config.json

{
  "architectures": [
    "Fgclip2Model"
  ],
  "auto_map": {
    "AutoConfig": "modeling_fgclip2.Fgclip2Config",
    "AutoModelForCausalLM": "modeling_fgclip2.Fgclip2Model"
  },
  "dtype": "float32",
  "initializer_factor": 1.0,
  "model_type": "fgclip2",
  "text_config": {
    "attention_dropout": 0.0,
    "dtype": "float32",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 768,
    "intermediate_size": 3072,
    "keep_len": 20,
    "layer_norm_eps": 1e-06,
    "longtext_len": 196,
    "max_position_embeddings": 64,
    "model_type": "fgclip2_text_model",
    "num_attention_heads": 12,
    "num_hidden_layers": 12,
    "projection_size": 768,
    "vocab_size": 256000
  },
  "transformers_version": "4.57.0.dev0",
  "vision_config": {
    "attention_dropout": 0.0,
    "dtype": "float32",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 768,
    "intermediate_size": 3072,
    "layer_norm_eps": 1e-06,
    "model_type": "fgclip2_vision_model",
    "num_attention_heads": 12,
    "num_channels": 3,
    "num_hidden_layers": 12,
    "num_patches": 256,
    "patch_size": 16
  }
}