Files
FF_3/config.json
ModelHub XC 2e9dc165e0 初始化项目,由ModelHub XC社区提供模型
Model: francescofiamingo1/FF_3
Source: Original Platform
2026-05-31 05:07:17 +08:00

23 lines
492 B
JSON

{
"architectures": ["GPT2LMHeadModel"],
"model_type": "gpt2",
"vocab_size": 50257,
"n_embd": 2048,
"n_head": 16,
"n_inner": 8192,
"n_layer": 38,
"n_positions": 2048,
"n_ctx": 2048,
"bos_token_id": 50256,
"eos_token_id": 50256,
"activation_function": "gelu_new",
"attn_pdrop": 0.0,
"embd_pdrop": 0.0,
"resid_pdrop": 0.0,
"layer_norm_epsilon": 1e-05,
"initializer_range": 0.02,
"torch_dtype": "bfloat16",
"use_cache": true,
"tie_word_embeddings": true
}