AndroYD84

> The inference tutorial instruction point you to S3 bucket with a model. This model is new and shiny. Is this model as accurate as your new [SOTA 2019](https://github.com/facebookresearch/wav2letter/tree/master/recipes/models/sota/2019) model?...

@avidov >Out of interest, which ESPnet module and tutorial did you use? After building ESPnet, I ran the [ASR Demo example](https://github.com/espnet/espnet#asr-demo) on my WAV audio files, I used their pretrained...

@tlikhomanenko >However, could you also test our transformer S2S model I pointed above to compare with ESPnet? Here are the results using "[decode_transformer_s2s_ngram_other.cfg](https://github.com/facebookresearch/wav2letter/blob/master/recipes/models/sota/2019/librivox/decode_transformer_s2s_ngram_other.cfg)" : ``` tspeak0001.wav ESPNET= THE UNLIMITED BE...

@tlikhomanenko Sorry but I'm on holiday now, I can access my PC remotely but can't do much as now it's using 100% of my GPU and CPU on another task,...

I was going to report a much less troublesome issue than this, but seeing how even very important features like these are left ignored, I lost hope. Yet another open...

 I'm pretty confident you forgot to change this default option.

In my experience, Dupeguru throws this error when a single file have 1000 or more duplicates (I don't know the exact number), the way I worked around it, is to...

Having the same problem. The GUI has been stuck like this for 3 days while doing a mass symlink operation. It takes about 3 days to complete a scan (with...

Made this python script to convert a saved duplicate finder result from Czkawka into a Dupeguru file. So you can keep working on Dupeguru if Czkawka crashes without wasting time...

Are people actually trying before opening issues? It's right click and "Make Selected into Reference" on the file you want to preserve.