Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,14 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: llama3.1
|
| 3 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: llama3.1
|
| 3 |
+
---
|
| 4 |
+
| | Wiki | C4 | PIQA | ARC-E | ARC-C | HellaSwag | Wino | Avg. |
|
| 5 |
+
| ----------- | ---- | ----- | ----- | ----- | ----- | --------- | ----- | ----- |
|
| 6 |
+
| Unquantized | TBA | TBA | TBA | TBA | TBA | TBA | TBA | TBA |
|
| 7 |
+
| W4G64 | TBA | TBA | TBA | TBA | TBA | TBA | TBA | TBA |
|
| 8 |
+
| W3G64 | TBA | TBA | TBA | TBA | TBA | TBA | TBA | TBA |
|
| 9 |
+
|
| 10 |
+
Revisions available in this repository:
|
| 11 |
+
- `main` (W4G64, scales learned);
|
| 12 |
+
- `nfl_w3g64` (W3G64, scales learned);
|
| 13 |
+
|
| 14 |
+
Evaluations are provided for models with learned scales.<br>Benchmark scores (zero-shot) are computed with [`lm-evaluation-harness`](https://github.com/EleutherAI/lm-evaluation-harness).
|