初始化项目,由ModelHub XC社区提供模型

Model: ibm-granite/granite-8b-code-base-128k
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-05-02 21:05:34 +08:00
commit ccad2c5a9e
13 changed files with 99429 additions and 0 deletions

29
config.json Normal file
View File

@@ -0,0 +1,29 @@
{
"architectures": [
"LlamaForCausalLM"
],
"attention_bias": true,
"attention_dropout": 0.1,
"bos_token_id": 0,
"eos_token_id": 0,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 128000,
"mlp_bias": true,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 36,
"num_key_value_heads": 8,
"pad_token_id": 0,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_theta": 10000000,
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.41.0",
"use_cache": true,
"vocab_size": 49152,
"max_model_len": 128000
}