Update models/xverse-13B/README.md
Browse files- models/xverse-13B/README.md +8 -10
models/xverse-13B/README.md
CHANGED
@@ -15,13 +15,11 @@ For specific information, consult original developers or refer to the Original l
|
|
15 |
* Split and Isolated
|
16 |
* Numbers : (code & math, compression ratios)
|
17 |
* Split and Isolated(invidiual_digits=true)
|
18 |
-
##
|
19 |
-
* 100257
|
20 |
-
* Around 96567 are single width character unicode characters (WIP)
|
21 |
-
|
22 |
-
*
|
23 |
-
*
|
24 |
-
|
25 |
-
*
|
26 |
-
* Most multi-char tokens are english or chinese
|
27 |
-
|
|
|
15 |
* Split and Isolated
|
16 |
* Numbers : (code & math, compression ratios)
|
17 |
* Split and Isolated(invidiual_digits=true)
|
18 |
+
## Vocabulary
|
19 |
+
* Size : 100257
|
20 |
+
* Around 96567 are single width character unicode characters (WIP)
|
21 |
+
* Average Length : WIP
|
22 |
+
* Median Length : 1 char
|
23 |
+
* Languages
|
24 |
+
* "[multiple languages including](https://huggingface.co/xverse/XVERSE-13B/discussions/5) ar, bg, ca, cs, da, de, el, en, es, et, fa, fi, fr, hi, hu, id, it, iw, ja, kk, ko, lt, lv, mr, ms, nl, no, pl, pt, ro, ru, sk, sl, sr, sv, ta, th, tr, uk, vi, and zh"
|
25 |
+
* Most multi-char tokens are english or chinese
|
|
|
|