Chess Datasets Translation

Currently, Qwen3-30B-A3B-Instruct-2507-FP8 is used for fast inference (> 1000 samples/min)

Getting started

1. Install vLLM:
   pip install vllm
2. Update paths in the translate.py file (global variables
   DATASET_PATH
   ,
   TRANSLATED_DATASET_PATH
   )
3. Run:
   vllm serve Qwen/Qwen3-30B-A3B-Instruct-2507-FP8 --gpu-memory-utilization 0.9
4. When server is ready, launch translation:
   python translate.py