speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

[fix] better solution for handling empty result (#1796)

Browse Source
This commit is contained in:
Xingchen Song(宋星辰) 2024-06-11 14:00:10 +08:00 committed by GitHub
parent 5c507c32d2
commit 0bd1a4d6a9
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 5 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
funasr/auto/auto_model.py
View File

@ -388,7 +388,6 @@ class AutoModel:
): ):
max_len_in_batch = max(max_len_in_batch, sample_length) max_len_in_batch = max(max_len_in_batch, sample_length)
end_idx += 1 end_idx += 1
results_sorted.append({'key': 'bad_data', 'text': '', 'timestamp': []})
continue continue
speech_j, speech_lengths_j = slice_padding_audio_samples( speech_j, speech_lengths_j = slice_padding_audio_samples(
@ -418,7 +417,7 @@ class AutoModel:
end_idx += 1 end_idx += 1
max_len_in_batch = sample_length max_len_in_batch = sample_length
if len(results) < 1: if len(results) < 1:
results.append({'key': 'bad_data', 'text': '', 'timestamp': []}) continue
results_sorted.extend(results) results_sorted.extend(results)
# end_asr_total = time.time() # end_asr_total = time.time()
@ -428,6 +427,10 @@ class AutoModel:
# f"time_speech_total_per_sample: {time_speech_total_per_sample: 0.3f}, " # f"time_speech_total_per_sample: {time_speech_total_per_sample: 0.3f}, "
# f"time_escape_total_per_sample: {time_escape_total_per_sample:0.3f}") # f"time_escape_total_per_sample: {time_escape_total_per_sample:0.3f}")
if len(results_sorted) != n:
results_ret_list.append({"key": key, "text": "", "timestamp": []})
logging.info("decoding, utt: {}, empty result".format(key))
continue
restored_data = [0] * n restored_data = [0] * n
for j in range(n): for j in range(n):
index = sorted_data[j][1] index = sorted_data[j][1]