metadata
language:
- pl
license: apache-2.0
tags:
- automatic-speech-recognition
- mozilla-foundation/common_voice_8_0
- pl
- robust-speech-event
datasets:
- mozilla-foundation/common_voice_8_0
model-index:
- name: XLS-R-1B - Polish
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 8
type: mozilla-foundation/common_voice_8_0
args: pl
metrics:
- name: Test WER
type: wer
value: 11.01
- name: Test CER
type: cer
value: 2.55
- name: Test WER (+LM)
type: wer
value: 7.32
- name: Test CER (+LM)
type: cer
value: 1.95
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Robust Speech Event - Dev Data
type: speech-recognition-community-v2/dev_data
args: pl
metrics:
- name: Test WER
type: wer
value: 26.31
- name: Test CER
type: cer
value: 13.85
- name: Test WER (+LM)
type: wer
value: 20.33
- name: Test CER (+LM)
type: cer
value: 13
XLS-R-1B-POLISH
Fine-tuned facebook/wav2vec2-xls-r-1b on Polish using the Common Voice 8. When using this model, make sure that your speech input is sampled at 16kHz.
This model has been fine-tuned thanks to the GPU credits generously given by the OVHcloud :)
The script used for training can be found here: https://github.com/jonatasgrosman/wav2vec2-sprint
Evaluation Commands
- To evaluate on
mozilla-foundation/common_voice_8_0
with splittest
python eval.py --model_id jonatasgrosman/wav2vec2-xls-r-1b-polish --dataset mozilla-foundation/common_voice_8_0 --config pl --split test
- To evaluate on
speech-recognition-community-v2/dev_data
python eval.py --model_id jonatasgrosman/wav2vec2-xls-r-1b-polish --dataset speech-recognition-community-v2/dev_data --config pl --split validation --chunk_length_s 5.0 --stride_length_s 1.0