-
### 🐛 Describe the bug
Enable cpu offload and save FSDP model dict with local state dictionary fail with below error:
```
Traceback (most recent call last):
File "train_llama_fsdp_datasets.py", …
-
### 🐛 Describe the bug
I ran into a problem while permuting the following tensor (to convert into a complex dtype):
```python
>>> torch.view_as_complex(torch.empty(1,0,2,100,100).permute(0,1,3,…
-
## 🐛 Bug
Hi, I'm trying to download wikitext datasets using torchtext APIs but facing with an error:
```
---------------------------------------------------------------------------
HTTPError…
-
### 🐛 Describe the bug
My code is the following:
```
resnet50 = torch.hub.load('NVIDIA/DeepLearningExamples:torchhub', 'nvidia_resnet50', pretrained=True)
utils = torch.hub.load('NVIDIA/DeepLear…
-
### 😵 Describe the installation problem
Building from the master on powerpc64 Iget the following error from:
```console
$ pip install git+https://github.com/pyg-team/pyg-lib.git --config-setting…
-
### 🐛 Describe the bug
`torch.where` function seems to deviate from `np.where` when it comes to handling the dtype of the condition used.
For instance, `torch.where` only accepts a tensor of dtype…
-
When i run whisperx in google colab it always says ,
UnboundLocalError: local variable 'word_start' referenced before assignment.
My command was..
![Screenshot_20230502-051355](https://user-i…
-
### 🐛 Describe the bug
I was using `torchtext` to construct a vocabulary (v) and tokenize the input text. I just wanted to check the indices of the tokens and therefore used the method `v.vocab.loo…
-
Seems like torch is already at 2.0 , which was the reason for the --upgrade flag mentioned in the description. Is the --upgrade flag still needed?
If I do this instead: !pip3 install torch
I get…
-
您好,我将zero数据处理后约7000w,存储到lmdb文件,但是lmdb仅能生成一个data.mdb,存放在一个device上,导致我训练模型时数据读取io瓶颈,CPU与GPU利用不充分,请问有遇到这个问题吗?