Update README.md
Browse files
README.md
CHANGED
|
@@ -59,11 +59,11 @@ This is a transformer-based question-answering model fine-tuned from `deepseek-a
|
|
| 59 |
To generate accurate completions, prompt the model in the following structure:
|
| 60 |
|
| 61 |
```
|
| 62 |
-
|
| 63 |
-
Question: If Sally has 3 apples and buys 2 more, how many does she have in total? <think>
|
| 64 |
-
|
| 65 |
```
|
| 66 |
|
|
|
|
|
|
|
| 67 |
The model will continue reasoning within `<think>...</think>` and provide a final answer inside `<answer>...</answer>`.
|
| 68 |
|
| 69 |
---
|
|
|
|
| 59 |
To generate accurate completions, prompt the model in the following structure:
|
| 60 |
|
| 61 |
```
|
| 62 |
+
<question>Question: If Sally has 3 apples and buys 2 more, how many does she have in total?</question>
|
|
|
|
|
|
|
| 63 |
```
|
| 64 |
|
| 65 |
+
Be aware that this token `</question>` will prompt the answer to start with `<think>` which is trained into the model based on training data.
|
| 66 |
+
|
| 67 |
The model will continue reasoning within `<think>...</think>` and provide a final answer inside `<answer>...</answer>`.
|
| 68 |
|
| 69 |
---
|