config.json

{
    "architectures": [
        "ArlowGPTForImageTextToText"
    ],
    "auto_map": {
        "AutoConfig": "configuration_arlow_gpt.ArlowGPTConfig",
        "AutoModelForImageTextToText": "modeling_arlow_gpt.ArlowGPTForImageTextToText"
    },
    "clip_config": null,
    "clip_model_name": "yuchenxie/CLiP",
    "gpt2_config": null,
    "gpt2_model_name": "yuchenxie/GPT-2",
    "model_type": "arlow_gpt",
    "projection_dim": 768,
    "torch_dtype": "float32",
    "transformers_version": "4.46.2",
    "vocab_size": 50257,
    "_name_or_path": "yuchenxie/GPT-2V"
}