Spaces:
Sleeping
Sleeping
File size: 2,542 Bytes
b9ffe6a 500c1ba b9ffe6a 500c1ba a45abf6 500c1ba b9ffe6a 1963268 3bfdeb9 b9ffe6a 46fb1dc 0c1bd23 46fb1dc 500c1ba b687ff9 500c1ba b687ff9 c6afeca 500c1ba b9ffe6a |
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 |
import logging
import os
from openai import OpenAI
from openai import AzureOpenAI
from openai import OpenAIError, RateLimitError
# from config import OPENAI_API_KEY, AZURE_OPENAI_KEY
# Initialize the OpenAI client with the API key from the environment variable
#api_key = os.getenv('OPENAI_API_KEY')
# client = OpenAI(api_key=OPENAI_API_KEY)
AZURE_OPENAI_ENDPOINT = os.getenv('AZURE_OPENAI_ENDPOINT')
AZURE_API_VERSION = "2023-03-15-preview" # API version for Azure OpenAI
AZURE_OPENAI_KEY = os.getenv('AZURE_OPENAI_API_KEY')
# Initialize the Azure OpenAI client with the endpoint and API key
try:
client = AzureOpenAI(
azure_endpoint=AZURE_OPENAI_ENDPOINT,
api_key=AZURE_OPENAI_KEY,
api_version=AZURE_API_VERSION
)
print('Azure OpenAI working')
except Exception as e:
logging.error(f"Azure OpenAI API error: {e}")
def generate_rag_response(json_output, user_query):
logging.info("Generating RAG response")
# Extract text from the JSON output
context_texts = [hit['chunk_text'] for hit in json_output]
# Create the context for the prompt
context = "\n".join(context_texts)
prompt = f"Based on the given context, answer the user query: {user_query}\nContext:\n{context} and Employ references to the ID of articles provided [ID], ensuring their relevance to the query. The referencing should always be in the format of [1][2]... etc. </instructions> "
main_prompt = [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": prompt}
]
try:
# Create a chat completion request
chat_completion = client.chat.completions.create(
messages=main_prompt,
model="urdu-llama", # Use the gpt-4o-mini model
max_tokens=2000, # Limit the maximum number of tokens in the response
temperature=0.5
)
# Log the response from the model
logging.info("RAG response generation completed")
logging.info(f"RAG response: {chat_completion.choices[0].message.content}")
return chat_completion.choices[0].message.content, None
except RateLimitError as e:
logging.error(f"Rate limit exceeded: {e}")
return None, "Rate limit exceeded. Please try again later."
except OpenAIError as e:
logging.error(f"OpenAI API error: {e}")
return None, f"An error occurred: {str(e)}"
except Exception as e:
logging.error(f"Unexpected error: {e}")
return None, str(e)
|