Spaces:
Running
Running
Update model_api.py
Browse files- model_api.py +24 -8
model_api.py
CHANGED
|
@@ -1,21 +1,37 @@
|
|
| 1 |
from huggingface_hub import InferenceClient
|
| 2 |
import os
|
|
|
|
|
|
|
| 3 |
|
| 4 |
def query_model(prompt):
|
| 5 |
try:
|
| 6 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
| 7 |
-
|
| 8 |
-
client = InferenceClient(
|
| 9 |
-
|
|
|
|
|
|
|
|
|
|
| 10 |
response = client.chat.completions.create(
|
| 11 |
model="Qwen/Qwen2.5-7B-Instruct",
|
| 12 |
messages=[
|
| 13 |
-
{"role": "system", "content": "You are a professional fitness trainer."},
|
| 14 |
{"role": "user", "content": prompt}
|
| 15 |
],
|
| 16 |
-
max_tokens=
|
| 17 |
-
temperature=0.
|
| 18 |
)
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
except Exception as e:
|
| 21 |
-
return
|
|
|
|
| 1 |
from huggingface_hub import InferenceClient
|
| 2 |
import os
|
| 3 |
+
import json
|
| 4 |
+
import re
|
| 5 |
|
| 6 |
def query_model(prompt):
|
| 7 |
try:
|
| 8 |
HF_TOKEN = os.getenv("HF_TOKEN")
|
| 9 |
+
|
| 10 |
+
client = InferenceClient(
|
| 11 |
+
provider="auto",
|
| 12 |
+
api_key=HF_TOKEN
|
| 13 |
+
)
|
| 14 |
+
|
| 15 |
response = client.chat.completions.create(
|
| 16 |
model="Qwen/Qwen2.5-7B-Instruct",
|
| 17 |
messages=[
|
| 18 |
+
{"role": "system", "content": "You are a professional fitness trainer that returns ONLY JSON workout plans."},
|
| 19 |
{"role": "user", "content": prompt}
|
| 20 |
],
|
| 21 |
+
max_tokens=1200,
|
| 22 |
+
temperature=0.6
|
| 23 |
)
|
| 24 |
+
|
| 25 |
+
result = response.choices[0].message.content
|
| 26 |
+
|
| 27 |
+
# Remove markdown JSON blocks if model returns ```json ```
|
| 28 |
+
result = re.sub(r"```json|```", "", result).strip()
|
| 29 |
+
|
| 30 |
+
try:
|
| 31 |
+
result_json = json.loads(result)
|
| 32 |
+
return result_json
|
| 33 |
+
except json.JSONDecodeError:
|
| 34 |
+
return {"error": "Model returned invalid JSON", "raw_output": result}
|
| 35 |
+
|
| 36 |
except Exception as e:
|
| 37 |
+
return {"error": str(e)}
|