Files
LLaMA-1B-dj-refine-100B/config.json
ModelHub XC be7ce7aa3e 初始化项目,由ModelHub XC社区提供模型
Model: Data-Juicer/LLaMA-1B-dj-refine-100B
Source: Original Platform
2026-06-29 20:43:12 +08:00

27 lines
613 B
JSON

{
"architectures": [
"LLaMAForCausalLM"
],
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 2048,
"initializer_range": 0.02,
"intermediate_size": 5504,
"max_position_embeddings": 2048,
"max_sequence_length": 2048,
"model_type": "llama",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"pad_token_id": 32004,
"pretraining_tp": 1,
"rms_norm_eps": 1e-06,
"rope_scaling": null,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.31.0",
"use_cache": true,
"vocab_size": 32128
}