Update README.md
Browse files
README.md
CHANGED
|
@@ -20,7 +20,7 @@ base_model: CompVis/stable-diffusion-v1-4
|
|
| 20 |
|
| 21 |
# Model Card for Renaissance Stable Diffusion
|
| 22 |
|
| 23 |
-
<!-- Provide a quick summary of what the model is/does. [
|
| 24 |
This is a Fine-tuned Stable Diffusion model on a custom dataset of {image, caption} pairs. This model has been built on top of the fine-tuning script provided by Hugging Face. This model uses the KerasCV implementation of stability.ai's text-to-image model. Unlike other open-source alternatives like Hugging Face's Diffusers, KerasCV offers advantages such as XLA compilation and mixed precision support, resulting in state-of-the-art generation speed. Fine-tuning the Stable Diffusion model for generating high-quality Renaissance-style portraits.
|
| 25 |
|
| 26 |
|
|
@@ -35,7 +35,6 @@ This is a Fine-tuned Stable Diffusion model on a custom dataset of {image, capti
|
|
| 35 |
- [Model Description](#model-description)
|
| 36 |
- [Uses](#uses)
|
| 37 |
- [Direct Use](#direct-use)
|
| 38 |
-
- [Downstream Use [Optional]](#downstream-use-optional)
|
| 39 |
- [Out-of-Scope Use](#out-of-scope-use)
|
| 40 |
- [Bias, Risks, and Limitations](#bias-risks-and-limitations)
|
| 41 |
- [Recommendations](#recommendations)
|
|
@@ -52,15 +51,15 @@ This is a Fine-tuned Stable Diffusion model on a custom dataset of {image, capti
|
|
| 52 |
- [Results](#results)
|
| 53 |
- [Model Examination](#model-examination)
|
| 54 |
- [Environmental Impact](#environmental-impact)
|
| 55 |
-
- [Technical Specifications [
|
| 56 |
- [Model Architecture and Objective](#model-architecture-and-objective)
|
| 57 |
- [Compute Infrastructure](#compute-infrastructure)
|
| 58 |
- [Hardware](#hardware)
|
| 59 |
- [Software](#software)
|
| 60 |
- [Citation](#citation)
|
| 61 |
-
- [Glossary [
|
| 62 |
-
- [More Information [
|
| 63 |
-
- [Model Card Authors [
|
| 64 |
- [Model Card Contact](#model-card-contact)
|
| 65 |
- [How to Get Started with the Model](#how-to-get-started-with-the-model)
|
| 66 |
|
|
@@ -101,14 +100,6 @@ The model is intended for research purposes only. Possible research areas and ta
|
|
| 101 |
|
| 102 |
Excluded uses are described below.
|
| 103 |
|
| 104 |
-
|
| 105 |
-
## Downstream Use [Optional]
|
| 106 |
-
|
| 107 |
-
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 108 |
-
<!-- If the user enters content, print that. If not, but they enter a task in the list, use that. If neither, say "more info needed." -->
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
## Out-of-Scope Use
|
| 113 |
|
| 114 |
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
|
@@ -131,7 +122,6 @@ Using the model to generate content that is cruel to individuals is a misuse of
|
|
| 131 |
|
| 132 |
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 133 |
|
| 134 |
-
Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
|
| 135 |
## Limitations
|
| 136 |
- The model does not achieve perfect photorealism
|
| 137 |
- The model cannot render legible text
|
|
@@ -182,31 +172,6 @@ Batch: 1
|
|
| 182 |
|
| 183 |
Learning rate: warmup to 0.0001 for 10,000 steps and then kept constant
|
| 184 |
|
| 185 |
-
# Evaluation
|
| 186 |
-
|
| 187 |
-
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 188 |
-
|
| 189 |
-
## Testing Data, Factors & Metrics
|
| 190 |
-
|
| 191 |
-
### Testing Data
|
| 192 |
-
|
| 193 |
-
<!-- This should link to a Data Card if possible. -->
|
| 194 |
-
|
| 195 |
-
More information needed
|
| 196 |
-
|
| 197 |
-
|
| 198 |
-
### Factors
|
| 199 |
-
|
| 200 |
-
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 201 |
-
|
| 202 |
-
More information needed
|
| 203 |
-
|
| 204 |
-
### Metrics
|
| 205 |
-
|
| 206 |
-
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 207 |
-
|
| 208 |
-
More information needed
|
| 209 |
-
|
| 210 |
## Results
|
| 211 |
|
| 212 |
Please Check out the Github Repo at https://github.com/martingasparyan/Fine-Tune-Stable-Diffusion/wiki
|
|
@@ -257,7 +222,7 @@ A100 PCIe 40/80GB
|
|
| 257 |
|
| 258 |
Google Colab, Jupyter Lab
|
| 259 |
|
| 260 |
-
# Model Card Authors [
|
| 261 |
|
| 262 |
<!-- This section provides another layer of transparency and accountability. Whose views is this model card representing? How many voices were included in its construction? Etc. -->
|
| 263 |
|
|
|
|
| 20 |
|
| 21 |
# Model Card for Renaissance Stable Diffusion
|
| 22 |
|
| 23 |
+
<!-- Provide a quick summary of what the model is/does. [] -->
|
| 24 |
This is a Fine-tuned Stable Diffusion model on a custom dataset of {image, caption} pairs. This model has been built on top of the fine-tuning script provided by Hugging Face. This model uses the KerasCV implementation of stability.ai's text-to-image model. Unlike other open-source alternatives like Hugging Face's Diffusers, KerasCV offers advantages such as XLA compilation and mixed precision support, resulting in state-of-the-art generation speed. Fine-tuning the Stable Diffusion model for generating high-quality Renaissance-style portraits.
|
| 25 |
|
| 26 |
|
|
|
|
| 35 |
- [Model Description](#model-description)
|
| 36 |
- [Uses](#uses)
|
| 37 |
- [Direct Use](#direct-use)
|
|
|
|
| 38 |
- [Out-of-Scope Use](#out-of-scope-use)
|
| 39 |
- [Bias, Risks, and Limitations](#bias-risks-and-limitations)
|
| 40 |
- [Recommendations](#recommendations)
|
|
|
|
| 51 |
- [Results](#results)
|
| 52 |
- [Model Examination](#model-examination)
|
| 53 |
- [Environmental Impact](#environmental-impact)
|
| 54 |
+
- [Technical Specifications []](#technical-specifications-)
|
| 55 |
- [Model Architecture and Objective](#model-architecture-and-objective)
|
| 56 |
- [Compute Infrastructure](#compute-infrastructure)
|
| 57 |
- [Hardware](#hardware)
|
| 58 |
- [Software](#software)
|
| 59 |
- [Citation](#citation)
|
| 60 |
+
- [Glossary []](#glossary-)
|
| 61 |
+
- [More Information []](#more-information-)
|
| 62 |
+
- [Model Card Authors []](#model-card-authors-)
|
| 63 |
- [Model Card Contact](#model-card-contact)
|
| 64 |
- [How to Get Started with the Model](#how-to-get-started-with-the-model)
|
| 65 |
|
|
|
|
| 100 |
|
| 101 |
Excluded uses are described below.
|
| 102 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
## Out-of-Scope Use
|
| 104 |
|
| 105 |
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
|
|
|
| 122 |
|
| 123 |
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 124 |
|
|
|
|
| 125 |
## Limitations
|
| 126 |
- The model does not achieve perfect photorealism
|
| 127 |
- The model cannot render legible text
|
|
|
|
| 172 |
|
| 173 |
Learning rate: warmup to 0.0001 for 10,000 steps and then kept constant
|
| 174 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 175 |
## Results
|
| 176 |
|
| 177 |
Please Check out the Github Repo at https://github.com/martingasparyan/Fine-Tune-Stable-Diffusion/wiki
|
|
|
|
| 222 |
|
| 223 |
Google Colab, Jupyter Lab
|
| 224 |
|
| 225 |
+
# Model Card Authors []
|
| 226 |
|
| 227 |
<!-- This section provides another layer of transparency and accountability. Whose views is this model card representing? How many voices were included in its construction? Etc. -->
|
| 228 |
|