Merge remote-tracking branch 'origin/master' into vision

Author: dirkgr

CHANGELOG.md

@@ -38,6 +38,11 @@ dataset at every epoch) and a `MultiTaskScheduler` (for ordering the instances w
 
 ## Unreleased (1.x branch)
 
+### Fixed
+
+- `GumbelSampler` now sorts the beams by their true log prob.
+
+
 ## [v1.2.1](https://github.com/allenai/allennlp/releases/tag/v1.2.1) - 2020-11-10
 
 ### Added
@@ -48,7 +53,7 @@ dataset at every epoch) and a `MultiTaskScheduler` (for ordering the instances w
 - Added more documentation about plugins.
 - Added sampler class and parameter in beam search for non-deterministic search, with several
   implementations, including `MultinomialSampler`, `TopKSampler`, `TopPSampler`, and
-  `GumbelMaxSampler`. Utilizing `GumbelMaxSampler` will give [Stochastic Beam Search](https://api.semanticscholar.org/CorpusID:76662039).
+  `GumbelSampler`. Utilizing `GumbelSampler` will give [Stochastic Beam Search](https://api.semanticscholar.org/CorpusID:76662039).
 
 ### Changed
 
@@ -67,6 +72,8 @@ dataset at every epoch) and a `MultiTaskScheduler` (for ordering the instances w
 - Fixed typo with registered name of ROUGE metric. Previously was `rogue`, fixed to `rouge`.
 - Fixed default masks that were erroneously created on the CPU even when a GPU is available.
 - Fixed pretrained embeddings for transformers that don't use end tokens.
+- Fixed the transformer tokenizer cache when the tokenizers are initialized with custom kwargs.
+
 
 ## [v1.2.0](https://github.com/allenai/allennlp/releases/tag/v1.2.0) - 2020-10-29
 

allennlp/common/cached_transformers.py

@@ -94,11 +94,13 @@ def strip_prefix(s):
         return transformer
 
 
-_tokenizer_cache: Dict[Tuple[str, frozenset], transformers.PreTrainedTokenizer] = {}
+_tokenizer_cache: Dict[Tuple[str, str], transformers.PreTrainedTokenizer] = {}
 
 
 def get_tokenizer(model_name: str, **kwargs) -> transformers.PreTrainedTokenizer:
-    cache_key = (model_name, frozenset(kwargs.items()))
+    from allennlp.common.util import hash_object
+
+    cache_key = (model_name, hash_object(kwargs))
 
     global _tokenizer_cache
     tokenizer = _tokenizer_cache.get(cache_key, None)

allennlp/common/util.py

@@ -1,6 +1,9 @@
 """
 Various utilities that don't fit anywhere else.
 """
+import hashlib
+import io
+import pickle
 from datetime import timedelta
 import importlib
 import json
@@ -679,3 +682,12 @@ def cycle_iterator_function(iterator_function: Callable[[], Iterable[T]]) -> Ite
             yield next(iterator)
         except StopIteration:
             iterator = iter(iterator_function())
+
+
+def hash_object(o: Any) -> str:
+    """Returns a 32-character hash code of arbitrary Python objects."""
+    m = hashlib.blake2b()
+    with io.BytesIO() as buffer:
+        pickle.dump(o, buffer)
+        m.update(buffer.getbuffer())
+        return m.hexdigest()

allennlp/nn/beam_search.py

@@ -385,12 +385,18 @@ def sample_beams(
         # shape (both): (batch_size, beam_size)
         G_phi_S_new, selected_indices = torch.topk(G_phi_S, beam_size, dim=-1)
 
-        # shape: (batch_size * beam_size,)
-        G_phi_S_new = G_phi_S_new.reshape(batch_size * beam_size)
-
         # shape: (batch_size, beam_size)
         selected_log_probs = log_probs.gather(1, selected_indices)
 
+        # Now sort the selected beams by their true log prob.
+        # shape (all): (batch_size, beam_size)
+        selected_log_probs, sort_indices = selected_log_probs.sort(dim=-1, descending=True)
+        selected_indices = selected_indices.gather(1, sort_indices)
+        G_phi_S_new = G_phi_S_new.gather(1, sort_indices)
+
+        # shape: (batch_size * beam_size,)
+        G_phi_S_new = G_phi_S_new.reshape(batch_size * beam_size)
+
         # shape: (batch_size * beam_size,)
         phi_S = selected_log_probs.reshape(batch_size * beam_size)
 

tests/nn/beam_search_test.py

@@ -434,10 +434,14 @@ def test_gumbel_sampler(self):
         num_classes = len(log_probabilities[0])
         sampler_state = sampler.init_state(log_probabilities, batch_size=2, num_classes=num_classes)
 
-        probabilities, classes, state = sampler.sample_beams(log_probabilities, 3, sampler_state)
+        log_probs, indices, state = sampler.sample_beams(log_probabilities, 3, sampler_state)
 
-        assert probabilities.size() == classes.size()
-        assert classes.size() == (2, 3)
+        assert log_probs.size() == indices.size()
+        assert indices.size() == (2, 3)
+
+        # Make sure the probabilities are sorted.
+        _, sorted_indices = log_probs.sort(dim=-1, descending=True)
+        assert (sorted_indices == torch.arange(3).unsqueeze(0)).all()
 
-        assert all([x >= 0 and x < 4 for x in classes[0]])
-        assert all([x > 1 and x <= 5 for x in classes[1]])
+        assert all([x >= 0 and x < 4 for x in indices[0]])
+        assert all([x > 1 and x <= 5 for x in indices[1]])