Introduce model configs to abstract pairings of models and hardware

[yk]

• introduces the concept of "model config" defined in model_configs.py that can be referenced by the frontend by name
• adjusted dockerfiles to automatically determine model id and quantization from a given config name
• introduced an OOM test to find out sequence length and batch size limits on different GPUs, results in notion

[github-actions]

❌ pre-commit failed.
Please run pre-commit run --all-files locally and commit the changes.
Find more information in the repository's CONTRIBUTING.md