Update README.md
Browse files
README.md
CHANGED
|
@@ -2,8 +2,10 @@
|
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
|
| 5 |
-
Use rwkv pip package 0.8.
|
| 6 |
|
| 7 |
Very interesting. RWKV-5 is great at benchmarks (excellent zeroshot performance), but generates quite worse music (just like GPT models) despite lower loss.
|
| 8 |
|
| 9 |
This fits my theory: Dot-product is good for uncreative work, while Channelwise is good for creative work.
|
|
|
|
|
|
|
|
|
| 2 |
license: apache-2.0
|
| 3 |
---
|
| 4 |
|
| 5 |
+
Use rwkv pip package 0.8.6+ for RWKV-5. Might overflow in fp16. Use fp32.
|
| 6 |
|
| 7 |
Very interesting. RWKV-5 is great at benchmarks (excellent zeroshot performance), but generates quite worse music (just like GPT models) despite lower loss.
|
| 8 |
|
| 9 |
This fits my theory: Dot-product is good for uncreative work, while Channelwise is good for creative work.
|
| 10 |
+
|
| 11 |
+
Therefore use https://huggingface.co/BlinkDL/rwkv-4-music for better music. Or let's see if someone can find a better sampling method to improve RWKV-5 results.
|