Dutch Masters: Jan (Johannes) Vermeer Style - V1.2
by ModelsLabRemoved all the curtains from the training set images because they are appearing too often. Fixed the Florence 2 generated captions by hand for better accuracy, this seems to have improved prompt following somewhat.This is epoch 13, the rest can be found here: tensor. art/models/806185152939254291/Jan-Vermeer-V3-HCpD6Cos5R30E13-2024-12-12-01:01:25-Ep-13FLUX.1 - dev-fp8Trigger: janver1 paintingRepeat: 30 Epoch: 13 Trained on just ten 512x512 images for otal of 3900 stepsUnet LR: 0.0005 Scheduler: cosine Optimizer: AdamWNetwork Dim: 6 Alpha: 3Epoch Loss4 0.3185 0.3016 0.2717 0.2348 0.1939 0.16310 0.14711 0.13212 0.12213 0.121 <- lowest
dutch-masters-jan-johannes-vermeer-style-v1-2Input
Per image generation will cost 0.0047$
For premium plan image generation will cost 0.00$ i.e Free.
Output
Unknown content type
Related Models
Discover similar models you might be interested in
![LOVE Booster for Rendered Romance Contest [FLUX+IL] - FLUX v0.1](https://images.stablediffusionapi.com/?Image=https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/5e6b0b93-af51-460a-b225-a4e37ebacf93/width=720/55335286.mp4)
![Skynet Cyberdyne Systems - by ChronoKnight - [FLUX] - v2.0](https://images.stablediffusionapi.com/?Image=https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/49421db0-bd69-4dd8-8aff-b96a2b7e5efa/width=992/47699167.jpeg)
Dutch Masters: Jan (Johannes) Vermeer Style - V1.2 Readme
Removed all the curtains from the training set images because they are appearing too often. Fixed the Florence 2 generated captions by hand for better accuracy, this seems to have improved prompt following somewhat.This is epoch 13, the rest can be found here: tensor. art/models/806185152939254291/Jan-Vermeer-V3-HCpD6Cos5R30E13-2024-12-12-01:01:25-Ep-13FLUX.1 - dev-fp8Trigger: janver1 paintingRepeat: 30 Epoch: 13 Trained on just ten 512x512 images for otal of 3900 stepsUnet LR: 0.0005 Scheduler: cosine Optimizer: AdamWNetwork Dim: 6 Alpha: 3Epoch Loss4 0.3185 0.3016 0.2717 0.2348 0.1939 0.16310 0.14711 0.13212 0.12213 0.121 <- lowest