Rejnald Lleshi comments

Results 74 comments of


                                            Rejnald Lleshi

high number of concurrent requests causes MLServer microservice to timeout

Yep, the CPU cycles are getting throttled whenever there is an influx of requests. I am running a worker for each core.

WSL Ubuntu - " Build Error cp: cannot stat 'CMakeFiles/ggml.dir/ggml-cuda.cu.o' "

I can second this error on Ubuntu. The error message points to a potential version mismatch between the Cuda toolkit and Nvidia drivers though. I'm also wondering if it could...

phi 3 error

@arnavgarg1 I'm also running into this issue. Any updates?

Error running inference on Llama3 model

Same issue. @arnavgarg1 kind ping to perhaps take a look on this too Env Ludwig==0.9.1 Transformers==4.42.3 peft==0.11.1 Python 3.8.0

get timestamps in split on silence

This is currently doable with `pydub.silence.detect_nonsilent` if anyone is seeking a solution with the current version.

Is there a way to change language for LLama 2?

Best to use [LeoLM ](https://laion.ai/blog/leo-lm/). This is llama-2 trained on high-quality German corpuses.

[Feature] messaging users

Great, ty for the tip!

MMpose integration

Ok, ty! Any rough timeline in mind?

`poetry add` does not support paths without a directory component

Getting the same error on Windows. @will-tha-mise did you find a work around for this?

`poetry add` does not support paths without a directory component

@will-tha-mise had u installed that python using pyenv?