text_encoder/model_config.json

{
  "max_text_length": 77,
  "vocab_size": 49408,
  "text_embed_dim": 768,
  "text_heads": 12,
  "text_layers": 12,
  "text_hidden_act": "quick_gelu",
  "projection_dim": 768,
  "initializer_range": 0.02,
  "initializer_factor": 1.0,
  "init_class": "CLIPTextModel"
}