Wojciech Prazuch

TypeError: 'NoneType' object cannot be interpreted as an integer

17

I tried to implement LSTM explainability using shap and used this[ notebook](https://github.com/slundberg/shap/blob/master/notebooks/deep_explainer/Keras%20LSTM%20for%20IMDB%20Sentiment%20Classification.ipynb) . However, in line: `shap_values = explainer.shap_values(x_test[:10])` I receive following error: `site-packages\shap\explainers\_deep\deep_tf.py in anon() 352 shape = list(self.model_inputs[i].shape)...

### Bug description `TransformerEnginePrecision.convert_module` function seems to not work for the the FSDP-wrapped model. ### What version are you seeing the problem on? master ### How to reproduce the bug...

RuntimeError: Attempted to call `variable.set_data(tensor)`, but `variable` and `tensor` have incompatible tensor type.

9

## 🐛 Bug There is an error when running fp8 for FSDP in `benchmark_litgpt.py` , which was also reported in the [pytorch-lightning repo](https://github.com/Lightning-AI/pytorch-lightning/issues/19989) ### To Reproduce Steps to reproduce the...

[NeVa] [rank0]: TypeError: matmul(): argument 'input' (position 1) must be Tensor, not TensorProxy

2

While preparing the benchmark for eager and dynamo using the code from the fork: https://github.com/tfogal/NeMo I get errors for dynamo case. ## 🐛 Bug After fixing [1187](https://github.com/Lightning-AI/lightning-thunder/issues/1187) NeMo NeVa dynamo...

[Regressions] ThunderFX is slower than 2 weeks ago for 2 models

2

## 🐛 Bug Recently found regressions:  ### To Reproduce All parameters to `benchmark_litgpt.py` are visible in the attached image. ### Environment Tested on `pjnl-20241122`...