Added info about context size
Browse files
README.md
CHANGED
|
@@ -15,6 +15,11 @@ for example:
|
|
| 15 |
```
|
| 16 |
cat technical.txt multilingual.txt wiki.txt >custom.matrix
|
| 17 |
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
|
| 19 |
You will find below descriptions for the various input files provided, to help you choose the correct one.
|
| 20 |
|
|
|
|
| 15 |
```
|
| 16 |
cat technical.txt multilingual.txt wiki.txt >custom.matrix
|
| 17 |
```
|
| 18 |
+
Note on **context size** when generating the matrix: in general, a small context size such as 512 is recommended, and community
|
| 19 |
+
tests have shown it usually performs than a larger one such as 4096. However, I would argue this is is highly dependent on the
|
| 20 |
+
source data you are using: with random tokens or short text a small context makes sense; but when using larger texts, a larger
|
| 21 |
+
context matching the size of the texts might be a better choice. Remember that the size is in tokens, which roughly translates
|
| 22 |
+
to number of words, not characters.
|
| 23 |
|
| 24 |
You will find below descriptions for the various input files provided, to help you choose the correct one.
|
| 25 |
|