19 lines
728 B
YAML
19 lines
728 B
YAML
|
{
|
||
|
|
"experiment_name": "qwen3gen-chemistry-RLRT_TR-Qwen-Qwen3-8B-mbs8-tr0.1-decay0-train32-rollout8-lr1e-6-vllm0.8-epsw0.5",
|
||
|
|
"base_model": "Qwen/Qwen3-8B",
|
||
|
|
"dataset": "chemistry",
|
||
|
|
"method": "RLRT-TR",
|
||
|
|
"checkpoint_dir": "/mnt/mole/SDPO/L2T/checkpoints/datasets/sciknoweval/chemistry/qwen3gen-chemistry-RLRT_TR-Qwen-Qwen3-8B-mbs8-tr0.1-decay0-train32-rollout8-lr1e-6-vllm0.8-epsw0.5",
|
||
|
|
"best_checkpoint": "global_step_100",
|
||
|
|
"last_checkpoint": "global_step_100",
|
||
|
|
"train_batch_size": 32,
|
||
|
|
"train_max_samples": 3200,
|
||
|
|
"teacher_regularization": "trust-region",
|
||
|
|
"teacher_update_rate": 0.1,
|
||
|
|
"rollout_n": 8,
|
||
|
|
"validation_n": 16,
|
||
|
|
"learning_rate": "1e-6",
|
||
|
|
"vllm_gpu_memory_utilization": 0.8,
|
||
|
|
"command_args": ""
|
||
|
|
}
|