Commit
•
1f0de3a
1
Parent(s):
30e4893
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# flan-ul2 4-bit 128-groupsize GPTQ
|
| 2 |
Quantized using qwopqwop200's GPTQ-for-Llama repo on the t5 branch.<br>
|
| 3 |
Original model can be found here: [Google/flan-ul2](https://huggingface.co/google/flan-ul2)
|
|
@@ -36,6 +48,3 @@ Average accuracy 0.613 - other (business, health, misc.)
|
|
| 36 |
MMLU Average accuracy: 0.540
|
| 37 |
```
|
| 38 |
|
| 39 |
-
---
|
| 40 |
-
license: apache-2.0
|
| 41 |
-
---
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
- fr
|
| 5 |
+
- ro
|
| 6 |
+
- de
|
| 7 |
+
- multilingual
|
| 8 |
+
thumbnail: "url to a thumbnail used in social sharing"
|
| 9 |
+
license: apache-2.0
|
| 10 |
+
metrics:
|
| 11 |
+
- mmlu
|
| 12 |
+
---
|
| 13 |
# flan-ul2 4-bit 128-groupsize GPTQ
|
| 14 |
Quantized using qwopqwop200's GPTQ-for-Llama repo on the t5 branch.<br>
|
| 15 |
Original model can be found here: [Google/flan-ul2](https://huggingface.co/google/flan-ul2)
|
|
|
|
| 48 |
MMLU Average accuracy: 0.540
|
| 49 |
```
|
| 50 |
|
|
|
|
|
|
|
|
|