Update README.md
Browse files
README.md
CHANGED
|
@@ -5,6 +5,9 @@ language:
|
|
| 5 |
pipeline_tag: text-generation
|
| 6 |
---
|
| 7 |
|
|
|
|
|
|
|
|
|
|
| 8 |
|
| 9 |
|
| 10 |
# AURORA-Tiny 🌅✨
|
|
@@ -76,16 +79,6 @@ AURORA-Tiny uses a novel combination of:
|
|
| 76 |
- **Multi-Head Attention**: Standard transformer attention with time modulation
|
| 77 |
- **Output Projection**: Maps back to vocabulary space
|
| 78 |
|
| 79 |
-
## 📊 Performance
|
| 80 |
-
|
| 81 |
-
AURORA-Tiny achieves competitive results on various text generation tasks despite its compact size:
|
| 82 |
-
|
| 83 |
-
| Dataset | Perplexity | BLEU Score | Training Time | Parameters |
|
| 84 |
-
|---------|------------|------------|---------------|------------|
|
| 85 |
-
| Movie Reviews | 28.1 | 0.38 | ~15 min | 2.4M |
|
| 86 |
-
| News Articles | 35.2 | 0.34 | ~20 min | 2.4M |
|
| 87 |
-
| Poetry | 23.6 | 0.31 | ~12 min | 2.4M |
|
| 88 |
-
|
| 89 |
*Tested on RTX 3060, batch_size=16, 15 epochs. Model size: ~2.4M parameters*
|
| 90 |
|
| 91 |
## 🎛️ Configuration
|
|
|
|
| 5 |
pipeline_tag: text-generation
|
| 6 |
---
|
| 7 |
|
| 8 |
+
> [!NOTE]
|
| 9 |
+
> Hii!!! This is a side project, so is not the best.
|
| 10 |
+
|
| 11 |
|
| 12 |
|
| 13 |
# AURORA-Tiny 🌅✨
|
|
|
|
| 79 |
- **Multi-Head Attention**: Standard transformer attention with time modulation
|
| 80 |
- **Output Projection**: Maps back to vocabulary space
|
| 81 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
*Tested on RTX 3060, batch_size=16, 15 epochs. Model size: ~2.4M parameters*
|
| 83 |
|
| 84 |
## 🎛️ Configuration
|