Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,35 @@
|
|
1 |
---
|
2 |
license: cc-by-nc-sa-4.0
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
license: cc-by-nc-sa-4.0
|
3 |
+
datasets:
|
4 |
+
- roneneldan/TinyStories
|
5 |
---
|
6 |
+
This is a character (english a-z 0-9 and so on) trained model following Andrej karpathy's llama.c project https://github.com/karpathy/llama2.c on both TinyStories and my own internal similar dataset I made.
|
7 |
+
|
8 |
+
for it to see/output Uppercase letters this model uses a Shift-Key modifier before the letter to become uppercase, and has never been trained on actual uppercase letters.
|
9 |
+
|
10 |
+
This modifier is ↨ and here are the functions I use to convert from straight text to the modified format and back.
|
11 |
+
```
|
12 |
+
def add_caseifer(text):
|
13 |
+
# Using list comprehension for more efficient concatenation
|
14 |
+
return ''.join(['↨' + char.lower() if char.isupper() else char for char in text
|
15 |
+
|
16 |
+
def remove_caseifer(text):
|
17 |
+
new_text = ""
|
18 |
+
i = 0
|
19 |
+
while i < len(text):
|
20 |
+
if text[i] == "↨":
|
21 |
+
if i+1 < len(text):
|
22 |
+
new_text += text[i+1].upper()
|
23 |
+
i += 1
|
24 |
+
else:
|
25 |
+
pass # skip this index
|
26 |
+
else:
|
27 |
+
new_text += text[i]
|
28 |
+
i += 1
|
29 |
+
return new_text
|
30 |
+
```
|
31 |
+
|
32 |
+
As such for test strings to use in chat try using somthing like:
|
33 |
+
```
|
34 |
+
↨hello, my name is ↨clara and ↨i like
|
35 |
+
```
|