| adapter: lora | |
| base_model: Qwen/Qwen2.5-14B-Instruct | |
| batch_size: 64 | |
| bf16: true | |
| chat_template: tokenizer_default_fallback_alpaca | |
| datasets: | |
| - format: custom | |
| path: prithivMLmods/Song-Catalogue-Long-Thought | |
| type: | |
| field_input: song | |
| field_instruction: artist | |
| field_output: link | |
| format: '{instruction} {input}' | |
| no_input_format: '{instruction}' | |
| system_format: '{system}' | |
| system_prompt: '' | |
| device_map: auto | |
| eval_sample_packing: false | |
| eval_steps: 0.1 | |
| flash_attention: true | |
| gradient_checkpointing: true | |
| group_by_length: true | |
| hub_model_id: SystemAdmin123/Qwen2.5-14B-Instruct_36619497 | |
| hub_strategy: checkpoint | |
| learning_rate: 0.0001 | |
| logging_steps: 10 | |
| lora_alpha: 256 | |
| lora_dropout: 0.1 | |
| lora_r: 128 | |
| lora_target_linear: true | |
| lr_scheduler: cosine | |
| max_steps: 1600.0 | |
| micro_batch_size: 1 | |
| model_type: AutoModelForCausalLM | |
| num_epochs: 10000 | |
| optimizer: adamw_bnb_8bit | |
| output_dir: /root/.sn56/axolotl/tmp/Qwen2.5-14B-Instruct_36619497 | |
| pad_to_sequence_len: true | |
| resize_token_embeddings_to_32x: false | |
| sample_packing: true | |
| save_steps: 40 | |
| save_total_limit: 1 | |
| sequence_len: 2048 | |
| tokenizer_type: Qwen2TokenizerFast | |
| torch_dtype: bf16 | |
| training_args_kwargs: | |
| disable_tqdm: true | |
| hub_private_repo: true | |
| save_only_model: true | |
| trust_remote_code: true | |
| val_set_size: 0.01 | |
| wandb_entity: '' | |
| wandb_mode: online | |
| wandb_name: Qwen/Qwen2.5-14B-Instruct-prithivMLmods/Song-Catalogue-Long-Thought | |
| wandb_project: Gradients-On-Demand | |
| wandb_run: your_name | |
| wandb_runid: default | |
| warmup_ratio: 0.05 | |