MiniMaxAI/MiniMax-Text-01
Text Generation
•
Updated
•
2.38k
•
406
Valeriy Selitskiy
WaveCut
AI & ML interests
Looking to switch from hobby to career
Recent Activity
liked
a model
5 days ago
MiniMaxAI/MiniMax-Text-01
liked
a model
6 days ago
nvidia/Cosmos-1.0-Diffusion-7B-Text2World
upvoted
a
collection
6 days ago
HIGGS
Organizations
WaveCut's activity
liked
a
model
5 days ago
liked
a
model
6 days ago
upvoted
a
collection
6 days ago
updated
a
model
6 days ago
Update README.md
#1 opened 6 days ago
by
WaveCut
liked
a
model
6 days ago
liked
a
model
9 days ago
liked
a
Space
9 days ago
liked
a
model
14 days ago
liked
a
Space
14 days ago
liked
2
models
15 days ago
liked
2
Spaces
15 days ago
liked
a
model
15 days ago
liked
2
Spaces
16 days ago
upvoted
a
collection
16 days ago
reacted to
grimjim's
post with 👍
16 days ago
Post
2691
I'm (finally) releasing a Python script that trims excess weights in Gemma2 full-weight models that bloated by ~1B parameters due to an early mergekit bug.
https://github.com/jim-plus/Gemma2-mergekit-remediation
I'd noticed something was off when merges of Gemma2 9B models ended up having ~10B parameters. The current mergekit package is fine, but there are still bloated models on HF that could stand to be fixed.
The script assumes that it will be run from the same directory as the model weights, and will trim the unnecessary lm_head.weight tensor and corresponding index entry.
https://github.com/jim-plus/Gemma2-mergekit-remediation
I'd noticed something was off when merges of Gemma2 9B models ended up having ~10B parameters. The current mergekit package is fine, but there are still bloated models on HF that could stand to be fixed.
The script assumes that it will be run from the same directory as the model weights, and will trim the unnecessary lm_head.weight tensor and corresponding index entry.
- 2 replies
liked
a
model
16 days ago
