File size: 7,363 Bytes
c0658b2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cb0bf83
 
f88a286
cb0bf83
e1d21ef
cb0bf83
ad93aea
cb0bf83
 
f88a286
 
 
 
cb0bf83
 
 
c0658b2
 
cb0bf83
 
 
 
c0658b2
cb0bf83
 
 
f88a286
cb0bf83
c0658b2
cb0bf83
ad93aea
 
 
 
 
c0658b2
ad93aea
 
 
 
cb0bf83
c0658b2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ad93aea
 
 
c0658b2
ad93aea
 
c0658b2
ad93aea
0cef40a
 
ad93aea
 
 
 
 
0cef40a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cb0bf83
f88a286
cb0bf83
 
 
 
 
f88a286
c0658b2
 
 
 
f88a286
cb0bf83
f88a286
cb0bf83
 
f88a286
cb0bf83
 
 
 
c0658b2
f88a286
 
 
e1d21ef
 
 
 
f88a286
 
 
e1d21ef
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
# # app.py
# import os
# import logging
# from fastapi import FastAPI, HTTPException
# from fastapi.responses import JSONResponse
# from pydantic import BaseModel
# from huggingface_hub import InferenceClient
# from typing import Optional

# # Set up logging
# logging.basicConfig(level=logging.INFO)
# logger = logging.getLogger(__name__)

# # Initialize FastAPI app
# app = FastAPI(
#     title="LLM Chat API",
#     description="API for getting chat responses from Llama model",
#     version="1.0.0"
# )

# class ChatRequest(BaseModel):
#     text: str

# class ChatResponse(BaseModel):
#     response: str
#     status: str

# def llm_chat_response(text: str) -> str:
#     try:
#         HF_TOKEN = os.getenv("HF_TOKEN")
#         logger.info("Checking HF_TOKEN...")
#         if not HF_TOKEN:
#             logger.error("HF_TOKEN not found in environment variables")
#             raise HTTPException(status_code=500, detail="HF_TOKEN not configured")
        
#         logger.info("Initializing InferenceClient...")
#         client = InferenceClient(
#             provider="sambanova",
#             api_key=HF_TOKEN
#         )
        
#         messages = [
#             {
#                 "role": "user",
#                 "content": [
#                     {
#                         "type": "text",
#                         "text": text + " describe in one line only"
#                     }
#                 ]
#             }
#         ]
        
#         logger.info("Sending request to model...")
#         completion = client.chat.completions.create(
#             model="meta-llama/Llama-3.2-11B-Vision-Instruct",
#             messages=messages,
#             max_tokens=500
#         )
#         return completion.choices[0].message['content']

#     except Exception as e:
#         logger.error(f"Error in llm_chat_response: {str(e)}")
#         raise HTTPException(status_code=500, detail=str(e))

# @app.post("/chat", response_model=ChatResponse)
# async def chat(request: ChatRequest):
#     try:
#         logger.info(f"Received chat request with text: {request.text}")
#         response = llm_chat_response(request.text)
#         return ChatResponse(response=response, status="success")
#     except HTTPException as he:
#         logger.error(f"HTTP Exception in chat endpoint: {str(he)}")
#         raise he
#     except Exception as e:
#         logger.error(f"Unexpected error in chat endpoint: {str(e)}")
#         raise HTTPException(status_code=500, detail=str(e))

# @app.get("/")
# async def root():
#     return {"message": "Welcome to the LLM Chat API. Use POST /chat endpoint to get responses."}

# @app.exception_handler(404)
# async def not_found_handler(request, exc):
#     return JSONResponse(
#         status_code=404,
#         content={"error": "Endpoint not found. Please use POST /chat for queries."}
#     )

# @app.exception_handler(405)
# async def method_not_allowed_handler(request, exc):
#     return JSONResponse(
#         status_code=405,
#         content={"error": "Method not allowed. Please check the API documentation."}
#     )


# app.py
import os
import logging
from fastapi import FastAPI, HTTPException
from fastapi.responses import JSONResponse
from pydantic import BaseModel
from huggingface_hub import InferenceClient
from typing import Optional

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

# Initialize FastAPI app
app = FastAPI(
    title="LLM Chat API",
    description="API for getting chat responses from Llama model with image support",
    version="1.1.0"
)

class ChatRequest(BaseModel):
    text: str
    image_url: Optional[str] = None

class ChatResponse(BaseModel):
    response: str
    status: str

def llm_chat_response(text: str, image_url: Optional[str] = None) -> str:
    try:
        HF_TOKEN = os.getenv("HF_TOKEN")
        logger.info("Checking HF_TOKEN...")
        if not HF_TOKEN:
            logger.error("HF_TOKEN not found in environment variables")
            raise HTTPException(status_code=500, detail="HF_TOKEN not configured")

        logger.info("Initializing InferenceClient...")
        client = InferenceClient(
            provider="sambanova",
            api_key=HF_TOKEN
        )

        # Prepare content list for the message
        content = [
            {
                "type": "text",
                "text": text + " describe in one line only"
            }
        ]

        # Add image to content if provided
        if image_url:
            logger.info(f"Adding image URL to request: {image_url}")
            content.append({
                "type": "image_url",
                "image_url": {
                    "url": image_url
                }
            })

        messages = [
            {
                "role": "user",
                "content": content
            }
        ]

        logger.info("Sending request to model...")
        logger.info(f"Request payload: {messages}")
        
        completion = client.chat.completions.create(
            model="meta-llama/Llama-3.2-11B-Vision-Instruct",
            messages=messages,
            max_tokens=500
        )
        
        logger.info(f"Response received: {completion}")
        
        # Check the structure of the response and extract content
        if hasattr(completion, 'choices') and len(completion.choices) > 0:
            message = completion.choices[0].message
            # Handle different response formats
            if isinstance(message, dict) and 'content' in message:
                return message['content']
            elif hasattr(message, 'content'):
                return message.content
            else:
                logger.error(f"Unexpected message format: {message}")
                return str(message)
        else:
            logger.error(f"Unexpected completion format: {completion}")
            return str(completion)
            
    except Exception as e:
        logger.error(f"Error in llm_chat_response: {str(e)}")
        raise HTTPException(status_code=500, detail=str(e))

@app.post("/chat", response_model=ChatResponse)
async def chat(request: ChatRequest):
    try:
        logger.info(f"Received chat request with text: {request.text}")
        if request.image_url:
            logger.info(f"Image URL included: {request.image_url}")
        
        response = llm_chat_response(request.text, request.image_url)
        return ChatResponse(response=response, status="success")
    except HTTPException as he:
        logger.error(f"HTTP Exception in chat endpoint: {str(he)}")
        raise he
    except Exception as e:
        logger.error(f"Unexpected error in chat endpoint: {str(e)}")
        raise HTTPException(status_code=500, detail=str(e))

@app.get("/")
async def root():
    return {"message": "Welcome to the LLM Chat API with image support. Use POST /chat endpoint to get responses."}

@app.exception_handler(404)
async def not_found_handler(request, exc):
    return JSONResponse(
        status_code=404,
        content={"error": "Endpoint not found. Please use POST /chat for queries."}
    )

@app.exception_handler(405)
async def method_not_allowed_handler(request, exc):
    return JSONResponse(
        status_code=405,
        content={"error": "Method not allowed. Please check the API documentation."}
    )