progress

build.py

@@ -168,11 +168,10 @@ def build_dataset(test_split=0.10, dev_split=0.10):
                     tar.add(audio_dir_path, arcname=os.path.basename(audio_dir_path))
 
                 pbar.set_description(f"Saving transcripts ({dialect}/{split})")
-                # if len(transcripts) > 0:
+
                 with open(os.path.join(transcript_dir_path, f"{split}.tsv"), 'w', newline='') as f:
-                    writer = csv.DictWriter(f, fieldnames=transcripts[0].keys() if len(transcripts) > 0 else [
-                        "client_id","transcription","translation","path","locale","proficiency","age","crowdsourced","unlabelled","interrupted"
-                    ], delimiter='\t')
+                    fieldnames = [] if len(transcripts) == 0 else transcripts[0].keys()
+                    writer = csv.DictWriter(f, fieldnames=fieldnames, delimiter='\t')
                     writer.writeheader()
                     writer.writerows(transcripts)
 

nena_speech_1_0_test.py

@@ -64,7 +64,7 @@ class NENASpeechConfig(datasets.BuilderConfig):
 class NENASpeech(datasets.GeneratorBasedBuilder):
     DEFAULT_WRITER_BATCH_SIZE = 1000
 
-    BUILDER_CONFIGS = [
+    BUILDER_CONFIGS = sorted([
         NENASpeechConfig(
             name=dialect,
             version=STATS["version"],
@@ -78,7 +78,7 @@ class NENASpeech(datasets.GeneratorBasedBuilder):
             # size_bytes=int(dialect_stats["size"]) if dialect_stats["size"] else None,
         )
         for dialect, dialect_stats in STATS["dialects"].items()
-    ]
+    ], key=lambda config: config.num_examples)
 
     def _info(self):
         total_dialects = len(STATS["dialects"])