base_model: meta-llama/Llama-3.2-3B-Instruct

datasets:
  - path: axolotl_format_data_llama_wm.json
    type: input_output
dataset_prepared_path: last_run_prepared
    
output_dir: ./models/llama_wm
sequence_length: 2048

wandb_project: agent-v0
wandb_name: llama-3b_wm

train_on_inputs: false
gradient_checkpointing: true
gradient_accumulation_steps: 4
micro_batch_size: 1
num_epochs: 3
optimizer: adamw_torch
learning_rate: 2e-5

logging_steps: 5

warmup_steps: 10
saves_per_epoch: 1
weight_decay: 0.0

deepspeed: axolotl/deepspeed_configs/zero3_bf16_cpuoffload_all.json

special_tokens:
  pad_token: <|end_of_text|>

models/llama_wm

This model is a fine-tuned version of meta-llama/Llama-3.2-3B-Instruct on the axolotl_format_data_llama_wm.json dataset.

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: 2e-05
train_batch_size: 1
eval_batch_size: 1
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 4
total_train_batch_size: 32
total_eval_batch_size: 8
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: 10
num_epochs: 3

Training results

Framework versions

Transformers 4.46.3
Pytorch 2.5.1+cu124
Datasets 3.1.0
Tokenizers 0.20.3

mfirth/l3t_wm

models/llama_wm

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

mfirth/
l3t_wm