|
{ |
|
"_name_or_path": "openbmb/RankCPM-R", |
|
"architectures": [ |
|
"MiniCPM" |
|
], |
|
"auto_map": { |
|
"AutoConfig": "configuration_minicpm.MiniCPMConfig", |
|
"AutoModel": "modeling_minicpm.MiniCPMModel", |
|
"AutoModelForCausalLM": "modeling_minicpm.MiniCPMForCausalLM", |
|
"AutoModelForSeq2SeqLM": "modeling_minicpm.MiniCPMForCausalLM", |
|
"AutoModelForSequenceClassification": "modeling_minicpm.MiniCPMForSequenceClassification" |
|
}, |
|
"bos_token_id": 1, |
|
"eos_token_id": 2, |
|
"hidden_act": "silu", |
|
"hidden_size": 2304, |
|
"initializer_range": 0.1, |
|
"intermediate_size": 5760, |
|
"is_causal": false, |
|
"max_position_embeddings": 2048, |
|
"num_attention_heads": 36, |
|
"num_hidden_layers": 40, |
|
"num_key_value_heads": 36, |
|
"rms_norm_eps": 1e-05, |
|
"rope_scaling": null, |
|
"torch_dtype": "bfloat16", |
|
"transformers_version": "4.36.0", |
|
"use_cache": true, |
|
"vocab_size": 122753, |
|
"scale_emb": 12, |
|
"dim_model_base": 256, |
|
"scale_depth": 1.4, |
|
"num_labels": 1 |
|
} |