Per-channel baseline#26516
Closed
raghuramank100 wants to merge 19 commits intogh/raghuramank100/26/basefrom
Closed
Per-channel baseline#26516raghuramank100 wants to merge 19 commits intogh/raghuramank100/26/basefrom
raghuramank100 wants to merge 19 commits intogh/raghuramank100/26/basefrom
Conversation
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
raghuramank100
pushed a commit
that referenced
this pull request
Sep 20, 2019
Pull Request resolved: #26516 ghstack-source-id: 90471845 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
This was referenced Sep 20, 2019
raghuramank100
pushed a commit
that referenced
this pull request
Sep 20, 2019
Pull Request resolved: #26516 ghstack-source-id: 90517083 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
This was referenced Sep 22, 2019
jerryzh168
reviewed
Sep 23, 2019
test/test_quantized_models.py
Outdated
| def test_float_quant_compare_per_channel(self): | ||
| # Test for per-channel Quant | ||
| torch.manual_seed(67) | ||
| myModel = ModelMultipleOps().to(torch.float32) |
Contributor
There was a problem hiding this comment.
nit: can you use '_' naming e.g. 'my_model', or maybe just 'model'? so that we have a consistent naming?
jerryzh168
reviewed
Sep 23, 2019
jerryzh168
reviewed
Sep 23, 2019
torch/nn/quantized/modules/conv.py
Outdated
| # regular QTensor form for serialization. Packed weights should not live | ||
| # outside the process in which they were created, rather they should be derived | ||
| # from the QTensor weight. | ||
|
|
jerryzh168
reviewed
Sep 23, 2019
jerryzh168
reviewed
Sep 23, 2019
torch/nn/quantized/modules/conv.py
Outdated
| act_scale, act_zp = activation_observer.calculate_qparams() | ||
| assert weight_observer.dtype == torch.qint8, 'Weight observer must have a dtype of qint8' | ||
| wt_scale, wt_zp = weight_observer.calculate_qparams() | ||
| if weight_observer.qscheme in set([torch.per_tensor_symmetric, torch.per_tensor_affine]): |
jerryzh168
reviewed
Sep 23, 2019
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
raghuramank100
pushed a commit
that referenced
this pull request
Sep 27, 2019
Pull Request resolved: #26516 ghstack-source-id: 90895960 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
jerryzh168
reviewed
Sep 27, 2019
torch/nn/quantized/modules/utils.py
Outdated
|
|
||
| def _quantize_weight(float_wt, observer): | ||
| wt_scale, wt_zp = observer.calculate_qparams() | ||
| if observer.qscheme in {torch.per_tensor_symmetric, torch.per_tensor_affine}: |
Contributor
There was a problem hiding this comment.
nit: this can be a list
jerryzh168
reviewed
Sep 27, 2019
| @@ -4,26 +4,47 @@ | |||
| from common_quantization import QuantizationTestCase, ModelMultipleOps | |||
Contributor
There was a problem hiding this comment.
why is this not in test_quantization.py?
Contributor
There was a problem hiding this comment.
this looks like an integration test
Contributor
Author
There was a problem hiding this comment.
Let's chat. test_quantized_models covers tests at a higher level: more complex models and comparison of float with quantized numerics at the model level. I think its better to have it as a separate file as test_quantization is really large already
jerryzh168
approved these changes
Sep 27, 2019
Contributor
jerryzh168
left a comment
There was a problem hiding this comment.
maybe remove test_quantized_model.py and move the code to test_quantization.py? could be in another PR
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
raghuramank100
pushed a commit
that referenced
this pull request
Sep 27, 2019
Pull Request resolved: #26516 ghstack-source-id: 90926724 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
raghuramank100
pushed a commit
that referenced
this pull request
Sep 27, 2019
Pull Request resolved: #26516 ghstack-source-id: 90966545 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/) [ghstack-poisoned]
Contributor
|
This pull request has been merged in 2ccbdb7. |
jamesr66a
pushed a commit
that referenced
this pull request
Oct 3, 2019
Summary: Pull Request resolved: #26516 ghstack-source-id: 90982010 Test Plan: Integrate per-channel support into conv and linear modules. The following tests pass: buck test caffe2/test:quantized -- 'test_linear_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_conv_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_float_quant_compare_per_channel \(test_quantized_models\.ModelNumerics\)' --print-passing-details Differential Revision: D17342622 fbshipit-source-id: f0d618928e3d9348672c589a6b7a47049c372a2e
jamesr66a
pushed a commit
that referenced
this pull request
Oct 3, 2019
Summary: Pull Request resolved: #26516 ghstack-source-id: 90982010 Test Plan: Integrate per-channel support into conv and linear modules. The following tests pass: buck test caffe2/test:quantized -- 'test_linear_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_conv_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_float_quant_compare_per_channel \(test_quantized_models\.ModelNumerics\)' --print-passing-details Differential Revision: D17342622 fbshipit-source-id: f0d618928e3d9348672c589a6b7a47049c372a2e
jamesr66a
pushed a commit
that referenced
this pull request
Oct 4, 2019
Summary: Pull Request resolved: #26516 ghstack-source-id: 90982010 Test Plan: Integrate per-channel support into conv and linear modules. The following tests pass: buck test caffe2/test:quantized -- 'test_linear_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_conv_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_float_quant_compare_per_channel \(test_quantized_models\.ModelNumerics\)' --print-passing-details Differential Revision: D17342622 fbshipit-source-id: f0d618928e3d9348672c589a6b7a47049c372a2e
jamesr66a
pushed a commit
that referenced
this pull request
Oct 4, 2019
Summary: Pull Request resolved: #26516 ghstack-source-id: 90982010 Test Plan: Integrate per-channel support into conv and linear modules. The following tests pass: buck test caffe2/test:quantized -- 'test_linear_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_conv_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_float_quant_compare_per_channel \(test_quantized_models\.ModelNumerics\)' --print-passing-details Differential Revision: D17342622 fbshipit-source-id: f0d618928e3d9348672c589a6b7a47049c372a2e
soumith
pushed a commit
that referenced
this pull request
Oct 7, 2019
Summary: Pull Request resolved: #26516 ghstack-source-id: 90982010 Test Plan: Integrate per-channel support into conv and linear modules. The following tests pass: buck test caffe2/test:quantized -- 'test_linear_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_conv_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_float_quant_compare_per_channel \(test_quantized_models\.ModelNumerics\)' --print-passing-details Differential Revision: D17342622 fbshipit-source-id: f0d618928e3d9348672c589a6b7a47049c372a2e
pdlive215
pushed a commit
to pdlive215/pytorch
that referenced
this pull request
Nov 27, 2019
Summary: Pull Request resolved: pytorch#26516 ghstack-source-id: 90982010 Test Plan: Integrate per-channel support into conv and linear modules. The following tests pass: buck test caffe2/test:quantized -- 'test_linear_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_conv_api \(test_quantized_nn_mods\.ModuleAPITest\)' --print-passing-details buck test caffe2/test:quantized -- 'test_float_quant_compare_per_channel \(test_quantized_models\.ModelNumerics\)' --print-passing-details Differential Revision: D17342622 fbshipit-source-id: f0d618928e3d9348672c589a6b7a47049c372a2e
xxtEchjovs44
pushed a commit
to xxtEchjovs44/pytorch
that referenced
this pull request
Jan 29, 2020
Pull Request resolved: pytorch/pytorch#26516 ghstack-source-id: 90704045 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
xxtEchjovs44
pushed a commit
to xxtEchjovs44/pytorch
that referenced
this pull request
Jan 29, 2020
Pull Request resolved: pytorch/pytorch#26516 ghstack-source-id: 90988292 Differential Revision: [D17342622](https://our.internmc.facebook.com/intern/diff/D17342622/)
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
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:
Differential Revision: D17342622