ceos/unsloth
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Update README.md

Browse source
This commit is contained in:
Daniel Han 2023-12-01 13:49:53 +11:00 • committed by GitHub
commit f383e5dbd7
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 2 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

5
README.md
View file

@ -1,13 +1,12 @@
<img src="./images/unsloth new logo.png" width="400" />
## 50% faster 50% less memory LLM local finetuning
## 5X faster 50% less memory LLM local finetuning
* Manual autograd engine.
* All kernels written in OpenAI's Triton language.
* 0% loss in accuracy.
* No change of hardware necessary. Supports Tesla T4, RTX 20, 30, 40 series, A100, H100s
* Flash Attention
* Train Alpaca **fully locally in 13 hours from 23 hours.**
* Train Open Assistant **fully locally in 4 hours from 8 hours.**
* Train Slim Orca **fully locally in 260 hours from 1301 hours (54 days).**
* Check out [Unsloth Pro and Max](https://unsloth.ai/) codepaths for **30x faster training**!
<div class="align-center">