Upload README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,25 @@
|
|
| 1 |
-
---
|
| 2 |
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
license: apache-2.0
|
| 2 |
+
|
| 3 |
+
a DPO LoRA fine-tuned model with preference dataset
|
| 4 |
+
|
| 5 |
+
LoRA Experiment
|
| 6 |
+
|
| 7 |
+
RWKV-5.2-3b-World-DPO is merged model with base
|
| 8 |
+
|
| 9 |
+
Base Model
|
| 10 |
+
|
| 11 |
+
RWKV-5-World-3B-v2-20231113-ctx4096
|
| 12 |
+
|
| 13 |
+
Parameters:
|
| 14 |
+
Lora Rank 8
|
| 15 |
+
Lora Alpha 16
|
| 16 |
+
ctx length 4096
|
| 17 |
+
epoch:19
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
Dataset
|
| 21 |
+
Randomly chosed 1000pairs
|
| 22 |
+
https://huggingface.co/datasets/HuggingFaceH4/ultrafeedback_binarized
|
| 23 |
+
|
| 24 |
+
|
| 25 |
+
|