[CK][CK TILE] Dispatcher kernel selection heuristic for
grouped conv (#6327)
## Motivation
The ML heuristic in dispatcher does not support grouped-conv operator
yet. In this PR, the support for fwd, bdw-data, and bwd-weight
grouped-conv kernels have been added. A tile_engine utility has also
been added to compile and run any selected kernel configuration through
dispatcher infrastructure.
## Technical Details
1. Tile engine utility is added to benchmark each shape with all the
possible kernel+tile_size combinations here -
[https://github.com/ROCm/rocm-libraries/blob/users/yraparti/ck/dispatcher-grouped-conv-heuristics/projects/composablekernel/tile_engine/ops/grouped_conv/grouped_conv_full_benchmark.py](url)
2. New LGBM regressor models for grouped conv are added to models
directory. We have 3 separate models for fwd, bwd-data, and bwd-weights
[https://github.com/ROCm/rocm-libraries/tree/users/yraparti/ck/dispatcher-grouped-conv-heuristics/projects/composablekernel/dispatcher/heuristics/models](url)
3. Implemented lazy GPU initialization (dispatcher/python)
- **Issue**: ProcessPoolExecutor fork() + GPU context caused memory
access faults
- **Solution**: Mirror FMHA pattern - defer GPU initialization until
first run()
- **Changes**:
- setup_multiple_grouped_conv_dispatchers() returns List[Path], not
loaded libs
- GpuGroupedConvRunner.__init__() no longer calls ctypes.CDLL
- Added _ensure_initialized() method for lazy GPU loading
- GPU context created only on first run() call
- **Benefit**: Parallel compilation now works without GPU conflicts
4. Addressed few miscellaneous issues such as:
- Fixed BF16->FP16 naming bug in the dispatcher wrapper
- Added new tile sizes, and comp_v5 pipeline to the arch spec to expand
the kernel selection
- Added automatic padding support for unsupported shapes in dispatcher
runner
- Created a single source of truth between tile_engine and dispatcher
about the architecture and tile_size details
- Build a validation scripts to compare oracle_best vs ml_heuristic
comparison
## Test Plan
1. Validated fwd, bwd-data, and bwd-weight kernels with both known and
unseen data sets with up to 300 problems.
2. Ensured that test cases are added in both dispatcher and tile_engine
to validate the heuristic.
## Test Result
Results on Unseen shapes validated on gfx950
#### Forward Pass Model
- **Training Data**: 48,845 measurements across 1,372 unique problem
shapes
- **Validation Set**: 300 unseen problems from model crawler
- **Validation Performance** (vs. oracle):
- Mean Efficiency: **93.05%**
- Median Efficiency: **96.8%**
- P10 Efficiency: **79.9%**
#### Backward Data Gradient (bwd_data) Model
- **Training Data**: 18,773 measurements across 891 unique problem
shapes
- **Validation Set**: 300 unseen problems from model crawler
- **Validation Performance** (vs. oracle):
- Mean Efficiency: **93.8%**
- Median Efficiency: **96.5%**
- P10 Efficiency: **82.9%**
#### Backward Weight Gradient (bwd_weight) Model
- **Training Data**: 34,900 measurements across 1,508 unique problem
shapes
- **Validation Set**: 300 unseen problems from model crawler
- **Validation Performance** (vs. oracle):
- Mean Efficiency: **96.1%**
- Median Efficiency: **99.2%**
- P10 Efficiency: **89.4%**
## Submission Checklist
- [ x] Look over the contributing guidelines at
https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.
[CK] [CK_Tile] Add GroupConv to Kernel Dispatcher
## Motivation
This PR adds CK Tile group convolution (forward, backward-data,
backward-weight) support to the kernel dispatcher, matching and unifying
with the existing dispatcher GEMM infrastructure in architecture and
usability. The dispatcher provides a unified kernel dispatch system with
both C++ and Python frontends, and until now only supported GEMM
operations. This PR enables framework integrators to use the same
declarative kernel workflow for convolutions as they do for GEMM:
declare kernels, build a registry JIT, select kernels within the
registry at runtime, and dispatch to GPU. Future PRs will include
runtime kernel selection heuristics for autotuning of kernel parameters
based on (problem, hardware arch).
## Technical Details
Grouped convolution support has been added to the CK Tile Dispatcher
with generated_conv_backend.hpp enabling dispatcher.run(in, wei, out,
problem) for all 6 conv variants (fwd/bwdd/bwdw x 2D/3D), runtime
heuristic kernel selection, and GroupedConvKernelKey with full
ConvConfigBase fields. Python side adds parallel JIT via
registry.build(max_workers) and heuristic registry.select(). Includes 7
C++ and 6 Python examples covering all directions with CPU reference
validation, and shared infrastructure improvements (BaseRegistry CRTP,
structured exceptions). As a sanity check, JIT compile times for a
single kernel remains the same and for multiple kernels there is better
parallelism:
Kernels | 1 worker | 8 workers
1 | 7.7 s | 7.7 s
2 | 15.9 s | 8.2 s
4 | 33.4 s | 9.7 s
6 | 52.3 s | 10.2 s
## Test Plan
145 ephemeral unit tests have been added to test basic functionality.
All 30 examples/integration tests run end-to-end on gfx950 (MI350): 7
C++ conv, 7 C++ GEMM, 6 Python conv, 10 Python GEMM. CPU reference
validation for forward, backward-data, and backward-weight (2D) in both
C++ and Python examples pass.
## Test Result
30 examples pass. Peak performance: 132 TFLOPS (Batch-32 forward 56x56),
53 TFLOPS (pointwise 1x1). CPU reference accuracy: max_abs_diff < 0.002
for all directions (fp16 vs fp32 reference).
## Submission Checklist
- [x] Look over the contributing guidelines at
https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.
* WIP POC of dispatcher
* Dispatcher python workflow setup.
* Dispatcher cleanup and updates.
Further dispatcher cleanup and updates.
Build fixes
Improvements and python to CK example
Improvements to readme
* Fixes to python paths
* Cleaning up code
* Improving dispatcher support for different arch
Fixing typos
* Fix formatting errors
* Cleaning up examples
* Improving codegeneration
* Improving and fixing C++ examples
* Adding conv functionality (fwd,bwd,bwdw) and examples.
* Fixes based on feedback.
* Further fixes based on feedback.
* Adding stress test for autogeneration and autocorrection, and fixing preshuffle bug.
* Another round of improvements based on feedback.
* Trimming out unnecessary code.
* Fixing the multi-D implementation.
* Using gpu verification for gemms and fixing convolutions tflops calculation.
* Fix counter usage issue and arch filtering per ops.
* Adding changelog and other fixes.
* Improve examples and resolve critical bugs.
* Reduce build time for python examples.
* Fixing minor bug.
* Fix compilation error.
* Improve installation instructions for dispatcher.
* Add docker based installation instructions for dispatcher.
* Fixing arch-based filtering to match tile engine.
* Remove dead code and fix arch filtering.
* Minor bugfix.
* Updates after rebase.
* Trimming code.
* Fix copyright headers.
* Consolidate examples, cut down code.
* Minor fixes.
* Improving python examples.
* Update readmes.
* Remove conv functionality.
* Cleanup following conv removable.
