-
The following script throws cuda error - CUDA error: invalid device function
(when swapping to nonautoregressive_transformer, there is no CUDA error)
export CUDA_VISIBLE_DEVICES=0
fairseq-trai…
-
My SMTP URL is `smtp://AKIA:@email-smtp.us-west-2.amazonaws.com:465`. Attempting to send a test email gives an Internal Server Error with the following stack trace in the logs:
```
Exception while in…
-
Allow multiple makelunch instances to exist concurrently, with users able to correspond to an eater in several different instances.
Rather than filtering data at the app level, we should have one cen…
-
Hi there! I hope you are having a nice day.
I have been trying to use @rocket.chat/sdk in my React app.
### SERVER DETAILS
Deployment
Version
3.15.0
Deployment ID
RFJ4fpYQA2vC2cNLA
Apps …
-
# [rfc] Trigger callback when backwards begins for DDP with custom autograd function
with @zhaojuanmao
*Context*
A variety of improvements to `DistributedDataParallel`, namely around ensurin…
-
@fangwei123456
**Issue type**
- [x] Bug Report
- [ ] Feature Request
- [ ] Help wanted
- [ ] Other
**SpikingJelly version**
`0.0.0.0.14`
**Description**
使用 pytorch DDP 单机多卡训练时,无法使用…
-
Torch/torch-ccl/ipex version 1.13.0
cluster node: 2
World_size: 2
All nodes have password-less connections set, and mpirun works well as the readme says:
```
mpirun -f ./hosts -n 2 -ppn 1 -genv O…
-
### 🚀 The feature, motivation and pitch
I did not see an example for this, but I am trying to run a PyG model using pytorch DDP with nccl as in this example
[multigpu](https://github.com/pytorch/exa…
-
Tasmota supports art-net, but this project doesn't seem to. Art-net would be super useful for using smart bulbs in installation art, for example
-
Hi there,
Great work with dMoE! I'm trying to test dMoE with regular DDP + pytorch AMP(BF16) and I get the following error:
```bash
optimizer_state["found_inf_per_device"] = self._unscale_…