Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -136,6 +136,27 @@ Question:
|
|
| 136 |
return response.choices[0].message.content
|
| 137 |
|
| 138 |
# ... (keep all previous code until chat function)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 139 |
|
| 140 |
def chat(user_input, history):
|
| 141 |
answer = generate_answer(user_input)
|
|
|
|
| 136 |
return response.choices[0].message.content
|
| 137 |
|
| 138 |
# ... (keep all previous code until chat function)
|
| 139 |
+
# Add to app.py
|
| 140 |
+
MODELS = {
|
| 141 |
+
"Fast (Groq)": "llama-3.1-8b-instant",
|
| 142 |
+
"Smart (Groq)": "llama-3.1-70b-versatile",
|
| 143 |
+
"Your Bot": "yourusername/finance-bot" # After fine-tuning
|
| 144 |
+
}
|
| 145 |
+
|
| 146 |
+
model_dropdown = gr.Dropdown(choices=list(MODELS.keys()), value="Fast (Groq)", label="🤖 AI Model")
|
| 147 |
+
|
| 148 |
+
def generate_answer(query, conversation_context="", model_name="Fast (Groq)"):
|
| 149 |
+
# ... RAG logic ...
|
| 150 |
+
model_id = MODELS[model_name]
|
| 151 |
+
|
| 152 |
+
if "Your Bot" in model_name:
|
| 153 |
+
# Use HF Inference API for your model
|
| 154 |
+
response = requests.post("https://api-inference.huggingface.co/models/yourusername/finance-bot",
|
| 155 |
+
json={"inputs": full_prompt})
|
| 156 |
+
return response.json()[0]["generated_text"]
|
| 157 |
+
else:
|
| 158 |
+
# Groq API
|
| 159 |
+
response = client.chat.completions.create(model=model_id, ...)
|
| 160 |
|
| 161 |
def chat(user_input, history):
|
| 162 |
answer = generate_answer(user_input)
|