Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Add required example_args argument to prepare_fx and prepare_qat_fx #791

Closed
wants to merge 1 commit into from
Closed

Add required example_args argument to prepare_fx and prepare_qat_fx #791

wants to merge 1 commit into from

Conversation

jerryzh168
Copy link
Contributor

Summary:
FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

@facebook-github-bot facebook-github-bot added CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported labels May 16, 2022
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

@jerryzh168 jerryzh168 mentioned this pull request May 16, 2022
Closed
jerryzh168 added a commit to jerryzh168/ClassyVision-1 that referenced this pull request May 17, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
f8a5187 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: ef5536ff98a3e621ab0d10341940dcb4a2dfcd32
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/d2go that referenced this pull request May 17, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
b941adc 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

Pull Request resolved: facebookresearch#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 2fc9c06805d443fc1478d530232cdbcfeef39f67
jerryzh168 added a commit to jerryzh168/mobile-vision that referenced this pull request May 17, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
97ecdc4 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: facebookresearch#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 2a9df6332f24650b26dfbc4c754b9156d38ea890
jerryzh168 added a commit to jerryzh168/benchmark that referenced this pull request May 17, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
a5b8463 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

Pull Request resolved: pytorch#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 7e1ce6dc13a1ecc4d46939c8e3b3f3721248c727
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/ClassyVision-1 that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
e7c046c 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 58c1e0afa7421ce79c164a31e88bb7dc4541f42b
jerryzh168 added a commit to jerryzh168/d2go that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
fcd795b 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

Pull Request resolved: facebookresearch#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 7150e372404a9a6a9352163b7dce8963a7a3293b
jerryzh168 added a commit to jerryzh168/mobile-vision that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
1981806 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: facebookresearch#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: c01860fe846684bb1e781dac19a7b2d89d004329
jerryzh168 added a commit to jerryzh168/benchmark that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
54c8b9a 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

Pull Request resolved: pytorch#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 7abfc1c5c57633e7a7e38060d9552e45659cb2a1
jerryzh168 added a commit to jerryzh168/ClassyVision-1 that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
d43cc8f 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: bc7b108b768293a74561825b2df95d84fb4822ee
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/d2go that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
56701e8 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

Pull Request resolved: facebookresearch#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 5b7837005a34a095b331dbca7d6a8c2d6fa5ee51
jerryzh168 added a commit to jerryzh168/ClassyVision-1 that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
a88bbc1 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: f1690c0d02f6ce719b935843bf5bcb4a881da993
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/mobile-vision that referenced this pull request May 19, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
c2b0fb3 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: facebookresearch#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)

After:
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 038fd19ade462e3c4b6a4f8db1661bf5e85b762b
jerryzh168 added a commit to jerryzh168/mobile-vision that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx
def5ce3 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: facebookresearch#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 06a3020780ab9745abad3f069f35a66a8bec58be
jerryzh168 added a commit to jerryzh168/benchmark that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
9d6d95e 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

Pull Request resolved: pytorch#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: cb722cb1fd4e3973d5e222451b3b64c87c52d31d
jerryzh168 added a commit to jerryzh168/ClassyVision-1 that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_args argument to prepare_fx and prepare_qat_fx (…
166f2dd 
…#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 558ab28a46ad855aa147b25ed0dd6aafa22f6cbd
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/mobile-vision that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx
5d29a94 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: facebookresearch#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: f55cbbe418d3f3a4e91123eba1c18e9487f720db
jerryzh168 added a commit to jerryzh168/ClassyVision-1 that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx…
d645c6c 
… (#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: d3d9d81cd9fbb191ff43471e0f1a20906925d24a
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/d2go that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx
78e3df1 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

Pull Request resolved: facebookresearch#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 44a77a37397618ccd665fda22f3a365d0117c421
jerryzh168 added a commit to jerryzh168/benchmark that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx…
87a0f10 
… (#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

Pull Request resolved: pytorch#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: c7b82d836e7b5cde291b310d2ba6a3ff1803b661
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx…
4846183 
… (#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

Pull Request resolved: facebookresearch#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 68861512ebb3dd1ae6e069a144c30eee2d63e5da
jerryzh168 added a commit to jerryzh168/d2go that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx
f37828d 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

Pull Request resolved: facebookresearch#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: db85afad4af5a37a834cc84aa581fbfce7462bef
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D35984526

jerryzh168 added a commit to jerryzh168/mobile-vision that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx
651c11f 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: facebookresearch#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: dbb586adac94951d8ea2cb257df38f413ae97647
jerryzh168 added a commit to jerryzh168/benchmark that referenced this pull request May 20, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx…
3e6f57d 
… (#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

Pull Request resolved: pytorch#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: e598f5df1134ddf28a7ca0e346e9ea88c06d370d
jerryzh168 added a commit to jerryzh168/pytorch that referenced this pull request May 21, 2022
@jerryzh168 @facebook-github-bot
[quant][fx][bc-breaking] Add required example_inputs argument to prep…
9e0b8bb 
…are_fx and prepare_qat_fx (pytorch#77608)

Summary:
Pull Request resolved: pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Test Plan:
python test/test_quantization.py TestQuantizeFx
python test/test_quantization.py TestQuantizeFxOps
python test/test_quantization.py TestQuantizeFxModels

Imported from OSS

**Static Docs Preview: classyvision**
|[Full Site](https://our.intern.facebook.com/intern/staticdocs/eph/D35984526/V44/classyvision/)|

|**Modified Pages**|

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 716e5992ebe99cfb90be669357f56b214d692aef
facebook-github-bot pushed a commit to facebookresearch/d2go that referenced this pull request May 21, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx
9f74615 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

Pull Request resolved: #249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 706c8df71722c9aa5082a6491734f0144f0dd670
facebook-github-bot pushed a commit to pytorch/benchmark that referenced this pull request May 21, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx…
6f0d2fb 
… (#77608)

Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

Pull Request resolved: #916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 706c8df71722c9aa5082a6491734f0144f0dd670
facebook-github-bot pushed a commit to facebookresearch/mobile-vision that referenced this pull request May 21, 2022
@jerryzh168 @facebook-github-bot
Add required example_inputs argument to prepare_fx and prepare_qat_fx
659d501 
Summary:
X-link: pytorch/pytorch#77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

Pull Request resolved: #68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 706c8df71722c9aa5082a6491734f0144f0dd670
pytorchmergebot pushed a commit to pytorch/pytorch that referenced this pull request May 21, 2022
@jerryzh168 @pytorchmergebot
[quant][fx][bc-breaking] Add required example_args argument to prepar…
416899d 
…e_fx and prepare_qat_fx (#249) (#77608)

Summary:
X-link: facebookresearch/d2go#249

X-link: https://github.com/fairinternal/ClassyVision/pull/104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make #76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Test Plan:
python test/test_quantization.py TestQuantizeFx
python test/test_quantization.py TestQuantizeFxOps
python test/test_quantization.py TestQuantizeFxModels

Imported from OSS

**Static Docs Preview: classyvision**
|[Full Site](https://our.intern.facebook.com/intern/staticdocs/eph/D35984526/V30/classyvision/)|

|**Modified Pages**|

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

Pull Request resolved: #77608
Approved by: https://github.com/dzdang
facebook-github-bot pushed a commit to pytorch/pytorch that referenced this pull request May 21, 2022
@jerryzh168 @facebook-github-bot
[quant][fx][bc-breaking] Add required example_inputs argument to prep…
f64a8dc 
…are_fx and prepare_qat_fx (#77608)

Summary:
Pull Request resolved: #77608

X-link: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make #76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Test Plan:
python test/test_quantization.py TestQuantizeFx
python test/test_quantization.py TestQuantizeFxOps
python test/test_quantization.py TestQuantizeFxModels

Imported from OSS

**Static Docs Preview: classyvision**
|[Full Site](https://our.intern.facebook.com/intern/staticdocs/eph/D35984526/V44/classyvision/)|

|**Modified Pages**|

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 706c8df71722c9aa5082a6491734f0144f0dd670
swang392 pushed a commit to pytorch/pytorch that referenced this pull request May 25, 2022
@jerryzh168 @swang392
[quant][fx][bc-breaking] Add required example_args argument to prepar…
804c100 
…e_fx and prepare_qat_fx (#249) (#77608)

Summary:
X-link: facebookresearch/d2go#249

X-link: https://github.com/fairinternal/ClassyVision/pull/104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make #76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Test Plan:
python test/test_quantization.py TestQuantizeFx
python test/test_quantization.py TestQuantizeFxOps
python test/test_quantization.py TestQuantizeFxModels

Imported from OSS

**Static Docs Preview: classyvision**
|[Full Site](https://our.intern.facebook.com/intern/staticdocs/eph/D35984526/V30/classyvision/)|

|**Modified Pages**|

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

Pull Request resolved: #77608
Approved by: https://github.com/dzdang
frank-wei pushed a commit to pytorch/TensorRT that referenced this pull request Jun 4, 2022
@jerryzh168
[quant][fx][bc-breaking] Add required example_inputs argument to prep…
134de54 
…are_fx and prepare_qat_fx (#77608)

Summary:
X-link: pytorch/pytorch#77608

Pull Request resolved: pytorch/fx2trt#76

X-link: facebookresearch/d2go#249

X-link: fairinternal/ClassyVision#104

X-link: pytorch/benchmark#916

X-link: facebookresearch/ClassyVision#791

X-link: facebookresearch/mobile-vision#68

FX Graph Mode Quantization needs to know whether an fx node is a floating point Tensor before it can decide whether to
insert observer/fake_quantize module or not, since we only insert observer/fake_quantize module for floating point Tensors.
Currently we have some hacks to support this by defining some rules like NON_OBSERVABLE_ARG_DICT (https://github.com/pytorch/pytorch/blob/master/torch/ao/quantization/fx/utils.py#L496), but this approach is fragile and we do not plan to maintain it long term in the pytorch code base.

As we discussed in the design review, we'd need to ask users to provide sample args and sample keyword args
so that we can infer the type in a more robust way. This PR starts with changing the prepare_fx and prepare_qat_fx api to require user to either provide
example arguments thrugh example_inputs, Note this api doesn't support kwargs, kwargs can make pytorch/pytorch#76496 (comment) (comment) simpler, but
it will be rare, and even then we can still workaround with positional arguments, also torch.jit.trace(https://pytorch.org/docs/stable/generated/torch.jit.trace.html) and ShapeProp: https://github.com/pytorch/pytorch/blob/master/torch/fx/passes/shape_prop.py#L140 just have single positional args, we'll just use a single example_inputs argument for now.

If needed, we can extend the api with an optional example_kwargs. e.g. in case when there are a lot of arguments for forward and it makes more sense to
pass the arguments by keyword

BC-breaking Note:
Before:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict)
# or
m = prepare_qat_fx(m, qconfig_dict)
```
After:
```python
m = resnet18(...)
m = prepare_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
# or
m = prepare_qat_fx(m, qconfig_dict, example_inputs=(torch.randn(1, 3, 224, 224),))
```

Reviewed By: vkuzo, andrewor14

Differential Revision: D35984526

fbshipit-source-id: 706c8df71722c9aa5082a6491734f0144f0dd670
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

None yet
2 participants
@jerryzh168 @facebook-github-bot