Spaces:
Sleeping
Sleeping
40 (10%) of token overlap in the tiktkon(izer)
Browse files
app.py
CHANGED
|
@@ -29,7 +29,7 @@ def tiktoken_len(text):
|
|
| 29 |
|
| 30 |
text_splitter = RecursiveCharacterTextSplitter(
|
| 31 |
chunk_size=400,
|
| 32 |
-
chunk_overlap=
|
| 33 |
length_function=tiktoken_len,
|
| 34 |
separators=["\n\n", "\n", " ", ""]
|
| 35 |
)
|
|
|
|
| 29 |
|
| 30 |
text_splitter = RecursiveCharacterTextSplitter(
|
| 31 |
chunk_size=400,
|
| 32 |
+
chunk_overlap=40,
|
| 33 |
length_function=tiktoken_len,
|
| 34 |
separators=["\n\n", "\n", " ", ""]
|
| 35 |
)
|