                                                      

 61%|██████    | 6550/10708 [1:30:31<34:09,  2.03it/s]
 61%|██████    | 6551/10708 [1:30:32<34:12,  2.03it/s]
 61%|██████    | 6552/10708 [1:30:32<34:11,  2.03it/s]
 61%|██████    | 6553/10708 [1:30:33<34:09,  2.03it/s]
 61%|██████    | 6554/10708 [1:30:33<34:07,  2.03it/s]
 61%|██████    | 6555/10708 [1:30:34<34:07,  2.03it/s]
 61%|██████    | 6556/10708 [1:30:34<34:05,  2.03it/s]
 61%|██████    | 6557/10708 [1:30:35<34:02,  2.03it/s]
 61%|██████    | 6558/10708 [1:30:35<34:05,  2.03it/s]
 61%|██████▏   | 6559/10708 [1:30:36<34:03,  2.03it/s]
 61%|██████▏   | 6560/10708 [1:30:36<34:02,  2.03it/s]
 61%|██████▏   | 6561/10708 [1:30:37<34:00,  2.03it/s]
 61%|██████▏   | 6562/10708 [1:30:37<34:01,  2.03it/s]
 61%|██████▏   | 6563/10708 [1:30:38<34:00,  2.03it/s]
 61%|██████▏   | 6564/10708 [1:30:38<33:59,  2.03it/s]
 61%|██████▏   | 6565/10708 [1:30:38<33:59,  2.03it/s]
 61%|██████▏   | 6566/10708 [1:30:39<33:59,  2.03it/s]
 61%|██████▏   | 6567/10708 [1:30:39<33:56,  2.03it/s]
 61%|██████▏   | 6568/10708 [1:30:40<33:59,  2.03it/s]
 61%|██████▏   | 6569/10708 [1:30:40<33:55,  2.03it/s]
 61%|██████▏   | 6570/10708 [1:30:41<33:56,  2.03it/s]
 61%|██████▏   | 6571/10708 [1:30:41<33:56,  2.03it/s]
 61%|██████▏   | 6572/10708 [1:30:42<33:54,  2.03it/s]
 61%|██████▏   | 6573/10708 [1:30:42<33:52,  2.03it/s]
 61%|██████▏   | 6574/10708 [1:30:43<33:58,  2.03it/s]
 61%|██████▏   | 6575/10708 [1:30:43<34:06,  2.02it/s]
                                                      
{'loss': 3.615, 'grad_norm': 0.2090175747871399, 'learning_rate': 0.0003891936090507763, 'epoch': 0.61}

 61%|██████▏   | 6575/10708 [1:30:43<34:06,  2.02it/s]
 61%|██████▏   | 6576/10708 [1:30:44<34:01,  2.02it/s]
 61%|██████▏   | 6577/10708 [1:30:44<33:57,  2.03it/s]
 61%|██████▏   | 6578/10708 [1:30:45<33:55,  2.03it/s]
 61%|██████▏   | 6579/10708 [1:30:45<33:51,  2.03it/s]
 61%|██████▏   | 6580/10708 [1:30:46<33:52,  2.03it/s]
 61%|██████▏   | 6581/10708 [1:30:46<33:50,  2.03it/s]
 61%|██████▏   | 6582/10708 [1:30:47<33:47,  2.03it/s]
 61%|██████▏   | 6583/10708 [1:30:47<33:49,  2.03it/s]
 61%|██████▏   | 6584/10708 [1:30:48<33:49,  2.03it/s]
 61%|██████▏   | 6585/10708 [1:30:48<33:48,  2.03it/s]
 62%|██████▏   | 6586/10708 [1:30:49<33:48,  2.03it/s]
 62%|██████▏   | 6587/10708 [1:30:49<33:45,  2.04it/s]
 62%|██████▏   | 6588/10708 [1:30:50<33:47,  2.03it/s]
 62%|██████▏   | 6589/10708 [1:30:50<33:47,  2.03it/s]
 62%|██████▏   | 6590/10708 [1:30:51<33:45,  2.03it/s]
 62%|██████▏   | 6591/10708 [1:30:51<33:44,  2.03it/s]
 62%|██████▏   | 6592/10708 [1:30:52<33:46,  2.03it/s]
 62%|██████▏   | 6593/10708 [1:30:52<33:44,  2.03it/s]
 62%|██████▏   | 6594/10708 [1:30:53<33:43,  2.03it/s]
 62%|██████▏   | 6595/10708 [1:30:53<33:42,  2.03it/s]
 62%|██████▏   | 6596/10708 [1:30:54<33:42,  2.03it/s]
 62%|██████▏   | 6597/10708 [1:30:54<33:42,  2.03it/s]
 62%|██████▏   | 6598/10708 [1:30:55<33:40,  2.03it/s]
 62%|██████▏   | 6599/10708 [1:30:55<33:39,  2.03it/s]
 62%|██████▏   | 6600/10708 [1:30:56<33:40,  2.03it/s]
                                                      
{'loss': 3.6132, 'grad_norm': 0.20929305255413055, 'learning_rate': 0.0003852237464912001, 'epoch': 0.62}

 62%|██████▏   | 6600/10708 [1:30:56<33:40,  2.03it/s]
 62%|██████▏   | 6601/10708 [1:30:56<33:41,  2.03it/s]
 62%|██████▏   | 6602/10708 [1:30:57<33:40,  2.03it/s]
 62%|██████▏   | 6603/10708 [1:30:57<33:39,  2.03it/s]
 62%|██████▏   | 6604/10708 [1:30:58<33:38,  2.03it/s]
 62%|██████▏   | 6605/10708 [1:30:58<33:37,  2.03it/s]
 62%|██████▏   | 6606/10708 [1:30:59<33:36,  2.03it/s]
 62%|██████▏   | 6607/10708 [1:30:59<33:35,  2.04it/s]
 62%|██████▏   | 6608/10708 [1:31:00<33:35,  2.03it/s]
 62%|██████▏   | 6609/10708 [1:31:00<33:35,  2.03it/s]
 62%|██████▏   | 6610/10708 [1:31:01<33:35,  2.03it/s]
 62%|██████▏   | 6611/10708 [1:31:01<33:35,  2.03it/s]
 62%|██████▏   | 6612/10708 [1:31:02<33:35,  2.03it/s]
 62%|██████▏   | 6613/10708 [1:31:02<33:35,  2.03it/s]
 62%|██████▏   | 6614/10708 [1:31:03<33:36,  2.03it/s]
 62%|██████▏   | 6615/10708 [1:31:03<33:32,  2.03it/s]
 62%|██████▏   | 6616/10708 [1:31:04<33:34,  2.03it/s]
 62%|██████▏   | 6617/10708 [1:31:04<33:33,  2.03it/s]
 62%|██████▏   | 6618/10708 [1:31:05<33:33,  2.03it/s]
 62%|██████▏   | 6619/10708 [1:31:05<33:35,  2.03it/s]
 62%|██████▏   | 6620/10708 [1:31:06<33:31,  2.03it/s]
 62%|██████▏   | 6621/10708 [1:31:06<33:29,  2.03it/s]
 62%|██████▏   | 6622/10708 [1:31:07<33:30,  2.03it/s]
 62%|██████▏   | 6623/10708 [1:31:07<33:31,  2.03it/s]
 62%|██████▏   | 6624/10708 [1:31:08<33:32,  2.03it/s]
 62%|██████▏   | 6625/10708 [1:31:08<33:30,  2.03it/s]
                                                      
{'loss': 3.6095, 'grad_norm': 0.22022207081317902, 'learning_rate': 0.00038126150727857156, 'epoch': 0.62}

 62%|██████▏   | 6625/10708 [1:31:08<33:30,  2.03it/s]
 62%|██████▏   | 6626/10708 [1:31:09<33:33,  2.03it/s]
 62%|██████▏   | 6627/10708 [1:31:09<33:31,  2.03it/s]
 62%|██████▏   | 6628/10708 [1:31:10<33:28,  2.03it/s]
 62%|██████▏   | 6629/10708 [1:31:10<33:25,  2.03it/s]
 62%|██████▏   | 6630/10708 [1:31:10<33:24,  2.03it/s]
 62%|██████▏   | 6631/10708 [1:31:11<33:26,  2.03it/s]
 62%|██████▏   | 6632/10708 [1:31:11<33:23,  2.03it/s]
 62%|██████▏   | 6633/10708 [1:31:12<33:28,  2.03it/s]
 62%|██████▏   | 6634/10708 [1:31:12<33:23,  2.03it/s]
 62%|██████▏   | 6635/10708 [1:31:13<33:28,  2.03it/s]
 62%|██████▏   | 6636/10708 [1:31:13<33:29,  2.03it/s]
 62%|██████▏   | 6637/10708 [1:31:14<33:25,  2.03it/s]
 62%|██████▏   | 6638/10708 [1:31:14<33:23,  2.03it/s]
 62%|██████▏   | 6639/10708 [1:31:15<33:20,  2.03it/s]
 62%|██████▏   | 6640/10708 [1:31:15<33:19,  2.03it/s]
 62%|██████▏   | 6641/10708 [1:31:16<33:20,  2.03it/s]
 62%|██████▏   | 6642/10708 [1:31:16<33:20,  2.03it/s]
 62%|██████▏   | 6643/10708 [1:31:17<33:21,  2.03it/s]
 62%|██████▏   | 6644/10708 [1:31:17<33:24,  2.03it/s]
 62%|██████▏   | 6645/10708 [1:31:18<33:23,  2.03it/s]
 62%|██████▏   | 6646/10708 [1:31:18<33:22,  2.03it/s]
 62%|██████▏   | 6647/10708 [1:31:19<33:20,  2.03it/s]
 62%|██████▏   | 6648/10708 [1:31:19<33:18,  2.03it/s]
 62%|██████▏   | 6649/10708 [1:31:20<33:16,  2.03it/s]
 62%|██████▏   | 6650/10708 [1:31:20<33:15,  2.03it/s]{'loss': 3.6072, 'grad_norm': 0.20062419772148132, 'learning_rate': 0.00037730715458164995, 'epoch': 0.62}

                                                      

 62%|██████▏   | 6650/10708 [1:31:20<33:15,  2.03it/s]
 62%|██████▏   | 6651/10708 [1:31:21<33:15,  2.03it/s]
 62%|██████▏   | 6652/10708 [1:31:21<33:14,  2.03it/s]
 62%|██████▏   | 6653/10708 [1:31:22<33:15,  2.03it/s]
 62%|██████▏   | 6654/10708 [1:31:22<33:12,  2.03it/s]
 62%|██████▏   | 6655/10708 [1:31:23<33:13,  2.03it/s]
 62%|██████▏   | 6656/10708 [1:31:23<33:13,  2.03it/s]
 62%|██████▏   | 6657/10708 [1:31:24<33:12,  2.03it/s]
 62%|██████▏   | 6658/10708 [1:31:24<33:12,  2.03it/s]
 62%|██████▏   | 6659/10708 [1:31:25<33:11,  2.03it/s]
 62%|██████▏   | 6660/10708 [1:31:25<33:13,  2.03it/s]
 62%|██████▏   | 6661/10708 [1:31:26<33:13,  2.03it/s]
 62%|██████▏   | 6662/10708 [1:31:26<33:12,  2.03it/s]
 62%|██████▏   | 6663/10708 [1:31:27<33:14,  2.03it/s]
 62%|██████▏   | 6664/10708 [1:31:27<33:11,  2.03it/s]
 62%|██████▏   | 6665/10708 [1:31:28<33:10,  2.03it/s]
 62%|██████▏   | 6666/10708 [1:31:28<33:11,  2.03it/s]
 62%|██████▏   | 6667/10708 [1:31:29<33:07,  2.03it/s]
 62%|██████▏   | 6668/10708 [1:31:29<33:10,  2.03it/s]
 62%|██████▏   | 6669/10708 [1:31:30<33:08,  2.03it/s]
 62%|██████▏   | 6670/10708 [1:31:30<33:05,  2.03it/s]
 62%|██████▏   | 6671/10708 [1:31:31<33:06,  2.03it/s]
 62%|██████▏   | 6672/10708 [1:31:31<33:04,  2.03it/s]
 62%|██████▏   | 6673/10708 [1:31:32<33:03,  2.03it/s]
 62%|██████▏   | 6674/10708 [1:31:32<33:04,  2.03it/s]
 62%|██████▏   | 6675/10708 [1:31:33<33:04,  2.03it/s]
                                                      
{'loss': 3.6114, 'grad_norm': 0.20616786181926727, 'learning_rate': 0.0003733609510453785, 'epoch': 0.62}

 62%|██████▏   | 6675/10708 [1:31:33<33:04,  2.03it/s]
 62%|██████▏   | 6676/10708 [1:31:33<33:08,  2.03it/s]
 62%|██████▏   | 6677/10708 [1:31:34<33:07,  2.03it/s]
 62%|██████▏   | 6678/10708 [1:31:34<33:03,  2.03it/s]
 62%|██████▏   | 6679/10708 [1:31:35<33:03,  2.03it/s]
 62%|██████▏   | 6680/10708 [1:31:35<33:03,  2.03it/s]
 62%|██████▏   | 6681/10708 [1:31:36<33:02,  2.03it/s]
 62%|██████▏   | 6682/10708 [1:31:36<33:00,  2.03it/s]
 62%|██████▏   | 6683/10708 [1:31:37<33:00,  2.03it/s]
 62%|██████▏   | 6684/10708 [1:31:37<32:58,  2.03it/s]
 62%|██████▏   | 6685/10708 [1:31:38<32:58,  2.03it/s]
 62%|██████▏   | 6686/10708 [1:31:38<32:58,  2.03it/s]
 62%|██████▏   | 6687/10708 [1:31:39<32:57,  2.03it/s]
 62%|██████▏   | 6688/10708 [1:31:39<32:57,  2.03it/s]
 62%|██████▏   | 6689/10708 [1:31:40<33:02,  2.03it/s]
 62%|██████▏   | 6690/10708 [1:31:40<32:59,  2.03it/s]
 62%|██████▏   | 6691/10708 [1:31:41<32:56,  2.03it/s]
 62%|██████▏   | 6692/10708 [1:31:41<32:56,  2.03it/s]
 63%|██████▎   | 6693/10708 [1:31:41<32:56,  2.03it/s]
 63%|██████▎   | 6694/10708 [1:31:42<32:55,  2.03it/s]
 63%|██████▎   | 6695/10708 [1:31:42<33:00,  2.03it/s]
 63%|██████▎   | 6696/10708 [1:31:43<33:02,  2.02it/s]
 63%|██████▎   | 6697/10708 [1:31:43<33:00,  2.03it/s]
 63%|██████▎   | 6698/10708 [1:31:44<38:32,  1.73it/s]
 63%|██████▎   | 6699/10708 [1:31:45<36:48,  1.82it/s]
 63%|██████▎   | 6700/10708 [1:31:45<35:37,  1.88it/s]{'loss': 3.6176, 'grad_norm': 0.20766480267047882, 'learning_rate': 0.0003694231587734391, 'epoch': 0.63}

                                                      

 63%|██████▎   | 6700/10708 [1:31:45<35:37,  1.88it/s]
 63%|██████▎   | 6701/10708 [1:31:46<34:51,  1.92it/s]
 63%|██████▎   | 6702/10708 [1:31:46<34:15,  1.95it/s]
 63%|██████▎   | 6703/10708 [1:31:47<33:49,  1.97it/s]
 63%|██████▎   | 6704/10708 [1:31:47<33:31,  1.99it/s]
 63%|██████▎   | 6705/10708 [1:31:48<33:19,  2.00it/s]
 63%|██████▎   | 6706/10708 [1:31:48<33:09,  2.01it/s]
 63%|██████▎   | 6707/10708 [1:31:49<33:03,  2.02it/s]
 63%|██████▎   | 6708/10708 [1:31:49<32:57,  2.02it/s]
 63%|██████▎   | 6709/10708 [1:31:50<32:52,  2.03it/s]
 63%|██████▎   | 6710/10708 [1:31:50<32:48,  2.03it/s]
 63%|██████▎   | 6711/10708 [1:31:51<32:47,  2.03it/s]
 63%|██████▎   | 6712/10708 [1:31:51<32:44,  2.03it/s]
 63%|██████▎   | 6713/10708 [1:31:52<32:44,  2.03it/s]
 63%|██████▎   | 6714/10708 [1:31:52<32:44,  2.03it/s]
 63%|██████▎   | 6715/10708 [1:31:53<32:42,  2.03it/s]
 63%|██████▎   | 6716/10708 [1:31:53<32:41,  2.03it/s]
 63%|██████▎   | 6717/10708 [1:31:54<32:42,  2.03it/s]
 63%|██████▎   | 6718/10708 [1:31:54<32:40,  2.03it/s]
 63%|██████▎   | 6719/10708 [1:31:55<32:39,  2.04it/s]
 63%|██████▎   | 6720/10708 [1:31:55<32:39,  2.03it/s]
 63%|██████▎   | 6721/10708 [1:31:56<32:37,  2.04it/s]
 63%|██████▎   | 6722/10708 [1:31:56<32:36,  2.04it/s]
 63%|██████▎   | 6723/10708 [1:31:57<32:37,  2.04it/s]
 63%|██████▎   | 6724/10708 [1:31:57<32:38,  2.03it/s]
 63%|██████▎   | 6725/10708 [1:31:58<32:38,  2.03it/s]
                                                      {'loss': 3.6025, 'grad_norm': 0.20862355828285217, 'learning_rate': 0.00036549403931084484, 'epoch': 0.63}


 63%|██████▎   | 6725/10708 [1:31:58<32:38,  2.03it/s]
 63%|██████▎   | 6726/10708 [1:31:58<32:40,  2.03it/s]
 63%|██████▎   | 6727/10708 [1:31:59<32:37,  2.03it/s]
 63%|██████▎   | 6728/10708 [1:31:59<38:10,  1.74it/s]
 63%|██████▎   | 6729/10708 [1:32:00<36:29,  1.82it/s]
 63%|██████▎   | 6730/10708 [1:32:00<35:16,  1.88it/s]
 63%|██████▎   | 6731/10708 [1:32:01<34:24,  1.93it/s]
 63%|██████▎   | 6732/10708 [1:32:01<33:50,  1.96it/s]
 63%|██████▎   | 6733/10708 [1:32:02<33:28,  1.98it/s]
 63%|██████▎   | 6734/10708 [1:32:02<33:12,  1.99it/s]
 63%|██████▎   | 6735/10708 [1:32:03<33:03,  2.00it/s]
 63%|██████▎   | 6736/10708 [1:32:03<32:51,  2.01it/s]
 63%|██████▎   | 6737/10708 [1:32:04<32:44,  2.02it/s]
 63%|██████▎   | 6738/10708 [1:32:04<32:41,  2.02it/s]
 63%|██████▎   | 6739/10708 [1:32:05<32:36,  2.03it/s]
 63%|██████▎   | 6740/10708 [1:32:05<32:36,  2.03it/s]
 63%|██████▎   | 6741/10708 [1:32:06<32:34,  2.03it/s]
 63%|██████▎   | 6742/10708 [1:32:06<32:31,  2.03it/s]
 63%|██████▎   | 6743/10708 [1:32:07<32:28,  2.03it/s]
 63%|██████▎   | 6744/10708 [1:32:07<32:27,  2.04it/s]
 63%|██████▎   | 6745/10708 [1:32:08<32:27,  2.04it/s]
 63%|██████▎   | 6746/10708 [1:32:08<32:26,  2.04it/s]
 63%|██████▎   | 6747/10708 [1:32:09<32:27,  2.03it/s]
 63%|██████▎   | 6748/10708 [1:32:09<32:24,  2.04it/s]
 63%|██████▎   | 6749/10708 [1:32:10<32:26,  2.03it/s]
 63%|██████▎   | 6750/10708 [1:32:10<32:28,  2.03it/s]
                                                      
{'loss': 3.6025, 'grad_norm': 0.2081010937690735, 'learning_rate': 0.0003615738536265674, 'epoch': 0.63}

 63%|██████▎   | 6750/10708 [1:32:10<32:28,  2.03it/s]
 63%|██████▎   | 6751/10708 [1:32:11<32:27,  2.03it/s]
 63%|██████▎   | 6752/10708 [1:32:11<32:25,  2.03it/s]
 63%|██████▎   | 6753/10708 [1:32:12<32:26,  2.03it/s]
 63%|██████▎   | 6754/10708 [1:32:12<32:24,  2.03it/s]
 63%|██████▎   | 6755/10708 [1:32:13<32:25,  2.03it/s]
 63%|██████▎   | 6756/10708 [1:32:13<32:27,  2.03it/s]
 63%|██████▎   | 6757/10708 [1:32:14<32:28,  2.03it/s]
 63%|██████▎   | 6758/10708 [1:32:14<32:25,  2.03it/s]
 63%|██████▎   | 6759/10708 [1:32:15<32:25,  2.03it/s]
 63%|██████▎   | 6760/10708 [1:32:15<32:24,  2.03it/s]
 63%|██████▎   | 6761/10708 [1:32:16<32:23,  2.03it/s]
 63%|██████▎   | 6762/10708 [1:32:16<32:19,  2.03it/s]
 63%|██████▎   | 6763/10708 [1:32:16<32:17,  2.04it/s]
 63%|██████▎   | 6764/10708 [1:32:17<32:21,  2.03it/s]
 63%|██████▎   | 6765/10708 [1:32:17<32:21,  2.03it/s]
 63%|██████▎   | 6766/10708 [1:32:18<32:19,  2.03it/s]
 63%|██████▎   | 6767/10708 [1:32:18<32:20,  2.03it/s]
 63%|██████▎   | 6768/10708 [1:32:19<32:20,  2.03it/s]
 63%|██████▎   | 6769/10708 [1:32:19<32:18,  2.03it/s]
 63%|██████▎   | 6770/10708 [1:32:20<32:17,  2.03it/s]
 63%|██████▎   | 6771/10708 [1:32:20<32:15,  2.03it/s]
 63%|██████▎   | 6772/10708 [1:32:21<32:18,  2.03it/s]
 63%|██████▎   | 6773/10708 [1:32:21<32:16,  2.03it/s]
 63%|██████▎   | 6774/10708 [1:32:22<32:16,  2.03it/s]
 63%|██████▎   | 6775/10708 [1:32:22<32:16,  2.03it/s]{'loss': 3.6049, 'grad_norm': 0.21015094220638275, 'learning_rate': 0.00035766286209620445, 'epoch': 0.63}

                                                      

 63%|██████▎   | 6775/10708 [1:32:22<32:16,  2.03it/s]
 63%|██████▎   | 6776/10708 [1:32:23<32:17,  2.03it/s]
 63%|██████▎   | 6777/10708 [1:32:23<32:15,  2.03it/s]
 63%|██████▎   | 6778/10708 [1:32:24<32:15,  2.03it/s]
 63%|██████▎   | 6779/10708 [1:32:24<32:12,  2.03it/s]
 63%|██████▎   | 6780/10708 [1:32:25<32:09,  2.04it/s]
 63%|██████▎   | 6781/10708 [1:32:25<32:13,  2.03it/s]
 63%|██████▎   | 6782/10708 [1:32:26<32:10,  2.03it/s]
 63%|██████▎   | 6783/10708 [1:32:26<32:08,  2.04it/s]
 63%|██████▎   | 6784/10708 [1:32:27<32:11,  2.03it/s]
 63%|██████▎   | 6785/10708 [1:32:27<32:08,  2.03it/s]
 63%|██████▎   | 6786/10708 [1:32:28<32:07,  2.03it/s]
 63%|██████▎   | 6787/10708 [1:32:28<32:09,  2.03it/s]
 63%|██████▎   | 6788/10708 [1:32:29<32:06,  2.03it/s]
 63%|██████▎   | 6789/10708 [1:32:29<32:06,  2.03it/s]
 63%|██████▎   | 6790/10708 [1:32:30<32:05,  2.03it/s]
 63%|██████▎   | 6791/10708 [1:32:30<32:03,  2.04it/s]
 63%|██████▎   | 6792/10708 [1:32:31<32:03,  2.04it/s]
 63%|██████▎   | 6793/10708 [1:32:31<32:05,  2.03it/s]
 63%|██████▎   | 6794/10708 [1:32:32<32:04,  2.03it/s]
 63%|██████▎   | 6795/10708 [1:32:32<32:04,  2.03it/s]
 63%|██████▎   | 6796/10708 [1:32:33<32:02,  2.03it/s]
 63%|██████▎   | 6797/10708 [1:32:33<32:00,  2.04it/s]
 63%|██████▎   | 6798/10708 [1:32:34<32:01,  2.03it/s]
 63%|██████▎   | 6799/10708 [1:32:34<32:00,  2.04it/s]
 64%|██████▎   | 6800/10708 [1:32:35<31:59,  2.04it/s]{'loss': 3.5999, 'grad_norm': 0.21331951022148132, 'learning_rate': 0.0003537613244846846, 'epoch': 0.64}

                                                      

 64%|██████▎   | 6800/10708 [1:32:35<31:59,  2.04it/s]
 64%|██████▎   | 6801/10708 [1:32:35<32:02,  2.03it/s]
 64%|██████▎   | 6802/10708 [1:32:36<32:00,  2.03it/s]
 64%|██████▎   | 6803/10708 [1:32:36<31:59,  2.03it/s]
 64%|██████▎   | 6804/10708 [1:32:37<32:00,  2.03it/s]
 64%|██████▎   | 6805/10708 [1:32:37<31:59,  2.03it/s]
 64%|██████▎   | 6806/10708 [1:32:38<31:59,  2.03it/s]
 64%|██████▎   | 6807/10708 [1:32:38<31:58,  2.03it/s]
 64%|██████▎   | 6808/10708 [1:32:39<31:56,  2.04it/s]
 64%|██████▎   | 6809/10708 [1:32:39<31:54,  2.04it/s]
 64%|██████▎   | 6810/10708 [1:32:40<31:55,  2.03it/s]
 64%|██████▎   | 6811/10708 [1:32:40<31:56,  2.03it/s]
 64%|██████▎   | 6812/10708 [1:32:41<31:53,  2.04it/s]
 64%|██████▎   | 6813/10708 [1:32:41<31:53,  2.04it/s]
 64%|██████▎   | 6814/10708 [1:32:42<31:54,  2.03it/s]
 64%|██████▎   | 6815/10708 [1:32:42<31:54,  2.03it/s]
 64%|██████▎   | 6816/10708 [1:32:43<31:54,  2.03it/s]
 64%|██████▎   | 6817/10708 [1:32:43<31:59,  2.03it/s]
 64%|██████▎   | 6818/10708 [1:32:44<32:02,  2.02it/s]
 64%|██████▎   | 6819/10708 [1:32:44<32:00,  2.03it/s]
 64%|██████▎   | 6820/10708 [1:32:45<31:58,  2.03it/s]
 64%|██████▎   | 6821/10708 [1:32:45<31:52,  2.03it/s]
 64%|██████▎   | 6822/10708 [1:32:46<31:50,  2.03it/s]
 64%|██████▎   | 6823/10708 [1:32:46<31:52,  2.03it/s]
 64%|██████▎   | 6824/10708 [1:32:46<31:49,  2.03it/s]
 64%|██████▎   | 6825/10708 [1:32:47<31:50,  2.03it/s]
                                                      
{'loss': 3.5988, 'grad_norm': 0.20647749304771423, 'learning_rate': 0.0003498694999290153, 'epoch': 0.64}

 64%|██████▎   | 6825/10708 [1:32:47<31:50,  2.03it/s]
 64%|██████▎   | 6826/10708 [1:32:47<31:52,  2.03it/s]
 64%|██████▍   | 6827/10708 [1:32:48<31:48,  2.03it/s]
 64%|██████▍   | 6828/10708 [1:32:48<31:52,  2.03it/s]
 64%|██████▍   | 6829/10708 [1:32:49<31:50,  2.03it/s]
 64%|██████▍   | 6830/10708 [1:32:49<31:46,  2.03it/s]
 64%|██████▍   | 6831/10708 [1:32:50<31:44,  2.04it/s]
 64%|██████▍   | 6832/10708 [1:32:50<31:45,  2.03it/s]
 64%|██████▍   | 6833/10708 [1:32:51<31:44,  2.03it/s]
 64%|██████▍   | 6834/10708 [1:32:51<31:42,  2.04it/s]
 64%|██████▍   | 6835/10708 [1:32:52<31:44,  2.03it/s]
 64%|██████▍   | 6836/10708 [1:32:52<31:44,  2.03it/s]
 64%|██████▍   | 6837/10708 [1:32:53<31:45,  2.03it/s]
 64%|██████▍   | 6838/10708 [1:32:53<31:46,  2.03it/s]
 64%|██████▍   | 6839/10708 [1:32:54<31:42,  2.03it/s]
 64%|██████▍   | 6840/10708 [1:32:54<31:42,  2.03it/s]
 64%|██████▍   | 6841/10708 [1:32:55<31:40,  2.04it/s]
 64%|██████▍   | 6842/10708 [1:32:55<31:38,  2.04it/s]
 64%|██████▍   | 6843/10708 [1:32:56<31:40,  2.03it/s]
 64%|██████▍   | 6844/10708 [1:32:56<31:40,  2.03it/s]
 64%|██████▍   | 6845/10708 [1:32:57<31:40,  2.03it/s]
 64%|██████▍   | 6846/10708 [1:32:57<31:39,  2.03it/s]
 64%|██████▍   | 6847/10708 [1:32:58<31:38,  2.03it/s]
 64%|██████▍   | 6848/10708 [1:32:58<31:36,  2.04it/s]
 64%|██████▍   | 6849/10708 [1:32:59<31:37,  2.03it/s]
 64%|██████▍   | 6850/10708 [1:32:59<31:38,  2.03it/s]{'loss': 3.5918, 'grad_norm': 0.19739456474781036, 'learning_rate': 0.00034598764692107044, 'epoch': 0.64}
                                                      

 64%|██████▍   | 6850/10708 [1:32:59<31:38,  2.03it/s]
 64%|██████▍   | 6851/10708 [1:33:00<31:41,  2.03it/s]
 64%|██████▍   | 6852/10708 [1:33:00<31:39,  2.03it/s]
 64%|██████▍   | 6853/10708 [1:33:01<31:35,  2.03it/s]
 64%|██████▍   | 6854/10708 [1:33:01<31:34,  2.03it/s]
 64%|██████▍   | 6855/10708 [1:33:02<31:35,  2.03it/s]
 64%|██████▍   | 6856/10708 [1:33:02<31:33,  2.03it/s]
 64%|██████▍   | 6857/10708 [1:33:03<31:33,  2.03it/s]
 64%|██████▍   | 6858/10708 [1:33:03<31:33,  2.03it/s]
 64%|██████▍   | 6859/10708 [1:33:04<31:30,  2.04it/s]
 64%|██████▍   | 6860/10708 [1:33:04<31:30,  2.04it/s]
 64%|██████▍   | 6861/10708 [1:33:05<31:29,  2.04it/s]
 64%|██████▍   | 6862/10708 [1:33:05<31:32,  2.03it/s]
 64%|██████▍   | 6863/10708 [1:33:06<31:30,  2.03it/s]
 64%|██████▍   | 6864/10708 [1:33:06<31:30,  2.03it/s]
 64%|██████▍   | 6865/10708 [1:33:07<31:31,  2.03it/s]
 64%|██████▍   | 6866/10708 [1:33:07<31:32,  2.03it/s]
 64%|██████▍   | 6867/10708 [1:33:08<31:30,  2.03it/s]
 64%|██████▍   | 6868/10708 [1:33:08<31:30,  2.03it/s]
 64%|██████▍   | 6869/10708 [1:33:09<31:28,  2.03it/s]
 64%|██████▍   | 6870/10708 [1:33:09<31:29,  2.03it/s]
 64%|██████▍   | 6871/10708 [1:33:10<31:27,  2.03it/s]
 64%|██████▍   | 6872/10708 [1:33:10<31:26,  2.03it/s]
 64%|██████▍   | 6873/10708 [1:33:11<31:28,  2.03it/s]
 64%|██████▍   | 6874/10708 [1:33:11<31:28,  2.03it/s]
 64%|██████▍   | 6875/10708 [1:33:12<31:26,  2.03it/s]
                                                      
{'loss': 3.5935, 'grad_norm': 0.1955113708972931, 'learning_rate': 0.0003421160232904219, 'epoch': 0.64}

 64%|██████▍   | 6875/10708 [1:33:12<31:26,  2.03it/s]
 64%|██████▍   | 6876/10708 [1:33:12<31:29,  2.03it/s]
 64%|██████▍   | 6877/10708 [1:33:13<31:28,  2.03it/s]
 64%|██████▍   | 6878/10708 [1:33:13<31:29,  2.03it/s]
 64%|██████▍   | 6879/10708 [1:33:14<31:28,  2.03it/s]
 64%|██████▍   | 6880/10708 [1:33:14<31:29,  2.03it/s]
 64%|██████▍   | 6881/10708 [1:33:15<31:25,  2.03it/s]
 64%|██████▍   | 6882/10708 [1:33:15<31:24,  2.03it/s]
 64%|██████▍   | 6883/10708 [1:33:16<31:24,  2.03it/s]
 64%|██████▍   | 6884/10708 [1:33:16<31:21,  2.03it/s]
 64%|██████▍   | 6885/10708 [1:33:17<31:21,  2.03it/s]
 64%|██████▍   | 6886/10708 [1:33:17<31:21,  2.03it/s]
 64%|██████▍   | 6887/10708 [1:33:17<31:19,  2.03it/s]
 64%|██████▍   | 6888/10708 [1:33:18<31:17,  2.03it/s]
 64%|██████▍   | 6889/10708 [1:33:18<31:19,  2.03it/s]
 64%|██████▍   | 6890/10708 [1:33:19<31:18,  2.03it/s]
 64%|██████▍   | 6891/10708 [1:33:19<31:15,  2.04it/s]
 64%|██████▍   | 6892/10708 [1:33:20<31:15,  2.03it/s]
 64%|██████▍   | 6893/10708 [1:33:20<31:14,  2.04it/s]
 64%|██████▍   | 6894/10708 [1:33:21<31:17,  2.03it/s]
 64%|██████▍   | 6895/10708 [1:33:21<31:16,  2.03it/s]
 64%|██████▍   | 6896/10708 [1:33:22<31:13,  2.03it/s]
 64%|██████▍   | 6897/10708 [1:33:22<31:12,  2.04it/s]
 64%|██████▍   | 6898/10708 [1:33:23<31:14,  2.03it/s]
 64%|██████▍   | 6899/10708 [1:33:23<31:11,  2.04it/s]
 64%|██████▍   | 6900/10708 [1:33:24<31:09,  2.04it/s]{'loss': 3.5959, 'grad_norm': 0.19794796407222748, 'learning_rate': 0.0003382548861872139, 'epoch': 0.64}
                                                      

 64%|██████▍   | 6900/10708 [1:33:24<31:09,  2.04it/s]
 64%|██████▍   | 6901/10708 [1:33:24<31:12,  2.03it/s]
 64%|██████▍   | 6902/10708 [1:33:25<31:11,  2.03it/s]
 64%|██████▍   | 6903/10708 [1:33:25<31:11,  2.03it/s]
 64%|██████▍   | 6904/10708 [1:33:26<31:13,  2.03it/s]
 64%|██████▍   | 6905/10708 [1:33:26<31:11,  2.03it/s]
 64%|██████▍   | 6906/10708 [1:33:27<31:10,  2.03it/s]
 65%|██████▍   | 6907/10708 [1:33:27<31:10,  2.03it/s]
 65%|██████▍   | 6908/10708 [1:33:28<31:08,  2.03it/s]
 65%|██████▍   | 6909/10708 [1:33:28<31:08,  2.03it/s]
 65%|██████▍   | 6910/10708 [1:33:29<31:07,  2.03it/s]
 65%|██████▍   | 6911/10708 [1:33:29<31:06,  2.03it/s]
 65%|██████▍   | 6912/10708 [1:33:30<31:07,  2.03it/s]
 65%|██████▍   | 6913/10708 [1:33:30<31:08,  2.03it/s]
 65%|██████▍   | 6914/10708 [1:33:31<31:06,  2.03it/s]
 65%|██████▍   | 6915/10708 [1:33:31<31:03,  2.04it/s]
 65%|██████▍   | 6916/10708 [1:33:32<31:02,  2.04it/s]
 65%|██████▍   | 6917/10708 [1:33:32<31:00,  2.04it/s]
 65%|██████▍   | 6918/10708 [1:33:33<31:00,  2.04it/s]
 65%|██████▍   | 6919/10708 [1:33:33<31:04,  2.03it/s]
 65%|██████▍   | 6920/10708 [1:33:34<31:05,  2.03it/s]
 65%|██████▍   | 6921/10708 [1:33:34<31:05,  2.03it/s]
 65%|██████▍   | 6922/10708 [1:33:35<31:01,  2.03it/s]
 65%|██████▍   | 6923/10708 [1:33:35<31:05,  2.03it/s]
 65%|██████▍   | 6924/10708 [1:33:36<31:03,  2.03it/s]
 65%|██████▍   | 6925/10708 [1:33:36<31:04,  2.03it/s]
                                                      {'loss': 3.5987, 'grad_norm': 0.2107304483652115, 'learning_rate': 0.0003344044920650846, 'epoch': 0.65}

 65%|██████▍   | 6925/10708 [1:33:36<31:04,  2.03it/s]
 65%|██████▍   | 6926/10708 [1:33:37<31:05,  2.03it/s]
 65%|██████▍   | 6927/10708 [1:33:37<31:02,  2.03it/s]
 65%|██████▍   | 6928/10708 [1:33:38<31:03,  2.03it/s]
 65%|██████▍   | 6929/10708 [1:33:38<31:02,  2.03it/s]
 65%|██████▍   | 6930/10708 [1:33:39<31:02,  2.03it/s]
 65%|██████▍   | 6931/10708 [1:33:39<31:02,  2.03it/s]
 65%|██████▍   | 6932/10708 [1:33:40<31:01,  2.03it/s]
 65%|██████▍   | 6933/10708 [1:33:40<30:58,  2.03it/s]
 65%|██████▍   | 6934/10708 [1:33:41<30:57,  2.03it/s]
 65%|██████▍   | 6935/10708 [1:33:41<30:55,  2.03it/s]
 65%|██████▍   | 6936/10708 [1:33:42<30:53,  2.04it/s]
 65%|██████▍   | 6937/10708 [1:33:42<30:53,  2.03it/s]
 65%|██████▍   | 6938/10708 [1:33:43<30:53,  2.03it/s]
 65%|██████▍   | 6939/10708 [1:33:43<30:59,  2.03it/s]
 65%|██████▍   | 6940/10708 [1:33:44<30:56,  2.03it/s]
 65%|██████▍   | 6941/10708 [1:33:44<30:56,  2.03it/s]
 65%|██████▍   | 6942/10708 [1:33:45<30:54,  2.03it/s]
 65%|██████▍   | 6943/10708 [1:33:45<30:52,  2.03it/s]
 65%|██████▍   | 6944/10708 [1:33:46<30:54,  2.03it/s]
 65%|██████▍   | 6945/10708 [1:33:46<30:56,  2.03it/s]
 65%|██████▍   | 6946/10708 [1:33:47<30:53,  2.03it/s]
 65%|██████▍   | 6947/10708 [1:33:47<30:52,  2.03it/s]
 65%|██████▍   | 6948/10708 [1:33:48<30:50,  2.03it/s]
 65%|██████▍   | 6949/10708 [1:33:48<30:46,  2.04it/s]
 65%|██████▍   | 6950/10708 [1:33:48<30:47,  2.03it/s]
                                                      
{'loss': 3.6019, 'grad_norm': 0.21149395406246185, 'learning_rate': 0.00033056509666413135, 'epoch': 0.65}

 65%|██████▍   | 6950/10708 [1:33:48<30:47,  2.03it/s]
 65%|██████▍   | 6951/10708 [1:33:49<30:52,  2.03it/s]
 65%|██████▍   | 6952/10708 [1:33:49<30:48,  2.03it/s]
 65%|██████▍   | 6953/10708 [1:33:50<30:48,  2.03it/s]
 65%|██████▍   | 6954/10708 [1:33:50<30:49,  2.03it/s]
 65%|██████▍   | 6955/10708 [1:33:51<30:48,  2.03it/s]
 65%|██████▍   | 6956/10708 [1:33:51<30:45,  2.03it/s]
 65%|██████▍   | 6957/10708 [1:33:52<30:45,  2.03it/s]
 65%|██████▍   | 6958/10708 [1:33:52<30:44,  2.03it/s]
 65%|██████▍   | 6959/10708 [1:33:53<30:42,  2.03it/s]
 65%|██████▍   | 6960/10708 [1:33:53<30:41,  2.04it/s]
 65%|██████▌   | 6961/10708 [1:33:54<30:44,  2.03it/s]
 65%|██████▌   | 6962/10708 [1:33:54<30:43,  2.03it/s]
 65%|██████▌   | 6963/10708 [1:33:55<30:40,  2.03it/s]
 65%|██████▌   | 6964/10708 [1:33:55<30:43,  2.03it/s]
 65%|██████▌   | 6965/10708 [1:33:56<30:41,  2.03it/s]
 65%|██████▌   | 6966/10708 [1:33:56<30:37,  2.04it/s]
 65%|██████▌   | 6967/10708 [1:33:57<30:38,  2.03it/s]
 65%|██████▌   | 6968/10708 [1:33:57<30:37,  2.03it/s]
 65%|██████▌   | 6969/10708 [1:33:58<30:36,  2.04it/s]
 65%|██████▌   | 6970/10708 [1:33:58<30:40,  2.03it/s]
 65%|██████▌   | 6971/10708 [1:33:59<30:41,  2.03it/s]
 65%|██████▌   | 6972/10708 [1:33:59<30:40,  2.03it/s]
 65%|██████▌   | 6973/10708 [1:34:00<30:38,  2.03it/s]
 65%|██████▌   | 6974/10708 [1:34:00<30:37,  2.03it/s]
 65%|██████▌   | 6975/10708 [1:34:01<30:40,  2.03it/s]{'loss': 3.5957, 'grad_norm': 0.1999390721321106, 'learning_rate': 0.00032673695499392606, 'epoch': 0.65}

                                                      

 65%|██████▌   | 6975/10708 [1:34:01<30:40,  2.03it/s]
 65%|██████▌   | 6976/10708 [1:34:01<30:38,  2.03it/s]
 65%|██████▌   | 6977/10708 [1:34:02<30:36,  2.03it/s]
 65%|██████▌   | 6978/10708 [1:34:02<30:38,  2.03it/s]
 65%|██████▌   | 6979/10708 [1:34:03<30:36,  2.03it/s]
 65%|██████▌   | 6980/10708 [1:34:03<30:35,  2.03it/s]
 65%|██████▌   | 6981/10708 [1:34:04<30:34,  2.03it/s]
 65%|██████▌   | 6982/10708 [1:34:04<30:32,  2.03it/s]
 65%|██████▌   | 6983/10708 [1:34:05<30:34,  2.03it/s]
 65%|██████▌   | 6984/10708 [1:34:05<30:35,  2.03it/s]
 65%|██████▌   | 6985/10708 [1:34:06<30:35,  2.03it/s]
 65%|██████▌   | 6986/10708 [1:34:06<30:33,  2.03it/s]
 65%|██████▌   | 6987/10708 [1:34:07<30:33,  2.03it/s]
 65%|██████▌   | 6988/10708 [1:34:07<30:31,  2.03it/s]
 65%|██████▌   | 6989/10708 [1:34:08<30:29,  2.03it/s]
 65%|██████▌   | 6990/10708 [1:34:08<30:29,  2.03it/s]
 65%|██████▌   | 6991/10708 [1:34:09<30:28,  2.03it/s]
 65%|██████▌   | 6992/10708 [1:34:09<30:27,  2.03it/s]
 65%|██████▌   | 6993/10708 [1:34:10<30:26,  2.03it/s]
 65%|██████▌   | 6994/10708 [1:34:10<30:25,  2.03it/s]
 65%|██████▌   | 6995/10708 [1:34:11<30:24,  2.03it/s]
 65%|██████▌   | 6996/10708 [1:34:11<30:24,  2.03it/s]
 65%|██████▌   | 6997/10708 [1:34:12<30:23,  2.04it/s]
 65%|██████▌   | 6998/10708 [1:34:12<30:23,  2.03it/s]
 65%|██████▌   | 6999/10708 [1:34:13<30:25,  2.03it/s]
 65%|██████▌   | 7000/10708 [1:34:13<30:27,  2.03it/s]{'loss': 3.5847, 'grad_norm': 0.20677202939987183, 'learning_rate': 0.00032292032131657565, 'epoch': 0.65}

                                                      

 65%|██████▌   | 7000/10708 [1:34:13<30:27,  2.03it/s]
 65%|██████▌   | 7001/10708 [1:34:14<30:31,  2.02it/s]
 65%|██████▌   | 7002/10708 [1:34:14<30:28,  2.03it/s]
 65%|██████▌   | 7003/10708 [1:34:15<30:28,  2.03it/s]
 65%|██████▌   | 7004/10708 [1:34:15<30:24,  2.03it/s]
 65%|██████▌   | 7005/10708 [1:34:16<30:23,  2.03it/s]
 65%|██████▌   | 7006/10708 [1:34:16<30:22,  2.03it/s]
 65%|██████▌   | 7007/10708 [1:34:17<30:22,  2.03it/s]
 65%|██████▌   | 7008/10708 [1:34:17<30:21,  2.03it/s]
 65%|██████▌   | 7009/10708 [1:34:18<30:19,  2.03it/s]
 65%|██████▌   | 7010/10708 [1:34:18<30:20,  2.03it/s]
 65%|██████▌   | 7011/10708 [1:34:19<30:20,  2.03it/s]
 65%|██████▌   | 7012/10708 [1:34:19<30:18,  2.03it/s]
 65%|██████▌   | 7013/10708 [1:34:20<30:18,  2.03it/s]
 66%|██████▌   | 7014/10708 [1:34:20<30:17,  2.03it/s]
 66%|██████▌   | 7015/10708 [1:34:20<30:15,  2.03it/s]
 66%|██████▌   | 7016/10708 [1:34:21<30:12,  2.04it/s]
 66%|██████▌   | 7017/10708 [1:34:21<30:15,  2.03it/s]
 66%|██████▌   | 7018/10708 [1:34:22<30:14,  2.03it/s]
 66%|██████▌   | 7019/10708 [1:34:22<30:11,  2.04it/s]
 66%|██████▌   | 7020/10708 [1:34:23<30:11,  2.04it/s]
 66%|██████▌   | 7021/10708 [1:34:23<30:12,  2.03it/s]
 66%|██████▌   | 7022/10708 [1:34:24<30:09,  2.04it/s]
 66%|██████▌   | 7023/10708 [1:34:24<30:13,  2.03it/s]
 66%|██████▌   | 7024/10708 [1:34:25<30:12,  2.03it/s]
 66%|██████▌   | 7025/10708 [1:34:25<30:10,  2.03it/s]{'loss': 3.5957, 'grad_norm': 0.212770015001297, 'learning_rate': 0.00031911544912983654, 'epoch': 0.66}
                                                      

 66%|██████▌   | 7025/10708 [1:34:25<30:10,  2.03it/s]
 66%|██████▌   | 7026/10708 [1:34:26<30:13,  2.03it/s]
 66%|██████▌   | 7027/10708 [1:34:26<30:14,  2.03it/s]
 66%|██████▌   | 7028/10708 [1:34:27<30:14,  2.03it/s]
 66%|██████▌   | 7029/10708 [1:34:27<30:09,  2.03it/s]
 66%|██████▌   | 7030/10708 [1:34:28<30:10,  2.03it/s]
 66%|██████▌   | 7031/10708 [1:34:28<30:11,  2.03it/s]
 66%|██████▌   | 7032/10708 [1:34:29<30:08,  2.03it/s]
 66%|██████▌   | 7033/10708 [1:34:29<30:08,  2.03it/s]
 66%|██████▌   | 7034/10708 [1:34:30<30:09,  2.03it/s]
 66%|██████▌   | 7035/10708 [1:34:30<30:08,  2.03it/s]
 66%|██████▌   | 7036/10708 [1:34:31<30:06,  2.03it/s]
 66%|██████▌   | 7037/10708 [1:34:31<30:06,  2.03it/s]
 66%|██████▌   | 7038/10708 [1:34:32<30:05,  2.03it/s]
 66%|██████▌   | 7039/10708 [1:34:32<30:03,  2.03it/s]
 66%|██████▌   | 7040/10708 [1:34:33<30:05,  2.03it/s]
 66%|██████▌   | 7041/10708 [1:34:33<30:03,  2.03it/s]
 66%|██████▌   | 7042/10708 [1:34:34<30:05,  2.03it/s]
 66%|██████▌   | 7043/10708 [1:34:34<30:06,  2.03it/s]
 66%|██████▌   | 7044/10708 [1:34:35<30:02,  2.03it/s]
 66%|██████▌   | 7045/10708 [1:34:35<30:03,  2.03it/s]
 66%|██████▌   | 7046/10708 [1:34:36<30:02,  2.03it/s]
 66%|██████▌   | 7047/10708 [1:34:36<29:59,  2.03it/s]
 66%|██████▌   | 7048/10708 [1:34:37<30:01,  2.03it/s]
 66%|██████▌   | 7049/10708 [1:34:37<29:59,  2.03it/s]
 66%|██████▌   | 7050/10708 [1:34:38<29:58,  2.03it/s]
                                                      
{'loss': 3.5915, 'grad_norm': 0.20401093363761902, 'learning_rate': 0.0003153225911502755, 'epoch': 0.66}

 66%|██████▌   | 7050/10708 [1:34:38<29:58,  2.03it/s]
 66%|██████▌   | 7051/10708 [1:34:38<30:00,  2.03it/s]
 66%|██████▌   | 7052/10708 [1:34:39<29:58,  2.03it/s]
 66%|██████▌   | 7053/10708 [1:34:39<29:57,  2.03it/s]
 66%|██████▌   | 7054/10708 [1:34:40<29:56,  2.03it/s]
 66%|██████▌   | 7055/10708 [1:34:40<29:56,  2.03it/s]
 66%|██████▌   | 7056/10708 [1:34:41<29:55,  2.03it/s]
 66%|██████▌   | 7057/10708 [1:34:41<29:58,  2.03it/s]
 66%|██████▌   | 7058/10708 [1:34:42<29:56,  2.03it/s]
 66%|██████▌   | 7059/10708 [1:34:42<29:56,  2.03it/s]
 66%|██████▌   | 7060/10708 [1:34:43<29:57,  2.03it/s]
 66%|██████▌   | 7061/10708 [1:34:43<30:01,  2.02it/s]
 66%|██████▌   | 7062/10708 [1:34:44<29:58,  2.03it/s]
 66%|██████▌   | 7063/10708 [1:34:44<29:55,  2.03it/s]
 66%|██████▌   | 7064/10708 [1:34:45<29:54,  2.03it/s]
 66%|██████▌   | 7065/10708 [1:34:45<29:51,  2.03it/s]
 66%|██████▌   | 7066/10708 [1:34:46<29:50,  2.03it/s]
 66%|██████▌   | 7067/10708 [1:34:46<29:51,  2.03it/s]
 66%|██████▌   | 7068/10708 [1:34:47<29:49,  2.03it/s]
 66%|██████▌   | 7069/10708 [1:34:47<29:49,  2.03it/s]
 66%|██████▌   | 7070/10708 [1:34:48<29:50,  2.03it/s]
 66%|██████▌   | 7071/10708 [1:34:48<29:48,  2.03it/s]
 66%|██████▌   | 7072/10708 [1:34:49<29:48,  2.03it/s]
 66%|██████▌   | 7073/10708 [1:34:49<29:48,  2.03it/s]
 66%|██████▌   | 7074/10708 [1:34:50<29:44,  2.04it/s]
 66%|██████▌   | 7075/10708 [1:34:50<29:46,  2.03it/s]{'loss': 3.5942, 'grad_norm': 0.20173300802707672, 'learning_rate': 0.00031154199929648587, 'epoch': 0.66}

                                                      

 66%|██████▌   | 7075/10708 [1:34:50<29:46,  2.03it/s]
 66%|██████▌   | 7076/10708 [1:34:51<29:47,  2.03it/s]
 66%|██████▌   | 7077/10708 [1:34:51<29:47,  2.03it/s]
 66%|██████▌   | 7078/10708 [1:34:51<29:46,  2.03it/s]
 66%|██████▌   | 7079/10708 [1:34:52<29:46,  2.03it/s]
 66%|██████▌   | 7080/10708 [1:34:52<29:45,  2.03it/s]
 66%|██████▌   | 7081/10708 [1:34:53<29:45,  2.03it/s]
 66%|██████▌   | 7082/10708 [1:34:53<29:44,  2.03it/s]
 66%|██████▌   | 7083/10708 [1:34:54<29:44,  2.03it/s]
 66%|██████▌   | 7084/10708 [1:34:54<29:44,  2.03it/s]
 66%|██████▌   | 7085/10708 [1:34:55<29:42,  2.03it/s]
 66%|██████▌   | 7086/10708 [1:34:55<29:43,  2.03it/s]
 66%|██████▌   | 7087/10708 [1:34:56<29:42,  2.03it/s]
 66%|██████▌   | 7088/10708 [1:34:56<29:40,  2.03it/s]
 66%|██████▌   | 7089/10708 [1:34:57<29:40,  2.03it/s]
 66%|██████▌   | 7090/10708 [1:34:57<29:41,  2.03it/s]
 66%|██████▌   | 7091/10708 [1:34:58<29:41,  2.03it/s]
 66%|██████▌   | 7092/10708 [1:34:58<29:37,  2.03it/s]
 66%|██████▌   | 7093/10708 [1:34:59<29:35,  2.04it/s]
 66%|██████▌   | 7094/10708 [1:34:59<29:38,  2.03it/s]
 66%|██████▋   | 7095/10708 [1:35:00<29:37,  2.03it/s]
 66%|██████▋   | 7096/10708 [1:35:00<29:34,  2.04it/s]
 66%|██████▋   | 7097/10708 [1:35:01<29:37,  2.03it/s]
 66%|██████▋   | 7098/10708 [1:35:01<29:35,  2.03it/s]
 66%|██████▋   | 7099/10708 [1:35:02<29:34,  2.03it/s]
 66%|██████▋   | 7100/10708 [1:35:02<29:34,  2.03it/s]
                                                      {'loss': 3.5931, 'grad_norm': 0.20071572065353394, 'learning_rate': 0.0003077739246723543, 'epoch': 0.66}


 66%|██████▋   | 7100/10708 [1:35:02<29:34,  2.03it/s]
 66%|██████▋   | 7101/10708 [1:35:03<29:34,  2.03it/s]
 66%|██████▋   | 7102/10708 [1:35:03<29:34,  2.03it/s]
 66%|██████▋   | 7103/10708 [1:35:04<29:33,  2.03it/s]
 66%|██████▋   | 7104/10708 [1:35:04<29:33,  2.03it/s]
 66%|██████▋   | 7105/10708 [1:35:05<29:32,  2.03it/s]
 66%|██████▋   | 7106/10708 [1:35:05<29:33,  2.03it/s]
 66%|██████▋   | 7107/10708 [1:35:06<29:32,  2.03it/s]
 66%|██████▋   | 7108/10708 [1:35:06<29:32,  2.03it/s]
 66%|██████▋   | 7109/10708 [1:35:07<29:31,  2.03it/s]
 66%|██████▋   | 7110/10708 [1:35:07<29:28,  2.03it/s]
 66%|██████▋   | 7111/10708 [1:35:08<29:32,  2.03it/s]
 66%|██████▋   | 7112/10708 [1:35:08<29:29,  2.03it/s]
 66%|██████▋   | 7113/10708 [1:35:09<29:27,  2.03it/s]
 66%|██████▋   | 7114/10708 [1:35:09<29:29,  2.03it/s]
 66%|██████▋   | 7115/10708 [1:35:10<29:26,  2.03it/s]
 66%|██████▋   | 7116/10708 [1:35:10<29:25,  2.03it/s]
 66%|██████▋   | 7117/10708 [1:35:11<29:26,  2.03it/s]
 66%|██████▋   | 7118/10708 [1:35:11<29:25,  2.03it/s]
 66%|██████▋   | 7119/10708 [1:35:12<29:24,  2.03it/s]
 66%|██████▋   | 7120/10708 [1:35:12<29:23,  2.04it/s]
 67%|██████▋   | 7121/10708 [1:35:13<29:24,  2.03it/s]
 67%|██████▋   | 7122/10708 [1:35:13<29:28,  2.03it/s]
 67%|██████▋   | 7123/10708 [1:35:14<29:25,  2.03it/s]
 67%|██████▋   | 7124/10708 [1:35:14<29:23,  2.03it/s]
 67%|██████▋   | 7125/10708 [1:35:15<29:22,  2.03it/s]
                                                      
{'loss': 3.5859, 'grad_norm': 0.20278993248939514, 'learning_rate': 0.0003040186175503832, 'epoch': 0.67}

 67%|██████▋   | 7125/10708 [1:35:15<29:22,  2.03it/s]
 67%|██████▋   | 7126/10708 [1:35:15<29:26,  2.03it/s]
 67%|██████▋   | 7127/10708 [1:35:16<29:26,  2.03it/s]
 67%|██████▋   | 7128/10708 [1:35:16<29:24,  2.03it/s]
 67%|██████▋   | 7129/10708 [1:35:17<29:25,  2.03it/s]
 67%|██████▋   | 7130/10708 [1:35:17<29:23,  2.03it/s]
 67%|██████▋   | 7131/10708 [1:35:18<29:18,  2.03it/s]
 67%|██████▋   | 7132/10708 [1:35:18<29:19,  2.03it/s]
 67%|██████▋   | 7133/10708 [1:35:19<29:18,  2.03it/s]
 67%|██████▋   | 7134/10708 [1:35:19<29:19,  2.03it/s]
 67%|██████▋   | 7135/10708 [1:35:20<29:21,  2.03it/s]
 67%|██████▋   | 7136/10708 [1:35:20<29:18,  2.03it/s]
 67%|██████▋   | 7137/10708 [1:35:21<29:21,  2.03it/s]
 67%|██████▋   | 7138/10708 [1:35:21<29:19,  2.03it/s]
 67%|██████▋   | 7139/10708 [1:35:22<29:19,  2.03it/s]
 67%|██████▋   | 7140/10708 [1:35:22<29:18,  2.03it/s]
 67%|██████▋   | 7141/10708 [1:35:22<29:15,  2.03it/s]
 67%|██████▋   | 7142/10708 [1:35:23<29:17,  2.03it/s]
 67%|██████▋   | 7143/10708 [1:35:23<29:15,  2.03it/s]
 67%|██████▋   | 7144/10708 [1:35:24<29:13,  2.03it/s]
 67%|██████▋   | 7145/10708 [1:35:24<29:14,  2.03it/s]
 67%|██████▋   | 7146/10708 [1:35:25<29:13,  2.03it/s]
 67%|██████▋   | 7147/10708 [1:35:25<29:11,  2.03it/s]
 67%|██████▋   | 7148/10708 [1:35:26<29:11,  2.03it/s]
 67%|██████▋   | 7149/10708 [1:35:26<29:09,  2.03it/s]
 67%|██████▋   | 7150/10708 [1:35:27<29:09,  2.03it/s]{'loss': 3.5966, 'grad_norm': 0.1976289004087448, 'learning_rate': 0.00030027632735506773, 'epoch': 0.67}

                                                      

 67%|██████▋   | 7150/10708 [1:35:27<29:09,  2.03it/s]
 67%|██████▋   | 7151/10708 [1:35:27<29:12,  2.03it/s]
 67%|██████▋   | 7152/10708 [1:35:28<29:11,  2.03it/s]
 67%|██████▋   | 7153/10708 [1:35:28<29:10,  2.03it/s]
 67%|██████▋   | 7154/10708 [1:35:29<29:06,  2.03it/s]
 67%|██████▋   | 7155/10708 [1:35:29<29:08,  2.03it/s]
 67%|██████▋   | 7156/10708 [1:35:30<29:10,  2.03it/s]
 67%|██████▋   | 7157/10708 [1:35:30<29:07,  2.03it/s]
 67%|██████▋   | 7158/10708 [1:35:31<29:08,  2.03it/s]
 67%|██████▋   | 7159/10708 [1:35:31<29:07,  2.03it/s]
 67%|██████▋   | 7160/10708 [1:35:32<29:05,  2.03it/s]
 67%|██████▋   | 7161/10708 [1:35:32<29:09,  2.03it/s]
 67%|██████▋   | 7162/10708 [1:35:33<29:07,  2.03it/s]
 67%|██████▋   | 7163/10708 [1:35:33<29:05,  2.03it/s]
 67%|██████▋   | 7164/10708 [1:35:34<29:04,  2.03it/s]
 67%|██████▋   | 7165/10708 [1:35:34<29:02,  2.03it/s]
 67%|██████▋   | 7166/10708 [1:35:35<29:00,  2.03it/s]
 67%|██████▋   | 7167/10708 [1:35:35<29:00,  2.03it/s]
 67%|██████▋   | 7168/10708 [1:35:36<29:00,  2.03it/s]
 67%|██████▋   | 7169/10708 [1:35:36<28:58,  2.04it/s]
 67%|██████▋   | 7170/10708 [1:35:37<28:58,  2.03it/s]
 67%|██████▋   | 7171/10708 [1:35:37<28:59,  2.03it/s]
 67%|██████▋   | 7172/10708 [1:35:38<28:58,  2.03it/s]
 67%|██████▋   | 7173/10708 [1:35:38<29:01,  2.03it/s]
 67%|██████▋   | 7174/10708 [1:35:39<28:58,  2.03it/s]
 67%|██████▋   | 7175/10708 [1:35:39<28:58,  2.03it/s]{'loss': 3.589, 'grad_norm': 0.20862643420696259, 'learning_rate': 0.00029654730264632915, 'epoch': 0.67}

                                                      

 67%|██████▋   | 7175/10708 [1:35:39<28:58,  2.03it/s]
 67%|██████▋   | 7176/10708 [1:35:40<29:00,  2.03it/s]
 67%|██████▋   | 7177/10708 [1:35:40<28:56,  2.03it/s]
 67%|██████▋   | 7178/10708 [1:35:41<28:58,  2.03it/s]
 67%|██████▋   | 7179/10708 [1:35:41<28:58,  2.03it/s]
 67%|██████▋   | 7180/10708 [1:35:42<28:54,  2.03it/s]
 67%|██████▋   | 7181/10708 [1:35:42<28:56,  2.03it/s]
 67%|██████▋   | 7182/10708 [1:35:43<28:56,  2.03it/s]
 67%|██████▋   | 7183/10708 [1:35:43<29:01,  2.02it/s]
 67%|██████▋   | 7184/10708 [1:35:44<28:58,  2.03it/s]
 67%|██████▋   | 7185/10708 [1:35:44<28:54,  2.03it/s]
 67%|██████▋   | 7186/10708 [1:35:45<28:55,  2.03it/s]
 67%|██████▋   | 7187/10708 [1:35:45<28:52,  2.03it/s]
 67%|██████▋   | 7188/10708 [1:35:46<28:50,  2.03it/s]
 67%|██████▋   | 7189/10708 [1:35:46<28:50,  2.03it/s]
 67%|██████▋   | 7190/10708 [1:35:47<28:51,  2.03it/s]
 67%|██████▋   | 7191/10708 [1:35:47<28:51,  2.03it/s]
 67%|██████▋   | 7192/10708 [1:35:48<28:50,  2.03it/s]
 67%|██████▋   | 7193/10708 [1:35:48<28:49,  2.03it/s]
 67%|██████▋   | 7194/10708 [1:35:49<28:48,  2.03it/s]
 67%|██████▋   | 7195/10708 [1:35:49<28:47,  2.03it/s]
 67%|██████▋   | 7196/10708 [1:35:50<28:45,  2.04it/s]
 67%|██████▋   | 7197/10708 [1:35:50<28:46,  2.03it/s]
 67%|██████▋   | 7198/10708 [1:35:51<28:45,  2.03it/s]
 67%|██████▋   | 7199/10708 [1:35:51<28:43,  2.04it/s]
 67%|██████▋   | 7200/10708 [1:35:52<28:43,  2.03it/s]
                                                      
{'loss': 3.5898, 'grad_norm': 0.208867609500885, 'learning_rate': 0.000292831791103006, 'epoch': 0.67}

 67%|██████▋   | 7200/10708 [1:35:52<28:43,  2.03it/s]
 67%|██████▋   | 7201/10708 [1:35:52<28:46,  2.03it/s]
 67%|██████▋   | 7202/10708 [1:35:53<28:47,  2.03it/s]
 67%|██████▋   | 7203/10708 [1:35:53<28:46,  2.03it/s]
 67%|██████▋   | 7204/10708 [1:35:54<28:45,  2.03it/s]
 67%|██████▋   | 7205/10708 [1:35:54<28:45,  2.03it/s]
 67%|██████▋   | 7206/10708 [1:35:54<28:45,  2.03it/s]
 67%|██████▋   | 7207/10708 [1:35:55<28:46,  2.03it/s]
 67%|██████▋   | 7208/10708 [1:35:55<28:43,  2.03it/s]
 67%|██████▋   | 7209/10708 [1:35:56<28:40,  2.03it/s]
 67%|██████▋   | 7210/10708 [1:35:56<28:40,  2.03it/s]
 67%|██████▋   | 7211/10708 [1:35:57<28:40,  2.03it/s]
 67%|██████▋   | 7212/10708 [1:35:57<28:39,  2.03it/s]
 67%|██████▋   | 7213/10708 [1:35:58<28:39,  2.03it/s]
 67%|██████▋   | 7214/10708 [1:35:58<28:38,  2.03it/s]
 67%|██████▋   | 7215/10708 [1:35:59<28:37,  2.03it/s]
 67%|██████▋   | 7216/10708 [1:35:59<28:38,  2.03it/s]
 67%|██████▋   | 7217/10708 [1:36:00<28:37,  2.03it/s]
 67%|██████▋   | 7218/10708 [1:36:00<28:36,  2.03it/s]
 67%|██████▋   | 7219/10708 [1:36:01<28:35,  2.03it/s]
 67%|██████▋   | 7220/10708 [1:36:01<28:34,  2.03it/s]
 67%|██████▋   | 7221/10708 [1:36:02<28:33,  2.03it/s]
 67%|██████▋   | 7222/10708 [1:36:02<28:31,  2.04it/s]
 67%|██████▋   | 7223/10708 [1:36:03<28:30,  2.04it/s]
 67%|██████▋   | 7224/10708 [1:36:03<28:33,  2.03it/s]
 67%|██████▋   | 7225/10708 [1:36:04<28:32,  2.03it/s]
                                                      
{'loss': 3.5836, 'grad_norm': 0.20241212844848633, 'learning_rate': 0.0002891300395064026, 'epoch': 0.67}

 67%|██████▋   | 7225/10708 [1:36:04<28:32,  2.03it/s]
 67%|██████▋   | 7226/10708 [1:36:04<28:35,  2.03it/s]
 67%|██████▋   | 7227/10708 [1:36:05<28:34,  2.03it/s]
 68%|██████▊   | 7228/10708 [1:36:05<28:32,  2.03it/s]
 68%|██████▊   | 7229/10708 [1:36:06<28:32,  2.03it/s]
 68%|██████▊   | 7230/10708 [1:36:06<28:32,  2.03it/s]
 68%|██████▊   | 7231/10708 [1:36:07<28:30,  2.03it/s]
 68%|██████▊   | 7232/10708 [1:36:07<28:31,  2.03it/s]
 68%|██████▊   | 7233/10708 [1:36:08<28:31,  2.03it/s]
 68%|██████▊   | 7234/10708 [1:36:08<28:29,  2.03it/s]
 68%|██████▊   | 7235/10708 [1:36:09<28:31,  2.03it/s]
 68%|██████▊   | 7236/10708 [1:36:09<28:32,  2.03it/s]
 68%|██████▊   | 7237/10708 [1:36:10<28:30,  2.03it/s]
 68%|██████▊   | 7238/10708 [1:36:10<28:27,  2.03it/s]
 68%|██████▊   | 7239/10708 [1:36:11<28:27,  2.03it/s]
 68%|██████▊   | 7240/10708 [1:36:11<28:26,  2.03it/s]
 68%|██████▊   | 7241/10708 [1:36:12<28:27,  2.03it/s]
 68%|██████▊   | 7242/10708 [1:36:12<28:27,  2.03it/s]
 68%|██████▊   | 7243/10708 [1:36:13<28:27,  2.03it/s]
 68%|██████▊   | 7244/10708 [1:36:13<28:29,  2.03it/s]
 68%|██████▊   | 7245/10708 [1:36:14<28:27,  2.03it/s]
 68%|██████▊   | 7246/10708 [1:36:14<28:28,  2.03it/s]
 68%|██████▊   | 7247/10708 [1:36:15<28:26,  2.03it/s]
 68%|██████▊   | 7248/10708 [1:36:15<28:25,  2.03it/s]
 68%|██████▊   | 7249/10708 [1:36:16<28:23,  2.03it/s]
 68%|██████▊   | 7250/10708 [1:36:16<28:25,  2.03it/s]{'loss': 3.5809, 'grad_norm': 0.20804046094417572, 'learning_rate': 0.0002854422937238992, 'epoch': 0.68}

                                                      

 68%|██████▊   | 7250/10708 [1:36:16<28:25,  2.03it/s]
 68%|██████▊   | 7251/10708 [1:36:17<28:25,  2.03it/s]
 68%|██████▊   | 7252/10708 [1:36:17<28:25,  2.03it/s]
 68%|██████▊   | 7253/10708 [1:36:18<28:25,  2.03it/s]
 68%|██████▊   | 7254/10708 [1:36:18<28:22,  2.03it/s]
 68%|██████▊   | 7255/10708 [1:36:19<28:20,  2.03it/s]
 68%|██████▊   | 7256/10708 [1:36:19<28:19,  2.03it/s]
 68%|██████▊   | 7257/10708 [1:36:20<28:16,  2.03it/s]
 68%|██████▊   | 7258/10708 [1:36:20<28:15,  2.03it/s]
 68%|██████▊   | 7259/10708 [1:36:21<28:15,  2.03it/s]
 68%|██████▊   | 7260/10708 [1:36:21<28:12,  2.04it/s]
 68%|██████▊   | 7261/10708 [1:36:22<28:13,  2.04it/s]
 68%|██████▊   | 7262/10708 [1:36:22<28:13,  2.03it/s]
 68%|██████▊   | 7263/10708 [1:36:23<28:13,  2.03it/s]
 68%|██████▊   | 7264/10708 [1:36:23<28:14,  2.03it/s]
 68%|██████▊   | 7265/10708 [1:36:24<28:12,  2.03it/s]
 68%|██████▊   | 7266/10708 [1:36:24<28:13,  2.03it/s]
 68%|██████▊   | 7267/10708 [1:36:25<28:14,  2.03it/s]
 68%|██████▊   | 7268/10708 [1:36:25<28:11,  2.03it/s]
 68%|██████▊   | 7269/10708 [1:36:25<28:11,  2.03it/s]
 68%|██████▊   | 7270/10708 [1:36:26<28:10,  2.03it/s]
 68%|██████▊   | 7271/10708 [1:36:26<28:10,  2.03it/s]
 68%|██████▊   | 7272/10708 [1:36:27<28:08,  2.03it/s]
 68%|██████▊   | 7273/10708 [1:36:27<28:09,  2.03it/s]
 68%|██████▊   | 7274/10708 [1:36:28<28:05,  2.04it/s]
 68%|██████▊   | 7275/10708 [1:36:28<28:06,  2.04it/s]{'loss': 3.5825, 'grad_norm': 0.21013985574245453, 'learning_rate': 0.0002817687986926206, 'epoch': 0.68}

                                                      

 68%|██████▊   | 7275/10708 [1:36:28<28:06,  2.04it/s]
 68%|██████▊   | 7276/10708 [1:36:29<28:09,  2.03it/s]
 68%|██████▊   | 7277/10708 [1:36:29<28:07,  2.03it/s]
 68%|██████▊   | 7278/10708 [1:36:30<28:10,  2.03it/s]
 68%|██████▊   | 7279/10708 [1:36:30<28:07,  2.03it/s]
 68%|██████▊   | 7280/10708 [1:36:31<28:08,  2.03it/s]
 68%|██████▊   | 7281/10708 [1:36:31<28:08,  2.03it/s]
 68%|██████▊   | 7282/10708 [1:36:32<28:05,  2.03it/s]
 68%|██████▊   | 7283/10708 [1:36:32<28:06,  2.03it/s]
 68%|██████▊   | 7284/10708 [1:36:33<28:05,  2.03it/s]
 68%|██████▊   | 7285/10708 [1:36:33<28:01,  2.04it/s]
 68%|██████▊   | 7286/10708 [1:36:34<28:02,  2.03it/s]
 68%|██████▊   | 7287/10708 [1:36:34<28:02,  2.03it/s]
 68%|██████▊   | 7288/10708 [1:36:35<28:00,  2.03it/s]
 68%|██████▊   | 7289/10708 [1:36:35<28:00,  2.03it/s]
 68%|██████▊   | 7290/10708 [1:36:36<27:59,  2.03it/s]
 68%|██████▊   | 7291/10708 [1:36:36<27:59,  2.03it/s]
 68%|██████▊   | 7292/10708 [1:36:37<28:00,  2.03it/s]
 68%|██████▊   | 7293/10708 [1:36:37<28:04,  2.03it/s]
 68%|██████▊   | 7294/10708 [1:36:38<28:02,  2.03it/s]
 68%|██████▊   | 7295/10708 [1:36:38<28:02,  2.03it/s]
 68%|██████▊   | 7296/10708 [1:36:39<28:00,  2.03it/s]
 68%|██████▊   | 7297/10708 [1:36:39<27:57,  2.03it/s]
 68%|██████▊   | 7298/10708 [1:36:40<27:57,  2.03it/s]
 68%|██████▊   | 7299/10708 [1:36:40<27:57,  2.03it/s]
 68%|██████▊   | 7300/10708 [1:36:41<27:57,  2.03it/s]
                                                      
{'loss': 3.5761, 'grad_norm': 0.19962915778160095, 'learning_rate': 0.0002781097984031684, 'epoch': 0.68}

 68%|██████▊   | 7300/10708 [1:36:41<27:57,  2.03it/s]
 68%|██████▊   | 7301/10708 [1:36:41<27:58,  2.03it/s]
 68%|██████▊   | 7302/10708 [1:36:42<27:59,  2.03it/s]
 68%|██████▊   | 7303/10708 [1:36:42<27:57,  2.03it/s]
 68%|██████▊   | 7304/10708 [1:36:43<28:01,  2.02it/s]
 68%|██████▊   | 7305/10708 [1:36:43<28:03,  2.02it/s]
 68%|██████▊   | 7306/10708 [1:36:44<28:00,  2.02it/s]
 68%|██████▊   | 7307/10708 [1:36:44<27:57,  2.03it/s]
 68%|██████▊   | 7308/10708 [1:36:45<27:54,  2.03it/s]
 68%|██████▊   | 7309/10708 [1:36:45<27:55,  2.03it/s]
 68%|██████▊   | 7310/10708 [1:36:46<27:54,  2.03it/s]
 68%|██████▊   | 7311/10708 [1:36:46<27:52,  2.03it/s]
 68%|██████▊   | 7312/10708 [1:36:47<27:50,  2.03it/s]
 68%|██████▊   | 7313/10708 [1:36:47<27:51,  2.03it/s]
 68%|██████▊   | 7314/10708 [1:36:48<27:51,  2.03it/s]
 68%|██████▊   | 7315/10708 [1:36:48<27:48,  2.03it/s]
 68%|██████▊   | 7316/10708 [1:36:49<27:48,  2.03it/s]
 68%|██████▊   | 7317/10708 [1:36:49<27:48,  2.03it/s]
 68%|██████▊   | 7318/10708 [1:36:50<27:47,  2.03it/s]
 68%|██████▊   | 7319/10708 [1:36:50<27:46,  2.03it/s]
 68%|██████▊   | 7320/10708 [1:36:51<27:45,  2.03it/s]
 68%|██████▊   | 7321/10708 [1:36:51<27:44,  2.03it/s]
 68%|██████▊   | 7322/10708 [1:36:52<27:43,  2.04it/s]
 68%|██████▊   | 7323/10708 [1:36:52<27:44,  2.03it/s]
 68%|██████▊   | 7324/10708 [1:36:53<27:42,  2.04it/s]
 68%|██████▊   | 7325/10708 [1:36:53<27:42,  2.04it/s]{'loss': 3.5804, 'grad_norm': 0.20632390677928925, 'learning_rate': 0.000274465535883415, 'epoch': 0.68}

                                                      

 68%|██████▊   | 7325/10708 [1:36:53<27:42,  2.04it/s]
 68%|██████▊   | 7326/10708 [1:36:54<27:44,  2.03it/s]
 68%|██████▊   | 7327/10708 [1:36:54<27:42,  2.03it/s]
 68%|██████▊   | 7328/10708 [1:36:55<27:44,  2.03it/s]
 68%|██████▊   | 7329/10708 [1:36:55<27:42,  2.03it/s]
 68%|██████▊   | 7330/10708 [1:36:56<27:42,  2.03it/s]
 68%|██████▊   | 7331/10708 [1:36:56<27:41,  2.03it/s]
 68%|██████▊   | 7332/10708 [1:36:57<27:39,  2.03it/s]
 68%|██████▊   | 7333/10708 [1:36:57<27:40,  2.03it/s]
 68%|██████▊   | 7334/10708 [1:36:57<27:39,  2.03it/s]
 69%|██████▊   | 7335/10708 [1:36:58<27:39,  2.03it/s]
 69%|██████▊   | 7336/10708 [1:36:58<27:38,  2.03it/s]
 69%|██████▊   | 7337/10708 [1:36:59<27:36,  2.04it/s]
 69%|██████▊   | 7338/10708 [1:36:59<27:35,  2.04it/s]
 69%|██████▊   | 7339/10708 [1:37:00<27:39,  2.03it/s]
 69%|██████▊   | 7340/10708 [1:37:00<27:37,  2.03it/s]
 69%|██████▊   | 7341/10708 [1:37:01<27:35,  2.03it/s]
 69%|██████▊   | 7342/10708 [1:37:01<27:37,  2.03it/s]
 69%|██████▊   | 7343/10708 [1:37:02<27:35,  2.03it/s]
 69%|██████▊   | 7344/10708 [1:37:02<27:35,  2.03it/s]
 69%|██████▊   | 7345/10708 [1:37:03<27:33,  2.03it/s]
 69%|██████▊   | 7346/10708 [1:37:03<27:32,  2.03it/s]
 69%|██████▊   | 7347/10708 [1:37:04<27:32,  2.03it/s]
 69%|██████▊   | 7348/10708 [1:37:04<27:34,  2.03it/s]
 69%|██████▊   | 7349/10708 [1:37:05<27:36,  2.03it/s]
 69%|██████▊   | 7350/10708 [1:37:05<27:34,  2.03it/s]{'loss': 3.5814, 'grad_norm': 0.2004912942647934, 'learning_rate': 0.0002708362531823621, 'epoch': 0.69}

                                                      

 69%|██████▊   | 7350/10708 [1:37:05<27:34,  2.03it/s]
 69%|██████▊   | 7351/10708 [1:37:06<27:37,  2.03it/s]
 69%|██████▊   | 7352/10708 [1:37:06<27:36,  2.03it/s]
 69%|██████▊   | 7353/10708 [1:37:07<27:33,  2.03it/s]
 69%|██████▊   | 7354/10708 [1:37:07<27:31,  2.03it/s]
 69%|██████▊   | 7355/10708 [1:37:08<27:30,  2.03it/s]
 69%|██████▊   | 7356/10708 [1:37:08<27:28,  2.03it/s]
 69%|██████▊   | 7357/10708 [1:37:09<27:29,  2.03it/s]
 69%|██████▊   | 7358/10708 [1:37:09<27:28,  2.03it/s]
 69%|██████▊   | 7359/10708 [1:37:10<27:28,  2.03it/s]
 69%|██████▊   | 7360/10708 [1:37:10<27:27,  2.03it/s]
 69%|██████▊   | 7361/10708 [1:37:11<27:27,  2.03it/s]
 69%|██████▉   | 7362/10708 [1:37:11<27:28,  2.03it/s]
 69%|██████▉   | 7363/10708 [1:37:12<27:29,  2.03it/s]
 69%|██████▉   | 7364/10708 [1:37:12<27:27,  2.03it/s]
 69%|██████▉   | 7365/10708 [1:37:13<27:26,  2.03it/s]
 69%|██████▉   | 7366/10708 [1:37:13<27:26,  2.03it/s]
 69%|██████▉   | 7367/10708 [1:37:14<27:27,  2.03it/s]
 69%|██████▉   | 7368/10708 [1:37:14<27:24,  2.03it/s]
 69%|██████▉   | 7369/10708 [1:37:15<27:26,  2.03it/s]
 69%|██████▉   | 7370/10708 [1:37:15<27:24,  2.03it/s]
 69%|██████▉   | 7371/10708 [1:37:16<27:23,  2.03it/s]
 69%|██████▉   | 7372/10708 [1:37:16<27:21,  2.03it/s]
 69%|██████▉   | 7373/10708 [1:37:17<27:21,  2.03it/s]
 69%|██████▉   | 7374/10708 [1:37:17<27:20,  2.03it/s]
 69%|██████▉   | 7375/10708 [1:37:18<27:21,  2.03it/s]{'loss': 3.5782, 'grad_norm': 0.19705772399902344, 'learning_rate': 0.00026722219135406325, 'epoch': 0.69}
                                                      

 69%|██████▉   | 7375/10708 [1:37:18<27:21,  2.03it/s]
 69%|██████▉   | 7376/10708 [1:37:18<27:23,  2.03it/s]
 69%|██████▉   | 7377/10708 [1:37:19<27:21,  2.03it/s]
 69%|██████▉   | 7378/10708 [1:37:19<27:22,  2.03it/s]
 69%|██████▉   | 7379/10708 [1:37:20<27:20,  2.03it/s]
 69%|██████▉   | 7380/10708 [1:37:20<27:20,  2.03it/s]
 69%|██████▉   | 7381/10708 [1:37:21<27:19,  2.03it/s]
 69%|██████▉   | 7382/10708 [1:37:21<27:16,  2.03it/s]
 69%|██████▉   | 7383/10708 [1:37:22<27:19,  2.03it/s]
 69%|██████▉   | 7384/10708 [1:37:22<27:16,  2.03it/s]
 69%|██████▉   | 7385/10708 [1:37:23<27:18,  2.03it/s]
 69%|██████▉   | 7386/10708 [1:37:23<27:16,  2.03it/s]
 69%|██████▉   | 7387/10708 [1:37:24<27:12,  2.03it/s]
 69%|██████▉   | 7388/10708 [1:37:24<27:15,  2.03it/s]
 69%|██████▉   | 7389/10708 [1:37:25<27:12,  2.03it/s]
 69%|██████▉   | 7390/10708 [1:37:25<27:10,  2.03it/s]
 69%|██████▉   | 7391/10708 [1:37:26<27:13,  2.03it/s]
 69%|██████▉   | 7392/10708 [1:37:26<27:12,  2.03it/s]
 69%|██████▉   | 7393/10708 [1:37:27<27:12,  2.03it/s]
 69%|██████▉   | 7394/10708 [1:37:27<27:12,  2.03it/s]
 69%|██████▉   | 7395/10708 [1:37:28<27:10,  2.03it/s]
 69%|██████▉   | 7396/10708 [1:37:28<27:10,  2.03it/s]
 69%|██████▉   | 7397/10708 [1:37:29<27:08,  2.03it/s]
 69%|██████▉   | 7398/10708 [1:37:29<27:09,  2.03it/s]
 69%|██████▉   | 7399/10708 [1:37:29<27:08,  2.03it/s]
 69%|██████▉   | 7400/10708 [1:37:30<27:09,  2.03it/s]
                                                      
{'loss': 3.5757, 'grad_norm': 0.19908393919467926, 'learning_rate': 0.00026362359044161454, 'epoch': 0.69}

 69%|██████▉   | 7400/10708 [1:37:30<27:09,  2.03it/s]
 69%|██████▉   | 7401/10708 [1:37:30<27:10,  2.03it/s]
 69%|██████▉   | 7402/10708 [1:37:31<27:08,  2.03it/s]
 69%|██████▉   | 7403/10708 [1:37:31<27:09,  2.03it/s]
 69%|██████▉   | 7404/10708 [1:37:32<27:08,  2.03it/s]
 69%|██████▉   | 7405/10708 [1:37:32<27:06,  2.03it/s]
 69%|██████▉   | 7406/10708 [1:37:33<31:39,  1.74it/s]
 69%|██████▉   | 7407/10708 [1:37:34<30:18,  1.82it/s]
 69%|██████▉   | 7408/10708 [1:37:34<29:17,  1.88it/s]
 69%|██████▉   | 7409/10708 [1:37:35<28:37,  1.92it/s]
 69%|██████▉   | 7410/10708 [1:37:35<28:07,  1.95it/s]
 69%|██████▉   | 7411/10708 [1:37:36<27:47,  1.98it/s]
 69%|██████▉   | 7412/10708 [1:37:36<27:32,  1.99it/s]
 69%|██████▉   | 7413/10708 [1:37:37<27:22,  2.01it/s]
 69%|██████▉   | 7414/10708 [1:37:37<27:15,  2.01it/s]
 69%|██████▉   | 7415/10708 [1:37:38<27:09,  2.02it/s]
 69%|██████▉   | 7416/10708 [1:37:38<27:05,  2.02it/s]
 69%|██████▉   | 7417/10708 [1:37:39<27:03,  2.03it/s]
 69%|██████▉   | 7418/10708 [1:37:39<27:01,  2.03it/s]
 69%|██████▉   | 7419/10708 [1:37:40<27:02,  2.03it/s]
 69%|██████▉   | 7420/10708 [1:37:40<26:57,  2.03it/s]
 69%|██████▉   | 7421/10708 [1:37:41<26:58,  2.03it/s]
 69%|██████▉   | 7422/10708 [1:37:41<26:58,  2.03it/s]
 69%|██████▉   | 7423/10708 [1:37:42<26:56,  2.03it/s]
 69%|██████▉   | 7424/10708 [1:37:42<26:54,  2.03it/s]
 69%|██████▉   | 7425/10708 [1:37:43<26:54,  2.03it/s]
                                                      {'loss': 3.578, 'grad_norm': 0.20535682141780853, 'learning_rate': 0.0002600406894612096, 'epoch': 0.69}

 69%|██████▉   | 7425/10708 [1:37:43<26:54,  2.03it/s]
 69%|██████▉   | 7426/10708 [1:37:43<26:59,  2.03it/s]
 69%|██████▉   | 7427/10708 [1:37:44<26:57,  2.03it/s]
 69%|██████▉   | 7428/10708 [1:37:44<26:56,  2.03it/s]
 69%|██████▉   | 7429/10708 [1:37:45<26:55,  2.03it/s]
 69%|██████▉   | 7430/10708 [1:37:45<26:53,  2.03it/s]
 69%|██████▉   | 7431/10708 [1:37:46<26:53,  2.03it/s]
 69%|██████▉   | 7432/10708 [1:37:46<26:54,  2.03it/s]
 69%|██████▉   | 7433/10708 [1:37:47<26:50,  2.03it/s]
 69%|██████▉   | 7434/10708 [1:37:47<26:50,  2.03it/s]
 69%|██████▉   | 7435/10708 [1:37:47<26:49,  2.03it/s]
 69%|██████▉   | 7436/10708 [1:37:48<26:46,  2.04it/s]
 69%|██████▉   | 7437/10708 [1:37:49<31:28,  1.73it/s]
 69%|██████▉   | 7438/10708 [1:37:49<30:03,  1.81it/s]
 69%|██████▉   | 7439/10708 [1:37:50<29:03,  1.87it/s]
 69%|██████▉   | 7440/10708 [1:37:50<28:23,  1.92it/s]
 69%|██████▉   | 7441/10708 [1:37:51<27:53,  1.95it/s]
 69%|██████▉   | 7442/10708 [1:37:51<27:31,  1.98it/s]
 70%|██████▉   | 7443/10708 [1:37:52<27:17,  1.99it/s]
 70%|██████▉   | 7444/10708 [1:37:52<27:07,  2.01it/s]
 70%|██████▉   | 7445/10708 [1:37:53<27:02,  2.01it/s]
 70%|██████▉   | 7446/10708 [1:37:53<26:57,  2.02it/s]
 70%|██████▉   | 7447/10708 [1:37:54<26:51,  2.02it/s]
 70%|██████▉   | 7448/10708 [1:37:54<26:48,  2.03it/s]
 70%|██████▉   | 7449/10708 [1:37:55<26:46,  2.03it/s]
 70%|██████▉   | 7450/10708 [1:37:55<26:43,  2.03it/s]
                                                      
{'loss': 3.5741, 'grad_norm': 0.20758800208568573, 'learning_rate': 0.0002564737263862661, 'epoch': 0.7}

 70%|██████▉   | 7450/10708 [1:37:55<26:43,  2.03it/s]
 70%|██████▉   | 7451/10708 [1:37:56<26:44,  2.03it/s]
 70%|██████▉   | 7452/10708 [1:37:56<26:44,  2.03it/s]
 70%|██████▉   | 7453/10708 [1:37:57<26:41,  2.03it/s]
 70%|██████▉   | 7454/10708 [1:37:57<26:40,  2.03it/s]
 70%|██████▉   | 7455/10708 [1:37:58<26:40,  2.03it/s]
 70%|██████▉   | 7456/10708 [1:37:58<26:38,  2.03it/s]
 70%|██████▉   | 7457/10708 [1:37:59<26:38,  2.03it/s]
 70%|██████▉   | 7458/10708 [1:37:59<26:39,  2.03it/s]
 70%|██████▉   | 7459/10708 [1:38:00<26:37,  2.03it/s]
 70%|██████▉   | 7460/10708 [1:38:00<26:36,  2.03it/s]
 70%|██████▉   | 7461/10708 [1:38:01<26:36,  2.03it/s]
 70%|██████▉   | 7462/10708 [1:38:01<26:35,  2.03it/s]
 70%|██████▉   | 7463/10708 [1:38:02<26:34,  2.03it/s]
 70%|██████▉   | 7464/10708 [1:38:02<26:33,  2.04it/s]
 70%|██████▉   | 7465/10708 [1:38:03<26:33,  2.04it/s]
 70%|██████▉   | 7466/10708 [1:38:03<26:34,  2.03it/s]
 70%|██████▉   | 7467/10708 [1:38:04<26:32,  2.03it/s]
 70%|██████▉   | 7468/10708 [1:38:04<26:32,  2.03it/s]
 70%|██████▉   | 7469/10708 [1:38:04<26:32,  2.03it/s]
 70%|██████▉   | 7470/10708 [1:38:05<26:31,  2.04it/s]
 70%|██████▉   | 7471/10708 [1:38:05<26:29,  2.04it/s]
 70%|██████▉   | 7472/10708 [1:38:06<26:28,  2.04it/s]
 70%|██████▉   | 7473/10708 [1:38:06<26:28,  2.04it/s]
 70%|██████▉   | 7474/10708 [1:38:07<26:28,  2.04it/s]
 70%|██████▉   | 7475/10708 [1:38:07<26:28,  2.04it/s]{'loss': 3.5796, 'grad_norm': 0.2186162918806076, 'learning_rate': 0.00025292293813161773, 'epoch': 0.7}
                                                      

 70%|██████▉   | 7475/10708 [1:38:07<26:28,  2.04it/s]
 70%|██████▉   | 7476/10708 [1:38:08<26:31,  2.03it/s]
 70%|██████▉   | 7477/10708 [1:38:08<26:28,  2.03it/s]
 70%|██████▉   | 7478/10708 [1:38:09<26:30,  2.03it/s]
 70%|██████▉   | 7479/10708 [1:38:09<26:28,  2.03it/s]
 70%|██████▉   | 7480/10708 [1:38:10<26:30,  2.03it/s]
 70%|██████▉   | 7481/10708 [1:38:10<26:30,  2.03it/s]
 70%|██████▉   | 7482/10708 [1:38:11<26:27,  2.03it/s]
 70%|██████▉   | 7483/10708 [1:38:11<26:29,  2.03it/s]
 70%|██████▉   | 7484/10708 [1:38:12<26:26,  2.03it/s]
 70%|██████▉   | 7485/10708 [1:38:12<26:25,  2.03it/s]
 70%|██████▉   | 7486/10708 [1:38:13<26:28,  2.03it/s]
 70%|██████▉   | 7487/10708 [1:38:13<26:26,  2.03it/s]
 70%|██████▉   | 7488/10708 [1:38:14<26:26,  2.03it/s]
 70%|██████▉   | 7489/10708 [1:38:14<26:25,  2.03it/s]
 70%|██████▉   | 7490/10708 [1:38:15<26:24,  2.03it/s]
 70%|██████▉   | 7491/10708 [1:38:15<26:23,  2.03it/s]
 70%|██████▉   | 7492/10708 [1:38:16<26:21,  2.03it/s]
 70%|██████▉   | 7493/10708 [1:38:16<26:20,  2.03it/s]
 70%|██████▉   | 7494/10708 [1:38:17<26:18,  2.04it/s]
 70%|██████▉   | 7495/10708 [1:38:17<26:18,  2.03it/s]
 70%|███████   | 7496/10708 [1:38:18<26:19,  2.03it/s]
 70%|███████   | 7497/10708 [1:38:18<26:16,  2.04it/s]
 70%|███████   | 7498/10708 [1:38:19<26:17,  2.04it/s]
 70%|███████   | 7499/10708 [1:38:19<26:18,  2.03it/s]
 70%|███████   | 7500/10708 [1:38:20<26:16,  2.03it/s]
                                                      
{'loss': 3.5775, 'grad_norm': 0.2034405767917633, 'learning_rate': 0.0002493885605377802, 'epoch': 0.7}

 70%|███████   | 7500/10708 [1:38:20<26:16,  2.03it/s]
 70%|███████   | 7501/10708 [1:38:20<26:19,  2.03it/s]
 70%|███████   | 7502/10708 [1:38:21<26:19,  2.03it/s]
 70%|███████   | 7503/10708 [1:38:21<26:17,  2.03it/s]
 70%|███████   | 7504/10708 [1:38:22<26:17,  2.03it/s]
 70%|███████   | 7505/10708 [1:38:22<26:17,  2.03it/s]
 70%|███████   | 7506/10708 [1:38:23<26:17,  2.03it/s]
 70%|███████   | 7507/10708 [1:38:23<26:17,  2.03it/s]
 70%|███████   | 7508/10708 [1:38:24<26:16,  2.03it/s]
 70%|███████   | 7509/10708 [1:38:24<26:15,  2.03it/s]
 70%|███████   | 7510/10708 [1:38:25<26:15,  2.03it/s]
 70%|███████   | 7511/10708 [1:38:25<26:12,  2.03it/s]
 70%|███████   | 7512/10708 [1:38:26<26:11,  2.03it/s]
 70%|███████   | 7513/10708 [1:38:26<26:12,  2.03it/s]
 70%|███████   | 7514/10708 [1:38:27<26:10,  2.03it/s]
 70%|███████   | 7515/10708 [1:38:27<26:08,  2.04it/s]
 70%|███████   | 7516/10708 [1:38:28<26:08,  2.03it/s]
 70%|███████   | 7517/10708 [1:38:28<26:07,  2.04it/s]
 70%|███████   | 7518/10708 [1:38:29<26:06,  2.04it/s]
 70%|███████   | 7519/10708 [1:38:29<26:06,  2.04it/s]
 70%|███████   | 7520/10708 [1:38:30<26:05,  2.04it/s]
 70%|███████   | 7521/10708 [1:38:30<26:04,  2.04it/s]
 70%|███████   | 7522/10708 [1:38:31<26:04,  2.04it/s]
 70%|███████   | 7523/10708 [1:38:31<26:04,  2.04it/s]
 70%|███████   | 7524/10708 [1:38:32<26:03,  2.04it/s]
 70%|███████   | 7525/10708 [1:38:32<26:05,  2.03it/s]
                                                      
{'loss': 3.5754, 'grad_norm': 0.19979241490364075, 'learning_rate': 0.00024587082835528647, 'epoch': 0.7}

 70%|███████   | 7525/10708 [1:38:32<26:05,  2.03it/s]
 70%|███████   | 7526/10708 [1:38:33<26:05,  2.03it/s]
 70%|███████   | 7527/10708 [1:38:33<26:05,  2.03it/s]
 70%|███████   | 7528/10708 [1:38:34<26:05,  2.03it/s]
 70%|███████   | 7529/10708 [1:38:34<26:03,  2.03it/s]
 70%|███████   | 7530/10708 [1:38:34<26:01,  2.03it/s]
 70%|███████   | 7531/10708 [1:38:35<26:00,  2.04it/s]
 70%|███████   | 7532/10708 [1:38:35<25:59,  2.04it/s]
 70%|███████   | 7533/10708 [1:38:36<25:58,  2.04it/s]
 70%|███████   | 7534/10708 [1:38:36<25:59,  2.04it/s]
 70%|███████   | 7535/10708 [1:38:37<26:00,  2.03it/s]
 70%|███████   | 7536/10708 [1:38:37<25:59,  2.03it/s]
 70%|███████   | 7537/10708 [1:38:38<25:59,  2.03it/s]
 70%|███████   | 7538/10708 [1:38:38<25:57,  2.03it/s]
 70%|███████   | 7539/10708 [1:38:39<26:00,  2.03it/s]
 70%|███████   | 7540/10708 [1:38:39<25:58,  2.03it/s]
 70%|███████   | 7541/10708 [1:38:40<25:56,  2.03it/s]
 70%|███████   | 7542/10708 [1:38:40<25:59,  2.03it/s]
 70%|███████   | 7543/10708 [1:38:41<25:58,  2.03it/s]
 70%|███████   | 7544/10708 [1:38:41<25:59,  2.03it/s]
 70%|███████   | 7545/10708 [1:38:42<25:58,  2.03it/s]
 70%|███████   | 7546/10708 [1:38:42<25:55,  2.03it/s]
 70%|███████   | 7547/10708 [1:38:43<25:59,  2.03it/s]
 70%|███████   | 7548/10708 [1:38:43<25:59,  2.03it/s]
 70%|███████   | 7549/10708 [1:38:44<25:58,  2.03it/s]
 71%|███████   | 7550/10708 [1:38:44<25:55,  2.03it/s]
                                                      
{'loss': 3.5803, 'grad_norm': 0.20403750240802765, 'learning_rate': 0.0002423699752290938, 'epoch': 0.71}

 71%|███████   | 7550/10708 [1:38:44<25:55,  2.03it/s]
 71%|███████   | 7551/10708 [1:38:45<25:53,  2.03it/s]
 71%|███████   | 7552/10708 [1:38:45<25:53,  2.03it/s]
 71%|███████   | 7553/10708 [1:38:46<25:52,  2.03it/s]
 71%|███████   | 7554/10708 [1:38:46<25:49,  2.04it/s]
 71%|███████   | 7555/10708 [1:38:47<25:50,  2.03it/s]
 71%|███████   | 7556/10708 [1:38:47<25:49,  2.03it/s]
 71%|███████   | 7557/10708 [1:38:48<25:48,  2.04it/s]
 71%|███████   | 7558/10708 [1:38:48<25:49,  2.03it/s]
 71%|███████   | 7559/10708 [1:38:49<25:49,  2.03it/s]
 71%|███████   | 7560/10708 [1:38:49<25:50,  2.03it/s]
 71%|███████   | 7561/10708 [1:38:50<25:49,  2.03it/s]
 71%|███████   | 7562/10708 [1:38:50<25:47,  2.03it/s]
 71%|███████   | 7563/10708 [1:38:51<25:46,  2.03it/s]
 71%|███████   | 7564/10708 [1:38:51<25:44,  2.04it/s]
 71%|███████   | 7565/10708 [1:38:52<25:44,  2.04it/s]
 71%|███████   | 7566/10708 [1:38:52<25:43,  2.04it/s]
 71%|███████   | 7567/10708 [1:38:53<25:44,  2.03it/s]
 71%|███████   | 7568/10708 [1:38:53<25:47,  2.03it/s]
 71%|███████   | 7569/10708 [1:38:54<25:44,  2.03it/s]
 71%|███████   | 7570/10708 [1:38:54<25:41,  2.04it/s]
 71%|███████   | 7571/10708 [1:38:55<25:43,  2.03it/s]
 71%|███████   | 7572/10708 [1:38:55<25:41,  2.03it/s]
 71%|███████   | 7573/10708 [1:38:56<25:40,  2.04it/s]
 71%|███████   | 7574/10708 [1:38:56<25:39,  2.04it/s]
 71%|███████   | 7575/10708 [1:38:57<25:41,  2.03it/s]
                                                      
{'loss': 3.5761, 'grad_norm': 0.20690347254276276, 'learning_rate': 0.00023888623368306688, 'epoch': 0.71}

 71%|███████   | 7575/10708 [1:38:57<25:41,  2.03it/s]
 71%|███████   | 7576/10708 [1:38:57<25:43,  2.03it/s]
 71%|███████   | 7577/10708 [1:38:58<25:41,  2.03it/s]
 71%|███████   | 7578/10708 [1:38:58<25:39,  2.03it/s]
 71%|███████   | 7579/10708 [1:38:59<25:37,  2.04it/s]
 71%|███████   | 7580/10708 [1:38:59<25:36,  2.04it/s]
 71%|███████   | 7581/10708 [1:39:00<25:38,  2.03it/s]
 71%|███████   | 7582/10708 [1:39:00<25:35,  2.04it/s]
 71%|███████   | 7583/10708 [1:39:01<25:36,  2.03it/s]
 71%|███████   | 7584/10708 [1:39:01<25:37,  2.03it/s]
 71%|███████   | 7585/10708 [1:39:02<25:35,  2.03it/s]
 71%|███████   | 7586/10708 [1:39:02<25:35,  2.03it/s]
 71%|███████   | 7587/10708 [1:39:03<25:34,  2.03it/s]
 71%|███████   | 7588/10708 [1:39:03<25:32,  2.04it/s]
 71%|███████   | 7589/10708 [1:39:04<25:32,  2.03it/s]
 71%|███████   | 7590/10708 [1:39:04<25:33,  2.03it/s]
 71%|███████   | 7591/10708 [1:39:04<25:33,  2.03it/s]
 71%|███████   | 7592/10708 [1:39:05<25:31,  2.03it/s]
 71%|███████   | 7593/10708 [1:39:05<25:32,  2.03it/s]
 71%|███████   | 7594/10708 [1:39:06<25:31,  2.03it/s]
 71%|███████   | 7595/10708 [1:39:06<25:30,  2.03it/s]
 71%|███████   | 7596/10708 [1:39:07<25:33,  2.03it/s]
 71%|███████   | 7597/10708 [1:39:07<25:31,  2.03it/s]
 71%|███████   | 7598/10708 [1:39:08<25:31,  2.03it/s]
 71%|███████   | 7599/10708 [1:39:08<25:29,  2.03it/s]
 71%|███████   | 7600/10708 [1:39:09<25:27,  2.03it/s]{'loss': 3.5701, 'grad_norm': 0.21807318925857544, 'learning_rate': 0.00023541983510453208, 'epoch': 0.71}

                                                      

 71%|███████   | 7600/10708 [1:39:09<25:27,  2.03it/s]
 71%|███████   | 7601/10708 [1:39:09<25:34,  2.02it/s]
 71%|███████   | 7602/10708 [1:39:10<25:31,  2.03it/s]
 71%|███████   | 7603/10708 [1:39:10<25:31,  2.03it/s]
 71%|███████   | 7604/10708 [1:39:11<25:28,  2.03it/s]
 71%|███████   | 7605/10708 [1:39:11<25:28,  2.03it/s]
 71%|███████   | 7606/10708 [1:39:12<25:27,  2.03it/s]
 71%|███████   | 7607/10708 [1:39:12<25:24,  2.03it/s]
 71%|███████   | 7608/10708 [1:39:13<25:29,  2.03it/s]
 71%|███████   | 7609/10708 [1:39:13<25:28,  2.03it/s]
 71%|███████   | 7610/10708 [1:39:14<25:26,  2.03it/s]
 71%|███████   | 7611/10708 [1:39:14<25:26,  2.03it/s]
 71%|███████   | 7612/10708 [1:39:15<25:22,  2.03it/s]
 71%|███████   | 7613/10708 [1:39:15<25:24,  2.03it/s]
 71%|███████   | 7614/10708 [1:39:16<25:22,  2.03it/s]
 71%|███████   | 7615/10708 [1:39:16<25:22,  2.03it/s]
 71%|███████   | 7616/10708 [1:39:17<25:23,  2.03it/s]
 71%|███████   | 7617/10708 [1:39:17<25:22,  2.03it/s]
 71%|███████   | 7618/10708 [1:39:18<25:21,  2.03it/s]
 71%|███████   | 7619/10708 [1:39:18<25:19,  2.03it/s]
 71%|███████   | 7620/10708 [1:39:19<25:17,  2.03it/s]
 71%|███████   | 7621/10708 [1:39:19<25:15,  2.04it/s]
 71%|███████   | 7622/10708 [1:39:20<25:17,  2.03it/s]
 71%|███████   | 7623/10708 [1:39:20<25:17,  2.03it/s]
 71%|███████   | 7624/10708 [1:39:21<25:16,  2.03it/s]
 71%|███████   | 7625/10708 [1:39:21<25:17,  2.03it/s]{'loss': 3.5759, 'grad_norm': 0.20434308052062988, 'learning_rate': 0.00023197100972891066, 'epoch': 0.71}

                                                      

 71%|███████   | 7625/10708 [1:39:21<25:17,  2.03it/s]
 71%|███████   | 7626/10708 [1:39:22<25:17,  2.03it/s]
 71%|███████   | 7627/10708 [1:39:22<25:15,  2.03it/s]
 71%|███████   | 7628/10708 [1:39:23<25:15,  2.03it/s]
 71%|███████   | 7629/10708 [1:39:23<25:15,  2.03it/s]
 71%|███████▏  | 7630/10708 [1:39:24<25:14,  2.03it/s]
 71%|███████▏  | 7631/10708 [1:39:24<25:15,  2.03it/s]
 71%|███████▏  | 7632/10708 [1:39:25<25:12,  2.03it/s]
 71%|███████▏  | 7633/10708 [1:39:25<25:11,  2.03it/s]
 71%|███████▏  | 7634/10708 [1:39:26<25:11,  2.03it/s]
 71%|███████▏  | 7635/10708 [1:39:26<25:12,  2.03it/s]
 71%|███████▏  | 7636/10708 [1:39:27<25:11,  2.03it/s]
 71%|███████▏  | 7637/10708 [1:39:27<25:12,  2.03it/s]
 71%|███████▏  | 7638/10708 [1:39:28<25:10,  2.03it/s]
 71%|███████▏  | 7639/10708 [1:39:28<25:08,  2.03it/s]
 71%|███████▏  | 7640/10708 [1:39:29<25:06,  2.04it/s]
 71%|███████▏  | 7641/10708 [1:39:29<25:07,  2.03it/s]
 71%|███████▏  | 7642/10708 [1:39:30<25:06,  2.04it/s]
 71%|███████▏  | 7643/10708 [1:39:30<25:06,  2.03it/s]
 71%|███████▏  | 7644/10708 [1:39:31<25:05,  2.04it/s]
 71%|███████▏  | 7645/10708 [1:39:31<25:05,  2.03it/s]
 71%|███████▏  | 7646/10708 [1:39:32<25:06,  2.03it/s]
 71%|███████▏  | 7647/10708 [1:39:32<25:05,  2.03it/s]
 71%|███████▏  | 7648/10708 [1:39:33<25:06,  2.03it/s]
 71%|███████▏  | 7649/10708 [1:39:33<25:02,  2.04it/s]
 71%|███████▏  | 7650/10708 [1:39:34<25:00,  2.04it/s]{'loss': 3.5734, 'grad_norm': 0.206534743309021, 'learning_rate': 0.00022853998662442494, 'epoch': 0.71}

                                                      

 71%|███████▏  | 7650/10708 [1:39:34<25:00,  2.04it/s]
 71%|███████▏  | 7651/10708 [1:39:34<25:02,  2.03it/s]
 71%|███████▏  | 7652/10708 [1:39:35<25:01,  2.04it/s]
 71%|███████▏  | 7653/10708 [1:39:35<25:00,  2.04it/s]
 71%|███████▏  | 7654/10708 [1:39:35<25:00,  2.03it/s]
 71%|███████▏  | 7655/10708 [1:39:36<24:59,  2.04it/s]
 71%|███████▏  | 7656/10708 [1:39:36<24:58,  2.04it/s]
 72%|███████▏  | 7657/10708 [1:39:37<25:00,  2.03it/s]
 72%|███████▏  | 7658/10708 [1:39:37<24:58,  2.03it/s]
 72%|███████▏  | 7659/10708 [1:39:38<24:57,  2.04it/s]
 72%|███████▏  | 7660/10708 [1:39:38<24:57,  2.04it/s]
 72%|███████▏  | 7661/10708 [1:39:39<24:57,  2.04it/s]
 72%|███████▏  | 7662/10708 [1:39:39<24:55,  2.04it/s]
 72%|███████▏  | 7663/10708 [1:39:40<24:55,  2.04it/s]
 72%|███████▏  | 7664/10708 [1:39:40<24:54,  2.04it/s]
 72%|███████▏  | 7665/10708 [1:39:41<24:54,  2.04it/s]
 72%|███████▏  | 7666/10708 [1:39:41<24:55,  2.03it/s]
 72%|███████▏  | 7667/10708 [1:39:42<24:54,  2.03it/s]
 72%|███████▏  | 7668/10708 [1:39:42<24:53,  2.04it/s]
 72%|███████▏  | 7669/10708 [1:39:43<24:58,  2.03it/s]
 72%|███████▏  | 7670/10708 [1:39:43<24:56,  2.03it/s]
 72%|███████▏  | 7671/10708 [1:39:44<24:56,  2.03it/s]
 72%|███████▏  | 7672/10708 [1:39:44<24:54,  2.03it/s]
 72%|███████▏  | 7673/10708 [1:39:45<24:52,  2.03it/s]
 72%|███████▏  | 7674/10708 [1:39:45<24:51,  2.03it/s]
 72%|███████▏  | 7675/10708 [1:39:46<24:51,  2.03it/s]{'loss': 3.5745, 'grad_norm': 0.1973312795162201, 'learning_rate': 0.00022512699367688532, 'epoch': 0.72}

                                                      

 72%|███████▏  | 7675/10708 [1:39:46<24:51,  2.03it/s]
 72%|███████▏  | 7676/10708 [1:39:46<24:55,  2.03it/s]
 72%|███████▏  | 7677/10708 [1:39:47<24:53,  2.03it/s]
 72%|███████▏  | 7678/10708 [1:39:47<24:51,  2.03it/s]
 72%|███████▏  | 7679/10708 [1:39:48<24:50,  2.03it/s]
 72%|███████▏  | 7680/10708 [1:39:48<24:50,  2.03it/s]
 72%|███████▏  | 7681/10708 [1:39:49<24:48,  2.03it/s]
 72%|███████▏  | 7682/10708 [1:39:49<24:51,  2.03it/s]
 72%|███████▏  | 7683/10708 [1:39:50<24:48,  2.03it/s]
 72%|███████▏  | 7684/10708 [1:39:50<24:50,  2.03it/s]
 72%|███████▏  | 7685/10708 [1:39:51<24:48,  2.03it/s]
 72%|███████▏  | 7686/10708 [1:39:51<24:49,  2.03it/s]
 72%|███████▏  | 7687/10708 [1:39:52<24:51,  2.03it/s]
 72%|███████▏  | 7688/10708 [1:39:52<24:48,  2.03it/s]
 72%|███████▏  | 7689/10708 [1:39:53<24:48,  2.03it/s]
 72%|███████▏  | 7690/10708 [1:39:53<24:46,  2.03it/s]
 72%|███████▏  | 7691/10708 [1:39:54<24:47,  2.03it/s]
 72%|███████▏  | 7692/10708 [1:39:54<24:46,  2.03it/s]
 72%|███████▏  | 7693/10708 [1:39:55<24:44,  2.03it/s]
 72%|███████▏  | 7694/10708 [1:39:55<24:43,  2.03it/s]
 72%|███████▏  | 7695/10708 [1:39:56<24:43,  2.03it/s]
 72%|███████▏  | 7696/10708 [1:39:56<24:40,  2.03it/s]
 72%|███████▏  | 7697/10708 [1:39:57<24:42,  2.03it/s]
 72%|███████▏  | 7698/10708 [1:39:57<24:41,  2.03it/s]
 72%|███████▏  | 7699/10708 [1:39:58<24:40,  2.03it/s]
 72%|███████▏  | 7700/10708 [1:39:58<24:39,  2.03it/s]{'loss': 3.5648, 'grad_norm': 0.20736274123191833, 'learning_rate': 0.00022173225757455306, 'epoch': 0.72}
                                                      

 72%|███████▏  | 7700/10708 [1:39:58<24:39,  2.03it/s]
 72%|███████▏  | 7701/10708 [1:39:59<24:42,  2.03it/s]
 72%|███████▏  | 7702/10708 [1:39:59<24:40,  2.03it/s]
 72%|███████▏  | 7703/10708 [1:40:00<24:39,  2.03it/s]
 72%|███████▏  | 7704/10708 [1:40:00<24:37,  2.03it/s]
 72%|███████▏  | 7705/10708 [1:40:01<24:36,  2.03it/s]
 72%|███████▏  | 7706/10708 [1:40:01<24:35,  2.03it/s]
 72%|███████▏  | 7707/10708 [1:40:02<24:38,  2.03it/s]
 72%|███████▏  | 7708/10708 [1:40:02<24:38,  2.03it/s]
 72%|███████▏  | 7709/10708 [1:40:03<24:37,  2.03it/s]
 72%|███████▏  | 7710/10708 [1:40:03<24:34,  2.03it/s]
 72%|███████▏  | 7711/10708 [1:40:04<24:34,  2.03it/s]
 72%|███████▏  | 7712/10708 [1:40:04<24:33,  2.03it/s]
 72%|███████▏  | 7713/10708 [1:40:05<24:30,  2.04it/s]
 72%|███████▏  | 7714/10708 [1:40:05<24:31,  2.04it/s]
 72%|███████▏  | 7715/10708 [1:40:06<24:32,  2.03it/s]
 72%|███████▏  | 7716/10708 [1:40:06<24:29,  2.04it/s]
 72%|███████▏  | 7717/10708 [1:40:06<24:34,  2.03it/s]
 72%|███████▏  | 7718/10708 [1:40:07<24:31,  2.03it/s]
 72%|███████▏  | 7719/10708 [1:40:07<24:29,  2.03it/s]
 72%|███████▏  | 7720/10708 [1:40:08<24:30,  2.03it/s]
 72%|███████▏  | 7721/10708 [1:40:08<24:28,  2.03it/s]
 72%|███████▏  | 7722/10708 [1:40:09<24:26,  2.04it/s]
 72%|███████▏  | 7723/10708 [1:40:09<24:29,  2.03it/s]
 72%|███████▏  | 7724/10708 [1:40:10<24:27,  2.03it/s]
 72%|███████▏  | 7725/10708 [1:40:10<24:26,  2.03it/s]{'loss': 3.5707, 'grad_norm': 0.1961323320865631, 'learning_rate': 0.00021835600379308467, 'epoch': 0.72}

                                                      

 72%|███████▏  | 7725/10708 [1:40:10<24:26,  2.03it/s]
 72%|███████▏  | 7726/10708 [1:40:11<24:28,  2.03it/s]
 72%|███████▏  | 7727/10708 [1:40:11<24:26,  2.03it/s]
 72%|███████▏  | 7728/10708 [1:40:12<24:26,  2.03it/s]
 72%|███████▏  | 7729/10708 [1:40:12<24:26,  2.03it/s]
 72%|███████▏  | 7730/10708 [1:40:13<24:26,  2.03it/s]
 72%|███████▏  | 7731/10708 [1:40:13<24:28,  2.03it/s]
 72%|███████▏  | 7732/10708 [1:40:14<24:25,  2.03it/s]
 72%|███████▏  | 7733/10708 [1:40:14<24:26,  2.03it/s]
 72%|███████▏  | 7734/10708 [1:40:15<24:24,  2.03it/s]
 72%|███████▏  | 7735/10708 [1:40:15<24:23,  2.03it/s]
 72%|███████▏  | 7736/10708 [1:40:16<24:22,  2.03it/s]
 72%|███████▏  | 7737/10708 [1:40:16<24:22,  2.03it/s]
 72%|███████▏  | 7738/10708 [1:40:17<24:21,  2.03it/s]
 72%|███████▏  | 7739/10708 [1:40:17<24:18,  2.04it/s]
 72%|███████▏  | 7740/10708 [1:40:18<24:19,  2.03it/s]
 72%|███████▏  | 7741/10708 [1:40:18<24:19,  2.03it/s]
 72%|███████▏  | 7742/10708 [1:40:19<24:20,  2.03it/s]
 72%|███████▏  | 7743/10708 [1:40:19<24:21,  2.03it/s]
 72%|███████▏  | 7744/10708 [1:40:20<24:18,  2.03it/s]
 72%|███████▏  | 7745/10708 [1:40:20<24:17,  2.03it/s]
 72%|███████▏  | 7746/10708 [1:40:21<24:16,  2.03it/s]
 72%|███████▏  | 7747/10708 [1:40:21<24:15,  2.03it/s]
 72%|███████▏  | 7748/10708 [1:40:22<24:15,  2.03it/s]
 72%|███████▏  | 7749/10708 [1:40:22<24:13,  2.04it/s]
 72%|███████▏  | 7750/10708 [1:40:23<24:12,  2.04it/s]
                                                      {'loss': 3.5703, 'grad_norm': 0.2021539807319641, 'learning_rate': 0.00021499845658055572, 'epoch': 0.72}

 72%|███████▏  | 7750/10708 [1:40:23<24:12,  2.04it/s]
 72%|███████▏  | 7751/10708 [1:40:23<24:15,  2.03it/s]
 72%|███████▏  | 7752/10708 [1:40:24<24:14,  2.03it/s]
 72%|███████▏  | 7753/10708 [1:40:24<24:16,  2.03it/s]
 72%|███████▏  | 7754/10708 [1:40:25<24:15,  2.03it/s]
 72%|███████▏  | 7755/10708 [1:40:25<24:14,  2.03it/s]
 72%|███████▏  | 7756/10708 [1:40:26<24:15,  2.03it/s]
 72%|███████▏  | 7757/10708 [1:40:26<24:13,  2.03it/s]
 72%|███████▏  | 7758/10708 [1:40:27<24:12,  2.03it/s]
 72%|███████▏  | 7759/10708 [1:40:27<24:14,  2.03it/s]
 72%|███████▏  | 7760/10708 [1:40:28<24:12,  2.03it/s]
 72%|███████▏  | 7761/10708 [1:40:28<24:11,  2.03it/s]
 72%|███████▏  | 7762/10708 [1:40:29<24:11,  2.03it/s]
 72%|███████▏  | 7763/10708 [1:40:29<24:13,  2.03it/s]
 73%|███████▎  | 7764/10708 [1:40:30<24:10,  2.03it/s]
 73%|███████▎  | 7765/10708 [1:40:30<24:09,  2.03it/s]
 73%|███████▎  | 7766/10708 [1:40:31<24:08,  2.03it/s]
 73%|███████▎  | 7767/10708 [1:40:31<24:08,  2.03it/s]
 73%|███████▎  | 7768/10708 [1:40:32<24:06,  2.03it/s]
 73%|███████▎  | 7769/10708 [1:40:32<24:05,  2.03it/s]
 73%|███████▎  | 7770/10708 [1:40:33<24:05,  2.03it/s]
 73%|███████▎  | 7771/10708 [1:40:33<24:04,  2.03it/s]
 73%|███████▎  | 7772/10708 [1:40:34<24:02,  2.03it/s]
 73%|███████▎  | 7773/10708 [1:40:34<24:03,  2.03it/s]
 73%|███████▎  | 7774/10708 [1:40:35<24:01,  2.04it/s]
 73%|███████▎  | 7775/10708 [1:40:35<24:02,  2.03it/s]{'loss': 3.5652, 'grad_norm': 0.21359223127365112, 'learning_rate': 0.00021165983894256646, 'epoch': 0.73}
                                                      

 73%|███████▎  | 7775/10708 [1:40:35<24:02,  2.03it/s]
 73%|███████▎  | 7776/10708 [1:40:36<24:04,  2.03it/s]
 73%|███████▎  | 7777/10708 [1:40:36<24:03,  2.03it/s]
 73%|███████▎  | 7778/10708 [1:40:37<24:00,  2.03it/s]
 73%|███████▎  | 7779/10708 [1:40:37<23:59,  2.03it/s]
 73%|███████▎  | 7780/10708 [1:40:38<24:00,  2.03it/s]
 73%|███████▎  | 7781/10708 [1:40:38<23:58,  2.03it/s]
 73%|███████▎  | 7782/10708 [1:40:38<23:57,  2.04it/s]
 73%|███████▎  | 7783/10708 [1:40:39<23:58,  2.03it/s]
 73%|███████▎  | 7784/10708 [1:40:39<23:57,  2.03it/s]
 73%|███████▎  | 7785/10708 [1:40:40<23:58,  2.03it/s]
 73%|███████▎  | 7786/10708 [1:40:40<23:57,  2.03it/s]
 73%|███████▎  | 7787/10708 [1:40:41<23:57,  2.03it/s]
 73%|███████▎  | 7788/10708 [1:40:41<23:54,  2.04it/s]
 73%|███████▎  | 7789/10708 [1:40:42<23:57,  2.03it/s]
 73%|███████▎  | 7790/10708 [1:40:42<23:55,  2.03it/s]
 73%|███████▎  | 7791/10708 [1:40:43<24:00,  2.02it/s]
 73%|███████▎  | 7792/10708 [1:40:43<24:03,  2.02it/s]
 73%|███████▎  | 7793/10708 [1:40:44<24:02,  2.02it/s]
 73%|███████▎  | 7794/10708 [1:40:44<23:59,  2.02it/s]
 73%|███████▎  | 7795/10708 [1:40:45<23:58,  2.03it/s]
 73%|███████▎  | 7796/10708 [1:40:45<23:57,  2.03it/s]
 73%|███████▎  | 7797/10708 [1:40:46<23:56,  2.03it/s]
 73%|███████▎  | 7798/10708 [1:40:46<23:53,  2.03it/s]
 73%|███████▎  | 7799/10708 [1:40:47<23:53,  2.03it/s]
 73%|███████▎  | 7800/10708 [1:40:47<23:52,  2.03it/s]
                                                      {'loss': 3.5637, 'grad_norm': 0.20908664166927338, 'learning_rate': 0.00020834037262742956, 'epoch': 0.73}


 73%|███████▎  | 7800/10708 [1:40:47<23:52,  2.03it/s]
 73%|███████▎  | 7801/10708 [1:40:48<23:54,  2.03it/s]
 73%|███████▎  | 7802/10708 [1:40:48<23:54,  2.03it/s]
 73%|███████▎  | 7803/10708 [1:40:49<23:52,  2.03it/s]
 73%|███████▎  | 7804/10708 [1:40:49<23:50,  2.03it/s]
 73%|███████▎  | 7805/10708 [1:40:50<23:49,  2.03it/s]
 73%|███████▎  | 7806/10708 [1:40:50<23:48,  2.03it/s]
 73%|███████▎  | 7807/10708 [1:40:51<23:47,  2.03it/s]
 73%|███████▎  | 7808/10708 [1:40:51<23:47,  2.03it/s]
 73%|███████▎  | 7809/10708 [1:40:52<23:46,  2.03it/s]
 73%|███████▎  | 7810/10708 [1:40:52<23:45,  2.03it/s]
 73%|███████▎  | 7811/10708 [1:40:53<23:45,  2.03it/s]
 73%|███████▎  | 7812/10708 [1:40:53<23:42,  2.04it/s]
 73%|███████▎  | 7813/10708 [1:40:54<23:44,  2.03it/s]
 73%|███████▎  | 7814/10708 [1:40:54<23:42,  2.03it/s]
 73%|███████▎  | 7815/10708 [1:40:55<23:41,  2.03it/s]
 73%|███████▎  | 7816/10708 [1:40:55<23:42,  2.03it/s]
 73%|███████▎  | 7817/10708 [1:40:56<23:42,  2.03it/s]
 73%|███████▎  | 7818/10708 [1:40:56<23:41,  2.03it/s]
 73%|███████▎  | 7819/10708 [1:40:57<23:43,  2.03it/s]
 73%|███████▎  | 7820/10708 [1:40:57<23:40,  2.03it/s]
 73%|███████▎  | 7821/10708 [1:40:58<23:42,  2.03it/s]
 73%|███████▎  | 7822/10708 [1:40:58<23:40,  2.03it/s]
 73%|███████▎  | 7823/10708 [1:40:59<23:41,  2.03it/s]
 73%|███████▎  | 7824/10708 [1:40:59<23:40,  2.03it/s]
 73%|███████▎  | 7825/10708 [1:41:00<23:37,  2.03it/s]
                                                      {'loss': 3.5623, 'grad_norm': 0.2024209201335907, 'learning_rate': 0.00020504027811144265, 'epoch': 0.73}


 73%|███████▎  | 7825/10708 [1:41:00<23:37,  2.03it/s]
 73%|███████▎  | 7826/10708 [1:41:00<23:41,  2.03it/s]
 73%|███████▎  | 7827/10708 [1:41:01<23:38,  2.03it/s]
 73%|███████▎  | 7828/10708 [1:41:01<23:39,  2.03it/s]
 73%|███████▎  | 7829/10708 [1:41:02<23:39,  2.03it/s]
 73%|███████▎  | 7830/10708 [1:41:02<23:38,  2.03it/s]
 73%|███████▎  | 7831/10708 [1:41:03<23:37,  2.03it/s]
 73%|███████▎  | 7832/10708 [1:41:03<23:35,  2.03it/s]
 73%|███████▎  | 7833/10708 [1:41:04<23:35,  2.03it/s]
 73%|███████▎  | 7834/10708 [1:41:04<23:34,  2.03it/s]
 73%|███████▎  | 7835/10708 [1:41:05<23:31,  2.04it/s]
 73%|███████▎  | 7836/10708 [1:41:05<23:31,  2.03it/s]
 73%|███████▎  | 7837/10708 [1:41:06<23:32,  2.03it/s]
 73%|███████▎  | 7838/10708 [1:41:06<23:31,  2.03it/s]
 73%|███████▎  | 7839/10708 [1:41:07<23:31,  2.03it/s]
 73%|███████▎  | 7840/10708 [1:41:07<23:29,  2.03it/s]
 73%|███████▎  | 7841/10708 [1:41:08<23:28,  2.03it/s]
 73%|███████▎  | 7842/10708 [1:41:08<23:27,  2.04it/s]
 73%|███████▎  | 7843/10708 [1:41:09<23:29,  2.03it/s]
 73%|███████▎  | 7844/10708 [1:41:09<23:28,  2.03it/s]
 73%|███████▎  | 7845/10708 [1:41:10<23:26,  2.04it/s]
 73%|███████▎  | 7846/10708 [1:41:10<23:26,  2.03it/s]
 73%|███████▎  | 7847/10708 [1:41:10<23:26,  2.03it/s]
 73%|███████▎  | 7848/10708 [1:41:11<23:26,  2.03it/s]
 73%|███████▎  | 7849/10708 [1:41:11<23:25,  2.03it/s]
 73%|███████▎  | 7850/10708 [1:41:12<23:24,  2.04it/s]{'loss': 3.5631, 'grad_norm': 0.20639784634113312, 'learning_rate': 0.00020175977458424343, 'epoch': 0.73}
                                                      

 73%|███████▎  | 7850/10708 [1:41:12<23:24,  2.04it/s]
 73%|███████▎  | 7851/10708 [1:41:12<23:27,  2.03it/s]
 73%|███████▎  | 7852/10708 [1:41:13<23:27,  2.03it/s]
 73%|███████▎  | 7853/10708 [1:41:13<23:29,  2.03it/s]
 73%|███████▎  | 7854/10708 [1:41:14<23:27,  2.03it/s]
 73%|███████▎  | 7855/10708 [1:41:14<23:25,  2.03it/s]
 73%|███████▎  | 7856/10708 [1:41:15<23:23,  2.03it/s]
 73%|███████▎  | 7857/10708 [1:41:15<23:25,  2.03it/s]
 73%|███████▎  | 7858/10708 [1:41:16<23:23,  2.03it/s]
 73%|███████▎  | 7859/10708 [1:41:16<23:21,  2.03it/s]
 73%|███████▎  | 7860/10708 [1:41:17<23:23,  2.03it/s]
 73%|███████▎  | 7861/10708 [1:41:17<23:21,  2.03it/s]
 73%|███████▎  | 7862/10708 [1:41:18<23:21,  2.03it/s]
 73%|███████▎  | 7863/10708 [1:41:18<23:23,  2.03it/s]
 73%|███████▎  | 7864/10708 [1:41:19<23:23,  2.03it/s]
 73%|███████▎  | 7865/10708 [1:41:19<23:22,  2.03it/s]
 73%|███████▎  | 7866/10708 [1:41:20<23:20,  2.03it/s]
 73%|███████▎  | 7867/10708 [1:41:20<23:20,  2.03it/s]
 73%|███████▎  | 7868/10708 [1:41:21<23:19,  2.03it/s]
 73%|███████▎  | 7869/10708 [1:41:21<23:17,  2.03it/s]
 73%|███████▎  | 7870/10708 [1:41:22<23:16,  2.03it/s]
 74%|███████▎  | 7871/10708 [1:41:22<23:15,  2.03it/s]
 74%|███████▎  | 7872/10708 [1:41:23<23:13,  2.04it/s]
 74%|███████▎  | 7873/10708 [1:41:23<23:13,  2.03it/s]
 74%|███████▎  | 7874/10708 [1:41:24<23:13,  2.03it/s]
 74%|███████▎  | 7875/10708 [1:41:24<23:11,  2.04it/s]{'loss': 3.5636, 'grad_norm': 0.2069689780473709, 'learning_rate': 0.00019849907993425243, 'epoch': 0.74}

                                                      

 74%|███████▎  | 7875/10708 [1:41:24<23:11,  2.04it/s]
 74%|███████▎  | 7876/10708 [1:41:25<23:15,  2.03it/s]
 74%|███████▎  | 7877/10708 [1:41:25<23:13,  2.03it/s]
 74%|███████▎  | 7878/10708 [1:41:26<23:12,  2.03it/s]
 74%|███████▎  | 7879/10708 [1:41:26<23:12,  2.03it/s]
 74%|███████▎  | 7880/10708 [1:41:27<23:09,  2.03it/s]
 74%|███████▎  | 7881/10708 [1:41:27<23:09,  2.03it/s]
 74%|███████▎  | 7882/10708 [1:41:28<23:09,  2.03it/s]
 74%|███████▎  | 7883/10708 [1:41:28<23:08,  2.03it/s]
 74%|███████▎  | 7884/10708 [1:41:29<23:09,  2.03it/s]
 74%|███████▎  | 7885/10708 [1:41:29<23:08,  2.03it/s]
 74%|███████▎  | 7886/10708 [1:41:30<23:09,  2.03it/s]
 74%|███████▎  | 7887/10708 [1:41:30<23:09,  2.03it/s]
 74%|███████▎  | 7888/10708 [1:41:31<23:07,  2.03it/s]
 74%|███████▎  | 7889/10708 [1:41:31<23:06,  2.03it/s]
 74%|███████▎  | 7890/10708 [1:41:32<23:07,  2.03it/s]
 74%|███████▎  | 7891/10708 [1:41:32<23:04,  2.03it/s]
 74%|███████▎  | 7892/10708 [1:41:33<23:04,  2.03it/s]
 74%|███████▎  | 7893/10708 [1:41:33<23:05,  2.03it/s]
 74%|███████▎  | 7894/10708 [1:41:34<23:04,  2.03it/s]
 74%|███████▎  | 7895/10708 [1:41:34<23:04,  2.03it/s]
 74%|███████▎  | 7896/10708 [1:41:35<23:04,  2.03it/s]
 74%|███████▎  | 7897/10708 [1:41:35<23:02,  2.03it/s]
 74%|███████▍  | 7898/10708 [1:41:36<23:01,  2.03it/s]
 74%|███████▍  | 7899/10708 [1:41:36<22:59,  2.04it/s]
 74%|███████▍  | 7900/10708 [1:41:37<22:58,  2.04it/s]
                                                      
{'loss': 3.5532, 'grad_norm': 0.20628918707370758, 'learning_rate': 0.00019525841073419966, 'epoch': 0.74}

 74%|███████▍  | 7900/10708 [1:41:37<22:58,  2.04it/s]
 74%|███████▍  | 7901/10708 [1:41:37<23:00,  2.03it/s]
 74%|███████▍  | 7902/10708 [1:41:38<23:01,  2.03it/s]
 74%|███████▍  | 7903/10708 [1:41:38<23:00,  2.03it/s]
 74%|███████▍  | 7904/10708 [1:41:39<23:01,  2.03it/s]
 74%|███████▍  | 7905/10708 [1:41:39<22:58,  2.03it/s]
 74%|███████▍  | 7906/10708 [1:41:40<22:59,  2.03it/s]
 74%|███████▍  | 7907/10708 [1:41:40<22:59,  2.03it/s]
 74%|███████▍  | 7908/10708 [1:41:41<22:57,  2.03it/s]
 74%|███████▍  | 7909/10708 [1:41:41<22:56,  2.03it/s]
 74%|███████▍  | 7910/10708 [1:41:41<22:55,  2.03it/s]
 74%|███████▍  | 7911/10708 [1:41:42<22:55,  2.03it/s]
 74%|███████▍  | 7912/10708 [1:41:42<22:55,  2.03it/s]
 74%|███████▍  | 7913/10708 [1:41:43<23:00,  2.03it/s]
 74%|███████▍  | 7914/10708 [1:41:43<22:59,  2.03it/s]
 74%|███████▍  | 7915/10708 [1:41:44<22:58,  2.03it/s]
 74%|███████▍  | 7916/10708 [1:41:44<22:56,  2.03it/s]
 74%|███████▍  | 7917/10708 [1:41:45<22:55,  2.03it/s]
 74%|███████▍  | 7918/10708 [1:41:45<22:52,  2.03it/s]
 74%|███████▍  | 7919/10708 [1:41:46<22:51,  2.03it/s]
 74%|███████▍  | 7920/10708 [1:41:46<22:52,  2.03it/s]
 74%|███████▍  | 7921/10708 [1:41:47<22:50,  2.03it/s]
 74%|███████▍  | 7922/10708 [1:41:47<22:51,  2.03it/s]
 74%|███████▍  | 7923/10708 [1:41:48<22:50,  2.03it/s]
 74%|███████▍  | 7924/10708 [1:41:48<22:50,  2.03it/s]
 74%|███████▍  | 7925/10708 [1:41:49<22:49,  2.03it/s]
                                                      
{'loss': 3.5556, 'grad_norm': 0.21290211379528046, 'learning_rate': 0.0001920379822267414, 'epoch': 0.74}

 74%|███████▍  | 7925/10708 [1:41:49<22:49,  2.03it/s]
 74%|███████▍  | 7926/10708 [1:41:49<22:50,  2.03it/s]
 74%|███████▍  | 7927/10708 [1:41:50<22:49,  2.03it/s]
 74%|███████▍  | 7928/10708 [1:41:50<22:48,  2.03it/s]
 74%|███████▍  | 7929/10708 [1:41:51<22:45,  2.03it/s]
 74%|███████▍  | 7930/10708 [1:41:51<22:45,  2.04it/s]
 74%|███████▍  | 7931/10708 [1:41:52<22:46,  2.03it/s]
 74%|███████▍  | 7932/10708 [1:41:52<22:45,  2.03it/s]
 74%|███████▍  | 7933/10708 [1:41:53<22:47,  2.03it/s]
 74%|███████▍  | 7934/10708 [1:41:53<22:46,  2.03it/s]
 74%|███████▍  | 7935/10708 [1:41:54<22:45,  2.03it/s]
 74%|███████▍  | 7936/10708 [1:41:54<22:46,  2.03it/s]
 74%|███████▍  | 7937/10708 [1:41:55<22:44,  2.03it/s]
 74%|███████▍  | 7938/10708 [1:41:55<22:43,  2.03it/s]
 74%|███████▍  | 7939/10708 [1:41:56<22:42,  2.03it/s]
 74%|███████▍  | 7940/10708 [1:41:56<22:42,  2.03it/s]
 74%|███████▍  | 7941/10708 [1:41:57<22:42,  2.03it/s]
 74%|███████▍  | 7942/10708 [1:41:57<22:41,  2.03it/s]
 74%|███████▍  | 7943/10708 [1:41:58<22:40,  2.03it/s]
 74%|███████▍  | 7944/10708 [1:41:58<22:41,  2.03it/s]
 74%|███████▍  | 7945/10708 [1:41:59<22:40,  2.03it/s]
 74%|███████▍  | 7946/10708 [1:41:59<22:38,  2.03it/s]
 74%|███████▍  | 7947/10708 [1:42:00<22:37,  2.03it/s]
 74%|███████▍  | 7948/10708 [1:42:00<22:35,  2.04it/s]
 74%|███████▍  | 7949/10708 [1:42:01<22:35,  2.03it/s]
 74%|███████▍  | 7950/10708 [1:42:01<22:35,  2.04it/s]{'loss': 3.5597, 'grad_norm': 0.19825972616672516, 'learning_rate': 0.00018883800831016322, 'epoch': 0.74}

                                                      

 74%|███████▍  | 7950/10708 [1:42:01<22:35,  2.04it/s]
 74%|███████▍  | 7951/10708 [1:42:02<22:36,  2.03it/s]
 74%|███████▍  | 7952/10708 [1:42:02<22:35,  2.03it/s]
 74%|███████▍  | 7953/10708 [1:42:03<22:34,  2.03it/s]
 74%|███████▍  | 7954/10708 [1:42:03<22:34,  2.03it/s]
 74%|███████▍  | 7955/10708 [1:42:04<22:34,  2.03it/s]
 74%|███████▍  | 7956/10708 [1:42:04<22:33,  2.03it/s]
 74%|███████▍  | 7957/10708 [1:42:05<22:33,  2.03it/s]
 74%|███████▍  | 7958/10708 [1:42:05<22:34,  2.03it/s]
 74%|███████▍  | 7959/10708 [1:42:06<22:35,  2.03it/s]
 74%|███████▍  | 7960/10708 [1:42:06<22:33,  2.03it/s]
 74%|███████▍  | 7961/10708 [1:42:07<22:33,  2.03it/s]
 74%|███████▍  | 7962/10708 [1:42:07<22:32,  2.03it/s]
 74%|███████▍  | 7963/10708 [1:42:08<22:30,  2.03it/s]
 74%|███████▍  | 7964/10708 [1:42:08<22:29,  2.03it/s]
 74%|███████▍  | 7965/10708 [1:42:09<22:30,  2.03it/s]
 74%|███████▍  | 7966/10708 [1:42:09<22:29,  2.03it/s]
 74%|███████▍  | 7967/10708 [1:42:10<22:28,  2.03it/s]
 74%|███████▍  | 7968/10708 [1:42:10<22:27,  2.03it/s]
 74%|███████▍  | 7969/10708 [1:42:11<22:26,  2.03it/s]
 74%|███████▍  | 7970/10708 [1:42:11<22:27,  2.03it/s]
 74%|███████▍  | 7971/10708 [1:42:12<22:26,  2.03it/s]
 74%|███████▍  | 7972/10708 [1:42:12<22:27,  2.03it/s]
 74%|███████▍  | 7973/10708 [1:42:13<22:27,  2.03it/s]
 74%|███████▍  | 7974/10708 [1:42:13<22:29,  2.03it/s]
 74%|███████▍  | 7975/10708 [1:42:13<22:28,  2.03it/s]{'loss': 3.5576, 'grad_norm': 0.20794692635536194, 'learning_rate': 0.00018565870152417263, 'epoch': 0.74}

                                                      

 74%|███████▍  | 7975/10708 [1:42:13<22:28,  2.03it/s]
 74%|███████▍  | 7976/10708 [1:42:14<22:27,  2.03it/s]
 74%|███████▍  | 7977/10708 [1:42:14<22:26,  2.03it/s]
 75%|███████▍  | 7978/10708 [1:42:15<22:24,  2.03it/s]
 75%|███████▍  | 7979/10708 [1:42:15<22:24,  2.03it/s]
 75%|███████▍  | 7980/10708 [1:42:16<22:22,  2.03it/s]
 75%|███████▍  | 7981/10708 [1:42:16<22:21,  2.03it/s]
 75%|███████▍  | 7982/10708 [1:42:17<22:20,  2.03it/s]
 75%|███████▍  | 7983/10708 [1:42:17<22:20,  2.03it/s]
 75%|███████▍  | 7984/10708 [1:42:18<22:20,  2.03it/s]
 75%|███████▍  | 7985/10708 [1:42:18<22:18,  2.03it/s]
 75%|███████▍  | 7986/10708 [1:42:19<22:19,  2.03it/s]
 75%|███████▍  | 7987/10708 [1:42:19<22:18,  2.03it/s]
 75%|███████▍  | 7988/10708 [1:42:20<22:17,  2.03it/s]
 75%|███████▍  | 7989/10708 [1:42:20<22:18,  2.03it/s]
 75%|███████▍  | 7990/10708 [1:42:21<22:17,  2.03it/s]
 75%|███████▍  | 7991/10708 [1:42:21<22:15,  2.03it/s]
 75%|███████▍  | 7992/10708 [1:42:22<22:16,  2.03it/s]
 75%|███████▍  | 7993/10708 [1:42:22<22:15,  2.03it/s]
 75%|███████▍  | 7994/10708 [1:42:23<22:16,  2.03it/s]
 75%|███████▍  | 7995/10708 [1:42:23<22:14,  2.03it/s]
 75%|███████▍  | 7996/10708 [1:42:24<22:13,  2.03it/s]
 75%|███████▍  | 7997/10708 [1:42:24<22:15,  2.03it/s]
 75%|███████▍  | 7998/10708 [1:42:25<22:13,  2.03it/s]
 75%|███████▍  | 7999/10708 [1:42:25<22:13,  2.03it/s]
 75%|███████▍  | 8000/10708 [1:42:26<22:13,  2.03it/s]
                                                      
{'loss': 3.5574, 'grad_norm': 0.20874689519405365, 'learning_rate': 0.00018250027303578353, 'epoch': 0.75}

 75%|███████▍  | 8000/10708 [1:42:26<22:13,  2.03it/s]
 75%|███████▍  | 8001/10708 [1:42:26<22:12,  2.03it/s]
 75%|███████▍  | 8002/10708 [1:42:27<22:17,  2.02it/s]
 75%|███████▍  | 8003/10708 [1:42:27<22:14,  2.03it/s]
 75%|███████▍  | 8004/10708 [1:42:28<22:14,  2.03it/s]
 75%|███████▍  | 8005/10708 [1:42:28<22:12,  2.03it/s]
 75%|███████▍  | 8006/10708 [1:42:29<22:10,  2.03it/s]
 75%|███████▍  | 8007/10708 [1:42:29<22:10,  2.03it/s]
 75%|███████▍  | 8008/10708 [1:42:30<22:11,  2.03it/s]
 75%|███████▍  | 8009/10708 [1:42:30<22:11,  2.03it/s]
 75%|███████▍  | 8010/10708 [1:42:31<22:10,  2.03it/s]
 75%|███████▍  | 8011/10708 [1:42:31<22:09,  2.03it/s]
 75%|███████▍  | 8012/10708 [1:42:32<22:07,  2.03it/s]
 75%|███████▍  | 8013/10708 [1:42:32<22:05,  2.03it/s]
 75%|███████▍  | 8014/10708 [1:42:33<22:04,  2.03it/s]
 75%|███████▍  | 8015/10708 [1:42:33<22:04,  2.03it/s]
 75%|███████▍  | 8016/10708 [1:42:34<22:03,  2.03it/s]
 75%|███████▍  | 8017/10708 [1:42:34<22:03,  2.03it/s]
 75%|███████▍  | 8018/10708 [1:42:35<22:04,  2.03it/s]
 75%|███████▍  | 8019/10708 [1:42:35<22:02,  2.03it/s]
 75%|███████▍  | 8020/10708 [1:42:36<22:01,  2.03it/s]
 75%|███████▍  | 8021/10708 [1:42:36<22:01,  2.03it/s]
 75%|███████▍  | 8022/10708 [1:42:37<21:59,  2.04it/s]
 75%|███████▍  | 8023/10708 [1:42:37<21:59,  2.03it/s]
 75%|███████▍  | 8024/10708 [1:42:38<21:59,  2.03it/s]
 75%|███████▍  | 8025/10708 [1:42:38<21:58,  2.04it/s]
                                                      
{'loss': 3.5573, 'grad_norm': 0.20225083827972412, 'learning_rate': 0.0001793629326252894, 'epoch': 0.75}

 75%|███████▍  | 8025/10708 [1:42:38<21:58,  2.04it/s]
 75%|███████▍  | 8026/10708 [1:42:39<22:02,  2.03it/s]
 75%|███████▍  | 8027/10708 [1:42:39<22:01,  2.03it/s]
 75%|███████▍  | 8028/10708 [1:42:40<21:59,  2.03it/s]
 75%|███████▍  | 8029/10708 [1:42:40<21:57,  2.03it/s]
 75%|███████▍  | 8030/10708 [1:42:41<21:56,  2.03it/s]
 75%|███████▌  | 8031/10708 [1:42:41<21:56,  2.03it/s]
 75%|███████▌  | 8032/10708 [1:42:42<21:56,  2.03it/s]
 75%|███████▌  | 8033/10708 [1:42:42<21:55,  2.03it/s]
 75%|███████▌  | 8034/10708 [1:42:43<21:54,  2.03it/s]
 75%|███████▌  | 8035/10708 [1:42:43<21:58,  2.03it/s]
 75%|███████▌  | 8036/10708 [1:42:44<21:57,  2.03it/s]
 75%|███████▌  | 8037/10708 [1:42:44<21:57,  2.03it/s]
 75%|███████▌  | 8038/10708 [1:42:45<21:55,  2.03it/s]
 75%|███████▌  | 8039/10708 [1:42:45<21:52,  2.03it/s]
 75%|███████▌  | 8040/10708 [1:42:45<21:52,  2.03it/s]
 75%|███████▌  | 8041/10708 [1:42:46<21:51,  2.03it/s]
 75%|███████▌  | 8042/10708 [1:42:46<21:50,  2.03it/s]
 75%|███████▌  | 8043/10708 [1:42:47<21:50,  2.03it/s]
 75%|███████▌  | 8044/10708 [1:42:47<21:49,  2.03it/s]
 75%|███████▌  | 8045/10708 [1:42:48<21:49,  2.03it/s]
 75%|███████▌  | 8046/10708 [1:42:48<21:47,  2.04it/s]
 75%|███████▌  | 8047/10708 [1:42:49<21:47,  2.03it/s]
 75%|███████▌  | 8048/10708 [1:42:49<21:46,  2.04it/s]
 75%|███████▌  | 8049/10708 [1:42:50<21:44,  2.04it/s]
 75%|███████▌  | 8050/10708 [1:42:50<21:45,  2.04it/s]{'loss': 3.5583, 'grad_norm': 0.2041643261909485, 'learning_rate': 0.00017624688867233103, 'epoch': 0.75}

                                                      

 75%|███████▌  | 8050/10708 [1:42:50<21:45,  2.04it/s]
 75%|███████▌  | 8051/10708 [1:42:51<21:47,  2.03it/s]
 75%|███████▌  | 8052/10708 [1:42:51<21:45,  2.04it/s]
 75%|███████▌  | 8053/10708 [1:42:52<21:45,  2.03it/s]
 75%|███████▌  | 8054/10708 [1:42:52<21:45,  2.03it/s]
 75%|███████▌  | 8055/10708 [1:42:53<21:45,  2.03it/s]
 75%|███████▌  | 8056/10708 [1:42:53<21:45,  2.03it/s]
 75%|███████▌  | 8057/10708 [1:42:54<21:45,  2.03it/s]
 75%|███████▌  | 8058/10708 [1:42:54<21:43,  2.03it/s]
 75%|███████▌  | 8059/10708 [1:42:55<21:41,  2.03it/s]
 75%|███████▌  | 8060/10708 [1:42:55<21:42,  2.03it/s]
 75%|███████▌  | 8061/10708 [1:42:56<21:41,  2.03it/s]
 75%|███████▌  | 8062/10708 [1:42:56<21:41,  2.03it/s]
 75%|███████▌  | 8063/10708 [1:42:57<21:41,  2.03it/s]
 75%|███████▌  | 8064/10708 [1:42:57<21:41,  2.03it/s]
 75%|███████▌  | 8065/10708 [1:42:58<21:41,  2.03it/s]
 75%|███████▌  | 8066/10708 [1:42:58<21:39,  2.03it/s]
 75%|███████▌  | 8067/10708 [1:42:59<21:39,  2.03it/s]
 75%|███████▌  | 8068/10708 [1:42:59<21:38,  2.03it/s]
 75%|███████▌  | 8069/10708 [1:43:00<21:37,  2.03it/s]
 75%|███████▌  | 8070/10708 [1:43:00<21:37,  2.03it/s]
 75%|███████▌  | 8071/10708 [1:43:01<21:36,  2.03it/s]
 75%|███████▌  | 8072/10708 [1:43:01<21:37,  2.03it/s]
 75%|███████▌  | 8073/10708 [1:43:02<21:37,  2.03it/s]
 75%|███████▌  | 8074/10708 [1:43:02<21:37,  2.03it/s]
 75%|███████▌  | 8075/10708 [1:43:03<21:38,  2.03it/s]
                                                      
{'loss': 3.5472, 'grad_norm': 0.2039419263601303, 'learning_rate': 0.00017315234814205483, 'epoch': 0.75}

 75%|███████▌  | 8075/10708 [1:43:03<21:38,  2.03it/s]
 75%|███████▌  | 8076/10708 [1:43:03<21:40,  2.02it/s]
 75%|███████▌  | 8077/10708 [1:43:04<21:37,  2.03it/s]
 75%|███████▌  | 8078/10708 [1:43:04<21:38,  2.03it/s]
 75%|███████▌  | 8079/10708 [1:43:05<21:35,  2.03it/s]
 75%|███████▌  | 8080/10708 [1:43:05<21:35,  2.03it/s]
 75%|███████▌  | 8081/10708 [1:43:06<21:34,  2.03it/s]
 75%|███████▌  | 8082/10708 [1:43:06<21:33,  2.03it/s]
 75%|███████▌  | 8083/10708 [1:43:07<21:32,  2.03it/s]
 75%|███████▌  | 8084/10708 [1:43:07<21:29,  2.03it/s]
 76%|███████▌  | 8085/10708 [1:43:08<21:30,  2.03it/s]
 76%|███████▌  | 8086/10708 [1:43:08<21:28,  2.03it/s]
 76%|███████▌  | 8087/10708 [1:43:09<21:27,  2.04it/s]
 76%|███████▌  | 8088/10708 [1:43:09<21:27,  2.03it/s]
 76%|███████▌  | 8089/10708 [1:43:10<21:27,  2.03it/s]
 76%|███████▌  | 8090/10708 [1:43:10<21:27,  2.03it/s]
 76%|███████▌  | 8091/10708 [1:43:11<21:28,  2.03it/s]
 76%|███████▌  | 8092/10708 [1:43:11<21:28,  2.03it/s]
 76%|███████▌  | 8093/10708 [1:43:12<21:27,  2.03it/s]
 76%|███████▌  | 8094/10708 [1:43:12<21:28,  2.03it/s]
 76%|███████▌  | 8095/10708 [1:43:13<21:27,  2.03it/s]
 76%|███████▌  | 8096/10708 [1:43:13<21:31,  2.02it/s]
 76%|███████▌  | 8097/10708 [1:43:14<21:31,  2.02it/s]
 76%|███████▌  | 8098/10708 [1:43:14<21:28,  2.03it/s]
 76%|███████▌  | 8099/10708 [1:43:15<21:24,  2.03it/s]
 76%|███████▌  | 8100/10708 [1:43:15<21:23,  2.03it/s]{'loss': 3.5532, 'grad_norm': 0.20255491137504578, 'learning_rate': 0.00017007951657136778, 'epoch': 0.76}

                                                      

 76%|███████▌  | 8100/10708 [1:43:15<21:23,  2.03it/s]
 76%|███████▌  | 8101/10708 [1:43:16<21:24,  2.03it/s]
 76%|███████▌  | 8102/10708 [1:43:16<21:21,  2.03it/s]
 76%|███████▌  | 8103/10708 [1:43:16<21:21,  2.03it/s]
 76%|███████▌  | 8104/10708 [1:43:17<21:21,  2.03it/s]
 76%|███████▌  | 8105/10708 [1:43:17<21:21,  2.03it/s]
 76%|███████▌  | 8106/10708 [1:43:18<21:21,  2.03it/s]
 76%|███████▌  | 8107/10708 [1:43:18<21:20,  2.03it/s]
 76%|███████▌  | 8108/10708 [1:43:19<21:19,  2.03it/s]
 76%|███████▌  | 8109/10708 [1:43:19<21:20,  2.03it/s]
 76%|███████▌  | 8110/10708 [1:43:20<21:17,  2.03it/s]
 76%|███████▌  | 8111/10708 [1:43:20<21:16,  2.03it/s]
 76%|███████▌  | 8112/10708 [1:43:21<21:18,  2.03it/s]
 76%|███████▌  | 8113/10708 [1:43:21<21:18,  2.03it/s]
 76%|███████▌  | 8114/10708 [1:43:22<21:18,  2.03it/s]
 76%|███████▌  | 8115/10708 [1:43:23<24:51,  1.74it/s]
 76%|███████▌  | 8116/10708 [1:43:23<23:44,  1.82it/s]
 76%|███████▌  | 8117/10708 [1:43:24<22:58,  1.88it/s]
 76%|███████▌  | 8118/10708 [1:43:24<22:56,  1.88it/s]
 76%|███████▌  | 8119/10708 [1:43:25<22:25,  1.92it/s]
 76%|███████▌  | 8120/10708 [1:43:25<22:02,  1.96it/s]
 76%|███████▌  | 8121/10708 [1:43:26<21:46,  1.98it/s]
 76%|███████▌  | 8122/10708 [1:43:26<21:36,  1.99it/s]
 76%|███████▌  | 8123/10708 [1:43:27<21:28,  2.01it/s]
 76%|███████▌  | 8124/10708 [1:43:27<21:23,  2.01it/s]
 76%|███████▌  | 8125/10708 [1:43:28<21:18,  2.02it/s]
                                                      {'loss': 3.5569, 'grad_norm': 0.21344296634197235, 'learning_rate': 0.00016702859805528468, 'epoch': 0.76}

 76%|███████▌  | 8125/10708 [1:43:28<21:18,  2.02it/s]
 76%|███████▌  | 8126/10708 [1:43:28<21:17,  2.02it/s]
 76%|███████▌  | 8127/10708 [1:43:29<21:14,  2.03it/s]
 76%|███████▌  | 8128/10708 [1:43:29<21:12,  2.03it/s]
 76%|███████▌  | 8129/10708 [1:43:30<21:11,  2.03it/s]
 76%|███████▌  | 8130/10708 [1:43:30<21:09,  2.03it/s]
 76%|███████▌  | 8131/10708 [1:43:31<21:08,  2.03it/s]
 76%|███████▌  | 8132/10708 [1:43:31<21:06,  2.03it/s]
 76%|███████▌  | 8133/10708 [1:43:32<21:38,  1.98it/s]
 76%|███████▌  | 8134/10708 [1:43:32<21:28,  2.00it/s]
 76%|███████▌  | 8135/10708 [1:43:33<21:20,  2.01it/s]
 76%|███████▌  | 8136/10708 [1:43:33<21:15,  2.02it/s]
 76%|███████▌  | 8137/10708 [1:43:34<21:12,  2.02it/s]
 76%|███████▌  | 8138/10708 [1:43:34<21:08,  2.03it/s]
 76%|███████▌  | 8139/10708 [1:43:35<21:06,  2.03it/s]
 76%|███████▌  | 8140/10708 [1:43:35<21:05,  2.03it/s]
 76%|███████▌  | 8141/10708 [1:43:36<21:04,  2.03it/s]
 76%|███████▌  | 8142/10708 [1:43:36<21:02,  2.03it/s]
 76%|███████▌  | 8143/10708 [1:43:37<21:02,  2.03it/s]
 76%|███████▌  | 8144/10708 [1:43:37<21:00,  2.03it/s]
 76%|███████▌  | 8145/10708 [1:43:38<20:57,  2.04it/s]
 76%|███████▌  | 8146/10708 [1:43:38<21:35,  1.98it/s]
 76%|███████▌  | 8147/10708 [1:43:39<21:21,  2.00it/s]
 76%|███████▌  | 8148/10708 [1:43:39<21:14,  2.01it/s]
 76%|███████▌  | 8149/10708 [1:43:40<21:09,  2.02it/s]
 76%|███████▌  | 8150/10708 [1:43:40<21:04,  2.02it/s]
                                                      
{'loss': 3.5551, 'grad_norm': 0.2016773372888565, 'learning_rate': 0.00016399979523337328, 'epoch': 0.76}

 76%|███████▌  | 8150/10708 [1:43:40<21:04,  2.02it/s]
 76%|███████▌  | 8151/10708 [1:43:41<24:33,  1.74it/s]
 76%|███████▌  | 8152/10708 [1:43:41<23:28,  1.82it/s]
 76%|███████▌  | 8153/10708 [1:43:42<22:42,  1.87it/s]
 76%|███████▌  | 8154/10708 [1:43:42<22:09,  1.92it/s]
 76%|███████▌  | 8155/10708 [1:43:43<21:48,  1.95it/s]
 76%|███████▌  | 8156/10708 [1:43:43<21:36,  1.97it/s]
 76%|███████▌  | 8157/10708 [1:43:44<21:23,  1.99it/s]
 76%|███████▌  | 8158/10708 [1:43:44<21:16,  2.00it/s]
 76%|███████▌  | 8159/10708 [1:43:45<21:08,  2.01it/s]
 76%|███████▌  | 8160/10708 [1:43:45<21:04,  2.02it/s]
 76%|███████▌  | 8161/10708 [1:43:46<20:59,  2.02it/s]
 76%|███████▌  | 8162/10708 [1:43:46<20:57,  2.03it/s]
 76%|███████▌  | 8163/10708 [1:43:47<20:55,  2.03it/s]
 76%|███████▌  | 8164/10708 [1:43:47<20:53,  2.03it/s]
 76%|███████▋  | 8165/10708 [1:43:48<20:51,  2.03it/s]
 76%|███████▋  | 8166/10708 [1:43:48<20:52,  2.03it/s]
 76%|███████▋  | 8167/10708 [1:43:49<20:50,  2.03it/s]
 76%|███████▋  | 8168/10708 [1:43:49<20:49,  2.03it/s]
 76%|███████▋  | 8169/10708 [1:43:50<20:49,  2.03it/s]
 76%|███████▋  | 8170/10708 [1:43:50<20:46,  2.04it/s]
 76%|███████▋  | 8171/10708 [1:43:51<20:46,  2.03it/s]
 76%|███████▋  | 8172/10708 [1:43:51<20:47,  2.03it/s]
 76%|███████▋  | 8173/10708 [1:43:52<20:46,  2.03it/s]
 76%|███████▋  | 8174/10708 [1:43:52<20:45,  2.03it/s]
 76%|███████▋  | 8175/10708 [1:43:53<20:44,  2.03it/s]
                                                      {'loss': 3.554, 'grad_norm': 0.20271672308444977, 'learning_rate': 0.00016099330927629374, 'epoch': 0.76}

 76%|███████▋  | 8175/10708 [1:43:53<20:44,  2.03it/s]
 76%|███████▋  | 8176/10708 [1:43:53<20:45,  2.03it/s]
 76%|███████▋  | 8177/10708 [1:43:54<20:45,  2.03it/s]
 76%|███████▋  | 8178/10708 [1:43:54<20:45,  2.03it/s]
 76%|███████▋  | 8179/10708 [1:43:55<20:44,  2.03it/s]
 76%|███████▋  | 8180/10708 [1:43:55<20:43,  2.03it/s]
 76%|███████▋  | 8181/10708 [1:43:56<20:45,  2.03it/s]
 76%|███████▋  | 8182/10708 [1:43:56<20:43,  2.03it/s]
 76%|███████▋  | 8183/10708 [1:43:57<20:44,  2.03it/s]
 76%|███████▋  | 8184/10708 [1:43:57<20:44,  2.03it/s]
 76%|███████▋  | 8185/10708 [1:43:58<20:42,  2.03it/s]
 76%|███████▋  | 8186/10708 [1:43:58<20:40,  2.03it/s]
 76%|███████▋  | 8187/10708 [1:43:59<20:42,  2.03it/s]
 76%|███████▋  | 8188/10708 [1:43:59<20:39,  2.03it/s]
 76%|███████▋  | 8189/10708 [1:43:59<20:37,  2.04it/s]
 76%|███████▋  | 8190/10708 [1:44:00<20:37,  2.03it/s]
 76%|███████▋  | 8191/10708 [1:44:00<20:38,  2.03it/s]
 77%|███████▋  | 8192/10708 [1:44:01<20:37,  2.03it/s]
 77%|███████▋  | 8193/10708 [1:44:01<20:38,  2.03it/s]
 77%|███████▋  | 8194/10708 [1:44:02<20:37,  2.03it/s]
 77%|███████▋  | 8195/10708 [1:44:02<20:36,  2.03it/s]
 77%|███████▋  | 8196/10708 [1:44:03<20:35,  2.03it/s]
 77%|███████▋  | 8197/10708 [1:44:03<20:35,  2.03it/s]
 77%|███████▋  | 8198/10708 [1:44:04<20:33,  2.03it/s]
 77%|███████▋  | 8199/10708 [1:44:04<20:32,  2.04it/s]
 77%|███████▋  | 8200/10708 [1:44:05<20:33,  2.03it/s]
                                                      {'loss': 3.5524, 'grad_norm': 0.22203558683395386, 'learning_rate': 0.0001580093398724388, 'epoch': 0.77}


 77%|███████▋  | 8200/10708 [1:44:05<20:33,  2.03it/s]
 77%|███████▋  | 8201/10708 [1:44:05<20:34,  2.03it/s]
 77%|███████▋  | 8202/10708 [1:44:06<20:34,  2.03it/s]
 77%|███████▋  | 8203/10708 [1:44:06<20:32,  2.03it/s]
 77%|███████▋  | 8204/10708 [1:44:07<20:30,  2.03it/s]
 77%|███████▋  | 8205/10708 [1:44:07<20:31,  2.03it/s]
 77%|███████▋  | 8206/10708 [1:44:08<20:30,  2.03it/s]
 77%|███████▋  | 8207/10708 [1:44:08<20:28,  2.04it/s]
 77%|███████▋  | 8208/10708 [1:44:09<20:28,  2.03it/s]
 77%|███████▋  | 8209/10708 [1:44:09<20:27,  2.04it/s]
 77%|███████▋  | 8210/10708 [1:44:10<20:26,  2.04it/s]
 77%|███████▋  | 8211/10708 [1:44:10<20:28,  2.03it/s]
 77%|███████▋  | 8212/10708 [1:44:11<20:27,  2.03it/s]
 77%|███████▋  | 8213/10708 [1:44:11<20:27,  2.03it/s]
 77%|███████▋  | 8214/10708 [1:44:12<20:27,  2.03it/s]
 77%|███████▋  | 8215/10708 [1:44:12<20:25,  2.03it/s]
 77%|███████▋  | 8216/10708 [1:44:13<20:26,  2.03it/s]
 77%|███████▋  | 8217/10708 [1:44:13<20:26,  2.03it/s]
 77%|███████▋  | 8218/10708 [1:44:14<20:24,  2.03it/s]
 77%|███████▋  | 8219/10708 [1:44:14<20:25,  2.03it/s]
 77%|███████▋  | 8220/10708 [1:44:15<20:23,  2.03it/s]
 77%|███████▋  | 8221/10708 [1:44:15<20:21,  2.04it/s]
 77%|███████▋  | 8222/10708 [1:44:16<20:21,  2.04it/s]
 77%|███████▋  | 8223/10708 [1:44:16<20:21,  2.03it/s]
 77%|███████▋  | 8224/10708 [1:44:17<20:20,  2.04it/s]
 77%|███████▋  | 8225/10708 [1:44:17<20:21,  2.03it/s]
                                                      {'loss': 3.5544, 'grad_norm': 0.20285533368587494, 'learning_rate': 0.0001550480852146691, 'epoch': 0.77}


 77%|███████▋  | 8225/10708 [1:44:17<20:21,  2.03it/s]
 77%|███████▋  | 8226/10708 [1:44:18<20:20,  2.03it/s]
 77%|███████▋  | 8227/10708 [1:44:18<20:20,  2.03it/s]
 77%|███████▋  | 8228/10708 [1:44:19<20:21,  2.03it/s]
 77%|███████▋  | 8229/10708 [1:44:19<20:20,  2.03it/s]
 77%|███████▋  | 8230/10708 [1:44:20<20:21,  2.03it/s]
 77%|███████▋  | 8231/10708 [1:44:20<20:18,  2.03it/s]
 77%|███████▋  | 8232/10708 [1:44:21<20:18,  2.03it/s]
 77%|███████▋  | 8233/10708 [1:44:21<20:17,  2.03it/s]
 77%|███████▋  | 8234/10708 [1:44:22<20:16,  2.03it/s]
 77%|███████▋  | 8235/10708 [1:44:22<20:16,  2.03it/s]
 77%|███████▋  | 8236/10708 [1:44:23<20:15,  2.03it/s]
 77%|███████▋  | 8237/10708 [1:44:23<20:14,  2.04it/s]
 77%|███████▋  | 8238/10708 [1:44:24<20:14,  2.03it/s]
 77%|███████▋  | 8239/10708 [1:44:24<20:13,  2.03it/s]
 77%|███████▋  | 8240/10708 [1:44:25<20:12,  2.04it/s]
 77%|███████▋  | 8241/10708 [1:44:25<20:10,  2.04it/s]
 77%|███████▋  | 8242/10708 [1:44:26<20:13,  2.03it/s]
 77%|███████▋  | 8243/10708 [1:44:26<20:12,  2.03it/s]
 77%|███████▋  | 8244/10708 [1:44:27<20:09,  2.04it/s]
 77%|███████▋  | 8245/10708 [1:44:27<20:10,  2.03it/s]
 77%|███████▋  | 8246/10708 [1:44:28<20:10,  2.03it/s]
 77%|███████▋  | 8247/10708 [1:44:28<20:09,  2.03it/s]
 77%|███████▋  | 8248/10708 [1:44:28<20:09,  2.03it/s]
 77%|███████▋  | 8249/10708 [1:44:29<20:08,  2.03it/s]
 77%|███████▋  | 8250/10708 [1:44:29<20:09,  2.03it/s]
                                                      {'loss': 3.5493, 'grad_norm': 0.21280908584594727, 'learning_rate': 0.0001521097419871506, 'epoch': 0.77}

 77%|███████▋  | 8250/10708 [1:44:29<20:09,  2.03it/s]
 77%|███████▋  | 8251/10708 [1:44:30<20:08,  2.03it/s]
 77%|███████▋  | 8252/10708 [1:44:30<20:08,  2.03it/s]
 77%|███████▋  | 8253/10708 [1:44:31<20:08,  2.03it/s]
 77%|███████▋  | 8254/10708 [1:44:31<20:06,  2.03it/s]
 77%|███████▋  | 8255/10708 [1:44:32<20:05,  2.03it/s]
 77%|███████▋  | 8256/10708 [1:44:32<20:06,  2.03it/s]
 77%|███████▋  | 8257/10708 [1:44:33<20:04,  2.03it/s]
 77%|███████▋  | 8258/10708 [1:44:33<20:03,  2.04it/s]
 77%|███████▋  | 8259/10708 [1:44:34<20:03,  2.03it/s]
 77%|███████▋  | 8260/10708 [1:44:34<20:03,  2.03it/s]
 77%|███████▋  | 8261/10708 [1:44:35<20:01,  2.04it/s]
 77%|███████▋  | 8262/10708 [1:44:35<20:02,  2.03it/s]
 77%|███████▋  | 8263/10708 [1:44:36<20:03,  2.03it/s]
 77%|███████▋  | 8264/10708 [1:44:36<20:01,  2.03it/s]
 77%|███████▋  | 8265/10708 [1:44:37<20:01,  2.03it/s]
 77%|███████▋  | 8266/10708 [1:44:37<19:59,  2.04it/s]
 77%|███████▋  | 8267/10708 [1:44:38<19:59,  2.04it/s]
 77%|███████▋  | 8268/10708 [1:44:38<20:01,  2.03it/s]
 77%|███████▋  | 8269/10708 [1:44:39<20:00,  2.03it/s]
 77%|███████▋  | 8270/10708 [1:44:39<19:58,  2.03it/s]
 77%|███████▋  | 8271/10708 [1:44:40<20:00,  2.03it/s]
 77%|███████▋  | 8272/10708 [1:44:40<19:58,  2.03it/s]
 77%|███████▋  | 8273/10708 [1:44:41<19:58,  2.03it/s]
 77%|███████▋  | 8274/10708 [1:44:41<19:56,  2.03it/s]
 77%|███████▋  | 8275/10708 [1:44:42<19:56,  2.03it/s]
                                                      
{'loss': 3.5447, 'grad_norm': 0.22011396288871765, 'learning_rate': 0.00014919450535228973, 'epoch': 0.77}

 77%|███████▋  | 8275/10708 [1:44:42<19:56,  2.03it/s]
 77%|███████▋  | 8276/10708 [1:44:42<19:57,  2.03it/s]
 77%|███████▋  | 8277/10708 [1:44:43<19:58,  2.03it/s]
 77%|███████▋  | 8278/10708 [1:44:43<19:59,  2.03it/s]
 77%|███████▋  | 8279/10708 [1:44:44<19:58,  2.03it/s]
 77%|███████▋  | 8280/10708 [1:44:44<19:56,  2.03it/s]
 77%|███████▋  | 8281/10708 [1:44:45<19:55,  2.03it/s]
 77%|███████▋  | 8282/10708 [1:44:45<19:55,  2.03it/s]
 77%|███████▋  | 8283/10708 [1:44:46<19:53,  2.03it/s]
 77%|███████▋  | 8284/10708 [1:44:46<19:52,  2.03it/s]
 77%|███████▋  | 8285/10708 [1:44:47<19:52,  2.03it/s]
 77%|███████▋  | 8286/10708 [1:44:47<19:51,  2.03it/s]
 77%|███████▋  | 8287/10708 [1:44:48<19:50,  2.03it/s]
 77%|███████▋  | 8288/10708 [1:44:48<19:50,  2.03it/s]
 77%|███████▋  | 8289/10708 [1:44:49<19:49,  2.03it/s]
 77%|███████▋  | 8290/10708 [1:44:49<19:48,  2.03it/s]
 77%|███████▋  | 8291/10708 [1:44:50<19:48,  2.03it/s]
 77%|███████▋  | 8292/10708 [1:44:50<19:47,  2.03it/s]
 77%|███████▋  | 8293/10708 [1:44:51<19:47,  2.03it/s]
 77%|███████▋  | 8294/10708 [1:44:51<19:47,  2.03it/s]
 77%|███████▋  | 8295/10708 [1:44:52<19:46,  2.03it/s]
 77%|███████▋  | 8296/10708 [1:44:52<19:45,  2.03it/s]
 77%|███████▋  | 8297/10708 [1:44:53<19:45,  2.03it/s]
 77%|███████▋  | 8298/10708 [1:44:53<19:45,  2.03it/s]
 78%|███████▊  | 8299/10708 [1:44:54<19:44,  2.03it/s]
 78%|███████▊  | 8300/10708 [1:44:54<19:43,  2.03it/s]
                                                      {'loss': 3.5492, 'grad_norm': 0.2016398012638092, 'learning_rate': 0.00014630256893777218, 'epoch': 0.78}

 78%|███████▊  | 8300/10708 [1:44:54<19:43,  2.03it/s]
 78%|███████▊  | 8301/10708 [1:44:55<19:44,  2.03it/s]
 78%|███████▊  | 8302/10708 [1:44:55<19:43,  2.03it/s]
 78%|███████▊  | 8303/10708 [1:44:56<19:43,  2.03it/s]
 78%|███████▊  | 8304/10708 [1:44:56<19:41,  2.03it/s]
 78%|███████▊  | 8305/10708 [1:44:57<19:41,  2.03it/s]
 78%|███████▊  | 8306/10708 [1:44:57<19:40,  2.04it/s]
 78%|███████▊  | 8307/10708 [1:44:58<19:39,  2.04it/s]
 78%|███████▊  | 8308/10708 [1:44:58<19:39,  2.03it/s]
 78%|███████▊  | 8309/10708 [1:44:58<19:38,  2.04it/s]
 78%|███████▊  | 8310/10708 [1:44:59<19:36,  2.04it/s]
 78%|███████▊  | 8311/10708 [1:44:59<19:37,  2.04it/s]
 78%|███████▊  | 8312/10708 [1:45:00<19:37,  2.03it/s]
 78%|███████▊  | 8313/10708 [1:45:00<19:36,  2.04it/s]
 78%|███████▊  | 8314/10708 [1:45:01<19:36,  2.03it/s]
 78%|███████▊  | 8315/10708 [1:45:01<19:36,  2.03it/s]
 78%|███████▊  | 8316/10708 [1:45:02<19:34,  2.04it/s]
 78%|███████▊  | 8317/10708 [1:45:02<19:34,  2.04it/s]
 78%|███████▊  | 8318/10708 [1:45:03<19:34,  2.03it/s]
 78%|███████▊  | 8319/10708 [1:45:03<19:34,  2.03it/s]
 78%|███████▊  | 8320/10708 [1:45:04<19:32,  2.04it/s]
 78%|███████▊  | 8321/10708 [1:45:04<19:32,  2.04it/s]
 78%|███████▊  | 8322/10708 [1:45:05<19:33,  2.03it/s]
 78%|███████▊  | 8323/10708 [1:45:05<19:34,  2.03it/s]
 78%|███████▊  | 8324/10708 [1:45:06<19:33,  2.03it/s]
 78%|███████▊  | 8325/10708 [1:45:06<19:32,  2.03it/s]
                                                      
{'loss': 3.5527, 'grad_norm': 0.20495714247226715, 'learning_rate': 0.00014343412482370095, 'epoch': 0.78}

 78%|███████▊  | 8325/10708 [1:45:06<19:32,  2.03it/s]
 78%|███████▊  | 8326/10708 [1:45:07<19:36,  2.02it/s]
 78%|███████▊  | 8327/10708 [1:45:07<19:35,  2.02it/s]
 78%|███████▊  | 8328/10708 [1:45:08<19:34,  2.03it/s]
 78%|███████▊  | 8329/10708 [1:45:08<19:30,  2.03it/s]
 78%|███████▊  | 8330/10708 [1:45:09<19:30,  2.03it/s]
 78%|███████▊  | 8331/10708 [1:45:09<19:29,  2.03it/s]
 78%|███████▊  | 8332/10708 [1:45:10<19:27,  2.03it/s]
 78%|███████▊  | 8333/10708 [1:45:10<19:26,  2.04it/s]
 78%|███████▊  | 8334/10708 [1:45:11<19:27,  2.03it/s]
 78%|███████▊  | 8335/10708 [1:45:11<19:27,  2.03it/s]
 78%|███████▊  | 8336/10708 [1:45:12<19:25,  2.03it/s]
 78%|███████▊  | 8337/10708 [1:45:12<19:26,  2.03it/s]
 78%|███████▊  | 8338/10708 [1:45:13<19:26,  2.03it/s]
 78%|███████▊  | 8339/10708 [1:45:13<19:28,  2.03it/s]
 78%|███████▊  | 8340/10708 [1:45:14<19:26,  2.03it/s]
 78%|███████▊  | 8341/10708 [1:45:14<19:24,  2.03it/s]
 78%|███████▊  | 8342/10708 [1:45:15<19:24,  2.03it/s]
 78%|███████▊  | 8343/10708 [1:45:15<19:23,  2.03it/s]
 78%|███████▊  | 8344/10708 [1:45:16<19:21,  2.04it/s]
 78%|███████▊  | 8345/10708 [1:45:16<19:20,  2.04it/s]
 78%|███████▊  | 8346/10708 [1:45:17<19:20,  2.04it/s]
 78%|███████▊  | 8347/10708 [1:45:17<19:21,  2.03it/s]
 78%|███████▊  | 8348/10708 [1:45:18<19:21,  2.03it/s]
 78%|███████▊  | 8349/10708 [1:45:18<19:21,  2.03it/s]
 78%|███████▊  | 8350/10708 [1:45:19<19:21,  2.03it/s]
                                                      {'loss': 3.5394, 'grad_norm': 0.20390868186950684, 'learning_rate': 0.00014058936352983988, 'epoch': 0.78}

 78%|███████▊  | 8350/10708 [1:45:19<19:21,  2.03it/s]
 78%|███████▊  | 8351/10708 [1:45:19<19:21,  2.03it/s]
 78%|███████▊  | 8352/10708 [1:45:20<19:23,  2.03it/s]
 78%|███████▊  | 8353/10708 [1:45:20<19:21,  2.03it/s]
 78%|███████▊  | 8354/10708 [1:45:21<19:20,  2.03it/s]
 78%|███████▊  | 8355/10708 [1:45:21<19:19,  2.03it/s]
 78%|███████▊  | 8356/10708 [1:45:22<19:17,  2.03it/s]
 78%|███████▊  | 8357/10708 [1:45:22<19:16,  2.03it/s]
 78%|███████▊  | 8358/10708 [1:45:23<19:16,  2.03it/s]
 78%|███████▊  | 8359/10708 [1:45:23<19:15,  2.03it/s]
 78%|███████▊  | 8360/10708 [1:45:24<19:16,  2.03it/s]
 78%|███████▊  | 8361/10708 [1:45:24<19:17,  2.03it/s]
 78%|███████▊  | 8362/10708 [1:45:25<19:17,  2.03it/s]
 78%|███████▊  | 8363/10708 [1:45:25<19:15,  2.03it/s]
 78%|███████▊  | 8364/10708 [1:45:26<19:14,  2.03it/s]
 78%|███████▊  | 8365/10708 [1:45:26<19:13,  2.03it/s]
 78%|███████▊  | 8366/10708 [1:45:27<19:12,  2.03it/s]
 78%|███████▊  | 8367/10708 [1:45:27<19:12,  2.03it/s]
 78%|███████▊  | 8368/10708 [1:45:28<19:11,  2.03it/s]
 78%|███████▊  | 8369/10708 [1:45:28<19:08,  2.04it/s]
 78%|███████▊  | 8370/10708 [1:45:29<19:07,  2.04it/s]
 78%|███████▊  | 8371/10708 [1:45:29<19:08,  2.03it/s]
 78%|███████▊  | 8372/10708 [1:45:30<19:07,  2.04it/s]
 78%|███████▊  | 8373/10708 [1:45:30<19:06,  2.04it/s]
 78%|███████▊  | 8374/10708 [1:45:30<19:06,  2.04it/s]
 78%|███████▊  | 8375/10708 [1:45:31<19:05,  2.04it/s]{'loss': 3.5456, 'grad_norm': 0.20154650509357452, 'learning_rate': 0.0001377684740029586, 'epoch': 0.78}
                                                      

 78%|███████▊  | 8375/10708 [1:45:31<19:05,  2.04it/s]
 78%|███████▊  | 8376/10708 [1:45:31<19:08,  2.03it/s]
 78%|███████▊  | 8377/10708 [1:45:32<19:07,  2.03it/s]
 78%|███████▊  | 8378/10708 [1:45:32<19:06,  2.03it/s]
 78%|███████▊  | 8379/10708 [1:45:33<19:08,  2.03it/s]
 78%|███████▊  | 8380/10708 [1:45:33<19:06,  2.03it/s]
 78%|███████▊  | 8381/10708 [1:45:34<19:05,  2.03it/s]
 78%|███████▊  | 8382/10708 [1:45:34<19:03,  2.03it/s]
 78%|███████▊  | 8383/10708 [1:45:35<19:02,  2.04it/s]
 78%|███████▊  | 8384/10708 [1:45:35<19:03,  2.03it/s]
 78%|███████▊  | 8385/10708 [1:45:36<19:02,  2.03it/s]
 78%|███████▊  | 8386/10708 [1:45:36<19:02,  2.03it/s]
 78%|███████▊  | 8387/10708 [1:45:37<19:02,  2.03it/s]
 78%|███████▊  | 8388/10708 [1:45:37<19:00,  2.03it/s]
 78%|███████▊  | 8389/10708 [1:45:38<18:59,  2.04it/s]
 78%|███████▊  | 8390/10708 [1:45:38<19:00,  2.03it/s]
 78%|███████▊  | 8391/10708 [1:45:39<18:59,  2.03it/s]
 78%|███████▊  | 8392/10708 [1:45:39<18:57,  2.04it/s]
 78%|███████▊  | 8393/10708 [1:45:40<18:57,  2.03it/s]
 78%|███████▊  | 8394/10708 [1:45:40<18:57,  2.03it/s]
 78%|███████▊  | 8395/10708 [1:45:41<18:57,  2.03it/s]
 78%|███████▊  | 8396/10708 [1:45:41<18:57,  2.03it/s]
 78%|███████▊  | 8397/10708 [1:45:42<18:56,  2.03it/s]
 78%|███████▊  | 8398/10708 [1:45:42<18:55,  2.04it/s]
 78%|███████▊  | 8399/10708 [1:45:43<18:57,  2.03it/s]
 78%|███████▊  | 8400/10708 [1:45:43<18:59,  2.03it/s]
                                                      
{'loss': 3.5493, 'grad_norm': 0.20327302813529968, 'learning_rate': 0.00013497164360428288, 'epoch': 0.78}

 78%|███████▊  | 8400/10708 [1:45:43<18:59,  2.03it/s]
 78%|███████▊  | 8401/10708 [1:45:44<18:59,  2.02it/s]
 78%|███████▊  | 8402/10708 [1:45:44<18:57,  2.03it/s]
 78%|███████▊  | 8403/10708 [1:45:45<18:56,  2.03it/s]
 78%|███████▊  | 8404/10708 [1:45:45<18:56,  2.03it/s]
 78%|███████▊  | 8405/10708 [1:45:46<18:53,  2.03it/s]
 79%|███████▊  | 8406/10708 [1:45:46<18:53,  2.03it/s]
 79%|███████▊  | 8407/10708 [1:45:47<18:51,  2.03it/s]
 79%|███████▊  | 8408/10708 [1:45:47<18:52,  2.03it/s]
 79%|███████▊  | 8409/10708 [1:45:48<18:51,  2.03it/s]
 79%|███████▊  | 8410/10708 [1:45:48<18:49,  2.03it/s]
 79%|███████▊  | 8411/10708 [1:45:49<18:48,  2.04it/s]
 79%|███████▊  | 8412/10708 [1:45:49<18:48,  2.03it/s]
 79%|███████▊  | 8413/10708 [1:45:50<18:49,  2.03it/s]
 79%|███████▊  | 8414/10708 [1:45:50<18:47,  2.03it/s]
 79%|███████▊  | 8415/10708 [1:45:51<18:48,  2.03it/s]
 79%|███████▊  | 8416/10708 [1:45:51<18:46,  2.03it/s]
 79%|███████▊  | 8417/10708 [1:45:52<18:46,  2.03it/s]
 79%|███████▊  | 8418/10708 [1:45:52<18:46,  2.03it/s]
 79%|███████▊  | 8419/10708 [1:45:53<18:45,  2.03it/s]
 79%|███████▊  | 8420/10708 [1:45:53<18:45,  2.03it/s]
 79%|███████▊  | 8421/10708 [1:45:54<18:45,  2.03it/s]
 79%|███████▊  | 8422/10708 [1:45:54<18:43,  2.03it/s]
 79%|███████▊  | 8423/10708 [1:45:55<18:43,  2.03it/s]
 79%|███████▊  | 8424/10708 [1:45:55<18:43,  2.03it/s]
 79%|███████▊  | 8425/10708 [1:45:56<18:41,  2.04it/s]{'loss': 3.5443, 'grad_norm': 0.1998712122440338, 'learning_rate': 0.0001321990580970508, 'epoch': 0.79}

                                                      

 79%|███████▊  | 8425/10708 [1:45:56<18:41,  2.04it/s]
 79%|███████▊  | 8426/10708 [1:45:56<18:42,  2.03it/s]
 79%|███████▊  | 8427/10708 [1:45:57<18:42,  2.03it/s]
 79%|███████▊  | 8428/10708 [1:45:57<18:43,  2.03it/s]
 79%|███████▊  | 8429/10708 [1:45:58<18:42,  2.03it/s]
 79%|███████▊  | 8430/10708 [1:45:58<18:40,  2.03it/s]
 79%|███████▊  | 8431/10708 [1:45:59<18:40,  2.03it/s]
 79%|███████▊  | 8432/10708 [1:45:59<18:39,  2.03it/s]
 79%|███████▉  | 8433/10708 [1:46:00<18:40,  2.03it/s]
 79%|███████▉  | 8434/10708 [1:46:00<18:38,  2.03it/s]
 79%|███████▉  | 8435/10708 [1:46:00<18:37,  2.03it/s]
 79%|███████▉  | 8436/10708 [1:46:01<18:37,  2.03it/s]
 79%|███████▉  | 8437/10708 [1:46:01<18:35,  2.04it/s]
 79%|███████▉  | 8438/10708 [1:46:02<18:37,  2.03it/s]
 79%|███████▉  | 8439/10708 [1:46:02<18:36,  2.03it/s]
 79%|███████▉  | 8440/10708 [1:46:03<18:35,  2.03it/s]
 79%|███████▉  | 8441/10708 [1:46:03<18:36,  2.03it/s]
 79%|███████▉  | 8442/10708 [1:46:04<18:35,  2.03it/s]
 79%|███████▉  | 8443/10708 [1:46:04<18:34,  2.03it/s]
 79%|███████▉  | 8444/10708 [1:46:05<18:35,  2.03it/s]
 79%|███████▉  | 8445/10708 [1:46:05<18:34,  2.03it/s]
 79%|███████▉  | 8446/10708 [1:46:06<18:35,  2.03it/s]
 79%|███████▉  | 8447/10708 [1:46:06<18:33,  2.03it/s]
 79%|███████▉  | 8448/10708 [1:46:07<18:33,  2.03it/s]
 79%|███████▉  | 8449/10708 [1:46:07<18:33,  2.03it/s]
 79%|███████▉  | 8450/10708 [1:46:08<18:31,  2.03it/s]{'loss': 3.5397, 'grad_norm': 0.21266348659992218, 'learning_rate': 0.00012945090163417366, 'epoch': 0.79}

                                                      

 79%|███████▉  | 8450/10708 [1:46:08<18:31,  2.03it/s]
 79%|███████▉  | 8451/10708 [1:46:08<18:32,  2.03it/s]
 79%|███████▉  | 8452/10708 [1:46:09<18:31,  2.03it/s]
 79%|███████▉  | 8453/10708 [1:46:09<18:30,  2.03it/s]
 79%|███████▉  | 8454/10708 [1:46:10<18:31,  2.03it/s]
 79%|███████▉  | 8455/10708 [1:46:10<18:30,  2.03it/s]
 79%|███████▉  | 8456/10708 [1:46:11<18:28,  2.03it/s]
 79%|███████▉  | 8457/10708 [1:46:11<18:27,  2.03it/s]
 79%|███████▉  | 8458/10708 [1:46:12<18:29,  2.03it/s]
 79%|███████▉  | 8459/10708 [1:46:12<18:28,  2.03it/s]
 79%|███████▉  | 8460/10708 [1:46:13<18:30,  2.02it/s]
 79%|███████▉  | 8461/10708 [1:46:13<18:29,  2.03it/s]
 79%|███████▉  | 8462/10708 [1:46:14<18:26,  2.03it/s]
 79%|███████▉  | 8463/10708 [1:46:14<18:26,  2.03it/s]
 79%|███████▉  | 8464/10708 [1:46:15<18:25,  2.03it/s]
 79%|███████▉  | 8465/10708 [1:46:15<18:23,  2.03it/s]
 79%|███████▉  | 8466/10708 [1:46:16<18:23,  2.03it/s]
 79%|███████▉  | 8467/10708 [1:46:16<18:22,  2.03it/s]
 79%|███████▉  | 8468/10708 [1:46:17<18:21,  2.03it/s]
 79%|███████▉  | 8469/10708 [1:46:17<18:21,  2.03it/s]
 79%|███████▉  | 8470/10708 [1:46:18<18:20,  2.03it/s]
 79%|███████▉  | 8471/10708 [1:46:18<18:18,  2.04it/s]
 79%|███████▉  | 8472/10708 [1:46:19<18:20,  2.03it/s]
 79%|███████▉  | 8473/10708 [1:46:19<18:19,  2.03it/s]
 79%|███████▉  | 8474/10708 [1:46:20<18:19,  2.03it/s]
 79%|███████▉  | 8475/10708 [1:46:20<18:19,  2.03it/s]
                                                      {'loss': 3.5423, 'grad_norm': 0.2006271332502365, 'learning_rate': 0.00012672735674600576, 'epoch': 0.79}


 79%|███████▉  | 8475/10708 [1:46:20<18:19,  2.03it/s]
 79%|███████▉  | 8476/10708 [1:46:21<18:18,  2.03it/s]
 79%|███████▉  | 8477/10708 [1:46:21<18:17,  2.03it/s]
 79%|███████▉  | 8478/10708 [1:46:22<18:16,  2.03it/s]
 79%|███████▉  | 8479/10708 [1:46:22<18:15,  2.04it/s]
 79%|███████▉  | 8480/10708 [1:46:23<18:15,  2.03it/s]
 79%|███████▉  | 8481/10708 [1:46:23<18:15,  2.03it/s]
 79%|███████▉  | 8482/10708 [1:46:24<18:13,  2.04it/s]
 79%|███████▉  | 8483/10708 [1:46:24<18:14,  2.03it/s]
 79%|███████▉  | 8484/10708 [1:46:25<18:15,  2.03it/s]
 79%|███████▉  | 8485/10708 [1:46:25<18:15,  2.03it/s]
 79%|███████▉  | 8486/10708 [1:46:26<18:13,  2.03it/s]
 79%|███████▉  | 8487/10708 [1:46:26<18:12,  2.03it/s]
 79%|███████▉  | 8488/10708 [1:46:27<18:13,  2.03it/s]
 79%|███████▉  | 8489/10708 [1:46:27<18:13,  2.03it/s]
 79%|███████▉  | 8490/10708 [1:46:28<18:13,  2.03it/s]
 79%|███████▉  | 8491/10708 [1:46:28<18:11,  2.03it/s]
 79%|███████▉  | 8492/10708 [1:46:29<18:11,  2.03it/s]
 79%|███████▉  | 8493/10708 [1:46:29<18:11,  2.03it/s]
 79%|███████▉  | 8494/10708 [1:46:30<18:09,  2.03it/s]
 79%|███████▉  | 8495/10708 [1:46:30<18:10,  2.03it/s]
 79%|███████▉  | 8496/10708 [1:46:31<18:08,  2.03it/s]
 79%|███████▉  | 8497/10708 [1:46:31<18:07,  2.03it/s]
 79%|███████▉  | 8498/10708 [1:46:32<18:08,  2.03it/s]
 79%|███████▉  | 8499/10708 [1:46:32<18:07,  2.03it/s]
 79%|███████▉  | 8500/10708 [1:46:33<18:06,  2.03it/s]
                                                      {'loss': 3.542, 'grad_norm': 0.19671335816383362, 'learning_rate': 0.00012402860432821972, 'epoch': 0.79}


 79%|███████▉  | 8500/10708 [1:46:33<18:06,  2.03it/s]
 79%|███████▉  | 8501/10708 [1:46:33<18:07,  2.03it/s]
 79%|███████▉  | 8502/10708 [1:46:33<18:06,  2.03it/s]
 79%|███████▉  | 8503/10708 [1:46:34<18:06,  2.03it/s]
 79%|███████▉  | 8504/10708 [1:46:34<18:05,  2.03it/s]
 79%|███████▉  | 8505/10708 [1:46:35<18:05,  2.03it/s]
 79%|███████▉  | 8506/10708 [1:46:35<18:03,  2.03it/s]
 79%|███████▉  | 8507/10708 [1:46:36<18:02,  2.03it/s]
 79%|███████▉  | 8508/10708 [1:46:36<18:03,  2.03it/s]
 79%|███████▉  | 8509/10708 [1:46:37<18:04,  2.03it/s]
 79%|███████▉  | 8510/10708 [1:46:37<18:02,  2.03it/s]
 79%|███████▉  | 8511/10708 [1:46:38<18:00,  2.03it/s]
 79%|███████▉  | 8512/10708 [1:46:38<18:00,  2.03it/s]
 80%|███████▉  | 8513/10708 [1:46:39<18:02,  2.03it/s]
 80%|███████▉  | 8514/10708 [1:46:39<17:59,  2.03it/s]
 80%|███████▉  | 8515/10708 [1:46:40<17:59,  2.03it/s]
 80%|███████▉  | 8516/10708 [1:46:40<17:57,  2.03it/s]
 80%|███████▉  | 8517/10708 [1:46:41<17:57,  2.03it/s]
 80%|███████▉  | 8518/10708 [1:46:41<17:55,  2.04it/s]
 80%|███████▉  | 8519/10708 [1:46:42<17:55,  2.03it/s]
 80%|███████▉  | 8520/10708 [1:46:42<17:55,  2.03it/s]
 80%|███████▉  | 8521/10708 [1:46:43<17:57,  2.03it/s]
 80%|███████▉  | 8522/10708 [1:46:43<17:58,  2.03it/s]
 80%|███████▉  | 8523/10708 [1:46:44<17:58,  2.03it/s]
 80%|███████▉  | 8524/10708 [1:46:44<17:57,  2.03it/s]
 80%|███████▉  | 8525/10708 [1:46:45<17:56,  2.03it/s]
                                                      
{'loss': 3.5433, 'grad_norm': 0.20292304456233978, 'learning_rate': 0.0001213548236297925, 'epoch': 0.8}

 80%|███████▉  | 8525/10708 [1:46:45<17:56,  2.03it/s]
 80%|███████▉  | 8526/10708 [1:46:45<17:56,  2.03it/s]
 80%|███████▉  | 8527/10708 [1:46:46<17:55,  2.03it/s]
 80%|███████▉  | 8528/10708 [1:46:46<17:53,  2.03it/s]
 80%|███████▉  | 8529/10708 [1:46:47<17:52,  2.03it/s]
 80%|███████▉  | 8530/10708 [1:46:47<17:51,  2.03it/s]
 80%|███████▉  | 8531/10708 [1:46:48<17:50,  2.03it/s]
 80%|███████▉  | 8532/10708 [1:46:48<17:50,  2.03it/s]
 80%|███████▉  | 8533/10708 [1:46:49<17:50,  2.03it/s]
 80%|███████▉  | 8534/10708 [1:46:49<17:51,  2.03it/s]
 80%|███████▉  | 8535/10708 [1:46:50<17:50,  2.03it/s]
 80%|███████▉  | 8536/10708 [1:46:50<17:48,  2.03it/s]
 80%|███████▉  | 8537/10708 [1:46:51<17:48,  2.03it/s]
 80%|███████▉  | 8538/10708 [1:46:51<17:47,  2.03it/s]
 80%|███████▉  | 8539/10708 [1:46:52<17:46,  2.03it/s]
 80%|███████▉  | 8540/10708 [1:46:52<17:47,  2.03it/s]
 80%|███████▉  | 8541/10708 [1:46:53<17:47,  2.03it/s]
 80%|███████▉  | 8542/10708 [1:46:53<17:45,  2.03it/s]
 80%|███████▉  | 8543/10708 [1:46:54<17:45,  2.03it/s]
 80%|███████▉  | 8544/10708 [1:46:54<17:44,  2.03it/s]
 80%|███████▉  | 8545/10708 [1:46:55<17:43,  2.03it/s]
 80%|███████▉  | 8546/10708 [1:46:55<17:43,  2.03it/s]
 80%|███████▉  | 8547/10708 [1:46:56<17:42,  2.03it/s]
 80%|███████▉  | 8548/10708 [1:46:56<17:41,  2.04it/s]
 80%|███████▉  | 8549/10708 [1:46:57<17:41,  2.03it/s]
 80%|███████▉  | 8550/10708 [1:46:57<17:40,  2.03it/s]{'loss': 3.5377, 'grad_norm': 0.20108270645141602, 'learning_rate': 0.00011870619224109935, 'epoch': 0.8}

                                                      

 80%|███████▉  | 8550/10708 [1:46:57<17:40,  2.03it/s]
 80%|███████▉  | 8551/10708 [1:46:58<17:40,  2.03it/s]
 80%|███████▉  | 8552/10708 [1:46:58<17:40,  2.03it/s]
 80%|███████▉  | 8553/10708 [1:46:59<17:40,  2.03it/s]
 80%|███████▉  | 8554/10708 [1:46:59<17:39,  2.03it/s]
 80%|███████▉  | 8555/10708 [1:47:00<17:39,  2.03it/s]
 80%|███████▉  | 8556/10708 [1:47:00<17:38,  2.03it/s]
 80%|███████▉  | 8557/10708 [1:47:01<17:36,  2.04it/s]
 80%|███████▉  | 8558/10708 [1:47:01<17:38,  2.03it/s]
 80%|███████▉  | 8559/10708 [1:47:02<17:36,  2.03it/s]
 80%|███████▉  | 8560/10708 [1:47:02<17:36,  2.03it/s]
 80%|███████▉  | 8561/10708 [1:47:03<17:36,  2.03it/s]
 80%|███████▉  | 8562/10708 [1:47:03<17:34,  2.03it/s]
 80%|███████▉  | 8563/10708 [1:47:04<17:36,  2.03it/s]
 80%|███████▉  | 8564/10708 [1:47:04<17:34,  2.03it/s]
 80%|███████▉  | 8565/10708 [1:47:04<17:33,  2.03it/s]
 80%|███████▉  | 8566/10708 [1:47:05<17:34,  2.03it/s]
 80%|████████  | 8567/10708 [1:47:05<17:32,  2.03it/s]
 80%|████████  | 8568/10708 [1:47:06<17:32,  2.03it/s]
 80%|████████  | 8569/10708 [1:47:06<17:32,  2.03it/s]
 80%|████████  | 8570/10708 [1:47:07<17:31,  2.03it/s]
 80%|████████  | 8571/10708 [1:47:07<17:31,  2.03it/s]
 80%|████████  | 8572/10708 [1:47:08<17:29,  2.03it/s]
 80%|████████  | 8573/10708 [1:47:08<17:28,  2.04it/s]
 80%|████████  | 8574/10708 [1:47:09<17:29,  2.03it/s]
 80%|████████  | 8575/10708 [1:47:09<17:30,  2.03it/s]
                                                      
{'loss': 3.5447, 'grad_norm': 0.20357897877693176, 'learning_rate': 0.00011608288608211853, 'epoch': 0.8}

 80%|████████  | 8575/10708 [1:47:09<17:30,  2.03it/s]
 80%|████████  | 8576/10708 [1:47:10<17:30,  2.03it/s]
 80%|████████  | 8577/10708 [1:47:10<17:29,  2.03it/s]
 80%|████████  | 8578/10708 [1:47:11<17:28,  2.03it/s]
 80%|████████  | 8579/10708 [1:47:11<17:27,  2.03it/s]
 80%|████████  | 8580/10708 [1:47:12<17:27,  2.03it/s]
 80%|████████  | 8581/10708 [1:47:12<17:25,  2.03it/s]
 80%|████████  | 8582/10708 [1:47:13<17:27,  2.03it/s]
 80%|████████  | 8583/10708 [1:47:13<17:27,  2.03it/s]
 80%|████████  | 8584/10708 [1:47:14<17:26,  2.03it/s]
 80%|████████  | 8585/10708 [1:47:14<17:26,  2.03it/s]
 80%|████████  | 8586/10708 [1:47:15<17:25,  2.03it/s]
 80%|████████  | 8587/10708 [1:47:15<17:24,  2.03it/s]
 80%|████████  | 8588/10708 [1:47:16<17:24,  2.03it/s]
 80%|████████  | 8589/10708 [1:47:16<17:24,  2.03it/s]
 80%|████████  | 8590/10708 [1:47:17<17:22,  2.03it/s]
 80%|████████  | 8591/10708 [1:47:17<17:23,  2.03it/s]
 80%|████████  | 8592/10708 [1:47:18<17:21,  2.03it/s]
 80%|████████  | 8593/10708 [1:47:18<17:21,  2.03it/s]
 80%|████████  | 8594/10708 [1:47:19<17:21,  2.03it/s]
 80%|████████  | 8595/10708 [1:47:19<17:20,  2.03it/s]
 80%|████████  | 8596/10708 [1:47:20<17:18,  2.03it/s]
 80%|████████  | 8597/10708 [1:47:20<17:19,  2.03it/s]
 80%|████████  | 8598/10708 [1:47:21<17:18,  2.03it/s]
 80%|████████  | 8599/10708 [1:47:21<17:16,  2.04it/s]
 80%|████████  | 8600/10708 [1:47:22<17:17,  2.03it/s]{'loss': 3.5401, 'grad_norm': 0.2022303193807602, 'learning_rate': 0.00011348507939074654, 'epoch': 0.8}

                                                      

 80%|████████  | 8600/10708 [1:47:22<17:17,  2.03it/s]
 80%|████████  | 8601/10708 [1:47:22<17:17,  2.03it/s]
 80%|████████  | 8602/10708 [1:47:23<17:17,  2.03it/s]
 80%|████████  | 8603/10708 [1:47:23<17:17,  2.03it/s]
 80%|████████  | 8604/10708 [1:47:24<17:16,  2.03it/s]
 80%|████████  | 8605/10708 [1:47:24<17:15,  2.03it/s]
 80%|████████  | 8606/10708 [1:47:25<17:14,  2.03it/s]
 80%|████████  | 8607/10708 [1:47:25<17:13,  2.03it/s]
 80%|████████  | 8608/10708 [1:47:26<17:13,  2.03it/s]
 80%|████████  | 8609/10708 [1:47:26<17:12,  2.03it/s]
 80%|████████  | 8610/10708 [1:47:27<17:12,  2.03it/s]
 80%|████████  | 8611/10708 [1:47:27<17:12,  2.03it/s]
 80%|████████  | 8612/10708 [1:47:28<17:10,  2.03it/s]
 80%|████████  | 8613/10708 [1:47:28<17:11,  2.03it/s]
 80%|████████  | 8614/10708 [1:47:29<17:10,  2.03it/s]
 80%|████████  | 8615/10708 [1:47:29<17:08,  2.03it/s]
 80%|████████  | 8616/10708 [1:47:30<17:08,  2.03it/s]
 80%|████████  | 8617/10708 [1:47:30<17:07,  2.03it/s]
 80%|████████  | 8618/10708 [1:47:31<17:07,  2.03it/s]
 80%|████████  | 8619/10708 [1:47:31<17:07,  2.03it/s]
 81%|████████  | 8620/10708 [1:47:32<17:06,  2.03it/s]
 81%|████████  | 8621/10708 [1:47:32<17:05,  2.04it/s]
 81%|████████  | 8622/10708 [1:47:33<17:04,  2.04it/s]
 81%|████████  | 8623/10708 [1:47:33<17:06,  2.03it/s]
 81%|████████  | 8624/10708 [1:47:34<17:04,  2.03it/s]
 81%|████████  | 8625/10708 [1:47:34<17:04,  2.03it/s]{'loss': 3.5411, 'grad_norm': 0.19806449115276337, 'learning_rate': 0.00011091294471122598, 'epoch': 0.81}
                                                      

 81%|████████  | 8625/10708 [1:47:34<17:04,  2.03it/s]
 81%|████████  | 8626/10708 [1:47:35<17:04,  2.03it/s]
 81%|████████  | 8627/10708 [1:47:35<17:05,  2.03it/s]
 81%|████████  | 8628/10708 [1:47:36<17:04,  2.03it/s]
 81%|████████  | 8629/10708 [1:47:36<17:02,  2.03it/s]
 81%|████████  | 8630/10708 [1:47:36<17:02,  2.03it/s]
 81%|████████  | 8631/10708 [1:47:37<17:02,  2.03it/s]
 81%|████████  | 8632/10708 [1:47:37<17:00,  2.03it/s]
 81%|████████  | 8633/10708 [1:47:38<17:00,  2.03it/s]
 81%|████████  | 8634/10708 [1:47:38<16:59,  2.03it/s]
 81%|████████  | 8635/10708 [1:47:39<16:59,  2.03it/s]
 81%|████████  | 8636/10708 [1:47:39<16:58,  2.03it/s]
 81%|████████  | 8637/10708 [1:47:40<16:58,  2.03it/s]
 81%|████████  | 8638/10708 [1:47:40<16:57,  2.04it/s]
 81%|████████  | 8639/10708 [1:47:41<16:57,  2.03it/s]
 81%|████████  | 8640/10708 [1:47:41<16:56,  2.03it/s]
 81%|████████  | 8641/10708 [1:47:42<16:56,  2.03it/s]
 81%|████████  | 8642/10708 [1:47:42<16:56,  2.03it/s]
 81%|████████  | 8643/10708 [1:47:43<16:59,  2.03it/s]
 81%|████████  | 8644/10708 [1:47:43<16:58,  2.03it/s]
 81%|████████  | 8645/10708 [1:47:44<16:57,  2.03it/s]
 81%|████████  | 8646/10708 [1:47:44<16:56,  2.03it/s]
 81%|████████  | 8647/10708 [1:47:45<16:54,  2.03it/s]
 81%|████████  | 8648/10708 [1:47:45<16:53,  2.03it/s]
 81%|████████  | 8649/10708 [1:47:46<16:53,  2.03it/s]
 81%|████████  | 8650/10708 [1:47:46<16:51,  2.03it/s]
                                                      {'loss': 3.532, 'grad_norm': 0.19799315929412842, 'learning_rate': 0.0001083666528826846, 'epoch': 0.81}


 81%|████████  | 8650/10708 [1:47:46<16:51,  2.03it/s]
 81%|████████  | 8651/10708 [1:47:47<16:52,  2.03it/s]
 81%|████████  | 8652/10708 [1:47:47<16:51,  2.03it/s]
 81%|████████  | 8653/10708 [1:47:48<16:52,  2.03it/s]
 81%|████████  | 8654/10708 [1:47:48<16:51,  2.03it/s]
 81%|████████  | 8655/10708 [1:47:49<16:51,  2.03it/s]
 81%|████████  | 8656/10708 [1:47:49<16:50,  2.03it/s]
 81%|████████  | 8657/10708 [1:47:50<16:50,  2.03it/s]
 81%|████████  | 8658/10708 [1:47:50<16:50,  2.03it/s]
 81%|████████  | 8659/10708 [1:47:51<16:49,  2.03it/s]
 81%|████████  | 8660/10708 [1:47:51<16:47,  2.03it/s]
 81%|████████  | 8661/10708 [1:47:52<16:46,  2.03it/s]
 81%|████████  | 8662/10708 [1:47:52<16:46,  2.03it/s]
 81%|████████  | 8663/10708 [1:47:53<16:46,  2.03it/s]
 81%|████████  | 8664/10708 [1:47:53<16:45,  2.03it/s]
 81%|████████  | 8665/10708 [1:47:54<16:45,  2.03it/s]
 81%|████████  | 8666/10708 [1:47:54<16:44,  2.03it/s]
 81%|████████  | 8667/10708 [1:47:55<16:44,  2.03it/s]
 81%|████████  | 8668/10708 [1:47:55<16:44,  2.03it/s]
 81%|████████  | 8669/10708 [1:47:56<16:43,  2.03it/s]
 81%|████████  | 8670/10708 [1:47:56<16:43,  2.03it/s]
 81%|████████  | 8671/10708 [1:47:57<16:42,  2.03it/s]
 81%|████████  | 8672/10708 [1:47:57<16:41,  2.03it/s]
 81%|████████  | 8673/10708 [1:47:58<16:43,  2.03it/s]
 81%|████████  | 8674/10708 [1:47:58<16:41,  2.03it/s]
 81%|████████  | 8675/10708 [1:47:59<16:41,  2.03it/s]
                                                      
{'loss': 3.5374, 'grad_norm': 0.2083081752061844, 'learning_rate': 0.0001058463730277891, 'epoch': 0.81}

 81%|████████  | 8675/10708 [1:47:59<16:41,  2.03it/s]
 81%|████████  | 8676/10708 [1:47:59<16:40,  2.03it/s]
 81%|████████  | 8677/10708 [1:48:00<16:39,  2.03it/s]
 81%|████████  | 8678/10708 [1:48:00<16:39,  2.03it/s]
 81%|████████  | 8679/10708 [1:48:01<16:38,  2.03it/s]
 81%|████████  | 8680/10708 [1:48:01<16:38,  2.03it/s]
 81%|████████  | 8681/10708 [1:48:02<16:38,  2.03it/s]
 81%|████████  | 8682/10708 [1:48:02<16:37,  2.03it/s]
 81%|████████  | 8683/10708 [1:48:03<16:38,  2.03it/s]
 81%|████████  | 8684/10708 [1:48:03<16:36,  2.03it/s]
 81%|████████  | 8685/10708 [1:48:04<16:35,  2.03it/s]
 81%|████████  | 8686/10708 [1:48:04<16:35,  2.03it/s]
 81%|████████  | 8687/10708 [1:48:05<16:34,  2.03it/s]
 81%|████████  | 8688/10708 [1:48:05<16:33,  2.03it/s]
 81%|████████  | 8689/10708 [1:48:06<16:33,  2.03it/s]
 81%|████████  | 8690/10708 [1:48:06<16:32,  2.03it/s]
 81%|████████  | 8691/10708 [1:48:07<16:31,  2.03it/s]
 81%|████████  | 8692/10708 [1:48:07<16:31,  2.03it/s]
 81%|████████  | 8693/10708 [1:48:07<16:31,  2.03it/s]
 81%|████████  | 8694/10708 [1:48:08<16:29,  2.03it/s]
 81%|████████  | 8695/10708 [1:48:08<16:30,  2.03it/s]
 81%|████████  | 8696/10708 [1:48:09<16:29,  2.03it/s]
 81%|████████  | 8697/10708 [1:48:09<16:28,  2.03it/s]
 81%|████████  | 8698/10708 [1:48:10<16:28,  2.03it/s]
 81%|████████  | 8699/10708 [1:48:10<16:27,  2.04it/s]
 81%|████████  | 8700/10708 [1:48:11<16:28,  2.03it/s]{'loss': 3.5414, 'grad_norm': 0.20587189495563507, 'learning_rate': 0.00010335227254151125, 'epoch': 0.81}
                                                      

 81%|████████  | 8700/10708 [1:48:11<16:28,  2.03it/s]
 81%|████████▏ | 8701/10708 [1:48:11<16:29,  2.03it/s]
 81%|████████▏ | 8702/10708 [1:48:12<16:26,  2.03it/s]
 81%|████████▏ | 8703/10708 [1:48:12<16:26,  2.03it/s]
 81%|████████▏ | 8704/10708 [1:48:13<16:27,  2.03it/s]
 81%|████████▏ | 8705/10708 [1:48:13<16:28,  2.03it/s]
 81%|████████▏ | 8706/10708 [1:48:14<16:26,  2.03it/s]
 81%|████████▏ | 8707/10708 [1:48:14<16:25,  2.03it/s]
 81%|████████▏ | 8708/10708 [1:48:15<16:24,  2.03it/s]
 81%|████████▏ | 8709/10708 [1:48:15<16:23,  2.03it/s]
 81%|████████▏ | 8710/10708 [1:48:16<16:23,  2.03it/s]
 81%|████████▏ | 8711/10708 [1:48:16<16:23,  2.03it/s]
 81%|████████▏ | 8712/10708 [1:48:17<16:22,  2.03it/s]
 81%|████████▏ | 8713/10708 [1:48:17<16:23,  2.03it/s]
 81%|████████▏ | 8714/10708 [1:48:18<16:22,  2.03it/s]
 81%|████████▏ | 8715/10708 [1:48:18<16:21,  2.03it/s]
 81%|████████▏ | 8716/10708 [1:48:19<16:20,  2.03it/s]
 81%|████████▏ | 8717/10708 [1:48:19<16:19,  2.03it/s]
 81%|████████▏ | 8718/10708 [1:48:20<16:19,  2.03it/s]
 81%|████████▏ | 8719/10708 [1:48:20<16:19,  2.03it/s]
 81%|████████▏ | 8720/10708 [1:48:21<16:17,  2.03it/s]
 81%|████████▏ | 8721/10708 [1:48:21<16:18,  2.03it/s]
 81%|████████▏ | 8722/10708 [1:48:22<16:17,  2.03it/s]
 81%|████████▏ | 8723/10708 [1:48:22<16:15,  2.03it/s]
 81%|████████▏ | 8724/10708 [1:48:23<16:15,  2.03it/s]
 81%|████████▏ | 8725/10708 [1:48:23<16:14,  2.03it/s]
                                                      
{'loss': 3.5301, 'grad_norm': 0.2102261483669281, 'learning_rate': 0.00010088451708001034, 'epoch': 0.81}

 81%|████████▏ | 8725/10708 [1:48:23<16:14,  2.03it/s]
 81%|████████▏ | 8726/10708 [1:48:24<16:14,  2.03it/s]
 81%|████████▏ | 8727/10708 [1:48:24<16:15,  2.03it/s]
 82%|████████▏ | 8728/10708 [1:48:25<16:14,  2.03it/s]
 82%|████████▏ | 8729/10708 [1:48:25<16:15,  2.03it/s]
 82%|████████▏ | 8730/10708 [1:48:26<16:15,  2.03it/s]
 82%|████████▏ | 8731/10708 [1:48:26<16:13,  2.03it/s]
 82%|████████▏ | 8732/10708 [1:48:27<16:12,  2.03it/s]
 82%|████████▏ | 8733/10708 [1:48:27<16:10,  2.04it/s]
 82%|████████▏ | 8734/10708 [1:48:28<16:10,  2.03it/s]
 82%|████████▏ | 8735/10708 [1:48:28<16:09,  2.04it/s]
 82%|████████▏ | 8736/10708 [1:48:29<16:08,  2.04it/s]
 82%|████████▏ | 8737/10708 [1:48:29<16:07,  2.04it/s]
 82%|████████▏ | 8738/10708 [1:48:30<16:09,  2.03it/s]
 82%|████████▏ | 8739/10708 [1:48:30<16:07,  2.03it/s]
 82%|████████▏ | 8740/10708 [1:48:31<16:06,  2.04it/s]
 82%|████████▏ | 8741/10708 [1:48:31<16:07,  2.03it/s]
 82%|████████▏ | 8742/10708 [1:48:32<16:07,  2.03it/s]
 82%|████████▏ | 8743/10708 [1:48:32<16:06,  2.03it/s]
 82%|████████▏ | 8744/10708 [1:48:33<16:06,  2.03it/s]
 82%|████████▏ | 8745/10708 [1:48:33<16:05,  2.03it/s]
 82%|████████▏ | 8746/10708 [1:48:34<16:06,  2.03it/s]
 82%|████████▏ | 8747/10708 [1:48:34<16:03,  2.03it/s]
 82%|████████▏ | 8748/10708 [1:48:35<16:03,  2.03it/s]
 82%|████████▏ | 8749/10708 [1:48:35<16:03,  2.03it/s]
 82%|████████▏ | 8750/10708 [1:48:36<16:02,  2.04it/s]
                                                      {'loss': 3.5401, 'grad_norm': 0.1963588297367096, 'learning_rate': 9.84432705496302e-05, 'epoch': 0.82}

 82%|████████▏ | 8750/10708 [1:48:36<16:02,  2.04it/s]
 82%|████████▏ | 8751/10708 [1:48:36<16:03,  2.03it/s]
 82%|████████▏ | 8752/10708 [1:48:37<16:02,  2.03it/s]
 82%|████████▏ | 8753/10708 [1:48:37<16:01,  2.03it/s]
 82%|████████▏ | 8754/10708 [1:48:38<16:01,  2.03it/s]
 82%|████████▏ | 8755/10708 [1:48:38<16:00,  2.03it/s]
 82%|████████▏ | 8756/10708 [1:48:38<15:58,  2.04it/s]
 82%|████████▏ | 8757/10708 [1:48:39<15:59,  2.03it/s]
 82%|████████▏ | 8758/10708 [1:48:39<15:59,  2.03it/s]
 82%|████████▏ | 8759/10708 [1:48:40<15:58,  2.03it/s]
 82%|████████▏ | 8760/10708 [1:48:40<15:59,  2.03it/s]
 82%|████████▏ | 8761/10708 [1:48:41<15:58,  2.03it/s]
 82%|████████▏ | 8762/10708 [1:48:41<15:58,  2.03it/s]
 82%|████████▏ | 8763/10708 [1:48:42<15:58,  2.03it/s]
 82%|████████▏ | 8764/10708 [1:48:42<15:58,  2.03it/s]
 82%|████████▏ | 8765/10708 [1:48:43<15:59,  2.02it/s]
 82%|████████▏ | 8766/10708 [1:48:43<15:59,  2.02it/s]
 82%|████████▏ | 8767/10708 [1:48:44<15:57,  2.03it/s]
 82%|████████▏ | 8768/10708 [1:48:44<15:54,  2.03it/s]
 82%|████████▏ | 8769/10708 [1:48:45<15:55,  2.03it/s]
 82%|████████▏ | 8770/10708 [1:48:45<15:54,  2.03it/s]
 82%|████████▏ | 8771/10708 [1:48:46<15:53,  2.03it/s]
 82%|████████▏ | 8772/10708 [1:48:46<15:53,  2.03it/s]
 82%|████████▏ | 8773/10708 [1:48:47<15:53,  2.03it/s]
 82%|████████▏ | 8774/10708 [1:48:47<15:52,  2.03it/s]
 82%|████████▏ | 8775/10708 [1:48:48<15:53,  2.03it/s]
                                                      
{'loss': 3.5395, 'grad_norm': 0.1970537006855011, 'learning_rate': 9.602869509601275e-05, 'epoch': 0.82}

 82%|████████▏ | 8775/10708 [1:48:48<15:53,  2.03it/s]
 82%|████████▏ | 8776/10708 [1:48:48<15:54,  2.02it/s]
 82%|████████▏ | 8777/10708 [1:48:49<15:52,  2.03it/s]
 82%|████████▏ | 8778/10708 [1:48:49<15:52,  2.03it/s]
 82%|████████▏ | 8779/10708 [1:48:50<15:51,  2.03it/s]
 82%|████████▏ | 8780/10708 [1:48:50<15:49,  2.03it/s]
 82%|████████▏ | 8781/10708 [1:48:51<15:48,  2.03it/s]
 82%|████████▏ | 8782/10708 [1:48:51<15:47,  2.03it/s]
 82%|████████▏ | 8783/10708 [1:48:52<15:46,  2.03it/s]
 82%|████████▏ | 8784/10708 [1:48:52<15:46,  2.03it/s]
 82%|████████▏ | 8785/10708 [1:48:53<15:45,  2.03it/s]
 82%|████████▏ | 8786/10708 [1:48:53<15:44,  2.04it/s]
 82%|████████▏ | 8787/10708 [1:48:54<15:44,  2.03it/s]
 82%|████████▏ | 8788/10708 [1:48:54<15:43,  2.03it/s]
 82%|████████▏ | 8789/10708 [1:48:55<15:43,  2.03it/s]
 82%|████████▏ | 8790/10708 [1:48:55<15:43,  2.03it/s]
 82%|████████▏ | 8791/10708 [1:48:56<15:43,  2.03it/s]
 82%|████████▏ | 8792/10708 [1:48:56<15:43,  2.03it/s]
 82%|████████▏ | 8793/10708 [1:48:57<15:42,  2.03it/s]
 82%|████████▏ | 8794/10708 [1:48:57<15:41,  2.03it/s]
 82%|████████▏ | 8795/10708 [1:48:58<15:40,  2.03it/s]
 82%|████████▏ | 8796/10708 [1:48:58<15:38,  2.04it/s]
 82%|████████▏ | 8797/10708 [1:48:59<15:41,  2.03it/s]
 82%|████████▏ | 8798/10708 [1:48:59<15:40,  2.03it/s]
 82%|████████▏ | 8799/10708 [1:49:00<15:40,  2.03it/s]
 82%|████████▏ | 8800/10708 [1:49:00<15:40,  2.03it/s]{'loss': 3.5343, 'grad_norm': 0.20690546929836273, 'learning_rate': 9.364095109332793e-05, 'epoch': 0.82}
                                                      

 82%|████████▏ | 8800/10708 [1:49:00<15:40,  2.03it/s]
 82%|████████▏ | 8801/10708 [1:49:01<15:39,  2.03it/s]
 82%|████████▏ | 8802/10708 [1:49:01<15:38,  2.03it/s]
 82%|████████▏ | 8803/10708 [1:49:02<15:37,  2.03it/s]
 82%|████████▏ | 8804/10708 [1:49:02<15:36,  2.03it/s]
 82%|████████▏ | 8805/10708 [1:49:03<15:35,  2.03it/s]
 82%|████████▏ | 8806/10708 [1:49:03<15:34,  2.04it/s]
 82%|████████▏ | 8807/10708 [1:49:04<15:34,  2.04it/s]
 82%|████████▏ | 8808/10708 [1:49:04<15:34,  2.03it/s]
 82%|████████▏ | 8809/10708 [1:49:05<15:34,  2.03it/s]
 82%|████████▏ | 8810/10708 [1:49:05<15:34,  2.03it/s]
 82%|████████▏ | 8811/10708 [1:49:06<15:33,  2.03it/s]
 82%|████████▏ | 8812/10708 [1:49:06<15:32,  2.03it/s]
 82%|████████▏ | 8813/10708 [1:49:07<15:32,  2.03it/s]
 82%|████████▏ | 8814/10708 [1:49:07<15:32,  2.03it/s]
 82%|████████▏ | 8815/10708 [1:49:08<15:32,  2.03it/s]
 82%|████████▏ | 8816/10708 [1:49:08<15:30,  2.03it/s]
 82%|████████▏ | 8817/10708 [1:49:09<15:30,  2.03it/s]
 82%|████████▏ | 8818/10708 [1:49:09<15:28,  2.03it/s]
 82%|████████▏ | 8819/10708 [1:49:10<15:27,  2.04it/s]
 82%|████████▏ | 8820/10708 [1:49:10<15:28,  2.03it/s]
 82%|████████▏ | 8821/10708 [1:49:10<15:27,  2.03it/s]
 82%|████████▏ | 8822/10708 [1:49:11<15:26,  2.04it/s]
 82%|████████▏ | 8823/10708 [1:49:12<18:00,  1.75it/s]
 82%|████████▏ | 8824/10708 [1:49:12<17:12,  1.82it/s]
 82%|████████▏ | 8825/10708 [1:49:13<16:42,  1.88it/s]{'loss': 3.5371, 'grad_norm': 0.2250635176897049, 'learning_rate': 9.128019713362262e-05, 'epoch': 0.82}

                                                      

 82%|████████▏ | 8825/10708 [1:49:13<16:42,  1.88it/s]
 82%|████████▏ | 8826/10708 [1:49:13<16:21,  1.92it/s]
 82%|████████▏ | 8827/10708 [1:49:14<16:03,  1.95it/s]
 82%|████████▏ | 8828/10708 [1:49:14<15:52,  1.97it/s]
 82%|████████▏ | 8829/10708 [1:49:15<15:42,  1.99it/s]
 82%|████████▏ | 8830/10708 [1:49:15<15:37,  2.00it/s]
 82%|████████▏ | 8831/10708 [1:49:16<15:32,  2.01it/s]
 82%|████████▏ | 8832/10708 [1:49:16<15:28,  2.02it/s]
 82%|████████▏ | 8833/10708 [1:49:17<15:27,  2.02it/s]
 82%|████████▏ | 8834/10708 [1:49:17<15:24,  2.03it/s]
 83%|████████▎ | 8835/10708 [1:49:18<15:23,  2.03it/s]
 83%|████████▎ | 8836/10708 [1:49:18<15:23,  2.03it/s]
 83%|████████▎ | 8837/10708 [1:49:19<15:20,  2.03it/s]
 83%|████████▎ | 8838/10708 [1:49:19<15:19,  2.03it/s]
 83%|████████▎ | 8839/10708 [1:49:20<15:19,  2.03it/s]
 83%|████████▎ | 8840/10708 [1:49:20<15:19,  2.03it/s]
 83%|████████▎ | 8841/10708 [1:49:21<15:18,  2.03it/s]
 83%|████████▎ | 8842/10708 [1:49:21<15:17,  2.03it/s]
 83%|████████▎ | 8843/10708 [1:49:22<15:16,  2.03it/s]
 83%|████████▎ | 8844/10708 [1:49:22<15:16,  2.03it/s]
 83%|████████▎ | 8845/10708 [1:49:23<15:16,  2.03it/s]
 83%|████████▎ | 8846/10708 [1:49:23<15:15,  2.03it/s]
 83%|████████▎ | 8847/10708 [1:49:24<15:15,  2.03it/s]
 83%|████████▎ | 8848/10708 [1:49:24<15:13,  2.04it/s]
 83%|████████▎ | 8849/10708 [1:49:25<15:12,  2.04it/s]
 83%|████████▎ | 8850/10708 [1:49:25<15:13,  2.03it/s]{'loss': 3.5361, 'grad_norm': 0.202717125415802, 'learning_rate': 8.894659001628641e-05, 'epoch': 0.83}

                                                      

 83%|████████▎ | 8850/10708 [1:49:25<15:13,  2.03it/s]
 83%|████████▎ | 8851/10708 [1:49:26<15:13,  2.03it/s]
 83%|████████▎ | 8852/10708 [1:49:26<15:13,  2.03it/s]
 83%|████████▎ | 8853/10708 [1:49:27<15:12,  2.03it/s]
 83%|████████▎ | 8854/10708 [1:49:27<15:11,  2.03it/s]
 83%|████████▎ | 8855/10708 [1:49:27<15:11,  2.03it/s]
 83%|████████▎ | 8856/10708 [1:49:28<15:11,  2.03it/s]
 83%|████████▎ | 8857/10708 [1:49:28<15:09,  2.03it/s]
 83%|████████▎ | 8858/10708 [1:49:29<15:09,  2.03it/s]
 83%|████████▎ | 8859/10708 [1:49:29<15:08,  2.03it/s]
 83%|████████▎ | 8860/10708 [1:49:30<17:36,  1.75it/s]
 83%|████████▎ | 8861/10708 [1:49:31<16:50,  1.83it/s]
 83%|████████▎ | 8862/10708 [1:49:31<16:21,  1.88it/s]
 83%|████████▎ | 8863/10708 [1:49:32<15:58,  1.93it/s]
 83%|████████▎ | 8864/10708 [1:49:32<15:44,  1.95it/s]
 83%|████████▎ | 8865/10708 [1:49:33<15:32,  1.98it/s]
 83%|████████▎ | 8866/10708 [1:49:33<15:23,  1.99it/s]
 83%|████████▎ | 8867/10708 [1:49:34<15:16,  2.01it/s]
 83%|████████▎ | 8868/10708 [1:49:34<15:13,  2.01it/s]
 83%|████████▎ | 8869/10708 [1:49:35<15:10,  2.02it/s]
 83%|████████▎ | 8870/10708 [1:49:35<15:07,  2.03it/s]
 83%|████████▎ | 8871/10708 [1:49:36<15:05,  2.03it/s]
 83%|████████▎ | 8872/10708 [1:49:36<15:04,  2.03it/s]
 83%|████████▎ | 8873/10708 [1:49:37<15:03,  2.03it/s]
 83%|████████▎ | 8874/10708 [1:49:37<15:01,  2.04it/s]
 83%|████████▎ | 8875/10708 [1:49:38<15:00,  2.04it/s]
                                                      
{'loss': 3.5329, 'grad_norm': 0.20164251327514648, 'learning_rate': 8.664028473763719e-05, 'epoch': 0.83}

 83%|████████▎ | 8875/10708 [1:49:38<15:00,  2.04it/s]
 83%|████████▎ | 8876/10708 [1:49:38<15:01,  2.03it/s]
 83%|████████▎ | 8877/10708 [1:49:39<15:00,  2.03it/s]
 83%|████████▎ | 8878/10708 [1:49:39<15:00,  2.03it/s]
 83%|████████▎ | 8879/10708 [1:49:40<15:00,  2.03it/s]
 83%|████████▎ | 8880/10708 [1:49:40<14:59,  2.03it/s]
 83%|████████▎ | 8881/10708 [1:49:41<14:58,  2.03it/s]
 83%|████████▎ | 8882/10708 [1:49:41<14:58,  2.03it/s]
 83%|████████▎ | 8883/10708 [1:49:42<14:56,  2.04it/s]
 83%|████████▎ | 8884/10708 [1:49:42<14:56,  2.03it/s]
 83%|████████▎ | 8885/10708 [1:49:43<14:56,  2.03it/s]
 83%|████████▎ | 8886/10708 [1:49:43<14:57,  2.03it/s]
 83%|████████▎ | 8887/10708 [1:49:43<14:58,  2.03it/s]
 83%|████████▎ | 8888/10708 [1:49:44<14:56,  2.03it/s]
 83%|████████▎ | 8889/10708 [1:49:44<14:56,  2.03it/s]
 83%|████████▎ | 8890/10708 [1:49:45<14:55,  2.03it/s]
 83%|████████▎ | 8891/10708 [1:49:45<14:53,  2.03it/s]
 83%|████████▎ | 8892/10708 [1:49:46<14:53,  2.03it/s]
 83%|████████▎ | 8893/10708 [1:49:46<14:52,  2.03it/s]
 83%|████████▎ | 8894/10708 [1:49:47<14:50,  2.04it/s]
 83%|████████▎ | 8895/10708 [1:49:47<14:50,  2.04it/s]
 83%|████████▎ | 8896/10708 [1:49:48<14:50,  2.03it/s]
 83%|████████▎ | 8897/10708 [1:49:48<14:48,  2.04it/s]
 83%|████████▎ | 8898/10708 [1:49:49<14:49,  2.03it/s]
 83%|████████▎ | 8899/10708 [1:49:49<14:48,  2.04it/s]
 83%|████████▎ | 8900/10708 [1:49:50<14:47,  2.04it/s]{'loss': 3.5449, 'grad_norm': 0.20338138937950134, 'learning_rate': 8.4361434480627e-05, 'epoch': 0.83}

                                                      

 83%|████████▎ | 8900/10708 [1:49:50<14:47,  2.04it/s]
 83%|████████▎ | 8901/10708 [1:49:50<14:50,  2.03it/s]
 83%|████████▎ | 8902/10708 [1:49:51<14:48,  2.03it/s]
 83%|████████▎ | 8903/10708 [1:49:51<14:47,  2.03it/s]
 83%|████████▎ | 8904/10708 [1:49:52<14:47,  2.03it/s]
 83%|████████▎ | 8905/10708 [1:49:52<14:45,  2.04it/s]
 83%|████████▎ | 8906/10708 [1:49:53<14:45,  2.04it/s]
 83%|████████▎ | 8907/10708 [1:49:53<14:45,  2.03it/s]
 83%|████████▎ | 8908/10708 [1:49:54<14:44,  2.03it/s]
 83%|████████▎ | 8909/10708 [1:49:54<14:44,  2.03it/s]
 83%|████████▎ | 8910/10708 [1:49:55<14:44,  2.03it/s]
 83%|████████▎ | 8911/10708 [1:49:55<14:43,  2.03it/s]
 83%|████████▎ | 8912/10708 [1:49:56<14:43,  2.03it/s]
 83%|████████▎ | 8913/10708 [1:49:56<14:42,  2.03it/s]
 83%|████████▎ | 8914/10708 [1:49:57<14:41,  2.04it/s]
 83%|████████▎ | 8915/10708 [1:49:57<14:40,  2.04it/s]
 83%|████████▎ | 8916/10708 [1:49:58<14:41,  2.03it/s]
 83%|████████▎ | 8917/10708 [1:49:58<14:40,  2.03it/s]
 83%|████████▎ | 8918/10708 [1:49:59<14:40,  2.03it/s]
 83%|████████▎ | 8919/10708 [1:49:59<14:40,  2.03it/s]
 83%|████████▎ | 8920/10708 [1:50:00<14:39,  2.03it/s]
 83%|████████▎ | 8921/10708 [1:50:00<14:39,  2.03it/s]
 83%|████████▎ | 8922/10708 [1:50:01<14:37,  2.03it/s]
 83%|████████▎ | 8923/10708 [1:50:01<14:37,  2.03it/s]
 83%|████████▎ | 8924/10708 [1:50:02<14:37,  2.03it/s]
 83%|████████▎ | 8925/10708 [1:50:02<14:36,  2.04it/s]{'loss': 3.536, 'grad_norm': 0.20314031839370728, 'learning_rate': 8.211019060466679e-05, 'epoch': 0.83}
                                                      

 83%|████████▎ | 8925/10708 [1:50:02<14:36,  2.04it/s]
 83%|████████▎ | 8926/10708 [1:50:03<14:37,  2.03it/s]
 83%|████████▎ | 8927/10708 [1:50:03<14:36,  2.03it/s]
 83%|████████▎ | 8928/10708 [1:50:04<14:35,  2.03it/s]
 83%|████████▎ | 8929/10708 [1:50:04<14:34,  2.03it/s]
 83%|████████▎ | 8930/10708 [1:50:05<14:35,  2.03it/s]
 83%|████████▎ | 8931/10708 [1:50:05<14:34,  2.03it/s]
 83%|████████▎ | 8932/10708 [1:50:06<14:34,  2.03it/s]
 83%|████████▎ | 8933/10708 [1:50:06<14:33,  2.03it/s]
 83%|████████▎ | 8934/10708 [1:50:07<14:32,  2.03it/s]
 83%|████████▎ | 8935/10708 [1:50:07<14:31,  2.04it/s]
 83%|████████▎ | 8936/10708 [1:50:08<14:31,  2.03it/s]
 83%|████████▎ | 8937/10708 [1:50:08<14:32,  2.03it/s]
 83%|████████▎ | 8938/10708 [1:50:09<14:32,  2.03it/s]
 83%|████████▎ | 8939/10708 [1:50:09<14:32,  2.03it/s]
 83%|████████▎ | 8940/10708 [1:50:10<14:32,  2.03it/s]
 83%|████████▎ | 8941/10708 [1:50:10<14:31,  2.03it/s]
 84%|████████▎ | 8942/10708 [1:50:11<14:30,  2.03it/s]
 84%|████████▎ | 8943/10708 [1:50:11<14:29,  2.03it/s]
 84%|████████▎ | 8944/10708 [1:50:12<14:27,  2.03it/s]
 84%|████████▎ | 8945/10708 [1:50:12<14:26,  2.03it/s]
 84%|████████▎ | 8946/10708 [1:50:13<14:28,  2.03it/s]
 84%|████████▎ | 8947/10708 [1:50:13<14:27,  2.03it/s]
 84%|████████▎ | 8948/10708 [1:50:14<14:29,  2.02it/s]
 84%|████████▎ | 8949/10708 [1:50:14<14:28,  2.03it/s]
 84%|████████▎ | 8950/10708 [1:50:14<14:27,  2.03it/s]
                                                      
{'loss': 3.5254, 'grad_norm': 0.199686661362648, 'learning_rate': 7.988670263557413e-05, 'epoch': 0.84}

 84%|████████▎ | 8950/10708 [1:50:14<14:27,  2.03it/s]
 84%|████████▎ | 8951/10708 [1:50:15<14:27,  2.02it/s]
 84%|████████▎ | 8952/10708 [1:50:15<14:26,  2.03it/s]
 84%|████████▎ | 8953/10708 [1:50:16<14:25,  2.03it/s]
 84%|████████▎ | 8954/10708 [1:50:16<14:23,  2.03it/s]
 84%|████████▎ | 8955/10708 [1:50:17<14:22,  2.03it/s]
 84%|████████▎ | 8956/10708 [1:50:17<14:22,  2.03it/s]
 84%|████████▎ | 8957/10708 [1:50:18<14:22,  2.03it/s]
 84%|████████▎ | 8958/10708 [1:50:18<14:22,  2.03it/s]
 84%|████████▎ | 8959/10708 [1:50:19<14:22,  2.03it/s]
 84%|████████▎ | 8960/10708 [1:50:19<14:21,  2.03it/s]
 84%|████████▎ | 8961/10708 [1:50:20<14:20,  2.03it/s]
 84%|████████▎ | 8962/10708 [1:50:20<14:19,  2.03it/s]
 84%|████████▎ | 8963/10708 [1:50:21<14:17,  2.04it/s]
 84%|████████▎ | 8964/10708 [1:50:21<14:17,  2.03it/s]
 84%|████████▎ | 8965/10708 [1:50:22<14:17,  2.03it/s]
 84%|████████▎ | 8966/10708 [1:50:22<14:16,  2.03it/s]
 84%|████████▎ | 8967/10708 [1:50:23<14:15,  2.04it/s]
 84%|████████▍ | 8968/10708 [1:50:23<14:15,  2.03it/s]
 84%|████████▍ | 8969/10708 [1:50:24<14:14,  2.03it/s]
 84%|████████▍ | 8970/10708 [1:50:24<14:14,  2.03it/s]
 84%|████████▍ | 8971/10708 [1:50:25<14:13,  2.03it/s]
 84%|████████▍ | 8972/10708 [1:50:25<14:12,  2.04it/s]
 84%|████████▍ | 8973/10708 [1:50:26<14:13,  2.03it/s]
 84%|████████▍ | 8974/10708 [1:50:26<14:11,  2.04it/s]
 84%|████████▍ | 8975/10708 [1:50:27<14:11,  2.04it/s]{'loss': 3.5377, 'grad_norm': 0.19752657413482666, 'learning_rate': 7.76911182556414e-05, 'epoch': 0.84}

                                                      

 84%|████████▍ | 8975/10708 [1:50:27<14:11,  2.04it/s]
 84%|████████▍ | 8976/10708 [1:50:27<14:12,  2.03it/s]
 84%|████████▍ | 8977/10708 [1:50:28<14:10,  2.03it/s]
 84%|████████▍ | 8978/10708 [1:50:28<14:09,  2.04it/s]
 84%|████████▍ | 8979/10708 [1:50:29<14:10,  2.03it/s]
 84%|████████▍ | 8980/10708 [1:50:29<14:09,  2.03it/s]
 84%|████████▍ | 8981/10708 [1:50:30<14:08,  2.03it/s]
 84%|████████▍ | 8982/10708 [1:50:30<14:09,  2.03it/s]
 84%|████████▍ | 8983/10708 [1:50:31<14:07,  2.03it/s]
 84%|████████▍ | 8984/10708 [1:50:31<14:07,  2.03it/s]
 84%|████████▍ | 8985/10708 [1:50:32<14:07,  2.03it/s]
 84%|████████▍ | 8986/10708 [1:50:32<14:06,  2.03it/s]
 84%|████████▍ | 8987/10708 [1:50:33<14:05,  2.04it/s]
 84%|████████▍ | 8988/10708 [1:50:33<14:05,  2.03it/s]
 84%|████████▍ | 8989/10708 [1:50:34<14:05,  2.03it/s]
 84%|████████▍ | 8990/10708 [1:50:34<14:05,  2.03it/s]
 84%|████████▍ | 8991/10708 [1:50:35<14:05,  2.03it/s]
 84%|████████▍ | 8992/10708 [1:50:35<14:03,  2.03it/s]
 84%|████████▍ | 8993/10708 [1:50:36<14:03,  2.03it/s]
 84%|████████▍ | 8994/10708 [1:50:36<14:03,  2.03it/s]
 84%|████████▍ | 8995/10708 [1:50:37<14:02,  2.03it/s]
 84%|████████▍ | 8996/10708 [1:50:37<14:02,  2.03it/s]
 84%|████████▍ | 8997/10708 [1:50:38<14:02,  2.03it/s]
 84%|████████▍ | 8998/10708 [1:50:38<14:02,  2.03it/s]
 84%|████████▍ | 8999/10708 [1:50:39<14:00,  2.03it/s]
 84%|████████▍ | 9000/10708 [1:50:39<14:00,  2.03it/s]
                                                      
{'loss': 3.532, 'grad_norm': 0.2044970691204071, 'learning_rate': 7.552358329382691e-05, 'epoch': 0.84}

 84%|████████▍ | 9000/10708 [1:50:39<14:00,  2.03it/s]
 84%|████████▍ | 9001/10708 [1:50:40<14:00,  2.03it/s]
 84%|████████▍ | 9002/10708 [1:50:40<13:59,  2.03it/s]
 84%|████████▍ | 9003/10708 [1:50:41<13:58,  2.03it/s]
 84%|████████▍ | 9004/10708 [1:50:41<13:58,  2.03it/s]
 84%|████████▍ | 9005/10708 [1:50:42<13:58,  2.03it/s]
 84%|████████▍ | 9006/10708 [1:50:42<13:57,  2.03it/s]
 84%|████████▍ | 9007/10708 [1:50:43<13:56,  2.03it/s]
 84%|████████▍ | 9008/10708 [1:50:43<13:58,  2.03it/s]
 84%|████████▍ | 9009/10708 [1:50:44<13:58,  2.03it/s]
 84%|████████▍ | 9010/10708 [1:50:44<13:56,  2.03it/s]
 84%|████████▍ | 9011/10708 [1:50:45<13:56,  2.03it/s]
 84%|████████▍ | 9012/10708 [1:50:45<13:56,  2.03it/s]
 84%|████████▍ | 9013/10708 [1:50:45<13:55,  2.03it/s]
 84%|████████▍ | 9014/10708 [1:50:46<13:53,  2.03it/s]
 84%|████████▍ | 9015/10708 [1:50:46<13:53,  2.03it/s]
 84%|████████▍ | 9016/10708 [1:50:47<13:52,  2.03it/s]
 84%|████████▍ | 9017/10708 [1:50:47<13:51,  2.03it/s]
 84%|████████▍ | 9018/10708 [1:50:48<13:52,  2.03it/s]
 84%|████████▍ | 9019/10708 [1:50:48<13:51,  2.03it/s]
 84%|████████▍ | 9020/10708 [1:50:49<13:51,  2.03it/s]
 84%|████████▍ | 9021/10708 [1:50:49<13:50,  2.03it/s]
 84%|████████▍ | 9022/10708 [1:50:50<13:50,  2.03it/s]
 84%|████████▍ | 9023/10708 [1:50:50<13:49,  2.03it/s]
 84%|████████▍ | 9024/10708 [1:50:51<13:49,  2.03it/s]
 84%|████████▍ | 9025/10708 [1:50:51<13:49,  2.03it/s]
                                                      
{'loss': 3.5351, 'grad_norm': 0.19974032044410706, 'learning_rate': 7.338424171606878e-05, 'epoch': 0.84}

 84%|████████▍ | 9025/10708 [1:50:51<13:49,  2.03it/s]
 84%|████████▍ | 9026/10708 [1:50:52<13:48,  2.03it/s]
 84%|████████▍ | 9027/10708 [1:50:52<13:48,  2.03it/s]
 84%|████████▍ | 9028/10708 [1:50:53<13:47,  2.03it/s]
 84%|████████▍ | 9029/10708 [1:50:53<13:46,  2.03it/s]
 84%|████████▍ | 9030/10708 [1:50:54<13:46,  2.03it/s]
 84%|████████▍ | 9031/10708 [1:50:54<13:45,  2.03it/s]
 84%|████████▍ | 9032/10708 [1:50:55<13:45,  2.03it/s]
 84%|████████▍ | 9033/10708 [1:50:55<13:44,  2.03it/s]
 84%|████████▍ | 9034/10708 [1:50:56<13:43,  2.03it/s]
 84%|████████▍ | 9035/10708 [1:50:56<13:41,  2.04it/s]
 84%|████████▍ | 9036/10708 [1:50:57<13:41,  2.04it/s]
 84%|████████▍ | 9037/10708 [1:50:57<13:41,  2.03it/s]
 84%|████████▍ | 9038/10708 [1:50:58<13:41,  2.03it/s]
 84%|████████▍ | 9039/10708 [1:50:58<13:40,  2.03it/s]
 84%|████████▍ | 9040/10708 [1:50:59<13:39,  2.04it/s]
 84%|████████▍ | 9041/10708 [1:50:59<13:38,  2.04it/s]
 84%|████████▍ | 9042/10708 [1:51:00<13:39,  2.03it/s]
 84%|████████▍ | 9043/10708 [1:51:00<13:37,  2.04it/s]
 84%|████████▍ | 9044/10708 [1:51:01<13:36,  2.04it/s]
 84%|████████▍ | 9045/10708 [1:51:01<13:37,  2.03it/s]
 84%|████████▍ | 9046/10708 [1:51:02<13:36,  2.04it/s]
 84%|████████▍ | 9047/10708 [1:51:02<13:36,  2.03it/s]
 84%|████████▍ | 9048/10708 [1:51:03<13:36,  2.03it/s]
 85%|████████▍ | 9049/10708 [1:51:03<13:35,  2.04it/s]
 85%|████████▍ | 9050/10708 [1:51:04<13:34,  2.03it/s]
                                                      
{'loss': 3.5278, 'grad_norm': 0.20373842120170593, 'learning_rate': 7.127323561572363e-05, 'epoch': 0.85}

 85%|████████▍ | 9050/10708 [1:51:04<13:34,  2.03it/s]
 85%|████████▍ | 9051/10708 [1:51:04<13:35,  2.03it/s]
 85%|████████▍ | 9052/10708 [1:51:05<13:34,  2.03it/s]
 85%|████████▍ | 9053/10708 [1:51:05<13:34,  2.03it/s]
 85%|████████▍ | 9054/10708 [1:51:06<13:33,  2.03it/s]
 85%|████████▍ | 9055/10708 [1:51:06<13:31,  2.04it/s]
 85%|████████▍ | 9056/10708 [1:51:07<13:32,  2.03it/s]
 85%|████████▍ | 9057/10708 [1:51:07<13:32,  2.03it/s]
 85%|████████▍ | 9058/10708 [1:51:08<13:31,  2.03it/s]
 85%|████████▍ | 9059/10708 [1:51:08<13:31,  2.03it/s]
 85%|████████▍ | 9060/10708 [1:51:09<13:30,  2.03it/s]
 85%|████████▍ | 9061/10708 [1:51:09<13:30,  2.03it/s]
 85%|████████▍ | 9062/10708 [1:51:10<13:29,  2.03it/s]
 85%|████████▍ | 9063/10708 [1:51:10<13:28,  2.04it/s]
 85%|████████▍ | 9064/10708 [1:51:11<13:26,  2.04it/s]
 85%|████████▍ | 9065/10708 [1:51:11<13:26,  2.04it/s]
 85%|████████▍ | 9066/10708 [1:51:12<13:26,  2.03it/s]
 85%|████████▍ | 9067/10708 [1:51:12<13:26,  2.03it/s]
 85%|████████▍ | 9068/10708 [1:51:13<13:27,  2.03it/s]
 85%|████████▍ | 9069/10708 [1:51:13<13:28,  2.03it/s]
 85%|████████▍ | 9070/10708 [1:51:14<13:28,  2.02it/s]
 85%|████████▍ | 9071/10708 [1:51:14<13:26,  2.03it/s]
 85%|████████▍ | 9072/10708 [1:51:15<13:26,  2.03it/s]
 85%|████████▍ | 9073/10708 [1:51:15<13:25,  2.03it/s]
 85%|████████▍ | 9074/10708 [1:51:16<13:24,  2.03it/s]
 85%|████████▍ | 9075/10708 [1:51:16<13:24,  2.03it/s]
                                                      
{'loss': 3.5311, 'grad_norm': 0.19978515803813934, 'learning_rate': 6.919070520412768e-05, 'epoch': 0.85}

 85%|████████▍ | 9075/10708 [1:51:16<13:24,  2.03it/s]
 85%|████████▍ | 9076/10708 [1:51:16<13:23,  2.03it/s]
 85%|████████▍ | 9077/10708 [1:51:17<13:23,  2.03it/s]
 85%|████████▍ | 9078/10708 [1:51:17<13:22,  2.03it/s]
 85%|████████▍ | 9079/10708 [1:51:18<13:22,  2.03it/s]
 85%|████████▍ | 9080/10708 [1:51:18<13:22,  2.03it/s]
 85%|████████▍ | 9081/10708 [1:51:19<13:21,  2.03it/s]
 85%|████████▍ | 9082/10708 [1:51:19<13:21,  2.03it/s]
 85%|████████▍ | 9083/10708 [1:51:20<13:20,  2.03it/s]
 85%|████████▍ | 9084/10708 [1:51:20<13:19,  2.03it/s]
 85%|████████▍ | 9085/10708 [1:51:21<13:18,  2.03it/s]
 85%|████████▍ | 9086/10708 [1:51:21<13:17,  2.03it/s]
 85%|████████▍ | 9087/10708 [1:51:22<13:17,  2.03it/s]
 85%|████████▍ | 9088/10708 [1:51:22<13:17,  2.03it/s]
 85%|████████▍ | 9089/10708 [1:51:23<13:16,  2.03it/s]
 85%|████████▍ | 9090/10708 [1:51:23<13:16,  2.03it/s]
 85%|████████▍ | 9091/10708 [1:51:24<13:15,  2.03it/s]
 85%|████████▍ | 9092/10708 [1:51:24<13:15,  2.03it/s]
 85%|████████▍ | 9093/10708 [1:51:25<13:15,  2.03it/s]
 85%|████████▍ | 9094/10708 [1:51:25<13:14,  2.03it/s]
 85%|████████▍ | 9095/10708 [1:51:26<13:14,  2.03it/s]
 85%|████████▍ | 9096/10708 [1:51:26<13:12,  2.03it/s]
 85%|████████▍ | 9097/10708 [1:51:27<13:11,  2.04it/s]
 85%|████████▍ | 9098/10708 [1:51:27<13:11,  2.03it/s]
 85%|████████▍ | 9099/10708 [1:51:28<13:10,  2.03it/s]
 85%|████████▍ | 9100/10708 [1:51:28<13:09,  2.04it/s]
                                                      
{'loss': 3.5318, 'grad_norm': 0.1989358365535736, 'learning_rate': 6.713678880128543e-05, 'epoch': 0.85}

 85%|████████▍ | 9100/10708 [1:51:28<13:09,  2.04it/s]
 85%|████████▍ | 9101/10708 [1:51:29<13:10,  2.03it/s]
 85%|████████▌ | 9102/10708 [1:51:29<13:09,  2.03it/s]
 85%|████████▌ | 9103/10708 [1:51:30<13:07,  2.04it/s]
 85%|████████▌ | 9104/10708 [1:51:30<13:07,  2.04it/s]
 85%|████████▌ | 9105/10708 [1:51:31<13:08,  2.03it/s]
 85%|████████▌ | 9106/10708 [1:51:31<13:06,  2.04it/s]
 85%|████████▌ | 9107/10708 [1:51:32<13:05,  2.04it/s]
 85%|████████▌ | 9108/10708 [1:51:32<13:06,  2.04it/s]
 85%|████████▌ | 9109/10708 [1:51:33<13:05,  2.04it/s]
 85%|████████▌ | 9110/10708 [1:51:33<13:05,  2.03it/s]
 85%|████████▌ | 9111/10708 [1:51:34<13:05,  2.03it/s]
 85%|████████▌ | 9112/10708 [1:51:34<13:05,  2.03it/s]
 85%|████████▌ | 9113/10708 [1:51:35<13:05,  2.03it/s]
 85%|████████▌ | 9114/10708 [1:51:35<13:04,  2.03it/s]
 85%|████████▌ | 9115/10708 [1:51:36<13:03,  2.03it/s]
 85%|████████▌ | 9116/10708 [1:51:36<13:02,  2.03it/s]
 85%|████████▌ | 9117/10708 [1:51:37<13:02,  2.03it/s]
 85%|████████▌ | 9118/10708 [1:51:37<13:03,  2.03it/s]
 85%|████████▌ | 9119/10708 [1:51:38<13:03,  2.03it/s]
 85%|████████▌ | 9120/10708 [1:51:38<13:03,  2.03it/s]
 85%|████████▌ | 9121/10708 [1:51:39<13:02,  2.03it/s]
 85%|████████▌ | 9122/10708 [1:51:39<13:00,  2.03it/s]
 85%|████████▌ | 9123/10708 [1:51:40<12:59,  2.03it/s]
 85%|████████▌ | 9124/10708 [1:51:40<12:58,  2.03it/s]
 85%|████████▌ | 9125/10708 [1:51:41<12:58,  2.03it/s]{'loss': 3.5314, 'grad_norm': 0.19509272277355194, 'learning_rate': 6.511162282668104e-05, 'epoch': 0.85}

                                                      

 85%|████████▌ | 9125/10708 [1:51:41<12:58,  2.03it/s]
 85%|████████▌ | 9126/10708 [1:51:41<12:58,  2.03it/s]
 85%|████████▌ | 9127/10708 [1:51:42<12:58,  2.03it/s]
 85%|████████▌ | 9128/10708 [1:51:42<12:57,  2.03it/s]
 85%|████████▌ | 9129/10708 [1:51:43<12:57,  2.03it/s]
 85%|████████▌ | 9130/10708 [1:51:43<12:59,  2.03it/s]
 85%|████████▌ | 9131/10708 [1:51:44<12:59,  2.02it/s]
 85%|████████▌ | 9132/10708 [1:51:44<12:57,  2.03it/s]
 85%|████████▌ | 9133/10708 [1:51:45<12:57,  2.03it/s]
 85%|████████▌ | 9134/10708 [1:51:45<12:55,  2.03it/s]
 85%|████████▌ | 9135/10708 [1:51:46<12:54,  2.03it/s]
 85%|████████▌ | 9136/10708 [1:51:46<12:53,  2.03it/s]
 85%|████████▌ | 9137/10708 [1:51:47<12:52,  2.03it/s]
 85%|████████▌ | 9138/10708 [1:51:47<12:51,  2.03it/s]
 85%|████████▌ | 9139/10708 [1:51:47<12:52,  2.03it/s]
 85%|████████▌ | 9140/10708 [1:51:48<12:51,  2.03it/s]
 85%|████████▌ | 9141/10708 [1:51:48<12:50,  2.03it/s]
 85%|████████▌ | 9142/10708 [1:51:49<12:51,  2.03it/s]
 85%|████████▌ | 9143/10708 [1:51:49<12:51,  2.03it/s]
 85%|████████▌ | 9144/10708 [1:51:50<12:50,  2.03it/s]
 85%|████████▌ | 9145/10708 [1:51:50<12:52,  2.02it/s]
 85%|████████▌ | 9146/10708 [1:51:51<12:50,  2.03it/s]
 85%|████████▌ | 9147/10708 [1:51:51<12:48,  2.03it/s]
 85%|████████▌ | 9148/10708 [1:51:52<12:49,  2.03it/s]
 85%|████████▌ | 9149/10708 [1:51:52<12:48,  2.03it/s]
 85%|████████▌ | 9150/10708 [1:51:53<12:46,  2.03it/s]
                                                      {'loss': 3.5305, 'grad_norm': 0.196262389421463, 'learning_rate': 6.311534179021878e-05, 'epoch': 0.85}

 85%|████████▌ | 9150/10708 [1:51:53<12:46,  2.03it/s]
 85%|████████▌ | 9151/10708 [1:51:53<12:47,  2.03it/s]
 85%|████████▌ | 9152/10708 [1:51:54<12:46,  2.03it/s]
 85%|████████▌ | 9153/10708 [1:51:54<12:44,  2.03it/s]
 85%|████████▌ | 9154/10708 [1:51:55<12:43,  2.03it/s]
 85%|████████▌ | 9155/10708 [1:51:55<12:43,  2.03it/s]
 86%|████████▌ | 9156/10708 [1:51:56<12:43,  2.03it/s]
 86%|████████▌ | 9157/10708 [1:51:56<12:42,  2.03it/s]
 86%|████████▌ | 9158/10708 [1:51:57<12:42,  2.03it/s]
 86%|████████▌ | 9159/10708 [1:51:57<12:41,  2.03it/s]
 86%|████████▌ | 9160/10708 [1:51:58<12:41,  2.03it/s]
 86%|████████▌ | 9161/10708 [1:51:58<12:40,  2.03it/s]
 86%|████████▌ | 9162/10708 [1:51:59<12:39,  2.04it/s]
 86%|████████▌ | 9163/10708 [1:51:59<12:41,  2.03it/s]
 86%|████████▌ | 9164/10708 [1:52:00<12:39,  2.03it/s]
 86%|████████▌ | 9165/10708 [1:52:00<12:38,  2.03it/s]
 86%|████████▌ | 9166/10708 [1:52:01<12:38,  2.03it/s]
 86%|████████▌ | 9167/10708 [1:52:01<12:37,  2.03it/s]
 86%|████████▌ | 9168/10708 [1:52:02<12:36,  2.03it/s]
 86%|████████▌ | 9169/10708 [1:52:02<12:37,  2.03it/s]
 86%|████████▌ | 9170/10708 [1:52:03<12:36,  2.03it/s]
 86%|████████▌ | 9171/10708 [1:52:03<12:37,  2.03it/s]
 86%|████████▌ | 9172/10708 [1:52:04<12:35,  2.03it/s]
 86%|████████▌ | 9173/10708 [1:52:04<12:35,  2.03it/s]
 86%|████████▌ | 9174/10708 [1:52:05<12:35,  2.03it/s]
 86%|████████▌ | 9175/10708 [1:52:05<12:34,  2.03it/s]
                                                      
{'loss': 3.5335, 'grad_norm': 0.19770342111587524, 'learning_rate': 6.114807828328828e-05, 'epoch': 0.86}

 86%|████████▌ | 9175/10708 [1:52:05<12:34,  2.03it/s]
 86%|████████▌ | 9176/10708 [1:52:06<12:34,  2.03it/s]
 86%|████████▌ | 9177/10708 [1:52:06<12:33,  2.03it/s]
 86%|████████▌ | 9178/10708 [1:52:07<12:32,  2.03it/s]
 86%|████████▌ | 9179/10708 [1:52:07<12:32,  2.03it/s]
 86%|████████▌ | 9180/10708 [1:52:08<12:32,  2.03it/s]
 86%|████████▌ | 9181/10708 [1:52:08<12:30,  2.03it/s]
 86%|████████▌ | 9182/10708 [1:52:09<12:30,  2.03it/s]
 86%|████████▌ | 9183/10708 [1:52:09<12:29,  2.03it/s]
 86%|████████▌ | 9184/10708 [1:52:10<12:28,  2.04it/s]
 86%|████████▌ | 9185/10708 [1:52:10<12:28,  2.04it/s]
 86%|████████▌ | 9186/10708 [1:52:11<12:28,  2.03it/s]
 86%|████████▌ | 9187/10708 [1:52:11<12:26,  2.04it/s]
 86%|████████▌ | 9188/10708 [1:52:12<12:26,  2.04it/s]
 86%|████████▌ | 9189/10708 [1:52:12<12:27,  2.03it/s]
 86%|████████▌ | 9190/10708 [1:52:13<12:27,  2.03it/s]
 86%|████████▌ | 9191/10708 [1:52:13<12:28,  2.03it/s]
 86%|████████▌ | 9192/10708 [1:52:14<12:26,  2.03it/s]
 86%|████████▌ | 9193/10708 [1:52:14<12:26,  2.03it/s]
 86%|████████▌ | 9194/10708 [1:52:15<12:25,  2.03it/s]
 86%|████████▌ | 9195/10708 [1:52:15<12:23,  2.03it/s]
 86%|████████▌ | 9196/10708 [1:52:16<12:23,  2.03it/s]
 86%|████████▌ | 9197/10708 [1:52:16<12:22,  2.03it/s]
 86%|████████▌ | 9198/10708 [1:52:17<12:22,  2.03it/s]
 86%|████████▌ | 9199/10708 [1:52:17<12:22,  2.03it/s]
 86%|████████▌ | 9200/10708 [1:52:18<12:21,  2.03it/s]
                                                      
{'loss': 3.5275, 'grad_norm': 0.20367445051670074, 'learning_rate': 5.920996296995812e-05, 'epoch': 0.86}

 86%|████████▌ | 9200/10708 [1:52:18<12:21,  2.03it/s]
 86%|████████▌ | 9201/10708 [1:52:18<12:21,  2.03it/s]
 86%|████████▌ | 9202/10708 [1:52:18<12:21,  2.03it/s]
 86%|████████▌ | 9203/10708 [1:52:19<12:20,  2.03it/s]
 86%|████████▌ | 9204/10708 [1:52:19<12:20,  2.03it/s]
 86%|████████▌ | 9205/10708 [1:52:20<12:18,  2.03it/s]
 86%|████████▌ | 9206/10708 [1:52:20<12:19,  2.03it/s]
 86%|████████▌ | 9207/10708 [1:52:21<12:18,  2.03it/s]
 86%|████████▌ | 9208/10708 [1:52:21<12:17,  2.03it/s]
 86%|████████▌ | 9209/10708 [1:52:22<12:18,  2.03it/s]
 86%|████████▌ | 9210/10708 [1:52:22<12:17,  2.03it/s]
 86%|████████▌ | 9211/10708 [1:52:23<12:16,  2.03it/s]
 86%|████████▌ | 9212/10708 [1:52:23<12:15,  2.03it/s]
 86%|████████▌ | 9213/10708 [1:52:24<12:15,  2.03it/s]
 86%|████████▌ | 9214/10708 [1:52:24<12:13,  2.04it/s]
 86%|████████▌ | 9215/10708 [1:52:25<12:13,  2.04it/s]
 86%|████████▌ | 9216/10708 [1:52:25<12:13,  2.03it/s]
 86%|████████▌ | 9217/10708 [1:52:26<12:12,  2.04it/s]
 86%|████████▌ | 9218/10708 [1:52:26<12:12,  2.03it/s]
 86%|████████▌ | 9219/10708 [1:52:27<12:12,  2.03it/s]
 86%|████████▌ | 9220/10708 [1:52:27<12:11,  2.04it/s]
 86%|████████▌ | 9221/10708 [1:52:28<12:11,  2.03it/s]
 86%|████████▌ | 9222/10708 [1:52:28<12:11,  2.03it/s]
 86%|████████▌ | 9223/10708 [1:52:29<12:10,  2.03it/s]
 86%|████████▌ | 9224/10708 [1:52:29<12:09,  2.03it/s]
 86%|████████▌ | 9225/10708 [1:52:30<12:09,  2.03it/s]{'loss': 3.5299, 'grad_norm': 0.1972706913948059, 'learning_rate': 5.730112457829689e-05, 'epoch': 0.86}
                                                      

 86%|████████▌ | 9225/10708 [1:52:30<12:09,  2.03it/s]
 86%|████████▌ | 9226/10708 [1:52:30<12:09,  2.03it/s]
 86%|████████▌ | 9227/10708 [1:52:31<12:10,  2.03it/s]
 86%|████████▌ | 9228/10708 [1:52:31<12:09,  2.03it/s]
 86%|████████▌ | 9229/10708 [1:52:32<12:08,  2.03it/s]
 86%|████████▌ | 9230/10708 [1:52:32<12:07,  2.03it/s]
 86%|████████▌ | 9231/10708 [1:52:33<12:07,  2.03it/s]
 86%|████████▌ | 9232/10708 [1:52:33<12:06,  2.03it/s]
 86%|████████▌ | 9233/10708 [1:52:34<12:06,  2.03it/s]
 86%|████████▌ | 9234/10708 [1:52:34<12:05,  2.03it/s]
 86%|████████▌ | 9235/10708 [1:52:35<12:04,  2.03it/s]
 86%|████████▋ | 9236/10708 [1:52:35<12:04,  2.03it/s]
 86%|████████▋ | 9237/10708 [1:52:36<12:03,  2.03it/s]
 86%|████████▋ | 9238/10708 [1:52:36<12:04,  2.03it/s]
 86%|████████▋ | 9239/10708 [1:52:37<12:02,  2.03it/s]
 86%|████████▋ | 9240/10708 [1:52:37<12:03,  2.03it/s]
 86%|████████▋ | 9241/10708 [1:52:38<12:02,  2.03it/s]
 86%|████████▋ | 9242/10708 [1:52:38<12:01,  2.03it/s]
 86%|████████▋ | 9243/10708 [1:52:39<12:00,  2.03it/s]
 86%|████████▋ | 9244/10708 [1:52:39<11:59,  2.03it/s]
 86%|████████▋ | 9245/10708 [1:52:40<11:59,  2.03it/s]
 86%|████████▋ | 9246/10708 [1:52:40<11:58,  2.03it/s]
 86%|████████▋ | 9247/10708 [1:52:41<11:58,  2.03it/s]
 86%|████████▋ | 9248/10708 [1:52:41<11:58,  2.03it/s]
 86%|████████▋ | 9249/10708 [1:52:42<11:58,  2.03it/s]
 86%|████████▋ | 9250/10708 [1:52:42<11:57,  2.03it/s]
                                                      {'loss': 3.5298, 'grad_norm': 0.1976974904537201, 'learning_rate': 5.542168989182389e-05, 'epoch': 0.86}


 86%|████████▋ | 9250/10708 [1:52:42<11:57,  2.03it/s]
 86%|████████▋ | 9251/10708 [1:52:43<11:58,  2.03it/s]
 86%|████████▋ | 9252/10708 [1:52:43<11:58,  2.03it/s]
 86%|████████▋ | 9253/10708 [1:52:44<11:56,  2.03it/s]
 86%|████████▋ | 9254/10708 [1:52:44<11:56,  2.03it/s]
 86%|████████▋ | 9255/10708 [1:52:45<11:54,  2.03it/s]
 86%|████████▋ | 9256/10708 [1:52:45<11:54,  2.03it/s]
 86%|████████▋ | 9257/10708 [1:52:46<11:54,  2.03it/s]
 86%|████████▋ | 9258/10708 [1:52:46<11:54,  2.03it/s]
 86%|████████▋ | 9259/10708 [1:52:47<11:53,  2.03it/s]
 86%|████████▋ | 9260/10708 [1:52:47<11:52,  2.03it/s]
 86%|████████▋ | 9261/10708 [1:52:48<11:52,  2.03it/s]
 86%|████████▋ | 9262/10708 [1:52:48<11:52,  2.03it/s]
 87%|████████▋ | 9263/10708 [1:52:49<11:51,  2.03it/s]
 87%|████████▋ | 9264/10708 [1:52:49<11:51,  2.03it/s]
 87%|████████▋ | 9265/10708 [1:52:49<11:50,  2.03it/s]
 87%|████████▋ | 9266/10708 [1:52:50<11:49,  2.03it/s]
 87%|████████▋ | 9267/10708 [1:52:50<11:49,  2.03it/s]
 87%|████████▋ | 9268/10708 [1:52:51<11:49,  2.03it/s]
 87%|████████▋ | 9269/10708 [1:52:51<11:48,  2.03it/s]
 87%|████████▋ | 9270/10708 [1:52:52<11:47,  2.03it/s]
 87%|████████▋ | 9271/10708 [1:52:52<11:48,  2.03it/s]
 87%|████████▋ | 9272/10708 [1:52:53<11:46,  2.03it/s]
 87%|████████▋ | 9273/10708 [1:52:53<11:46,  2.03it/s]
 87%|████████▋ | 9274/10708 [1:52:54<11:45,  2.03it/s]
 87%|████████▋ | 9275/10708 [1:52:54<11:43,  2.04it/s]{'loss': 3.5248, 'grad_norm': 0.19870561361312866, 'learning_rate': 5.357178374108751e-05, 'epoch': 0.87}

                                                      

 87%|████████▋ | 9275/10708 [1:52:54<11:43,  2.04it/s]
 87%|████████▋ | 9276/10708 [1:52:55<11:45,  2.03it/s]
 87%|████████▋ | 9277/10708 [1:52:55<11:44,  2.03it/s]
 87%|████████▋ | 9278/10708 [1:52:56<11:43,  2.03it/s]
 87%|████████▋ | 9279/10708 [1:52:56<11:43,  2.03it/s]
 87%|████████▋ | 9280/10708 [1:52:57<11:42,  2.03it/s]
 87%|████████▋ | 9281/10708 [1:52:57<11:42,  2.03it/s]
 87%|████████▋ | 9282/10708 [1:52:58<11:42,  2.03it/s]
 87%|████████▋ | 9283/10708 [1:52:58<11:41,  2.03it/s]
 87%|████████▋ | 9284/10708 [1:52:59<11:39,  2.03it/s]
 87%|████████▋ | 9285/10708 [1:52:59<11:39,  2.04it/s]
 87%|████████▋ | 9286/10708 [1:53:00<11:39,  2.03it/s]
 87%|████████▋ | 9287/10708 [1:53:00<11:39,  2.03it/s]
 87%|████████▋ | 9288/10708 [1:53:01<11:38,  2.03it/s]
 87%|████████▋ | 9289/10708 [1:53:01<11:37,  2.03it/s]
 87%|████████▋ | 9290/10708 [1:53:02<11:36,  2.03it/s]
 87%|████████▋ | 9291/10708 [1:53:02<11:37,  2.03it/s]
 87%|████████▋ | 9292/10708 [1:53:03<11:36,  2.03it/s]
 87%|████████▋ | 9293/10708 [1:53:03<11:35,  2.03it/s]
 87%|████████▋ | 9294/10708 [1:53:04<11:35,  2.03it/s]
 87%|████████▋ | 9295/10708 [1:53:04<11:34,  2.03it/s]
 87%|████████▋ | 9296/10708 [1:53:05<11:34,  2.03it/s]
 87%|████████▋ | 9297/10708 [1:53:05<11:33,  2.03it/s]
 87%|████████▋ | 9298/10708 [1:53:06<11:33,  2.03it/s]
 87%|████████▋ | 9299/10708 [1:53:06<11:32,  2.03it/s]
 87%|████████▋ | 9300/10708 [1:53:07<11:32,  2.03it/s]
                                                      
{'loss': 3.5312, 'grad_norm': 0.1978793889284134, 'learning_rate': 5.1751528995374495e-05, 'epoch': 0.87}

 87%|████████▋ | 9300/10708 [1:53:07<11:32,  2.03it/s]
 87%|████████▋ | 9301/10708 [1:53:07<11:32,  2.03it/s]
 87%|████████▋ | 9302/10708 [1:53:08<11:31,  2.03it/s]
 87%|████████▋ | 9303/10708 [1:53:08<11:31,  2.03it/s]
 87%|████████▋ | 9304/10708 [1:53:09<11:30,  2.03it/s]
 87%|████████▋ | 9305/10708 [1:53:09<11:28,  2.04it/s]
 87%|████████▋ | 9306/10708 [1:53:10<11:29,  2.03it/s]
 87%|████████▋ | 9307/10708 [1:53:10<11:28,  2.03it/s]
 87%|████████▋ | 9308/10708 [1:53:11<11:28,  2.03it/s]
 87%|████████▋ | 9309/10708 [1:53:11<11:28,  2.03it/s]
 87%|████████▋ | 9310/10708 [1:53:12<11:27,  2.03it/s]
 87%|████████▋ | 9311/10708 [1:53:12<11:26,  2.03it/s]
 87%|████████▋ | 9312/10708 [1:53:13<11:27,  2.03it/s]
 87%|████████▋ | 9313/10708 [1:53:13<11:27,  2.03it/s]
 87%|████████▋ | 9314/10708 [1:53:14<11:27,  2.03it/s]
 87%|████████▋ | 9315/10708 [1:53:14<11:25,  2.03it/s]
 87%|████████▋ | 9316/10708 [1:53:15<11:24,  2.03it/s]
 87%|████████▋ | 9317/10708 [1:53:15<11:24,  2.03it/s]
 87%|████████▋ | 9318/10708 [1:53:16<11:23,  2.03it/s]
 87%|████████▋ | 9319/10708 [1:53:16<11:23,  2.03it/s]
 87%|████████▋ | 9320/10708 [1:53:17<11:22,  2.03it/s]
 87%|████████▋ | 9321/10708 [1:53:17<11:21,  2.03it/s]
 87%|████████▋ | 9322/10708 [1:53:18<11:21,  2.03it/s]
 87%|████████▋ | 9323/10708 [1:53:18<11:20,  2.04it/s]
 87%|████████▋ | 9324/10708 [1:53:19<11:19,  2.04it/s]
 87%|████████▋ | 9325/10708 [1:53:19<11:20,  2.03it/s]
                                                      
{'loss': 3.5275, 'grad_norm': 0.20168893039226532, 'learning_rate': 4.9961046554549026e-05, 'epoch': 0.87}

 87%|████████▋ | 9325/10708 [1:53:19<11:20,  2.03it/s]
 87%|████████▋ | 9326/10708 [1:53:20<11:20,  2.03it/s]
 87%|████████▋ | 9327/10708 [1:53:20<11:20,  2.03it/s]
 87%|████████▋ | 9328/10708 [1:53:20<11:19,  2.03it/s]
 87%|████████▋ | 9329/10708 [1:53:21<11:18,  2.03it/s]
 87%|████████▋ | 9330/10708 [1:53:21<11:18,  2.03it/s]
 87%|████████▋ | 9331/10708 [1:53:22<11:17,  2.03it/s]
 87%|████████▋ | 9332/10708 [1:53:22<11:15,  2.04it/s]
 87%|████████▋ | 9333/10708 [1:53:23<11:15,  2.04it/s]
 87%|████████▋ | 9334/10708 [1:53:23<11:14,  2.04it/s]
 87%|████████▋ | 9335/10708 [1:53:24<11:14,  2.04it/s]
 87%|████████▋ | 9336/10708 [1:53:24<11:13,  2.04it/s]
 87%|████████▋ | 9337/10708 [1:53:25<11:13,  2.03it/s]
 87%|████████▋ | 9338/10708 [1:53:25<11:12,  2.04it/s]
 87%|████████▋ | 9339/10708 [1:53:26<11:11,  2.04it/s]
 87%|████████▋ | 9340/10708 [1:53:26<11:11,  2.04it/s]
 87%|████████▋ | 9341/10708 [1:53:27<11:11,  2.04it/s]
 87%|████████▋ | 9342/10708 [1:53:27<11:11,  2.03it/s]
 87%|████████▋ | 9343/10708 [1:53:28<11:10,  2.04it/s]
 87%|████████▋ | 9344/10708 [1:53:28<11:10,  2.04it/s]
 87%|████████▋ | 9345/10708 [1:53:29<11:10,  2.03it/s]
 87%|████████▋ | 9346/10708 [1:53:29<11:09,  2.04it/s]
 87%|████████▋ | 9347/10708 [1:53:30<11:08,  2.04it/s]
 87%|████████▋ | 9348/10708 [1:53:30<11:07,  2.04it/s]
 87%|████████▋ | 9349/10708 [1:53:31<11:07,  2.04it/s]
 87%|████████▋ | 9350/10708 [1:53:31<11:06,  2.04it/s]
                                                      {'loss': 3.5311, 'grad_norm': 0.19817441701889038, 'learning_rate': 4.820045534102263e-05, 'epoch': 0.87}

 87%|████████▋ | 9350/10708 [1:53:31<11:06,  2.04it/s]
 87%|████████▋ | 9351/10708 [1:53:32<11:07,  2.03it/s]
 87%|████████▋ | 9352/10708 [1:53:32<11:06,  2.03it/s]
 87%|████████▋ | 9353/10708 [1:53:33<11:05,  2.04it/s]
 87%|████████▋ | 9354/10708 [1:53:33<11:04,  2.04it/s]
 87%|████████▋ | 9355/10708 [1:53:34<11:05,  2.03it/s]
 87%|████████▋ | 9356/10708 [1:53:34<11:03,  2.04it/s]
 87%|████████▋ | 9357/10708 [1:53:35<11:04,  2.03it/s]
 87%|████████▋ | 9358/10708 [1:53:35<11:03,  2.03it/s]
 87%|████████▋ | 9359/10708 [1:53:36<11:02,  2.04it/s]
 87%|████████▋ | 9360/10708 [1:53:36<11:02,  2.03it/s]
 87%|████████▋ | 9361/10708 [1:53:37<11:02,  2.03it/s]
 87%|████████▋ | 9362/10708 [1:53:37<11:01,  2.04it/s]
 87%|████████▋ | 9363/10708 [1:53:38<11:01,  2.03it/s]
 87%|████████▋ | 9364/10708 [1:53:38<11:01,  2.03it/s]
 87%|████████▋ | 9365/10708 [1:53:39<11:00,  2.03it/s]
 87%|████████▋ | 9366/10708 [1:53:39<10:59,  2.04it/s]
 87%|████████▋ | 9367/10708 [1:53:40<10:59,  2.03it/s]
 87%|████████▋ | 9368/10708 [1:53:40<10:59,  2.03it/s]
 87%|████████▋ | 9369/10708 [1:53:41<10:58,  2.03it/s]
 88%|████████▊ | 9370/10708 [1:53:41<10:58,  2.03it/s]
 88%|████████▊ | 9371/10708 [1:53:42<10:57,  2.03it/s]
 88%|████████▊ | 9372/10708 [1:53:42<10:57,  2.03it/s]
 88%|████████▊ | 9373/10708 [1:53:43<10:57,  2.03it/s]
 88%|████████▊ | 9374/10708 [1:53:43<10:59,  2.02it/s]
 88%|████████▊ | 9375/10708 [1:53:44<10:58,  2.02it/s]
                                                      
{'loss': 3.5248, 'grad_norm': 0.19329145550727844, 'learning_rate': 4.6469872291855506e-05, 'epoch': 0.88}

 88%|████████▊ | 9375/10708 [1:53:44<10:58,  2.02it/s]
 88%|████████▊ | 9376/10708 [1:53:44<10:57,  2.03it/s]
 88%|████████▊ | 9377/10708 [1:53:45<10:56,  2.03it/s]
 88%|████████▊ | 9378/10708 [1:53:45<10:55,  2.03it/s]
 88%|████████▊ | 9379/10708 [1:53:46<10:54,  2.03it/s]
 88%|████████▊ | 9380/10708 [1:53:46<10:54,  2.03it/s]
 88%|████████▊ | 9381/10708 [1:53:47<10:54,  2.03it/s]
 88%|████████▊ | 9382/10708 [1:53:47<10:53,  2.03it/s]
 88%|████████▊ | 9383/10708 [1:53:48<10:52,  2.03it/s]
 88%|████████▊ | 9384/10708 [1:53:48<10:52,  2.03it/s]
 88%|████████▊ | 9385/10708 [1:53:49<10:52,  2.03it/s]
 88%|████████▊ | 9386/10708 [1:53:49<10:50,  2.03it/s]
 88%|████████▊ | 9387/10708 [1:53:50<10:50,  2.03it/s]
 88%|████████▊ | 9388/10708 [1:53:50<10:50,  2.03it/s]
 88%|████████▊ | 9389/10708 [1:53:50<10:49,  2.03it/s]
 88%|████████▊ | 9390/10708 [1:53:51<10:49,  2.03it/s]
 88%|████████▊ | 9391/10708 [1:53:51<10:48,  2.03it/s]
 88%|████████▊ | 9392/10708 [1:53:52<10:46,  2.04it/s]
 88%|████████▊ | 9393/10708 [1:53:52<10:46,  2.03it/s]
 88%|████████▊ | 9394/10708 [1:53:53<10:46,  2.03it/s]
 88%|████████▊ | 9395/10708 [1:53:53<10:45,  2.04it/s]
 88%|████████▊ | 9396/10708 [1:53:54<10:44,  2.03it/s]
 88%|████████▊ | 9397/10708 [1:53:54<10:44,  2.03it/s]
 88%|████████▊ | 9398/10708 [1:53:55<10:43,  2.04it/s]
 88%|████████▊ | 9399/10708 [1:53:55<10:44,  2.03it/s]
 88%|████████▊ | 9400/10708 [1:53:56<10:43,  2.03it/s]
                                                      
{'loss': 3.52, 'grad_norm': 0.19647416472434998, 'learning_rate': 4.476941235098936e-05, 'epoch': 0.88}

 88%|████████▊ | 9400/10708 [1:53:56<10:43,  2.03it/s]
 88%|████████▊ | 9401/10708 [1:53:56<10:44,  2.03it/s]
 88%|████████▊ | 9402/10708 [1:53:57<10:43,  2.03it/s]
 88%|████████▊ | 9403/10708 [1:53:57<10:41,  2.03it/s]
 88%|████████▊ | 9404/10708 [1:53:58<10:42,  2.03it/s]
 88%|████████▊ | 9405/10708 [1:53:58<10:41,  2.03it/s]
 88%|████████▊ | 9406/10708 [1:53:59<10:39,  2.04it/s]
 88%|████████▊ | 9407/10708 [1:53:59<10:40,  2.03it/s]
 88%|████████▊ | 9408/10708 [1:54:00<10:39,  2.03it/s]
 88%|████████▊ | 9409/10708 [1:54:00<10:39,  2.03it/s]
 88%|████████▊ | 9410/10708 [1:54:01<10:39,  2.03it/s]
 88%|████████▊ | 9411/10708 [1:54:01<10:38,  2.03it/s]
 88%|████████▊ | 9412/10708 [1:54:02<10:38,  2.03it/s]
 88%|████████▊ | 9413/10708 [1:54:02<10:37,  2.03it/s]
 88%|████████▊ | 9414/10708 [1:54:03<10:36,  2.03it/s]
 88%|████████▊ | 9415/10708 [1:54:03<10:36,  2.03it/s]
 88%|████████▊ | 9416/10708 [1:54:04<10:34,  2.04it/s]
 88%|████████▊ | 9417/10708 [1:54:04<10:35,  2.03it/s]
 88%|████████▊ | 9418/10708 [1:54:05<10:35,  2.03it/s]
 88%|████████▊ | 9419/10708 [1:54:05<10:34,  2.03it/s]
 88%|████████▊ | 9420/10708 [1:54:06<10:34,  2.03it/s]
 88%|████████▊ | 9421/10708 [1:54:06<10:33,  2.03it/s]
 88%|████████▊ | 9422/10708 [1:54:07<10:32,  2.03it/s]
 88%|████████▊ | 9423/10708 [1:54:07<10:31,  2.03it/s]
 88%|████████▊ | 9424/10708 [1:54:08<10:31,  2.03it/s]
 88%|████████▊ | 9425/10708 [1:54:08<10:31,  2.03it/s]{'loss': 3.5216, 'grad_norm': 0.19379588961601257, 'learning_rate': 4.3099188461613315e-05, 'epoch': 0.88}
                                                      

 88%|████████▊ | 9425/10708 [1:54:08<10:31,  2.03it/s]
 88%|████████▊ | 9426/10708 [1:54:09<10:32,  2.03it/s]
 88%|████████▊ | 9427/10708 [1:54:09<10:30,  2.03it/s]
 88%|████████▊ | 9428/10708 [1:54:10<10:30,  2.03it/s]
 88%|████████▊ | 9429/10708 [1:54:10<10:29,  2.03it/s]
 88%|████████▊ | 9430/10708 [1:54:11<10:28,  2.03it/s]
 88%|████████▊ | 9431/10708 [1:54:11<10:28,  2.03it/s]
 88%|████████▊ | 9432/10708 [1:54:12<10:27,  2.03it/s]
 88%|████████▊ | 9433/10708 [1:54:12<10:27,  2.03it/s]
 88%|████████▊ | 9434/10708 [1:54:13<10:27,  2.03it/s]
 88%|████████▊ | 9435/10708 [1:54:13<10:29,  2.02it/s]
 88%|████████▊ | 9436/10708 [1:54:14<10:28,  2.02it/s]
 88%|████████▊ | 9437/10708 [1:54:14<10:27,  2.02it/s]
 88%|████████▊ | 9438/10708 [1:54:15<10:28,  2.02it/s]
 88%|████████▊ | 9439/10708 [1:54:15<10:26,  2.03it/s]
 88%|████████▊ | 9440/10708 [1:54:16<10:25,  2.03it/s]
 88%|████████▊ | 9441/10708 [1:54:16<10:24,  2.03it/s]
 88%|████████▊ | 9442/10708 [1:54:17<10:23,  2.03it/s]
 88%|████████▊ | 9443/10708 [1:54:17<10:23,  2.03it/s]
 88%|████████▊ | 9444/10708 [1:54:18<10:21,  2.03it/s]
 88%|████████▊ | 9445/10708 [1:54:18<10:20,  2.04it/s]
 88%|████████▊ | 9446/10708 [1:54:19<10:20,  2.03it/s]
 88%|████████▊ | 9447/10708 [1:54:19<10:19,  2.03it/s]
 88%|████████▊ | 9448/10708 [1:54:20<10:19,  2.03it/s]
 88%|████████▊ | 9449/10708 [1:54:20<10:19,  2.03it/s]
 88%|████████▊ | 9450/10708 [1:54:21<10:18,  2.03it/s]{'loss': 3.5236, 'grad_norm': 0.19360174238681793, 'learning_rate': 4.1459311558662005e-05, 'epoch': 0.88}
                                                      

 88%|████████▊ | 9450/10708 [1:54:21<10:18,  2.03it/s]
 88%|████████▊ | 9451/10708 [1:54:21<10:18,  2.03it/s]
 88%|████████▊ | 9452/10708 [1:54:22<10:17,  2.03it/s]
 88%|████████▊ | 9453/10708 [1:54:22<10:16,  2.03it/s]
 88%|████████▊ | 9454/10708 [1:54:22<10:16,  2.03it/s]
 88%|████████▊ | 9455/10708 [1:54:23<10:15,  2.03it/s]
 88%|████████▊ | 9456/10708 [1:54:23<10:14,  2.04it/s]
 88%|████████▊ | 9457/10708 [1:54:24<10:14,  2.04it/s]
 88%|████████▊ | 9458/10708 [1:54:24<10:14,  2.03it/s]
 88%|████████▊ | 9459/10708 [1:54:25<10:13,  2.04it/s]
 88%|████████▊ | 9460/10708 [1:54:25<10:13,  2.03it/s]
 88%|████████▊ | 9461/10708 [1:54:26<10:13,  2.03it/s]
 88%|████████▊ | 9462/10708 [1:54:26<10:12,  2.03it/s]
 88%|████████▊ | 9463/10708 [1:54:27<10:12,  2.03it/s]
 88%|████████▊ | 9464/10708 [1:54:27<10:11,  2.03it/s]
 88%|████████▊ | 9465/10708 [1:54:28<10:10,  2.04it/s]
 88%|████████▊ | 9466/10708 [1:54:28<10:09,  2.04it/s]
 88%|████████▊ | 9467/10708 [1:54:29<10:10,  2.03it/s]
 88%|████████▊ | 9468/10708 [1:54:29<10:09,  2.03it/s]
 88%|████████▊ | 9469/10708 [1:54:30<10:09,  2.03it/s]
 88%|████████▊ | 9470/10708 [1:54:30<10:08,  2.03it/s]
 88%|████████▊ | 9471/10708 [1:54:31<10:07,  2.04it/s]
 88%|████████▊ | 9472/10708 [1:54:31<10:08,  2.03it/s]
 88%|████████▊ | 9473/10708 [1:54:32<10:07,  2.03it/s]
 88%|████████▊ | 9474/10708 [1:54:32<10:06,  2.03it/s]
 88%|████████▊ | 9475/10708 [1:54:33<10:06,  2.03it/s]{'loss': 3.5239, 'grad_norm': 0.19673432409763336, 'learning_rate': 3.984989056144766e-05, 'epoch': 0.88}

                                                      

 88%|████████▊ | 9475/10708 [1:54:33<10:06,  2.03it/s]
 88%|████████▊ | 9476/10708 [1:54:33<10:06,  2.03it/s]
 89%|████████▊ | 9477/10708 [1:54:34<10:06,  2.03it/s]
 89%|████████▊ | 9478/10708 [1:54:34<10:05,  2.03it/s]
 89%|████████▊ | 9479/10708 [1:54:35<10:05,  2.03it/s]
 89%|████████▊ | 9480/10708 [1:54:35<10:05,  2.03it/s]
 89%|████████▊ | 9481/10708 [1:54:36<10:05,  2.03it/s]
 89%|████████▊ | 9482/10708 [1:54:36<10:04,  2.03it/s]
 89%|████████▊ | 9483/10708 [1:54:37<10:03,  2.03it/s]
 89%|████████▊ | 9484/10708 [1:54:37<10:02,  2.03it/s]
 89%|████████▊ | 9485/10708 [1:54:38<10:02,  2.03it/s]
 89%|████████▊ | 9486/10708 [1:54:38<10:01,  2.03it/s]
 89%|████████▊ | 9487/10708 [1:54:39<10:02,  2.03it/s]
 89%|████████▊ | 9488/10708 [1:54:39<10:00,  2.03it/s]
 89%|████████▊ | 9489/10708 [1:54:40<10:00,  2.03it/s]
 89%|████████▊ | 9490/10708 [1:54:40<09:59,  2.03it/s]
 89%|████████▊ | 9491/10708 [1:54:41<09:58,  2.03it/s]
 89%|████████▊ | 9492/10708 [1:54:41<09:58,  2.03it/s]
 89%|████████▊ | 9493/10708 [1:54:42<09:59,  2.03it/s]
 89%|████████▊ | 9494/10708 [1:54:42<09:57,  2.03it/s]
 89%|████████▊ | 9495/10708 [1:54:43<09:57,  2.03it/s]
 89%|████████▊ | 9496/10708 [1:54:43<09:58,  2.03it/s]
 89%|████████▊ | 9497/10708 [1:54:44<09:56,  2.03it/s]
 89%|████████▊ | 9498/10708 [1:54:44<09:55,  2.03it/s]
 89%|████████▊ | 9499/10708 [1:54:45<09:54,  2.03it/s]
 89%|████████▊ | 9500/10708 [1:54:45<09:54,  2.03it/s]
                                                      
{'loss': 3.5305, 'grad_norm': 0.1924370676279068, 'learning_rate': 3.8271032366425485e-05, 'epoch': 0.89}

 89%|████████▊ | 9500/10708 [1:54:45<09:54,  2.03it/s]
 89%|████████▊ | 9501/10708 [1:54:46<09:54,  2.03it/s]
 89%|████████▊ | 9502/10708 [1:54:46<09:53,  2.03it/s]
 89%|████████▊ | 9503/10708 [1:54:47<09:53,  2.03it/s]
 89%|████████▉ | 9504/10708 [1:54:47<09:52,  2.03it/s]
 89%|████████▉ | 9505/10708 [1:54:48<09:51,  2.03it/s]
 89%|████████▉ | 9506/10708 [1:54:48<09:51,  2.03it/s]
 89%|████████▉ | 9507/10708 [1:54:49<09:51,  2.03it/s]
 89%|████████▉ | 9508/10708 [1:54:49<09:51,  2.03it/s]
 89%|████████▉ | 9509/10708 [1:54:50<09:51,  2.03it/s]
 89%|████████▉ | 9510/10708 [1:54:50<09:50,  2.03it/s]
 89%|████████▉ | 9511/10708 [1:54:51<09:49,  2.03it/s]
 89%|████████▉ | 9512/10708 [1:54:51<09:48,  2.03it/s]
 89%|████████▉ | 9513/10708 [1:54:52<09:49,  2.03it/s]
 89%|████████▉ | 9514/10708 [1:54:52<09:48,  2.03it/s]
 89%|████████▉ | 9515/10708 [1:54:53<09:47,  2.03it/s]
 89%|████████▉ | 9516/10708 [1:54:53<09:47,  2.03it/s]
 89%|████████▉ | 9517/10708 [1:54:54<09:46,  2.03it/s]
 89%|████████▉ | 9518/10708 [1:54:54<09:45,  2.03it/s]
 89%|████████▉ | 9519/10708 [1:54:54<09:45,  2.03it/s]
 89%|████████▉ | 9520/10708 [1:54:55<09:44,  2.03it/s]
 89%|████████▉ | 9521/10708 [1:54:55<09:43,  2.03it/s]
 89%|████████▉ | 9522/10708 [1:54:56<09:43,  2.03it/s]
 89%|████████▉ | 9523/10708 [1:54:56<09:43,  2.03it/s]
 89%|████████▉ | 9524/10708 [1:54:57<09:42,  2.03it/s]
 89%|████████▉ | 9525/10708 [1:54:57<09:41,  2.03it/s]{'loss': 3.5292, 'grad_norm': 0.1975420117378235, 'learning_rate': 3.672284184009389e-05, 'epoch': 0.89}
                                                      

 89%|████████▉ | 9525/10708 [1:54:57<09:41,  2.03it/s]
 89%|████████▉ | 9526/10708 [1:54:58<09:41,  2.03it/s]
 89%|████████▉ | 9527/10708 [1:54:58<09:41,  2.03it/s]
 89%|████████▉ | 9528/10708 [1:54:59<09:41,  2.03it/s]
 89%|████████▉ | 9529/10708 [1:54:59<09:40,  2.03it/s]
 89%|████████▉ | 9530/10708 [1:55:00<09:40,  2.03it/s]
 89%|████████▉ | 9531/10708 [1:55:00<09:39,  2.03it/s]
 89%|████████▉ | 9532/10708 [1:55:01<11:16,  1.74it/s]
 89%|████████▉ | 9533/10708 [1:55:02<10:45,  1.82it/s]
 89%|████████▉ | 9534/10708 [1:55:02<10:25,  1.88it/s]
 89%|████████▉ | 9535/10708 [1:55:03<10:09,  1.92it/s]
 89%|████████▉ | 9536/10708 [1:55:03<09:59,  1.96it/s]
 89%|████████▉ | 9537/10708 [1:55:04<09:51,  1.98it/s]
 89%|████████▉ | 9538/10708 [1:55:04<09:46,  2.00it/s]
 89%|████████▉ | 9539/10708 [1:55:05<09:42,  2.01it/s]
 89%|████████▉ | 9540/10708 [1:55:05<09:39,  2.02it/s]
 89%|████████▉ | 9541/10708 [1:55:06<09:37,  2.02it/s]
 89%|████████▉ | 9542/10708 [1:55:06<09:35,  2.03it/s]
 89%|████████▉ | 9543/10708 [1:55:07<09:35,  2.02it/s]
 89%|████████▉ | 9544/10708 [1:55:07<09:34,  2.03it/s]
 89%|████████▉ | 9545/10708 [1:55:08<09:32,  2.03it/s]
 89%|████████▉ | 9546/10708 [1:55:08<09:31,  2.03it/s]
 89%|████████▉ | 9547/10708 [1:55:09<09:31,  2.03it/s]
 89%|████████▉ | 9548/10708 [1:55:09<09:30,  2.03it/s]
 89%|████████▉ | 9549/10708 [1:55:10<09:29,  2.03it/s]
 89%|████████▉ | 9550/10708 [1:55:10<09:29,  2.03it/s]
                                                      
{'loss': 3.532, 'grad_norm': 0.19647075235843658, 'learning_rate': 3.5205421812029125e-05, 'epoch': 0.89}

 89%|████████▉ | 9550/10708 [1:55:10<09:29,  2.03it/s]
 89%|████████▉ | 9551/10708 [1:55:11<09:30,  2.03it/s]
 89%|████████▉ | 9552/10708 [1:55:11<09:28,  2.03it/s]
 89%|████████▉ | 9553/10708 [1:55:11<09:29,  2.03it/s]
 89%|████████▉ | 9554/10708 [1:55:12<09:28,  2.03it/s]
 89%|████████▉ | 9555/10708 [1:55:12<09:27,  2.03it/s]
 89%|████████▉ | 9556/10708 [1:55:13<09:28,  2.03it/s]
 89%|████████▉ | 9557/10708 [1:55:13<09:28,  2.02it/s]
 89%|████████▉ | 9558/10708 [1:55:14<09:28,  2.02it/s]
 89%|████████▉ | 9559/10708 [1:55:14<09:28,  2.02it/s]
 89%|████████▉ | 9560/10708 [1:55:15<09:26,  2.03it/s]
 89%|████████▉ | 9561/10708 [1:55:15<09:26,  2.03it/s]
 89%|████████▉ | 9562/10708 [1:55:16<09:24,  2.03it/s]
 89%|████████▉ | 9563/10708 [1:55:16<09:24,  2.03it/s]
 89%|████████▉ | 9564/10708 [1:55:17<09:23,  2.03it/s]
 89%|████████▉ | 9565/10708 [1:55:17<09:21,  2.03it/s]
 89%|████████▉ | 9566/10708 [1:55:18<09:21,  2.03it/s]
 89%|████████▉ | 9567/10708 [1:55:18<09:21,  2.03it/s]
 89%|████████▉ | 9568/10708 [1:55:19<09:19,  2.04it/s]
 89%|████████▉ | 9569/10708 [1:55:19<09:19,  2.03it/s]
 89%|████████▉ | 9570/10708 [1:55:20<09:19,  2.03it/s]
 89%|████████▉ | 9571/10708 [1:55:20<09:19,  2.03it/s]
 89%|████████▉ | 9572/10708 [1:55:21<09:18,  2.03it/s]
 89%|████████▉ | 9573/10708 [1:55:21<09:18,  2.03it/s]
 89%|████████▉ | 9574/10708 [1:55:22<10:48,  1.75it/s]
 89%|████████▉ | 9575/10708 [1:55:23<10:20,  1.83it/s]{'loss': 3.524, 'grad_norm': 0.19625818729400635, 'learning_rate': 3.371887306805577e-05, 'epoch': 0.89}

                                                      

 89%|████████▉ | 9575/10708 [1:55:23<10:20,  1.83it/s]
 89%|████████▉ | 9576/10708 [1:55:23<10:00,  1.88it/s]
 89%|████████▉ | 9577/10708 [1:55:24<09:47,  1.93it/s]
 89%|████████▉ | 9578/10708 [1:55:24<09:37,  1.96it/s]
 89%|████████▉ | 9579/10708 [1:55:25<09:30,  1.98it/s]
 89%|████████▉ | 9580/10708 [1:55:25<09:25,  1.99it/s]
 89%|████████▉ | 9581/10708 [1:55:26<09:22,  2.00it/s]
 89%|████████▉ | 9582/10708 [1:55:26<09:19,  2.01it/s]
 89%|████████▉ | 9583/10708 [1:55:27<09:17,  2.02it/s]
 90%|████████▉ | 9584/10708 [1:55:27<09:15,  2.02it/s]
 90%|████████▉ | 9585/10708 [1:55:28<09:13,  2.03it/s]
 90%|████████▉ | 9586/10708 [1:55:28<09:12,  2.03it/s]
 90%|████████▉ | 9587/10708 [1:55:28<09:12,  2.03it/s]
 90%|████████▉ | 9588/10708 [1:55:29<09:11,  2.03it/s]
 90%|████████▉ | 9589/10708 [1:55:29<09:11,  2.03it/s]
 90%|████████▉ | 9590/10708 [1:55:30<09:10,  2.03it/s]
 90%|████████▉ | 9591/10708 [1:55:30<09:09,  2.03it/s]
 90%|████████▉ | 9592/10708 [1:55:31<09:08,  2.03it/s]
 90%|████████▉ | 9593/10708 [1:55:31<09:09,  2.03it/s]
 90%|████████▉ | 9594/10708 [1:55:32<09:08,  2.03it/s]
 90%|████████▉ | 9595/10708 [1:55:32<09:07,  2.03it/s]
 90%|████████▉ | 9596/10708 [1:55:33<09:06,  2.03it/s]
 90%|████████▉ | 9597/10708 [1:55:33<09:06,  2.03it/s]
 90%|████████▉ | 9598/10708 [1:55:34<09:05,  2.03it/s]
 90%|████████▉ | 9599/10708 [1:55:34<09:06,  2.03it/s]
 90%|████████▉ | 9600/10708 [1:55:35<09:05,  2.03it/s]
                                                      
{'loss': 3.5277, 'grad_norm': 0.19690194725990295, 'learning_rate': 3.2263294343552244e-05, 'epoch': 0.9}

 90%|████████▉ | 9600/10708 [1:55:35<09:05,  2.03it/s]
 90%|████████▉ | 9601/10708 [1:55:35<09:07,  2.02it/s]
 90%|████████▉ | 9602/10708 [1:55:36<09:05,  2.03it/s]
 90%|████████▉ | 9603/10708 [1:55:36<09:05,  2.03it/s]
 90%|████████▉ | 9604/10708 [1:55:37<09:04,  2.03it/s]
 90%|████████▉ | 9605/10708 [1:55:37<09:03,  2.03it/s]
 90%|████████▉ | 9606/10708 [1:55:38<09:03,  2.03it/s]
 90%|████████▉ | 9607/10708 [1:55:38<09:01,  2.03it/s]
 90%|████████▉ | 9608/10708 [1:55:39<09:01,  2.03it/s]
 90%|████████▉ | 9609/10708 [1:55:39<08:59,  2.04it/s]
 90%|████████▉ | 9610/10708 [1:55:40<09:00,  2.03it/s]
 90%|████████▉ | 9611/10708 [1:55:40<08:59,  2.03it/s]
 90%|████████▉ | 9612/10708 [1:55:41<08:58,  2.03it/s]
 90%|████████▉ | 9613/10708 [1:55:41<08:58,  2.03it/s]
 90%|████████▉ | 9614/10708 [1:55:42<08:58,  2.03it/s]
 90%|████████▉ | 9615/10708 [1:55:42<08:57,  2.04it/s]
 90%|████████▉ | 9616/10708 [1:55:43<08:57,  2.03it/s]
 90%|████████▉ | 9617/10708 [1:55:43<08:57,  2.03it/s]
 90%|████████▉ | 9618/10708 [1:55:44<08:57,  2.03it/s]
 90%|████████▉ | 9619/10708 [1:55:44<08:57,  2.03it/s]
 90%|████████▉ | 9620/10708 [1:55:45<08:55,  2.03it/s]
 90%|████████▉ | 9621/10708 [1:55:45<08:54,  2.03it/s]
 90%|████████▉ | 9622/10708 [1:55:46<08:53,  2.03it/s]
 90%|████████▉ | 9623/10708 [1:55:46<08:53,  2.03it/s]
 90%|████████▉ | 9624/10708 [1:55:47<08:52,  2.04it/s]
 90%|████████▉ | 9625/10708 [1:55:47<08:52,  2.03it/s]{'loss': 3.5238, 'grad_norm': 0.19845744967460632, 'learning_rate': 3.083878231689319e-05, 'epoch': 0.9}
                                                      

 90%|████████▉ | 9625/10708 [1:55:47<08:52,  2.03it/s]
 90%|████████▉ | 9626/10708 [1:55:48<08:52,  2.03it/s]
 90%|████████▉ | 9627/10708 [1:55:48<08:51,  2.03it/s]
 90%|████████▉ | 9628/10708 [1:55:49<08:51,  2.03it/s]
 90%|████████▉ | 9629/10708 [1:55:49<08:50,  2.03it/s]
 90%|████████▉ | 9630/10708 [1:55:50<08:49,  2.03it/s]
 90%|████████▉ | 9631/10708 [1:55:50<08:49,  2.03it/s]
 90%|████████▉ | 9632/10708 [1:55:51<08:49,  2.03it/s]
 90%|████████▉ | 9633/10708 [1:55:51<08:48,  2.03it/s]
 90%|████████▉ | 9634/10708 [1:55:52<08:48,  2.03it/s]
 90%|████████▉ | 9635/10708 [1:55:52<08:47,  2.03it/s]
 90%|████████▉ | 9636/10708 [1:55:53<08:46,  2.04it/s]
 90%|████████▉ | 9637/10708 [1:55:53<08:46,  2.04it/s]
 90%|█████████ | 9638/10708 [1:55:54<08:45,  2.03it/s]
 90%|█████████ | 9639/10708 [1:55:54<08:44,  2.04it/s]
 90%|█████████ | 9640/10708 [1:55:55<08:44,  2.04it/s]
 90%|█████████ | 9641/10708 [1:55:55<08:44,  2.03it/s]
 90%|█████████ | 9642/10708 [1:55:56<08:43,  2.04it/s]
 90%|█████████ | 9643/10708 [1:55:56<08:44,  2.03it/s]
 90%|█████████ | 9644/10708 [1:55:57<08:43,  2.03it/s]
 90%|█████████ | 9645/10708 [1:55:57<08:42,  2.04it/s]
 90%|█████████ | 9646/10708 [1:55:58<08:42,  2.03it/s]
 90%|█████████ | 9647/10708 [1:55:58<08:41,  2.03it/s]
 90%|█████████ | 9648/10708 [1:55:59<08:40,  2.04it/s]
 90%|█████████ | 9649/10708 [1:55:59<08:40,  2.03it/s]
 90%|█████████ | 9650/10708 [1:55:59<08:39,  2.04it/s]{'loss': 3.5211, 'grad_norm': 0.1968899965286255, 'learning_rate': 2.9445431603027874e-05, 'epoch': 0.9}
                                                      

 90%|█████████ | 9650/10708 [1:55:59<08:39,  2.04it/s]
 90%|█████████ | 9651/10708 [1:56:00<08:41,  2.03it/s]
 90%|█████████ | 9652/10708 [1:56:00<08:40,  2.03it/s]
 90%|█████████ | 9653/10708 [1:56:01<08:39,  2.03it/s]
 90%|█████████ | 9654/10708 [1:56:01<08:39,  2.03it/s]
 90%|█████████ | 9655/10708 [1:56:02<08:38,  2.03it/s]
 90%|█████████ | 9656/10708 [1:56:02<08:38,  2.03it/s]
 90%|█████████ | 9657/10708 [1:56:03<08:37,  2.03it/s]
 90%|█████████ | 9658/10708 [1:56:03<08:36,  2.03it/s]
 90%|█████████ | 9659/10708 [1:56:04<08:36,  2.03it/s]
 90%|█████████ | 9660/10708 [1:56:04<08:35,  2.03it/s]
 90%|█████████ | 9661/10708 [1:56:05<08:34,  2.03it/s]
 90%|█████████ | 9662/10708 [1:56:05<08:34,  2.03it/s]
 90%|█████████ | 9663/10708 [1:56:06<08:34,  2.03it/s]
 90%|█████████ | 9664/10708 [1:56:06<08:33,  2.03it/s]
 90%|█████████ | 9665/10708 [1:56:07<08:33,  2.03it/s]
 90%|█████████ | 9666/10708 [1:56:07<08:32,  2.03it/s]
 90%|█████████ | 9667/10708 [1:56:08<08:32,  2.03it/s]
 90%|█████████ | 9668/10708 [1:56:08<08:32,  2.03it/s]
 90%|█████████ | 9669/10708 [1:56:09<08:30,  2.03it/s]
 90%|█████████ | 9670/10708 [1:56:09<08:30,  2.03it/s]
 90%|█████████ | 9671/10708 [1:56:10<08:30,  2.03it/s]
 90%|█████████ | 9672/10708 [1:56:10<08:28,  2.04it/s]
 90%|█████████ | 9673/10708 [1:56:11<08:28,  2.03it/s]
 90%|█████████ | 9674/10708 [1:56:11<08:28,  2.03it/s]
 90%|█████████ | 9675/10708 [1:56:12<08:27,  2.04it/s]{'loss': 3.5197, 'grad_norm': 0.1999296098947525, 'learning_rate': 2.808333474719621e-05, 'epoch': 0.9}

                                                      

 90%|█████████ | 9675/10708 [1:56:12<08:27,  2.04it/s]
 90%|█████████ | 9676/10708 [1:56:12<08:28,  2.03it/s]
 90%|█████████ | 9677/10708 [1:56:13<08:27,  2.03it/s]
 90%|█████████ | 9678/10708 [1:56:13<08:28,  2.02it/s]
 90%|█████████ | 9679/10708 [1:56:14<08:27,  2.03it/s]
 90%|█████████ | 9680/10708 [1:56:14<08:27,  2.02it/s]
 90%|█████████ | 9681/10708 [1:56:15<08:26,  2.03it/s]
 90%|█████████ | 9682/10708 [1:56:15<08:25,  2.03it/s]
 90%|█████████ | 9683/10708 [1:56:16<08:24,  2.03it/s]
 90%|█████████ | 9684/10708 [1:56:16<08:23,  2.03it/s]
 90%|█████████ | 9685/10708 [1:56:17<08:22,  2.03it/s]
 90%|█████████ | 9686/10708 [1:56:17<08:22,  2.03it/s]
 90%|█████████ | 9687/10708 [1:56:18<08:21,  2.04it/s]
 90%|█████████ | 9688/10708 [1:56:18<08:20,  2.04it/s]
 90%|█████████ | 9689/10708 [1:56:19<08:21,  2.03it/s]
 90%|█████████ | 9690/10708 [1:56:19<08:20,  2.03it/s]
 91%|█████████ | 9691/10708 [1:56:20<08:20,  2.03it/s]
 91%|█████████ | 9692/10708 [1:56:20<08:19,  2.03it/s]
 91%|█████████ | 9693/10708 [1:56:21<08:18,  2.03it/s]
 91%|█████████ | 9694/10708 [1:56:21<08:18,  2.04it/s]
 91%|█████████ | 9695/10708 [1:56:22<08:17,  2.04it/s]
 91%|█████████ | 9696/10708 [1:56:22<08:17,  2.03it/s]
 91%|█████████ | 9697/10708 [1:56:23<08:16,  2.03it/s]
 91%|█████████ | 9698/10708 [1:56:23<08:16,  2.03it/s]
 91%|█████████ | 9699/10708 [1:56:24<08:16,  2.03it/s]
 91%|█████████ | 9700/10708 [1:56:24<08:16,  2.03it/s]
                                                      
{'loss': 3.5206, 'grad_norm': 0.1957554817199707, 'learning_rate': 2.6752582218781742e-05, 'epoch': 0.91}

 91%|█████████ | 9700/10708 [1:56:24<08:16,  2.03it/s]
 91%|█████████ | 9701/10708 [1:56:25<08:16,  2.03it/s]
 91%|█████████ | 9702/10708 [1:56:25<08:15,  2.03it/s]
 91%|█████████ | 9703/10708 [1:56:26<08:14,  2.03it/s]
 91%|█████████ | 9704/10708 [1:56:26<08:14,  2.03it/s]
 91%|█████████ | 9705/10708 [1:56:27<08:12,  2.04it/s]
 91%|█████████ | 9706/10708 [1:56:27<08:12,  2.03it/s]
 91%|█████████ | 9707/10708 [1:56:28<08:12,  2.03it/s]
 91%|█████████ | 9708/10708 [1:56:28<08:11,  2.04it/s]
 91%|█████████ | 9709/10708 [1:56:29<08:10,  2.04it/s]
 91%|█████████ | 9710/10708 [1:56:29<08:10,  2.04it/s]
 91%|█████████ | 9711/10708 [1:56:29<08:09,  2.04it/s]
 91%|█████████ | 9712/10708 [1:56:30<08:08,  2.04it/s]
 91%|█████████ | 9713/10708 [1:56:30<08:09,  2.03it/s]
 91%|█████████ | 9714/10708 [1:56:31<08:08,  2.04it/s]
 91%|█████████ | 9715/10708 [1:56:31<08:08,  2.03it/s]
 91%|█████████ | 9716/10708 [1:56:32<08:08,  2.03it/s]
 91%|█████████ | 9717/10708 [1:56:32<08:07,  2.03it/s]
 91%|█████████ | 9718/10708 [1:56:33<08:07,  2.03it/s]
 91%|█████████ | 9719/10708 [1:56:33<08:06,  2.03it/s]
 91%|█████████ | 9720/10708 [1:56:34<08:05,  2.04it/s]
 91%|█████████ | 9721/10708 [1:56:34<08:04,  2.04it/s]
 91%|█████████ | 9722/10708 [1:56:35<08:04,  2.03it/s]
 91%|█████████ | 9723/10708 [1:56:35<08:03,  2.04it/s]
 91%|█████████ | 9724/10708 [1:56:36<08:03,  2.03it/s]
 91%|█████████ | 9725/10708 [1:56:36<08:03,  2.03it/s]
                                                      
{'loss': 3.5221, 'grad_norm': 0.1975858509540558, 'learning_rate': 2.5453262405302936e-05, 'epoch': 0.91}

 91%|█████████ | 9725/10708 [1:56:36<08:03,  2.03it/s]
 91%|█████████ | 9726/10708 [1:56:37<08:03,  2.03it/s]
 91%|█████████ | 9727/10708 [1:56:37<08:02,  2.03it/s]
 91%|█████████ | 9728/10708 [1:56:38<08:02,  2.03it/s]
 91%|█████████ | 9729/10708 [1:56:38<08:01,  2.03it/s]
 91%|█████████ | 9730/10708 [1:56:39<08:01,  2.03it/s]
 91%|█████████ | 9731/10708 [1:56:39<08:00,  2.03it/s]
 91%|█████████ | 9732/10708 [1:56:40<08:00,  2.03it/s]
 91%|█████████ | 9733/10708 [1:56:40<08:00,  2.03it/s]
 91%|█████████ | 9734/10708 [1:56:41<07:59,  2.03it/s]
 91%|█████████ | 9735/10708 [1:56:41<07:58,  2.03it/s]
 91%|█████████ | 9736/10708 [1:56:42<07:58,  2.03it/s]
 91%|█████████ | 9737/10708 [1:56:42<07:57,  2.03it/s]
 91%|█████████ | 9738/10708 [1:56:43<07:58,  2.03it/s]
 91%|█████████ | 9739/10708 [1:56:43<07:58,  2.03it/s]
 91%|█████████ | 9740/10708 [1:56:44<07:57,  2.03it/s]
 91%|█████████ | 9741/10708 [1:56:44<07:56,  2.03it/s]
 91%|█████████ | 9742/10708 [1:56:45<07:55,  2.03it/s]
 91%|█████████ | 9743/10708 [1:56:45<07:54,  2.03it/s]
 91%|█████████ | 9744/10708 [1:56:46<07:53,  2.03it/s]
 91%|█████████ | 9745/10708 [1:56:46<07:53,  2.04it/s]
 91%|█████████ | 9746/10708 [1:56:47<07:52,  2.03it/s]
 91%|█████████ | 9747/10708 [1:56:47<07:51,  2.04it/s]
 91%|█████████ | 9748/10708 [1:56:48<07:51,  2.04it/s]
 91%|█████████ | 9749/10708 [1:56:48<07:50,  2.04it/s]
 91%|█████████ | 9750/10708 [1:56:49<07:50,  2.03it/s]
                                                      
{'loss': 3.5167, 'grad_norm': 0.20004171133041382, 'learning_rate': 2.4185461606542446e-05, 'epoch': 0.91}

 91%|█████████ | 9750/10708 [1:56:49<07:50,  2.03it/s]
 91%|█████████ | 9751/10708 [1:56:49<07:51,  2.03it/s]
 91%|█████████ | 9752/10708 [1:56:50<07:50,  2.03it/s]
 91%|█████████ | 9753/10708 [1:56:50<07:49,  2.03it/s]
 91%|█████████ | 9754/10708 [1:56:51<07:49,  2.03it/s]
 91%|█████████ | 9755/10708 [1:56:51<07:48,  2.03it/s]
 91%|█████████ | 9756/10708 [1:56:52<07:48,  2.03it/s]
 91%|█████████ | 9757/10708 [1:56:52<07:48,  2.03it/s]
 91%|█████████ | 9758/10708 [1:56:53<07:47,  2.03it/s]
 91%|█████████ | 9759/10708 [1:56:53<07:47,  2.03it/s]
 91%|█████████ | 9760/10708 [1:56:54<07:47,  2.03it/s]
 91%|█████████ | 9761/10708 [1:56:54<07:46,  2.03it/s]
 91%|█████████ | 9762/10708 [1:56:55<07:46,  2.03it/s]
 91%|█████████ | 9763/10708 [1:56:55<07:45,  2.03it/s]
 91%|█████████ | 9764/10708 [1:56:56<07:43,  2.03it/s]
 91%|█████████ | 9765/10708 [1:56:56<07:43,  2.03it/s]
 91%|█████████ | 9766/10708 [1:56:57<07:43,  2.03it/s]
 91%|█████████ | 9767/10708 [1:56:57<07:42,  2.03it/s]
 91%|█████████ | 9768/10708 [1:56:58<07:42,  2.03it/s]
 91%|█████████ | 9769/10708 [1:56:58<07:41,  2.03it/s]
 91%|█████████ | 9770/10708 [1:56:59<07:41,  2.03it/s]
 91%|█████████ | 9771/10708 [1:56:59<07:42,  2.03it/s]
 91%|█████████▏| 9772/10708 [1:57:00<07:41,  2.03it/s]
 91%|█████████▏| 9773/10708 [1:57:00<07:41,  2.03it/s]
 91%|█████████▏| 9774/10708 [1:57:00<07:39,  2.03it/s]
 91%|█████████▏| 9775/10708 [1:57:01<07:38,  2.03it/s]{'loss': 3.5186, 'grad_norm': 0.1931161731481552, 'learning_rate': 2.2949264028815253e-05, 'epoch': 0.91}

                                                      

 91%|█████████▏| 9775/10708 [1:57:01<07:38,  2.03it/s]
 91%|█████████▏| 9776/10708 [1:57:01<07:39,  2.03it/s]
 91%|█████████▏| 9777/10708 [1:57:02<07:38,  2.03it/s]
 91%|█████████▏| 9778/10708 [1:57:02<07:38,  2.03it/s]
 91%|█████████▏| 9779/10708 [1:57:03<07:37,  2.03it/s]
 91%|█████████▏| 9780/10708 [1:57:03<07:37,  2.03it/s]
 91%|█████████▏| 9781/10708 [1:57:04<07:36,  2.03it/s]
 91%|█████████▏| 9782/10708 [1:57:04<07:35,  2.03it/s]
 91%|█████████▏| 9783/10708 [1:57:05<07:35,  2.03it/s]
 91%|█████████▏| 9784/10708 [1:57:05<07:34,  2.03it/s]
 91%|█████████▏| 9785/10708 [1:57:06<07:34,  2.03it/s]
 91%|█████████▏| 9786/10708 [1:57:06<07:33,  2.03it/s]
 91%|█████████▏| 9787/10708 [1:57:07<07:33,  2.03it/s]
 91%|█████████▏| 9788/10708 [1:57:07<07:32,  2.03it/s]
 91%|█████████▏| 9789/10708 [1:57:08<07:31,  2.03it/s]
 91%|█████████▏| 9790/10708 [1:57:08<07:31,  2.03it/s]
 91%|█████████▏| 9791/10708 [1:57:09<07:30,  2.03it/s]
 91%|█████████▏| 9792/10708 [1:57:09<07:29,  2.04it/s]
 91%|█████████▏| 9793/10708 [1:57:10<07:29,  2.03it/s]
 91%|█████████▏| 9794/10708 [1:57:10<07:29,  2.03it/s]
 91%|█████████▏| 9795/10708 [1:57:11<07:29,  2.03it/s]
 91%|█████████▏| 9796/10708 [1:57:11<07:28,  2.03it/s]
 91%|█████████▏| 9797/10708 [1:57:12<07:27,  2.03it/s]
 92%|█████████▏| 9798/10708 [1:57:12<07:27,  2.03it/s]
 92%|█████████▏| 9799/10708 [1:57:13<07:28,  2.03it/s]
 92%|█████████▏| 9800/10708 [1:57:13<07:28,  2.02it/s]{'loss': 3.5172, 'grad_norm': 0.19623972475528717, 'learning_rate': 2.1744751779375738e-05, 'epoch': 0.92}
                                                      

 92%|█████████▏| 9800/10708 [1:57:13<07:28,  2.02it/s]
 92%|█████████▏| 9801/10708 [1:57:14<07:27,  2.02it/s]
 92%|█████████▏| 9802/10708 [1:57:14<07:27,  2.03it/s]
 92%|█████████▏| 9803/10708 [1:57:15<07:26,  2.03it/s]
 92%|█████████▏| 9804/10708 [1:57:15<07:25,  2.03it/s]
 92%|█████████▏| 9805/10708 [1:57:16<07:24,  2.03it/s]
 92%|█████████▏| 9806/10708 [1:57:16<07:24,  2.03it/s]
 92%|█████████▏| 9807/10708 [1:57:17<07:23,  2.03it/s]
 92%|█████████▏| 9808/10708 [1:57:17<07:22,  2.03it/s]
 92%|█████████▏| 9809/10708 [1:57:18<07:22,  2.03it/s]
 92%|█████████▏| 9810/10708 [1:57:18<07:21,  2.04it/s]
 92%|█████████▏| 9811/10708 [1:57:19<07:20,  2.04it/s]
 92%|█████████▏| 9812/10708 [1:57:19<07:20,  2.03it/s]
 92%|█████████▏| 9813/10708 [1:57:20<07:19,  2.04it/s]
 92%|█████████▏| 9814/10708 [1:57:20<07:18,  2.04it/s]
 92%|█████████▏| 9815/10708 [1:57:21<07:18,  2.03it/s]
 92%|█████████▏| 9816/10708 [1:57:21<07:17,  2.04it/s]
 92%|█████████▏| 9817/10708 [1:57:22<07:17,  2.04it/s]
 92%|█████████▏| 9818/10708 [1:57:22<07:17,  2.04it/s]
 92%|█████████▏| 9819/10708 [1:57:23<07:17,  2.03it/s]
 92%|█████████▏| 9820/10708 [1:57:23<07:17,  2.03it/s]
 92%|█████████▏| 9821/10708 [1:57:24<07:16,  2.03it/s]
 92%|█████████▏| 9822/10708 [1:57:24<07:16,  2.03it/s]
 92%|█████████▏| 9823/10708 [1:57:25<07:15,  2.03it/s]
 92%|█████████▏| 9824/10708 [1:57:25<07:15,  2.03it/s]
 92%|█████████▏| 9825/10708 [1:57:26<07:14,  2.03it/s]{'loss': 3.5177, 'grad_norm': 0.19945721328258514, 'learning_rate': 2.0572004860963834e-05, 'epoch': 0.92}
                                                      

 92%|█████████▏| 9825/10708 [1:57:26<07:14,  2.03it/s]
 92%|█████████▏| 9826/10708 [1:57:26<07:14,  2.03it/s]
 92%|█████████▏| 9827/10708 [1:57:27<07:13,  2.03it/s]
 92%|█████████▏| 9828/10708 [1:57:27<07:13,  2.03it/s]
 92%|█████████▏| 9829/10708 [1:57:28<07:12,  2.03it/s]
 92%|█████████▏| 9830/10708 [1:57:28<07:11,  2.03it/s]
 92%|█████████▏| 9831/10708 [1:57:29<07:11,  2.03it/s]
 92%|█████████▏| 9832/10708 [1:57:29<07:10,  2.03it/s]
 92%|█████████▏| 9833/10708 [1:57:30<07:10,  2.03it/s]
 92%|█████████▏| 9834/10708 [1:57:30<07:09,  2.03it/s]
 92%|█████████▏| 9835/10708 [1:57:31<07:09,  2.03it/s]
 92%|█████████▏| 9836/10708 [1:57:31<07:08,  2.03it/s]
 92%|█████████▏| 9837/10708 [1:57:31<07:07,  2.04it/s]
 92%|█████████▏| 9838/10708 [1:57:32<07:07,  2.03it/s]
 92%|█████████▏| 9839/10708 [1:57:32<07:06,  2.04it/s]
 92%|█████████▏| 9840/10708 [1:57:33<07:06,  2.03it/s]
 92%|█████████▏| 9841/10708 [1:57:33<07:06,  2.03it/s]
 92%|█████████▏| 9842/10708 [1:57:34<07:05,  2.04it/s]
 92%|█████████▏| 9843/10708 [1:57:34<07:04,  2.04it/s]
 92%|█████████▏| 9844/10708 [1:57:35<07:04,  2.03it/s]
 92%|█████████▏| 9845/10708 [1:57:35<07:03,  2.04it/s]
 92%|█████████▏| 9846/10708 [1:57:36<07:03,  2.04it/s]
 92%|█████████▏| 9847/10708 [1:57:36<07:03,  2.03it/s]
 92%|█████████▏| 9848/10708 [1:57:37<07:02,  2.03it/s]
 92%|█████████▏| 9849/10708 [1:57:37<07:02,  2.04it/s]
 92%|█████████▏| 9850/10708 [1:57:38<07:01,  2.03it/s]
                                                      
{'loss': 3.5186, 'grad_norm': 0.19391712546348572, 'learning_rate': 1.9431101166491882e-05, 'epoch': 0.92}

 92%|█████████▏| 9850/10708 [1:57:38<07:01,  2.03it/s]
 92%|█████████▏| 9851/10708 [1:57:38<07:01,  2.03it/s]
 92%|█████████▏| 9852/10708 [1:57:39<07:01,  2.03it/s]
 92%|█████████▏| 9853/10708 [1:57:39<07:00,  2.03it/s]
 92%|█████████▏| 9854/10708 [1:57:40<07:00,  2.03it/s]
 92%|█████████▏| 9855/10708 [1:57:40<06:59,  2.03it/s]
 92%|█████████▏| 9856/10708 [1:57:41<06:59,  2.03it/s]
 92%|█████████▏| 9857/10708 [1:57:41<06:59,  2.03it/s]
 92%|█████████▏| 9858/10708 [1:57:42<06:58,  2.03it/s]
 92%|█████████▏| 9859/10708 [1:57:42<06:57,  2.03it/s]
 92%|█████████▏| 9860/10708 [1:57:43<06:57,  2.03it/s]
 92%|█████████▏| 9861/10708 [1:57:43<06:57,  2.03it/s]
 92%|█████████▏| 9862/10708 [1:57:44<06:57,  2.03it/s]
 92%|█████████▏| 9863/10708 [1:57:44<06:56,  2.03it/s]
 92%|█████████▏| 9864/10708 [1:57:45<06:55,  2.03it/s]
 92%|█████████▏| 9865/10708 [1:57:45<06:55,  2.03it/s]
 92%|█████████▏| 9866/10708 [1:57:46<06:54,  2.03it/s]
 92%|█████████▏| 9867/10708 [1:57:46<06:53,  2.03it/s]
 92%|█████████▏| 9868/10708 [1:57:47<06:53,  2.03it/s]
 92%|█████████▏| 9869/10708 [1:57:47<06:52,  2.04it/s]
 92%|█████████▏| 9870/10708 [1:57:48<06:52,  2.03it/s]
 92%|█████████▏| 9871/10708 [1:57:48<06:51,  2.03it/s]
 92%|█████████▏| 9872/10708 [1:57:49<06:51,  2.03it/s]
 92%|█████████▏| 9873/10708 [1:57:49<06:50,  2.03it/s]
 92%|█████████▏| 9874/10708 [1:57:50<06:49,  2.03it/s]
 92%|█████████▏| 9875/10708 [1:57:50<06:49,  2.03it/s]
                                                      
{'loss': 3.5266, 'grad_norm': 0.19379350543022156, 'learning_rate': 1.8322116473870555e-05, 'epoch': 0.92}

 92%|█████████▏| 9875/10708 [1:57:50<06:49,  2.03it/s]
 92%|█████████▏| 9876/10708 [1:57:51<06:50,  2.03it/s]
 92%|█████████▏| 9877/10708 [1:57:51<06:48,  2.03it/s]
 92%|█████████▏| 9878/10708 [1:57:52<06:48,  2.03it/s]
 92%|█████████▏| 9879/10708 [1:57:52<06:47,  2.03it/s]
 92%|█████████▏| 9880/10708 [1:57:53<06:47,  2.03it/s]
 92%|█████████▏| 9881/10708 [1:57:53<06:46,  2.03it/s]
 92%|█████████▏| 9882/10708 [1:57:54<06:45,  2.03it/s]
 92%|█████████▏| 9883/10708 [1:57:54<06:45,  2.04it/s]
 92%|█████████▏| 9884/10708 [1:57:55<06:44,  2.04it/s]
 92%|█████████▏| 9885/10708 [1:57:55<06:44,  2.03it/s]
 92%|█████████▏| 9886/10708 [1:57:56<06:44,  2.03it/s]
 92%|█████████▏| 9887/10708 [1:57:56<06:44,  2.03it/s]
 92%|█████████▏| 9888/10708 [1:57:57<06:43,  2.03it/s]
 92%|█████████▏| 9889/10708 [1:57:57<06:42,  2.03it/s]
 92%|█████████▏| 9890/10708 [1:57:58<06:43,  2.03it/s]
 92%|█████████▏| 9891/10708 [1:57:58<06:42,  2.03it/s]
 92%|█████████▏| 9892/10708 [1:57:59<06:41,  2.03it/s]
 92%|█████████▏| 9893/10708 [1:57:59<06:40,  2.03it/s]
 92%|█████████▏| 9894/10708 [1:58:00<06:39,  2.04it/s]
 92%|█████████▏| 9895/10708 [1:58:00<06:39,  2.03it/s]
 92%|█████████▏| 9896/10708 [1:58:01<06:39,  2.03it/s]
 92%|█████████▏| 9897/10708 [1:58:01<06:38,  2.03it/s]
 92%|█████████▏| 9898/10708 [1:58:02<06:39,  2.03it/s]
 92%|█████████▏| 9899/10708 [1:58:02<06:38,  2.03it/s]
 92%|█████████▏| 9900/10708 [1:58:02<06:38,  2.03it/s]{'loss': 3.522, 'grad_norm': 0.19417846202850342, 'learning_rate': 1.7245124440976202e-05, 'epoch': 0.92}
                                                      

 92%|█████████▏| 9900/10708 [1:58:02<06:38,  2.03it/s]
 92%|█████████▏| 9901/10708 [1:58:03<06:37,  2.03it/s]
 92%|█████████▏| 9902/10708 [1:58:03<06:36,  2.03it/s]
 92%|█████████▏| 9903/10708 [1:58:04<06:35,  2.03it/s]
 92%|█████████▏| 9904/10708 [1:58:04<06:35,  2.03it/s]
 93%|█████████▎| 9905/10708 [1:58:05<06:34,  2.04it/s]
 93%|█████████▎| 9906/10708 [1:58:05<06:34,  2.03it/s]
 93%|█████████▎| 9907/10708 [1:58:06<06:33,  2.03it/s]
 93%|█████████▎| 9908/10708 [1:58:06<06:32,  2.04it/s]
 93%|█████████▎| 9909/10708 [1:58:07<06:32,  2.04it/s]
 93%|█████████▎| 9910/10708 [1:58:07<06:32,  2.03it/s]
 93%|█████████▎| 9911/10708 [1:58:08<06:31,  2.04it/s]
 93%|█████████▎| 9912/10708 [1:58:08<06:30,  2.04it/s]
 93%|█████████▎| 9913/10708 [1:58:09<06:30,  2.04it/s]
 93%|█████████▎| 9914/10708 [1:58:09<06:29,  2.04it/s]
 93%|█████████▎| 9915/10708 [1:58:10<06:29,  2.04it/s]
 93%|█████████▎| 9916/10708 [1:58:10<06:29,  2.03it/s]
 93%|█████████▎| 9917/10708 [1:58:11<06:29,  2.03it/s]
 93%|█████████▎| 9918/10708 [1:58:11<06:28,  2.03it/s]
 93%|█████████▎| 9919/10708 [1:58:12<06:28,  2.03it/s]
 93%|█████████▎| 9920/10708 [1:58:12<06:27,  2.03it/s]
 93%|█████████▎| 9921/10708 [1:58:13<06:27,  2.03it/s]
 93%|█████████▎| 9922/10708 [1:58:13<06:28,  2.03it/s]
 93%|█████████▎| 9923/10708 [1:58:14<06:27,  2.03it/s]
 93%|█████████▎| 9924/10708 [1:58:14<06:26,  2.03it/s]
 93%|█████████▎| 9925/10708 [1:58:15<06:25,  2.03it/s]{'loss': 3.5239, 'grad_norm': 0.19219036400318146, 'learning_rate': 1.6200196600758e-05, 'epoch': 0.93}
                                                      

 93%|█████████▎| 9925/10708 [1:58:15<06:25,  2.03it/s]
 93%|█████████▎| 9926/10708 [1:58:15<06:25,  2.03it/s]
 93%|█████████▎| 9927/10708 [1:58:16<06:25,  2.02it/s]
 93%|█████████▎| 9928/10708 [1:58:16<06:24,  2.03it/s]
 93%|█████████▎| 9929/10708 [1:58:17<06:23,  2.03it/s]
 93%|█████████▎| 9930/10708 [1:58:17<06:22,  2.03it/s]
 93%|█████████▎| 9931/10708 [1:58:18<06:22,  2.03it/s]
 93%|█████████▎| 9932/10708 [1:58:18<06:21,  2.03it/s]
 93%|█████████▎| 9933/10708 [1:58:19<06:20,  2.03it/s]
 93%|█████████▎| 9934/10708 [1:58:19<06:20,  2.04it/s]
 93%|█████████▎| 9935/10708 [1:58:20<06:19,  2.04it/s]
 93%|█████████▎| 9936/10708 [1:58:20<06:19,  2.03it/s]
 93%|█████████▎| 9937/10708 [1:58:21<06:19,  2.03it/s]
 93%|█████████▎| 9938/10708 [1:58:21<06:18,  2.03it/s]
 93%|█████████▎| 9939/10708 [1:58:22<06:18,  2.03it/s]
 93%|█████████▎| 9940/10708 [1:58:22<06:17,  2.03it/s]
 93%|█████████▎| 9941/10708 [1:58:23<06:17,  2.03it/s]
 93%|█████████▎| 9942/10708 [1:58:23<06:16,  2.03it/s]
 93%|█████████▎| 9943/10708 [1:58:24<06:16,  2.03it/s]
 93%|█████████▎| 9944/10708 [1:58:24<06:15,  2.03it/s]
 93%|█████████▎| 9945/10708 [1:58:25<06:14,  2.04it/s]
 93%|█████████▎| 9946/10708 [1:58:25<06:14,  2.04it/s]
 93%|█████████▎| 9947/10708 [1:58:26<06:13,  2.04it/s]
 93%|█████████▎| 9948/10708 [1:58:26<06:13,  2.04it/s]
 93%|█████████▎| 9949/10708 [1:58:27<06:13,  2.03it/s]
 93%|█████████▎| 9950/10708 [1:58:27<06:12,  2.03it/s]
                                                      
{'loss': 3.5211, 'grad_norm': 0.19292192161083221, 'learning_rate': 1.5187402356487357e-05, 'epoch': 0.93}

 93%|█████████▎| 9950/10708 [1:58:27<06:12,  2.03it/s]
 93%|█████████▎| 9951/10708 [1:58:28<06:12,  2.03it/s]
 93%|█████████▎| 9952/10708 [1:58:28<06:12,  2.03it/s]
 93%|█████████▎| 9953/10708 [1:58:29<06:11,  2.03it/s]
 93%|█████████▎| 9954/10708 [1:58:29<06:11,  2.03it/s]
 93%|█████████▎| 9955/10708 [1:58:30<06:10,  2.03it/s]
 93%|█████████▎| 9956/10708 [1:58:30<06:09,  2.04it/s]
 93%|█████████▎| 9957/10708 [1:58:31<06:08,  2.04it/s]
 93%|█████████▎| 9958/10708 [1:58:31<06:08,  2.03it/s]
 93%|█████████▎| 9959/10708 [1:58:32<06:08,  2.03it/s]
 93%|█████████▎| 9960/10708 [1:58:32<06:07,  2.03it/s]
 93%|█████████▎| 9961/10708 [1:58:32<06:07,  2.03it/s]
 93%|█████████▎| 9962/10708 [1:58:33<06:06,  2.04it/s]
 93%|█████████▎| 9963/10708 [1:58:33<06:05,  2.04it/s]
 93%|█████████▎| 9964/10708 [1:58:34<06:05,  2.04it/s]
 93%|█████████▎| 9965/10708 [1:58:34<06:04,  2.04it/s]
 93%|█████████▎| 9966/10708 [1:58:35<06:05,  2.03it/s]
 93%|█████████▎| 9967/10708 [1:58:35<06:04,  2.03it/s]
 93%|█████████▎| 9968/10708 [1:58:36<06:03,  2.03it/s]
 93%|█████████▎| 9969/10708 [1:58:36<06:03,  2.03it/s]
 93%|█████████▎| 9970/10708 [1:58:37<06:03,  2.03it/s]
 93%|█████████▎| 9971/10708 [1:58:37<06:02,  2.03it/s]
 93%|█████████▎| 9972/10708 [1:58:38<06:02,  2.03it/s]
 93%|█████████▎| 9973/10708 [1:58:38<06:01,  2.03it/s]
 93%|█████████▎| 9974/10708 [1:58:39<06:01,  2.03it/s]
 93%|█████████▎| 9975/10708 [1:58:39<06:00,  2.03it/s]
                                                      
{'loss': 3.5251, 'grad_norm': 0.19535967707633972, 'learning_rate': 1.4206808977147767e-05, 'epoch': 0.93}

 93%|█████████▎| 9975/10708 [1:58:39<06:00,  2.03it/s]
 93%|█████████▎| 9976/10708 [1:58:40<06:00,  2.03it/s]
 93%|█████████▎| 9977/10708 [1:58:40<06:00,  2.03it/s]
 93%|█████████▎| 9978/10708 [1:58:41<05:59,  2.03it/s]
 93%|█████████▎| 9979/10708 [1:58:41<05:58,  2.03it/s]
 93%|█████████▎| 9980/10708 [1:58:42<05:59,  2.03it/s]
 93%|█████████▎| 9981/10708 [1:58:42<05:58,  2.03it/s]
 93%|█████████▎| 9982/10708 [1:58:43<05:58,  2.03it/s]
 93%|█████████▎| 9983/10708 [1:58:43<05:57,  2.03it/s]
 93%|█████████▎| 9984/10708 [1:58:44<05:57,  2.03it/s]
 93%|█████████▎| 9985/10708 [1:58:44<05:56,  2.03it/s]
 93%|█████████▎| 9986/10708 [1:58:45<05:55,  2.03it/s]
 93%|█████████▎| 9987/10708 [1:58:45<05:55,  2.03it/s]
 93%|█████████▎| 9988/10708 [1:58:46<05:54,  2.03it/s]
 93%|█████████▎| 9989/10708 [1:58:46<05:53,  2.03it/s]
 93%|█████████▎| 9990/10708 [1:58:47<05:53,  2.03it/s]
 93%|█████████▎| 9991/10708 [1:58:47<05:52,  2.03it/s]
 93%|█████████▎| 9992/10708 [1:58:48<05:52,  2.03it/s]
 93%|█████████▎| 9993/10708 [1:58:48<05:52,  2.03it/s]
 93%|█████████▎| 9994/10708 [1:58:49<05:51,  2.03it/s]
 93%|█████████▎| 9995/10708 [1:58:49<05:51,  2.03it/s]
 93%|█████████▎| 9996/10708 [1:58:50<05:50,  2.03it/s]
 93%|█████████▎| 9997/10708 [1:58:50<05:50,  2.03it/s]
 93%|█████████▎| 9998/10708 [1:58:51<05:49,  2.03it/s]
 93%|█████████▎| 9999/10708 [1:58:51<05:48,  2.03it/s]
 93%|█████████▎| 10000/10708 [1:58:52<05:48,  2.03it/s]
                                                       
{'loss': 3.5206, 'grad_norm': 0.19180549681186676, 'learning_rate': 1.3258481592967165e-05, 'epoch': 0.93}

 93%|█████████▎| 10000/10708 [1:58:52<05:48,  2.03it/s]
 93%|█████████▎| 10001/10708 [1:58:52<05:48,  2.03it/s]
 93%|█████████▎| 10002/10708 [1:58:53<05:47,  2.03it/s]
 93%|█████████▎| 10003/10708 [1:58:53<05:47,  2.03it/s]
 93%|█████████▎| 10004/10708 [1:58:54<05:46,  2.03it/s]
 93%|█████████▎| 10005/10708 [1:58:54<05:45,  2.03it/s]
 93%|█████████▎| 10006/10708 [1:58:55<05:44,  2.04it/s]
 93%|█████████▎| 10007/10708 [1:58:55<05:44,  2.03it/s]
 93%|█████████▎| 10008/10708 [1:58:56<05:44,  2.03it/s]
 93%|█████████▎| 10009/10708 [1:58:56<05:44,  2.03it/s]
 93%|█████████▎| 10010/10708 [1:58:57<05:43,  2.03it/s]
 93%|█████████▎| 10011/10708 [1:58:57<05:42,  2.03it/s]
 94%|█████████▎| 10012/10708 [1:58:58<05:42,  2.03it/s]
 94%|█████████▎| 10013/10708 [1:58:58<05:41,  2.03it/s]
 94%|█████████▎| 10014/10708 [1:58:59<05:41,  2.03it/s]
 94%|█████████▎| 10015/10708 [1:58:59<05:40,  2.03it/s]
 94%|█████████▎| 10016/10708 [1:59:00<05:40,  2.03it/s]
 94%|█████████▎| 10017/10708 [1:59:00<05:39,  2.03it/s]
 94%|█████████▎| 10018/10708 [1:59:01<05:38,  2.04it/s]
 94%|█████████▎| 10019/10708 [1:59:01<05:38,  2.03it/s]
 94%|█████████▎| 10020/10708 [1:59:02<05:37,  2.04it/s]
 94%|█████████▎| 10021/10708 [1:59:02<05:37,  2.04it/s]
 94%|█████████▎| 10022/10708 [1:59:03<05:37,  2.03it/s]
 94%|█████████▎| 10023/10708 [1:59:03<05:36,  2.04it/s]
 94%|█████████▎| 10024/10708 [1:59:03<05:35,  2.04it/s]
 94%|█████████▎| 10025/10708 [1:59:04<05:35,  2.04it/s]
                                                       
{'loss': 3.5179, 'grad_norm': 0.1912202388048172, 'learning_rate': 1.2342483191091769e-05, 'epoch': 0.94}

 94%|█████████▎| 10025/10708 [1:59:04<05:35,  2.04it/s]
 94%|█████████▎| 10026/10708 [1:59:04<05:35,  2.03it/s]
 94%|█████████▎| 10027/10708 [1:59:05<05:35,  2.03it/s]
 94%|█████████▎| 10028/10708 [1:59:05<05:34,  2.03it/s]
 94%|█████████▎| 10029/10708 [1:59:06<05:34,  2.03it/s]
 94%|█████████▎| 10030/10708 [1:59:06<05:33,  2.03it/s]
 94%|█████████▎| 10031/10708 [1:59:07<05:32,  2.03it/s]
 94%|█████████▎| 10032/10708 [1:59:07<05:32,  2.03it/s]
 94%|█████████▎| 10033/10708 [1:59:08<05:31,  2.03it/s]
 94%|█████████▎| 10034/10708 [1:59:08<05:30,  2.04it/s]
 94%|█████████▎| 10035/10708 [1:59:09<05:30,  2.04it/s]
 94%|█████████▎| 10036/10708 [1:59:09<05:30,  2.04it/s]
 94%|█████████▎| 10037/10708 [1:59:10<05:29,  2.03it/s]
 94%|█████████▎| 10038/10708 [1:59:10<05:29,  2.04it/s]
 94%|█████████▍| 10039/10708 [1:59:11<05:28,  2.03it/s]
 94%|█████████▍| 10040/10708 [1:59:11<05:28,  2.04it/s]
 94%|█████████▍| 10041/10708 [1:59:12<05:27,  2.04it/s]
 94%|█████████▍| 10042/10708 [1:59:12<05:27,  2.04it/s]
 94%|█████████▍| 10043/10708 [1:59:13<05:27,  2.03it/s]
 94%|█████████▍| 10044/10708 [1:59:13<05:27,  2.03it/s]
 94%|█████████▍| 10045/10708 [1:59:14<05:26,  2.03it/s]
 94%|█████████▍| 10046/10708 [1:59:14<05:26,  2.03it/s]
 94%|█████████▍| 10047/10708 [1:59:15<05:25,  2.03it/s]
 94%|█████████▍| 10048/10708 [1:59:15<05:24,  2.03it/s]
 94%|█████████▍| 10049/10708 [1:59:16<05:24,  2.03it/s]
 94%|█████████▍| 10050/10708 [1:59:16<05:24,  2.03it/s]
                                                       
{'loss': 3.5185, 'grad_norm': 0.19382037222385406, 'learning_rate': 1.145887461140277e-05, 'epoch': 0.94}

 94%|█████████▍| 10050/10708 [1:59:16<05:24,  2.03it/s]
 94%|█████████▍| 10051/10708 [1:59:17<05:24,  2.03it/s]
 94%|█████████▍| 10052/10708 [1:59:17<05:23,  2.03it/s]
 94%|█████████▍| 10053/10708 [1:59:18<05:22,  2.03it/s]
 94%|█████████▍| 10054/10708 [1:59:18<05:22,  2.03it/s]
 94%|█████████▍| 10055/10708 [1:59:19<05:21,  2.03it/s]
 94%|█████████▍| 10056/10708 [1:59:19<05:20,  2.03it/s]
 94%|█████████▍| 10057/10708 [1:59:20<05:20,  2.03it/s]
 94%|█████████▍| 10058/10708 [1:59:20<05:19,  2.03it/s]
 94%|█████████▍| 10059/10708 [1:59:21<05:19,  2.03it/s]
 94%|█████████▍| 10060/10708 [1:59:21<05:18,  2.03it/s]
 94%|█████████▍| 10061/10708 [1:59:22<05:18,  2.03it/s]
 94%|█████████▍| 10062/10708 [1:59:22<05:17,  2.03it/s]
 94%|█████████▍| 10063/10708 [1:59:23<05:17,  2.03it/s]
 94%|█████████▍| 10064/10708 [1:59:23<05:16,  2.03it/s]
 94%|█████████▍| 10065/10708 [1:59:24<05:16,  2.03it/s]
 94%|█████████▍| 10066/10708 [1:59:24<05:15,  2.04it/s]
 94%|█████████▍| 10067/10708 [1:59:25<05:15,  2.03it/s]
 94%|█████████▍| 10068/10708 [1:59:25<05:14,  2.03it/s]
 94%|█████████▍| 10069/10708 [1:59:26<05:14,  2.04it/s]
 94%|█████████▍| 10070/10708 [1:59:26<05:13,  2.03it/s]
 94%|█████████▍| 10071/10708 [1:59:27<05:13,  2.03it/s]
 94%|█████████▍| 10072/10708 [1:59:27<05:12,  2.03it/s]
 94%|█████████▍| 10073/10708 [1:59:28<05:12,  2.03it/s]
 94%|█████████▍| 10074/10708 [1:59:28<05:12,  2.03it/s]
 94%|█████████▍| 10075/10708 [1:59:29<05:11,  2.03it/s]
                                                       
{'loss': 3.5188, 'grad_norm': 0.19549855589866638, 'learning_rate': 1.0607714542475222e-05, 'epoch': 0.94}

 94%|█████████▍| 10075/10708 [1:59:29<05:11,  2.03it/s]
 94%|█████████▍| 10076/10708 [1:59:29<05:11,  2.03it/s]
 94%|█████████▍| 10077/10708 [1:59:30<05:10,  2.03it/s]
 94%|█████████▍| 10078/10708 [1:59:30<05:10,  2.03it/s]
 94%|█████████▍| 10079/10708 [1:59:31<05:09,  2.03it/s]
 94%|█████████▍| 10080/10708 [1:59:31<05:08,  2.04it/s]
 94%|█████████▍| 10081/10708 [1:59:32<05:07,  2.04it/s]
 94%|█████████▍| 10082/10708 [1:59:32<05:07,  2.04it/s]
 94%|█████████▍| 10083/10708 [1:59:33<05:06,  2.04it/s]
 94%|█████████▍| 10084/10708 [1:59:33<05:06,  2.03it/s]
 94%|█████████▍| 10085/10708 [1:59:33<05:06,  2.03it/s]
 94%|█████████▍| 10086/10708 [1:59:34<05:05,  2.03it/s]
 94%|█████████▍| 10087/10708 [1:59:34<05:05,  2.03it/s]
 94%|█████████▍| 10088/10708 [1:59:35<05:04,  2.03it/s]
 94%|█████████▍| 10089/10708 [1:59:35<05:04,  2.03it/s]
 94%|█████████▍| 10090/10708 [1:59:36<05:03,  2.03it/s]
 94%|█████████▍| 10091/10708 [1:59:36<05:03,  2.03it/s]
 94%|█████████▍| 10092/10708 [1:59:37<05:02,  2.03it/s]
 94%|█████████▍| 10093/10708 [1:59:37<05:02,  2.03it/s]
 94%|█████████▍| 10094/10708 [1:59:38<05:01,  2.04it/s]
 94%|█████████▍| 10095/10708 [1:59:38<05:01,  2.03it/s]
 94%|█████████▍| 10096/10708 [1:59:39<05:01,  2.03it/s]
 94%|█████████▍| 10097/10708 [1:59:39<05:00,  2.03it/s]
 94%|█████████▍| 10098/10708 [1:59:40<04:59,  2.03it/s]
 94%|█████████▍| 10099/10708 [1:59:40<04:59,  2.03it/s]
 94%|█████████▍| 10100/10708 [1:59:41<04:58,  2.04it/s]
                                                       {'loss': 3.5114, 'grad_norm': 0.1959145963191986, 'learning_rate': 9.78905951768011e-06, 'epoch': 0.94}

 94%|█████████▍| 10100/10708 [1:59:41<04:58,  2.04it/s]
 94%|█████████▍| 10101/10708 [1:59:41<04:58,  2.03it/s]
 94%|█████████▍| 10102/10708 [1:59:42<04:58,  2.03it/s]
 94%|█████████▍| 10103/10708 [1:59:42<04:57,  2.03it/s]
 94%|█████████▍| 10104/10708 [1:59:43<04:58,  2.03it/s]
 94%|█████████▍| 10105/10708 [1:59:43<04:57,  2.03it/s]
 94%|█████████▍| 10106/10708 [1:59:44<04:57,  2.03it/s]
 94%|█████████▍| 10107/10708 [1:59:44<04:55,  2.03it/s]
 94%|█████████▍| 10108/10708 [1:59:45<04:55,  2.03it/s]
 94%|█████████▍| 10109/10708 [1:59:45<04:54,  2.03it/s]
 94%|█████████▍| 10110/10708 [1:59:46<04:54,  2.03it/s]
 94%|█████████▍| 10111/10708 [1:59:46<04:53,  2.03it/s]
 94%|█████████▍| 10112/10708 [1:59:47<04:53,  2.03it/s]
 94%|█████████▍| 10113/10708 [1:59:47<04:52,  2.03it/s]
 94%|█████████▍| 10114/10708 [1:59:48<04:52,  2.03it/s]
 94%|█████████▍| 10115/10708 [1:59:48<04:51,  2.03it/s]
 94%|█████████▍| 10116/10708 [1:59:49<04:51,  2.03it/s]
 94%|█████████▍| 10117/10708 [1:59:49<04:50,  2.03it/s]
 94%|█████████▍| 10118/10708 [1:59:50<04:50,  2.03it/s]
 94%|█████████▍| 10119/10708 [1:59:50<04:49,  2.03it/s]
 95%|█████████▍| 10120/10708 [1:59:51<04:49,  2.03it/s]
 95%|█████████▍| 10121/10708 [1:59:51<04:49,  2.03it/s]
 95%|█████████▍| 10122/10708 [1:59:52<04:48,  2.03it/s]
 95%|█████████▍| 10123/10708 [1:59:52<04:47,  2.03it/s]
 95%|█████████▍| 10124/10708 [1:59:53<04:47,  2.03it/s]
 95%|█████████▍| 10125/10708 [1:59:53<04:47,  2.03it/s]{'loss': 3.5197, 'grad_norm': 0.19443587958812714, 'learning_rate': 9.00296391142924e-06, 'epoch': 0.95}

                                                       

 95%|█████████▍| 10125/10708 [1:59:53<04:47,  2.03it/s]
 95%|█████████▍| 10126/10708 [1:59:54<04:46,  2.03it/s]
 95%|█████████▍| 10127/10708 [1:59:54<04:45,  2.03it/s]
 95%|█████████▍| 10128/10708 [1:59:55<04:45,  2.03it/s]
 95%|█████████▍| 10129/10708 [1:59:55<04:44,  2.03it/s]
 95%|█████████▍| 10130/10708 [1:59:56<04:43,  2.04it/s]
 95%|█████████▍| 10131/10708 [1:59:56<04:43,  2.04it/s]
 95%|█████████▍| 10132/10708 [1:59:57<04:42,  2.04it/s]
 95%|█████████▍| 10133/10708 [1:59:57<04:42,  2.04it/s]
 95%|█████████▍| 10134/10708 [1:59:58<04:42,  2.03it/s]
 95%|█████████▍| 10135/10708 [1:59:58<04:41,  2.03it/s]
 95%|█████████▍| 10136/10708 [1:59:59<04:41,  2.03it/s]
 95%|█████████▍| 10137/10708 [1:59:59<04:40,  2.03it/s]
 95%|█████████▍| 10138/10708 [2:00:00<04:40,  2.03it/s]
 95%|█████████▍| 10139/10708 [2:00:00<04:39,  2.04it/s]
 95%|█████████▍| 10140/10708 [2:00:01<04:39,  2.03it/s]
 95%|█████████▍| 10141/10708 [2:00:01<04:38,  2.04it/s]
 95%|█████████▍| 10142/10708 [2:00:02<04:38,  2.04it/s]
 95%|█████████▍| 10143/10708 [2:00:02<04:38,  2.03it/s]
 95%|█████████▍| 10144/10708 [2:00:03<04:37,  2.03it/s]
 95%|█████████▍| 10145/10708 [2:00:03<04:37,  2.03it/s]
 95%|█████████▍| 10146/10708 [2:00:04<04:36,  2.03it/s]
 95%|█████████▍| 10147/10708 [2:00:04<04:36,  2.03it/s]
 95%|█████████▍| 10148/10708 [2:00:04<04:35,  2.03it/s]
 95%|█████████▍| 10149/10708 [2:00:05<04:34,  2.03it/s]
 95%|█████████▍| 10150/10708 [2:00:05<04:34,  2.03it/s]
                                                       
{'loss': 3.5145, 'grad_norm': 0.19253169000148773, 'learning_rate': 8.249479935563964e-06, 'epoch': 0.95}

 95%|█████████▍| 10150/10708 [2:00:05<04:34,  2.03it/s]
 95%|█████████▍| 10151/10708 [2:00:06<04:34,  2.03it/s]
 95%|█████████▍| 10152/10708 [2:00:06<04:33,  2.03it/s]
 95%|█████████▍| 10153/10708 [2:00:07<04:33,  2.03it/s]
 95%|█████████▍| 10154/10708 [2:00:07<04:32,  2.03it/s]
 95%|█████████▍| 10155/10708 [2:00:08<04:31,  2.03it/s]
 95%|█████████▍| 10156/10708 [2:00:08<04:31,  2.03it/s]
 95%|█████████▍| 10157/10708 [2:00:09<04:31,  2.03it/s]
 95%|█████████▍| 10158/10708 [2:00:09<04:30,  2.04it/s]
 95%|█████████▍| 10159/10708 [2:00:10<04:29,  2.04it/s]
 95%|█████████▍| 10160/10708 [2:00:10<04:29,  2.04it/s]
 95%|█████████▍| 10161/10708 [2:00:11<04:28,  2.04it/s]
 95%|█████████▍| 10162/10708 [2:00:11<04:28,  2.04it/s]
 95%|█████████▍| 10163/10708 [2:00:12<04:28,  2.03it/s]
 95%|█████████▍| 10164/10708 [2:00:12<04:27,  2.03it/s]
 95%|█████████▍| 10165/10708 [2:00:13<04:27,  2.03it/s]
 95%|█████████▍| 10166/10708 [2:00:13<04:27,  2.02it/s]
 95%|█████████▍| 10167/10708 [2:00:14<04:26,  2.03it/s]
 95%|█████████▍| 10168/10708 [2:00:14<04:25,  2.03it/s]
 95%|█████████▍| 10169/10708 [2:00:15<04:25,  2.03it/s]
 95%|█████████▍| 10170/10708 [2:00:15<04:25,  2.03it/s]
 95%|█████████▍| 10171/10708 [2:00:16<04:24,  2.03it/s]
 95%|█████████▍| 10172/10708 [2:00:16<04:23,  2.03it/s]
 95%|█████████▌| 10173/10708 [2:00:17<04:23,  2.03it/s]
 95%|█████████▌| 10174/10708 [2:00:17<04:22,  2.04it/s]
 95%|█████████▌| 10175/10708 [2:00:18<04:22,  2.03it/s]
                                                       
{'loss': 3.5176, 'grad_norm': 0.2000400274991989, 'learning_rate': 7.528657635887282e-06, 'epoch': 0.95}

 95%|█████████▌| 10175/10708 [2:00:18<04:22,  2.03it/s]
 95%|█████████▌| 10176/10708 [2:00:18<04:22,  2.03it/s]
 95%|█████████▌| 10177/10708 [2:00:19<04:21,  2.03it/s]
 95%|█████████▌| 10178/10708 [2:00:19<04:21,  2.03it/s]
 95%|█████████▌| 10179/10708 [2:00:20<04:20,  2.03it/s]
 95%|█████████▌| 10180/10708 [2:00:20<04:20,  2.03it/s]
 95%|█████████▌| 10181/10708 [2:00:21<04:19,  2.03it/s]
 95%|█████████▌| 10182/10708 [2:00:21<04:18,  2.03it/s]
 95%|█████████▌| 10183/10708 [2:00:22<04:18,  2.03it/s]
 95%|█████████▌| 10184/10708 [2:00:22<04:17,  2.03it/s]
 95%|█████████▌| 10185/10708 [2:00:23<04:17,  2.03it/s]
 95%|█████████▌| 10186/10708 [2:00:23<04:17,  2.03it/s]
 95%|█████████▌| 10187/10708 [2:00:24<04:16,  2.03it/s]
 95%|█████████▌| 10188/10708 [2:00:24<04:15,  2.03it/s]
 95%|█████████▌| 10189/10708 [2:00:25<04:15,  2.03it/s]
 95%|█████████▌| 10190/10708 [2:00:25<04:14,  2.03it/s]
 95%|█████████▌| 10191/10708 [2:00:26<04:14,  2.03it/s]
 95%|█████████▌| 10192/10708 [2:00:26<04:13,  2.03it/s]
 95%|█████████▌| 10193/10708 [2:00:27<04:14,  2.03it/s]
 95%|█████████▌| 10194/10708 [2:00:27<04:13,  2.03it/s]
 95%|█████████▌| 10195/10708 [2:00:28<04:13,  2.03it/s]
 95%|█████████▌| 10196/10708 [2:00:28<04:12,  2.03it/s]
 95%|█████████▌| 10197/10708 [2:00:29<04:11,  2.03it/s]
 95%|█████████▌| 10198/10708 [2:00:29<04:10,  2.03it/s]
 95%|█████████▌| 10199/10708 [2:00:30<04:10,  2.03it/s]
 95%|█████████▌| 10200/10708 [2:00:30<04:10,  2.03it/s]
                                                       
{'loss': 3.5315, 'grad_norm': 0.19174309074878693, 'learning_rate': 6.840544888839783e-06, 'epoch': 0.95}

 95%|█████████▌| 10200/10708 [2:00:30<04:10,  2.03it/s]
 95%|█████████▌| 10201/10708 [2:00:31<04:09,  2.03it/s]
 95%|█████████▌| 10202/10708 [2:00:31<04:09,  2.03it/s]
 95%|█████████▌| 10203/10708 [2:00:32<04:08,  2.03it/s]
 95%|█████████▌| 10204/10708 [2:00:32<04:07,  2.03it/s]
 95%|█████████▌| 10205/10708 [2:00:33<04:08,  2.03it/s]
 95%|█████████▌| 10206/10708 [2:00:33<04:07,  2.03it/s]
 95%|█████████▌| 10207/10708 [2:00:34<04:06,  2.03it/s]
 95%|█████████▌| 10208/10708 [2:00:34<04:06,  2.03it/s]
 95%|█████████▌| 10209/10708 [2:00:35<04:05,  2.03it/s]
 95%|█████████▌| 10210/10708 [2:00:35<04:05,  2.03it/s]
 95%|█████████▌| 10211/10708 [2:00:36<04:04,  2.03it/s]
 95%|█████████▌| 10212/10708 [2:00:36<04:03,  2.03it/s]
 95%|█████████▌| 10213/10708 [2:00:36<04:03,  2.03it/s]
 95%|█████████▌| 10214/10708 [2:00:37<04:02,  2.03it/s]
 95%|█████████▌| 10215/10708 [2:00:37<04:02,  2.03it/s]
 95%|█████████▌| 10216/10708 [2:00:38<04:02,  2.03it/s]
 95%|█████████▌| 10217/10708 [2:00:38<04:01,  2.03it/s]
 95%|█████████▌| 10218/10708 [2:00:39<04:01,  2.03it/s]
 95%|█████████▌| 10219/10708 [2:00:39<04:00,  2.03it/s]
 95%|█████████▌| 10220/10708 [2:00:40<04:00,  2.03it/s]
 95%|█████████▌| 10221/10708 [2:00:40<03:59,  2.03it/s]
 95%|█████████▌| 10222/10708 [2:00:41<03:59,  2.03it/s]
 95%|█████████▌| 10223/10708 [2:00:41<03:58,  2.03it/s]
 95%|█████████▌| 10224/10708 [2:00:42<03:58,  2.03it/s]
 95%|█████████▌| 10225/10708 [2:00:42<03:57,  2.03it/s]
                                                       
{'loss': 3.521, 'grad_norm': 0.1979113072156906, 'learning_rate': 6.185187398319691e-06, 'epoch': 0.95}

 95%|█████████▌| 10225/10708 [2:00:42<03:57,  2.03it/s]
 95%|█████████▌| 10226/10708 [2:00:43<03:58,  2.02it/s]
 96%|█████████▌| 10227/10708 [2:00:43<03:57,  2.03it/s]
 96%|█████████▌| 10228/10708 [2:00:44<03:56,  2.03it/s]
 96%|█████████▌| 10229/10708 [2:00:44<03:56,  2.03it/s]
 96%|█████████▌| 10230/10708 [2:00:45<03:55,  2.03it/s]
 96%|█████████▌| 10231/10708 [2:00:45<03:54,  2.03it/s]
 96%|█████████▌| 10232/10708 [2:00:46<03:54,  2.03it/s]
 96%|█████████▌| 10233/10708 [2:00:46<03:54,  2.03it/s]
 96%|█████████▌| 10234/10708 [2:00:47<03:53,  2.03it/s]
 96%|█████████▌| 10235/10708 [2:00:47<03:53,  2.03it/s]
 96%|█████████▌| 10236/10708 [2:00:48<03:52,  2.03it/s]
 96%|█████████▌| 10237/10708 [2:00:48<03:51,  2.03it/s]
 96%|█████████▌| 10238/10708 [2:00:49<03:51,  2.03it/s]
 96%|█████████▌| 10239/10708 [2:00:49<03:50,  2.03it/s]
 96%|█████████▌| 10240/10708 [2:00:50<04:28,  1.74it/s]
 96%|█████████▌| 10241/10708 [2:00:51<04:16,  1.82it/s]
 96%|█████████▌| 10242/10708 [2:00:51<04:07,  1.88it/s]
 96%|█████████▌| 10243/10708 [2:00:52<04:01,  1.92it/s]
 96%|█████████▌| 10244/10708 [2:00:52<03:57,  1.95it/s]
 96%|█████████▌| 10245/10708 [2:00:53<03:53,  1.98it/s]
 96%|█████████▌| 10246/10708 [2:00:53<03:51,  1.99it/s]
 96%|█████████▌| 10247/10708 [2:00:54<03:49,  2.01it/s]
 96%|█████████▌| 10248/10708 [2:00:54<03:48,  2.02it/s]
 96%|█████████▌| 10249/10708 [2:00:54<03:47,  2.02it/s]
 96%|█████████▌| 10250/10708 [2:00:55<03:46,  2.02it/s]
                                                       
{'loss': 3.5247, 'grad_norm': 0.19292253255844116, 'learning_rate': 5.562628692647287e-06, 'epoch': 0.96}

 96%|█████████▌| 10250/10708 [2:00:55<03:46,  2.02it/s]
 96%|█████████▌| 10251/10708 [2:00:55<03:45,  2.02it/s]
 96%|█████████▌| 10252/10708 [2:00:56<03:45,  2.03it/s]
 96%|█████████▌| 10253/10708 [2:00:56<03:44,  2.03it/s]
 96%|█████████▌| 10254/10708 [2:00:57<03:43,  2.03it/s]
 96%|█████████▌| 10255/10708 [2:00:57<03:43,  2.03it/s]
 96%|█████████▌| 10256/10708 [2:00:58<03:42,  2.03it/s]
 96%|█████████▌| 10257/10708 [2:00:58<03:41,  2.03it/s]
 96%|█████████▌| 10258/10708 [2:00:59<03:41,  2.03it/s]
 96%|█████████▌| 10259/10708 [2:00:59<03:41,  2.03it/s]
 96%|█████████▌| 10260/10708 [2:01:00<03:40,  2.03it/s]
 96%|█████████▌| 10261/10708 [2:01:00<03:39,  2.03it/s]
 96%|█████████▌| 10262/10708 [2:01:01<03:39,  2.03it/s]
 96%|█████████▌| 10263/10708 [2:01:01<03:39,  2.03it/s]
 96%|█████████▌| 10264/10708 [2:01:02<03:38,  2.03it/s]
 96%|█████████▌| 10265/10708 [2:01:02<03:38,  2.03it/s]
 96%|█████████▌| 10266/10708 [2:01:03<03:37,  2.03it/s]
 96%|█████████▌| 10267/10708 [2:01:03<03:37,  2.03it/s]
 96%|█████████▌| 10268/10708 [2:01:04<03:36,  2.03it/s]
 96%|█████████▌| 10269/10708 [2:01:04<03:35,  2.03it/s]
 96%|█████████▌| 10270/10708 [2:01:05<03:35,  2.03it/s]
 96%|█████████▌| 10271/10708 [2:01:05<03:35,  2.03it/s]
 96%|█████████▌| 10272/10708 [2:01:06<03:34,  2.03it/s]
 96%|█████████▌| 10273/10708 [2:01:06<03:34,  2.03it/s]
 96%|█████████▌| 10274/10708 [2:01:07<03:33,  2.03it/s]
 96%|█████████▌| 10275/10708 [2:01:07<03:33,  2.03it/s]
                                                       
{'loss': 3.5221, 'grad_norm': 0.19652847945690155, 'learning_rate': 4.972910121673779e-06, 'epoch': 0.96}

 96%|█████████▌| 10275/10708 [2:01:07<03:33,  2.03it/s]
 96%|█████████▌| 10276/10708 [2:01:08<03:32,  2.03it/s]
 96%|█████████▌| 10277/10708 [2:01:08<03:31,  2.03it/s]
 96%|█████████▌| 10278/10708 [2:01:09<03:31,  2.03it/s]
 96%|█████████▌| 10279/10708 [2:01:09<03:31,  2.03it/s]
 96%|█████████▌| 10280/10708 [2:01:10<03:30,  2.03it/s]
 96%|█████████▌| 10281/10708 [2:01:10<03:30,  2.03it/s]
 96%|█████████▌| 10282/10708 [2:01:11<03:29,  2.03it/s]
 96%|█████████▌| 10283/10708 [2:01:11<04:03,  1.75it/s]
 96%|█████████▌| 10284/10708 [2:01:12<03:52,  1.82it/s]
 96%|█████████▌| 10285/10708 [2:01:12<03:44,  1.88it/s]
 96%|█████████▌| 10286/10708 [2:01:13<03:39,  1.92it/s]
 96%|█████████▌| 10287/10708 [2:01:13<03:35,  1.95it/s]
 96%|█████████▌| 10288/10708 [2:01:14<03:32,  1.98it/s]
 96%|█████████▌| 10289/10708 [2:01:14<03:30,  1.99it/s]
 96%|█████████▌| 10290/10708 [2:01:15<03:28,  2.00it/s]
 96%|█████████▌| 10291/10708 [2:01:15<03:27,  2.01it/s]
 96%|█████████▌| 10292/10708 [2:01:16<03:26,  2.02it/s]
 96%|█████████▌| 10293/10708 [2:01:16<03:24,  2.03it/s]
 96%|█████████▌| 10294/10708 [2:01:17<03:24,  2.03it/s]
 96%|█████████▌| 10295/10708 [2:01:17<03:23,  2.03it/s]
 96%|█████████▌| 10296/10708 [2:01:18<03:22,  2.03it/s]
 96%|█████████▌| 10297/10708 [2:01:18<03:22,  2.03it/s]
 96%|█████████▌| 10298/10708 [2:01:19<03:21,  2.03it/s]
 96%|█████████▌| 10299/10708 [2:01:19<03:21,  2.03it/s]
 96%|█████████▌| 10300/10708 [2:01:20<03:20,  2.03it/s]
                                                       
{'loss': 3.5236, 'grad_norm': 0.19428735971450806, 'learning_rate': 4.416070854035004e-06, 'epoch': 0.96}

 96%|█████████▌| 10300/10708 [2:01:20<03:20,  2.03it/s]
 96%|█████████▌| 10301/10708 [2:01:20<03:20,  2.03it/s]
 96%|█████████▌| 10302/10708 [2:01:21<03:19,  2.03it/s]
 96%|█████████▌| 10303/10708 [2:01:21<03:19,  2.03it/s]
 96%|█████████▌| 10304/10708 [2:01:22<03:18,  2.03it/s]
 96%|█████████▌| 10305/10708 [2:01:22<03:18,  2.03it/s]
 96%|█████████▌| 10306/10708 [2:01:23<03:17,  2.03it/s]
 96%|█████████▋| 10307/10708 [2:01:23<03:17,  2.04it/s]
 96%|█████████▋| 10308/10708 [2:01:24<03:16,  2.03it/s]
 96%|█████████▋| 10309/10708 [2:01:24<03:16,  2.03it/s]
 96%|█████████▋| 10310/10708 [2:01:25<03:15,  2.04it/s]
 96%|█████████▋| 10311/10708 [2:01:25<03:15,  2.03it/s]
 96%|█████████▋| 10312/10708 [2:01:26<03:15,  2.03it/s]
 96%|█████████▋| 10313/10708 [2:01:26<03:14,  2.03it/s]
 96%|█████████▋| 10314/10708 [2:01:27<03:13,  2.03it/s]
 96%|█████████▋| 10315/10708 [2:01:27<03:13,  2.03it/s]
 96%|█████████▋| 10316/10708 [2:01:28<03:12,  2.03it/s]
 96%|█████████▋| 10317/10708 [2:01:28<03:12,  2.03it/s]
 96%|█████████▋| 10318/10708 [2:01:29<03:11,  2.03it/s]
 96%|█████████▋| 10319/10708 [2:01:29<03:10,  2.04it/s]
 96%|█████████▋| 10320/10708 [2:01:30<03:10,  2.04it/s]
 96%|█████████▋| 10321/10708 [2:01:30<03:10,  2.03it/s]
 96%|█████████▋| 10322/10708 [2:01:31<03:09,  2.03it/s]
 96%|█████████▋| 10323/10708 [2:01:31<03:09,  2.03it/s]
 96%|█████████▋| 10324/10708 [2:01:32<03:08,  2.04it/s]
 96%|█████████▋| 10325/10708 [2:01:32<03:08,  2.03it/s]{'loss': 3.5173, 'grad_norm': 0.19080506265163422, 'learning_rate': 3.892147874549612e-06, 'epoch': 0.96}
                                                       

 96%|█████████▋| 10325/10708 [2:01:32<03:08,  2.03it/s]
 96%|█████████▋| 10326/10708 [2:01:33<03:08,  2.03it/s]
 96%|█████████▋| 10327/10708 [2:01:33<03:07,  2.03it/s]
 96%|█████████▋| 10328/10708 [2:01:34<03:06,  2.03it/s]
 96%|█████████▋| 10329/10708 [2:01:34<03:06,  2.03it/s]
 96%|█████████▋| 10330/10708 [2:01:35<03:05,  2.03it/s]
 96%|█████████▋| 10331/10708 [2:01:35<03:05,  2.03it/s]
 96%|█████████▋| 10332/10708 [2:01:36<03:04,  2.03it/s]
 96%|█████████▋| 10333/10708 [2:01:36<03:04,  2.03it/s]
 97%|█████████▋| 10334/10708 [2:01:37<03:03,  2.03it/s]
 97%|█████████▋| 10335/10708 [2:01:37<03:03,  2.03it/s]
 97%|█████████▋| 10336/10708 [2:01:38<03:02,  2.03it/s]
 97%|█████████▋| 10337/10708 [2:01:38<03:02,  2.03it/s]
 97%|█████████▋| 10338/10708 [2:01:39<03:01,  2.04it/s]
 97%|█████████▋| 10339/10708 [2:01:39<03:01,  2.04it/s]
 97%|█████████▋| 10340/10708 [2:01:40<03:00,  2.04it/s]
 97%|█████████▋| 10341/10708 [2:01:40<03:00,  2.04it/s]
 97%|█████████▋| 10342/10708 [2:01:41<02:59,  2.04it/s]
 97%|█████████▋| 10343/10708 [2:01:41<02:59,  2.04it/s]
 97%|█████████▋| 10344/10708 [2:01:41<02:58,  2.04it/s]
 97%|█████████▋| 10345/10708 [2:01:42<02:58,  2.03it/s]
 97%|█████████▋| 10346/10708 [2:01:42<02:58,  2.03it/s]
 97%|█████████▋| 10347/10708 [2:01:43<02:58,  2.03it/s]
 97%|█████████▋| 10348/10708 [2:01:43<02:57,  2.03it/s]
 97%|█████████▋| 10349/10708 [2:01:44<02:56,  2.03it/s]
 97%|█████████▋| 10350/10708 [2:01:44<02:56,  2.03it/s]
                                                       
{'loss': 3.5172, 'grad_norm': 0.19315126538276672, 'learning_rate': 3.401175981762872e-06, 'epoch': 0.97}

 97%|█████████▋| 10350/10708 [2:01:44<02:56,  2.03it/s]
 97%|█████████▋| 10351/10708 [2:01:45<02:55,  2.03it/s]
 97%|█████████▋| 10352/10708 [2:01:45<02:55,  2.03it/s]
 97%|█████████▋| 10353/10708 [2:01:46<02:54,  2.03it/s]
 97%|█████████▋| 10354/10708 [2:01:46<02:54,  2.03it/s]
 97%|█████████▋| 10355/10708 [2:01:47<02:53,  2.03it/s]
 97%|█████████▋| 10356/10708 [2:01:47<02:53,  2.03it/s]
 97%|█████████▋| 10357/10708 [2:01:48<02:52,  2.03it/s]
 97%|█████████▋| 10358/10708 [2:01:48<02:52,  2.03it/s]
 97%|█████████▋| 10359/10708 [2:01:49<02:51,  2.03it/s]
 97%|█████████▋| 10360/10708 [2:01:49<02:51,  2.03it/s]
 97%|█████████▋| 10361/10708 [2:01:50<02:50,  2.03it/s]
 97%|█████████▋| 10362/10708 [2:01:50<02:50,  2.04it/s]
 97%|█████████▋| 10363/10708 [2:01:51<02:49,  2.04it/s]
 97%|█████████▋| 10364/10708 [2:01:51<02:49,  2.03it/s]
 97%|█████████▋| 10365/10708 [2:01:52<02:48,  2.04it/s]
 97%|█████████▋| 10366/10708 [2:01:52<02:48,  2.03it/s]
 97%|█████████▋| 10367/10708 [2:01:53<02:47,  2.03it/s]
 97%|█████████▋| 10368/10708 [2:01:53<02:47,  2.03it/s]
 97%|█████████▋| 10369/10708 [2:01:54<02:46,  2.04it/s]
 97%|█████████▋| 10370/10708 [2:01:54<02:45,  2.04it/s]
 97%|█████████▋| 10371/10708 [2:01:55<02:45,  2.04it/s]
 97%|█████████▋| 10372/10708 [2:01:55<02:45,  2.03it/s]
 97%|█████████▋| 10373/10708 [2:01:56<02:44,  2.04it/s]
 97%|█████████▋| 10374/10708 [2:01:56<02:44,  2.03it/s]
 97%|█████████▋| 10375/10708 [2:01:57<02:44,  2.03it/s]
                                                       
{'loss': 3.5196, 'grad_norm': 0.196741983294487, 'learning_rate': 2.943187785635182e-06, 'epoch': 0.97}

 97%|█████████▋| 10375/10708 [2:01:57<02:44,  2.03it/s]
 97%|█████████▋| 10376/10708 [2:01:57<02:43,  2.03it/s]
 97%|█████████▋| 10377/10708 [2:01:58<02:43,  2.03it/s]
 97%|█████████▋| 10378/10708 [2:01:58<02:42,  2.03it/s]
 97%|█████████▋| 10379/10708 [2:01:59<02:42,  2.03it/s]
 97%|█████████▋| 10380/10708 [2:01:59<02:41,  2.03it/s]
 97%|█████████▋| 10381/10708 [2:02:00<02:41,  2.03it/s]
 97%|█████████▋| 10382/10708 [2:02:00<02:40,  2.03it/s]
 97%|█████████▋| 10383/10708 [2:02:01<02:39,  2.03it/s]
 97%|█████████▋| 10384/10708 [2:02:01<02:39,  2.03it/s]
 97%|█████████▋| 10385/10708 [2:02:02<02:38,  2.03it/s]
 97%|█████████▋| 10386/10708 [2:02:02<02:38,  2.03it/s]
 97%|█████████▋| 10387/10708 [2:02:03<02:37,  2.03it/s]
 97%|█████████▋| 10388/10708 [2:02:03<02:37,  2.03it/s]
 97%|█████████▋| 10389/10708 [2:02:04<02:36,  2.03it/s]
 97%|█████████▋| 10390/10708 [2:02:04<02:36,  2.03it/s]
 97%|█████████▋| 10391/10708 [2:02:05<02:35,  2.03it/s]
 97%|█████████▋| 10392/10708 [2:02:05<02:35,  2.03it/s]
 97%|█████████▋| 10393/10708 [2:02:06<02:34,  2.03it/s]
 97%|█████████▋| 10394/10708 [2:02:06<02:34,  2.03it/s]
 97%|█████████▋| 10395/10708 [2:02:07<02:33,  2.03it/s]
 97%|█████████▋| 10396/10708 [2:02:07<02:33,  2.03it/s]
 97%|█████████▋| 10397/10708 [2:02:08<02:33,  2.03it/s]
 97%|█████████▋| 10398/10708 [2:02:08<02:32,  2.03it/s]
 97%|█████████▋| 10399/10708 [2:02:09<02:31,  2.03it/s]
 97%|█████████▋| 10400/10708 [2:02:09<02:31,  2.03it/s]
                                                       {'loss': 3.5186, 'grad_norm': 0.19684532284736633, 'learning_rate': 2.518213705376082e-06, 'epoch': 0.97}


 97%|█████████▋| 10400/10708 [2:02:09<02:31,  2.03it/s]
 97%|█████████▋| 10401/10708 [2:02:10<02:31,  2.03it/s]
 97%|█████████▋| 10402/10708 [2:02:10<02:30,  2.03it/s]
 97%|█████████▋| 10403/10708 [2:02:11<02:30,  2.03it/s]
 97%|█████████▋| 10404/10708 [2:02:11<02:29,  2.03it/s]
 97%|█████████▋| 10405/10708 [2:02:11<02:28,  2.04it/s]
 97%|█████████▋| 10406/10708 [2:02:12<02:28,  2.03it/s]
 97%|█████████▋| 10407/10708 [2:02:12<02:27,  2.03it/s]
 97%|█████████▋| 10408/10708 [2:02:13<02:27,  2.03it/s]
 97%|█████████▋| 10409/10708 [2:02:13<02:27,  2.03it/s]
 97%|█████████▋| 10410/10708 [2:02:14<02:26,  2.03it/s]
 97%|█████████▋| 10411/10708 [2:02:14<02:26,  2.03it/s]
 97%|█████████▋| 10412/10708 [2:02:15<02:25,  2.03it/s]
 97%|█████████▋| 10413/10708 [2:02:15<02:25,  2.03it/s]
 97%|█████████▋| 10414/10708 [2:02:16<02:24,  2.03it/s]
 97%|█████████▋| 10415/10708 [2:02:16<02:24,  2.03it/s]
 97%|█████████▋| 10416/10708 [2:02:17<02:23,  2.03it/s]
 97%|█████████▋| 10417/10708 [2:02:17<02:23,  2.03it/s]
 97%|█████████▋| 10418/10708 [2:02:18<02:23,  2.03it/s]
 97%|█████████▋| 10419/10708 [2:02:18<02:22,  2.03it/s]
 97%|█████████▋| 10420/10708 [2:02:19<02:21,  2.03it/s]
 97%|█████████▋| 10421/10708 [2:02:19<02:21,  2.03it/s]
 97%|█████████▋| 10422/10708 [2:02:20<02:20,  2.03it/s]
 97%|█████████▋| 10423/10708 [2:02:20<02:20,  2.03it/s]
 97%|█████████▋| 10424/10708 [2:02:21<02:19,  2.03it/s]
 97%|█████████▋| 10425/10708 [2:02:21<02:19,  2.03it/s]{'loss': 3.5129, 'grad_norm': 0.19202320277690887, 'learning_rate': 2.126281967424204e-06, 'epoch': 0.97}

                                                       

 97%|█████████▋| 10425/10708 [2:02:21<02:19,  2.03it/s]
 97%|█████████▋| 10426/10708 [2:02:22<02:19,  2.03it/s]
 97%|█████████▋| 10427/10708 [2:02:22<02:18,  2.03it/s]
 97%|█████████▋| 10428/10708 [2:02:23<02:18,  2.03it/s]
 97%|█████████▋| 10429/10708 [2:02:23<02:17,  2.03it/s]
 97%|█████████▋| 10430/10708 [2:02:24<02:16,  2.03it/s]
 97%|█████████▋| 10431/10708 [2:02:24<02:16,  2.03it/s]
 97%|█████████▋| 10432/10708 [2:02:25<02:15,  2.03it/s]
 97%|█████████▋| 10433/10708 [2:02:25<02:15,  2.03it/s]
 97%|█████████▋| 10434/10708 [2:02:26<02:14,  2.03it/s]
 97%|█████████▋| 10435/10708 [2:02:26<02:14,  2.03it/s]
 97%|█████████▋| 10436/10708 [2:02:27<02:13,  2.03it/s]
 97%|█████████▋| 10437/10708 [2:02:27<02:13,  2.03it/s]
 97%|█████████▋| 10438/10708 [2:02:28<02:13,  2.03it/s]
 97%|█████████▋| 10439/10708 [2:02:28<02:12,  2.03it/s]
 97%|█████████▋| 10440/10708 [2:02:29<02:11,  2.04it/s]
 98%|█████████▊| 10441/10708 [2:02:29<02:11,  2.04it/s]
 98%|█████████▊| 10442/10708 [2:02:30<02:10,  2.03it/s]
 98%|█████████▊| 10443/10708 [2:02:30<02:10,  2.04it/s]
 98%|█████████▊| 10444/10708 [2:02:31<02:09,  2.04it/s]
 98%|█████████▊| 10445/10708 [2:02:31<02:09,  2.03it/s]
 98%|█████████▊| 10446/10708 [2:02:32<02:08,  2.04it/s]
 98%|█████████▊| 10447/10708 [2:02:32<02:08,  2.04it/s]
 98%|█████████▊| 10448/10708 [2:02:33<02:07,  2.03it/s]
 98%|█████████▊| 10449/10708 [2:02:33<02:07,  2.04it/s]
 98%|█████████▊| 10450/10708 [2:02:34<02:06,  2.04it/s]
                                                       
{'loss': 3.5157, 'grad_norm': 0.19161450862884521, 'learning_rate': 1.7674186035719353e-06, 'epoch': 0.98}

 98%|█████████▊| 10450/10708 [2:02:34<02:06,  2.04it/s]
 98%|█████████▊| 10451/10708 [2:02:34<02:06,  2.03it/s]
 98%|█████████▊| 10452/10708 [2:02:35<02:05,  2.03it/s]
 98%|█████████▊| 10453/10708 [2:02:35<02:05,  2.04it/s]
 98%|█████████▊| 10454/10708 [2:02:36<02:04,  2.03it/s]
 98%|█████████▊| 10455/10708 [2:02:36<02:04,  2.03it/s]
 98%|█████████▊| 10456/10708 [2:02:37<02:03,  2.04it/s]
 98%|█████████▊| 10457/10708 [2:02:37<02:03,  2.03it/s]
 98%|█████████▊| 10458/10708 [2:02:38<02:02,  2.03it/s]
 98%|█████████▊| 10459/10708 [2:02:38<02:02,  2.04it/s]
 98%|█████████▊| 10460/10708 [2:02:39<02:01,  2.04it/s]
 98%|█████████▊| 10461/10708 [2:02:39<02:01,  2.04it/s]
 98%|█████████▊| 10462/10708 [2:02:40<02:00,  2.04it/s]
 98%|█████████▊| 10463/10708 [2:02:40<02:00,  2.03it/s]
 98%|█████████▊| 10464/10708 [2:02:41<01:59,  2.04it/s]
 98%|█████████▊| 10465/10708 [2:02:41<01:59,  2.04it/s]
 98%|█████████▊| 10466/10708 [2:02:42<01:59,  2.03it/s]
 98%|█████████▊| 10467/10708 [2:02:42<01:58,  2.03it/s]
 98%|█████████▊| 10468/10708 [2:02:42<01:58,  2.03it/s]
 98%|█████████▊| 10469/10708 [2:02:43<01:57,  2.03it/s]
 98%|█████████▊| 10470/10708 [2:02:43<01:57,  2.03it/s]
 98%|█████████▊| 10471/10708 [2:02:44<01:56,  2.03it/s]
 98%|█████████▊| 10472/10708 [2:02:44<01:56,  2.03it/s]
 98%|█████████▊| 10473/10708 [2:02:45<01:55,  2.03it/s]
 98%|█████████▊| 10474/10708 [2:02:45<01:55,  2.03it/s]
 98%|█████████▊| 10475/10708 [2:02:46<01:54,  2.03it/s]
                                                       {'loss': 3.525, 'grad_norm': 0.19095587730407715, 'learning_rate': 1.4416474492369714e-06, 'epoch': 0.98}


 98%|█████████▊| 10475/10708 [2:02:46<01:54,  2.03it/s]
 98%|█████████▊| 10476/10708 [2:02:46<01:54,  2.03it/s]
 98%|█████████▊| 10477/10708 [2:02:47<01:53,  2.03it/s]
 98%|█████████▊| 10478/10708 [2:02:47<01:53,  2.03it/s]
 98%|█████████▊| 10479/10708 [2:02:48<01:52,  2.03it/s]
 98%|█████████▊| 10480/10708 [2:02:48<01:52,  2.03it/s]
 98%|█████████▊| 10481/10708 [2:02:49<01:51,  2.03it/s]
 98%|█████████▊| 10482/10708 [2:02:49<01:51,  2.03it/s]
 98%|█████████▊| 10483/10708 [2:02:50<01:50,  2.04it/s]
 98%|█████████▊| 10484/10708 [2:02:50<01:50,  2.03it/s]
 98%|█████████▊| 10485/10708 [2:02:51<01:49,  2.03it/s]
 98%|█████████▊| 10486/10708 [2:02:51<01:49,  2.03it/s]
 98%|█████████▊| 10487/10708 [2:02:52<01:48,  2.03it/s]
 98%|█████████▊| 10488/10708 [2:02:52<01:48,  2.03it/s]
 98%|█████████▊| 10489/10708 [2:02:53<01:47,  2.03it/s]
 98%|█████████▊| 10490/10708 [2:02:53<01:47,  2.03it/s]
 98%|█████████▊| 10491/10708 [2:02:54<01:46,  2.03it/s]
 98%|█████████▊| 10492/10708 [2:02:54<01:46,  2.04it/s]
 98%|█████████▊| 10493/10708 [2:02:55<01:45,  2.03it/s]
 98%|█████████▊| 10494/10708 [2:02:55<01:45,  2.03it/s]
 98%|█████████▊| 10495/10708 [2:02:56<01:44,  2.04it/s]
 98%|█████████▊| 10496/10708 [2:02:56<01:44,  2.03it/s]
 98%|█████████▊| 10497/10708 [2:02:57<01:43,  2.03it/s]
 98%|█████████▊| 10498/10708 [2:02:57<01:43,  2.04it/s]
 98%|█████████▊| 10499/10708 [2:02:58<01:42,  2.03it/s]
 98%|█████████▊| 10500/10708 [2:02:58<01:42,  2.04it/s]
                                                       
{'loss': 3.5163, 'grad_norm': 0.19529984891414642, 'learning_rate': 1.1489901418786364e-06, 'epoch': 0.98}

 98%|█████████▊| 10500/10708 [2:02:58<01:42,  2.04it/s]
 98%|█████████▊| 10501/10708 [2:02:59<01:41,  2.03it/s]
 98%|█████████▊| 10502/10708 [2:02:59<01:41,  2.03it/s]
 98%|█████████▊| 10503/10708 [2:03:00<01:40,  2.03it/s]
 98%|█████████▊| 10504/10708 [2:03:00<01:40,  2.03it/s]
 98%|█████████▊| 10505/10708 [2:03:01<01:39,  2.03it/s]
 98%|█████████▊| 10506/10708 [2:03:01<01:39,  2.03it/s]
 98%|█████████▊| 10507/10708 [2:03:02<01:38,  2.03it/s]
 98%|█████████▊| 10508/10708 [2:03:02<01:38,  2.03it/s]
 98%|█████████▊| 10509/10708 [2:03:03<01:37,  2.03it/s]
 98%|█████████▊| 10510/10708 [2:03:03<01:37,  2.03it/s]
 98%|█████████▊| 10511/10708 [2:03:04<01:36,  2.03it/s]
 98%|█████████▊| 10512/10708 [2:03:04<01:36,  2.04it/s]
 98%|█████████▊| 10513/10708 [2:03:05<01:35,  2.04it/s]
 98%|█████████▊| 10514/10708 [2:03:05<01:35,  2.04it/s]
 98%|█████████▊| 10515/10708 [2:03:06<01:34,  2.04it/s]
 98%|█████████▊| 10516/10708 [2:03:06<01:34,  2.03it/s]
 98%|█████████▊| 10517/10708 [2:03:07<01:33,  2.04it/s]
 98%|█████████▊| 10518/10708 [2:03:07<01:33,  2.04it/s]
 98%|█████████▊| 10519/10708 [2:03:08<01:32,  2.03it/s]
 98%|█████████▊| 10520/10708 [2:03:08<01:32,  2.03it/s]
 98%|█████████▊| 10521/10708 [2:03:09<01:31,  2.04it/s]
 98%|█████████▊| 10522/10708 [2:03:09<01:31,  2.03it/s]
 98%|█████████▊| 10523/10708 [2:03:10<01:30,  2.03it/s]
 98%|█████████▊| 10524/10708 [2:03:10<01:30,  2.04it/s]
 98%|█████████▊| 10525/10708 [2:03:11<01:29,  2.03it/s]{'loss': 3.5222, 'grad_norm': 0.19408635795116425, 'learning_rate': 8.894661195613663e-07, 'epoch': 0.98}

                                                       

 98%|█████████▊| 10525/10708 [2:03:11<01:29,  2.03it/s]
 98%|█████████▊| 10526/10708 [2:03:11<01:29,  2.03it/s]
 98%|█████████▊| 10527/10708 [2:03:12<01:29,  2.03it/s]
 98%|█████████▊| 10528/10708 [2:03:12<01:28,  2.03it/s]
 98%|█████████▊| 10529/10708 [2:03:12<01:28,  2.03it/s]
 98%|█████████▊| 10530/10708 [2:03:13<01:27,  2.03it/s]
 98%|█████████▊| 10531/10708 [2:03:13<01:27,  2.03it/s]
 98%|█████████▊| 10532/10708 [2:03:14<01:26,  2.03it/s]
 98%|█████████▊| 10533/10708 [2:03:14<01:26,  2.03it/s]
 98%|█████████▊| 10534/10708 [2:03:15<01:25,  2.03it/s]
 98%|█████████▊| 10535/10708 [2:03:15<01:25,  2.03it/s]
 98%|█████████▊| 10536/10708 [2:03:16<01:24,  2.03it/s]
 98%|█████████▊| 10537/10708 [2:03:16<01:24,  2.03it/s]
 98%|█████████▊| 10538/10708 [2:03:17<01:23,  2.03it/s]
 98%|█████████▊| 10539/10708 [2:03:17<01:23,  2.03it/s]
 98%|█████████▊| 10540/10708 [2:03:18<01:22,  2.03it/s]
 98%|█████████▊| 10541/10708 [2:03:18<01:22,  2.03it/s]
 98%|█████████▊| 10542/10708 [2:03:19<01:21,  2.03it/s]
 98%|█████████▊| 10543/10708 [2:03:19<01:21,  2.03it/s]
 98%|█████████▊| 10544/10708 [2:03:20<01:20,  2.03it/s]
 98%|█████████▊| 10545/10708 [2:03:20<01:20,  2.03it/s]
 98%|█████████▊| 10546/10708 [2:03:21<01:19,  2.04it/s]
 98%|█████████▊| 10547/10708 [2:03:21<01:19,  2.04it/s]
 99%|█████████▊| 10548/10708 [2:03:22<01:18,  2.03it/s]
 99%|█████████▊| 10549/10708 [2:03:22<01:18,  2.04it/s]
 99%|█████████▊| 10550/10708 [2:03:23<01:17,  2.04it/s]{'loss': 3.5186, 'grad_norm': 0.19241052865982056, 'learning_rate': 6.630926196629639e-07, 'epoch': 0.99}

                                                       

 99%|█████████▊| 10550/10708 [2:03:23<01:17,  2.04it/s]
 99%|█████████▊| 10551/10708 [2:03:23<01:17,  2.03it/s]
 99%|█████████▊| 10552/10708 [2:03:24<01:16,  2.03it/s]
 99%|█████████▊| 10553/10708 [2:03:24<01:16,  2.03it/s]
 99%|█████████▊| 10554/10708 [2:03:25<01:15,  2.03it/s]
 99%|█████████▊| 10555/10708 [2:03:25<01:15,  2.03it/s]
 99%|█████████▊| 10556/10708 [2:03:26<01:14,  2.03it/s]
 99%|█████████▊| 10557/10708 [2:03:26<01:14,  2.03it/s]
 99%|█████████▊| 10558/10708 [2:03:27<01:13,  2.04it/s]
 99%|█████████▊| 10559/10708 [2:03:27<01:13,  2.03it/s]
 99%|█████████▊| 10560/10708 [2:03:28<01:12,  2.03it/s]
 99%|█████████▊| 10561/10708 [2:03:28<01:12,  2.03it/s]
 99%|█████████▊| 10562/10708 [2:03:29<01:11,  2.03it/s]
 99%|█████████▊| 10563/10708 [2:03:29<01:11,  2.03it/s]
 99%|█████████▊| 10564/10708 [2:03:30<01:10,  2.03it/s]
 99%|█████████▊| 10565/10708 [2:03:30<01:10,  2.03it/s]
 99%|█████████▊| 10566/10708 [2:03:31<01:09,  2.03it/s]
 99%|█████████▊| 10567/10708 [2:03:31<01:09,  2.03it/s]
 99%|█████████▊| 10568/10708 [2:03:32<01:08,  2.03it/s]
 99%|█████████▊| 10569/10708 [2:03:32<01:08,  2.03it/s]
 99%|█████████▊| 10570/10708 [2:03:33<01:08,  2.03it/s]
 99%|█████████▊| 10571/10708 [2:03:33<01:07,  2.03it/s]
 99%|█████████▊| 10572/10708 [2:03:34<01:06,  2.03it/s]
 99%|█████████▊| 10573/10708 [2:03:34<01:06,  2.03it/s]
 99%|█████████▊| 10574/10708 [2:03:35<01:05,  2.03it/s]
 99%|█████████▉| 10575/10708 [2:03:35<01:05,  2.03it/s]{'loss': 3.5184, 'grad_norm': 0.19117413461208344, 'learning_rate': 4.698846777303478e-07, 'epoch': 0.99}
                                                       

 99%|█████████▉| 10575/10708 [2:03:35<01:05,  2.03it/s]
 99%|█████████▉| 10576/10708 [2:03:36<01:04,  2.03it/s]
 99%|█████████▉| 10577/10708 [2:03:36<01:04,  2.03it/s]
 99%|█████████▉| 10578/10708 [2:03:37<01:03,  2.03it/s]
 99%|█████████▉| 10579/10708 [2:03:37<01:03,  2.04it/s]
 99%|█████████▉| 10580/10708 [2:03:38<01:02,  2.04it/s]
 99%|█████████▉| 10581/10708 [2:03:38<01:02,  2.04it/s]
 99%|█████████▉| 10582/10708 [2:03:39<01:01,  2.03it/s]
 99%|█████████▉| 10583/10708 [2:03:39<01:01,  2.04it/s]
 99%|█████████▉| 10584/10708 [2:03:40<01:00,  2.03it/s]
 99%|█████████▉| 10585/10708 [2:03:40<01:00,  2.03it/s]
 99%|█████████▉| 10586/10708 [2:03:41<00:59,  2.03it/s]
 99%|█████████▉| 10587/10708 [2:03:41<00:59,  2.03it/s]
 99%|█████████▉| 10588/10708 [2:03:42<00:59,  2.03it/s]
 99%|█████████▉| 10589/10708 [2:03:42<00:58,  2.04it/s]
 99%|█████████▉| 10590/10708 [2:03:42<00:58,  2.03it/s]
 99%|█████████▉| 10591/10708 [2:03:43<00:57,  2.03it/s]
 99%|█████████▉| 10592/10708 [2:03:43<00:57,  2.03it/s]
 99%|█████████▉| 10593/10708 [2:03:44<00:56,  2.03it/s]
 99%|█████████▉| 10594/10708 [2:03:44<00:56,  2.03it/s]
 99%|█████████▉| 10595/10708 [2:03:45<00:55,  2.03it/s]
 99%|█████████▉| 10596/10708 [2:03:45<00:55,  2.03it/s]
 99%|█████████▉| 10597/10708 [2:03:46<00:54,  2.03it/s]
 99%|█████████▉| 10598/10708 [2:03:46<00:54,  2.03it/s]
 99%|█████████▉| 10599/10708 [2:03:47<00:53,  2.03it/s]
 99%|█████████▉| 10600/10708 [2:03:47<00:53,  2.03it/s]
                                                       
{'loss': 3.5148, 'grad_norm': 0.19353532791137695, 'learning_rate': 3.0985512648035175e-07, 'epoch': 0.99}

 99%|█████████▉| 10600/10708 [2:03:47<00:53,  2.03it/s]
 99%|█████████▉| 10601/10708 [2:03:48<00:52,  2.03it/s]
 99%|█████████▉| 10602/10708 [2:03:48<00:52,  2.03it/s]
 99%|█████████▉| 10603/10708 [2:03:49<00:51,  2.03it/s]
 99%|█████████▉| 10604/10708 [2:03:49<00:51,  2.03it/s]
 99%|█████████▉| 10605/10708 [2:03:50<00:50,  2.03it/s]
 99%|█████████▉| 10606/10708 [2:03:50<00:50,  2.03it/s]
 99%|█████████▉| 10607/10708 [2:03:51<00:49,  2.03it/s]
 99%|█████████▉| 10608/10708 [2:03:51<00:49,  2.03it/s]
 99%|█████████▉| 10609/10708 [2:03:52<00:48,  2.03it/s]
 99%|█████████▉| 10610/10708 [2:03:52<00:48,  2.03it/s]
 99%|█████████▉| 10611/10708 [2:03:53<00:47,  2.03it/s]
 99%|█████████▉| 10612/10708 [2:03:53<00:47,  2.03it/s]
 99%|█████████▉| 10613/10708 [2:03:54<00:46,  2.03it/s]
 99%|█████████▉| 10614/10708 [2:03:54<00:46,  2.03it/s]
 99%|█████████▉| 10615/10708 [2:03:55<00:45,  2.03it/s]
 99%|█████████▉| 10616/10708 [2:03:55<00:45,  2.03it/s]
 99%|█████████▉| 10617/10708 [2:03:56<00:44,  2.03it/s]
 99%|█████████▉| 10618/10708 [2:03:56<00:44,  2.03it/s]
 99%|█████████▉| 10619/10708 [2:03:57<00:43,  2.04it/s]
 99%|█████████▉| 10620/10708 [2:03:57<00:43,  2.03it/s]
 99%|█████████▉| 10621/10708 [2:03:58<00:42,  2.03it/s]
 99%|█████████▉| 10622/10708 [2:03:58<00:42,  2.03it/s]
 99%|█████████▉| 10623/10708 [2:03:59<00:41,  2.03it/s]
 99%|█████████▉| 10624/10708 [2:03:59<00:41,  2.03it/s]
 99%|█████████▉| 10625/10708 [2:04:00<00:40,  2.03it/s]{'loss': 3.5179, 'grad_norm': 0.19335314631462097, 'learning_rate': 1.830145949478501e-07, 'epoch': 0.99}
                                                       

 99%|█████████▉| 10625/10708 [2:04:00<00:40,  2.03it/s]
 99%|█████████▉| 10626/10708 [2:04:00<00:40,  2.03it/s]
 99%|█████████▉| 10627/10708 [2:04:01<00:39,  2.03it/s]
 99%|█████████▉| 10628/10708 [2:04:01<00:39,  2.03it/s]
 99%|█████████▉| 10629/10708 [2:04:02<00:38,  2.03it/s]
 99%|█████████▉| 10630/10708 [2:04:02<00:38,  2.03it/s]
 99%|█████████▉| 10631/10708 [2:04:03<00:37,  2.03it/s]
 99%|█████████▉| 10632/10708 [2:04:03<00:37,  2.03it/s]
 99%|█████████▉| 10633/10708 [2:04:04<00:36,  2.03it/s]
 99%|█████████▉| 10634/10708 [2:04:04<00:36,  2.03it/s]
 99%|█████████▉| 10635/10708 [2:04:05<00:35,  2.03it/s]
 99%|█████████▉| 10636/10708 [2:04:05<00:35,  2.03it/s]
 99%|█████████▉| 10637/10708 [2:04:06<00:34,  2.03it/s]
 99%|█████████▉| 10638/10708 [2:04:06<00:34,  2.03it/s]
 99%|█████████▉| 10639/10708 [2:04:07<00:33,  2.03it/s]
 99%|█████████▉| 10640/10708 [2:04:07<00:33,  2.03it/s]
 99%|█████████▉| 10641/10708 [2:04:08<00:32,  2.03it/s]
 99%|█████████▉| 10642/10708 [2:04:08<00:32,  2.03it/s]
 99%|█████████▉| 10643/10708 [2:04:09<00:32,  2.03it/s]
 99%|█████████▉| 10644/10708 [2:04:09<00:31,  2.03it/s]
 99%|█████████▉| 10645/10708 [2:04:10<00:31,  2.03it/s]
 99%|█████████▉| 10646/10708 [2:04:10<00:30,  2.03it/s]
 99%|█████████▉| 10647/10708 [2:04:11<00:29,  2.03it/s]
 99%|█████████▉| 10648/10708 [2:04:11<00:29,  2.03it/s]
 99%|█████████▉| 10649/10708 [2:04:12<00:29,  2.03it/s]
 99%|█████████▉| 10650/10708 [2:04:12<00:28,  2.03it/s]{'loss': 3.5188, 'grad_norm': 0.19253556430339813, 'learning_rate': 8.937150777948988e-08, 'epoch': 0.99}
                                                       

 99%|█████████▉| 10650/10708 [2:04:12<00:28,  2.03it/s]
 99%|█████████▉| 10651/10708 [2:04:13<00:28,  2.03it/s]
 99%|█████████▉| 10652/10708 [2:04:13<00:27,  2.02it/s]
 99%|█████████▉| 10653/10708 [2:04:14<00:27,  2.02it/s]
 99%|█████████▉| 10654/10708 [2:04:14<00:26,  2.03it/s]
100%|█████████▉| 10655/10708 [2:04:15<00:26,  2.02it/s]
100%|█████████▉| 10656/10708 [2:04:15<00:25,  2.03it/s]
100%|█████████▉| 10657/10708 [2:04:16<00:25,  2.03it/s]
100%|█████████▉| 10658/10708 [2:04:16<00:24,  2.03it/s]
100%|█████████▉| 10659/10708 [2:04:16<00:24,  2.03it/s]
100%|█████████▉| 10660/10708 [2:04:17<00:23,  2.03it/s]
100%|█████████▉| 10661/10708 [2:04:17<00:23,  2.03it/s]
100%|█████████▉| 10662/10708 [2:04:18<00:22,  2.03it/s]
100%|█████████▉| 10663/10708 [2:04:18<00:22,  2.03it/s]
100%|█████████▉| 10664/10708 [2:04:19<00:21,  2.03it/s]
100%|█████████▉| 10665/10708 [2:04:19<00:21,  2.03it/s]
100%|█████████▉| 10666/10708 [2:04:20<00:20,  2.03it/s]
100%|█████████▉| 10667/10708 [2:04:20<00:20,  2.03it/s]
100%|█████████▉| 10668/10708 [2:04:21<00:19,  2.03it/s]
100%|█████████▉| 10669/10708 [2:04:21<00:19,  2.03it/s]
100%|█████████▉| 10670/10708 [2:04:22<00:18,  2.03it/s]
100%|█████████▉| 10671/10708 [2:04:22<00:18,  2.03it/s]
100%|█████████▉| 10672/10708 [2:04:23<00:17,  2.03it/s]
100%|█████████▉| 10673/10708 [2:04:23<00:17,  2.03it/s]
100%|█████████▉| 10674/10708 [2:04:24<00:16,  2.03it/s]
100%|█████████▉| 10675/10708 [2:04:24<00:16,  2.03it/s]{'loss': 3.5204, 'grad_norm': 0.19177909195423126, 'learning_rate': 2.8932084674138105e-08, 'epoch': 1.0}

                                                       

100%|█████████▉| 10675/10708 [2:04:24<00:16,  2.03it/s]
100%|█████████▉| 10676/10708 [2:04:25<00:15,  2.03it/s]
100%|█████████▉| 10677/10708 [2:04:25<00:15,  2.03it/s]
100%|█████████▉| 10678/10708 [2:04:26<00:14,  2.03it/s]
100%|█████████▉| 10679/10708 [2:04:26<00:14,  2.03it/s]
100%|█████████▉| 10680/10708 [2:04:27<00:13,  2.03it/s]
100%|█████████▉| 10681/10708 [2:04:27<00:13,  2.03it/s]
100%|█████████▉| 10682/10708 [2:04:28<00:12,  2.03it/s]
100%|█████████▉| 10683/10708 [2:04:28<00:12,  2.03it/s]
100%|█████████▉| 10684/10708 [2:04:29<00:11,  2.03it/s]
100%|█████████▉| 10685/10708 [2:04:29<00:11,  2.03it/s]
100%|█████████▉| 10686/10708 [2:04:30<00:10,  2.03it/s]
100%|█████████▉| 10687/10708 [2:04:30<00:10,  2.03it/s]
100%|█████████▉| 10688/10708 [2:04:31<00:09,  2.03it/s]
100%|█████████▉| 10689/10708 [2:04:31<00:09,  2.03it/s]
100%|█████████▉| 10690/10708 [2:04:32<00:08,  2.03it/s]
100%|█████████▉| 10691/10708 [2:04:32<00:08,  2.03it/s]
100%|█████████▉| 10692/10708 [2:04:33<00:07,  2.03it/s]
100%|█████████▉| 10693/10708 [2:04:33<00:07,  2.04it/s]
100%|█████████▉| 10694/10708 [2:04:34<00:06,  2.04it/s]
100%|█████████▉| 10695/10708 [2:04:34<00:06,  2.03it/s]
100%|█████████▉| 10696/10708 [2:04:35<00:05,  2.03it/s]
100%|█████████▉| 10697/10708 [2:04:35<00:05,  2.03it/s]
100%|█████████▉| 10698/10708 [2:04:36<00:04,  2.03it/s]
100%|█████████▉| 10699/10708 [2:04:36<00:04,  2.03it/s]
100%|█████████▉| 10700/10708 [2:04:37<00:03,  2.03it/s]
                                                       
{'loss': 3.5134, 'grad_norm': 0.19409115612506866, 'learning_rate': 1.7003399699899725e-09, 'epoch': 1.0}

100%|█████████▉| 10700/10708 [2:04:37<00:03,  2.03it/s]
100%|█████████▉| 10701/10708 [2:04:37<00:03,  2.03it/s]
100%|█████████▉| 10702/10708 [2:04:38<00:02,  2.03it/s]
100%|█████████▉| 10703/10708 [2:04:38<00:02,  2.03it/s]
100%|█████████▉| 10704/10708 [2:04:39<00:01,  2.03it/s]
100%|█████████▉| 10705/10708 [2:04:39<00:01,  2.03it/s]
100%|█████████▉| 10706/10708 [2:04:40<00:00,  2.03it/s]
100%|█████████▉| 10707/10708 [2:04:40<00:00,  2.03it/s]
100%|██████████| 10708/10708 [2:04:42<00:00,  1.30it/s]
                                                       
{'train_runtime': 7515.2039, 'train_samples_per_second': 1459.054, 'train_steps_per_second': 1.425, 'train_loss': 3.8992972186932513, 'epoch': 1.0}

100%|██████████| 10708/10708 [2:05:15<00:00,  1.30it/s]
100%|██████████| 10708/10708 [2:05:15<00:00,  1.42it/s]
Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.