Riccardo Felluga
Riccardo Felluga
On windows, the cast to `long` from the type `npy_intp` returned by [`PyArray_NDIMS`](https://numpy.org/doc/stable/reference/c-api/array.html?highlight=pyarray_ndims#c.PyArray_DIMS) results in incorrect parsing of numpy array dimensions. I propose the [use of `intptr_t`](https://stackoverflow.com/a/6326362) as a solution...
Before submitting - [x] Was this discussed/approved via a Github issue? (no need for typos and docs improvements) - [x] Did you read the [contributor guideline](https://github.com/Lightning-AI/pytorch-lightning/blob/main/.github/CONTRIBUTING.md), Pull Request section? -...
Before submitting - [ ] Was this discussed/approved via a Github issue? (no need for typos and docs improvements) - [x ] Did you read the [contributor guideline](https://github.com/Lightning-AI/pytorch-lightning/blob/main/.github/CONTRIBUTING.md), Pull Request...
## 🚀 Feature For debugging purposes, I would like to be able to quickly retrieve input information for a fusion. ### Motivation When debugging performance issues an important aspect is...
## 🐛 Bug It seems that when passing a `torch.dtype` as argument of a jitted function, Thunder caches the value and the resulting rerun of the function outputs wrong values....
## 🚀 Feature Mixtral 8x7B is a mixture-of-experts LLM that splits the parameters in 8 distinct groups an I would like to do both training and inference with Thunder. ###...
## 🚀 Feature I would like to have automated benchmarks for selected models to allow for performance tracking. ### Work items Automated benchmarking in this context means two things: -...
## 🚀 Feature I would like Thunder to support [`torch.nonzero`](https://pytorch.org/docs/stable/generated/torch.nonzero.html) as it can be used in Mixtral instead of `torch.where`. ### Pitch Support `torch.nonzero` as per https://pytorch.org/docs/stable/generated/torch.nonzero.html ### Additional context...
Before submitting - [x] Was this discussed/approved via a Github issue? (no need for typos and docs improvements) - [x] Did you read the [contributor guideline](https://github.com/Lightning-AI/pytorch-lightning/blob/main/.github/CONTRIBUTING.md), Pull Request section? -...
## What does this PR do? After discussing around some ways to implement a partial fix for #183 and #224, this option seems like a good compromise for the moment....