-
Define processes for decision making within the oneCCL project. This should include a RFC process for new design and feature proposals
(see oneDNN for example - PR tagged RFC that links to the propos…
-
- Link to the UXL Foundation Slack workspace and channel from the project README
- Assign team member(s) to monitor the Slack channel
- Set up notifications in the oneCCL Slack channel when new RFCs …
-
### Describe the issue
Hi IPEX team,
I have an application where I want to serve multiple models concurrently, and I want to share weights across concurrent instances. I normally do this with `tor…
-
Hello there,
I wanted to preface this by saying that I am running in a Conda environment for python 3.9.18 within Ubuntu 23.10 inside of visual studio code, the command line for attempting inferenc…
-
### System Info
```Shell
Ubuntu 22.04.3 LTS
Intel OneAPI 2024.1
python 3.11.9
accelerate 0.31.0
transformers 4.42.3
numpy 1.26.4
```
### Information
- [ ] The official example scripts
- [X] M…
-
Update the project pages with the UXL Foundation links, logo and identity.
-
### Describe the bug
I found in jupyter notebook, `to('xpu')` makes the Jupyter kernel die.
### Notebook to reproduce
![image](https://github.com/intel/intel-extension-for-pytorch/assets/105281…
-
problem:
run command bash run_benchmark.sh -m llama-2-13b -d bf16 -s 2 -bs 1 -in 2048 -out 2 -w 10 -I 1
meet Segmentation fault (core dumped) FIRST_TOKEN_WEIGHT_LOCATION=$1 NEXT_TOKEN_WEIGHT_…
-
## Summary
Calling `torch.distributed.all_gather()` when using the `'ccl'` backend results in a SEGFAULT if the tensors being gathered are larger than a few megabytes.
This problem also seems to…
-
I have installed **intel-extension-for-transformer** using `pip install intel-extension-for-transformers` but trying a little script to see if it worked I got this error :
Traceback (most recent c…