[qconv/qlinear] Enabling intra-op parallelism #26692
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
pytorchbot
added
module: operators
oncall: quantization
dskhudia
added a commit
that referenced
this pull request
Sep 24, 2019
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) ghstack-source-id: 90635816 Pull Request resolved: #26692
dskhudia
requested review from
jianyuh,
dzhulgakov and
raghuramank100
and removed request for
jianyuh
|
Yeah seems reasonable, but definitely need perf numbers before approval |
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear ~~TODO: Performance numbers.~~ qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.  Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
dskhudia
added a commit
that referenced
this pull request
Sep 24, 2019
Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 90712827 Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/)
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear ~~TODO: Performance numbers.~~ qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.  Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
dskhudia
added a commit
that referenced
this pull request
Sep 25, 2019
Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 90776466 Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/)
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear ~~TODO: Performance numbers.~~ qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.  Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
dskhudia
added a commit
that referenced
this pull request
Sep 28, 2019
Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 90976812 Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/)
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear ~~TODO: Performance numbers.~~ qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.  Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear ~~TODO: Performance numbers.~~ qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.  Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
dskhudia
added a commit
that referenced
this pull request
Sep 30, 2019
Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91050435 Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/)
ilia-cher
approved these changes
Sep 30, 2019
Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear ~~TODO: Performance numbers.~~ qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.  Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/) [ghstack-poisoned]
dskhudia
added a commit
that referenced
this pull request
Oct 1, 2019
Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Differential Revision: [D17540567](https://our.internmc.facebook.com/intern/diff/D17540567/)
|
This pull request has been merged in 3eefc54. |
zdevito
pushed a commit
to zdevito/ATen
that referenced
this pull request
Oct 2, 2019
Summary: Pull Request resolved: pytorch/pytorch#26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
jamesr66a
pushed a commit
that referenced
this pull request
Oct 3, 2019
Summary: Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
jamesr66a
pushed a commit
that referenced
this pull request
Oct 3, 2019
Summary: Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
jamesr66a
pushed a commit
that referenced
this pull request
Oct 3, 2019
Summary: Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
jamesr66a
pushed a commit
that referenced
this pull request
Oct 4, 2019
Summary: Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
jamesr66a
pushed a commit
that referenced
this pull request
Oct 4, 2019
Summary: Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
soumith
pushed a commit
that referenced
this pull request
Oct 7, 2019
Summary: Pull Request resolved: #26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
jianyuh
added a commit
that referenced
this pull request
Oct 22, 2019
Similar to #26692, we would like to enable the intra-op parallelism for dynamic Linear op. Differential Revision: [D18074757](https://our.internmc.facebook.com/intern/diff/D18074757/) [ghstack-poisoned]
jianyuh
added a commit
that referenced
this pull request
Oct 22, 2019
… operator" Similar to #26692, we would like to enable the intra-op parallelism for dynamic Linear op. Differential Revision: [D18074757](https://our.internmc.facebook.com/intern/diff/D18074757/) [ghstack-poisoned]
jianyuh
added a commit
that referenced
this pull request
Oct 22, 2019
Pull Request resolved: #28477 Similar to #26692, we would like to enable the intra-op parallelism for dynamic Linear op. ghstack-source-id: 92419573 Differential Revision: [D18074757](https://our.internmc.facebook.com/intern/diff/D18074757/)
facebook-github-bot
pushed a commit
that referenced
this pull request
Oct 28, 2019
…28477) Summary: Pull Request resolved: #28477 Similar to #26692, we would like to enable the intra-op parallelism for dynamic Linear op. ghstack-source-id: 92419573 Test Plan: CI Test Benchmark: ``` import time import torch K, N = 1024, 1024 print('M', 'nthread=1', 'nthread=2', 'nthread=4', 'nthread=8', 'nthread=16', sep=', ') for M in range(512, 2049, 512): print(M, sep=',', end=', ') for num_threads in (1, 2, 4, 8, 16,): torch.set_num_threads(num_threads) x = torch.rand(M, K) w = torch.rand(K, N) NITER = 20 # Test dynamic quantized q_w = torch.quantize_per_tensor(w, 0.01, 0, dtype=torch.qint8) packed_w = torch.ops.quantized.linear_prepack(q_w, None) s = time.time() for i in range(NITER): torch.ops.quantized.linear_dynamic(x, packed_w) elapsed_per_iter_dyn_quant = (time.time() - s) / NITER print("{:0.2f}".format(2.0*M*N*K/elapsed_per_iter_dyn_quant/1E9), end=', ') print("\n", end='') ``` Before this Diff: ``` (base) [root@rtptest10054.frc2 ~/jhuang_test/dynamic_quant]# python benchmark_quantize_dynamic.py M, nthread=1, nthread=2, nthread=4, nthread=8, nthread=16 512, 119.28, 139.50, 141.66, 141.58, 141.42, 1024, 122.42, 141.21, 123.09, 141.85, 123.03, 1536, 122.80, 122.18, 141.39, 123.25, 141.35, 2048, 123.41, 141.34, 123.62, 140.55, 123.76, ``` After this Diff: ``` (base) [root@rtptest10054.frc2 ~/jhuang_test/dynamic_quant]# python benchmark_quantize_dynamic.py M, nthread=1, nthread=2, nthread=4, nthread=8, nthread=16 512, 123.29, 271.99, 508.66, 882.83, 1295.07, 1024, 126.05, 273.15, 515.42, 914.11, 877.63, 1536, 142.48, 236.85, 524.10, 481.32, 970.81, 2048, 124.76, 279.03, 433.73, 958.67, 1045.82, ``` Differential Revision: D18074757 fbshipit-source-id: ad5b43477d2187c818c137093c6d6af02d5ca1d5
zdevito
pushed a commit
to zdevito/ATen
that referenced
this pull request
Oct 28, 2019
…#28477) Summary: Pull Request resolved: pytorch/pytorch#28477 Similar to pytorch/pytorch#26692, we would like to enable the intra-op parallelism for dynamic Linear op. ghstack-source-id: 92419573 Test Plan: CI Test Benchmark: ``` import time import torch K, N = 1024, 1024 print('M', 'nthread=1', 'nthread=2', 'nthread=4', 'nthread=8', 'nthread=16', sep=', ') for M in range(512, 2049, 512): print(M, sep=',', end=', ') for num_threads in (1, 2, 4, 8, 16,): torch.set_num_threads(num_threads) x = torch.rand(M, K) w = torch.rand(K, N) NITER = 20 # Test dynamic quantized q_w = torch.quantize_per_tensor(w, 0.01, 0, dtype=torch.qint8) packed_w = torch.ops.quantized.linear_prepack(q_w, None) s = time.time() for i in range(NITER): torch.ops.quantized.linear_dynamic(x, packed_w) elapsed_per_iter_dyn_quant = (time.time() - s) / NITER print("{:0.2f}".format(2.0*M*N*K/elapsed_per_iter_dyn_quant/1E9), end=', ') print("\n", end='') ``` Before this Diff: ``` (base) [root@rtptest10054.frc2 ~/jhuang_test/dynamic_quant]# python benchmark_quantize_dynamic.py M, nthread=1, nthread=2, nthread=4, nthread=8, nthread=16 512, 119.28, 139.50, 141.66, 141.58, 141.42, 1024, 122.42, 141.21, 123.09, 141.85, 123.03, 1536, 122.80, 122.18, 141.39, 123.25, 141.35, 2048, 123.41, 141.34, 123.62, 140.55, 123.76, ``` After this Diff: ``` (base) [root@rtptest10054.frc2 ~/jhuang_test/dynamic_quant]# python benchmark_quantize_dynamic.py M, nthread=1, nthread=2, nthread=4, nthread=8, nthread=16 512, 123.29, 271.99, 508.66, 882.83, 1295.07, 1024, 126.05, 273.15, 515.42, 914.11, 877.63, 1536, 142.48, 236.85, 524.10, 481.32, 970.81, 2048, 124.76, 279.03, 433.73, 958.67, 1045.82, ``` Differential Revision: D18074757 fbshipit-source-id: ad5b43477d2187c818c137093c6d6af02d5ca1d5
pdlive215
pushed a commit
to pdlive215/pytorch
that referenced
this pull request
Nov 27, 2019
Summary: Pull Request resolved: pytorch#26692 Adding intra-op parallelism for qconv and qlinear. export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear TODO: Performance numbers. ghstack-source-id: 91135613 Test Plan: export OMP_NUM_THREADS=4 python test/test_quantized.py TestQuantizedConv.test_qconv python test/test_quantized.py TestQuantizedLinear.test_qlinear Differential Revision: D17540567 fbshipit-source-id: e9962bdf0c25fd3ac4bd0673eee1edd697924406
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack:
Adding intra-op parallelism for qconv and qlinear.
export OMP_NUM_THREADS=4
python test/test_quantized.py TestQuantizedConv.test_qconv
python test/test_quantized.py TestQuantizedLinear.test_qlinear
TODO: Performance numbers.qconv performance ( I had to set manually with at::set_num_threads(4)). The op idxs which see the same performance are groupwise convolutions and parallelism for groupwise is not yet supported.
Differential Revision: D17540567