AI & ML interests

Instruction residuals (task vectors) allow to continue pretraining LLMs while benefitting from SOTA instruction-following capabilities. This organization contains Residuals adapters for various popular open-weight models.

omarkamaliĀ 
posted an update about 4 hours ago
view post
Post
33
Picomon v0.2.0 released! šŸ’«

- Supports all of AMD, Nvidia and Apple Silicon šŸ§‘ā€šŸ§‘ā€šŸ§’ā€šŸ§’
- Beautiful TUI with themes (who said monitoring should be boring?) šŸ’…
- Shareable Rig Cards! Boast to friends, family and foes alike 🫨

Get it now! uvx picomon or pip install picomon then picomon
  • 1 reply
Ā·
omarkamaliĀ 
posted an update 20 days ago
view post
Post
3465
Hello picomon! AMD GPU Monitoring made easy

Just run uvx picomon and behold:
ā”Œā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”  ā”Œā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”
│ GPU 0  GFX  42%  UMC  21%                │  │ GPU 1  GFX  78%  UMC  66%                │
│ PWR 135/250W (54%)  VRAM 10.0/16.0GB 62% │  │ PWR 210/250W (84%)  VRAM 14.5/16.0GB 90% │
│                                          │  │                                          │
│ GFX ā–ā–‚ā–‚ā–ƒā–„ā–„ā–…ā–†ā–†ā–‡ā–ˆā–‡ā–†ā–…ā–„ā–ƒā–‚ā–                   │  │ GFX ā–‚ā–ƒā–„ā–…ā–†ā–‡ā–ˆā–ˆā–‡ā–†ā–…ā–„ā–‚ā–‚ā–ƒā–…ā–†                    │
│ PWR ā–ā–ā–‚ā–‚ā–ƒā–„ā–„ā–…ā–†ā–‡ā–ˆā–ˆā–‡ā–†ā–…ā–„ā–‚ā–                   │  │ PWR ā–‚ā–‚ā–ƒā–„ā–…ā–†ā–‡ā–ˆā–ˆā–‡ā–†ā–…ā–„ā–ƒā–‚ā–‚ā–ƒ                    │
│ VRM ā–ā–ā–‚ā–‚ā–ƒā–„ā–„ā–…ā–†ā–‡ā–ˆā–ˆā–ˆā–‡ā–†ā–…ā–„ā–‚                   │  │ VRM ā–‚ā–ƒā–„ā–…ā–†ā–†ā–‡ā–ˆā–ˆā–ˆā–‡ā–†ā–…ā–„ā–ƒā–‚ā–‚                    │
ā””ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”˜  ā””ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”€ā”˜


Repo at https://github.com/omarkamali/picomon
Or pypi at https://pypi.org/project/picomon
omarkamaliĀ 
posted an update 26 days ago
view post
Post
5216
Exciting updates to the Wikipedia Monthly dataset for November! šŸš€

惻 Fixed a bug to remove infobox leftovers and other wiki markers such as __TOC__
惻 New python package https://pypi.org/project/wikisets: a dataset builder with efficient sampling so you can combine the languages you want seamlessly for any date (ideal for pretraining data but works for any purpose)
惻 Moved the pipeline to a large server. Much higher costs but with better reliability and predictability (let me know if you'd like to sponsor this!).
惻 Dataset sizes are unfortunately missing for this month due to shenanigans with the migration, but should be back in December's update.

Check out the dataset:
omarkamali/wikipedia-monthly