Comments (3)
from pytorch.
hi @rosario-purple , https://github.com/pytorch/FBGEMM/blob/d17fc49978aeb3f1516938838ab0b2616275cfc5/fbgemm_gpu/experimental/gen_ai/bench/ck_fp8_bench.py is hardcoded to use the torch.float8_e4m3fnuz
dtype which is not supported on NVIDIA hardware (AMD is one type of hardware which supports this dtype), and you are running on NVIDIA H100 hardware. You could try updating the benchmark script to be configurable between torch.float8_e4m3fnuz
and torch.float8_e4m3fn
.
from pytorch.
closing since this is working as expected, I think the fbgemm repo would be a good place to track the resolution
from pytorch.
Related Issues (20)
- partitioner doesn't appear to respect SAC region
- Missing support for deterministic algorithm for torch.float8_e4m3fn and torch.float8_e5m2 HOT 3
- Flaky test page should include retry runs
- _foreach_vector_norm HOT 2
- Unable to assign `nn.Parameter(DTensor)` (created outside of compile region) to an nn.Module param attribute during Dynamo tracing HOT 3
- DSD for TorchTune LoRA
- Outdated ncclResult code HOT 2
- ONNX export for gelu at version 20
- Link https://pytorch.org/docs/stable/nn.html#torch.nn.EmbeddingBag may not exist anymore HOT 1
- torch 2.5.0.dev failing with tons of IMA issues
- False INTERNAL ASSERT FAILED bug whilst training Neural Network HOT 1
- The name of the function `nn.L1Loss()` should be `nn.MAE()` or the name of the function `MSELoss()` should be `nn.L2Loss()`
- Any plans for a "torch.minmax" (min-max normalization) function? HOT 1
- Assign `torch.Generator` in APIs like `torch.randn_like()`
- CUDA error in torch.cdist with compute_mode=donot_use_mm_for_euclid_dist HOT 1
- GradType: a subset of dtype that is differentiable, containing all float and complex dtypes HOT 2
- add x/0 gradient behaviour to documentation
- AssertionError: <class 'torch._dynamo.variables.torch.TorchInGraphFunctionVariable'> when compiling `torch.nn.functional.layer_norm`
- `RuntimeError: invalid dtype for bias - should match query's dtype` when using torch.compile + FSDP + hf transformer
- [C10d]: Work state in dump trace file is not deterministic. HOT 1
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 pytorch.