Created by: hello-web

coding=gbk

import os import re import time import logging import json from random import random from tqdm import tqdm from functools import reduce, partial

import numpy as np import logging import argparse

import paddle import paddle.fluid as F import paddle.fluid.dygraph as FD import paddle.fluid.layers as L

from propeller import log import propeller.paddle as propeller

log.setLevel(logging.DEBUG) logging.getLogger().setLevel(logging.DEBUG)

#from model.bert import BertConfig, BertModelLayer #from ernie.modeling_ernie import ErnieModel, ErnieModelForSequenceClassification from ernie.tokenizing_ernie import ErnieTokenizer, ErnieTinyTokenizer from ernie.optimization import AdamW, LinearDecay from ernie.modeling_ernie import ErnieModelForPretraining, ErnieModel

if name == 'main': parser = argparse.ArgumentParser('save_inference_model with ERNIE') parser.add_argument('--from_pretrained', type=str, required=True, help='pretrained model directory or tag') parser.add_argument('--inference_model_dir', type=str, default=None, help='inference model output directory')

args = parser.parse_args()

tokenizer = ErnieTokenizer.from_pretrained(args.from_pretrained)
rev_dict = {v: k for k, v in tokenizer.vocab.items()}
rev_dict[tokenizer.pad_id] = '' # replace [PAD]
rev_dict[tokenizer.sep_id] = '' # replace [PAD]
rev_dict[tokenizer.unk_id] = '' # replace [PAD]

@np.vectorize
def rev_lookup(i):
    return rev_dict[i]

place = F.CPUPlace()
with FD.guard(place):
    model = ErnieModelForPretraining.from_pretrained(args.from_pretrained)

    src, _ = tokenizer.encode('戊戌变法，又称百日维新，是 [MASK] [MASK] [MASK] 、梁启超等维新派人士通过光绪帝进行 的一场资产阶级改良。')
    print(src)
    src_ids = np.expand_dims(src, 0)
    src_ids = FD.to_variable(src_ids)

    if args.inference_model_dir is not None:
        log.debug('saving inference model')
        class InferemceModel(ErnieModelForPretraining):
            def __init__(self, *args, **kwargs):
                super(InferemceModel, self).__init__(*args, **kwargs)
                #del self.pooler_heads
            def forward(self, src_ids, *args, **kwargs):
                #mlm_pos = kwargs.pop('mlm_pos', np.int64(3))
                pooled, encoded = ErnieModel.forward(self, src_ids, *args, **kwargs)
                encoded_2d = L.gather_nd(encoded, L.where(src_ids == 3))
                encoded_2d = self.mlm(encoded_2d)
                encoded_2d = self.mlm_ln(encoded_2d)
                logits_2d = L.matmul(encoded_2d, self.word_emb.weight, transpose_y=True) + self.mlm_bias
                return logits_2d

        model.__class__ = InferemceModel 
        logits = model(src_ids)
        _, static_model = FD.TracedLayer.trace(model, inputs=[src_ids])
        static_model.save_inference_model(args.inference_model_dir)
        
        logits = logits.numpy()
        output_ids = np.argmax(logits, -1)
        seg_txt = rev_lookup(output_ids)
        print(''.join(seg_txt))

以上为复现代码 问题：完型填空的例子能正确运行，导出推理模型不报错，但是推理模型为无效模型，不能进行正确推理。 导出的模型文件参考文件名： model create_parameter_0.b_0 embedding_[0-2].w_0 layer_norm_[0-25].b_0 layer_norm_[0-25].w_0 linear_[0-74].b_0 linear_[0-74].w_0