Explicitly pass serialization directory and local rank to trainer in train command

CHANGELOG.md

@@ -22,10 +22,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
   See [PR #5172](https://github.com/allenai/allennlp/pull/5172) for more details.
 - Added `SpanExtractorWithSpanWidthEmbedding`, putting specific span embedding computations into the `_embed_spans` method and leaving the common code in `SpanExtractorWithSpanWidthEmbedding` to unify the arguments, and modified `BidirectionalEndpointSpanExtractor`, `EndpointSpanExtractor` and `SelfAttentiveSpanExtractor` accordingly. Now, `SelfAttentiveSpanExtractor` can also embed span widths.
 
-
 ### Fixed
 
 - When `PretrainedTransformerIndexer` folds long sequences, it no longer loses the information from token type ids.
+- Fixed a bug where `local_rank` wasn't passed to the trainer during distributed trainer.
 
 
 ## [v2.4.0](https://github.com/allenai/allennlp/releases/tag/v2.4.0) - 2021-04-22

allennlp/commands/train.py

@@ -426,6 +426,7 @@ def _train_worker(
 
         # Till now, "cuda_device" might not be set in the trainer params.
         # But a worker trainer needs to only know about its specific GPU id.
+        params["trainer"]["local_rank"] = process_rank
         params["trainer"]["cuda_device"] = gpu_id
         params["trainer"]["world_size"] = world_size
         params["trainer"]["distributed"] = True