Parcollet Titouan
Parcollet Titouan
Our current tensor board logger is not DDP compliant and can't be resumed if the resume crash or if we restart from a different epoch. This PR fixes all this.
Dear all, I feel like you have made a mistake (or let a test value on the code :) ) on at the line 128 of training.py, indeed I think...
### Describe the bug While profiling a conformer transducer, I realised that a few functions were taking way more time than expected. I can't share the profiling trace here due...
## What does this PR do? Based on this #2526, this PR is a first attempt at adding Adapters to any SB model. This will only work if PreTrainer is...
### Describe the bug Starting any recipe with two or + processes will fail using DDP as the run_on_main function does not catch (no barrier?) the other processes. From the...
### Describe the bug It's not a bug, just a discussion. I think that people including @Adel-Moumen @poonehmousavi @mravanelli and maybe @pplantinga @asumagic may want to engage. Adapters, or, more...
Hello there! SpeechBrain 1.0.0 is out, and the API is still relying on 0.5.13 could it be possible to bump this requirement? I realise that we also need to change...
This PR adds the summarymixing for streaming conformer. I think this PR should be merged before the SSL one #9. Once both PR have been merged, we should create a...
### Describe the bug Putting a line like `mask2 = torch.rand(5, requires_grad=False, device=self.device)` in the on_stage_start function on a recipe litteraly breaks determinism. Let's say that mask2 isn't used, at...
### Describe the bug Samsung AI Center Cambridge (Shucong) may want to try to implement these embeddings for the v1. ### Expected behaviour New feature. ### To Reproduce _No response_...