HanjoonBespinGlobal's picture
3 epochs with 256 batch
1dcccb4
{
"_name_or_path": "BAAI/bge-reranker-v2-minicpm-layerwise",
"architectures": [
"LayerWiseMiniCPMForCausalLM"
],
"attention_bias": false,
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "BAAI/bge-reranker-v2-minicpm-layerwise--configuration_minicpm_reranker.LayerWiseMiniCPMConfig",
"AutoModel": "BAAI/bge-reranker-v2-minicpm-layerwise--modeling_minicpm_reranker.LayerWiseMiniCPMModel",
"AutoModelForCausalLM": "BAAI/bge-reranker-v2-minicpm-layerwise--modeling_minicpm_reranker.LayerWiseMiniCPMForCausalLM"
},
"bos_token_id": 1,
"dim_model_base": 256,
"eos_token_id": 2,
"head_multi": true,
"head_type": "simple",
"hidden_act": "silu",
"hidden_size": 2304,
"initializer_range": 0.1,
"intermediate_size": 5760,
"max_position_embeddings": 2048,
"model_type": "minicpm",
"num_attention_heads": 36,
"num_hidden_layers": 40,
"num_key_value_heads": 36,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 10000.0,
"scale_depth": 1.4,
"scale_emb": 12,
"start_layer": 8,
"torch_dtype": "float32",
"transformers_version": "4.41.1",
"use_cache": false,
"vocab_size": 122753
}