Merge 7494bd9 into a1e4942

danlp/models/allennlp_models/coref/dataset_readers/conll.py

@@ -5,7 +5,7 @@
 from overrides import overrides
 
 from allennlp.common.file_utils import cached_path
-from allennlp.data.dataset_readers.dataset_reader import DatasetReader, AllennlpDataset
+from allennlp.data.dataset_readers.dataset_reader import DatasetReader
 from allennlp.data.instance import Instance
 from allennlp.data.tokenizers import PretrainedTransformerTokenizer
 from allennlp.data.token_indexers import SingleIdTokenIndexer, TokenIndexer
@@ -112,7 +112,7 @@ def load_dataset(self, dataset: str):
         if not isinstance(instances, list):
             instances = list(instances)
 
-        return AllennlpDataset(instances)
+        return instances
 
     def _process_sentences(self, sentences):
         clusters: DefaultDict[int, List[Tuple[int, int]]] = collections.defaultdict(list)

examples/benchmarks/coreference_benchmarks.py

@@ -1,10 +1,12 @@
 
 import time
-from utils import print_speed_performance
+from .utils import print_speed_performance
 
 from danlp.datasets import Dacoref
 from danlp.models import load_xlmr_coref_model
 
+from allennlp.data.data_loaders import SimpleDataLoader
+
 import os
 
 # load the data
@@ -21,11 +23,10 @@ def benchmark_xlmr_mdl():
     from allennlp.training.util import evaluate
 
     xlmr = load_xlmr_coref_model()
-    data_loader_params = xlmr.config.pop("data_loader")
-
+
     instances = xlmr.dataset_reader.load_dataset(testset)
-    instances.index_with(xlmr.model.vocab)
-    data_loader = DataLoader.from_params(dataset=instances, params=data_loader_params)
+    data_loader = SimpleDataLoader(instances, 1)
+    data_loader.index_with(xlmr.model.vocab)
 
     start = time.time()
 

examples/benchmarks/requirements_benchmarks.txt

@@ -1,11 +1,11 @@
 tqdm==4.42.1
 spacy==2.2.3
 gensim==3.8.1
-flair==0.4.5
+flair==0.8
 pyconll==2.2.1
 pandas==1.0.1
-transformers==3.5.1
-torch==1.6.0
+transformers==4.3.3
+torch==1.7.1
 srsly==1.0.2
 sentida==0.5.0
 Morfessor==2.0.6
@@ -18,5 +18,5 @@ seqeval
 tweepy
 conllu
 NERDA
-allennlp==1.3.0
+allennlp==2.5.0
 stanza

requirements.txt

@@ -1,12 +1,12 @@
 tqdm==4.42.1
 spacy==2.2.3
 gensim==3.8.1
-torch==1.6.0
-flair==0.4.5
+torch==1.7.1
+flair==0.8
 pyconll==2.2.1
 conllu==4.4
 pandas==1.0.1
-transformers==4.0.1
+transformers==4.3.3
 srsly==1.0.2
 tweepy
-allennlp==1.3.0
+allennlp==2.5.0

setup.py

@@ -26,12 +26,12 @@
     ],
     extras_require={
         'all' : [
-            'transformers<=4.0.1',
+            'transformers<=4.3.3',
             'gensim<=3.8.1',
-            'torch<=1.6.0',
-            'flair<=0.4.5',
+            'torch<=1.7.1',
+            'flair<=0.8',
             'spacy<=2.2.3',
-            'allennlp<=1.3.0'
+            'allennlp<=2.5.0'
         ]
     },
     classifiers=[

tests/test_xlmr_models.py

@@ -1,10 +1,11 @@
+
 import unittest
 
 from danlp.models import load_xlmr_coref_model
 from danlp.download import DEFAULT_CACHE_DIR, download_model, \
     _unzip_process_func
 
-from allennlp.data import DataLoader
+from allennlp.data.data_loaders import SimpleDataLoader 
 from allennlp.training.util import evaluate
 
 import os
@@ -56,8 +57,8 @@ def test_model(self):
             ]
 
         instances = xlmr_model.dataset_reader.load_dataset(sentences)
-        instances.index_with(xlmr_model.model.vocab)
-        data_loader = DataLoader.from_params(dataset=instances, params=data_loader_params)
+        data_loader = SimpleDataLoader(instances, 1)
+        data_loader.index_with(xlmr_model.model.vocab)
 
         metrics = evaluate(
             xlmr_model.model,