Open stingerpk opened 7 years ago
Hello, I am using evaluating using the following command:
python evaluate.py --manifest val:/path/to/manifest.csv --model_file /path/to/librispeech_16_epochs.prm --use_wer --inference_file /path/to/results.pkl
I get the following error:
Traceback (most recent call last): File "evaluate.py", line 103, in <module> use_wer=args.use_wer) File "/home/ubuntu/nervana/deepspeech/speech/utils.py", line 75, in get_wer progress_string = get_progress_string(be.bsz, batchcount, nbatches) File "/home/ubuntu/nervana/deepspeech/speech/utils.py", line 23, in get_progress_string bar_width = int(float(bcount) / nbatches * max_bar_width) ZeroDivisionError: float division by zero
I am using current master versions of aeon, neon and deepspeech.
Could you show a few lines of your manifest file? That error indicates that nbatches is 0, but that implies that your dataset has 0 items in it.
nbatches
Hello, I am using evaluating using the following command:
python evaluate.py --manifest val:/path/to/manifest.csv --model_file /path/to/librispeech_16_epochs.prm --use_wer --inference_file /path/to/results.pkl
I get the following error:
I am using current master versions of aeon, neon and deepspeech.