-
For now it was disabled because of threading issues and memory corruption:
https://github.com/AvaloniaUI/Avalonia/pull/2615
It can be easily reproduced on Linux with GLX and requires some less tr…
-
```
load train state from /home/ubuntu/MyFiles/xiaodi/training/output/flux/sd-scripts/flux_lora_200k/flux_lora_200k-step00002250-state/train_state.json: {'current_epoch': 1, 'current_step': 2250}
[r…
-
### System Info
```Shell
- `Accelerate` version: 0.33.0
- Platform: Linux-5.15.133+-x86_64-with-glibc2.35
- `accelerate` bash location: /opt/conda/bin/accelerate
- Python version: 3.10.14
- Nu…
-
Hello, unlock-hf 伟大的贡献者们
我是一位 DataWhaler,我最近在学习 peft,发现 peft 框架中和 hf 的 TRF、Accelerate 实现深度集成。
我认为伟大的 unlock-hf 项目可以增加相关的介绍以及 demo,帮助更多的开发者理解和应用 PEFT 框架。如果我以后有空余时间的话也会参与贡献!
希望这个建议能够帮助仓库的内容更加完善!🤗🤗🤗
-
If we try to accelerate a tx that is already being accelerated, the backend returns an error and we don't display the invoice.
-
GPU acceleration for VRAD would help immensely when compiling maps, or even getting better lighting results.
This might help: https://github.com/darthryking/SilkRAD
-
Today I realised my mental model of how our Raft library works is subtly wrong.
I assumed that when a node restarts, it loads the most recent snapshot into the FSM _and replays committed logs_ from…
banks updated
4 weeks ago
-
I downloaded the log from Travis CI and searched for unit tests that take more than 10 seconds to run.
```
yi@WangYis-iMac:~/goyaccfmt (reformat_readme)$ grep '\---.*[0-9][0-9]\..*s)' /tmp/log.txt…
-
...PROFILE! it is almost inconceivable that the 3D drawing takes as long as the 4D transformations...
lene updated
10 years ago
-
yzshi updated
6 years ago