are there any additional curl parameters for aws-whisper-large inference end point?
otherwise any uploaded wav or flac files that are larger than a few words get truncated:
“The Witsber model was proposed in robust speech recognition via large scale weak supervision.”
instead of:
“The Witsber model was proposed in robust speech recognition via large scale weak supervision. The first paragraphs of the abstract read as follows: We study the capabilities of speech processing systems trained simply to predict large amounts of transcripts of audio on the internet. When scaled to 680,000 hours of multilingual and multitask supervision, the resulting models generalize well to standard benchmarks and are often competitive with prior fully supervised results but in a zeroshot transfer setting without the need for any finetuning. When compared to humans, the models approach their accuracy and robustness. We are releasing models and inference code to serve as a foundation for further work on robust speech processing.”
kind of a bummer if we cannot fix this.
logs say this:
lc8gt 2022-10-16T23:43:04.560Z /opt/conda/lib/python3.9/site-packages/transformers/pipelines/base.py:1043: UserWarning: You seem to be using the pipelines sequentially on GPU. In order to maximize efficiency please use a dataset
lc8gt 2022-10-16T23:43:04.560Z warnings.warn(
lc8gt 2022-10-16T23:43:04.829Z warnings.warn(
lc8gt 2022-10-16T23:43:04.829Z /opt/conda/lib/python3.9/site-packages/transformers/generation_utils.py:1296: UserWarning: Neither max_length
nor max_new_tokens
has been set, max_length
will default to 20 (self.config.max_length
). Controlling max_length
via the config is deprecated and max_length
will be removed from the config in v5 of Transformers – we recommend using max_new_tokens
to control the maximum length of the generation.
lc8gt 2022-10-16T23:43:05.314Z 2022-10-16 23:43:05,314 | INFO | POST / | Duration: 757.65 ms