-
Notifications
You must be signed in to change notification settings - Fork 7
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Bodo is 6x faster in processing, even with only 3 prompts. This uses Gemini Flash, but you can swap other models and try too.
- Loading branch information
1 parent
88d15d9
commit 63252df
Showing
1 changed file
with
158 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,158 @@ | ||
| { | ||
| "cells": [ | ||
| { | ||
| "cell_type": "code", | ||
| "execution_count": 4, | ||
| "metadata": {}, | ||
| "outputs": [ | ||
| { | ||
| "name": "stdout", | ||
| "output_type": "stream", | ||
| "text": [ | ||
| "Processing time: 3.0881738662719727\n", | ||
| " prompt \\\n", | ||
| "0 What is the capital of France? \n", | ||
| "1 What is the meaning of life? \n", | ||
| "2 What is the smoothest part of the universe? \n", | ||
| "\n", | ||
| " response \n", | ||
| "0 Paris\\n \n", | ||
| "1 There's no single, universally accepted answer... \n", | ||
| "2 The smoothest part of the universe, on the lar... \n" | ||
| ] | ||
| } | ||
| ], | ||
| "source": [ | ||
| "\"\"\"Preprocess and query LLMs:\n", | ||
| "- Install llm package using pip install llm\n", | ||
| "- Set keys for the model you'd like to use, using llm keys set [MODEL]\n", | ||
| "\"\"\"\n", | ||
| "\n", | ||
| "import pandas as pd\n", | ||
| "import time\n", | ||
| "from dotenv import load_dotenv\n", | ||
| "import llm\n", | ||
| "\n", | ||
| "load_dotenv()\n", | ||
| "MODEL = \"gemini-1.5-flash-8b-latest\"\n", | ||
| "model = llm.get_model(MODEL) \n", | ||
| "\n", | ||
| "def query_model(prompt):\n", | ||
| " \"\"\"\n", | ||
| " Sends a prompt to the AI Suite and returns the response.\n", | ||
| " \"\"\"\n", | ||
| " response = model.prompt(prompt) \n", | ||
| " return response.text()\n", | ||
| "\n", | ||
| "def query_model_all(df):\n", | ||
| " \"\"\"Clean up prompts and query the model for all prompts in the dataframe.\"\"\"\n", | ||
| " t0 = time.time()\n", | ||
| " cleaned_prompts = df[\"prompt\"].str.strip().str.lower()\n", | ||
| " df[\"response\"] = cleaned_prompts.map(query_model)\n", | ||
| " print(\"Processing time:\", time.time() - t0)\n", | ||
| " return df\n", | ||
| "\n", | ||
| "\n", | ||
| "if __name__ == \"__main__\":\n", | ||
| " raw_prompts = [\n", | ||
| " \" What is the capital of France? \",\n", | ||
| " \" What is the meaning of life? \",\n", | ||
| " \" What is the smoothest part of the universe?\",\n", | ||
| " ]\n", | ||
| "\n", | ||
| " # Repeat prompts 10 times for testing purposes\n", | ||
| " df = pd.DataFrame({\"prompt\": raw_prompts})\n", | ||
| " out_df = query_model_all(df)\n", | ||
| " print(out_df)" | ||
| ] | ||
| }, | ||
| { | ||
| "cell_type": "code", | ||
| "execution_count": 3, | ||
| "metadata": {}, | ||
| "outputs": [ | ||
| { | ||
| "name": "stdout", | ||
| "output_type": "stream", | ||
| "text": [ | ||
| "Processing time: 0.4763470000000325\n", | ||
| " prompt response\n", | ||
| "0 What is the capital of France? Paris\n", | ||
| "\n", | ||
| "1 What is the meaning of life in one word? Purpose\n", | ||
| "\n", | ||
| "2 What is the smoothest part of the universe, i... Cosmic\n", | ||
| "\n" | ||
| ] | ||
| } | ||
| ], | ||
| "source": [ | ||
| "\"\"\"Preprocess and query LLMs, Bodo edition:\n", | ||
| "- Install llm package using pip install llm\n", | ||
| "- Set keys for the model you'd like to use, using llm keys set [MODEL]\n", | ||
| "\"\"\"\n", | ||
| "\n", | ||
| "import pandas as pd\n", | ||
| "import bodo\n", | ||
| "import time\n", | ||
| "from dotenv import load_dotenv\n", | ||
| "import llm\n", | ||
| "\n", | ||
| "load_dotenv()\n", | ||
| "MODEL = \"gemini-1.5-flash-8b-latest\"\n", | ||
| "model = llm.get_model(MODEL) \n", | ||
| "\n", | ||
| "@bodo.wrap_python(bodo.string_type)\n", | ||
| "def query_model(prompt):\n", | ||
| " \"\"\"\n", | ||
| " Sends a prompt to the AI Suite and returns the response.\n", | ||
| " \"\"\"\n", | ||
| " response = model.prompt(prompt) \n", | ||
| " return response.text()\n", | ||
| "\n", | ||
| "@bodo.jit\n", | ||
| "def query_model_all(df):\n", | ||
| " \"\"\"Clean up prompts and query the model for all prompts in the dataframe.\"\"\"\n", | ||
| " t0 = time.time()\n", | ||
| " cleaned_prompts = df[\"prompt\"].str.strip().str.lower()\n", | ||
| " df[\"response\"] = cleaned_prompts.map(query_model)\n", | ||
| " print(\"Processing time:\", time.time() - t0)\n", | ||
| " return df\n", | ||
| "\n", | ||
| "\n", | ||
| "if __name__ == \"__main__\":\n", | ||
| " raw_prompts = [\n", | ||
| " \" What is the capital of France? \",\n", | ||
| " \" What is the meaning of life? \",\n", | ||
| " \" What is the smoothest part of the universe?\",\n", | ||
| " ]\n", | ||
| "\n", | ||
| " # Repeat prompts 10 times for testing purposes\n", | ||
| " df = pd.DataFrame({\"prompt\": raw_prompts})\n", | ||
| " out_df = query_model_all(df)\n", | ||
| " print(out_df)" | ||
| ] | ||
| } | ||
| ], | ||
| "metadata": { | ||
| "kernelspec": { | ||
| "display_name": "global_env", | ||
| "language": "python", | ||
| "name": "python3" | ||
| }, | ||
| "language_info": { | ||
| "codemirror_mode": { | ||
| "name": "ipython", | ||
| "version": 3 | ||
| }, | ||
| "file_extension": ".py", | ||
| "mimetype": "text/x-python", | ||
| "name": "python", | ||
| "nbconvert_exporter": "python", | ||
| "pygments_lexer": "ipython3", | ||
| "version": "3.11.6" | ||
| } | ||
| }, | ||
| "nbformat": 4, | ||
| "nbformat_minor": 2 | ||
| } |