license, tags, library_name, language, base_model, model-index

license	tags	library_name	language	base_model	model-index
apache-2.0	automatic-speech-recognition fi finnish	transformers	fi	GetmanY1/wav2vec2-large-fi-150k	name results wav2vec2-large-fi-150k-finetuned task dataset metrics name type Automatic Speech Recognition automatic-speech-recognition name type args Lahjoita puhetta (Donate Speech) lahjoita-puhetta fi name type value Dev WER wer 15.34 name type value Dev CER cer 4.14 name type value Test WER wer 16.86 name type value Test CER cer 5.07 task dataset metrics name type Automatic Speech Recognition automatic-speech-recognition name type args Finnish Parliament FinParl fi name type value Dev16 WER wer 11.3 name type value Dev16 CER cer 4.75 name type value Test16 WER wer 8.29 name type value Test16 CER cer 3.34 name type value Test20 WER wer 6.94 name type value Test20 CER cer 2.15 task dataset metrics name type Automatic Speech Recognition automatic-speech-recognition name type args Common Voice 16.1 mozilla-foundation/common_voice_16_1 fi name type value Dev WER wer 7.17 name type value Dev CER cer 1.11 name type value Test WER wer 5.86 name type value Test CER cer 0.91 task dataset metrics name type Automatic Speech Recognition automatic-speech-recognition name type args FLEURS google/fleurs fi_fi name type value Dev WER wer 9.2 name type value Dev CER cer 5.23 name type value Test WER wer 10.69 name type value Test CER cer 5.79

Finnish Wav2vec2-Large ASR

GetmanY1/wav2vec2-large-fi-150k fine-tuned on 4600 hours of Finnish speech on 16kHz sampled speech audio:

• 1500 hours of Lahjoita puhetta (Donate Speech) (colloquial Finnish)
• 3100 hours of the Finnish Parliament dataset

When using the model make sure that your speech input is also sampled at 16Khz.

Model description

The Finnish Wav2Vec2 Large has the same architecture and uses the same training objective as the English and multilingual one described in Paper.

GetmanY1/wav2vec2-large-fi-150k is a large-scale, 317-million parameter monolingual model pre-trained on 158k hours of unlabeled Finnish speech, including KAVI radio and television archive materials, Lahjoita puhetta (Donate Speech), Finnish Parliament, Finnish VoxPopuli.

You can read more about the pre-trained model from this paper. The training scripts are available on GitHub.

Intended uses

You can use this model for Finnish ASR (speech-to-text).

How to use

To transcribe audio files the model can be used as a standalone acoustic model as follows:

from transformers import Wav2Vec2Processor, Wav2Vec2ForCTC
from datasets import load_dataset
import torch

# load model and processor
processor = Wav2Vec2Processor.from_pretrained("GetmanY1/wav2vec2-large-fi-150k-finetuned")
model = Wav2Vec2ForCTC.from_pretrained("GetmanY1/wav2vec2-large-fi-150k-finetuned")

# load dummy dataset and read soundfiles
ds = load_dataset("mozilla-foundation/common_voice_16_1", "fi", split='test')

# tokenize
input_values = processor(ds[0]["audio"]["array"], return_tensors="pt", padding="longest").input_values  # Batch size 1

# retrieve logits
logits = model(input_values).logits

# take argmax and decode
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)

Citation

If you use our models or scripts, please cite our article as:

@inproceedings{getman25_interspeech,
  title     = {{Is your model big enough? Training and interpreting large-scale monolingual speech foundation models}},
  author    = {{Yaroslav Getman and Tamás Grósz and Tommi Lehtonen and Mikko Kurimo}},
  year      = {{2025}},
  booktitle = {{Interspeech 2025}},
  pages     = {{231--235}},
  doi       = {{10.21437/Interspeech.2025-46}},
  issn      = {{2958-1796}},
}

Team Members

• Yaroslav Getman, Hugging Face profile, LinkedIn profile
• Tamas Grosz, Hugging Face profile, LinkedIn profile

Feel free to contact us for more details 🤗