Update README.md
Browse files
README.md
CHANGED
|
@@ -3,8 +3,6 @@ license: mit
|
|
| 3 |
base_model:
|
| 4 |
- stabilityai/stable-diffusion-xl-base-1.0
|
| 5 |
pipeline_tag: text-to-image
|
| 6 |
-
datasets:
|
| 7 |
-
- neuralcatcher/hateful_memes
|
| 8 |
---
|
| 9 |
# SarcasmDiffusion — SDXL Fused Meme Generator
|
| 10 |
|
|
@@ -54,7 +52,7 @@ SarcasmDiffusion is a diffusion-based generative model focused on producing **cl
|
|
| 54 |
- **Frozen:** VAE, text_encoder, text_encoder_2
|
| 55 |
|
| 56 |
### Data
|
| 57 |
-
- Source: *Hateful Memes* (Facebook AI).
|
| 58 |
- We **excluded** labeled hateful samples and applied **NLP enrichment**:
|
| 59 |
- Emotion scoring (GoEmotions distilled) and irony scoring (RoBERTa‑irony).
|
| 60 |
- Heuristics + percentiles → tones: `humor / irony / neutral`.
|
|
@@ -137,7 +135,7 @@ pip install "diffusers==0.35.1" "transformers==4.45.2" "accelerate==1.10.1
|
|
| 137 |
|
| 138 |
## License
|
| 139 |
|
| 140 |
-
- **Code:** MIT
|
| 141 |
- **Model weights:** follow the base model’s license (Stability AI / SDXL Base 1.0).
|
| 142 |
- **Data:** Users must obtain *Hateful Memes* from its source and agree to its terms.
|
| 143 |
|
|
|
|
| 3 |
base_model:
|
| 4 |
- stabilityai/stable-diffusion-xl-base-1.0
|
| 5 |
pipeline_tag: text-to-image
|
|
|
|
|
|
|
| 6 |
---
|
| 7 |
# SarcasmDiffusion — SDXL Fused Meme Generator
|
| 8 |
|
|
|
|
| 52 |
- **Frozen:** VAE, text_encoder, text_encoder_2
|
| 53 |
|
| 54 |
### Data
|
| 55 |
+
- Source: *Hateful Memes* (Facebook AI).
|
| 56 |
- We **excluded** labeled hateful samples and applied **NLP enrichment**:
|
| 57 |
- Emotion scoring (GoEmotions distilled) and irony scoring (RoBERTa‑irony).
|
| 58 |
- Heuristics + percentiles → tones: `humor / irony / neutral`.
|
|
|
|
| 135 |
|
| 136 |
## License
|
| 137 |
|
| 138 |
+
- **Code:** MIT
|
| 139 |
- **Model weights:** follow the base model’s license (Stability AI / SDXL Base 1.0).
|
| 140 |
- **Data:** Users must obtain *Hateful Memes* from its source and agree to its terms.
|
| 141 |
|