initial commit for the model card
Browse files
README.md
ADDED
@@ -0,0 +1,84 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: diffusers
|
3 |
+
pipeline_tag: text-to-image
|
4 |
+
---
|
5 |
+
|
6 |
+
## Model Details
|
7 |
+
|
8 |
+
### Model Description
|
9 |
+
|
10 |
+
This model is fine-tuned from [stable-diffusion-v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5) on 110,000 image-text pairs from the MIMIC dataset using the attention-tuning PEFT method. Under this fine-tuning strategy, fine-tune only the attention weights in the U-Net while keeping everything else frozen. Attention tuning was first shown to be effective for fine-tuning vision transformers in this [paper](https://arxiv.org/abs/2203.09795)
|
11 |
+
|
12 |
+
- **Developed by:** [Raman Dutt](https://twitter.com/RamanDutt4)
|
13 |
+
- **Shared by:** [Raman Dutt](https://twitter.com/RamanDutt4)
|
14 |
+
- **Model type:** [Stable Diffusion fine-tuned using Parameter-Efficient Fine-Tuning]
|
15 |
+
- **Finetuned from model:** [stable-diffusion-v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5)
|
16 |
+
|
17 |
+
### Model Sources
|
18 |
+
|
19 |
+
|
20 |
+
- **Paper:** [Parameter-Efficient Fine-Tuning for Medical Image Analysis: The Missed Opportunity](https://arxiv.org/abs/2305.08252)
|
21 |
+
- **Demo:** [MIMIC-SD-PEFT-Demo](https://huggingface.co/spaces/raman07/MIMIC-SD-Demo-Memory-Optimized?logs=container)
|
22 |
+
|
23 |
+
## Direct Use
|
24 |
+
|
25 |
+
This model can be directly used to generate realistic medical images from text prompts.
|
26 |
+
|
27 |
+
|
28 |
+
## How to Get Started with the Model
|
29 |
+
|
30 |
+
```python
|
31 |
+
from diffusers.pipelines import StableDiffusionPipeline
|
32 |
+
|
33 |
+
pipe = StableDiffusionPipeline.from_pretrained(sd_folder_path, revision="fp16")
|
34 |
+
|
35 |
+
```
|
36 |
+
|
37 |
+
|
38 |
+
## Training Details
|
39 |
+
|
40 |
+
### Training Data
|
41 |
+
|
42 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
43 |
+
|
44 |
+
[More Information Needed]
|
45 |
+
|
46 |
+
### Training Procedure
|
47 |
+
|
48 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
49 |
+
|
50 |
+
|
51 |
+
#### Metrics
|
52 |
+
|
53 |
+
This model has been evaluated using the Fréchet inception distance (FID) Score on MIMIC dataset.
|
54 |
+
|
55 |
+
### Results
|
56 |
+
|
57 |
+
[More Information Needed]
|
58 |
+
|
59 |
+
|
60 |
+
## Environmental Impact
|
61 |
+
|
62 |
+
|
63 |
+
|
64 |
+
## Citation
|
65 |
+
|
66 |
+
|
67 |
+
**BibTeX:**
|
68 |
+
|
69 |
+
@article{dutt2023parameter,
|
70 |
+
title={Parameter-Efficient Fine-Tuning for Medical Image Analysis: The Missed Opportunity},
|
71 |
+
author={Dutt, Raman and Ericsson, Linus and Sanchez, Pedro and Tsaftaris, Sotirios A and Hospedales, Timothy},
|
72 |
+
journal={arXiv preprint arXiv:2305.08252},
|
73 |
+
year={2023}
|
74 |
+
}
|
75 |
+
|
76 |
+
**APA:**
|
77 |
+
Dutt, R., Ericsson, L., Sanchez, P., Tsaftaris, S. A., & Hospedales, T. (2023). Parameter-Efficient Fine-Tuning for Medical Image Analysis: The Missed Opportunity. arXiv preprint arXiv:2305.08252.
|
78 |
+
|
79 |
+
## Model Card Authors
|
80 |
+
|
81 |
+
Raman Dutt
|
82 |
+
[Twitter](https://twitter.com/RamanDutt4)
|
83 |
+
[LinkedIn](https://www.linkedin.com/in/raman-dutt/)
|
84 |
+
[Email](mailto:s2198939@ed.ac.uk)
|