40 lines
979 B
JSON
40 lines
979 B
JSON
|
{
|
||
|
|
"architectures": [
|
||
|
|
"QWenLMHeadModel"
|
||
|
|
],
|
||
|
|
"attn_dropout_prob": 0.1,
|
||
|
|
"auto_map": {
|
||
|
|
"AutoConfig": "configuration_qwen.QWenConfig",
|
||
|
|
"AutoModelForCausalLM": "modeling_qwen.QWenLMHeadModel"
|
||
|
|
},
|
||
|
|
"bf16": true,
|
||
|
|
"emb_dropout_prob": 0.1,
|
||
|
|
"fp16": false,
|
||
|
|
"fp32": false,
|
||
|
|
"hidden_size": 5120,
|
||
|
|
"initializer_range": 0.02,
|
||
|
|
"intermediate_size": 27392,
|
||
|
|
"kv_channels": 128,
|
||
|
|
"layer_norm_epsilon": 1e-05,
|
||
|
|
"max_position_embeddings": 16384,
|
||
|
|
"model_type": "qwen",
|
||
|
|
"no_bias": true,
|
||
|
|
"num_attention_heads": 40,
|
||
|
|
"num_hidden_layers": 40,
|
||
|
|
"onnx_safe": null,
|
||
|
|
"padded_vocab_size": 154112,
|
||
|
|
"params_dtype": "torch.bfloat16",
|
||
|
|
"rotary_emb_base": 10000,
|
||
|
|
"rotary_pct": 1.0,
|
||
|
|
"scale_attn_weights": true,
|
||
|
|
"seq_length": 16384,
|
||
|
|
"tie_word_embeddings": false,
|
||
|
|
"tokenizer_type": "QWenTokenizer",
|
||
|
|
"transformers_version": "4.28.1",
|
||
|
|
"use_cache": true,
|
||
|
|
"use_dynamic_ntk": false,
|
||
|
|
"use_flash_attn": false,
|
||
|
|
"use_logn_attn": false,
|
||
|
|
"vocab_size": 154112
|
||
|
|
}
|