Skip to content

Commit 1c7bfdd

Browse files
chensuyuechensuyue
committed
deploy: e22c61e 
1 parent 26107ff commit 1c7bfdd

File tree

804 files changed

+13496
-784
lines changed

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

804 files changed

+13496
-784
lines changed

‎latest/_sources/autoapi/neural_compressor/tensorflow/algorithms/static_quant/index.rst.txt

+9
Original file line numberDiff line numberDiff line change
@@ -6,3 +6,12 @@
66
.. py:module:: neural_compressor.tensorflow.algorithms.static_quant
77
88
9+
Submodules
10+
----------
11+
.. toctree::
12+
:titlesonly:
13+
:maxdepth: 1
14+
15+
tensorflow/index.rst
16+
17+

‎latest/_sources/autoapi/neural_compressor/tensorflow/algorithms/static_quant/tensorflow/index.rst.txt

+60
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,60 @@
1+
:py:mod:`neural_compressor.tensorflow.algorithms.static_quant.tensorflow`
2+
=========================================================================
3+
4+
.. py:module:: neural_compressor.tensorflow.algorithms.static_quant.tensorflow
5+
6+
.. autoapi-nested-parse::
7+
8+
Tensorflow Adaptor Classes.
9+
10+
11+
12+
Module Contents
13+
---------------
14+
15+
Classes
16+
~~~~~~~
17+
18+
.. autoapisummary::
19+
20+
neural_compressor.tensorflow.algorithms.static_quant.tensorflow.TensorFlowAdaptor
21+
neural_compressor.tensorflow.algorithms.static_quant.tensorflow.Tensorflow_ITEXAdaptor
22+
neural_compressor.tensorflow.algorithms.static_quant.tensorflow.TensorFlowConfig
23+
neural_compressor.tensorflow.algorithms.static_quant.tensorflow.TensorflowQuery
24+
neural_compressor.tensorflow.algorithms.static_quant.tensorflow.TensorflowConfigConverter
25+
26+
27+
28+
29+
.. py:class:: TensorFlowAdaptor(framework_specific_info)
30+
31+
32+
Adaptor Layer for stock tensorflow and spr-base.
33+
34+
35+
.. py:class:: Tensorflow_ITEXAdaptor(framework_specific_info)
36+
37+
38+
39+
40+
Tensorflow ITEX Adaptor Class.
41+
42+
43+
.. py:class:: TensorFlowConfig(precisions=None)
44+
45+
46+
Base config class for TensorFlow.
47+
48+
49+
.. py:class:: TensorflowQuery(local_config_file=None, performance_only=False, itex_mode=False, quant_mode='static')
50+
51+
52+
Tensorflow Query Capability Class.
53+
54+
55+
.. py:class:: TensorflowConfigConverter(quant_config: neural_compressor.tensorflow.quantization.config.StaticQuantConfig, capability: Dict)
56+
57+
58+
Convert `StaticQuantConfig` to the format used by static quant algo.
59+
60+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras`
4+
============================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/conv2d/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers.conv2d`
4+
==========================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers.conv2d
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/dense/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers.dense`
4+
=========================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers.dense
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/depthwise_conv2d/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers.depthwise_conv2d`
4+
====================================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers.depthwise_conv2d
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers`
4+
===================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/pool2d/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers.pool2d`
4+
==========================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers.pool2d
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/quantizer/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers.quantizer`
4+
=============================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers.quantizer
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/layers/separable_conv2d/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.layers.separable_conv2d`
4+
====================================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.layers.separable_conv2d
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/quantization/config/index.rst.txt

+50
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,50 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.quantization.config`
4+
================================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.quantization.config
7+
8+
9+
Module Contents
10+
---------------
11+
12+
Classes
13+
~~~~~~~
14+
15+
.. autoapisummary::
16+
17+
neural_compressor.tensorflow.keras.quantization.config.StaticQuantConfig
18+
19+
20+
21+
Functions
22+
~~~~~~~~~
23+
24+
.. autoapisummary::
25+
26+
neural_compressor.tensorflow.keras.quantization.config.get_all_registered_configs
27+
neural_compressor.tensorflow.keras.quantization.config.get_default_static_quant_config
28+
29+
30+
31+
.. py:class:: StaticQuantConfig(weight_dtype: str = 'int8', weight_sym: bool = True, weight_granularity: str = 'per_tensor', act_dtype: str = 'int8', act_sym: bool = True, act_granularity: str = 'per_tensor', white_list: Optional[List[neural_compressor.common.base_config.OP_NAME_OR_MODULE_TYPE]] = DEFAULT_WHITE_LIST)
32+
33+
34+
35+
36+
Config class for keras static quantization.
37+
38+
39+
.. py:function:: get_all_registered_configs() -> Dict[str, neural_compressor.common.base_config.BaseConfig]
40+
41+
Get all registered configs for keras framework.
42+
43+
44+
.. py:function:: get_default_static_quant_config() -> StaticQuantConfig
45+
46+
Generate the default static quant config.
47+
48+
:returns: the default keras config.
49+
50+

‎latest/_sources/autoapi/neural_compressor/tensorflow/keras/quantization/index.rst.txt

+8
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,8 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.keras.quantization`
4+
=========================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.keras.quantization
7+
8+

‎latest/_sources/autoapi/neural_compressor/tensorflow/quantization/algorithm_entry/index.rst.txt

+2-2
Original file line numberDiff line numberDiff line change
@@ -15,11 +15,11 @@ Functions
1515

1616
.. autoapisummary::
1717

18-
neural_compressor.tensorflow.quantization.algorithm_entry.static_quantize_entry
18+
neural_compressor.tensorflow.quantization.algorithm_entry.static_quant_entry
1919

2020

2121

22-
.. py:function:: static_quantize_entry(model: neural_compressor.tensorflow.utils.BaseModel, quant_config: neural_compressor.tensorflow.quantization.config.StaticQuantConfig, calib_dataloader: Callable = None, calib_iteration: int = 100)
22+
.. py:function:: static_quant_entry(model: neural_compressor.tensorflow.utils.BaseModel, quant_config: neural_compressor.common.base_config.BaseConfig, calib_dataloader: Callable = None, calib_iteration: int = 100)
2323
2424
The main entry to apply static quantization.
2525

‎latest/_sources/autoapi/neural_compressor/tensorflow/quantization/autotune/index.rst.txt

+28
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,28 @@
1+
:orphan:
2+
3+
:py:mod:`neural_compressor.tensorflow.quantization.autotune`
4+
============================================================
5+
6+
.. py:module:: neural_compressor.tensorflow.quantization.autotune
7+
8+
9+
Module Contents
10+
---------------
11+
12+
13+
Functions
14+
~~~~~~~~~
15+
16+
.. autoapisummary::
17+
18+
neural_compressor.tensorflow.quantization.autotune.get_all_config_set
19+
neural_compressor.tensorflow.quantization.autotune.autotune
20+
21+
22+
23+
24+
.. py:function:: autotune(model: Union[str, tensorflow.keras.Model, neural_compressor.tensorflow.utils.BaseModel], tune_config: neural_compressor.common.base_tuning.TuningConfig, eval_fns: Optional[Union[Dict, List[Dict]]] = None, calib_dataloader: Callable = None, calib_iteration: int = 100) -> Optional[neural_compressor.tensorflow.utils.BaseModel]
25+
26+
The main entry of auto-tune.
27+
28+

‎latest/_sources/autoapi/neural_compressor/tensorflow/quantization/config/index.rst.txt

+3-9
Original file line numberDiff line numberDiff line change
@@ -24,30 +24,24 @@ Functions
2424

2525
.. autoapisummary::
2626

27-
neural_compressor.tensorflow.quantization.config.get_all_registered_configs
2827
neural_compressor.tensorflow.quantization.config.get_default_static_quant_config
2928
neural_compressor.tensorflow.quantization.config.get_default_sq_config
3029

3130

3231

33-
.. py:class:: StaticQuantConfig(weight_dtype: str = 'int8', weight_sym: bool = True, weight_granularity: str = 'per_tensor', act_dtype: str = 'int8', act_sym: bool = True, act_granularity: str = 'per_tensor', white_list: Optional[List[neural_compressor.common.base_config.OP_NAME_OR_MODULE_TYPE]] = DEFAULT_WHITE_LIST)
32+
.. py:class:: StaticQuantConfig(weight_dtype: str = 'int8', weight_sym: bool = True, weight_granularity: str = 'per_tensor', weight_algorithm: str = 'minmax', act_dtype: str = 'int8', act_sym: bool = True, act_granularity: str = 'per_tensor', act_algorithm: str = 'minmax', white_list: Optional[List[neural_compressor.common.base_config.OP_NAME_OR_MODULE_TYPE]] = DEFAULT_WHITE_LIST)
3433
3534
3635
3736
38-
Config class for keras static quantization.
39-
40-
41-
.. py:function:: get_all_registered_configs() -> Dict[str, neural_compressor.common.base_config.BaseConfig]
42-
43-
Get all registered configs for keras framework.
37+
Config class for tf static quantization.
4438

4539

4640
.. py:function:: get_default_static_quant_config() -> StaticQuantConfig
4741
4842
Generate the default static quant config.
4943

50-
:returns: the default keras config.
44+
:returns: the default tf config.
5145

5246

5347
.. py:class:: SmoothQuantConfig(alpha: float = 0.5, folding: bool = False, percentile: float = 99.999, op_types: list = ['MatMul', 'Conv2D'], scales_per_op: bool = True, record_max_info: bool = False, weight_clip: bool = True, auto_alpha_args: Dict = DEFAULT_SQ_ALPHA_ARGS, white_list: Optional[List[neural_compressor.common.base_config.OP_NAME_OR_MODULE_TYPE]] = DEFAULT_WHITE_LIST)

‎latest/_sources/autoapi/neural_compressor/tensorflow/quantization/quantize/index.rst.txt

+15-1
Original file line numberDiff line numberDiff line change
@@ -16,14 +16,28 @@ Functions
1616
.. autoapisummary::
1717

1818
neural_compressor.tensorflow.quantization.quantize.quantize_model
19+
neural_compressor.tensorflow.quantization.quantize.quantize_model_with_single_config
1920

2021

2122

22-
.. py:function:: quantize_model(model: Union[str, tensorflow.keras.Model, neural_compressor.tensorflow.utils.BaseModel], quant_config: neural_compressor.common.base_config.BaseConfig, calib_dataloader: Callable = None, calib_iteration: int = 100)
23+
.. py:function:: quantize_model(model: Union[str, tensorflow.keras.Model, neural_compressor.tensorflow.utils.BaseModel], quant_config: Union[neural_compressor.common.base_config.BaseConfig, list], calib_dataloader: Callable = None, calib_iteration: int = 100)
2324
2425
The main entry to quantize model.
2526

2627
:param model: a fp32 model to be quantized.
28+
:param quant_config: single or lists of quantization configuration.
29+
:param calib_dataloader: a data loader for calibration.
30+
:param calib_iteration: the iteration of calibration.
31+
32+
:returns: the quantized model.
33+
:rtype: q_model
34+
35+
36+
.. py:function:: quantize_model_with_single_config(q_model: neural_compressor.tensorflow.utils.BaseModel, quant_config: neural_compressor.common.base_config.BaseConfig, calib_dataloader: Callable = None, calib_iteration: int = 100)
37+
38+
Quantize model using single config.
39+
40+
:param model: a model wrapped by INC TF model class.
2741
:param quant_config: a quantization configuration.
2842
:param calib_dataloader: a data loader for calibration.
2943
:param calib_iteration: the iteration of calibration.

‎latest/_sources/autoapi/neural_compressor/tensorflow/quantization/utils/graph_converter/index.rst.txt

+30
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,30 @@
1+
:py:mod:`neural_compressor.tensorflow.quantization.utils.graph_converter`
2+
=========================================================================
3+
4+
.. py:module:: neural_compressor.tensorflow.quantization.utils.graph_converter
5+
6+
.. autoapi-nested-parse::
7+
8+
Graph Converter Class.
9+
10+
11+
12+
Module Contents
13+
---------------
14+
15+
Classes
16+
~~~~~~~
17+
18+
.. autoapisummary::
19+
20+
neural_compressor.tensorflow.quantization.utils.graph_converter.GraphConverter
21+
22+
23+
24+
25+
.. py:class:: GraphConverter(model, qt_config={}, recipes={}, int8_sequences={}, fp32_ops=[], bf16_ops=[], data_loader=None, calib_func=None, fake_quant=False, itex_mode=False, qdq_enabled=False, new_api=False, performance_only=False, use_bf16=False)
26+
27+
28+
Graph Converter Class is used to generate the quantization graph.
29+
30+

‎latest/_sources/autoapi/neural_compressor/tensorflow/quantization/utils/graph_converter_without_calib/index.rst.txt

+30
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,30 @@
1+
:py:mod:`neural_compressor.tensorflow.quantization.utils.graph_converter_without_calib`
2+
=======================================================================================
3+
4+
.. py:module:: neural_compressor.tensorflow.quantization.utils.graph_converter_without_calib
5+
6+
.. autoapi-nested-parse::
7+
8+
Without calibration Graph Converter Class.
9+
10+
11+
12+
Module Contents
13+
---------------
14+
15+
Classes
16+
~~~~~~~
17+
18+
.. autoapisummary::
19+
20+
neural_compressor.tensorflow.quantization.utils.graph_converter_without_calib.GraphConverterWithoutCalib
21+
22+
23+
24+
25+
.. py:class:: GraphConverterWithoutCalib(model, data_loader=None, recover_config=None, new_api=False, performance_only=False, use_bf16=False)
26+
27+
28+
Graph Converter without calibration Class is used to generate the quantization graph without calibration.
29+
30+

0 commit comments

Comments
 (0)