See axolotl config

axolotl version: 0.4.0

adapter: qlora
base_model: meta-llama/Meta-Llama-3-8B-Instruct
base_model_config: meta-llama/Meta-Llama-3-8B-Instruct
datasets:
- path: ibivibiv/causal-lm-smaller_0
  type: alpaca
flash_attention: true
gradient_accumulation_steps: 4
gradient_checkpointing: true
hf_use_auth_token: true
hub_model_id: ibivibiv/causal-llama-3-8B-Instruct
learning_rate: 0.00025
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: 3
optimizer: paged_adamw_32bit
output_dir: /job/out
sample_packing: true
save_safetensors: true
sequence_len: 4096
special_tokens:
  pad_token: <|end_of_text|>
tokenizer_type: AutoTokenizer
wandb_project: TuneStudio
wandb_run_id: causalllama0
wandb_watch: 'true'
warmup_steps: 10

causal-llama-3-8B-Instruct

This model is a fine-tuned version of meta-llama/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: 0.00025
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• total_eval_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• num_epochs: 3

Training results

Framework versions

• PEFT 0.10.0
• Transformers 4.40.2
• Pytorch 2.1.2+cu118
• Datasets 2.19.1
• Tokenizers 0.19.1