What types of models can be submitted?

Hello HF team!

We tried to submit e.g. quantized models to the benchmark directly on [the web interface](https://huggingface.co/spaces/AIEnergyScore/submission_portal)? However, it mostly returns that the model is not compatible with the task while we use a LLM for the text generation task. As a concrete example, `PrunaAI/microsoft-phi-4-AWQ-4bit-smashed` or `PrunaAI/HuggingFaceTB-SmolLM-135M-Instruct-HQQ-1bit-smashed` returned this error.

- How can we submit the quantized models?
- Do you plan to support quantized models?

Looking forward to submit models here.