Search results

https://www.akiba-online.com/threads/post-your-jav-subtitle-files-here-jav-subtitle-repository-jsp-not-a-sub-request-thread.1466451/post-4662235 now with 26 subtitle files. All Whisper edits from myself and then edited lightly.

Here are 26 Whisper Edits for the Slave Color series. This is the first 20 entries + the first three Teacher and first three Fiance entries. This is usually considered the Golden Age of the series, featuring lots of crossovers between casts. I think #10 is the best in the franchise, and...

Existing version control and existing wiki formats just aren't good for subtitles. You really need something that presents SubtitleEdit's UI as a communal wiki, with there being three separate and linked columns of time, duration, and text. It is like a table, but with restrictions in place...

Cuda, in particular, seems like there's a whole lot of potential problems when using mish mashes of DLLs, which is likely to happen when you have to start using workarounds to deal with different pathing standards. All of these upgrades versions of Whisper have easily ten times the overall...

whisper-ctranslate2 does the Large-v2 model when I could only use medium before. Lots of downsides though. I spent a huge amount of time dealing with Cuda paths. I eventually had to just copy-paste all of the Cuda dll files from my torch site-packages to the ctranslate folder, which is bad...

My recommendation for inspecting subtitles is the program SubtitleEdit. Open a sub in that, and then click on the video player or settings to make it correspond to the video file. Being able to see all of text, click on it, and immediately jump to that part of the video is a killer feature...

That is too complicated, unfortunately. There should be a better source or UI for crowd edits, but Git specifically is both too complicated and I don't think Git version of merging will work well with SRT files that have each line given a specific number.

Attackers looks like a good month coming up, with new entries in some of their long series. Female Gambler https://attackers.net/works/detail/RBK071 Female Investigator, Until the Fall... https://attackers.net/works/detail/RBK073 Absolute r*** https://attackers.net/works/detail/SAME051

https://attackers.net/works/detail/SAME051?page_from=series&sys_code=4728 Absolute r*** returns with the first new entry since 2020. They changed the cover theme to a woman being held down on a bed, rather than the woman posing outside.

For Slave Color, only the first entry was SSPD. The rest are RBD, which after RBD-999 switched over to RBK-001. The switch from RBK-001 is a decent marker for when the series moved over to the Slave Color Days series.

I feel like the first six entires are lower quality recordings, so the audio is worse and that makes the transcriptions worse. The very first entry (SSPD) is probably just a bad transcription, even after my edits. I moved on rather than spending a lot of time on a film that does not interest me.

What I was trying to say was that, of the subtitles that I have uploaded, if there are basic plot elements or scenes that are still unclear, I appreciate notes on which scenes need more direct editing from me. I have not posted 208 or 260 yet. 208 is one that I personally just like a lot, so...

I find the commandline for regular Whisper way easier to use, and I want to avoid having to upload files due to how many subs I do. I will likely take a stab at using the WebUI at some point, or just wait for ctranslate2 to get fixed-- there is no reason for it to not just use whatever paths...

Is there a way to run this via the command line as simply as regular Whisper? I have a ton of problems with trying to do things via regular Python code. I encounter way more VRAM issues, etc and it doesn't spit out all of the different subtitle formats automatically. EDIT: Looks like that...

I have the scenes translated in the subtitles. I am open to people pointing out when plot scenes are wrong or hard to understand.

Thank you! If there's really plot related scenes that are gibberish I'm willing to edit them for future releases. I do not care at al for the other 1000+ lines of just banter, but the 10 or so important plot scenes per movie should be right. The peak of the series is the 10-22 segment coming...

The first batch is available here! https://www.akiba-online.com/threads/post-your-jav-subtitle-files-here-jav-subtitle-repository-jsp-not-a-sub-request-thread.1466451/post-4654476 The first 9 entries of the series, and the first entries of Slave Teacher and Slave Fiance. These are the ones I...

A first set of (lightly) edited Whisper subtitles for the Attackers Slave Color series. This is for volumes 1 through 9 of the main series, Slave Fiance 1, and Slave Teacher 1. I think this is before the golden age of the series, which starts with volume 10 or so. Note that I had a relatively...

It doesn't have the username of my PC/account, or anything like that? I'm not good at technology.

I think I have enough Whisper edits done to do a full Slave Color bundle of the first 25 entries. After that it pivots to the "gimp" part of the series that is way less interesting to me. When I post subs here in a zip, what happens to the metadata of the files? I wish I could just upload...

Slave Color Days 4 - RBK-001 is notable for having Ellie Akira appear as a newly numbered slave a decade after she appeared in Slave Color 15. Slave #56 and Slave #172, a decade apart. The rate comes out to almost exactly a new slave per month. Some months we get a view, and other months we...

I actually don't know if that is really associated with the franchise, or not. I have it categorized with the Slave Color series but I think I may have just not understood how many other "Slave" franchises were out that deal with other aspects of the trafficking community in the Attackers...

After going through with Whisper, I think the real peak of the series is the Slave Color 14-20, and the spinoffs from that era. That section of story is the peak of cross over plots and Yukiko Yabuta at her most villainous. Slave Teacher 5, Slave Teacher 6, Slave Color 14, and Slave Color 15...

I've been documenting some more of the plot connections for the Slave Color and spinoffs for this thread and been especially trying to document the films that have connections. More often than not those connections come from the female dominants. Most of the time they just taunt and trash...

My best luck so far has been no_speech_threshold at 0.6, log at -1.0, and condition_on_previous_text as False. Beam size something like 8 or 10. This is a good general use for recent HD videos. I imagine that lower video quality could use other solutions. From there, open it up in...

Thanks to Whisper, there will probably be English subtitles for all of these within the next few months. I may put a few of my own edits on SubtitleCat, but I assume there's going to be more translations off of Chinese translations via Whisper. It has been 8 months since the last entry (a...

False for "condition on previous text" lets Whisper start doing things like subtitling (crying) and (grunting), but I also get (speaking in foreign language). No shit, Whisper? That is what I am here for! That is some of the value in running Whisper multiple times per project. One of them...

If you are using the command line (which I assume is what you are doing based on the "--", it is "--condition_on_previous_text False". You do not do any equal sign on the command line.

I think this thread will be a good place to debug subtitle files, but it would be a lot easier if we were allowed to upload .vtt and or .srt files. I can't imagine there being a reason why not. Otherwise, standard practice should be to rename them to .txt so they can be opened and viewed...

As far as I can tell, you can't just run things with the simple command line after that. You need to start worrying about the different parts of the process. Not sure if that has changed or if it will change, but the documentation does not describe the process well. There is logit biasing or...

Has somebody done that though? Once you have a fine tuned model, Whisper is vastly more difficult to run as far as I can tell.

Damn, that probably is a result of the large-v2. I don't really understand the interplay between No Speech and Logprob. I think hallucination may end up solved via Logprob but I don't know what values to even guess at for it.

No Speech at 0.08 is very low. Do you get lots of phantom speech?

If you're trying to get away with just Small, and it only takes a few minutes, you can try to do runs with a low temperature (but not zero) and like best_of 10 or something. Part of the key to Whisper is doing settings that try to compensate for whatever your computational stuff is limited by.

I have a frequent problem of the first 30 seconds not having translations or having bad translations, and then the chunk after the 30 second mark to be rushed and horribly mistimed.

I made a thread for Whisper setup, parameters, and discussion https://www.akiba-online.com/threads/whisper-openai-automatic-english-subtitles-for-any-film-in-any-language-an-intro-guide-to-subtitling-jav.2115103/

OpenAI made massive waves in 2022 when they released Dall-E 2 and ChatGPT, but that year they also released Whisper for Automatic Speech Recognition, meaning transcripts. Not just transcripts, but direct-to-English transcripts from any of 97 different languages, including Japanese. In...

Set "condition_on_previous_text" or whatever it is to False. In theory, it is there to make it learn into consistent style and slightly bias the results towards specific topics that the entire source is about. In practice specifically for porn films of languages other than English, there's...

Do some parameters get better timings? There's the known issue with not getting good timing within the first 30 second window, but there's some pretty bad mismatches all through some transcripts while others are great.

DeepL is theoretically better, but there's probably some value in doing direct-to-English with the same deep learning model rather than taking the transcribed output and feeding into a second deep learning model that isn't specifically designed to work interact with the first. There's just an...

I looked at the help, and tried all combintaions of "" (), and []. Searching for help with Whisper online is just extremely difficult because it is new, the name sucks, and because virtually nobody seems to be using the command line. https://blog.deepgram.com/exploring-whisper/ has examples...

What is the syntax for Temperature on the command line? in Python it is temperature=(0.2, 0.4, 0.5) and best_of=3 to use those three different temps, but I can't get a specific spread like that to work on the command line.

Thank you, I had seen the Temperature discussion but not that optimization blog.

For some reason many subs generated via Whisper will have a bunch of subtitles start exactly at 30 seconds, and totally rushed through and offset. Most of the time the timing is great, but I've seen this initial rush a few times and I can't tell how/why it is happening like this. For the most...

It is worth messing around with --suppress_tokens feature to see if this can be adjusted. Argument of "food" or ["food"] (not sure when to use which) will make it avoid using the word food, or words that are something similar.

I totally missed that Slave Color Married Woman RBD-468 had a really good English subtitle. I don't know if I had missed out on the film the first time around, because wow does this feel like one of the better entries in the series. They trash talk her like hell, she puts out a lot of crying...

I think Whisper Transcribe + Google Translate is a sidegrade. Not seeing consistent advance to it. Some lines better, some are worse. An adjusted Whisper model will likely beat both.

I am getting close to the point where I could make a Whisper thread that discusses how to best use it. I want to see if I can get one of the JP focused models working though. "--conditon_on_previous_text False" will prevent some of the repeated lines, but lead to worse translations everywhere...

It depends what you're going for. The value of Whisper isn't to get the highest quality translation possible, it is about lowering the effort bar to get something good enough. Medium Whisper, with well-chosen parameters and some manual editing from just watching the move and editing as you...

Seriously, the gap in quality between Whisper and most JAV subtitles is huge. RBD-179's translation on Subtitle Cat's first 2 minutes: Now let's look at even the Medium basic model from Whisper. The only thing it gets wrong is who married Mr. Kazuo. Her mother married Kazuo. I see that...