{ "architectures": [ "CLIPTextModelWithProjection" ], "attention_dropout": 0.0, "bos_token_id": 0, "eos_token_id": 2, "hidden_act": "gelu", "hidden_size": 32, "initializer_factor": 1.0, "initializer_range": 0.02, "intermediate_size": 37, "layer_norm_eps": 1e-05, "max_position_embeddings": 77, "model_type": "clip_text_model", "num_attention_heads": 4, "num_hidden_layers": 5, "pad_token_id": 1, "projection_dim": 32, "torch_dtype": "float32", "transformers_version": "4.34.1", "vocab_size": 1000 }