NeonBohdan

Implementing [wav2vec 2.0](https://github.com/facebookresearch/fairseq/tree/main/examples/wav2vec#wav2vec-20) will be very useful I think CTranslate2 is the best CPU Transformer inference accelerator, on GPU to for low batch sizes, and comparable for higher If it...

I have tried this, but no luck ``` input_dummy = torch.randint(0, 24, (batch_size, 128)).long().to(device) input_lengths = torch.randint(100, 129, (batch_size,)).long().to(device) input_lengths[-1] = 128 spec = torch.rand(batch_size, filter_length // 2 + 1,...

Hi @guillaumekln is it possible to implement EOS sequence of tokens For example [this model](https://huggingface.co/togethercomputer/RedPajama-INCITE-Chat-3B-v1) don't use a token to stop generation but a sequence `[":"]`

@guillaumekln Thanks Looking at the problem from a development time point of view, it's easier to use the ":" token. You are right, I just notify you that such models...

This is because whisper was originally also trained on examples where you feed initial prompt too If this was absent in your dataset then this feature will die This is...

Your model solves a lot of problems of gemma-7b-it (With it's tokenizer it's great for Ukrainian) Can you please do the same for 2b one It's very interesting size Hope...