Files
LUFFY-Qwen-Math-1.5B-Zero/config.json
ModelHub XC 2428ca2ecb 初始化项目,由ModelHub XC社区提供模型
Model: Elliott/LUFFY-Qwen-Math-1.5B-Zero
Source: Original Platform
2026-06-17 07:43:17 +08:00

31 lines
863 B
JSON

{
"_name_or_path": "/mnt/petrelfs/yanjianhao/RL/rl4kd-ds/deepscaler/verl/checkpoints/baseline-rl-exps/MATH-1.5B_17_DRGRPO_v9_8K_P_DIV_P_0.1_NOCLIP/best/actor/",
"architectures": [
"Qwen2ForCausalLM"
],
"attention_dropout": 0.0,
"eos_token_id": 151643,
"hidden_act": "silu",
"hidden_size": 1536,
"initializer_range": 0.02,
"intermediate_size": 8960,
"max_position_embeddings": 16384,
"max_window_layers": 21,
"model_type": "qwen2",
"num_attention_heads": 12,
"num_hidden_layers": 28,
"num_key_value_heads": 2,
"pad_token_id": 151643,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"rope_theta": 40000.0,
"sliding_window": null,
"tie_word_embeddings": true,
"torch_dtype": "float32",
"transformers_version": "4.46.3",
"use_cache": true,
"use_mrope": false,
"use_sliding_window": false,
"vocab_size": 151936
}