FP8-Emulation-Toolkit
FP8-Emulation-Toolkit copied to clipboard
PyTorch extension for emulating FP8 data formats on standard FP32 Xeon/GPU hardware.
this line of code https://github.com/IntelLabs/FP8-Emulation-Toolkit/blob/main/examples/training/bert/run_qa_no_trainer.py#L725 cannot work since squad or squad_v2 didn't provide an evaluate subclass.
Hello, is there a hardware reason for support starting from V100 GPU or would it be possible to support fp8 emulation on Pascal (P6000 with GP102 in my case). Thank...
Hi, thanks for providing such complete toolkit, but I have some questions about this toolkit. I use this toolkit to evaluate resnet18 on cifar10 with FP8 in hybrid mode, I...
Hello, I intergrated your fp8 emulator with my Lenet (2 conv layers, 3 fc layers) training process. When I set `list_exempt_layers = ["conv1"]`, everything works well. However, when I set...
Hello, I am trying to set it up as per your instructions but I got an error like in the title. I ran this on Google Colab ``` !git clone...
请问下enable fp8后,性能相比与fp16 fp32提升吗,比如resnet?