-
I have this error:
```
outputs = self.base_model.generate(**kwargs)
File "/home/administrator/.local/lib/python3.8/site-packages/torch/autograd/grad_mode.py", line 27, in decorate_context
…
-
之前在v100 32G*4 上也报OOM,然后换到了一台A100 80G*1 就正常跑通了。
现在v100 32G*4上会报标题的错误。而两台机器,8张v100 32G,用scripts/multinode_run.sh 还报这个错误,请问是显存不够的原因吗?
没有打印其他日志
```
[INFO|modeling_utils.py:2263] 2023-05-29 11:08:21,…
-
With transformers 4.41.0., Ubuntu 22.0
Calling the training script scripts/train_jat_tokenized.py as given (with --per_device_train_batch_size 1 and one GPU) the following error comes when the syst…
-
View details in Rollbar: [https://rollbar.com/Azavea/MMW/items/206/](https://rollbar.com/Azavea/MMW/items/206/)
```
Traceback (most recent call last):
File "/usr/local/lib/python2.7/dist-packages/c…
-
Hey there, I can't find the source code for creating the './helm/hd/{model_name}/hd_act.json' files which are part of the HELM dataset. Although you haven't used them at the end, I'd like to know how …
-
### System Info
```shell
Run on a general Google colab notbook with the following installations:
!pip install --quiet transformers
!python -m pip install git+https://github.com/huggingface/opt…
-
When I run this code:
predictor = Predictor.from_path("https://s3-us-west-2.amazonaws.com/allennlp/models/bert-base-srl-2019.06.17.tar.gz", cuda_device=opts.cuda_device)
Send an error:
Traceback (m…
-
### 问题确认 Search before asking
- [X] 我已经查询[历史issue](https://github.com/PaddlePaddle/PaddleDetection/issues),没有发现相似的bug。I have searched the [issues](https://github.com/PaddlePaddle/PaddleDetection/issu…
-
When the adminmetadata template was added as a blank node to the instance & work templates, catalogers were very happy (they can now mostly ignore it), but it has caused issues with the local adminmet…
-
When the adminmetadata template was added as a blank node to the instance & work templates, catalogers were very happy (they can now mostly ignore it), but it has caused issues with the local adminmet…