config.json

{
  "architectures": [
    "CLIPSegModel"
  ],
  "conditional_layer": 0,
  "decoder_attention_dropout": 0.0,
  "decoder_hidden_act": "quick_gelu",
  "decoder_intermediate_size": 2048,
  "decoder_num_attention_heads": 4,
  "dtype": "float32",
  "extract_layers": [
    1
  ],
  "initializer_factor": 1.0,
  "logit_scale_init_value": 2.6592,
  "model_type": "clipseg",
  "projection_dim": 64,
  "reduce_dim": 32,
  "text_config": {
    "attention_dropout": 0.1,
    "bos_token_id": 49406,
    "dropout": 0.1,
    "eos_token_id": 49407,
    "hidden_act": "quick_gelu",
    "hidden_size": 32,
    "initializer_factor": 1.0,
    "initializer_range": 0.02,
    "intermediate_size": 37,
    "layer_norm_eps": 1e-05,
    "max_position_embeddings": 512,
    "model_type": "clipseg_text_model",
    "num_attention_heads": 4,
    "num_hidden_layers": 2,
    "pad_token_id": 49407,
    "vocab_size": 49408
  },
  "transformers_version": "5.16.0.dev0",
  "use_complex_transposed_convolution": false,
  "vision_config": {
    "attention_dropout": 0.1,
    "dropout": 0.1,
    "hidden_act": "quick_gelu",
    "hidden_size": 32,
    "image_size": 30,
    "initializer_factor": 1.0,
    "initializer_range": 0.02,
    "intermediate_size": 37,
    "layer_norm_eps": 1e-05,
    "model_type": "clipseg_vision_model",
    "num_attention_heads": 4,
    "num_channels": 3,
    "num_hidden_layers": 2,
    "patch_size": 2
  }
}