Cauchy-3B-preview / config.json
MeowFET
feat: update model names, make vocab_size divisible to 128
ac14fde
{
"_name_or_path": "/mnt/kaiwu-group-pp-sh/edge_llm/track2_pretrained_cauchy/3b-cauchy-chat-anchor",
"_ori_bos_token_id": 1,
"_ori_eos_token_id": 2,
"architectures": [
"CauchyForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "configuration_cauchy.CauchyConfig",
"AutoModel": "modeling_cauchy.CauchyModel",
"AutoModelForCausalLM": "modeling_cauchy.CauchyForCausalLM",
"AutoModelForSeq2SeqLM": "modeling_cauchy.CauchyForCausalLM",
"AutoModelForSequenceClassification": "modeling_cauchy.CauchyForSequenceClassification"
},
"bos_token_id": 151643,
"dim_model_base": 256,
"eos_token_id": 151645,
"hidden_act": "silu",
"hidden_size": 2304,
"initializer_range": 0.1,
"intermediate_size": 5760,
"max_position_embeddings": 4096,
"model_type": "cauchy",
"num_attention_heads": 36,
"num_hidden_layers": 52,
"num_key_value_heads": 9,
"pretraining_tp": 1,
"qk_norm": true,
"qkv_bias": true,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 1000000.0,
"scale_depth": 1.4,
"scale_emb": 12,
"torch_dtype": "bfloat16",
"transformers_version": "4.43.3",
"use_cache": true,
"vocab_size": 151680
}