Update for Transformers AWQ support

This commit is contained in:
ai-modelscope
2025-02-15 00:56:32 +08:00
parent 9df68ab171
commit 4e56254ecb
15 changed files with 93932 additions and 55 deletions

34
config.json Normal file
View File

@@ -0,0 +1,34 @@
{
"_name_or_path": "/workspace/hub/models--Gryphe--MythoMax-L2-13b/snapshots/faa4ef8c87dbb00d447904ceb048d49b6a463d07",
"architectures": [
"LlamaForCausalLM"
],
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 13824,
"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,
"rope_theta": 10000.0,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.33.0.dev0",
"use_cache": true,
"vocab_size": 32001,
"quantization_config": {
"quant_method": "awq",
"zero_point": true,
"group_size": 128,
"bits": 4,
"version": "gemm"
}
}