Comments (1)
Hey @asfiyab-nvidia, thanks so much for giving this a try!
I think the issues here are tracked by #455 and #457, which are more granular, so I'm going to close this in favor of those issues. You might also be interested in the more general #266.
But we're very interested in Stable Diffusion! If you hit any issues after one or both of those are fixed then please open new issues (ideally one issue per subnet). Thanks!
from lightning-thunder.
Related Issues (20)
- thunder/tests/test_nvfuser_remat.py::test_find_cut_dropout_nvfuser_cuda_None is over sensitive. HOT 7
- CSE should always modify subsymbols
- jit : if torch.device is an input to jitted fn, it's value is not checked in prologue.
- TransformerEngine: Communication and Computation is not overlapped in backward pass with FSDP Zero3 HOT 5
- [distributed] Enable transformed modules to load state dicts of the originals HOT 2
- reenable testing cudnn SDPA with PyTorch dev version / 2.4.0a0+ HOT 2
- Returning `dtype` or `device` from jitted function returns thunder's dtype or device (not torch.{dtype/device}). HOT 2
- Find a way to properly sort the communication operators for zero2/zero3
- `column_parallel` / `row_parallel` fail to transform `litgpt.model.CausalSelfAttention` with `TypeError: unshable type: 'TensorProxy'`
- nvFuser executor doesn't support prims.sum with symbolic dimensions HOT 2
- NotImplementedError: requires_grad=True is not yet supported within thunder.compile HOT 3
- CUDA error: CUDA_ERROR_ILLEGAL_ADDRESS failed when training falcon-7b HOT 4
- Unsupported – setattr(FSDPManagedNNModuleVariable(FullyShardedDataParallel), _is_root, ...)
- An error occurred: KeyError – 't5479' / HOT 4
- [Tensor Parallelism] Improve comm optimization logic for pair of column-wise parallel linear and row-wise parallel linear
- unhashable type: 'TensorProxy' error on NeVA model
- jit error: unpacking from nonconstant opaque function
- randn meta function should be able to accept shape of type list
- Add nvFuser support for thunder.torch.randn
- Revise memory clearing mechanism in the torch.autograd.Function integration
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from lightning-thunder.