Recent content by HixonStep

Attached (hopefully) is over 1300 Lesbian JAV .srt files. Attaching .7z is not allowed so these files are renamed, should still be able to use 7zip to decompress. Probably a better way to do this but I'm lazy and wanted to get something out there. All these created with LargeV3 whisper model...

New BBAN subs for titles 201-300.

BBAN Subs[151-200]. Whisper Large

BBAN Subs[101-150]. Whisper Large

BBAN Subs[051-100]. Whisper Large

BBAN Subs[001-050] (Whisper Large model based should be a better then my earlier post)

Thanks. Glad I'm not the only one that want's to be able to understand the the dialog. I'm just trying to tweak solutions that were already in place and automate it. I definitely ran into alot of the she vs he. DASS-122 is an example of one I couldn't get right. Medium would mess up the...

I've fixed my docker image to get better results and run faster. Won't bore with the details but there was some audio that only got picked up during a medium pass, while large large pass was better quality. I just hated the missing audio but still wanted the better quality. I've put together a...

Agreed. There's also the power bill of running this on your own GPU, and I can't use the GPU for anything else while its running. There's alot of reasons why google colab is better. Also the 10GB for vram on 3080 is just BARELY enough for large (and only it specific scenarios). If you...

Here's a guide along with the scripts/docker build files/etc. All the actual cool stuff was written by "ANonEntity", I just got it to run local on a large batch. Make sure you run this on unimportant test data first, since its a recursive search. I can't imagine a situation it would delete...

I'm using the same parameters from theydontwantto's guide. Vad_threshold 0.4,chuck threshold 1.1, max attempts 3. I haven't really played around with changing these, I'm pretty amazed by the results of whisper altogether.

I have a 3080 with 10GB of VRAM. it's faster as in less work (i can just point it at a folder a leave). I also like that its all being handled locally instead of uploading stuff. I used your guide to learn everything but just tweaked the Collab script to function in python. It takes me about...

I don't have UAS-001, but I can try to share a guide for how I got things working. Funny thing is it seems like I get better results running the linux docker in windows. Seems like windows provide additional "shared gpu memory" that is enough for me do run large models. (Something I couldn't...

I'm using about 6GB of vram with open AI medium models. My GPU has 10GB and while I can load the large model, I run out of memory trying to compare. Faster-Whisper takes alot less VRAM and I was able to make that work on my GPU as well. I really liked the results of its large-v2 model using...

Thanks. Theydontwantto's guide of google colab/whisper was what showed me everything. (I knew nothing about any of this). But I wanted run something locally and automatic (so I can have it run on all my files). I ended up rewriting some of the collab script and getting it all automated in a...