See axolotl config

axolotl version: 0.4.1

adapter: qlora
base_model: tuneai/Meta-Llama-3-8B-Instruct
base_model_config: tuneai/Meta-Llama-3-8B-Instruct
chat_template: llama3
datasets:
- conversation: llama3
  data_files: /root/.cache/model/chat/alpaca-jsonl-rfp-response-1.jsonl
  ds_type: json
  path: /root/.cache/model/chat/alpaca-jsonl-rfp-response-1.jsonl
  type: sharegpt
eval_sample_packing: false
eval_steps: 50
flash_attention: true
gradient_accumulation_steps: 4
gradient_checkpointing: true
hf_use_auth_token: true
hub_model_id: esha111/alpaca-2
learning_rate: 0.0002
load_in_4bit: true
logging_steps: 1
lora_alpha: 16
lora_dropout: 0.05
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
micro_batch_size: 2
model_type: AutoModelForCausalLM
num_epochs: 6
optimizer: paged_adamw_32bit
output_dir: /root/.cache/model/esha111/alpaca-2-model-5sbvomla
pad_to_sequence_len: true
sample_packing: true
save_safetensors: true
sequence_len: 4096
special_tokens:
  pad_token: <|end_of_text|>
tokenizer_type: AutoTokenizer
wandb_project: finetune-rfp-response-1-tune-studio
wandb_run_id: '3'
wandb_watch: 'true'
warmup_steps: 10

alpaca-2

This model is a fine-tuned version of tuneai/Meta-Llama-3-8B-Instruct on the None 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: 0.0002
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.PAGED_ADAMW 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: 6

Training results

Framework versions

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