Update README.md
Browse files
README.md
CHANGED
|
@@ -12,9 +12,12 @@ palmer is a series of ~1b parameters language models fine-tuned to be used as ba
|
|
| 12 |
|Model| ARC_C| HellaSwag| PIQA| Winogrande|
|
| 13 |
|------|-----|-----------|------|-------------|
|
| 14 |
|tinyllama-2t| 0.2807| 0.5463| 0.7067| 0.5683|
|
| 15 |
-
|palmer-001 | 0.2807| 0.5524| 0.7106|
|
| 16 |
-
|sheared-1.3b| 0.2910| 0.5935|
|
| 17 |
-
|palmer-002 |
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
### training
|
| 20 |
Training took ~5 P100 gpu hours. It was trained on 15,000 gpt-4 shuffled samples. palmer was fine-tuned using lower learning rates ensuring it keeps as much general knowledge as possible.
|
|
|
|
| 12 |
|Model| ARC_C| HellaSwag| PIQA| Winogrande|
|
| 13 |
|------|-----|-----------|------|-------------|
|
| 14 |
|tinyllama-2t| 0.2807| 0.5463| 0.7067| 0.5683|
|
| 15 |
+
|palmer-001 | 0.2807| 0.5524| 0.7106| 0.5896|
|
| 16 |
+
|sheared-1.3b| 0.2910| 0.5935| 0.7339| 0.5809|
|
| 17 |
+
|palmer-002 | 0.3157| **0.6022**| 0.7334| 0.5864|
|
| 18 |
+
|falcon-rw-1b-instruct-openorca (sota) | **0.3362**| 0.5997| **0.7394**| **0.6148**|
|
| 19 |
+
|
| 20 |
+
This model was trained on less than 25% of the dataset yet achieves competitive performance to current sota on open llm leaderboard. Wait for what it's coming!
|
| 21 |
|
| 22 |
### training
|
| 23 |
Training took ~5 P100 gpu hours. It was trained on 15,000 gpt-4 shuffled samples. palmer was fine-tuned using lower learning rates ensuring it keeps as much general knowledge as possible.
|