|
|
--- |
|
|
license: mit |
|
|
language: |
|
|
- en |
|
|
--- |
|
|
|
|
|
## LoRA Training Config |
|
|
An example (fetal abdomen) of LoRA training configuration. |
|
|
``` |
|
|
{ |
|
|
"pretrained_model_name_or_path": "/root/autodl-tmp/sd_ckpt/v1-5-pruned.safetensors", |
|
|
"v2": false, |
|
|
"v_parameterization": false, |
|
|
"logging_dir": "/root/autodl-tmp/fetal_us/datasets/log", |
|
|
"train_data_dir": "/root/autodl-tmp/fetal_us/datasets/img", |
|
|
"reg_data_dir": "", |
|
|
"output_dir": "/root/autodl-tmp/fetal_us/datasets/output", |
|
|
"max_resolution": "512,512", |
|
|
"learning_rate": "0.0001", |
|
|
"lr_scheduler": "constant", |
|
|
"lr_warmup": "0", |
|
|
"train_batch_size": 1, |
|
|
"epoch": "1", |
|
|
"save_every_n_epochs": "1", |
|
|
"mixed_precision": "fp16", |
|
|
"save_precision": "fp16", |
|
|
"seed": "1234", |
|
|
"num_cpu_threads_per_process": 2, |
|
|
"cache_latents": true, |
|
|
"caption_extension": ".txt", |
|
|
"enable_bucket": true, |
|
|
"gradient_checkpointing": false, |
|
|
"full_fp16": false, |
|
|
"no_token_padding": false, |
|
|
"stop_text_encoder_training": 0, |
|
|
"xformers": false, |
|
|
"save_model_as": "safetensors", |
|
|
"shuffle_caption": false, |
|
|
"save_state": true, |
|
|
"resume": "", |
|
|
"prior_loss_weight": 1.0, |
|
|
"text_encoder_lr": "5e-5", |
|
|
"unet_lr": "0.0001", |
|
|
"network_dim": 128, |
|
|
"lora_network_weights": "", |
|
|
"color_aug": false, |
|
|
"flip_aug": false, |
|
|
"clip_skip": 2, |
|
|
"gradient_accumulation_steps": 1.0, |
|
|
"mem_eff_attn": false, |
|
|
"output_name": "fetal_ultrasound_v1.0", |
|
|
"model_list": "custom", |
|
|
"max_token_length": "75", |
|
|
"max_train_epochs": "1", |
|
|
"max_data_loader_n_workers": "1", |
|
|
"network_alpha": 128, |
|
|
"training_comment": "", |
|
|
"keep_tokens": "0", |
|
|
"lr_scheduler_num_cycles": "", |
|
|
"lr_scheduler_power": "", |
|
|
"persistent_data_loader_workers": false, |
|
|
"bucket_no_upscale": true, |
|
|
"random_crop": false, |
|
|
"bucket_reso_steps": 64.0, |
|
|
"caption_dropout_every_n_epochs": 0.0, |
|
|
"caption_dropout_rate": 0, |
|
|
"optimizer": "AdamW8bit", |
|
|
"optimizer_args": "", |
|
|
"noise_offset": "", |
|
|
"LoRA_type": "Standard", |
|
|
"conv_dim": 1, |
|
|
"conv_alpha": 1 |
|
|
} |
|
|
``` |
|
|
|
|
|
## FU-LoRA: LoRA Models |
|
|
| LoRA Model | Rank | |
|
|
| :------: | :------: | |
|
|
| fetal_ultrasound_v1.0.safetensors | 128 | |
|
|
| fetal_ultrasound_v2.0.safetensors | 32 | |
|
|
| fetal_ultrasound_v3.0.safetensors | 8 | |
|
|
|
|
|
## Paper |
|
|
URL: https://arxiv.org/abs/2407.20072 |
|
|
|
|
|
## Citation |
|
|
``` |
|
|
@misc{wang2024generative, |
|
|
title={Generative Diffusion Model Bootstraps Zero-shot Classification of Fetal Ultrasound Images In Underrepresented African Populations}, |
|
|
author={Fangyijie Wang and Kevin Whelan and Guénolé Silvestre and Kathleen M. Curran}, |
|
|
year={2024}, |
|
|
eprint={2407.20072}, |
|
|
archivePrefix={arXiv}, |
|
|
primaryClass={eess.IV}, |
|
|
url={https://arxiv.org/abs/2407.20072}, |
|
|
} |
|
|
``` |
