60 lines
1.5 KiB
JSON
60 lines
1.5 KiB
JSON
|
{
|
||
|
|
"architectures": [
|
||
|
|
"mPLUGDocOwl2"
|
||
|
|
],
|
||
|
|
"auto_map": {
|
||
|
|
"AutoConfig": "configuration_mplug_docowl.MPLUGDocOwlConfig",
|
||
|
|
"AutoModel": "modeling_mplug_docowl.MPLUGDocOwl2",
|
||
|
|
"AutoModelForCausalLM": "modeling_mplug_docowl.MPLUGDocOwl2"
|
||
|
|
},
|
||
|
|
"attention_bias": false,
|
||
|
|
"bos_token_id": 1,
|
||
|
|
"eos_token_id": 2,
|
||
|
|
"hidden_act": "silu",
|
||
|
|
"hidden_size": 4096,
|
||
|
|
"initializer_range": 0.02,
|
||
|
|
"intermediate_size": 11008,
|
||
|
|
"max_position_embeddings": 2048,
|
||
|
|
"model_type": "mplug_docowl",
|
||
|
|
"num_attention_heads": 32,
|
||
|
|
"num_hidden_layers": 32,
|
||
|
|
"num_key_value_heads": 32,
|
||
|
|
"pretraining_tp": 1,
|
||
|
|
"rms_norm_eps": 1e-06,
|
||
|
|
"rope_scaling": null,
|
||
|
|
"rope_theta": 10000.0,
|
||
|
|
"tie_word_embeddings": false,
|
||
|
|
"transformers_version": "4.39.3",
|
||
|
|
"use_cache": true,
|
||
|
|
"visual_config": {
|
||
|
|
"visual_hrcompressor": {
|
||
|
|
"layer": 2,
|
||
|
|
"high_reso_cross_num_att_heads": 16,
|
||
|
|
"high_reso_cross_hid_size": 4096,
|
||
|
|
"high_reso_cross_dropout": 0.0
|
||
|
|
},
|
||
|
|
"visual_hreducer": {
|
||
|
|
"conv_shape": "1x4",
|
||
|
|
"hidden_size": 1024
|
||
|
|
},
|
||
|
|
"visual_model": {
|
||
|
|
"attention_dropout": 0.0,
|
||
|
|
"hidden_act": "quick_gelu",
|
||
|
|
"hidden_size": 1024,
|
||
|
|
"image_size": 504,
|
||
|
|
"initializer_factor": 1.0,
|
||
|
|
"initializer_range": 0.02,
|
||
|
|
"intermediate_size": 4096,
|
||
|
|
"layer_norm_eps": 1e-06,
|
||
|
|
"model_type": "mplug_owl_vision_model",
|
||
|
|
"num_attention_heads": 16,
|
||
|
|
"num_channels": 3,
|
||
|
|
"num_hidden_layers": 24,
|
||
|
|
"patch_size": 14,
|
||
|
|
"projection_dim": 768,
|
||
|
|
"use_flash_attn": false
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"vocab_size": 32000
|
||
|
|
}
|