Tal

Results 239 comments of Tal

The base example is bad (no ddp) ``` cd examples CUDA_VISIBLE_DEVICES=0,1 python training_wizardcoder_mydata.py --do_train --do_predict --num_epochs 1 --output_dir outputs-wizard --model_name WizardLM/WizardCoder-15B-V1.0 ``` and there are some discrepancies in the code,...

> Hi @mrT23 thanks for the PR - please can you add a test for this scenario I am not sure I have the technical understanding in litellm to do...

i am not sure you would like review to be run automatically every new commit. it can flood the git with a lot(!) of messages. my suggestion - use it...

Technically its possible, i agree. However, optional features still need to be viable. In practice, even as an option, would you really want 10-20 long review feedback messages in a...

What do you mean ? can you give a concrete use-case for using the [saved replies feature](https://docs.github.com/en/get-started/writing-on-github/working-with-saved-replies) ?

Thanks @lukefx for the PR. @okotek is on vacation, he will get to review it later this week

**[PR Description](https://github.com/Codium-ai/pr-agent/pull/557)** updated to latest commit (https://github.com/Codium-ai/pr-agent/commit/183e722bf76114f18b0ec80391f19a9f9b194f57)

## PR Review ⏱️ Estimated effort to review [1-5] 3, because the PR involves significant changes in dependency management and Docker configurations, which requires a thorough review of configurations and understanding of the Poetry...