Instructions to use logasja/instagram-dogpatch with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Keras
How to use logasja/instagram-dogpatch with Keras:
# Available backend options are: "jax", "torch", "tensorflow". import os os.environ["KERAS_BACKEND"] = "jax" import keras model = keras.saving.load_model("hf://logasja/instagram-dogpatch") - Notebooks
- Google Colab
- Kaggle
| library_name: keras | |
| pipeline_tag: image-to-image | |
| metrics: | |
| - TopIQ-FR | |
| - ArcFace Cosine Distance | |
| license: gpl-3.0 | |
| tags: | |
| - adversarial | |
| - aesthetic | |
| - quality | |
| - filter | |
| widget: | |
| - text: input | |
| output: | |
| url: ./assets/input.png | |
| - text: target | |
| output: | |
| url: ./assets/target.png | |
| - text: output | |
| output: | |
| url: ./assets/output.png | |
| base_model: | |
| - vnet | |
| - logasja/ArcFace | |
| datasets: | |
| - logasja/FDF | |
| new_version: logasja/auramask-ensemble-dogpatch | |
| <Gallery /> | |
| Training logs [here](https://wandb.ai/spuds/auramask/runs/2c8bc343cb920b6c5646d628d787f609) | |
| # Model Description | |
| This model uses a modified vnet for 2D input/output implemented [here](https://github.com/logasja/keras3-unets) with the following configuration. | |
| ```json | |
| { | |
| "activation": "ReLU", | |
| "batch_norm": false, | |
| "filter_num": [ | |
| 64, | |
| 128, | |
| 256, | |
| 512, | |
| 512 | |
| ], | |
| "n_labels": 3, | |
| "output_activation": "tanh", | |
| "pool": false, | |
| "res_num_ini": 1, | |
| "res_num_max": 3, | |
| "unpool": false | |
| } | |
| ``` | |
| ```json | |
| { | |
| "alpha": 0.0001, | |
| "batch": 64, | |
| "epochs": 500, | |
| "epsilon": 1, | |
| "input": "(256, 256)", | |
| "losses": { | |
| "FEAT_ArcFace": { | |
| "d": "cosine_similarity", | |
| "f": "ArcFace", | |
| "name": "FEAT_ArcFace", | |
| "reduction": "sum_over_batch_size", | |
| "threshold": 0.68, | |
| "weight": 0.1 | |
| }, | |
| "TopIQ": { | |
| "full_ref": true, | |
| "lower_better": false, | |
| "name": "TopIQ", | |
| "reduction": "sum_over_batch_size", | |
| "score_range": "~0, ~1", | |
| "weight": 0.9 | |
| }, | |
| "mean_squared_error": { | |
| "name": "mean_squared_error", | |
| "reduction": "sum_over_batch_size", | |
| "weight": 0.1 | |
| } | |
| }, | |
| "mixed_precision": true, | |
| "optimizer": { | |
| "amsgrad": false, | |
| "beta_1": 0.9, | |
| "beta_2": 0.999, | |
| "clipnorm": null, | |
| "clipvalue": null, | |
| "ema_momentum": 0.99, | |
| "ema_overwrite_frequency": null, | |
| "epsilon": 1e-07, | |
| "global_clipnorm": null, | |
| "gradient_accumulation_steps": null, | |
| "learning_rate": 9.999999747378752e-05, | |
| "loss_scale_factor": null, | |
| "name": "adamw", | |
| "use_ema": false, | |
| "weight_decay": 0.004 | |
| }, | |
| "seed": "BIIIIIGSTRETCH", | |
| "testing": 0.01, | |
| "training": 0.99 | |
| } | |
| ``` | |
| ## Model Architecture Plot | |
|  |