# Trainer.predict in parallel not supported! **URL:** \ **Category:** 🤗Transformers\ **Created:** [November 1, 2022, 5:22pm UTC](https://discuss.huggingface.co/t/trainer-predict-in-parallel-not-supported/25347 "2022-11-01T17:22:40Z")\ **Posts on this page:** 3\ **Page:** 1
**Author:** ![djaym7](https://avatars.discourse-cdn.com/v4/letter/d/b5a626/32.png) [@djaym7](https://discuss.huggingface.co/u/djaym7)\ **Post date:** [November 1, 2022, 5:22pm UTC](https://discuss.huggingface.co/t/trainer-predict-in-parallel-not-supported/25347/1 "2022-11-01T17:22:40Z")
Currently doing any inference via trainer.predict only uses 1 gpu to do all the computations. Has someone done any parallelization for this ? Split the data among all available gpus and do inference, aggregate all metrics once all processes are done ? ---
**Author:** ![sgugger](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/sgugger/32/2291_2.png) [@sgugger](https://discuss.huggingface.co/u/sgugger)\ **Post date:** [November 2, 2022, 1:56pm UTC](https://discuss.huggingface.co/t/trainer-predict-in-parallel-not-supported/25347/2 "2022-11-02T13:56:20Z")
It is supported and you can have a look at our [examples](https://github.com/huggingface/transformers/tree/main/examples) which all use distributed evaluation if you have several GPUs. ---
**Author:** ![djaym7](https://avatars.discourse-cdn.com/v4/letter/d/b5a626/32.png) [@djaym7](https://discuss.huggingface.co/u/djaym7)\ **Post date:** [November 2, 2022, 4:49pm UTC](https://discuss.huggingface.co/t/trainer-predict-in-parallel-not-supported/25347/3 "2022-11-02T16:49:33Z")
No, it doesn’t run it parallel. Trainer.predict() keeps using 1 gpu and rest of 7 sit idle. Different from that, I made my own which divides data and trainer.predicts on all N gpus and then gathers output back. This reduces time N times vs default usage of trainer.predict()