text_encoder/config.json

{
  "architectures": [
    "ClapModel"
  ],
  "audio_config": {
    "_attn_implementation_autoset": true,
    "depths": [
      1,
      1
    ],
    "hidden_size": 192,
    "intermediate_size": 37,
    "model_type": "clap_audio_model",
    "num_attention_heads": [
      1,
      1
    ],
    "num_hidden_layers": 1,
    "num_mel_bins": 8,
    "patch_embed_input_channels": 4,
    "patch_size": 2,
    "patch_stride": 2,
    "projection_dim": 16,
    "spec_size": 8,
    "window_size": 4
  },
  "hidden_size": 8,
  "initializer_factor": 1.0,
  "logit_scale_init_value": 14.285714285714285,
  "model_type": "clap",
  "num_hidden_layers": 3,
  "projection_dim": 16,
  "projection_hidden_act": "relu",
  "text_config": {
    "_attn_implementation_autoset": true,
    "hidden_size": 8,
    "intermediate_size": 37,
    "layer_norm_eps": 1e-05,
    "model_type": "clap_text_model",
    "num_attention_heads": 1,
    "num_hidden_layers": 1,
    "projection_dim": 16,
    "vocab_size": 1000
  },
  "torch_dtype": "float32",
  "transformers_version": "4.50.0.dev0"
}