Is this a nvidia TDT model?
Can you give more details about this ?(architecture, training dataset, model card etc..?)
please give individual language wise models if possible..
We are currently in the process of preparing detailed documentation, which we expect to share in a few weeks.
Are you looking for any particular language? We have some language-specific models available here: https://huggingface.co/ARTPARK-IISc/models
We would love to receive detailed feedback on the models and the Vaani dataset. Please feel free to connect with me at sujith@artpark.in
.
Looking for Malayalam, its not there as an individual weight .. Waiting for your documentations. It'll be good if provided an individual malayalam weight quick...
We would like to know about your use case, it if is aligned with our vision, happy to help you with individual malayalam model.
Doing my masters, doing data collection (audio interviews). It would be thousand times helpfull if a model is provided...
(much more helpfull, if the model is an MLX/metal/appleSilicon compatible)..
Have you tried Vaani-FastConformer-Multilingual model for malayalam?
tried, good model, but still some small small corrections..
any way am getting any .mlmodel of malayalam?
waiting..
please check ARTPARK-IISc/Vaani-FastConformer-Malayalam it would be great if can share the comparative performance report of the models on your data
thanks for the model, am using it now... ( here and there some accuracy issues, a fine tune will sort that out). it'll be good if you can give provide, some small size, onnx, safetensor models for browser and edge devices....