Commit
·
c308cb8
1
Parent(s):
909b781
Update README.md
Browse files
README.md
CHANGED
|
@@ -27,4 +27,10 @@ The "microsoft/Orca-2-13b" model fully fine-tuned on HuggingFaceH4/no_robots, to
|
|
| 27 |
|
| 28 |
Make sure to comply with the microsoft research license. Please read it before using this model.
|
| 29 |
|
| 30 |
-
This model was trained on the ChatML prompt template.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
|
| 28 |
Make sure to comply with the microsoft research license. Please read it before using this model.
|
| 29 |
|
| 30 |
+
This model was trained on the ChatML prompt template.
|
| 31 |
+
|
| 32 |
+
The responses seen in the inference API were generated using the following sampling parameters:
|
| 33 |
+
temperature = 0.1
|
| 34 |
+
top_p = 0.14
|
| 35 |
+
top_k = 41
|
| 36 |
+
repetition_penalty = 1.176
|