-
**Describe the bug/problem**
- From an engineering perspective, it's too much information if someone wants to add a skill or knowledge file. The documentation is very good but geared towar…
-
the `instructlab/sdg` is introducing more extensive data generation pipelines. These pipelines are defined in a format that's possible to express as configuration. To allow some customizations, we sho…
-
Download yaml feature download the yaml and name it "knowledge_qna.yaml" or "skill_qna.yaml". User needs to rename these files to qna.yaml to submit the PR to taxonomy. It would be better to rename th…
-
Add release automation for publishing to PyPI. We can copy most of this from `instructlab/instructlab`.
-
https://github.com/instructlab/dev-docs?tab=readme-ov-file#governance isn't very clear in it's examples - the wording here should be updated
-
I developed this yaml file originally for a PR:
```
task_description: |
The rot13 task enables converting from rot13 to plaintext.
seed_examples:
- question: Decode \"uryyb\" from rot13.
…
-
When mt_bench gets an error in an openai call, the current logic is to retry a few times before returning $ERROR and a score of -1. I am not sure what lm_eval is doing for mmlu yet. The -1 score is …
-
InstructLab 0.13 supports hardware acceleration for Apple Silicon (via `mlx`) and CUDA-like GPUs (NVIDIA CUDA and AMD ROCm via `torch.cuda`). I would like to add support for Intel Gaudi 2 hardware and…
tiran updated
3 months ago
-
vLLM emits the error:
`ERROR 07-05 19:50:10 serving_chat.py:225] Error in applying chat template from request: Conversation roles must alternate user/assistant/user/assistant/...
INFO: ::1:394…
-
The LLM models are trained and released as a version today for usage. We can also contribute through skills and knowledge. But what we see is if the data contains noise then it will get into the train…