30 lines
809 B
JSON
30 lines
809 B
JSON
|
{
|
||
|
|
"train_file": "data_processed/medrect_v2/mixed_sft_train.jsonl",
|
||
|
|
"eval_split": 0.05,
|
||
|
|
"limit": null,
|
||
|
|
"model_name": "Qwen/Qwen3-4B",
|
||
|
|
"output_dir": "outputs/local_training/qwen3-4b-medrect-mixed-v2",
|
||
|
|
"max_seq_length": 4096,
|
||
|
|
"per_device_train_batch_size": 2,
|
||
|
|
"gradient_accumulation_steps": 16,
|
||
|
|
"learning_rate": 0.0001,
|
||
|
|
"num_train_epochs": 3,
|
||
|
|
"warmup_ratio": 0.1,
|
||
|
|
"weight_decay": 0.1,
|
||
|
|
"logging_steps": 10,
|
||
|
|
"save_steps": 100,
|
||
|
|
"eval_steps": 50,
|
||
|
|
"bf16": true,
|
||
|
|
"gradient_checkpointing": true,
|
||
|
|
"lora_r": 64,
|
||
|
|
"lora_alpha": 128,
|
||
|
|
"lora_dropout": 0.05,
|
||
|
|
"lora_target_modules": "all-linear",
|
||
|
|
"early_stopping_patience": 3,
|
||
|
|
"no_early_stopping": false,
|
||
|
|
"wandb": true,
|
||
|
|
"wandb_project": "medrect-mixed-sft-v2",
|
||
|
|
"seed": 42,
|
||
|
|
"debug_samples": 2,
|
||
|
|
"dataloader_num_workers": 4
|
||
|
|
}
|