初始化项目,由ModelHub XC社区提供模型

Model: ahxt/LiteLlama-460M-1T
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-07-25 10:08:10 +08:00
commit 3ae0b5c0de
11 changed files with 100530 additions and 0 deletions

26
config.json Normal file
View File

@@ -0,0 +1,26 @@
{
"architectures": [
"LlamaForCausalLM"
],
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 4096,
"layernorm_epsilon": 1e-05,
"max_position_embeddings": 1024,
"model_type": "llama",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"num_key_value_heads": 2,
"pad_token_id": 0,
"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": 50304
}