Update README.md
Browse files
README.md
CHANGED
@@ -29,6 +29,8 @@ Q8_0 quant is maxed only, as Imatrix has no effect on this quant.
|
|
29 |
|
30 |
F16 is full precision.
|
31 |
|
|
|
|
|
32 |
NOTE:
|
33 |
|
34 |
If you are having issues with Jinja "auto template", use CHATML template.
|
|
|
29 |
|
30 |
F16 is full precision.
|
31 |
|
32 |
+
Context Length: 32 K + 8K output generation. (can be extended to 128k)
|
33 |
+
|
34 |
NOTE:
|
35 |
|
36 |
If you are having issues with Jinja "auto template", use CHATML template.
|