Ok, the build flag seems to have been broken for a while since the function it calls doesn't exist anymore.
Repurposed it to enable dispatcher printing (which requires a full (and slow) debug build otherwise).
Pull Request resolved: https://github.com/pytorch/pytorch/pull/145517
Approved by: https://github.com/bdhirsh
Useful for code reuse for Metal shader build both for eager mode and MPSInductor, but it requires one to implement `_cpp_embed_headers` tool that, as name suggests, would preprocess and embeds the for shader to be used in dynamic compilation.
Test using:
- `TestMetalLibrary.test_metal_include`
- Moving `i0`/`i1` implementation to `c10/util/metal_special_math.h` and call it from `SpecialOps.metal` shader, which now looks much more compact:
```metal
template <typename T, typename Tout = T>
void kernel
i0(constant T* input,
device Tout* output,
uint index [[thread_position_in_grid]]) {
output[index] = c10::i0(static_cast<Tout>(input[index]));
}
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/145087
Approved by: https://github.com/dcci
ghstack dependencies: #145023
Description:
1. Quantize Linear Layer Weights to 4-bits:
Quantize the weights of the Linear layer to 4 bits, using symmetric quantization.
Pack two 4-bit weights into one uint8 container.
Choose a quantization scheme (channel-wise or group-wise), with the group size being a multiple of 32.
2. Prepare Quantized Weights, Scales, and Optional Bias:
After quantizing, obtain the quantized_weights, scales, and groupsize.
If the original Linear layer has a bias, prepare it as well.
3. Pack the Weights Efficiently:
Use torch.ops.aten._dyn_quant_pack_4bit_weight to optimally pack the weights, scales, and optional bias.
```python
packed_weights = torch.ops.aten._dyn_quant_pack_4bit_weight(weight, scales_and_zeros, bias, groupsize, in_features, out_features)
```
Input parameters should include:
in_features and out_features (the same as the Linear layer’s corresponding parameters).
4. Perform Dynamic Quantized Matrix Multiplication:
Use torch.ops.aten._dyn_quant_matmul_4bit to perform matrix multiplication with quantized weights.
```python
output = torch.ops.aten._dyn_quant_matmul_4bit(input, packed_weights, groupsize, in_features, out_features)
```
Inputs required include:
The input tensor, packed_weights , groupsize, and the in_features and out_features.
API Usage: https://github.com/pytorch/pytorch/issues/143289
Model Perf :
7B Transformer model:
Prefill : 340 t/s
Decode : 40 t/s
2B Transformer model
Prefill : 747 t/s
Decode : 80 t/s
Tests:
python test/test_linalg.py -k test__dyn_quant_pack_4bit_weight
Ran 1 test in 0.016s
OK
python test/test_linalg.py -k test__dyn_quant_matmul_4bit
Ran 8 tests in 0.077s
OK
python test/test_linalg.py -k test_compile_dyn_quant_matmul_4bit
Ran 8 tests in 11.454s
Change-Id: Ia1672bad5e6ec94e64d8bb1971395d60f4b3a452
Fixes #ISSUE_NUMBER
Pull Request resolved: https://github.com/pytorch/pytorch/pull/134124
Approved by: https://github.com/digantdesai, https://github.com/malfet
Description:
1. Quantize Linear Layer Weights to 4-bits:
Quantize the weights of the Linear layer to 4 bits, using symmetric quantization.
Pack two 4-bit weights into one uint8 container.
Choose a quantization scheme (channel-wise or group-wise), with the group size being a multiple of 32.
2. Prepare Quantized Weights, Scales, and Optional Bias:
After quantizing, obtain the quantized_weights, scales, and groupsize.
If the original Linear layer has a bias, prepare it as well.
3. Pack the Weights Efficiently:
Use torch.ops.aten._dyn_quant_pack_4bit_weight to optimally pack the weights, scales, and optional bias.
```python
packed_weights = torch.ops.aten._dyn_quant_pack_4bit_weight(weight, scales_and_zeros, bias, groupsize, in_features, out_features)
```
Input parameters should include:
in_features and out_features (the same as the Linear layer’s corresponding parameters).
4. Perform Dynamic Quantized Matrix Multiplication:
Use torch.ops.aten._dyn_quant_matmul_4bit to perform matrix multiplication with quantized weights.
```python
output = torch.ops.aten._dyn_quant_matmul_4bit(input, packed_weights, groupsize, in_features, out_features)
```
Inputs required include:
The input tensor, packed_weights , groupsize, and the in_features and out_features.
API Usage: https://github.com/pytorch/pytorch/issues/143289
Model Perf :
7B Transformer model:
Prefill : 340 t/s
Decode : 40 t/s
2B Transformer model
Prefill : 747 t/s
Decode : 80 t/s
Tests:
python test/test_linalg.py -k test__dyn_quant_pack_4bit_weight
Ran 1 test in 0.016s
OK
python test/test_linalg.py -k test__dyn_quant_matmul_4bit
Ran 8 tests in 0.077s
OK
python test/test_linalg.py -k test_compile_dyn_quant_matmul_4bit
Ran 8 tests in 11.454s
Change-Id: Ia1672bad5e6ec94e64d8bb1971395d60f4b3a452
Fixes #ISSUE_NUMBER
Pull Request resolved: https://github.com/pytorch/pytorch/pull/134124
Approved by: https://github.com/digantdesai, https://github.com/malfet
Description:
1. Quantize Linear Layer Weights to 4-bits:
Quantize the weights of the Linear layer to 4 bits, using symmetric quantization.
Pack two 4-bit weights into one uint8 container.
Choose a quantization scheme (channel-wise or group-wise), with the group size being a multiple of 32.
2. Prepare Quantized Weights, Scales, and Optional Bias:
After quantizing, obtain the quantized_weights, scales, and groupsize.
If the original Linear layer has a bias, prepare it as well.
3. Pack the Weights Efficiently:
Use torch.ops.aten._dyn_quant_pack_4bit_weight to optimally pack the weights, scales, and optional bias.
```python
packed_weights = torch.ops.aten._dyn_quant_pack_4bit_weight(weight, scales_and_zeros, bias, groupsize, in_features, out_features)
```
Input parameters should include:
in_features and out_features (the same as the Linear layer’s corresponding parameters).
4. Perform Dynamic Quantized Matrix Multiplication:
Use torch.ops.aten._dyn_quant_matmul_4bit to perform matrix multiplication with quantized weights.
```python
output = torch.ops.aten._dyn_quant_matmul_4bit(input, packed_weights, groupsize, in_features, out_features)
```
Inputs required include:
The input tensor, packed_weights , groupsize, and the in_features and out_features.
API Usage: https://github.com/pytorch/pytorch/issues/143289
Model Perf :
7B Transformer model:
Prefill : 340 t/s
Decode : 40 t/s
2B Transformer model
Prefill : 747 t/s
Decode : 80 t/s
Tests:
python test/test_linalg.py -k test__dyn_quant_pack_4bit_weight
Ran 1 test in 0.016s
OK
python test/test_linalg.py -k test__dyn_quant_matmul_4bit
Ran 8 tests in 0.077s
OK
python test/test_linalg.py -k test_compile_dyn_quant_matmul_4bit
Ran 8 tests in 11.454s
Change-Id: Ia1672bad5e6ec94e64d8bb1971395d60f4b3a452
Fixes #ISSUE_NUMBER
Pull Request resolved: https://github.com/pytorch/pytorch/pull/134124
Approved by: https://github.com/digantdesai, https://github.com/malfet
Notable new features for SDPA operators on AMD systems from AOTriton 0.8b:
1. Nestedtensor support;
2. MQA/GQA support;
3. Restore Efficient attention support for causal=True and seqlen_q != seqlen_k cases;
+ The kernel should use top-left alignment, bottom right alignment will be added later
4. Move gfx1100 (RX7900/W7800/W7900) out of experimental support status.
However, users are strongly recommended to update to ROCM 6.2.4, notably for
its firmware updates.
Related unit tests are enabled as well.
Notable related changes from AOTriton 0.8b:
1. AOTriton 0.8b moves the GPU kernel out of libaotriton.so to a separate directory `aotriton.images`;
2. LZMA replaces ZSTD as GPU kernel compression algorithm for better compression ratio: aotriton0.8b (.so + aotriton.images take 350MB) compared to aotriton0.7b .so: 800MB
3. The compression cannot be disabled now, and `liblzma` is hard run-time dependency.
+ Should not be a problem, since `lzma` is part of Python Standard Library
Pull Request resolved: https://github.com/pytorch/pytorch/pull/140172
Approved by: https://github.com/jithunnair-amd, https://github.com/jeffdaily
Co-authored-by: Jithun Nair <37884920+jithunnair-amd@users.noreply.github.com>
Summary:
In this diff we implement a way to ensure the internal thrift schema from cfgr (configerator/structs/caffe2/torch/export/schema.thrift) and the schema in OSS (torch/_export/serde/schema.thrift) are in sync, by adding a unittest to reflect on the type names and fields from each schema and compare them field by field.
When we detect new fields/types from torch/_export/serde/schema.thrift, there'll be a test failure on the trunk and the error message hints people to add the missing field/type to the thrift schema from cfgr, so that they are always in sync in practice.
Test Plan: buck test mode/opt caffe2/test:test_export -- -r test_thrift_schema_in_sync
Differential Revision: D66716834
Pull Request resolved: https://github.com/pytorch/pytorch/pull/141989
Approved by: https://github.com/yiming0416
## Summary
We are currently [updating](https://github.com/conda-forge/pytorch-cpu-feedstock/pull/277) the [`conda-forge::pytorch`](https://anaconda.org/conda-forge/pytorch) package to version 2.5.0. This update includes a new dependency, the third_party/NVTX submodule. However, like other package management frameworks (e.g., apt), conda-forge prefers using system-installed packages instead of vendor-provided third-party packages.
This pull request aims to add an option, `USE_SYSTEM_NVTX`, to select whether to use the vendored nvtx or the system-installed one, with the default being the vendored one (which is the current behavior).
## Test Plan
The `USE_SYSTEM_NVTX` option is tested by building the `conda-forge::pytorch` package with the change applied as a [patch](cd1d2464dd/recipe/patches/0005-Use-system-nvtx3.patch).
Pull Request resolved: https://github.com/pytorch/pytorch/pull/138287
Approved by: https://github.com/albanD
**Motivation:**
In Pytorch, Aten vectorization supports multiple platforms, including x86 and Arm, as well as multiple data types. It provides a generic implementation of Vector (Vec) type that allows the programmer to write code packing various primitives (such as floats) within 256bit & 512bits registers. It can be extended to support other ISAs easily by adding more VecISA sub-classes.
**Reference Link:** https://github.com/pytorch/pytorch/tree/main/aten/src/ATen/cpu/vec
**This PR:**
* Our goal with this contribution is to add support for SVE backend for Vec in the Aten vectorization for CPU backend which can be benefitted by any ARM architecture supported CPU's that supports SVE.
* More about SVE ISA for ARM: [https://developer.arm.com/Architectures/Scalable Vector Extensions](https://developer.arm.com/Architectures/Scalable%20Vector%20Extensions)
* We are using the ARM C Language Extensions for SVE (https://developer.arm.com/documentation/102699/0100/Optimizing-with-intrinsics ) to accelerate performance for various operators in the SVE backend for Vec.
* Currently we are adding support only for SVE ISA with the vector length of 256 bits (SVE 256). In future, we plan to extend this SVE support for other vector lengths as well.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/119571
Approved by: https://github.com/malfet, https://github.com/snadampal
Co-authored-by: Divya Kotadiya <divya.kotadiya@fujitsu.com>
* Added a cpp loader, AOTIModelPackageLoader, which can load the .pt2, build the .so, and create a runner. The python-facing API is that users can directly call the `run` function, whereas in cpp users can directly access the `runner_` if they are more familiar with that. I couldn't figure out how to bind the `get_runner()` function to python...
* Added a new config, `aot_inductor.package_cpp_only` which will **not** package the so. This means that whenever the package is loaded, we will need to build the so. This is turned off by default so that new environments do not need to rebuild their so. The `package_cpp_only` is a feature which torchchat intends to use to provide flexibility to users.
* Added a new config, `aot_inductor.metadata` which stores user-provided metadata, serialized to the pt2 as a json file. It also stores the device used when exporting, "cuda" or "cpu", so that during load time, we can use that data to determine which AOTIModelContainerRunner to use. The metadata can be accessed through `loader.get_metadata()`. TODO is to move this metadata to the toplevel `package_aoti` function so that we can remove the metadata as a config.
* Separated out `package_aoti` as a standalone function, instead of it automatically being called in inductor. This is to prepare for the case where users will compile multiple models, and want to bundle it in one package. The specific use case is in torchchat, where we want to package the separately-exported encoder and decoder layers. An example of how to use this is in `test_multiple_methods`.
* `load_package` will load a singular model, given the model name.
* The loader doesn't support windows for now, I think I need to add some more casing to make the build commands work on windows?
Differential Revision: [D62329906](https://our.internmc.facebook.com/intern/diff/D62329906)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/135374
Approved by: https://github.com/desertfire, https://github.com/malfet
This PR switches to cuDSS library and has the same purpose of #127692, which is to add Sparse CSR tensor support to linalg.solve.
Fixes#69538
Minimum example of usage:
```
import torch
if __name__ == '__main__':
spd = torch.rand(4, 3)
A = spd.T @ spd
b = torch.rand(3).to(torch.float64).cuda()
A = A.to_sparse_csr().to(torch.float64).cuda()
x = torch.linalg.solve(A, b)
print((A @ x - b).norm())
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/129856
Approved by: https://github.com/amjames, https://github.com/lezcano, https://github.com/huydhn
Co-authored-by: Zihang Fang <zhfang1108@gmail.com>
Co-authored-by: Huy Do <huydhn@gmail.com>
Another attempt to update NVTX to NVTX3. We now avoid changing NVTX header inclusion of existing code. The advantage of NVTX3 over NVTX is that it is a header-only library so that linking with NVTX3 can greatly simplify our CMake and other building scripts for finding libraries in user environments. In addition, NVTX are indeed still present in the latest CUDA versions, but they're no longer a compiled library: It's now a header-only library. That's why there isn't a .lib file anymore.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/109843
Approved by: https://github.com/peterbell10, https://github.com/eqy
Co-authored-by: Ivan Zaitsev <108101595+izaitsevfb@users.noreply.github.com>
0.12.0 Major Updates:
- Add context manager to temporarily set the dictionary sorting mode
- Add accessor APIs
- Use `stable` tag for `pybind11` for Python 3.13 support
- Fix potential segmentation fault for pickling support
0.12.1 Updates:
- Fix warning regression during import when launch with strict warning filters
Closes#130155
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130139
Approved by: https://github.com/zou3519
ghstack dependencies: #130895
------
The opposite of #130836. Pin `sympy >= 1.13.0` for Python >= 3.9 and `sympy == 1.12.1` for Python 3.8.
- #130836
See the PR description of #130836 for more details.
`sympy` 1.13.0 introduces some breaking changes which break our tests. More specifically:
- Ref [Backwards compatibility breaks and deprecations](https://github.com/sympy/sympy/wiki/release-notes-for-1.13.0#backwards-compatibility-breaks-and-deprecations)
> BREAKING CHANGE: Float and Integer/Rational no longer compare equal with a == b. From now on Float(2.0) != Integer(2). Previously expressions involving Float would compare unequal e.g. x*2.0 != x*2 but an individual Float would compare equal to an Integer. In SymPy 1.7 a Float will always compare unequal to an Integer even if they have the same "value". Use sympy.numbers.int_valued(number) to test if a number is a concrete number with no decimal part. ([#25614](https://github.com/sympy/sympy/pull/25614) by [@smichr](https://github.com/smichr))
`sympy >= 1.13.0` is required to enable Python 3.13 support. This should be part of #130689.
- #130689
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130895
Approved by: https://github.com/ezyang
0.12.0 Major Updates:
- Add context manager to temporarily set the dictionary sorting mode
- Add accessor APIs
- Use `stable` tag for `pybind11` for Python 3.13 support
- Fix potential segmentation fault for pickling support
0.12.1 Updates:
- Fix warning regression during import when launch with strict warning filters
Closes#130155
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130139
Approved by: https://github.com/zou3519
Removes extraneous .a, .so, and .py files from the split build. From here we can also clean up the builder script which produces the binary to do this. That pr is https://github.com/pytorch/builder/pull/1912
Verification:
The built wheel with BUILD_LIBTORCH_WHL=1 has the following files only (with .a, .so, and .py extensions)
```
sahanp@devgpu086 ~/p/dist (viable/strict)> pwd (pytorch-3.10)
/home/sahanp/pytorch/dist
sahanp@devgpu086 ~/p/dist (viable/strict)> find . -type f \( -name "*.py" -o -name "*.a" -o -name "*.so" \) (pytorch-3.10)
./torch/__init__.py
./torch/lib/libbackend_with_compiler.so
./torch/lib/libc10.so
./torch/lib/libjitbackend_test.so
./torch/lib/libtorch.so
./torch/lib/libtorch_cpu.so
./torch/lib/libtorch_global_deps.so
./torch/lib/libtorchbind_test.so
sahanp@devgpu086 ~/p/dist (viable/strict)>
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130053
Approved by: https://github.com/atalman
This PR removes the second separate package we were using for the libtorch wheel.
In terms of testing that this works we will look use the PRs above this in the stack.
As for sanity checking these are the wheels that are produced by running
```
python setup.py clean && BUILD_LIBTORCH_WHL=1 with-proxy python setup.py bdist_whee
l && BUILD_PYTHON_ONLY=1 with-proxy python setup.py bdist_wheel --cmake
```
```
sahanp@devgpu086 ~/pytorch ((5f15e171…))> ls -al dist/ (pytorch-3.10)
total 677236
drwxr-xr-x 1 sahanp users 188 Jun 4 12:19 ./
drwxr-xr-x 1 sahanp users 1696 Jun 4 12:59 ../
-rw-r--r-- 1 sahanp users 81405742 Jun 4 12:19 torch-2.4.0a0+gitca0a73c-cp310-cp310-linux_x86_64.whl
-rw-r--r-- 1 sahanp users 612076919 Jun 4 12:19 libtorch-2.4.0a0+gitca0a73c-py3-none-any.whl
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/127934
Approved by: https://github.com/atalman
Summary:
1.Package public headers of kineto if USE_KINETO so that they can be used by PrivateUse1 user.
2.Add PrivateUse1 key to ActivityType.
3. Support PrivateUse1 key in function deviceTypeFromActivity and _supported_activities.
4. Fix some bugs when processing profiler results.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/124818
Approved by: https://github.com/aaronenyeshi
Summary:
1.Package public headers of kineto if USE_KINETO so that they can be used by PrivateUse1 user.
2.Add PrivateUse1 key to ActivityType.
3. Support PrivateUse1 key in function deviceTypeFromActivity and _supported_activities.
4. Fix some bugs when processing profiler results.
Co-authored-by: albanD <desmaison.alban@gmail.com>
Co-authored-by: Aaron Shi <enye.shi@gmail.com>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/120556
Approved by: https://github.com/aaronenyeshi
This PR adds a linker script optimization based on prioritized symbols that can be extracted from the profiles of popular workloads. The present linker script was generated to target ARM+CUDA and later can be extended if necessary. The reason we target ARM is shown below:
> PyTorch and other applications that access more than 24x 2MB code regions in quick succession can result in performance bottlenecks in the CPU front-end. The link-time optimization improves executable code locality and improve performance. We recommend turning on the optimization always for PyTorch and other application that behaves similarly.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/121975
Approved by: https://github.com/ptrblck, https://github.com/atalman
Summary: The current C shim layer manually implements a C interface for a handful of ops. Obviously that's not scalable if we want to extend it to cover all aten ops. This new torchgen script automatically generates C shim interfaces for CPU and CUDA backends. The interface follows the same parameter passing rules as the current C shim layer, such as
* Use plain C data types to pass parameters
* Use AtenTensorHandle to pass at::Tensor
* Use pointer type to pass optional parameter
* Use pointer+length to pass list
* Use device_type+device_index to pass device
* When a parameter is a pointer of pointer, e.g. AtenTensorHandle**, the script generates either a list of optional values or an optional list of values
https://gist.github.com/desertfire/83701532b126c6d34dae6ba68a1b074a is an example of the generated torch/csrc/inductor/aoti_torch/generated/c_shim_cuda.cpp file. The current version doesn't generate C shim wrappers for all aten ops, and probably generates more wrappers than needed on the other hand, but it should serve as a good basis.
This PR by itself won't change AOTI codegen and thus won't introduce any FC breakage. The actual wrapper codegen changes will come in another PR with some version control flag to avoid FC breakage.
Differential Revision: [D54258087](https://our.internmc.facebook.com/intern/diff/D54258087)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/120513
Approved by: https://github.com/jansel
Summary:
Expose an option to users to specify name of the LogsSpec implementation to use.
- Has to be defined in entrypoints under `torchrun.logs_specs` group.
- Must implement LogsSpec defined in prior PR/diff.
Test Plan: unit test+local tests
Reviewed By: ezyang
Differential Revision: D54180838
Pull Request resolved: https://github.com/pytorch/pytorch/pull/120942
Approved by: https://github.com/ezyang
Some operations, such as GEMMs, could be implemented using more than one library or more than one technique. For example, a GEMM could be implemented for CUDA or ROCm using either the blas or blasLt libraries. Further, ROCm's rocblas and hipblaslt libraries allow the user to query for all possible algorithms and then choose one. How does one know which implementation is the fastest and should be chosen? That's what TunableOp provides.
See the README.md for additional details.
TunableOp was ported from onnxruntime starting from commit 08dce54266. The content was significantly modified and reorganized for use within PyTorch. The files copied and their approximate new names or source content location within aten/src/ATen/cuda/tunable include the following:
- onnxruntime/core/framework/tunable.h -> Tunable.h
- onnxruntime/core/framework/tuning_context.h -> Tunable.h
- onnxruntime/core/framework/tuning_context_impl.h -> Tunable.cpp
- onnxruntime/core/providers/rocm/tunable/gemm_common.h -> GemmCommon.h
- onnxruntime/core/providers/rocm/tunable/gemm_hipblaslt.h -> GemmHipblaslt.h
- onnxruntime/core/providers/rocm/tunable/gemm_rocblas.h -> GemmRocblas.h
- onnxruntime/core/providers/rocm/tunable/gemm_tunable.cuh -> TunableGemm.h
- onnxruntime/core/providers/rocm/tunable/rocm_tuning_context.cc -> Tunable.cpp
- onnxruntime/core/providers/rocm/tunable/util.h -> StreamTimer.h
- onnxruntime/core/providers/rocm/tunable/util.cc -> StreamTimer.cpp
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114894
Approved by: https://github.com/xw285cornell, https://github.com/jianyuh
