DavidGOrtega

Results 267 comments of DavidGOrtega

@CharlieFRuan 🙏 > I personally do not know a < 3B model that supports >=8K context window size, but I may be wrong. Im working right now with 32K 7B...

Due to myriads of work I have not been able to review this lately (or try BNF!). Last time I was working on this, for some reason web-llm is not...

@CharlieFRuan I can help with testing. I have been doing it with Cypress functions using web-llm. Let me prepare it tomorrow reusing my code

WebGL is not worth at all for simple reasons like culling. WEBGPU is the way to go and will be greater when we move out the webassembly 4GB cage. Totally...

> WebGPU isn't supported by most browsers Can you name a few? Chromium based browsers supports WebGPU right now. To name a few Chrome, Opera, Edge and of course and...

@peq42 stats improves when you watch desktop, for mobile you have to go straight with TVM mobile. And this discussion might be a bit of futile, not because Webgpu is...

Not sure If you needed the screenshot of the error or the model working, Im guessing the latest so here it is. It is pretty amazing, just only a small...

:wave: @mikolajpabiszczak the reason is because the runner has been marked to do just one job with the parameter `--single` the option that you might be looking for is `--reuse`

@mikolajpabiszczak You have in your logs ``` Jul 22 11:37:30 ip-172-31-32-70 cml.sh[2440]: {"level":"error","message":"HttpError: Resource not accessible by integration","stack":"Error: HttpError: Resource not accessible by integration\n at process. (/snapshot/cml/bin/cml/runner.js:333:32)\n at process.emit (node:events:539:35)\n...

Just to be sure and move one step forward can you please your `REPO_TOKEN`? Does it have all all the permissions?