See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: EleutherAI/pythia-410m-deduped
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 591019ce86cf2dcb_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/591019ce86cf2dcb_train_data.json
  type:
    field_input: submenu
    field_instruction: top_section
    field_output: row2text
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_steps: null
eval_table_size: null
evals_per_epoch: 4
flash_attention: false
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: true
hub_model_id: diaenra/6a434f3b-2277-42ac-ba05-6a667a3e971b
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0001
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
max_memory:
  0: 70GB
micro_batch_size: 4
mlflow_experiment_name: /tmp/591019ce86cf2dcb_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 1
sequence_len: 512
special_tokens:
  pad_token: <|endoftext|>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: diaenra-tao-miner
wandb_mode: online
wandb_name: 17a240b4-ceb0-41a9-89f8-a59905875425
wandb_project: tao
wandb_run: diaenra
wandb_runid: 17a240b4-ceb0-41a9-89f8-a59905875425
warmup_steps: 100
weight_decay: 0.0
xformers_attention: true

6a434f3b-2277-42ac-ba05-6a667a3e971b

This model is a fine-tuned version of EleutherAI/pythia-410m-deduped on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.2394

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
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 100
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
4.0358	1.0	1	2.1946
4.0358	2.0	2	2.2394

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1