Const-me/Whisper on Hacker News

…just on nVidia. Whisper was the first, but I can’t say I’m happy because too much complicated C++ https://github.com/Const-me/Whisper For Mistral I decided to rely on C# as much as possible and I like the result much better…

comment by Const-me · 2024-08-12 · in Ask HN: What are some "toy" projects you used to learn neural networks hands-on?

True, and it’s not just games: https://github.com/Const-me/Whisper/issues/42

comment by Const-me · 2024-06-07 · in I learned Vulkan and wrote a small game engine with it

…kind of GPU, with reasonable performance… with a year or so I have implemented inference of Whisper https://github.com/Const-me/Whisper and Mistral https://github.com/Const-me/Cgml/tree/master/Mistral/Mistral... models on all GPUs which…

comment by Const-me · 2024-05-20 · in Llama3 implemented from scratch

I've been using WhisperDesktop ( https://github.com/Const-me/Whisper ) with great success on a 3090 for fast & accurate transcription of often poor quality euro-english hours long multispeaker audio files. If there's an easy way to compare…

comment by PeterStuer · 2024-04-18 · in Nvidia Speech and Translation AI Models Set Records for Speed and Accuracy

…D3D11 compute shaders gonna be adequate for a Torch backend, and they even work on Linux with Wine: https://github.com/Const-me/Whisper/issues/42 Native Vulkan compute shaders would be even better. Why would you want unified address space?…

comment by Const-me · 2023-12-17 · in AMD's CDNA 3 Compute Architecture

…to port open source ML models from CUDA/Python to a better technology stack. I have ported Whisper https://github.com/Const-me/Whisper/ and Mistral https://github.com/Const-me/Cgml/ to D3D11. I don’t remember how much time I spent, but…

comment by Const-me · 2023-12-17 · in AMD's CDNA 3 Compute Architecture

…My D3D11 port of Whisper outperformed original CUDA-based implementation running on the same GPU: https://github.com/Const-me/Whisper/

comment by Const-me · 2023-12-15 · in Intel CEO: 'The entire industry is motivated to eliminate the CUDA market'

https://github.com/ggerganov/whisper.cpp https://github.com/Const-me/Whisper I had fun with both of these. They will both do realtime transcription. Bit you will have to download the training data sets…

comment by CrimsonCape · 2023-09-28 · in Amazon Bedrock Is Now Generally Available

…[3] https://github.com/jstoecker/dxcompute-docs/tree/main [4] https://github.com/Const-me/Whisper/blob/master/ComputeShade...

comment by Const-me · 2023-07-26 · in Which GPU(s) to Get for Deep Learning

…the Valve’s work on DXVK 2.0, modern Linux runs Windows D3D11 software just fine. Here’s an example https://github.com/Const-me/Whisper/issues/42 BTW, a lot of BLAS in the compute shaders of that software.

comment by Const-me · 2023-07-26 · in Which GPU(s) to Get for Deep Learning

…open-source project unrelated to CAE, where I managed to do just that with decent results: https://github.com/Const-me/Whisper That software even works on Linux, probably due to Valve’s work on DXVK 2.0 (a compatibility layer which…

comment by Const-me · 2023-07-07 · in Why Nvidia Keeps Winning

▲ 3

Ask HN: What is your recommended speech to text/audio transcription tool?

story by elektor · 2023-06-12 · 6 comments

…5 minutes to process each hour on a 1080ti GPU. There are a few wrappers available with GUI like https://github.com/Const-me/Whisper

comment by cloudking · 2023-05-29 · in Ask HN: Any recommendations for cheap, high-quality transcription software

…the GPU instead with no luck. Running it on WSL was probably not helping. Then I found this gem: https://github.com/Const-me/Whisper A C++ Windows implementation of Whisper. I opened the program, fed it with the largest model and the file.…

comment by can_center_divs · 2023-03-13 · in How 'Open' Is OpenAI, Really?

…the most expensive compute shader is the one which computes product of 2 matrices. nVidia version: https://github.com/Const-me/Whisper/blob/master/ComputeShade... AMD version: https://github.com/Const-me/Whisper/blob/master/ComputeShade...…

comment by Const-me · 2023-02-14 · in ChatGPT Hardware: A look at 8x NVIDIA A100 powering the tool

…My DirectCompute port of that uses slightly over 4GB of VRAM, see the last column in that table: https://github.com/Const-me/Whisper/blob/master/SampleClips/... I haven’t actually optimized memory usage I only did the DirectCompute port,…

comment by Const-me · 2023-02-14 · in ChatGPT Hardware: A look at 8x NVIDIA A100 powering the tool

And there's a fork of that that uses DirectCompute to run it on GPUs without Cuda on Windows: https://github.com/Const-me/Whisper

comment by sorenjan · 2023-02-13 · in Show HN: Self-host Whisper As a Service with GUI and queueing

…will probably do the trick. I have ported ML inference to Windows desktops with DirectCompute https://github.com/Const-me/Whisper These NVIDIA A40 GPUs mentioned on the page you’ve linked cost $4000 (Amazon) to $10000 (Dell), deliver…

comment by Const-me · 2023-02-11 · in The Inference Cost of Search Disruption – Large Language Model Cost Analysis

> All they will ever offer is a paid prompt. They sometimes open their older models: https://github.com/Const-me/Whisper

comment by Const-me · 2023-02-09 · in Andrej Karpathy is joining OpenAI again

…matrices. Relatively easy to do with a compute shader. Here's an example of such shader in D3D11: https://github.com/Const-me/Whisper/blob/master/ComputeShade...

comment by Const-me · 2023-02-07 · in Introduction to FPGAs

…somewhere else. Here’s a vendor-agnostic DirectCompute re-implementation of their Whisper model: https://github.com/Const-me/Whisper On servers, DirectCompute is not great ‘coz Windows server licenses are expensive. Still, I did that port…

comment by Const-me · 2023-02-01 · in LLM discourse needs more nuance

…library for .NET: https://github.com/Const-me/Vrmac Recently, offline speech-to-text for Windows: https://github.com/Const-me/Whisper At this point, I consider side projects like that as a hobby.

comment by Const-me · 2023-01-27 · in Ask HN: Those making $0/month or less on side projects – Show and tell

Here's my D3D11 implementation of speech-to-text https://github.com/Const-me/Whisper With medium model it needs 1.43 GB of assets, 2 GB of VRAM, and on gaming GPUs works at 10x realtime speed. These performance figures might be good enough…

comment by Const-me · 2023-01-22 · in Ask HN: ChatGPT et al. and NPCs

…That's a single combobox in project settings. Here's an example, just running the .exe works: https://github.com/Const-me/Whisper

comment by Const-me · 2023-01-22 · in GNU Octave

…additional old gen GPU that nails a good price/VRAM ratio Based on my observations from the tests https://github.com/Const-me/Whisper/blob/master/SampleClips/... and also this thread about 3060Ti, it looks like the library is indeed bound…

comment by Const-me · 2023-01-20 · in Show HN: Windows port of OpenAI's Whisper automatic speech recognition model

My implementation of Whisper uses slightly over 4GB VRAM running their large multilingual model: https://github.com/Const-me/Whisper

comment by Const-me · 2023-01-19 · in Which GPU(s) to get for deep learning

…model implementation is GPU. The other two implementations are disabled with macros, there: https://github.com/Const-me/Whisper/blob/1.1.0/Whisper/stdaf... These implementations are lacking some UX features like callbacks and cancellation,…

comment by Const-me · 2023-01-17 · in Show HN: Windows port of OpenAI's Whisper automatic speech recognition model

I have uploaded two sample clips: https://github.com/Const-me/Whisper/tree/master/SampleClips The text files in that folder contain performance data from two computers: a desktop with nVidia 1080Ti, and a laptop with integrated AMD GPU. If…

comment by Const-me · 2023-01-16 · in Show HN: Windows port of OpenAI's Whisper automatic speech recognition model

…running the model to transcribe the buffer with these audio samples. The logic is in that method: https://github.com/Const-me/Whisper/blob/15aea5bc/Whisper/Wh... How long is “several seconds” controlled by these user-adjustable parameters:…

comment by Const-me · 2023-01-16 · in Show HN: Windows port of OpenAI's Whisper automatic speech recognition model

▲ 43

Show HN: Windows port of OpenAI's Whisper automatic speech recognition model

story by Const-me · 2023-01-16 · 20 comments