-
The prompt generation code is batching prompts into batches
The loop that sends those batches only seems to be using 1 prompt from each batch
It doesn't seem correct as most of the instructions in…
-
### Feature request
optimium currently limits transformers to `>= 4.38.0, < 4.39.0`. @regisss bumped the upper version limit in PR #1851 a month ago. Is there any technical reason to limit the upper …
tiran updated
4 months ago
-
As a follow-up to #968, I also discovered that documents with very long words (such as 2000 characters in a row with no newlines or spaces in between) do not get chunked properly.
The fix in #969 w…
-
`ilab` has a `generate` subcommand, but its implementation is independent from that of the backend pipeline (mostly here - https://github.com/instruct-lab/datagen-pipeline).
Further, I keep hearing…
-
We are generating these 3 files
```
1. generated_Mixtral-8x7B-Instruct-v0_2024-07-08T18_23_10.json
2. train_Mixtral-8x7B-Instruct-v0_2024-07-08T18_23_10.jsonl
3. test_Mixtral-8x7B-Instruct-v0_202…
-
I made changes to a MD file and found that the CI pipeline failed because of the Markdown linter. I didn't know how to run this locally on my machine which would make it easier to correct these lintin…
-
Costin Eugen Gament would like to access one of the 4 GPUs that we initially set up for use of folks working with Jeremy, His github name is kwozyman.
Jeremy says he is not using the GPUs now, s…
-
The PR https://github.com/instructlab/taxonomy/pull/1148 is throwing a 500 from SDG. Raised the issue with Raymond and it's being tshot on the server side. cc/ @jjasghar
```
cmd/generate.go:683…
-
### Is your enhancement related to a problem? Please describe
When I'm using AI Lab, I've got some imported GGUF model files from InstructLab, but sometimes need to delete them from the catalog. It's…
-
We don't have instructions on how to upgrade in the cli repo readme nor in the changelogs. It might be a good idea to have a thing in both places.
`pip install instructlab --upgrade` worked for me,…