Update README.md
Browse files
README.md
CHANGED
@@ -28,18 +28,19 @@ datasets:
|
|
28 |
|
29 |
微调数据集来源: ([GabrielCheng/Drone-flight-monitoring-reasoning-SFT](https://huggingface.co/datasets/GabrielCheng/Drone-flight-monitoring-reasoning-SFT))
|
30 |
|
31 |
-
|
32 |
|
33 |
### 模型调用示例
|
34 |
```python
|
35 |
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
|
|
36 |
model_id = "GabrielCheng/Deepseek-r1-finetuned-drone-safty"
|
37 |
model = AutoModelForCausalLM.from_pretrained(model_id).to("cuda")
|
38 |
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
39 |
|
40 |
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
|
41 |
|
42 |
-
question = "
|
43 |
|
44 |
prompt = f"""以下指令描述了一项任务,并附带了相关背景信息。
|
45 |
请用中文编写一个回复,以恰当地完成此任务请求。
|
|
|
28 |
|
29 |
微调数据集来源: ([GabrielCheng/Drone-flight-monitoring-reasoning-SFT](https://huggingface.co/datasets/GabrielCheng/Drone-flight-monitoring-reasoning-SFT))
|
30 |
|
31 |
+
注:微调数据集中只有问答文本数据,没有实时的飞行轨迹、环境信息等数据。所以模型不具备在真实动态场景中的实用性。仅用于展示形式化的微调训练的效果。
|
32 |
|
33 |
### 模型调用示例
|
34 |
```python
|
35 |
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
36 |
+
|
37 |
model_id = "GabrielCheng/Deepseek-r1-finetuned-drone-safty"
|
38 |
model = AutoModelForCausalLM.from_pretrained(model_id).to("cuda")
|
39 |
tokenizer = AutoTokenizer.from_pretrained(model_id)
|
40 |
|
41 |
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
|
42 |
|
43 |
+
question = "在低能见度环境下,如何利用飞行轨迹数据综合评估无人机的安全风险?"
|
44 |
|
45 |
prompt = f"""以下指令描述了一项任务,并附带了相关背景信息。
|
46 |
请用中文编写一个回复,以恰当地完成此任务请求。
|