-
```python
from thunder.benchmarks.targets import LitGPTConfig, LitGPTBenchmark, backward_only
import torch
import thunder
from thunder.executors.transformer_engineex import transformer_engine_ex
…
-
**Describe the bug**
Game crashed when Haschel was hit by Thunder Arrow. This could be an issue with Thunder Arrow itself, or it might be an issue with the status effect immunity animation since TA i…
-
*Note*: If you have a model or program that is not supported yet but should be, please use the program coverage template.
## 🐛 Bug
### To Reproduce
Steps to reproduce the behavior:
1. mo…
-
## 🐛 Bug
thunder's produces output with different dtype in compiled function
### To Reproduce
```
import thunder
import torch
def foo():
return torch.ones((1,), device="cuda")
jf…
-
Before the last update for war thunder the anti cheat would load up fine using proton 6.3 and all the newer ones. But since wind of change has come around the EAC will not load. It opens loads the bar…
-
## 🐛 Bug
`get_alloc_memory` has a potential to be a great tool for estimating the effect on memory usage after transforming traces. For those who don't know about this function it lives here:
http…
-
## 🐛 Bug
Before https://github.com/Lightning-AI/lightning-thunder/pull/705 Thunder applied autocast transform to all functions invoked within a torch.autocast region. In https://github.com/Lightn…
-
I think this can be a problem where part of the program is called with `thunder.jit` and it's output is used for non-compatible parts (eg. https://github.com/Lightning-AI/lightning-thunder/issues/461)…
-
## 🐛 Bug
There is not implemented error when running models:
Llama-3-70B
falcon-180B
longchat-13b-16k
CodeLlama-34b-hf
vicuna-7b-v1.5-16k
Mixtral-8x7B-v0.1
with `thunder_cudnn` and `th…
-
From the example below, the autocast is applied only for device cuda, however thunder.jit still applies it to CPU inputs.
```python
import thunder
import torch
def foo(x, w):
return torch…