Zero-Shot Image Classification
Transformers
Safetensors
siglip
vision
wirthual's picture
Add architecture to config file
00dd00c verified
raw
history blame
581 Bytes
{
"architectures": [
"SiglipModel"
],
"initializer_factor": 1.0,
"model_type": "siglip",
"text_config": {
"hidden_size": 1024,
"intermediate_size": 4096,
"model_type": "siglip_text_model",
"num_attention_heads": 16,
"num_hidden_layers": 24,
"projection_size": 1024,
"vocab_size": 256000
},
"transformers_version": "4.49.0.dev0",
"vision_config": {
"hidden_size": 1024,
"image_size": 384,
"intermediate_size": 4096,
"model_type": "siglip_vision_model",
"num_attention_heads": 16,
"num_hidden_layers": 24
}
}