Update for Transformers AWQ support

This commit is contained in:
ai-modelscope
2025-01-27 22:37:27 +08:00
parent 7b550f7fff
commit 79334b97df
14 changed files with 93970 additions and 60 deletions

34
config.json Normal file
View File

@@ -0,0 +1,34 @@
{
"_name_or_path": "vicuna-13b-v1.5",
"architectures": [
"LlamaForCausalLM"
],
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 13824,
"max_length": 4096,
"max_position_embeddings": 4096,
"model_type": "llama",
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 40,
"pad_token_id": 0,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.31.0",
"use_cache": true,
"vocab_size": 32000,
"quantization_config": {
"quant_method": "awq",
"zero_point": true,
"group_size": 128,
"bits": 4,
"version": "gemm"
}
}