| { | |
| "model_name": "unsloth/Llama-3.2-1B-Instruct", | |
| "max_seq_length": 2048, | |
| "dtype": null, | |
| "load_in_4bit": false, | |
| "lora_params": { | |
| "r": 16, | |
| "target_modules": [ | |
| "q_proj", | |
| "k_proj", | |
| "v_proj", | |
| "o_proj", | |
| "gate_proj", | |
| "up_proj", | |
| "down_proj" | |
| ], | |
| "lora_alpha": 16, | |
| "lora_dropout": 0, | |
| "bias": "none", | |
| "use_gradient_checkpointing": "unsloth", | |
| "random_state": 3407, | |
| "use_rslora": false | |
| }, | |
| "training_args": { | |
| "per_device_train_batch_size": 4, | |
| "gradient_accumulation_steps": 4, | |
| "warmup_steps": 5, | |
| "num_train_epochs": 1, | |
| "learning_rate": 0.0002, | |
| "fp16": false, | |
| "bf16": false, | |
| "logging_steps": 3, | |
| "optim": "adamw_8bit", | |
| "weight_decay": 0.01, | |
| "lr_scheduler_type": "linear", | |
| "seed": 3407, | |
| "output_dir": "outputs", | |
| "report_to": "none" | |
| }, | |
| "data_config": { | |
| "dataset_name": "mlabonne/FineTome-100k", | |
| "dataset_split": "train", | |
| "dataset_rows": 15000, | |
| "conversation_format": "llama-3.1", | |
| "max_seq_length": 2048, | |
| "train_on_responses_only": true | |
| }, | |
| "inference_config": { | |
| "temperature": 1.5, | |
| "min_p": 0.1, | |
| "max_new_tokens": 600 | |
| } | |
| } |