Plants of India SQL Agent Chatbot

.env.template

@@ -0,0 +1,5 @@
+OPENAI_API_KEY = 'Your OpenAI API Key Here'
+POI_SQL_URI = 'Your SQL URI Here'
+LANGCHAIN_TRACING_V2 = true
+LANGCHAIN_API_KEY = 'Your Langchain API Key Here'
+LANGCHAIN_PROJECT = 'Your Langchain Project Here'

.gitignore

@@ -4,6 +4,7 @@ coursera-dl/
 wandb
 *.ipynb
 *.pem
+qdrant_data/*
 
 # don't expose env files
 .env

.trunk/configs/.isort.cfg

@@ -1,2 +1,2 @@
 [settings]
-profile=black
+profile=google

.trunk/configs/.style.yapf

@@ -1,4 +1,4 @@
 [style]
 based_on_style = google
-column_limit = 120
+column_limit = 140
 indent_width = 2

.trunk/trunk.yaml

@@ -43,7 +43,6 @@ lint:
       paths:
         - .github/**/*
         - .trunk/**/*
-        - mkdocs.yml
         - .DS_Store
         - .vscode/**/*
         - README.md

Dockerfile

@@ -0,0 +1,27 @@
+# Use an official Python runtime as a parent image
+FROM python:3.10-slim
+
+# Set the working directory in the container
+WORKDIR /usr/src/app
+
+
+# Copy the requirements file first to leverage Docker cache
+COPY ai_ta_backend/requirements.txt .
+
+# Install any needed packages specified in requirements.txt
+RUN pip install -r requirements.txt
+
+# Mkdir for sqlite db
+RUN mkdir -p /usr/src/app/db
+
+# Copy the rest of the local directory contents into the container
+COPY . .
+
+# Set the Python path to include the ai_ta_backend directory
+ENV PYTHONPATH="${PYTHONPATH}:/usr/src/app/ai_ta_backend"
+
+# Make port 8000 available to the world outside this container
+EXPOSE 8000
+
+# Run the application using Gunicorn with specified configuration
+CMD ["gunicorn", "--workers=1", "--threads=100", "--worker-class=gthread", "ai_ta_backend.main:app", "--timeout=1800", "--bind=0.0.0.0:8000"]

README.md

@@ -14,6 +14,14 @@ Architecture diagram of Flask + Next.js & React hosted on Vercel.
 
 Automatic [API Reference](https://uiuc-chatbot.github.io/ai-ta-backend/reference/)
 
+## Docker Deployment
+
+1. Just run Docker Compose `docker compose up --build`
+
+Works on version: `Docker Compose version v2.27.1-desktop.1`
+
+Works on Apple Silicon M1 `aarch64`, and `x86`.
+
 ## 📣 Development
 
 1. Rename `.env.template` to `.env` and fill in the required variables
@@ -36,3 +44,4 @@ The docs are auto-built and deployed to [our docs website](https://uiuc-chatbot.
 'url': doc.metadata.get('url'), # wouldn't this error out?
 'base_url': doc.metadata.get('base_url'),
 ```
+

ai_ta_backend/beam/OpenaiEmbeddings.py

@@ -114,7 +114,6 @@
 # # from langchain.vectorstores import Qdrant
 # # from qdrant_client import QdrantClient, models
 
-
 # class OpenAIAPIProcessor:
 
 #   def __init__(self, input_prompts_list, request_url, api_key, max_requests_per_minute, max_tokens_per_minute,
@@ -225,9 +224,9 @@
 #           task_list.append(task)
 #           next_request = None  # reset next_request to empty
 
-#           # print("status_tracker.num_tasks_in_progress", status_tracker.num_tasks_in_progress)
+#           # logging.info("status_tracker.num_tasks_in_progress", status_tracker.num_tasks_in_progress)
 #           # one_task_result = task.result()
-#           # print("one_task_result", one_task_result)
+#           # logging.info("one_task_result", one_task_result)
 
 #       # if all tasks are finished, break
 #       if status_tracker.num_tasks_in_progress == 0:
@@ -263,7 +262,6 @@
 
 #     self.cleaned_results: List[str] = extract_context_from_results(self.results)
 
-
 # def extract_context_from_results(results: List[Any]) -> List[str]:
 #   assistant_contents = []
 #   total_prompt_tokens = 0
@@ -282,10 +280,8 @@
 
 #   return assistant_contents
 
-
 # # dataclasses
 
-
 # @dataclass
 # class StatusTracker:
 #   """Stores metadata about the script's progress. Only one instance is created."""
@@ -299,7 +295,6 @@
 #   num_other_errors: int = 0
 #   time_of_last_rate_limit_error: float = 0  # used to cool off after hitting rate limits
 
-
 # @dataclass
 # class APIRequest:
 #   """Stores an API request's inputs, outputs, and other metadata. Contains a method to make an API call."""
@@ -360,10 +355,8 @@
 
 #       return data
 
-
 # # functions
 
-
 # def api_endpoint_from_url(request_url: str):
 #   """Extract the API endpoint from the request URL."""
 #   if 'text-embedding-ada-002' in request_url:
@@ -372,14 +365,12 @@
 #     match = re.search('^https://[^/]+/v\\d+/(.+)$', request_url)
 #     return match[1]  # type: ignore
 
-
 # def append_to_jsonl(data, filename: str) -> None:
 #   """Append a json payload to the end of a jsonl file."""
 #   json_string = json.dumps(data)
 #   with open(filename, "a") as f:
 #     f.write(json_string + "\n")
 
-
 # def num_tokens_consumed_from_request(
 #     request_json: dict,
 #     api_endpoint: str,
@@ -432,15 +423,13 @@
 #   else:
 #     raise NotImplementedError(f'API endpoint "{api_endpoint}" not implemented in this script')
 
-
 # def task_id_generator_function():
 #   """Generate integers 0, 1, 2, and so on."""
 #   task_id = 0
 #   while True:
 #     yield task_id
 #     task_id += 1
 
-
 # if __name__ == '__main__':
 #   pass
 
@@ -496,8 +485,8 @@
 #   #   total_prompt_tokens = 0
 #   #   total_completion_tokens = 0
 
-#   #   print("Results, end of main: ", oai.results)
-#   #   print("-"*50)
+#   #   logging.info("Results, end of main: ", oai.results)
+#   #   logging.info("-"*50)
 
 #   #   # jsonObject = json.loads(oai.results)
 #   #   for element in oai.results:
@@ -509,21 +498,21 @@
 #   #               total_prompt_tokens += item['usage']['prompt_tokens']
 #   #               total_completion_tokens += item['usage']['completion_tokens']
 
-#   #   print("Assistant Contents:", assistant_contents)
-#   #   print("Total Prompt Tokens:", total_prompt_tokens)
-#   #   print("Total Completion Tokens:", total_completion_tokens)
+#   #   logging.info("Assistant Contents:", assistant_contents)
+#   #   logging.info("Total Prompt Tokens:", total_prompt_tokens)
+#   #   logging.info("Total Completion Tokens:", total_completion_tokens)
 #   #   turbo_total_cost = (total_prompt_tokens * 0.0015) + (total_completion_tokens * 0.002)
-#   #   print("Total cost (3.5-turbo):", (total_prompt_tokens * 0.0015), " + Completions: ", (total_completion_tokens * 0.002), " = ", turbo_total_cost)
+#   #   logging.info("Total cost (3.5-turbo):", (total_prompt_tokens * 0.0015), " + Completions: ", (total_completion_tokens * 0.002), " = ", turbo_total_cost)
 
 #   #   gpt4_total_cost = (total_prompt_tokens * 0.03) + (total_completion_tokens * 0.06)
-#   #   print("Hypothetical cost for GPT-4:", (total_prompt_tokens * 0.03), " + Completions: ", (total_completion_tokens * 0.06), " = ", gpt4_total_cost)
-#   #   print("GPT-4 cost premium: ", (gpt4_total_cost / turbo_total_cost), "x")
+#   #   logging.info("Hypothetical cost for GPT-4:", (total_prompt_tokens * 0.03), " + Completions: ", (total_completion_tokens * 0.06), " = ", gpt4_total_cost)
+#   #   logging.info("GPT-4 cost premium: ", (gpt4_total_cost / turbo_total_cost), "x")
 #   '''
 #   Pricing:
-#   GPT4: 
+#   GPT4:
 #     * $0.03 prompt
 #     * $0.06 completions
-#   3.5-turbo: 
+#   3.5-turbo:
 #     * $0.0015 prompt
 #     * $0.002 completions
 #   '''