use normalised dataset

audio_classification_scripts/run_mms_lid.sh

@@ -2,11 +2,11 @@
 
 python run_audio_classification.py \
     --model_name_or_path "facebook/mms-lid-126" \
-    --train_dataset_name "stable-speech/concatenated-accent-dataset" \
+    --train_dataset_name "stable-speech/concatenated-normalized-accent-dataset" \
     --train_dataset_config_name "default" \
     --train_split_name "train" \
     --train_label_column_name "labels" \
-    --eval_dataset_name "stable-speech/concatenated-accent-dataset" \
+    --eval_dataset_name "stable-speech/concatenated-normalized-accent-dataset" \
     --eval_dataset_config_name "default" \
     --eval_split_name "test" \
     --eval_label_column_name "labels" \
@@ -22,7 +22,7 @@ python run_audio_classification.py \
     --min_length_seconds 5 \
     --attention_mask \
     --warmup_steps 100 \
-    --max_steps 1000 \
+    --max_steps 2000 \
     --per_device_train_batch_size 32 \
     --per_device_eval_batch_size 32 \
     --preprocessing_num_workers 4 \
@@ -31,8 +31,8 @@ python run_audio_classification.py \
     --logging_steps 10 \
     --evaluation_strategy "steps" \
     --eval_steps 500 \
-    --save_strategy "steps" \
-    --save_steps 1000 \
-    --freeze_base_model False \
+    --save_strategy "no" \
+    --save_steps 2000 \
+    --freeze_base_model True \
     --push_to_hub False \
     --trust_remote_code

run_audio_classification.py

@@ -71,6 +71,21 @@ ACCENT_MAPPING = {
     "Northern irish": "Irish",
     "New zealand": "Australian",
     "Pakistani": "Indian",
+    "Mainstream us english": "American",
+    "Southern british english": "English",
+    "Indian english": "Indian",
+    "Scottish english": "Scottish",
+    "Don't know": "Unknown",
+    "Nigerian english": "Nigerian",
+    "Kenyan english": "Kenyan",
+    "Ghanain english": "Ghanain",
+    "Jamaican english": "Jamaican",
+    "Indonesian english": "Indonesian",
+    "South african english": "South african",
+    "Irish english": "Irish",
+    "Latin": "Latin American",
+    "European": "Unknown",  # Too general
+    "Eastern european": "Eastern european", # TODO(SG): keep for now, but maybe remove later
 }
 
 
@@ -599,17 +614,6 @@ def main():
     sampling_rate = feature_extractor.sampling_rate
     model_input_name = feature_extractor.model_input_names[0]
 
-    # filter training data with non-valid labels
-    def is_label_valid(label):
-        return label != "Unknown"
-
-    raw_datasets = raw_datasets.filter(
-        is_label_valid,
-        input_columns=["labels"],
-        num_proc=data_args.preprocessing_num_workers,
-        desc="Filtering by labels",
-    )
-
     def prepare_dataset(batch):
         batch["length"] = len(batch["audio"]["array"])
         batch["labels"] = preprocess_labels(batch["labels"])
@@ -634,6 +638,17 @@ def main():
         desc="Filtering by audio length",
     )
 
+    # filter training data with non-valid labels
+    def is_label_valid(label):
+        return label != "Unknown"
+
+    raw_datasets = raw_datasets.filter(
+        is_label_valid,
+        input_columns=["labels"],
+        num_proc=data_args.preprocessing_num_workers,
+        desc="Filtering by labels",
+    )
+
     # Print a summary of the labels to the stddout (helps identify low-label classes that could be filtered)
     # sort by freq
     count_labels_dict = Counter(raw_datasets["train"]["labels"])
@@ -650,11 +665,11 @@ def main():
         if freq < data_args.filter_threshold:
             labels_to_remove.append(lab)
 
-    # filter training data with label freq below threshold
-    def is_label_valid(label):
-        return label not in labels_to_remove
-
     if len(labels_to_remove):
+        # filter training data with label freq below threshold
+        def is_label_valid(label):
+            return label not in labels_to_remove
+
         raw_datasets = raw_datasets.filter(
             is_label_valid,
             input_columns=["labels"],