upd: added pre-training model setup

rwth_phoenix_weather_2014.py

@@ -69,6 +69,12 @@ class RWTHPhoenixWeather2014(datasets.GeneratorBasedBuilder):
             main_data_folder="phoenix-2014-signerindependent-SI5",
             corpus_file_suffix=".SI5.corpus.csv"
         ),
+        RWTHPhoenixWeather2014Config(
+            name="pre-training",
+            description="",
+            main_data_folder="phoenix-2014-multisigner",
+            corpus_file_suffix=".corpus.csv"
+        ),
     ]
 
     def _info(self):
@@ -115,6 +121,10 @@ class RWTHPhoenixWeather2014(datasets.GeneratorBasedBuilder):
 
             split_ids[split] = df.id
 
+            if self.config.name == 'pre-training':
+                split_frames[split] = None
+                continue
+
             frame_archive_urls = dl_manager.download([
                 f"{base_url}/features/fullFrame-210x260px/{dataDirMapper[split]}/{id}.tar"
                 for id in df.id
@@ -145,9 +155,11 @@ class RWTHPhoenixWeather2014(datasets.GeneratorBasedBuilder):
         for id, frames, tokens in zip(split_ids, split_frames, split_tokens):
             frames_as_list = [
                 {"path": p, "bytes": im.read()} for p, im in frames
-            ]
+            ] if frames is not None else None
 
-            yield id, {
+            result = {
                 "tokens": tokens,
-                "frames": frames_as_list,
+                "frames": frames_as_list if frames_as_list is not None else [],
             }
+
+            yield id,