Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
base_model:
|
| 6 |
+
- Qwen/Qwen3-30B-A3B
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
> [!NOTE]
|
| 11 |
+
> **Use "You are an assistant with reasoning capabilities." system message to trigger gemini-style thinking.**
|
| 12 |
+
|
| 13 |
+
# Training Dataset
|
| 14 |
+
|
| 15 |
+
- The fine-tuning dataset consists of ~450 diverse examples, 250 of which are directly from Gemini 2.5 Pro.
|
| 16 |
+
|
| 17 |
+
## Trained on:
|
| 18 |
+
- Unsloth version of Qwen3-30B-A3B (instruct).
|
| 19 |
+
- 32k seq_len with examples ranging from 1k to ~20k tokens.
|
| 20 |
+
- Up to 2 turns of conversations.
|
| 21 |
+
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
+
- No benchmark data for now.
|
| 25 |
+
|
| 26 |
+
**Keep in mind that it's slightly overfit since the training dataset was quite small. The model can be used to create more high quality examples for further training.**
|
| 27 |
+
|
| 28 |
+
|
| 29 |
+
|