Commit
•
8792875
1
Parent(s):
2648f81
Update README.md
Browse files
README.md
CHANGED
|
@@ -64,7 +64,6 @@ Zyda also outperforms Dolma, RefinedWeb, and Fineweb on 1.4B models trained on 5
|
|
| 64 |
|
| 65 |
According to our evaluations, Zyda is the most performant per-token open dataset available in its non-starcoder variant on language tasks and tying with fineweb otherwise.
|
| 66 |
|
| 67 |
-
|
| 68 |
|
| 69 |
|
| 70 |
## How to download
|
|
@@ -106,10 +105,8 @@ Dataset fields:
|
|
| 106 |
- `filtering_features`: precomputed values of different features that were used for filtering (converted to json string)
|
| 107 |
- `source_other`: metadata from the source dataset (converted to json string)
|
| 108 |
|
| 109 |
-
<center>
|
| 110 |
-
|
| 111 |
-
</center>
|
| 112 |
|
|
|
|
| 113 |
|
| 114 |
### Source Data
|
| 115 |
|
|
|
|
| 64 |
|
| 65 |
According to our evaluations, Zyda is the most performant per-token open dataset available in its non-starcoder variant on language tasks and tying with fineweb otherwise.
|
| 66 |
|
|
|
|
| 67 |
|
| 68 |
|
| 69 |
## How to download
|
|
|
|
| 105 |
- `filtering_features`: precomputed values of different features that were used for filtering (converted to json string)
|
| 106 |
- `source_other`: metadata from the source dataset (converted to json string)
|
| 107 |
|
|
|
|
|
|
|
|
|
|
| 108 |
|
| 109 |
+
|
| 110 |
|
| 111 |
### Source Data
|
| 112 |
|