model: Qwen2.5-7B-Instruct-reward-model
execution: remote_only
interface: scalar_score_http
tensor_parallel_size: 1
base_model_path: /data/lry_machine_learning/models/Qwen2.5-7B-Instruct
checkpoint_path: /data/lry_machine_learning/checkpoints/ttc_dense_verifier/verifier_qwen7b_rm_P10MULTI_20260604_055439/checkpoint-600
training_run_id: P10MULTI_20260604_055439
loader: llamafactory_value_head
score_semantics: higher_is_better