LI TIANCHENG

Apple Silicon Unified Memory usage

2

### 🚀 The feature, motivation and pitch Since Mac with apple silicon has unified memory, why in PyTorch we still need to copy tensors from cpu and mps? This will...

error email address when sign up to the early access list

1

**Describe the bug** When I want to sign up for the early access, I tried several email address. But they all showed that: Please use a valid email address. **To...

Add whisper.cpp support, which support various hardware architecture, including CPU, NVIDIA GPU, AMD GPU, Intel GPU and so on. Later I will write a doc about how to compile whisper.cpp,...

1. Add whisper.cpp support, which support various hardware architecture, including CPU, NVIDIA GPU, AMD GPU, Intel GPU and so on. Later I will write a doc about how to compile...

Does anyone know how much a GPU can accelerate in comparison to multicore CPU?

8

### Details Does anyone know how much a GPU can accelerate in comparison to multicore CPU? Since the CPU cores for GPU nodes are limited, it worthwhile to apply a...

### Describe the bug I use this INPUT file to test my build from source code: ```text INPUT_PARAMETERS #Parameters (General) calculation scf pseudo_dir . orbital_dir . ntype 1 #Parameters (Accuracy)...

Since whisperx cannot ultilize GPU on Mac, so I remove support for macos, use whisperkit instead

audioToSrt orchestrates the transcription, translation, and timestamping pipeline. Transcription: Sequential Translation: Parallel (Managed by errgroup) Timestamping: Sequential (Reads output from translation)

使用whisperkit转换该模型[Belle-whisper-larger-v3-turbo-zh](https://huggingface.co/BELLE-2/Belle-whisper-large-v3-turbo-zh)以后用不了，一直卡在加载模型的地方，转换是显示成功的。 机器：Mac mini M4Pro