codefuse-ai/CodeFuse-CodeGeeX2-6B

1{2  "_name_or_path": "/mnt/user/qumu/download_models/codegeex2-6b",3  "add_bias_linear": false,4  "add_qkv_bias": true,5  "apply_query_key_layer_scaling": true,6  "apply_residual_connection_post_layernorm": false,7  "architectures": [8    "ChatGLMForConditionalGeneration"9  ],10  "attention_dropout": 0.0,11  "attention_softmax_in_fp32": true,12  "auto_map": {13    "AutoConfig": "configuration_chatglm.ChatGLMConfig",14    "AutoModel": "modeling_chatglm.ChatGLMForConditionalGeneration",15    "AutoModelForCausalLM": "modeling_chatglm.ChatGLMForConditionalGeneration",16    "AutoModelForSeq2SeqLM": "modeling_chatglm.ChatGLMForConditionalGeneration"17  },18  "bias_dropout_fusion": true,19  "eos_token": "</s>",20  "eos_token_id": 2,21  "ffn_hidden_size": 13696,22  "fp32_residual_connection": false,23  "hidden_dropout": 0.0,24  "hidden_size": 4096,25  "interleaved_qkv": false,26  "kv_channels": 128,27  "layernorm_epsilon": 1e-05,28  "model_type": "chatglm",29  "multi_query_attention": true,30  "multi_query_group_num": 2,31  "num_attention_heads": 32,32  "num_layers": 28,33  "original_rope": true,34  "pad_token": "<unk>",35  "pad_token_id": 0,36  "padded_vocab_size": 65024,37  "post_layer_norm": true,38  "pre_seq_len": null,39  "prefix_projection": false,40  "quantization_bit": 0,41  "rmsnorm": true,42  "rotary_percent": 0.5,43  "seq_length": 8192,44  "tie_word_embeddings": false,45  "torch_dtype": "bfloat16",46  "transformers_version": "4.33.2",47  "use_cache": true,48  "vocab_size": 6502449}50