See axolotl config

axolotl version: 0.4.1

yaml

1adapter: lora 2base_model: trl-internal-testing/tiny-random-LlamaForCausalLM 3bf16: true 4chat_template: llama3 5dataset_prepared_path: null 6datasets: 7- data_files: 8 - 5dade9b72f38e9d0_train_data.json 9 ds_type: json 10 format: custom 11 path: /workspace/input_data/5dade9b72f38e9d0_train_data.json 12 type: 13 field_input: sent_1 14 field_instruction: original_l1 15 field_output: original_l2 16 format: '{instruction} {input}' 17 no_input_format: '{instruction}' 18 system_format: '{system}' 19 system_prompt: '' 20debug: null 21device_map: 22 ? '' 23 : 0,1,2,3,4,5,6,7 24early_stopping_patience: 2 25eval_max_new_tokens: 128 26eval_steps: 100 27eval_table_size: null 28flash_attention: true 29gradient_accumulation_steps: 8 30gradient_checkpointing: true 31group_by_length: false 32hub_model_id: Alphatao/bd8adeda-f612-4dc5-87de-1ffd9d384f26 33hub_repo: null 34hub_strategy: null 35hub_token: null 36learning_rate: 0.0002 37load_best_model_at_end: true 38load_in_4bit: false 39load_in_8bit: false 40local_rank: null 41logging_steps: 1 42lora_alpha: 128 43lora_dropout: 0.1 44lora_fan_in_fan_out: null 45lora_model_dir: null 46lora_r: 64 47lora_target_linear: true 48lora_target_modules: 49- q_proj 50- k_proj 51- v_proj 52- o_proj 53lr_scheduler: cosine 54max_grad_norm: 1.0 55max_steps: 3600 56micro_batch_size: 4 57mlflow_experiment_name: /tmp/5dade9b72f38e9d0_train_data.json 58model_type: AutoModelForCausalLM 59num_epochs: 2 60optimizer: adamw_bnb_8bit 61output_dir: miner_id_24 62pad_to_sequence_len: true 63resume_from_checkpoint: null 64s2_attention: null 65sample_packing: false 66save_steps: 100 67sequence_len: 2048 68strict: false 69tf32: true 70tokenizer_type: AutoTokenizer 71train_on_inputs: false 72trust_remote_code: true 73val_set_size: 0.05 74wandb_entity: null 75wandb_mode: online 76wandb_name: 64fe8567-d552-4c6d-bd9d-632e648d102a 77wandb_project: Gradients-On-Demand 78wandb_run: your_name 79wandb_runid: 64fe8567-d552-4c6d-bd9d-632e648d102a 80warmup_steps: 10 81weight_decay: 0.0 82xformers_attention: null 83

bd8adeda-f612-4dc5-87de-1ffd9d384f26

• Loss: 10.2407

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

• learning_rate: 0.0002
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_BNB 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
• training_steps: 3600

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