Is there a method to run this by way of the command line as simply just as common Whisper? I've a lot of issues with wanting to do items via standard Python code. I come across far more VRAM problems, and many others and it won't spit out all of the different subtitle formats routinely.
three. To fight the final trouble I parsed my whole map tree in a number of AI to question if it observed any duplicates that have this issue. Deepseek, Chatgpt could not handle it, They only invented copy names. Claud.AI context window was as well tiny (about 3000 actress names).
Again, I do not fully grasp Japanese so my re-interpretations may not be absolutely accurate but I try to match what is going on inside the scene. In any case, delight in and allow me to understand what you're thinking that.
Since it uses fewer VRAM, In addition, it means that folks who doesn't have 10 GB VRAM can use substantial-v2. RTX 2060 6GB seems to run it easily according to a comment on A lot quicker Whisper Webui Group tab
SPRD-1244 Eng Sub. I tied the knot at this age thanks to that stepson of mine… I desired to working experience the Uncooked, unfiltered satisfaction of a person getting me when my husband was away, and my son gave me that excuse.
three. To combat the final trouble I parsed my comprehensive map tree in two or three AI to request if it saw any duplicates which have this problem. Deepseek, Chatgpt click here couldn't cope with it, They only invented replicate names. Claud.AI context window was also smaller (about 3000 actress names).
⦁ Causes it to be possible for just one person to make subtitle information for what was a multi-man or woman position. 1 individual, with Whisper, can generate timed subtitle files with in English from any language.
After i have enough time I can even look at bundling my very own packs, apart from It truly is almost certainly a waste of my time as a consequence of overlapping with current packs. Click on to expand...
vorpal77 reported: There are several traces I needed to liberally interpret and a line I couldn't make legible to the top Regardless of how tough I tried so yeah sorry if that kills a boner. Click to increase...
Japanese is one of the greater languages for this by standard transcription and translation metrics, but dirtiness and pronouns will likely be a Repeated challenge to operate all over.
There are several lines I had to liberally interpret and a line I couldn't make legible to the top Regardless how hard I tried so yeah sorry if that kills a boner. Stating any translation faults could be appreciated, In particular the one I just pointed out (If replying in the thread is not Okay PMing will be wonderful likewise.)
I ought to point out that there are most likely some mistakes right here and there, as no automatic Resource is perfect. But on spot checking the gathering, the mistake fee looks relatively minimal, and the benefits appeared worthwhile.
There are many in subtitlecat, but they don't make any feeling. I'd personally get them, but I don't have the playing cards needed to get them. I understand it's not a request, but when anyone has them, you should post them here or privately.
There is certainly even now a huge difficulty, which can be aliases. 1 actress might have 2 distinctive directories but less than a unique alias. I realize a solution I could deal with it, but I want to determine how to query javstash graphql as they may have Practically all aliases.