model_config.json

{
    "type": "hfmodel",
    "config": {
        "model_path": "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
        "adapter_path": "/content/models/deepseek/en_ft/adapter",
        "task": "text-generation-with-past",
        "load_kwargs": {
            "torch_dtype": null,
            "device_map": null,
            "max_memory": null,
            "quantization_method": null,
            "quantization_config": null,
            "trust_remote_code": true,
            "extra_args": {
                "attn_implementation": "eager"
            }
        },
        "generative": true,
        "model_attributes": null,
        "io_config": null
    }
}