init

download_audio.py

@@ -102,6 +102,7 @@ def get_audio(dataframe: pd.DataFrame):
         features[f"{side}.path"] = str(p_join(cache_dir_audio, side, f"{features['line_no']}.{identifier}"))
         start, end = features[f"{side}.duration_start"], features[f"{side}.duration_end"]
         if not os.path.exists(features[f"{side}.path"]):
+            print(f"WGET {features[f'{side}.url']}")
             flag = wget(features[f"{side}.url"], output_file=features[f"{side}.path"])
             if not flag:
                 print("\n#### ERROR: wget failure ####\n")
@@ -112,9 +113,10 @@ def get_audio(dataframe: pd.DataFrame):
                 return None
             else:
                 try:
+                    print(f"LOAD AUDIO FROM {features[f'{side}.path']}")
                     wav = audio_loader.decode_example({"path": features[f"{side}.path"], "bytes": None})
                     if wav["sampling_rate"] != sampling_rate:
-                        print(f"RESAMPLING: {wav['array']} length audio")
+                        print(f"RESAMPLING: {len(wav['array'])} length audio")
                         if torch.cuda.is_available():
                             wav["array"] = julius.resample_frac(
                                 torch.as_tensor(wav["array"]).to("cuda"),

main.sh

@@ -2,6 +2,8 @@ rm -rf download/audio
 rm -rf download/feature
 python -c 'n=41; import os; from glob import glob; tmp = [int(os.path.basename(i).split(".")[0]) for i in glob("download/feature/enA-jaA/*.json")]; print(len([x for x in tmp if (n-1) * 2500 <= x < n * 2500]))'
 python -c 'n=41; import os; from glob import glob; tmp = [int(os.path.basename(i).split(".")[0]) for i in glob("download/audio/enA-jaA/enA/*")]; print(len([x for x in tmp if (n-1) * 2500 <= x < n * 2500]))'
+export CUDA_VISIBLE_DEVICES=
+file_name='download/audio/enA-jaA/jaA/100020.mp3'; from datasets import Audio; a=Audio(); a.decode_example({"path": file_name, "bytes": None})
 ####################
 # enA-jaA: 718_606 #
 ####################