|jvddlmZddlZddlmZddlmZddlmZer ddl m Z ddl m Z Gd d eZ dS) ) annotationsN) TYPE_CHECKING)fleet) Quantization)Layer) QuantConfigc4eZdZdZdfd ZdZddd ZxZS)PTQz; Applying post training quantization to the model. configr returnNonecJt|dS)N)super__init__)selfr __class__s g/lsinfo/ai/hellotax_ai/data_center/backend/venv/lib/python3.11/site-packages/paddle/quantization/ptq.pyrz PTQ.__init__"s!      c\ tjdkrdSdS#t$rYdSwxYw)NTF)r worker_num Exception)rs r_is_parallel_trainingzPTQ._is_parallel_training%sG !!A%%tu   55 s  ++Fmodelrinplaceboolc\|}|sF|r Jdtj|}||jr Jd|j||||j|||j|S)a9 Create a model for post-training quantization. The quantization configuration will be propagated in the model. And it will insert observers into the model to collect and compute quantization parameters. Args: model(Layer): The model to be quantized. inplace(bool): Whether to modify the model in-place. Return: The prepared model for post-training quantization. Examples: .. code-block:: python >>> from paddle.quantization import PTQ, QuantConfig >>> from paddle.quantization.observers import AbsmaxObserver >>> from paddle.vision.models import LeNet >>> observer = AbsmaxObserver() >>> q_config = QuantConfig(activation=observer, weight=observer) >>> ptq = PTQ(q_config) >>> model = LeNet() >>> model.eval() >>> quant_model = ptq.quantize(model) >>> print(quant_model) LeNet( (features): Sequential( (0): QuantedConv2D( (weight_quanter): AbsmaxObserverLayer() (activation_quanter): AbsmaxObserverLayer() ) (1): ObserveWrapper( (_observer): AbsmaxObserverLayer() (_observed): ReLU() ) (2): ObserveWrapper( (_observer): AbsmaxObserverLayer() (_observed): MaxPool2D(kernel_size=2, stride=2, padding=0) ) (3): QuantedConv2D( (weight_quanter): AbsmaxObserverLayer() (activation_quanter): AbsmaxObserverLayer() ) (4): ObserveWrapper( (_observer): AbsmaxObserverLayer() (_observed): ReLU() ) (5): ObserveWrapper( (_observer): AbsmaxObserverLayer() (_observed): MaxPool2D(kernel_size=2, stride=2, padding=0) ) ) (fc): Sequential( (0): QuantedLinear( (weight_quanter): AbsmaxObserverLayer() (activation_quanter): AbsmaxObserverLayer() ) (1): QuantedLinear( (weight_quanter): AbsmaxObserverLayer() (activation_quanter): AbsmaxObserverLayer() ) (2): QuantedLinear( (weight_quanter): AbsmaxObserverLayer() (activation_quanter): AbsmaxObserverLayer() ) ) ) z3'inplace' is not compatible with parallel training.zjPost-Training Quantization should not work on training models. Please set evaluation mode by model.eval().) rcopydeepcopyevaltraining_config_specify_convert_to_quant_layers_insert_activation_observers)rrr_models rquantizez PTQ.quantize.sN 1133  E  3]5))F KKMMM>  x  ! f%%% %%fdl;;; ))&$,??? r)r r r r)F)rrrrr r)__name__ __module__ __qualname____doc__rrr( __classcell__)rs@rr r sx!!!!!!TTTTTTTTTrr ) __future__rrtypingrpaddle.distributedrr(r paddle.nnrr r r rrr3s#""""" $$$$$$""""""$######eeeee,eeeeer