docs/source/en/perf_infer_gpu_one.md · 2cc8cf6ce7ae0416561acbb639df4bbc5f409b6f · chenpangpang / transformers · GitLab

Find file Blame History Permalink

Fix `torch.compile` with `fullgraph=True` when `attention_mask` input is used (#29211) · 2cc8cf6c
fxmarty authored Feb 22, 2024
```
* fix torch.export.export for llama

* do not change doc title

* make fix copies
```
2cc8cf6c

perf_infer_gpu_one.md 20.8 KB