uclaml / SPIN

The official implementation of Self-Play Fine-Tuning (SPIN)
https://uclaml.github.io/SPIN/
Apache License 2.0
995 stars 89 forks source link

small bugfixes for vllm #4

Closed sumo43 closed 7 months ago

sumo43 commented 7 months ago
  1. changed generate_vllm.sh to call generate_vllm.py
  2. changed batch_size back to world_size since it only controls the number of GPUs vllm uses
  3. added vllm to setup.py