Comments (9)
Hi! It's merged. Does pip install -U --pre triton
solve your problem?
from triton.
Hey! I am currently overhauling the entire build system to remove the build-time dependency on LLVM and PyTorch. I am also setting up some CI to automatically build nightly wheels that should just work. It should all be up in a few days. And I will be pushing Triton 1.0 on pip too when this happens.
from triton.
Great! Thanks for responding so quickly. I'm looking forward to it :)
from triton.
@ptillet It works for me! I am in a different problem though.
By using pip install triton
, setup cannot find llvm
ERROR: Command errored out with exit status 1:
command: /nas/longleaf/apps/python/3.7.9/bin/python3.7 -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-85ycizht/triton/setup.py'"'"'; __file__='"'"'/tmp/pip-install-85ycizht/triton/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(__file__);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' egg_info --egg-base /tmp/pip-pip-egg-info-qww7dwta
cwd: /tmp/pip-install-85ycizht/triton/
Complete output (7 lines):
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "/tmp/pip-install-85ycizht/triton/setup.py", line 98, in <module>
find_llvm()
File "/tmp/pip-install-85ycizht/triton/setup.py", line 28, in find_llvm
raise RuntimeError('Could not find llvm-config. ' + instructions)
RuntimeError: Could not find llvm-config. Please install llvm-{8, 9, 10}-dev
----------------------------------------
ERROR: Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.
pip install -U --pre triton
solves the problem.
from triton.
Yes, this is expected. The stable release on pypi will be updates to Triton 1.0 in a few weeks. Until then, installing the nightly build with pip install -U --pre triton
is the way to go.
The current stable version on pip (v0.3) is actually inconsistent with most of the current official documentation, so it is not recommended to use it unless you absolutely have to for the sake of interoperability.
from triton.
Yes it seems to work! Thanks for getting this out so fast. Sorry I took a bit to respond--I was afk for a week. Any idea on the timeline for updating the stable release?
from triton.
Checking back in here. Any idea idea of when the version of Pypi will be updated to 1.0.0?
Thanks!
from triton.
Hey!
Sorry for the delay. As you can see on the official documentation, there were major changes to the API, with depreciation of the C-like kernel language in favor of a @triton.jit python decorator (similar-ish to numba). This pushed things back a little. We want to make sure the Python API is really stable before we release 1.0.0
In the meantime, CUDA 11.3 came out; this comes with substantial compiler improvements that Triton cannot leverage at the moment. So CUTLASS is faster than Triton at the moment, and this doesn't quite satisfy our standard :p So this also pushed things back a little.
All in all, the new estimate is in a few weeks.
from triton.
Seems like the issue has been resolved. The recommended version for use with the C++-like kernel language is 0.4.2. All subsequent versions will use the Python API, starting with 1.0.0 which is just around the corner.
from triton.
Related Issues (20)
- How do I compile a Triton program? HOT 8
- Compile error with fp8 block pointer usage
- Jitting Error: can't pass bfloat16 as a tl.dype to my kernel! HOT 1
- Segmentation fault when DataLoader processes are launched after compiling Triton kernels HOT 6
- error when call tl.sort for data of torch.bfloat16 HOT 2
- Segfault on mixed mm HOT 4
- How to get the type of data from its ptr? HOT 1
- @LyricZhao Can you add a tutorial for the tile pointer? I think that could be helpful. HOT 1
- [Problem][Porposal] How does `make_block_ptr` work HOT 2
- FP8 conversion on SM89 (Ada Lovelace) fails with outdated PTX version HOT 4
- Initializing runtime driver overwrites root logger's verbosity level
- About Int8 Matrix Multiplication
- Why not allow JITFunction as parameter to another JITFunction(high-order jit function)?
- Add support for triangular solve operation.
- Inconsistency between constants as arguments and captured globals HOT 3
- The relationship between operator running time and number of runs HOT 2
- How to figure out layout/order for loads via explicit Tma? HOT 10
- Support for tl shift operator on tensor
- tl.clamp bug : AttributeError("module 'triton.language' has no attribute 'clamp'") HOT 4
- fp8 tensor core support on h100 HOT 2
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 triton.