@Dr. Furkan Gözükara What gpu is that for $1?
@Dr. Furkan Gözükara What gpu is that for $1?
Fsomething so wrong
FA6000
F3 hours
Fi will show durations at next tutorial hopefully
Fand will post entire info on patreon soon
Fpreparing
Fi will also compare 15 img vs 256, fine tune vs lora, batch size 1 vs 7 
Vdo vs dora since ontrainer can do doras for flux, maybe a finetune lora vs lora vs dora?
Fnice
Fbut this depends on dataset size
Fi do more steps with 256 images
Fit needs a special name
For you need to manually select
Fi shared in #tips-info channel
Falso here a tips for you
Ffine tuned model can do much higher high res upscale compared to lora
Fwithout stripe lines
V@Dr. Furkan Gözükara I had wanted to ask, I read some documentation that training captions requires commas to separate the prompt, does that still hold true for flux? I have my prompts structured as sentences and there are no commas only peroids at the end of the sentence.
Fi think not
Fbut you dont need captions for flux
Fat least for distilled dev
Vtoo late
V
VIt just finished and I am happy with the results.
Fnice
VFully captioned, including colors, with activation sentence and word.
V"painted of the style of "word"".
Fyes it works
Fbut a little bit subpar
Fi compared :d
Fbut depends on your dataset
Fit must be consisten
VI compare caption with activation vs caption without when making the image in comfyui. I prefer caption with activation word as quality is slightly better.
VI did caption by hand with gpt4
Fi need to start 4 lora training tonight
Fso i can compare best lora vs best fine tune :d
Vtry dora as well and compare all 3
Fit will be only later time
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=40)
Du reckon you will be finished with flux training by next christmas doc?
VThe only problem with this lora is that it took a long time to train - 25 hr at 2.61 it/s
Di think i could now recognise you blindfolded
HI had previously tried manual captioning with GPT Vision too, but it was really tedious and torturous because my internet was very limited and I couldn't download a BLIP model, especially with large datasets 
VIt doesn't do so well vs certain art styles like kawaii if it has lots of busy things happening, but for paintings its 95% great.
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=16)
Fbatch size 7 fails
Fmeanwhile fine tuning works :d
PThanks a lot, I am so stupid!
Fno issues