tanda-sequential-finetuning-with-asnq.diff

diff --git a/examples/run_glue.py b/examples/run_glue.py
index 19316cb..bef98bb 100644
--- a/examples/run_glue.py
+++ b/examples/run_glue.py
@@ -354,6 +354,8 @@ def main(): @@ this change adds `sequential fine-tuning` option
                              "than this will be truncated, sequences shorter will be padded.")
     parser.add_argument("--do_train", action='store_true',
                         help="Whether to run training.")
+    parser.add_argument("--sequential", action='store_true',
+                        help="Sequential fine-tune from a checkpoint.")
     parser.add_argument("--do_eval", action='store_true',
                         help="Whether to run eval on the dev set.")
     parser.add_argument("--evaluate_during_training", action='store_true',
@@ -488,7 +490,10 @@ def main(): @@ this change adds `sequential fine-tuning` option
     tokenizer = tokenizer_class.from_pretrained(args.tokenizer_name if args.tokenizer_name else args.model_name_or_path,
                                                 do_lower_case=args.do_lower_case,
                                                 cache_dir=args.cache_dir if args.cache_dir else None)
-    model = model_class.from_pretrained(args.model_name_or_path,
+    if args.sequential:
+        model = model_class.from_pretrained(args.output_dir)
+    else:
+        model = model_class.from_pretrained(args.model_name_or_path,
                                         from_tf=bool('.ckpt' in args.model_name_or_path),
                                         config=config,
                                         cache_dir=args.cache_dir if args.cache_dir else None)
diff --git a/transformers/data/metrics/__init__.py b/transformers/data/metrics/__init__.py
index c9ebaac..eddb931 100644
--- a/transformers/data/metrics/__init__.py
+++ b/transformers/data/metrics/__init__.py
@@ -79,5 +79,7 @@ if _has_sklearn: @@ this change adds processing for `asnq`
             return {"acc": simple_accuracy(preds, labels)}
         elif task_name == "wnli":
             return {"acc": simple_accuracy(preds, labels)}
+        elif task_name == "asnq":
+            return {"acc": simple_accuracy(preds, labels)}
         else:
             raise KeyError(task_name)
diff --git a/transformers/data/processors/glue.py b/transformers/data/processors/glue.py
index 518251b..de77dca 100644
--- a/transformers/data/processors/glue.py
+++ b/transformers/data/processors/glue.py
@@ -513,6 +513,46 @@ class WnliProcessor(DataProcessor): @@ this change adds processing for `asnq`
                 InputExample(guid=guid, text_a=text_a, text_b=text_b, label=label))
         return examples
 
+
+class AsnqProcessor(DataProcessor):
+    """Processor for the ASNQ data set (for TANDA)."""
+
+    def get_example_from_tensor_dict(self, tensor_dict):
+        """See base class."""
+        return InputExample(tensor_dict['idx'].numpy(),
+                            tensor_dict['sentence'].numpy().decode('utf-8'),
+                            None,
+                            str(tensor_dict['label'].numpy()))
+
+    def get_train_examples(self, data_dir):
+        """See base class."""
+        return self._create_examples(
+            self._read_tsv(os.path.join(data_dir, "train.tsv")), "train")
+
+    def get_dev_examples(self, data_dir):
+        """See base class."""
+        return self._create_examples(
+            self._read_tsv(os.path.join(data_dir, "dev.tsv")), "dev")
+
+    def get_labels(self):
+        """See base class."""
+        return ["0", "1"]
+
+    def _create_examples(self, lines, set_type):
+        """Creates examples for the training and dev sets."""
+        examples = []
+        for (i, line) in enumerate(lines):
+            guid = "%s-%s" % (set_type, i)
+            text_a = line[0]
+            text_b = line[1]
+            if eval(line[-1].strip()) == 4:
+                label = "1"
+            else:
+                label = "0"
+            examples.append(
+                InputExample(guid=guid, text_a=text_a, text_b=None, label=label))
+        return examples
+
 glue_tasks_num_labels = {
     "cola": 2,
     "mnli": 3,
@@ -523,6 +563,7 @@ glue_tasks_num_labels = {
     "qnli": 2,
     "rte": 2,
     "wnli": 2,
+    "asnq": 2,
 }
 
 glue_processors = {
@@ -536,6 +577,7 @@ glue_processors = {
     "qnli": QnliProcessor,
     "rte": RteProcessor,
     "wnli": WnliProcessor,
+    "asnq": AsnqProcessor,
 }
 
 glue_output_modes = {
@@ -549,4 +591,5 @@ glue_output_modes = {
     "qnli": "classification",
     "rte": "classification",
     "wnli": "classification",
+    "asnq": "classification",
 }