Downloads · 30 days
0
stremma-elm
sigrlami/stremma-elm
stremma-elm is a automatic speech recognition model from sigrlami. Use it when you need speech turned into text. The card lists the license as mit.
The initial model that was used for https://stremma.ai service, serves as a baseline for the current v2 in production, handling most of the ASR tasks.
Access
Public
Updated Jun 7, 2024
Repo size
—
Likes
1
Public
Repo makeup
Click a slice to open those files.
Other1.5 KB · 65%
At a glance
- Task
- Automatic Speech Recognition
- License
- mit
- Access
- Public
- Created
- Jan 17, 2024
- Updated
- Jun 7, 2024
- SHA
- d53231bf
Model card
From the Hugging Face model README
Stremma ELM(extendable language model)
The initial model that was used for https://stremma.ai service, serves as a baseline for the current v2 in production, handling most of the ASR tasks.
Detais
Stremma ELM is a transformer-based encoder-decoder model, also referred to as a sequence-to-sequence model. It was trained on 0.1M hours of labeled audio and 0.4M hours of pseudolabeled audio collected via Stremma SaaS during period of work in manual transcribing.
The models were trained on either English or Danish data. The English and Danish models were trained on the task of speech recognition. The multilingual extensions were trained on both speech recognition and speech translation.
Usage
TBD
- Task
- Automatic Speech Recognition
- License
- mit
- Languages
- en, de, da, no, sv
- Created
- Jan 17, 2024
- Updated
- Jun 7, 2024