I'm trying out massed compute with your tutorial and the thindrives option says it's broken. any ide
I'm trying out massed compute with your tutorial and the thindrives option says it's broken. any idea what that means and how to correct?
.@Furkan Gözükara SECourses I’m currently training Flex on my dataset and I feel it learns much quicker with half the steps compared to dev dedistilled. You were mentioning the learning rate might need adjustment. Do you think it needs to be lowered?
.After only 4000 steps. Not even finished the whole epoch.
A@Furkan Gözükara SECourses Hey so I experimented today a bit and got lost . Can you write what a segment syntax example looks like? Like is it 2 <segment> for -1 and -2?
What would the format look like ?
What would the format look like ?
AFor 2 Lora faces
Fif learning true fast yes
Flower needed
Fyes when connecting
Fselect end existing session
Fbut it terminates all running apps
F
Ai did didn't get an answer yet
Awill it be like <segment:yolov8.pt-1,0.6.05a photo of OHWX man <Lora:Testman-1>>. ?
Fthat seems wrong to me
Fprompt like this
F<segment:yolo-face_yolov9c.pt-1,0.7,0.5>photograph of ohwx man
Fand then click lora
Aand for Lora?
Aoh ok
Fbut actually i dont know how to use lora
Fon prompt 
Fbut it should come at the end
Fappended
Ait say in the document after the text prompt to type manually<Lora...>
A<segment:yolo-face_yolov9c.pt-1,0.7,0.5>photograph of ohwx man <Lora....>
Ayeah I think that's jt thanks doc !
Fye
TDo you meat at training? At inference L40S is only 8% slower but 0.3 second new prompt delay opposed to >3s on 5090
Tso for practical purposes L40S is still better
Fi think it depends if you are constantly changing prompt yes
Fbut L40S is 8000+ USD
Fand 5090 is at its slowest performance
Fit will only get better we dont even have official pytorch for windows
P@Furkan Gözükara SECourses Hello I just started training with the 24 GB flux dev model but my storage was less only 50gb so i think it can create only 2 checkpoints but I checked the training status and the second checkpoint epoch has already strated even before checkpoint one epochs are totally complete.How to save one checkpoint to huggingface if the second one has also started simultaneously
P@Furkan Gözükara SECourses and also once the storage is full what error will I get and how to resume the training
P@Furkan Gözükara SECourses actually I have set 4 checkpoints instead of 2
Fi never could make it save to hugging face before automatically from kohya
Fonce storage is full i dont remember behaviour sadly
P@Furkan Gözükara SECourses So can i just stop this process and increase the storage for the same gpu instance and the again start training
Fon runpod yes
PI am on masscompute
Fyou cant there
Fget L40 machine it has way more storage
For multiple A6000 GPU
Fby the way
Fyou can delete a lot of apps there
Fall will increase storage
Fyou can delete all except kohya
PAs you had said in the long tutorial that checkpoints can be saved midway so I proceeded with 4 checkpoints 
Fyou can. but you have to monitor
Fas soon as saved upload to huggin gface
