Llama-3-Ko-MiniCPM-8b / config.json
maywell's picture
Update config.json
2dcdcf5 verified
{
"_name_or_path": "instructkr/Llama-3-Ko-MiniCPM-8b",
"architectures": [
"MiniCPMV"
],
"attention_bias": false,
"attention_dropout": 0.0,
"auto_map": {
"AutoConfig": "instructkr/Llama-3-Ko-MiniCPM-8b--configuration_minicpm.MiniCPMVConfig",
"AutoModel": "instructkr/Llama-3-Ko-MiniCPM-8b--modeling_minicpmv.MiniCPMV",
"AutoModelForCausalLM": "instructkr/Llama-3-Ko-MiniCPM-8b--modeling_minicpmv.MiniCPMV"
},
"batch_vision_input": true,
"bos_token_id": 128000,
"drop_vision_last_layer": false,
"eos_token_id": 128001,
"hidden_act": "silu",
"hidden_size": 4096,
"image_size": 448,
"initializer_range": 0.02,
"intermediate_size": 14336,
"max_position_embeddings": 8192,
"mlp_bias": false,
"mm_use_im_start_end": true,
"model_type": "minicpmv",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 8,
"patch_size": 14,
"pretraining_tp": 1,
"query_num": 96,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 500000.0,
"slice_config": {
"max_slice_nums": 9,
"model_type": "minicpmv"
},
"slice_mode": true,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.41.1",
"use_cache": true,
"vision_config": {
"hidden_size": 1152,
"image_size": 980,
"intermediate_size": 4304,
"model_type": "idefics2",
"num_attention_heads": 16,
"num_hidden_layers": 27,
"patch_size": 14
},
"vocab_size": 128256
}