Model description
This model has been fine-tuned to learn the concept of Minecraft by ChienVM on this dataset
Intended uses & limitations
Concept token "mrf style". Example: "a photo of landscape in mrf style" This model only finetunes with the landscape image in Minecraft world. Therefore, it can not generate good iamges with other concepts.
Some examples prompt
<figure> <img src="https://huggingface.co/vumichien/dreambooth_diffusion_minercraft/resolve/main/demo_01.png" alt="Sample 1"> <figcaption>Prompt: a fishing village under a cherry blossom forest at sunset in mrf style. Negative prompt: bad anatomy, soft blurry. Steps: 50, CFG scale: 7.5, Size: 512x12 </figcaption> </figure>
<figure> <img src="https://huggingface.co/vumichien/dreambooth_diffusion_minercraft/resolve/main/demo_02.png" alt="Sample 2"> <figcaption>Prompt: a photo of landscape in mrf style. Negative prompt: bad anatomy, soft blurry. Steps: 50, CFG scale: 7.5, Size: 512x12 </figcaption> </figure>
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
| Hyperparameters | Value |
|---|---|
| inner_optimizer.class_name | Custom>RMSprop |
| inner_optimizer.config.name | RMSprop |
| inner_optimizer.config.weight_decay | None |
| inner_optimizer.config.clipnorm | None |
| inner_optimizer.config.global_clipnorm | None |
| inner_optimizer.config.clipvalue | None |
| inner_optimizer.config.use_ema | False |
| inner_optimizer.config.ema_momentum | 0.99 |
| inner_optimizer.config.ema_overwrite_frequency | 100 |
| inner_optimizer.config.jit_compile | True |
| inner_optimizer.config.is_legacy_optimizer | False |
| inner_optimizer.config.learning_rate | 0.0010000000474974513 |
| inner_optimizer.config.rho | 0.9 |
| inner_optimizer.config.momentum | 0.0 |
| inner_optimizer.config.epsilon | 1e-07 |
| inner_optimizer.config.centered | False |
| dynamic | True |
| initial_scale | 32768.0 |
| dynamic_growth_steps | 2000 |
| training_precision | mixed_float16 |
