-
## š Bug
Splitter for Thunder as a Dynamo backend should send regions of code under a "no_grad" context manager to Thunder. Currently, it chooses to send these computations to Inductor.
```py
fā¦
-
### š Documentation
The current lightning documentation is in dark mode https://lightning.ai/docs/pytorch/stable/.
Would it be possible to switch from dark to light mode (or vice versa) ?
It shoulā¦
-
*Note*: If you have a model or program that is not supported yet but should be, please use the program coverage template.
## š Bug
If traces include one or more `BoundSymbol`s of `torch.autogrā¦
-
It gets Network error, peer closed connection or connection reset by peer when trying Yi Lightning model on lmarena.ai.
-
### Bug description
I am trying to train a GAN model on multiple GPUs using DDP. I followed the tutorial at https://lightning.ai/docs/pytorch/stable/notebooks/lightning_examples/basic-gan.html, changā¦
-
### š Documentation
There's a lot of documentation out there about using the `resume_from_checkpoint` keyword in a pytorch trainer however this is wrong. In the latest pytorch version, one needs to pā¦
-
I am getting the following error:
> llama distribution install --spec local --name local-llama-8b
```
Using local-llama-8b as the Conda environment for this distribution
Conda environment 'locā¦
-
### Checklist
- [ ] The issue exists after disabling all extensions
- [X] The issue exists on a clean installation of webui
- [ ] The issue is caused by an extension, but I believe it is caused bā¦
-
## š Bug
We need to determine whether Thunder has real accuracy problems computing HF's Qwen 2 model.
The test added in https://github.com/Lightning-AI/lightning-thunder/pull/1406 might fail becā¦
-
Crosspost of https://github.com/Lightning-AI/lightning/issues/13381
pip install panel==0.13.1 bokeh==2.4.3 lightning==2022.6.15.post2 holoviews==1.14.9
-----------
I'm trying to create a Paneā¦