* dds layout now internally re-uses dsd code path for increased code
* at_mask and kp_mask related things are now dropped from the softmax API. I couldn't think of any case where it was needed beyond is_causal. And if there is any, we should probably find a way to get it implemented statically so that users don't have to materialize masks.
* fixed bug in blocksparse matmul that caused troubles when layout had a full row/col of zeros
* blocksparse softmax now no longer modifies any data in-place
* blocksparse softmax now takes an is_dense arguments that provides better performance. Passing is_dense=True, is_causal=True is the best way to achieve triangular attention.
* unit tests now test backward pass
I've been using this locally to find errors without running tests, and now that we're using autopep8, it passes with minimal suppressions. This is also what turned up the issues with the tutorials, which were fixed in #422.
- Fix meta-parameter usage on tutorials.
- Install tutorial dependencies on CI.
- Switch from `requirements-test.txt` to `extras_require` for test dependencies, and also use it for tutorial dependencies.
- Make some performance tests deterministic.
Run:
```
isort ./python
autopep8 -i --ignore E501,E701,E731 $(find ./python/ -name '*.py')
```
with an `.isort.cfg` and then clean up a few warts. This PR should be a no-op; the idea is that this is all boring whitespace changes, and any config file changes will be in a different change to make it easier to review.
A forthcoming PR will update the RNG to use these types.
Also:
- Add tests for the `//`, `<<`, and `>>` operators.
- Change `TensorWrapper` to unwrap objects when the resulting object would be simpler.
- Clean up `throw_unreachable`, since it was triggering compiler warnings.
Since numpy supports unsigned integers, and pytorch doesn't, this will make it easier to test unsigned integer support.
This adds an explicit requirement for numpy in tests, but we already required scipy, so it was already an implicit dependency.
- `BF16TyID` was missing a repr implementation.
- Throw a better exception on impossible casts.
- Add a few assertions. Tested with a debug build.
- Add `pointer_dtype.__str__` to aid kernel debugging.
Added an additional `repr` argument to the cache hook, which represents a human-readable string representation of the signature and argument attributes associated with the compiled binary.
Significantly improves the performance of `triton.ops.matmul` in memory-bound settings via the use of many more block configs coupled with a performance model to drive the auto-tuning process.
- Promote 16-bit floating-point `/` and `%` to 32-bit; we have to anyway.
- Do not force result of integer binary operations to be the LHS type. There used to be a bug in pytorch that did this, which Triton matched, but that bug is fixed now.
- When testing signed integer operations, use random numbers from the full range of the type.
- Add an optional `seed` argument to `triton.testing.random` so binary operations are not tested with both sides equal when the LHS and RHS have the same type.
- Fix a bad `CompilationError` invocation.
- Fix a warning suppression that causes tests to fail if you run them with `-W error` on python 3.8.
- Fix some whitespace.
- Make an undeclared dependency on `pytest` explicit.
- Fix deprecated `description-file` use.
- `#ifdef` out a deprecated `PyEval_InitThreads` call.
- Use a slightly different numpy invocation in `test_random.py` to quiet down overflow warnings in tests.
- Fix a deprecated cast in `test_core.py`.
- Suppress a warning about `visit_Constant` in Python 3.9+; we can't migrate yet because it'd break Python 3.6 and 3.7.
- Use chained exceptions for `CompilationError` rather than rolling our own; it makes the error messages nicer.
- Add a `__str__` for `tl.dtype` to make debugging kernels easier; it lets you `print` a dtype to see what type was inferred.
- Fix a few bad escapes.
- Removed driver module -- accelerator runtime is handled by pytorch
- Added basic support for ROCM based on @micmelesse 's PR -- now can execute empty kernel on AMD devices without any compile-time changes
- Now only using PREFER_SHARED for kernels when the size of shared memory is greater than 49k. Otherwise there can be poor L1 performance for broadcast tensors
This PR adds a BF16 data-type, along with FP32 <-> BF16 conversion instructions in the LLVM codegen. Other kinds of ops on bfloat16 are not yet supported.
Improved codegen for the Ampere GPUs.
* Make the layout pass recognize the multistage pipelined pattern.
* Now the pipeline pass can automate the multistage pipelining transformation.
* Remove extra barriers (from the prefetch pass & WAR) on Ampere.
* Update the code generator (generator.cc) to make Triton generate n-buffered shared memory loads/stores.
