prm-qwen3-8b-bf16-full / export_metadata.json
devangb4's picture
Upload PRM LoRA adapter + head + tokenizer from checkpoint
26ba3ce verified
raw
history blame contribute delete
237 Bytes
{
"base_model": "Qwen/Qwen3-8B",
"dtype": "bfloat16",
"lora": {
"r": 16,
"alpha": 32,
"dropout": 0.05,
"targets": [
"q_proj",
"k_proj",
"v_proj",
"o_proj"
]
},
"max_seq_length": 384
}