-
Version: Deno 1.45.2
I'm applying this guide: https://docs.deno.com/runtime/tutorials/aws_lambda/
I've got some permissions issue in the container. I can see what kind of problem this is, but wh…
-
If we can cache them, we can get rid of of some nasty latencies introduced by #706, which is important for a smooth user experience.
-
## Description
As I know, the optimizer decides the num_update according to its _index_update_count saved on each device, which means that If the trainer states on one GPU device and loaded into anot…
-
### Describe the bug
Hi there. I've reliably used the train_controlnet_sdxl.py on single gpu on GCP (A100 - 40 GB). I have had to switch to AWS and am presently using a p3.8xlarge which has 4 V100 …
-
### System Info / 系統信息
Transformer 4.43, 4.44, 4.33 都试了，modeling_chatglm.py 也替换了，运行最后的 .sh 文件是报了和其他人类似的错。
建议官方再把训练操作过程写的详细些。
### Who can help? / 谁可以帮助到您？
。
### Information / 问题信息
- [X] The offic…
-
-
# Summary
在报Bug前已尝试通过安装最新Nightly（oneflow-0.9.1.dev20230307+cu117）解决，但Bug仍出现
## Bug0: 开关无效
``` python
import oneflow as flow
flow.set_autocast_enabled = True
print(f"flow.is_autocast_enabled=…
-
## Type of issue
- Thanks guys for this awesome work. I was curious to run llama3-8B on my personal CPU, and the performance is quite impressive (nearly 2x llama.cpp for same model size on same HW).
…
-
目前看文档好像只有coco、imagenet等官方数据集，数据集太大了，现在没法进行测试，自己的数据集也没法加载
-
I have a config which was running on Kaggle fine in previous versions
Right now it is failing on 15 GB gpu
This should not happen
Same settings on OneTrainer uses lesser than 13.5 GB VRAM
…