Victor
Victor
@lontivero will need to add a GitHub secret named `NUGET_API_KEY` with their API key for NuGet. After that pushing any new tag will attempt to create and upload a NuGet...
I just want to run a benchmark, so I don't need the real model weights.
### Related Problem The well-known problem of NPM package bloat. ### The Elevator Pitch When I am looking for packages, information like download size of the package itself is insufficient...
### Description > CUDA error: operation failed due to a previous error during capture > current device: 0, in function ggml_cuda_op_mul_mat at D:\a\LLamaSharp\LLamaSharp\ggml\src\ggml-cuda.cu:1503 > cudaGetLastError() > GGML_ASSERT: D:\a\LLamaSharp\LLamaSharp\ggml\src\ggml-cuda.cu:101: !"CUDA error"...
### Description With asynchronous flows user might ask 2 questions in succession before model has time to answer the first one. This ends up in the following chat pattern: >...
No need to bother with run_server script. After package installation the server can be run using `python -m whisper_live.server` (similar to `python -m http.server`)
I wanted to use it to run on host machine to provide TURN for a VM.
This greatly simplifies debugging of SharpCompress source because it removes the need to setup nuget symbol downloading. It also simplifies debugging scenarios for anyone who is not using NuGet for...
Definitely one example with the best practice on error handling to give us the users a good start. And maybe one for passing .NET exception objects. ```[tasklist] ### Tasks ```
This is printed when I call `functional.scaled_dot_product_attention`: > [W914 13:25:36.000000000 sdp_utils.cpp:555] Warning: 1Torch was not compiled with flash attention. (function operator ()) I'm on Windows with `TorchSharp-cuda-windows=0.103.0`