Source code for easytransfer.postprocessors.comprehension_postprocessors

# coding=utf-8
# Copyright 2018 The Google AI Language Team Authors.
# Copyright (c) 2019 Alibaba PAI team.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import six
if six.PY2:
    import cPickle as pkl
else:
    import pickle as pkl
import collections
import math
import numpy as np
import tensorflow as tf
from easytransfer.engines.distribution import Process
from easytransfer.preprocessors.tokenization import BasicTokenizer, convert_to_unicode


[docs]class ComprehensionPostprocessor(Process): """ Postprocessor for text comprehension, search and improve the answer span """ def __init__(self, output_schema, n_best_size=20, max_answer_length=30, thread_num=None, input_queue=None, output_queue=None, job_name='ComprehentionPostprocessor'): super(ComprehensionPostprocessor, self).__init__( job_name, thread_num, input_queue, output_queue, batch_size=1) self.output_schema = output_schema self.n_best_size = n_best_size self.max_answer_length = max_answer_length @staticmethod def _get_best_indexes(logits, n_best_size): """Get the n-best logits from a list.""" index_and_score = sorted(enumerate(logits), key=lambda x: x[1], reverse=True) best_indexes = [] for i in range(len(index_and_score)): if i >= n_best_size: break best_indexes.append(index_and_score[i][0]) return best_indexes @staticmethod def _get_final_text(pred_text, orig_text, do_lower_case): """Project the tokenized prediction back to the original text.""" # When we created the data, we kept track of the alignment between original # (whitespace tokenized) tokens and our WordPiece tokenized tokens. So # now `orig_text` contains the span of our original text corresponding to the # span that we predicted. # # However, `orig_text` may contain extra characters that we don't want in # our prediction. # # For example, let's say: # pred_text = steve smith # orig_text = Steve Smith's # # We don't want to return `orig_text` because it contains the extra "'s". # # We don't want to return `pred_text` because it's already been normalized # (the SQuAD eval script also does punctuation stripping/lower casing but # our tokenizer does additional normalization like stripping accent # characters). # # What we really want to return is "Steve Smith". # # Therefore, we have to apply a semi-complicated alignment heruistic between # `pred_text` and `orig_text` to get a character-to-charcter alignment. This # can fail in certain cases in which case we just return `orig_text`. def _strip_spaces(text): ns_chars = [] ns_to_s_map = collections.OrderedDict() for (i, c) in enumerate(text): if c == " ": continue ns_to_s_map[len(ns_chars)] = i ns_chars.append(c) ns_text = "".join(ns_chars) return (ns_text, ns_to_s_map) # We first tokenize `orig_text`, strip whitespace from the result # and `pred_text`, and check if they are the same length. If they are # NOT the same length, the heuristic has failed. If they are the same # length, we assume the characters are one-to-one aligned. tokenizer = BasicTokenizer(do_lower_case=do_lower_case) tok_text = " ".join(tokenizer.tokenize(orig_text)) start_position = tok_text.find(pred_text) if start_position == -1: tf.logging.info( "Unable to find text: '%s' in '%s'" % (pred_text, orig_text)) return orig_text end_position = start_position + len(pred_text) - 1 (orig_ns_text, orig_ns_to_s_map) = _strip_spaces(orig_text) (tok_ns_text, tok_ns_to_s_map) = _strip_spaces(tok_text) if len(orig_ns_text) != len(tok_ns_text): tf.logging.info("Length not equal after stripping spaces: '%s' vs '%s'", orig_ns_text, tok_ns_text) return orig_text # We then project the characters in `pred_text` back to `orig_text` using # the character-to-character alignment. tok_s_to_ns_map = {} for (i, tok_index) in six.iteritems(tok_ns_to_s_map): tok_s_to_ns_map[tok_index] = i orig_start_position = None if start_position in tok_s_to_ns_map: ns_start_position = tok_s_to_ns_map[start_position] if ns_start_position in orig_ns_to_s_map: orig_start_position = orig_ns_to_s_map[ns_start_position] if orig_start_position is None: tf.logging.info("Couldn't map start position") return orig_text orig_end_position = None if end_position in tok_s_to_ns_map: ns_end_position = tok_s_to_ns_map[end_position] if ns_end_position in orig_ns_to_s_map: orig_end_position = orig_ns_to_s_map[ns_end_position] if orig_end_position is None: tf.logging.info("Couldn't map end position") return orig_text output_text = orig_text[orig_start_position:(orig_end_position + 1)] return output_text @staticmethod def _compute_softmax(scores): """Compute softmax probability over raw logits.""" if not scores: return [] max_score = None for score in scores: if max_score is None or score > max_score: max_score = score exp_scores = [] total_sum = 0.0 for score in scores: x = math.exp(score - max_score) exp_scores.append(x) total_sum += x probs = [] for score in exp_scores: probs.append(score / total_sum) return probs
[docs] def process(self, in_data): """ Post-process the model outputs Args: in_data (`dict`): a dict of model outputs Returns: ret (`dict`): a dict of post-processed model outputs """ ret = dict() for output_col_name in self.output_schema.split(","): if output_col_name in in_data: ret[output_col_name] = in_data[output_col_name] if "predictions" not in self.output_schema.split(",") and \ "probabilities" not in self.output_schema.split(","): return ret prediction_list = [] probability_list = [] for idx in range(len(in_data["start_logits"])): start_logits = in_data["start_logits"][idx] end_logits = in_data["end_logits"][idx] tokens = [convert_to_unicode(t) for t in in_data["tokens"][idx]] doc_tokens = [convert_to_unicode(t) for t in in_data["doc_tokens"][idx]] token_to_orig_map = {int(key): val for key, val in in_data["token_to_orig_map"][idx].items()} start_indexes = self._get_best_indexes(start_logits, self.n_best_size) end_indexes = self._get_best_indexes(end_logits, self.n_best_size) _PrelimPrediction = collections.namedtuple( # pylint: disable=invalid-name "PrelimPrediction", ["start_index", "end_index", "start_logit", "end_logit"]) prelim_predictions = [] for start_index in start_indexes: for end_index in end_indexes: # We could hypothetically create invalid predictions, e.g., predict # that the start of the span is in the question. We throw out all # invalid predictions. if start_index >= len(tokens): continue if end_index >= len(tokens): continue if start_index not in token_to_orig_map: continue if end_index not in token_to_orig_map: continue if end_index < start_index: continue length = end_index - start_index + 1 if length > self.max_answer_length: continue prelim_predictions.append( _PrelimPrediction( start_index=start_index, end_index=end_index, start_logit=start_logits[start_index], end_logit=end_logits[end_index])) prelim_predictions = sorted( prelim_predictions, key=lambda x: (x.start_logit + x.end_logit), reverse=True) _NbestPrediction = collections.namedtuple( # pylint: disable=invalid-name "NbestPrediction", ["text", "start_logit", "end_logit"]) seen_predictions = {} nbest = [] for pred in prelim_predictions: if len(nbest) >= self.n_best_size: break if pred.start_index > 0: # this is a non-null prediction tok_tokens = tokens[pred.start_index:(pred.end_index + 1)] orig_doc_start = token_to_orig_map[pred.start_index] orig_doc_end = token_to_orig_map[pred.end_index] orig_tokens = doc_tokens[orig_doc_start:(orig_doc_end + 1)] tok_text = " ".join(tok_tokens) # De-tokenize WordPieces that have been split off. tok_text = tok_text.replace(" ##", "") tok_text = tok_text.replace("##", "") # Clean whitespace tok_text = tok_text.strip() tok_text = " ".join(tok_text.split()) orig_text = " ".join(orig_tokens) final_text = self._get_final_text(tok_text, orig_text, True) if final_text in seen_predictions: continue seen_predictions[final_text] = True else: final_text = "" seen_predictions[final_text] = True nbest.append( _NbestPrediction( text=final_text, start_logit=pred.start_logit, end_logit=pred.end_logit)) # if we didn't inlude the empty option in the n-best, inlcude it if "" not in seen_predictions: null_start_logit = start_logits[0] null_end_logit = end_logits[0] nbest.append( _NbestPrediction( text="", start_logit=null_start_logit, end_logit=null_end_logit)) # In very rare edge cases we could have no valid predictions. So we # just create a nonce prediction in this case to avoid failure. if not nbest: nbest.append( _NbestPrediction(text="empty", start_logit=0.0, end_logit=0.0)) assert len(nbest) >= 1 total_scores = [] best_non_null_entry = None for entry in nbest: total_scores.append(entry.start_logit + entry.end_logit) if not best_non_null_entry: if entry.text: best_non_null_entry = entry probs = self._compute_softmax(total_scores) nbest_json = [] for (i, entry) in enumerate(nbest): output = dict() output["text"] = entry.text output["probability"] = str(probs[i]) output["start_logit"] = str(entry.start_logit) output["end_logit"] = str(entry.end_logit) nbest_json.append(output) assert len(nbest_json) >= 1 prediction_list.append(nbest_json[0]["text"].encode("utf8")) probability_list.append(nbest_json[0]["probability"]) if "predictions" in self.output_schema.split(","): ret["predictions"] = np.array(prediction_list) if "probabilities" in self.output_schema.split(","): ret["probabilities"] = np.array(probability_list) return ret
if __name__ == "__main__": with open("tmp.in_data.pkl") as f: in_data = pkl.load(f) obj = ComprehensionPostprocessor(output_schema="predictions,example_id,answer") obj.process(in_data)