See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/Llama-3.2-1B-Instruct
bf16: auto
chat_template: llama3
dataset_prepared_path: null
dataset_processes: 12
datasets:
- data_files:
  - /workspace/axolotl/data/ad9b0fa2-323a-4d04-be5e-1304b49c48da.json
  ds_type: json
  path: /workspace/axolotl/data/ad9b0fa2-323a-4d04-be5e-1304b49c48da.json
  type:
    field_input: problem
    field_instruction: type
    field_output: solution
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 512
eval_table_size: null
evals_per_epoch: 2
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: ncbateman/tuning-miner-testbed-ad9b0fa2-323a-4d04-be5e-1304b49c48da
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0001
load_in_4bit: false
load_in_8bit: true
local_rank: null
logging_steps: 1
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
micro_batch_size: 4
mlflow_experiment_name: https://5a301a635a9d0ac3cb7fcc3bf373c3c3.r2.cloudflarestorage.com/tuning/lighteval/MATH-Hard_train_data.json?X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=d49fdd0cc9750a097b58ba35b2d9fbed%2F20241024%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20241024T015513Z&X-Amz-Expires=604800&X-Amz-SignedHeaders=host&X-Amz-Signature=0879a56693622d0d57a6f707ace9e2a5e73d3647c1a227a4016e3e4011effe25
model_type: LlamaForCausalLM
num_epochs: 5
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 20
save_strategy: steps
sequence_len: 4096
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
val_set_size: 0.05
wandb_entity: breakfasthut
wandb_mode: online
wandb_project: tuning-miner
wandb_run: miner
wandb_runid: ad9b0fa2-323a-4d04-be5e-1304b49c48da
warmup_steps: 50
weight_decay: 0.0
xformers_attention: null

tuning-miner-testbed-ad9b0fa2-323a-4d04-be5e-1304b49c48da

This model is a fine-tuned version of unsloth/Llama-3.2-1B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7753

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0001
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• total_eval_batch_size: 8
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 50
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
0.8864	0.0095	1	0.9756
0.8585	0.4964	52	0.8117
0.7169	0.9928	104	0.7860
0.7558	1.4893	156	0.7723
0.7814	1.9857	208	0.7655
0.6849	2.4821	260	0.7670
0.7456	2.9785	312	0.7646
0.6037	3.4749	364	0.7706
0.6335	3.9714	416	0.7703
0.5835	4.4678	468	0.7749
0.7157	4.9642	520	0.7753

Framework versions

• PEFT 0.13.2
• Transformers 4.45.2
• Pytorch 2.3.1+cu121
• Datasets 3.0.1
• Tokenizers 0.20.1