Update README.md
Browse files
README.md
CHANGED
|
@@ -1,12 +1,11 @@
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
-
- Sweaterdog/Andy-4-base
|
| 4 |
-
- Sweaterdog/Andy-4-base-2
|
| 5 |
- Sweaterdog/Andy-4-ft
|
| 6 |
language:
|
| 7 |
- en
|
| 8 |
base_model:
|
| 9 |
-
- unsloth/
|
| 10 |
tags:
|
| 11 |
- gaming
|
| 12 |
- minecraft
|
|
@@ -34,7 +33,7 @@ tags:
|
|
| 34 |
- **Tokens:** 425,535,198
|
| 35 |
- **Conversations:** 62,149
|
| 36 |
|
| 37 |
-
- **Base Architecture:**
|
| 38 |
- **License:** [Andy 1.1 License](LICENSE)
|
| 39 |
- **Repository:** https://huggingface.co/Sweaterdog/Andy‑4
|
| 40 |
|
|
@@ -43,15 +42,10 @@ tags:
|
|
| 43 |
## 📊 Training Regimen
|
| 44 |
|
| 45 |
1. **Andy‑4‑base‑1** dataset
|
| 46 |
-
- **Epochs:**
|
| 47 |
- **Learning Rate:** 7e-5
|
| 48 |
- **Dataset Size:** 47.4k
|
| 49 |
|
| 50 |
-
2. **Andy‑4‑base‑2** dataset
|
| 51 |
-
- **Epochs:** 4
|
| 52 |
-
- **Learning Rate:** 3e-7
|
| 53 |
-
- **Dataset Size:** 48.9k
|
| 54 |
-
|
| 55 |
3. **Fine‑tune (FT) dataset**
|
| 56 |
- **Epochs:** 2.5
|
| 57 |
- **Learning Rate:** 2e-5
|
|
|
|
| 1 |
---
|
| 2 |
datasets:
|
| 3 |
+
- Sweaterdog/Andy-4-base
|
|
|
|
| 4 |
- Sweaterdog/Andy-4-ft
|
| 5 |
language:
|
| 6 |
- en
|
| 7 |
base_model:
|
| 8 |
+
- unsloth/Qwen3-8B-bnb-4bit
|
| 9 |
tags:
|
| 10 |
- gaming
|
| 11 |
- minecraft
|
|
|
|
| 33 |
- **Tokens:** 425,535,198
|
| 34 |
- **Conversations:** 62,149
|
| 35 |
|
| 36 |
+
- **Base Architecture:** Qwen 3 8b
|
| 37 |
- **License:** [Andy 1.1 License](LICENSE)
|
| 38 |
- **Repository:** https://huggingface.co/Sweaterdog/Andy‑4
|
| 39 |
|
|
|
|
| 42 |
## 📊 Training Regimen
|
| 43 |
|
| 44 |
1. **Andy‑4‑base‑1** dataset
|
| 45 |
+
- **Epochs:** 3
|
| 46 |
- **Learning Rate:** 7e-5
|
| 47 |
- **Dataset Size:** 47.4k
|
| 48 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
3. **Fine‑tune (FT) dataset**
|
| 50 |
- **Epochs:** 2.5
|
| 51 |
- **Learning Rate:** 2e-5
|