diff --git a/README.md b/README.md
index a06758ed..c0cf15a3 100644
--- a/README.md
+++ b/README.md
@@ -101,16 +101,17 @@ The AI gateway supports OpenAI compatible endpoints with added parameter support
## Gateway Cookbooks
-### π Trending Cookbooks
-* [Run Gateway on prompts from Langchain hub](/cookbook/use-cases/run-gateway-on-prompts-from-langchain-hub.md)
-* [Use Porkey Gateway with Vercel's AI SDK](/cookbook/integrations/vercel-ai.md)
-* [Set up fallback from SDXL to Dall-E-3](/cookbook/getting-started/fallback-from-stable-diffusion-to-dall-e.ipynb)
-
-### β¨ Latest Cookbooks
-* [Comparing Top 10 LMSYS Models with Portkey](/cookbook/use-cases/LMSYS%20Series/comparing-top10-LMSYS-models-with-Portkey.ipynb)
-* [Fallback from OpenAI to Azure OpenAI](/cookbook/getting-started/fallback-from-openai-to-azure.ipynb)
-* [Set up automatic retries for failed requests](/cookbook/getting-started/automatic-retries-on-failures.md)
-* [Call Llama 3 on Groq](/cookbook/use-cases/llama-3-on-groq.ipynb)
+### Trending Cookbooks
+- Use models from [Nvidia NIM](/cookbook/providers/nvidia.ipynb) with AI Gateway
+- Monitor [CrewAI Agents](/cookbook/monitoring-agents/CrewAI_with_Telemetry.ipynb) with Portkey!
+- Comparing [Top 10 LMSYS Models](./use-cases/LMSYS%20Series/comparing-top10-LMSYS-models-with-Portkey.ipynb) with AI Gateway.
+
+### Latest Cookbooks
+* [Create Synthetic Datasets using Nemotron](/cookbook/use-cases/Nemotron_GPT_Finetuning_Portkey.ipynb)
+* [Use Portkey Gateway with Vercel's AI SDK](/cookbook/integrations/vercel-ai.md)
+* [Monitor Llama Agents with Portkey](/cookbook/monitoring-agents/Llama_Agents_with_Telemetry.ipynb)
+
+
### [More Examples](https://github.com/Portkey-AI/gateway/tree/main/cookbook)
diff --git a/cookbook/README.md b/cookbook/README.md
index ec77916e..469088d0 100644
--- a/cookbook/README.md
+++ b/cookbook/README.md
@@ -7,36 +7,54 @@
[](https://portkey.ai/community)
[](https://twitter.com/portkeyai)
+## Table of Contents
+
+Please use the below table of contents to navigate through the cookbook.
+
+### Latest Notebooks
+
+- Use models from [Nvidia NIM](/cookbook/providers/nvidia.ipynb) with AI Gateway
+- Monitor [Llama Agents](/cookbook/monitoring-agents/Llama_Agents_with_Telemetry.ipynb) with Portkey!
+- Create [Synthetic Datasets](/cookbook/use-cases/Nemotron_GPT_Finetuning_Portkey.ipynb) using Nemotron-340B
+- Comparing [Top 10 LMSYS Models](./use-cases/LMSYS%20Series/comparing-top10-LMSYS-models-with-Portkey.ipynb) with AI Gateway.
+
+
## getting-started
+* [Gentle introduction to Portkey Gateway](./getting-started/gentle-introduction-to-portkey-gateway.ipynb)
* [Use Portkey cache to save LLM cost & time](./getting-started/enable-cache.md)
* [Retry automatically on LLM failures](./getting-started/automatic-retries-on-failures.md)
* [Image generation with Gateway](./getting-started/image-generation.ipynb)
* [Writing your first Gateway Config](./getting-started/writing-your-first-gateway-config.md)
-* [Gentle introduction to Portkey Gateway](./getting-started/gentle-introduction-to-portkey-gateway.ipynb)
* [Automatically Fallback from OpenAI to Azure](./getting-started/fallback-from-openai-to-azure.ipynb)
-* [Setting up resilient Load balancers with failure-mitigating Fallbacks](./getting-started/resilient-loadbalancing-with-failure-mitigating-fallbacks.md)
-* [Set up Fallback from Stable Diffusion to Dall-E](./getting-started/fallback-from-stable-diffusion-to-dall-e.ipynb)
+View the [official docs](https://portkey.ai/docs)
-## integrations
-* [OpenAI](./integrations/openai.ipynb)
-* [Anyscale](./integrations/anyscale.md)
-* [Mistral](./integrations/mistral.md)
+## providers
+* [OpenAI](./providers/openai.ipynb)
+* [Anthropic](./providers/anthropic.md)
+* [Mistral](./providers/mistral.md)
* [Vercel AI](./integrations/vercel-ai.md)
-* [Groq](./integrations/groq.ipynb)
-* [Mistral 8x22b](./integrations/mistral.md)
+* [Groq](./providers/groq.ipynb)
+* [Mistral](./providers/mistral.ipynb)
+* [Deepinfra](./providers/deepinfra.ipynb)
+* [Segmind](./providers/segmind.ipynb)
+* [Nvidia](./providers/nvidia.ipynb)
+
+View the [full list of providers here](https://portkey.ai/docs/welcome/integration-guides).
+
+## integrations
* [Langchain](./integrations/langchain.ipynb)
-* [Deepinfra](./integrations/deepinfra.ipynb)
-* [Segmind](./integrations/segmind.ipynb)
+* [Llama Index](./integrations/llama-index.ipynb)
+* [Instructor](./integrations/Instructor_with_Portkey.ipynb)
+* [Phidata](./integrations/Phidata_with_Portkey.ipynb)
+View the [full list of integrations here](https://portkey.ai/docs/welcome/integration-guides).
-## use-cases
-* [Comparing Top 10 LMSYS Models using Portkey](./use-cases/LMSYS%20Series/comparing-top10-LMSYS-models-with-Portkey.ipynb)
-* [Use OpenAI SDK with Portkey prompt templates](./use-cases/use-openai-sdk-with-portkey-prompt-templates.md)
-* [Run Gateway on prompts from Langchain hub](./use-cases/run-gateway-on-prompts-from-langchain-hub.md)
-* [Smart fallback with model optimized prompts](./use-cases/smart-fallback-with-model-optimized-prompts.md)
-* [Build an article suggestion app with Supabase pgvector, and Portkey](./use-cases/supabase-pgvector-and-portkey.md)
-* [Call Llama 3 on Groq](./use-cases/llama-3-on-groq.ipynb)
+## monitoring-agents
+* [Autogen](./monitoring-agents/Autogen_with_Telemetry.ipynb)
+* [CrewAI](./monitoring-agents/CrewAI_with_Telemetry.ipynb)
+* [Llama Agents](./monitoring-agents/Llama_Agents_with_Telemetry.ipynb)
+* [ControlFlow](./monitoring-agents/ControlFlow_with_Telemetry.ipynb)
## contributing
diff --git a/cookbook/integrations/Tool_Use_with_Portkey.ipynb b/cookbook/integrations/Tool_Use_with_Portkey.ipynb
new file mode 100644
index 00000000..8eeecb0a
--- /dev/null
+++ b/cookbook/integrations/Tool_Use_with_Portkey.ipynb
@@ -0,0 +1,507 @@
+{
+ "nbformat": 4,
+ "nbformat_minor": 0,
+ "metadata": {
+ "colab": {
+ "provenance": []
+ },
+ "kernelspec": {
+ "name": "python3",
+ "display_name": "Python 3"
+ },
+ "language_info": {
+ "name": "python"
+ }
+ },
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "source": [
+ ""
+ ],
+ "metadata": {
+ "id": "zld1izGSx2rl"
+ }
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "[Portkey](https://app.portkey.ai/) is the Control Panel for AI apps. With it's popular AI Gateway and Observability Suite, hundreds of teams ship reliable, cost-efficient, and fast apps.\n",
+ "\n",
+ "With Portkey, you can\n",
+ "\n",
+ "- Connect to 200+ models through a unified API,\n",
+ "- View 40+ metrics & logs for all requests\n",
+ "- Enable semantic cache to reduce latency & costs\n",
+ "- Implement automatic retries & fallbacks for failed requests"
+ ],
+ "metadata": {
+ "id": "VfWXytp_1EVj"
+ }
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "This notebook demonstrates how to use Portkey for function calling with various LLM providers including OpenAI, Anthropic, and Google's Gemini."
+ ],
+ "metadata": {
+ "id": "F1zYiehz1VAl"
+ }
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "[](https://colab.research.google.com/drive/1TeqYiXhlfZvFhoOHWEiyBycfWu0arWwp?usp=sharing)"
+ ],
+ "metadata": {
+ "id": "MosN-n4Wx4ZP"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "!pip install -qU openai portkey-ai"
+ ],
+ "metadata": {
+ "id": "lha5CiI-kvcu"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "tools = [\n",
+ " {\n",
+ " \"type\": \"function\",\n",
+ " \"function\": {\n",
+ " \"name\": \"get_current_weather\",\n",
+ " \"description\": \"Get the current weather in a given location\",\n",
+ " \"parameters\": {\n",
+ " \"type\": \"object\",\n",
+ " \"properties\": {\n",
+ " \"location\": {\n",
+ " \"type\": \"string\",\n",
+ " \"description\": \"The city and state, e.g. San Francisco, CA\",\n",
+ " },\n",
+ " \"unit\": {\"type\": \"string\", \"enum\": [\"celsius\", \"fahrenheit\"]},\n",
+ " },\n",
+ " \"required\": [\"location\"],\n",
+ " },\n",
+ " },\n",
+ " }\n",
+ " ]"
+ ],
+ "metadata": {
+ "id": "_wviaUs5pi8B"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "restaurant_recommendations = [{\n",
+ " \"type\": \"function\",\n",
+ " \"function\": {\n",
+ " \"name\": \"get_restaurant_recommendations\",\n",
+ " \"description\": \"Get restaurant recommendations for a given location and cuisine type\",\n",
+ " \"parameters\": {\n",
+ " \"type\": \"object\",\n",
+ " \"properties\": {\n",
+ " \"location\": {\n",
+ " \"type\": \"string\",\n",
+ " \"description\": \"The city and state, e.g. New York, NY\"\n",
+ " },\n",
+ " \"cuisine\": {\n",
+ " \"type\": \"string\",\n",
+ " \"description\": \"The type of cuisine, e.g. Italian, Chinese, Mexican\"\n",
+ " },\n",
+ " \"price_range\": {\n",
+ " \"type\": \"string\",\n",
+ " \"enum\": [\"$\", \"$$\", \"$$$\", \"$$$$\"],\n",
+ " \"description\": \"The price range of the restaurants\"\n",
+ " }\n",
+ " },\n",
+ " \"required\": [\"location\", \"cuisine\"]\n",
+ " }\n",
+ " }\n",
+ " },\n",
+ " ]"
+ ],
+ "metadata": {
+ "id": "bLeShK0bTj_r"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "latest_ai_news = [{\n",
+ " \"type\": \"function\",\n",
+ " \"function\": {\n",
+ " \"name\": \"get_latest_ai_news\",\n",
+ " \"description\": \"Retrieve the latest news articles about artificial intelligence\",\n",
+ " \"parameters\": {\n",
+ " \"type\": \"object\",\n",
+ " \"properties\": {\n",
+ " \"topic\": {\n",
+ " \"type\": \"string\",\n",
+ " \"description\": \"Specific AI topic (optional, e.g., 'machine learning', 'neural networks', 'robotics')\"\n",
+ " },\n",
+ " \"timeframe\": {\n",
+ " \"type\": \"string\",\n",
+ " \"enum\": [\"today\", \"this_week\", \"this_month\"],\n",
+ " \"description\": \"Time period for the news articles\"\n",
+ " },\n",
+ " \"source_type\": {\n",
+ " \"type\": \"string\",\n",
+ " \"enum\": [\"all\", \"academic\", \"industry\", \"mainstream\"],\n",
+ " \"description\": \"Type of news sources to include\"\n",
+ " },\n",
+ " \"max_results\": {\n",
+ " \"type\": \"integer\",\n",
+ " \"description\": \"Maximum number of news articles to return (1-50)\",\n",
+ " \"minimum\": 1,\n",
+ " \"maximum\": 50\n",
+ " },\n",
+ " \"language\": {\n",
+ " \"type\": \"string\",\n",
+ " \"description\": \"Preferred language for the articles (e.g., 'en' for English, 'es' for Spanish)\"\n",
+ " }\n",
+ " },\n",
+ " \"required\": [\"timeframe\"]\n",
+ " }\n",
+ " }\n",
+ "}]"
+ ],
+ "metadata": {
+ "id": "kG-wPaBvXDH3"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Initialize Portkey client\n",
+ "from portkey_ai import Portkey\n",
+ "from portkey_ai import PORTKEY_GATEWAY_URL, createHeaders\n",
+ "from google.colab import userdata\n",
+ "from tabulate import tabulate\n",
+ "\n",
+ "providers = [\n",
+ " (\"openai\", \"gpt-4o\"),\n",
+ " (\"openai\", \"gpt-4o-mini\"),\n",
+ " (\"anthropic\", \"claude-3-5-sonnet-20240620\"),\n",
+ " (\"anthropic\", \"claude-3-opus-20240229\"),\n",
+ " (\"anthropic\", \"claude-3-haiku-20240307\"),\n",
+ " (\"google\", \"gemini-1.5-flash-latest\"),\n",
+ " (\"google\", \"gemini-1.5-pro\"),\n",
+ " #(\"fireworks-ai\", \"accounts/yi-01-ai/models/yi-large\"),\n",
+ " (\"fireworks-ai\", \"accounts/fireworks/models/firefunction-v2\"),\n",
+ " # (\"fireworks-ai\", \"accounts/fireworks/models/codegemma-2b\"),\n",
+ " # (\"fireworks-ai\", \"accounts/fireworks/models/llama-v2-7b\"),\n",
+ " #(\"groq\", \"llama3-groq-70b-8192-tool-use-preview\"),\n",
+ "\n",
+ " # (\"together-ai\", \"meta-llama/Llama-2-70b-hf\"), #- Error code: 400 - {'error': {'message': 'together-ai error: google/gemma-2-27b-it is not supported for JSON mode/function calling\n",
+ "\n",
+ "\n",
+ " ]\n",
+ "\n",
+ "\n",
+ "portkey = Portkey(api_key=userdata.get('PORTKEY_API_KEY'))\n",
+ "\n",
+ "\n",
+ "def portkey_function_call(messages):\n",
+ " results = []\n",
+ "\n",
+ " for provider, model in providers:\n",
+ " config = {\n",
+ " \"provider\": provider.lower(),\n",
+ " \"api_key\": userdata.get(f'{provider.upper()}_API_KEY')\n",
+ " }\n",
+ "\n",
+ "\n",
+ " # Make the API call\n",
+ " response = portkey.with_options(config=config).chat.completions.create(\n",
+ " model=model,\n",
+ " messages=messages,\n",
+ " tools=latest_ai_news,\n",
+ " tool_choice=\"auto\",\n",
+ " max_tokens=512,\n",
+ " )\n",
+ "\n",
+ " response_message = response.choices[0].message\n",
+ " tool_call_response = str(response_message.tool_calls) if response_message.tool_calls else \"No tool calls\"\n",
+ "\n",
+ " results.append([model, provider, tool_call_response])\n",
+ "\n",
+ " # Print results using tabulate\n",
+ " print(tabulate(results, headers=[\"Model\", \"Provider\", \"Tool Call Response\"], tablefmt=\"grid\"))"
+ ],
+ "metadata": {
+ "id": "PRj-hPyO6jVt"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "#Testing of Get Latest AI News Tool"
+ ],
+ "metadata": {
+ "id": "2m_RKBxtYQFL"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Test message\n",
+ "messages = [{\"role\": \"user\", \"content\": \"give me latest ai news of this week\"}]\n",
+ "\n",
+ "# Run function calling for all providers\n",
+ "portkey_function_call(messages)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "ySeRNzxRXQ7a",
+ "outputId": "a353b7f3-babe-4d83-a459-09e4e9c932e2"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| Model | Provider | Tool Call Response |\n",
+ "+===========================================+==============+==========================================================================================================================================================================================================================================+\n",
+ "| gpt-4o | openai | [ChatCompletionMessageToolCall(id='call_9zk75e22eO5NVzq42vAFmawo', function=FunctionCall(arguments='{\"timeframe\":\"this_week\",\"max_results\":5}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gpt-4o-mini | openai | [ChatCompletionMessageToolCall(id='call_Geq0M1hYFgJmNmp6XbonxcmP', function=FunctionCall(arguments='{\"timeframe\":\"this_week\",\"max_results\":5}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-5-sonnet-20240620 | anthropic | [ChatCompletionMessageToolCall(id='toolu_01LmVxtbDtSaQXq27pX9zeCH', function=FunctionCall(arguments='{\"timeframe\":\"this_week\",\"max_results\":10,\"language\":\"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-opus-20240229 | anthropic | [ChatCompletionMessageToolCall(id='toolu_016DEvWVbfCYDNAvntbhCLE7', function=FunctionCall(arguments='{\"timeframe\":\"this_week\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-haiku-20240307 | anthropic | [ChatCompletionMessageToolCall(id='toolu_01UzArrKvobYz6Bo2CkG9AZW', function=FunctionCall(arguments='{\"timeframe\":\"this_week\",\"max_results\":10,\"source_type\":\"all\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-flash-latest | google | [ChatCompletionMessageToolCall(id='portkey-69bbbb43-9bd8-442f-a9be-aa926b46458c', function=FunctionCall(arguments='{\"timeframe\":\"this_week\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-pro | google | [ChatCompletionMessageToolCall(id='portkey-4424fd91-e7c1-4eed-ac7a-8be4b52e0325', function=FunctionCall(arguments='{\"timeframe\":\"this_week\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| accounts/fireworks/models/firefunction-v2 | fireworks-ai | [ChatCompletionMessageToolCall(id='call_927xwUgfmPh5vRKqsUoEMVeK', function=FunctionCall(arguments='{\"timeframe\": \"this_week\", \"max_results\": 5, \"source_type\": \"all\", \"language\": \"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "#Testing of restaurant_recommendations Tool"
+ ],
+ "metadata": {
+ "id": "-3xJxlWxYZUL"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Test message\n",
+ "messages = [{\"role\": \"user\", \"content\": \"Suggest latest and best restaurant of Aurangabad?\"}]\n",
+ "\n",
+ "# Run function calling for all providers\n",
+ "portkey_function_call(messages)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "Rjmzq6yXUKGc",
+ "outputId": "aa9b0fce-cc6e-4588-e9df-abce595d8abc"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| Model | Provider | Tool Call Response |\n",
+ "+===========================================+==============+=================================================================================================================================================================================================================================================================+\n",
+ "| gpt-4o | openai | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gpt-4o-mini | openai | [ChatCompletionMessageToolCall(id='call_LuP8f1JvDdmKO5hOI5xoqsNs', function=FunctionCall(arguments='{\"topic\":\"restaurant\",\"timeframe\":\"this_month\",\"source_type\":\"mainstream\",\"max_results\":5,\"language\":\"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-5-sonnet-20240620 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-opus-20240229 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-haiku-20240307 | anthropic | [ChatCompletionMessageToolCall(id='toolu_018RjfwRm3xTPzWffnCKid35', function=FunctionCall(arguments='{\"language\":\"en\",\"max_results\":5,\"source_type\":\"all\",\"timeframe\":\"today\",\"topic\":\"aurangabad restaurants\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-flash-latest | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-pro | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| accounts/fireworks/models/firefunction-v2 | fireworks-ai | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "#Testing of Current Weather Featching Tool"
+ ],
+ "metadata": {
+ "id": "60wn4j8DYjU_"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Test message\n",
+ "messages = [{\"role\": \"user\", \"content\": \"How's the weather like in San Francisco?\"}]\n",
+ "\n",
+ "# Run function calling for all providers\n",
+ "portkey_function_call(messages)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "kAYKGo137cwP",
+ "outputId": "81fd7e38-2c0f-49d5-e1a7-8d2d45b38fc7"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| Model | Provider | Tool Call Response |\n",
+ "+===========================================+==============+===============================================================================================================================================================================================================================================================================================+\n",
+ "| gpt-4o | openai | [ChatCompletionMessageToolCall(id='call_39osRa58C5UZfcZl2IwTeO8b', function=FunctionCall(arguments='{\\n \"topic\": \"weather forecasting\",\\n \"timeframe\": \"today\",\\n \"source_type\": \"mainstream\",\\n \"max_results\": 1,\\n \"language\": \"en\"\\n}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gpt-4o-mini | openai | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-5-sonnet-20240620 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-opus-20240229 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-haiku-20240307 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-flash-latest | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-pro | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| accounts/fireworks/models/firefunction-v2 | fireworks-ai | No tool calls |\n",
+ "+-------------------------------------------+--------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Test message\n",
+ "messages = [{\"role\": \"user\", \"content\": \"What's the weather like in San Francisco, Tokyo, and Paris?\"}]\n",
+ "\n",
+ "# Run function calling for all providers\n",
+ "portkey_function_call(messages)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "ZrDT-TFv62HW",
+ "outputId": "f149c47e-d4c6-4e2b-db97-c8aab2760dcc"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| Model | Provider | Tool Call Response |\n",
+ "+===========================================+==============+========================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================+\n",
+ "| gpt-4o | openai | [ChatCompletionMessageToolCall(id='call_jQM4yp3iVovjYBmBrdN74loM', function=FunctionCall(arguments='{\"location\": \"San Francisco\"}', name='get_current_weather'), type='function'), ChatCompletionMessageToolCall(id='call_w8ZYduT9EsP3NqJljv3T7Eyy', function=FunctionCall(arguments='{\"location\": \"Tokyo\"}', name='get_current_weather'), type='function'), ChatCompletionMessageToolCall(id='call_96F45BQiUOj0q4Gy8iKRGsZ7', function=FunctionCall(arguments='{\"location\": \"Paris\"}', name='get_current_weather'), type='function')] |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gpt-4o-mini | openai | [ChatCompletionMessageToolCall(id='call_FSiM3gfLKr108PsegkggycFw', function=FunctionCall(arguments='{\"topic\": \"weather\", \"timeframe\": \"today\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function'), ChatCompletionMessageToolCall(id='call_8Ovrkm2tDdLaauoUaYVqmNYv', function=FunctionCall(arguments='{\"topic\": \"weather\", \"timeframe\": \"today\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function'), ChatCompletionMessageToolCall(id='call_yKfJ6FFLY12JwiWzXen0LnLP', function=FunctionCall(arguments='{\"topic\": \"weather\", \"timeframe\": \"today\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-5-sonnet-20240620 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-opus-20240229 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-haiku-20240307 | anthropic | [ChatCompletionMessageToolCall(id='toolu_01EbJeB9eYvKG7ANQKkhS9bP', function=FunctionCall(arguments='{\"locations\":\"[\\\\\"San Francisco\\\\\", \\\\\"Tokyo\\\\\", \\\\\"Paris\\\\\"]\"}', name='get_weather'), type='function')] |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-flash-latest | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-pro | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| accounts/fireworks/models/firefunction-v2 | fireworks-ai | [ChatCompletionMessageToolCall(id='call_B2EiSi4DGpkwTjZhRrBKYDXJ', function=FunctionCall(arguments='{\"topic\": \"weather\", \"timeframe\": \"today\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Test message\n",
+ "messages = [{\"role\": \"user\", \"content\": \"What's the weather like in Mumbai ,Delhi and Banglore?\"}]\n",
+ "\n",
+ "# Run function calling for all providers\n",
+ "portkey_function_call(messages)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "jMv-2PMrVST6",
+ "outputId": "7a58e247-9a55-4d4e-e72e-98e8d1c469a6"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| Model | Provider | Tool Call Response |\n",
+ "+===========================================+==============+=========================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================================+\n",
+ "| gpt-4o | openai | [ChatCompletionMessageToolCall(id='call_RDwPeeHLHI9c5MBOxXnRkUlQ', function=FunctionCall(arguments='{\"location\": \"Mumbai\"}', name='get_current_weather'), type='function'), ChatCompletionMessageToolCall(id='call_nP0eZZISJeyzXI1utBG6alLj', function=FunctionCall(arguments='{\"location\": \"Delhi\"}', name='get_current_weather'), type='function'), ChatCompletionMessageToolCall(id='call_pOnq4tp9sSSPAaiarGsvykNd', function=FunctionCall(arguments='{\"location\": \"Bangalore\"}', name='get_current_weather'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gpt-4o-mini | openai | [ChatCompletionMessageToolCall(id='call_vS2RKIJ6Acm3l1lz4HhwOXss', function=FunctionCall(arguments='{\"topic\": \"Mumbai\", \"timeframe\": \"today\", \"source_type\": \"all\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function'), ChatCompletionMessageToolCall(id='call_i9gg5uKxQaq9XdBaPCcJnnAX', function=FunctionCall(arguments='{\"topic\": \"Delhi\", \"timeframe\": \"today\", \"source_type\": \"all\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function'), ChatCompletionMessageToolCall(id='call_cRQyEpj3OHoxOuqWnVHDF3Ey', function=FunctionCall(arguments='{\"topic\": \"Bangalore\", \"timeframe\": \"today\", \"source_type\": \"all\", \"max_results\": 1, \"language\": \"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-5-sonnet-20240620 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-opus-20240229 | anthropic | No tool calls |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| claude-3-haiku-20240307 | anthropic | [ChatCompletionMessageToolCall(id='toolu_018NHHkq5s62Rb8G4bmeZYo1', function=FunctionCall(arguments='{\"language\":\"en\",\"max_results\":3,\"source_type\":\"all\",\"timeframe\":\"today\",\"topic\":\"weather forecast mumbai delhi bangalore\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-flash-latest | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| gemini-1.5-pro | google | No tool calls |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n",
+ "| accounts/fireworks/models/firefunction-v2 | fireworks-ai | [ChatCompletionMessageToolCall(id='call_IqRs42JAXIxof45YAcAMWLIB', function=FunctionCall(arguments='{\"topic\": \"none\", \"timeframe\": \"today\", \"source_type\": \"all\", \"max_results\": 3, \"language\": \"en\"}', name='get_latest_ai_news'), type='function')] |\n",
+ "+-------------------------------------------+--------------+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+\n"
+ ]
+ }
+ ]
+ }
+ ]
+}
\ No newline at end of file
diff --git a/cookbook/integrations/Autogen_with_Telemetry.ipynb b/cookbook/monitoring-agents/Autogen_with_Telemetry.ipynb
similarity index 100%
rename from cookbook/integrations/Autogen_with_Telemetry.ipynb
rename to cookbook/monitoring-agents/Autogen_with_Telemetry.ipynb
diff --git a/cookbook/integrations/ControlFlow_with_Telemetry.ipynb b/cookbook/monitoring-agents/ControlFlow_with_Telemetry.ipynb
similarity index 100%
rename from cookbook/integrations/ControlFlow_with_Telemetry.ipynb
rename to cookbook/monitoring-agents/ControlFlow_with_Telemetry.ipynb
diff --git a/cookbook/integrations/CrewAI_with_Telemetry.ipynb b/cookbook/monitoring-agents/CrewAI_with_Telemetry.ipynb
similarity index 100%
rename from cookbook/integrations/CrewAI_with_Telemetry.ipynb
rename to cookbook/monitoring-agents/CrewAI_with_Telemetry.ipynb
diff --git a/cookbook/integrations/Llama_Agents_with_Telemetry.ipynb b/cookbook/monitoring-agents/Llama_Agents_with_Telemetry.ipynb
similarity index 100%
rename from cookbook/integrations/Llama_Agents_with_Telemetry.ipynb
rename to cookbook/monitoring-agents/Llama_Agents_with_Telemetry.ipynb
diff --git a/cookbook/providers/nvidia.ipynb b/cookbook/providers/nvidia.ipynb
new file mode 100644
index 00000000..f9148c07
--- /dev/null
+++ b/cookbook/providers/nvidia.ipynb
@@ -0,0 +1,198 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "bFj2ZAU6BtiR"
+ },
+ "source": [
+ ""
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "5Mfpp2FLAYEA"
+ },
+ "source": [
+ "[](https://colab.research.google.com/drive/1pykKqel2h6ltbVok4nKHhWljcs9_PKDD?usp=sharing)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "ekkvItFsWyQL"
+ },
+ "source": [
+ "# Portkey + Nvidia NIM"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "x8RrtT5yYEev"
+ },
+ "source": [
+ "[Portkey](https://app.portkey.ai/) is the Control Panel for AI apps. With it's popular AI Gateway and Observability Suite, hundreds of teams ship reliable, cost-efficient, and fast apps.\n",
+ "\n",
+ "With Portkey, you can\n",
+ "\n",
+ " - Connect to 150+ models through a unified API,\n",
+ " - View 40+ metrics & logs for all requests,\n",
+ " - Enable semantic cache to reduce latency & costs,\n",
+ " - Implement automatic retries & fallbacks for failed requests,\n",
+ " - Add custom tags to requests for better tracking and analysis and more.\n"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "n1kPDnimZIXb"
+ },
+ "source": [
+ "## Quickstart\n",
+ "\n",
+ "Since Portkey is fully compatible with the OpenAI signature, you can connect to the Portkey AI Gateway through OpenAI Client.\n",
+ "\n",
+ "- Set the `base_url` as `PORTKEY_GATEWAY_URL`\n",
+ "- Add `default_headers` to consume the headers needed by Portkey using the `createHeaders` helper method."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "rQSrrUBwkmOP"
+ },
+ "source": [
+ "You will need Portkey and Nvidia API keys to run this notebook.\n",
+ "\n",
+ "- Sign up for Portkey and generate your API key [here](https://app.portkey.ai/).\n",
+ "- Get your Nvidia NIM key [here](https://nvidia.com/dash/api_keys)\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "fffx7Tc2ghTR",
+ "outputId": "23ed5e5b-f4d6-424a-df17-547913f2ef81"
+ },
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m405.9/405.9 kB\u001b[0m \u001b[31m5.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m328.5/328.5 kB\u001b[0m \u001b[31m8.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m75.6/75.6 kB\u001b[0m \u001b[31m2.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m12.7/12.7 MB\u001b[0m \u001b[31m28.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m77.9/77.9 kB\u001b[0m \u001b[31m3.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m58.3/58.3 kB\u001b[0m \u001b[31m3.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[?25h"
+ ]
+ }
+ ],
+ "source": [
+ "!pip install -qU portkey-ai openai"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "ptP4L78HlBUL"
+ },
+ "source": [
+ "## With OpenAI Client"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "id": "VZ7zSwDAhc18"
+ },
+ "outputs": [],
+ "source": [
+ "from openai import OpenAI\n",
+ "from portkey_ai import PORTKEY_GATEWAY_URL, createHeaders\n",
+ "from google.colab import userdata\n",
+ "\n",
+ "client = OpenAI(\n",
+ " api_key= userdata.get('NVIDIA_API_KEY'), ## replace it your Mistral API key\n",
+ " base_url=PORTKEY_GATEWAY_URL,\n",
+ " default_headers=createHeaders(\n",
+ " provider = \"openai\",\n",
+ " custom_host = \"https://integrate.api.nvidia.com/v1\",\n",
+ " api_key= userdata.get('PORTKEY_API_KEY'), ## replace it your Portkey API key\n",
+ " )\n",
+ ")\n",
+ "\n",
+ "completion = client.chat.completions.create(\n",
+ " model=\"google/gemma-2-27b-it\",\n",
+ " messages=[{\"role\":\"user\",\"content\":\"Write a limerick about the wonders of GPU computing.\"}],\n",
+ ")\n",
+ "\n",
+ "print(completion.choices[0].message.content)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "wMhOOkaLqkSp"
+ },
+ "source": [
+ "## Observability with Portkey"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "LzECvXQaqr6Z"
+ },
+ "source": [
+ "By routing requests through Portkey you can track a number of metrics like - tokens used, latency, cost, etc.\n",
+ "\n",
+ "Here's a screenshot of the dashboard you get with Portkey!"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {
+ "id": "W80YWspqr7s0"
+ },
+ "source": [
+ "\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "id": "yqrIH5mY01BI"
+ },
+ "outputs": [],
+ "source": []
+ }
+ ],
+ "metadata": {
+ "colab": {
+ "provenance": []
+ },
+ "kernelspec": {
+ "display_name": "Python 3",
+ "name": "python3"
+ },
+ "language_info": {
+ "name": "python"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}
diff --git a/cookbook/use-cases/Creating_Artifacts_with_GPT_4o_.ipynb b/cookbook/use-cases/Creating_Artifacts_with_GPT_4o_.ipynb
new file mode 100644
index 00000000..ddb1d38e
--- /dev/null
+++ b/cookbook/use-cases/Creating_Artifacts_with_GPT_4o_.ipynb
@@ -0,0 +1,360 @@
+{
+ "nbformat": 4,
+ "nbformat_minor": 0,
+ "metadata": {
+ "colab": {
+ "provenance": []
+ },
+ "kernelspec": {
+ "name": "python3",
+ "display_name": "Python 3"
+ },
+ "language_info": {
+ "name": "python"
+ }
+ },
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "source": [
+ "# Artifacts with GPT-4o using Portkey"
+ ],
+ "metadata": {
+ "id": "Uxsc0IoBEbnS"
+ }
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "[](https://colab.research.google.com/drive/1-CyvKS3JlhkN-rV88AdZSdg5a5V7stzs?usp=sharing)"
+ ],
+ "metadata": {
+ "id": "nJ6TpzZ_Dcdz"
+ }
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ ""
+ ],
+ "metadata": {
+ "id": "fmhDqgPglA9T"
+ }
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "MmSOyeovkbZb",
+ "outputId": "c60179de-473f-4ff5-d597-ee54f1c70b64"
+ },
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m405.9/405.9 kB\u001b[0m \u001b[31m2.4 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m328.5/328.5 kB\u001b[0m \u001b[31m9.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m75.6/75.6 kB\u001b[0m \u001b[31m3.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m12.7/12.7 MB\u001b[0m \u001b[31m13.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m102.8/102.8 kB\u001b[0m \u001b[31m2.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m137.6/137.6 kB\u001b[0m \u001b[31m6.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m130.2/130.2 kB\u001b[0m \u001b[31m2.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m77.9/77.9 kB\u001b[0m \u001b[31m5.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m58.3/58.3 kB\u001b[0m \u001b[31m4.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
+ "\u001b[?25h"
+ ]
+ }
+ ],
+ "source": [
+ "!pip install -qU portkey-ai openai e2b_code_interpreter\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Get your API keys or save them to .env file.\n",
+ "import os\n",
+ "from google.colab import userdata\n",
+ "\n",
+ "# TODO: Get your E2B API key from https://e2b.dev/docs\n",
+ "E2B_API_KEY = userdata.get('E2B_API_KEY')\n",
+ "\n",
+ "\n",
+ "\n",
+ "\n",
+ "\n",
+ "SYSTEM_PROMPT = \"\"\"\n",
+ "## your job & context\n",
+ "you are a python data scientist. you are given tasks to complete and you run python code to solve them.\n",
+ "You DO NOT MAKE SYNTAX MISTAKES OR FORGET ANY IMPORTS\n",
+ "- the python code runs in jupyter notebook.\n",
+ "- every time you call `execute_python` tool, the python code is executed in a separate cell. it's okay to multiple calls to `execute_python`.\n",
+ "- display visualizations using matplotlib or any other visualization library directly in the notebook. don't worry about saving the visualizations to a file.\n",
+ "- you have access to the internet and can make api requests.\n",
+ "- you also have access to the filesystem and can read/write files.\n",
+ "- you can install any pip package (if it exists) if you need to but the usual packages for data analysis are already preinstalled.\n",
+ "- you can run any python code you want, everything is running in a secure sandbox environment.\n",
+ "\"\"\"\n",
+ "\n",
+ "tools = [\n",
+ " {\n",
+ " \"type\": \"function\",\n",
+ " \"function\": {\n",
+ " \"name\": \"execute_python\",\n",
+ " \"description\": \"Execute python code in a Jupyter notebook cell and returns any result, stdout, stderr, display_data, and error.\",\n",
+ " \"parameters\": {\n",
+ " \"type\": \"object\",\n",
+ " \"properties\": {\n",
+ " \"code\": {\n",
+ " \"type\": \"string\",\n",
+ " \"description\": \"The python code to execute in a single cell.\"\n",
+ " }\n",
+ " },\n",
+ " \"required\": [\"code\"]\n",
+ " }\n",
+ " },\n",
+ " }\n",
+ "]"
+ ],
+ "metadata": {
+ "id": "iWMO7iATmYTn"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "def code_interpret(e2b_code_interpreter, code):\n",
+ " print(\"Running code interpreter...\")\n",
+ " exec = e2b_code_interpreter.notebook.exec_cell(\n",
+ " code,\n",
+ " on_stderr=lambda stderr: print(\"[Code Interpreter]\", stderr),\n",
+ " on_stdout=lambda stdout: print(\"[Code Interpreter]\", stdout),\n",
+ " # You can also stream code execution results\n",
+ " # on_result=...\n",
+ " )\n",
+ "\n",
+ " if exec.error:\n",
+ " print(\"[Code Interpreter ERROR]\", exec.error)\n",
+ " else:\n",
+ " return exec.results"
+ ],
+ "metadata": {
+ "id": "VcObarUhm9HN"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "from openai import OpenAI\n",
+ "from portkey_ai import PORTKEY_GATEWAY_URL, createHeaders\n",
+ "from google.colab import userdata\n",
+ "import json\n",
+ "\n",
+ "gateway = OpenAI(\n",
+ " api_key=userdata.get(\"OPENAI_API_KEY\"),\n",
+ " base_url=PORTKEY_GATEWAY_URL, # Or http://localhost:8787/v1 when running locally\n",
+ " default_headers=createHeaders(\n",
+ " provider=\"openai\",\n",
+ " api_key=userdata.get(\"PORTKEY_API_KEY\") # Grab from https://app.portkey.ai # Not needed when running locally\n",
+ " )\n",
+ ")\n",
+ "\n",
+ "def chat(e2b_code_interpreter, user_message, base64_image = None, ):\n",
+ " print(f\"\\n{'='*50}\\nUser Message: {user_message}\\n{'='*50}\")\n",
+ "\n",
+ " messages = [\n",
+ " {\n",
+ " \"role\": \"system\",\n",
+ " \"content\": SYSTEM_PROMPT,\n",
+ " },\n",
+ " ]\n",
+ "\n",
+ " if base64_image is not None:\n",
+ " messages.append(\n",
+ " {\n",
+ " \"role\": \"user\",\n",
+ " \"content\": [\n",
+ " {\n",
+ " \"type\": \"text\",\n",
+ " \"text\": user_message,\n",
+ " },\n",
+ " {\n",
+ " \"type\": \"image_url\",\n",
+ " \"image_url\": {\n",
+ " \"url\": f\"data:image/jpeg;base64,{base64_image}\"\n",
+ " }\n",
+ " }\n",
+ " ]\n",
+ " }\n",
+ " )\n",
+ " else:\n",
+ " messages.append(\n",
+ " {\"role\": \"user\", \"content\": user_message},\n",
+ " )\n",
+ "\n",
+ " response = gateway.chat.completions.create(\n",
+ " model=\"gpt-4o\",\n",
+ " messages=messages,\n",
+ " tools=tools,\n",
+ " tool_choice=\"auto\"\n",
+ " )\n",
+ " for choice in response.choices:\n",
+ " if choice.message.tool_calls and len(choice.message.tool_calls) > 0:\n",
+ " for tool_call in choice.message.tool_calls:\n",
+ " if tool_call.function.name == \"execute_python\":\n",
+ " if \"code\" in tool_call.function.arguments:\n",
+ " code = tool_call.function.arguments[\"code\"]\n",
+ " else:\n",
+ " code = tool_call.function.arguments\n",
+ " print(\"CODE TO RUN\")\n",
+ " print(code)\n",
+ " code_interpreter_results = code_interpret(e2b_code_interpreter, code)\n",
+ " return code_interpreter_results\n",
+ " else:\n",
+ " print(\"Answer:\", choice.message.content)"
+ ],
+ "metadata": {
+ "id": "zYBSwUF7oVvX"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "# Example 1 :- distribution of salaries in the tech industry"
+ ],
+ "metadata": {
+ "id": "DtyYCr59C96Z"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "from e2b_code_interpreter import CodeInterpreter\n",
+ "code_interpreter = CodeInterpreter(api_key=E2B_API_KEY)\n",
+ "\n",
+ "# 1. Ask GPT-4o to generate chart\n",
+ "code_interpreter_results = chat(\n",
+ " code_interpreter,\n",
+ " \"Plot a chart visualizing the height distribution of men based on the data you know\",\n",
+ ")\n",
+ "print(code_interpreter_results)\n",
+ "plot1 = code_interpreter_results[0]\n",
+ "\n",
+ "\n",
+ "print(plot1.raw)\n",
+ "plot1"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/",
+ "height": 1000
+ },
+ "id": "ShQx9NdHoh_B",
+ "outputId": "c1b7e55f-10dc-4d38-86e9-7e508842ae4f"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "\n",
+ "==================================================\n",
+ "User Message: Plot a chart visualizing the height distribution of men based on the data you know\n",
+ "==================================================\n",
+ "CODE TO RUN\n",
+ "import numpy as np\n",
+ "import matplotlib.pyplot as plt\n",
+ "\n",
+ "# Set mean and standard deviation\n",
+ "mean_height = 69 # average height in inches\n",
+ "std_dev_height = 3 # standard deviation in inches\n",
+ "\n",
+ "# Generate synthetic dataset\n",
+ "np.random.seed(0)\n",
+ "heights = np.random.normal(mean_height, std_dev_height, 1000) # generate 1000 data points\n",
+ "\n",
+ "# Plotting the height distribution\n",
+ "plt.figure(figsize=(10, 6))\n",
+ "plt.hist(heights, bins=30, edgecolor='black', alpha=0.7)\n",
+ "plt.title(\"Height Distribution of Men\")\n",
+ "plt.xlabel(\"Height (inches)\")\n",
+ "plt.ylabel(\"Frequency\")\n",
+ "plt.grid(axis='y', linestyle='--', alpha=0.7)\n",
+ "plt.show()\n",
+ "Running code interpreter...\n",
+ "[Result()]\n",
+ "{'text/plain': '', 'image/png': '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'}\n"
+ ]
+ },
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "Result()"
+ ],
+ "image/png": "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"
+ },
+ "metadata": {},
+ "execution_count": 12
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "#Generate a histogram and a box plot to show the distribution of salaries"
+ ],
+ "metadata": {
+ "id": "XpM3dJL5DQ0D"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# 2. Feed the image back the chart to GPT-4o and ask question about the image\n",
+ "image = plot1.png\n",
+ "code_interpreter_results = chat(\n",
+ " code_interpreter,\n",
+ " \"Based on what you see, what's name of this distribution? Show me the distribution function\",\n",
+ " image,\n",
+ ")\n",
+ "\n",
+ "code_interpreter.close()\n",
+ "\n",
+ "print(code_interpreter_results)\n",
+ "plot2 = code_interpreter_results[0]\n",
+ "print(plot2.raw)\n",
+ "plot2"
+ ],
+ "metadata": {
+ "id": "P0-Z55q-szyA"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [],
+ "metadata": {
+ "id": "DaJQLK0zXLQn"
+ },
+ "execution_count": null,
+ "outputs": []
+ }
+ ]
+}
\ No newline at end of file
diff --git a/cookbook/use-cases/LLM_as_Judge.ipynb b/cookbook/use-cases/LLM_as_Judge.ipynb
deleted file mode 100644
index 1f67b8de..00000000
--- a/cookbook/use-cases/LLM_as_Judge.ipynb
+++ /dev/null
@@ -1,300 +0,0 @@
-{
- "nbformat": 4,
- "nbformat_minor": 0,
- "metadata": {
- "colab": {
- "provenance": []
- },
- "kernelspec": {
- "name": "python3",
- "display_name": "Python 3"
- },
- "language_info": {
- "name": "python"
- }
- },
- "cells": [
- {
- "cell_type": "markdown",
- "source": [
- ""
- ],
- "metadata": {
- "id": "75EKSrYn8hPe"
- }
- },
- {
- "cell_type": "markdown",
- "source": [
- "[](https://colab.research.google.com/drive/1K_DEXMlwdf6JPojDdquv6CbhqoME5mwd?usp=sharing)"
- ],
- "metadata": {
- "id": "Yvkxqeuo8ofe"
- }
- },
- {
- "cell_type": "markdown",
- "source": [
- "## LLM as Judge with Portkey Feedback API\n"
- ],
- "metadata": {
- "id": "z4yZQ3Yn1ZvK"
- }
- },
- {
- "cell_type": "markdown",
- "source": [
- "Ensuring the quality and correctness of the LLM's output is crucial for AI applications. One approach to address this challenge is to employ a second LLM as a judge to evaluate the output of the base LLM.\n",
- "\n"
- ],
- "metadata": {
- "id": "AqC-gpM9t4FN"
- }
- },
- {
- "cell_type": "code",
- "execution_count": null,
- "metadata": {
- "colab": {
- "base_uri": "https://localhost:8080/"
- },
- "id": "FnDNQmi_rvSi",
- "outputId": "fbef9011-d667-4f80-bd78-ace0b649ea28"
- },
- "outputs": [
- {
- "output_type": "stream",
- "name": "stdout",
- "text": [
- "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m86.3/86.3 kB\u001b[0m \u001b[31m1.5 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
- "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m75.6/75.6 kB\u001b[0m \u001b[31m5.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
- "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m12.7/12.7 MB\u001b[0m \u001b[31m52.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
- "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m320.7/320.7 kB\u001b[0m \u001b[31m21.9 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
- "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m77.9/77.9 kB\u001b[0m \u001b[31m6.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
- "\u001b[2K \u001b[90mββββββββββββββββββββββββββββββββββββββββ\u001b[0m \u001b[32m58.3/58.3 kB\u001b[0m \u001b[31m3.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
- "\u001b[?25h"
- ]
- }
- ],
- "source": [
- "!pip install -qU portkey-ai"
- ]
- },
- {
- "cell_type": "markdown",
- "source": [
- "### Quickstart\n",
- "\n",
- "Feedbacks in Portkey provide a simple way to get weighted feedback from customers on any request you served, at any stage in your app.\n",
- "\n",
- "You can capture this feedback on a generation or conversation level and analyze it based on custom tags by adding meta data to the relevant request."
- ],
- "metadata": {
- "id": "_pLsCS7r4S52"
- }
- },
- {
- "cell_type": "code",
- "source": [
- "def send_feedback():\n",
- " portkey = Portkey(\n",
- " api_key=userdata.get('PORTKEY_API_KEY'))\n",
- "\n",
- " portkey.feedback.create(\n",
- " trace_id= 'f4e93652-ac78-402e-b281-a4fb3b94b3f5', # trace ID of the request\n",
- " value= 0, # For thumbs down\n",
- " metadata = {\"review\" : \"Response should be less verbose\"}\n",
- " )\n",
- "\n",
- "send_feedback()"
- ],
- "metadata": {
- "id": "Lu1tqBgR4kqE"
- },
- "execution_count": null,
- "outputs": []
- },
- {
- "cell_type": "markdown",
- "source": [
- "### Using LLM as judge\n",
- "\n",
- "`LLM_as_judge` function using Portkey's Feedback API, which demonstrates a novel approach to improving the quality of text generated by LLMs. By employing a second LLM as a judge, we aim to evaluate and ensure the correctness of the generated text."
- ],
- "metadata": {
- "id": "FUlI02uq5GZ7"
- }
- },
- {
- "cell_type": "code",
- "source": [
- "from portkey_ai import Portkey\n",
- "from google.colab import userdata\n",
- "\n",
- "client = Portkey(\n",
- " api_key=userdata.get('PORTKEY_API_KEY')\n",
- ")\n",
- "\n",
- "llama3 = Portkey(api_key=userdata.get('PORTKEY_API_KEY'),\n",
- " virtual_key=\"groq-431005\",\n",
- " model = \"llama3-70b-8192\"\n",
- " )\n",
- "\n",
- "gemini_1_5_flash = Portkey(api_key=userdata.get('PORTKEY_API_KEY'),\n",
- " virtual_key=\"google-bb26f3\",\n",
- " model = \"gemini-1.5-flash\"\n",
- " )\n",
- "\n",
- "gpt_4o = Portkey(api_key=userdata.get('PORTKEY_API_KEY'),\n",
- " virtual_key=\"gpt3-8070a6\",\n",
- " model = \"gpt-4o\"\n",
- " )"
- ],
- "metadata": {
- "id": "gUTubcDdr03A"
- },
- "execution_count": null,
- "outputs": []
- },
- {
- "cell_type": "code",
- "source": [
- "##\n",
- "def LLM_as_judge(base_llm, judge_llm, prompt):\n",
- " pcompletion = base_llm.chat.completions.create(\n",
- " messages=[{\"role\": \"user\", \"content\": prompt}],\n",
- " )\n",
- "\n",
- " completion = pcompletion.choices[0].message.content\n",
- " req_trace = pcompletion.get_headers()['trace-id']\n",
- "\n",
- " print(completion)\n",
- " print(\"===================\")\n",
- " # check_response = f\"Check if the response is in correct JSON format or not {completion}. Give response as yes or no.\"\n",
- "\n",
- " check_response = \"\"\"Evaluate the following response:\n",
- " {}\n",
- " Score each metric from 1 (poor) to 5 (excellent):\n",
- " 1. Relevance and Completeness\n",
- " 2. Logical Coherence\n",
- " 3. Factual Accuracy and Consistency\n",
- " 4. Clarity and Comprehensibility\n",
- "\n",
- " Provide an overall score and a brief summary.\"\"\".format(completion)\n",
- "\n",
- "\n",
- " pfeedback = judge_llm.chat.completions.create(\n",
- " messages=[{\"role\": \"user\", \"content\": check_response}],\n",
- " )\n",
- "\n",
- " feedback = pfeedback.choices[0].message.content\n",
- "\n",
- " print(feedback)\n",
- " print(\"===================\")\n",
- "\n",
- " portkey = Portkey(\n",
- " api_key=userdata.get('PORTKEY_API_KEY')\n",
- " )\n",
- "\n",
- " portkey.feedback.create(\n",
- " trace_id= req_trace,\n",
- " value= 1,\n",
- " metadata = {\"review\" : feedback}\n",
- " )"
- ],
- "metadata": {
- "id": "SjSoWL-MsXYE"
- },
- "execution_count": null,
- "outputs": []
- },
- {
- "cell_type": "code",
- "source": [
- "LLM_as_judge(base_llm = llama3,\n",
- " judge_llm = gpt_4o,\n",
- " prompt= \"If 20 shirts take 5 hours to dry, how much time will 100 shirts take to dry?\"\n",
- " )"
- ],
- "metadata": {
- "colab": {
- "base_uri": "https://localhost:8080/"
- },
- "id": "BvDEnCZnv5AY",
- "outputId": "0c43930a-9cf3-47a0-d85f-a70de4c7ee90"
- },
- "execution_count": null,
- "outputs": [
- {
- "output_type": "stream",
- "name": "stdout",
- "text": [
- "Given that the drying rate seems to be consistent, we can determine the time required for 100 shirts based on the given information.\n",
- "\n",
- "First, let's find out how long it takes for one shirt to dry:\n",
- "1 shirt = Total time / Number of shirts\n",
- "1 shirt = 5 hours / 20 shirts\n",
- "1 shirt = 0.25 hours\n",
- "\n",
- "Now that we know it takes 0.25 hours (or 15 minutes) to dry one shirt, we can calculate the time required for 100 shirts:\n",
- "\n",
- "Time for 100 shirts = Time per shirt * Number of shirts\n",
- "Time for 100 shirts = 0.25 hours * 100 shirts\n",
- "Time for 100 shirts = 25 hours\n",
- "\n",
- "So, it will take 25 hours for 100 shirts to dry if the drying rate remains constant.\n",
- "===================\n",
- "1. Relevance and Completeness: 5 - The response addresses the specific question of determining the time required for drying 100 shirts based on the given information.\n",
- "2. Logical Coherence: 5 - The response follows a logical step-by-step process to calculate the time required for drying 100 shirts.\n",
- "3. Factual Accuracy and Consistency: 5 - The calculations for determining the time per shirt and the total time for 100 shirts are accurate and consistent.\n",
- "4. Clarity and Comprehensibility: 5 - The response is clear and easy to follow, with each step explained in a straightforward manner.\n",
- "\n",
- "Overall Score: 5\n",
- "Summary: The response provides a thorough and accurate calculation for determining the time required to dry 100 shirts based on the given information. The explanation is clear, logical, and easy to understand.\n",
- "===================\n"
- ]
- }
- ]
- },
- {
- "cell_type": "markdown",
- "source": [
- "### Feedback"
- ],
- "metadata": {
- "id": "RO6P1nSH6f-g"
- }
- },
- {
- "cell_type": "markdown",
- "source": [
- "Portkey provides a convenient way to view feedback on API calls. The feedback section displays both the request and the assistant's response, along with their respective token counts. Users can review the original prompt and the generated response. This feedback mechanism allows developers to assess the quality and effectiveness of the generated responses, enabling iterative improvements to the API and the underlying language models."
- ],
- "metadata": {
- "id": "Av-Bf6zV7K2l"
- }
- },
- {
- "cell_type": "markdown",
- "source": [
- "{prompt} \"\n",
+ " }\n",
+ " ],\n",
+ " temperature=0.3,\n",
+ " top_p=1,\n",
+ " stream=False,\n",
+ " response_format={\"type\": \"json_object\"},\n",
+ " stop=None,\n",
+ " )\n",
+ "\n",
+ "print(completion.choices[0].message.content)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "duGMAl0HmmH-",
+ "outputId": "aa0b2829-3eca-4622-e9b0-e33874b9f862"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "{\n",
+ " \"Preface\": \"An overview of the book's purpose and the author's connection to Bangalore Lights.\",\n",
+ " \"Chapter 1: The History of Bangalore Lights\": {\n",
+ " \"Introduction to Bangalore Lights\": \"Exploring the origins and evolution of Bangalore Lights.\",\n",
+ " \"Significance in Bangalore's Culture\": \"Discussing the cultural impact and importance of Bangalore Lights in the city.\",\n",
+ " \"Historical Landmarks\": \"Highlighting key historical events and landmarks associated with Bangalore Lights.\"\n",
+ " },\n",
+ " \"Chapter 2: The Technology Behind Bangalore Lights\": {\n",
+ " \"Evolution of Lighting Technology\": \"Tracing the development of lighting technology in Bangalore over the years.\",\n",
+ " \"Innovations and Trends\": \"Exploring the latest innovations and trends in lighting technology specific to Bangalore.\",\n",
+ " \"Sustainable Practices\": \"Examining the adoption of sustainable lighting practices in Bangalore Lights.\"\n",
+ " },\n",
+ " \"Chapter 3: Bangalore Lights in Art and Architecture\": {\n",
+ " \"Lighting in Architecture\": \"Exploring how Bangalore Lights have influenced architectural designs in the city.\",\n",
+ " \"Artistic Representations\": \"Showcasing how artists have captured Bangalore Lights in their works.\",\n",
+ " \"Light Festivals and Events\": \"Highlighting the various light festivals and events that celebrate Bangalore Lights.\"\n",
+ " },\n",
+ " \"Chapter 4: Social Impact of Bangalore Lights\": {\n",
+ " \"Community Engagement\": \"Discussing how Bangalore Lights have fostered community engagement and interaction.\",\n",
+ " \"Economic Contributions\": \"Exploring the economic impact of Bangalore Lights on businesses and tourism.\",\n",
+ " \"Social Initiatives\": \"Highlighting social initiatives and projects that use Bangalore Lights for positive change.\"\n",
+ " },\n",
+ " \"Chapter 5: Future Prospects and Sustainability\": {\n",
+ " \"Technological Advancements\": \"Predicting future technological advancements in Bangalore Lights.\",\n",
+ " \"Sustainability Practices\": \"Exploring the potential for further sustainability in the use of Bangalore Lights.\",\n",
+ " \"Community Development\": \"Discussing how Bangalore Lights can contribute to the future development of the city.\"\n",
+ " },\n",
+ " \"Appendix\": \"Additional resources, references, and further reading materials related to Bangalore Lights.\"\n",
+ "}\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "prompt = \"Bangalore Lights\"\n",
+ "\n",
+ "completion = groq.chat.completions.create(\n",
+ " model=\"llama3-70b-8192\",\n",
+ " messages=[\n",
+ " {\n",
+ " \"role\": \"system\",\n",
+ " \"content\": \"Write in JSON format:\\n\\n{\\\"Title of section goes here\\\":\\\"Description of section goes here\\\",\\n\\\"Title of section goes here\\\":{\\\"Title of section goes here\\\":\\\"Description of section goes here\\\",\\\"Title of section goes here\\\":\\\"Description of section goes here\\\",\\\"Title of section goes here\\\":\\\"Description of section goes here\\\"}}\"\n",
+ " },\n",
+ " {\n",
+ " \"role\": \"user\",\n",
+ " \"content\": f\"Write a comprehensive structure, omitting introduction and conclusion sections (forward, author's note, summary), for a book on the following subject:\\n\\n{prompt} \"\n",
+ " }\n",
+ " ],\n",
+ " temperature=0.3,\n",
+ " max_tokens=8000,\n",
+ " top_p=1,\n",
+ " stream=False,\n",
+ " response_format={\"type\": \"json_object\"},\n",
+ " stop=None,\n",
+ " )\n",
+ "\n",
+ "completion"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "5wqSOqc1irhG",
+ "outputId": "873210b2-ab1d-4a57-ec99-64bf2d3f47a4"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "ChatCompletion(id='chatcmpl-352927fe-7b1f-4d44-a0be-140434045a3a', choices=[Choice(finish_reason='stop', index=0, logprobs=None, message=ChatCompletionMessage(content='{\\n\"Bangalore Lights\": {\\n\"History of Bangalore\": {\\n\"Founding and Early Years\": \"Description of the founding and early years of Bangalore\",\\n\"British Colonial Era\": \"Description of Bangalore during the British colonial era\",\\n\"Post-Independence\": \"Description of Bangalore after India gained independence\"\\n},\\n\"Neighborhoods\": {\\n\"MG Road\": \"Description of MG Road neighborhood\",\\n\"Indiranagar\": \"Description of Indiranagar neighborhood\",\\n\"Koramangala\": \"Description of Koramangala neighborhood\",\\n\"Electronic City\": \"Description of Electronic City neighborhood\"\\n},\\n\"Culture\": {\\n\"Cuisine\": \"Description of Bangalore\\'s cuisine\",\\n\"Festivals and Celebrations\": \"Description of festivals and celebrations in Bangalore\",\\n\"Arts and Entertainment\": \"Description of arts and entertainment in Bangalore\"\\n},\\n\"Economy\": {\\n\"IT Industry\": \"Description of Bangalore\\'s IT industry\",\\n\"Startups and Entrepreneurship\": \"Description of startups and entrepreneurship in Bangalore\",\\n\"Other Industries\": \"Description of other industries in Bangalore\"\\n},\\n\"Tourism\": {\\n\"Popular Attractions\": \"Description of popular attractions in Bangalore\",\\n\"Day Trips\": \"Description of day trips from Bangalore\",\\n\"Accommodation and Dining\": \"Description of accommodation and dining options in Bangalore\"\\n},\\n\"Infrastructure\": {\\n\"Transportation\": \"Description of transportation options in Bangalore\",\\n\"Education\": \"Description of education system in Bangalore\",\\n\"Healthcare\": \"Description of healthcare system in Bangalore\"\\n}\\n}\\n}', role='assistant', function_call=None, tool_calls=None))], created=1719975827, model='llama3-70b-8192', object='chat.completion', service_tier=None, system_fingerprint='fp_753a4aecf6', usage=CompletionUsage(completion_tokens=302, prompt_tokens=117, total_tokens=419, prompt_time=0.033555685, completion_time=0.862857143, total_time=0.8964128280000001), x_groq={'id': 'req_01j1v8rmwzex1s3abe25c2wpyf'})"
+ ]
+ },
+ "metadata": {},
+ "execution_count": 3
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "print(completion.choices[0].message.content)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "x5sHAheSjFDu",
+ "outputId": "e303cdb5-3393-427b-cef4-2fd7df23d801"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "{\n",
+ "\"Bangalore Lights\": {\n",
+ "\"History of Bangalore\": {\n",
+ "\"Founding and Early Years\": \"Description of the founding and early years of Bangalore\",\n",
+ "\"British Colonial Era\": \"Description of Bangalore during the British colonial era\",\n",
+ "\"Post-Independence\": \"Description of Bangalore after India gained independence\"\n",
+ "},\n",
+ "\"Neighborhoods\": {\n",
+ "\"MG Road\": \"Description of MG Road neighborhood\",\n",
+ "\"Indiranagar\": \"Description of Indiranagar neighborhood\",\n",
+ "\"Koramangala\": \"Description of Koramangala neighborhood\",\n",
+ "\"Electronic City\": \"Description of Electronic City neighborhood\"\n",
+ "},\n",
+ "\"Culture\": {\n",
+ "\"Cuisine\": \"Description of Bangalore's cuisine\",\n",
+ "\"Festivals and Celebrations\": \"Description of festivals and celebrations in Bangalore\",\n",
+ "\"Arts and Entertainment\": \"Description of arts and entertainment in Bangalore\"\n",
+ "},\n",
+ "\"Economy\": {\n",
+ "\"IT Industry\": \"Description of Bangalore's IT industry\",\n",
+ "\"Startups and Entrepreneurship\": \"Description of startups and entrepreneurship in Bangalore\",\n",
+ "\"Other Industries\": \"Description of other industries in Bangalore\"\n",
+ "},\n",
+ "\"Tourism\": {\n",
+ "\"Popular Attractions\": \"Description of popular attractions in Bangalore\",\n",
+ "\"Day Trips\": \"Description of day trips from Bangalore\",\n",
+ "\"Accommodation and Dining\": \"Description of accommodation and dining options in Bangalore\"\n",
+ "},\n",
+ "\"Infrastructure\": {\n",
+ "\"Transportation\": \"Description of transportation options in Bangalore\",\n",
+ "\"Education\": \"Description of education system in Bangalore\",\n",
+ "\"Healthcare\": \"Description of healthcare system in Bangalore\"\n",
+ "}\n",
+ "}\n",
+ "}\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "source": [
+ "### Converting JSON output into a Pydantic format"
+ ],
+ "metadata": {
+ "id": "Flkv23O8mTdu"
+ }
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "from pydantic import BaseModel\n",
+ "from typing import List\n",
+ "\n",
+ "class MCQ(BaseModel):\n",
+ " \"\"\"A class representing a multiple choice question.\"\"\"\n",
+ " question: str\n",
+ " options: List[str]\n",
+ " correct_answer: str\n",
+ " explanation: str\n",
+ "\n",
+ "class MCQList(BaseModel):\n",
+ " questions: List[MCQ]\n"
+ ],
+ "metadata": {
+ "id": "EbUGbRkvjPDn"
+ },
+ "execution_count": null,
+ "outputs": []
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "topic = \"Quantum Entanglement\"\n",
+ "\n",
+ "completion = groq.chat.completions.create(\n",
+ " model=\"llama3-70b-8192\",\n",
+ " messages=[\n",
+ " {\n",
+ " \"role\": \"system\",\n",
+ " \"content\": \"\"\"Generate a list of multiple choice questions in the following JSON format:\n",
+ " {\n",
+ " \"questions\": [\n",
+ " {\n",
+ " \"question\": \"Question text goes here\",\n",
+ " \"options\": [\"Option A\", \"Option B\", \"Option C\", \"Option D\"],\n",
+ " \"correct_answer\": \"Correct Option in text\",\n",
+ " \"explanation\": \"Explanation for the correct answer\"\n",
+ " },\n",
+ " {\n",
+ " // Additional questions follow the same structure\n",
+ " }\n",
+ " ]\n",
+ " }\n",
+ " \"\"\"\n",
+ " },\n",
+ " {\n",
+ " \"role\": \"user\",\n",
+ " \"content\": f\"Generate 5 multiple choice questions on the following subject:\\n\\n{topic} \"\n",
+ " }\n",
+ " ],\n",
+ " temperature=0.3,\n",
+ " max_tokens=8000,\n",
+ " top_p=1,\n",
+ " stream=False,\n",
+ " response_format={\"type\": \"json_object\"},\n",
+ " stop=None,\n",
+ ")\n",
+ "\n",
+ "print(completion.choices[0].message.content)"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "9UP_7-rwlSec",
+ "outputId": "194cde0b-a109-4d67-e5ef-c1f1b2de7659"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "stream",
+ "name": "stdout",
+ "text": [
+ "{\n",
+ "\"questions\": [\n",
+ "{\n",
+ "\"question\": \"What is quantum entanglement?\",\n",
+ "\"options\": [\"A phenomenon where two particles are connected by a spring\", \"A process where two particles become correlated in such a way that the state of one particle cannot be described independently of the others\", \"A type of nuclear reaction\", \"A property of black holes\"],\n",
+ "\"correct_answer\": \"A process where two particles become correlated in such a way that the state of one particle cannot be described independently of the others\",\n",
+ "\"explanation\": \"Quantum entanglement is a phenomenon in which two or more particles become correlated in such a way that the state of one particle cannot be described independently of the others, even when they are separated by large distances.\"\n",
+ "},\n",
+ "{\n",
+ "\"question\": \"What is the key feature of entangled particles?\",\n",
+ "\"options\": [\"They are always in the same location\", \"They are always moving at the same speed\", \"They are connected by a physical medium\", \"Their properties are correlated, regardless of the distance between them\"],\n",
+ "\"correct_answer\": \"Their properties are correlated, regardless of the distance between them\",\n",
+ "\"explanation\": \"The key feature of entangled particles is that their properties, such as spin or polarization, are correlated, regardless of the distance between them.\"\n",
+ "},\n",
+ "{\n",
+ "\"question\": \"What is the concept that explains how entangled particles can affect each other instantaneously?\",\n",
+ "\"options\": [\"Quantum teleportation\", \"Wormholes\", \"Spooky action at a distance\", \"Quantum superposition\"],\n",
+ "\"correct_answer\": \"Spooky action at a distance\",\n",
+ "\"explanation\": \"The concept that explains how entangled particles can affect each other instantaneously is often referred to as 'spooky action at a distance', a term coined by Albert Einstein.\"\n",
+ "},\n",
+ "{\n",
+ "\"question\": \"What is the potential application of quantum entanglement?\",\n",
+ "\"options\": [\"Developing more efficient solar panels\", \"Creating more powerful magnets\", \"Enabling secure quantum communication\", \"Building faster-than-light spacecraft\"],\n",
+ "\"correct_answer\": \"Enabling secure quantum communication\",\n",
+ "\"explanation\": \"One of the potential applications of quantum entanglement is enabling secure quantum communication, as any attempt to measure or eavesdrop on the communication would disturb the entanglement and be detectable.\"\n",
+ "},\n",
+ "{\n",
+ "\"question\": \"Who is credited with the concept of quantum entanglement?\",\n",
+ "\"options\": [\"Albert Einstein\", \"Niels Bohr\", \"Erwin SchrΓΆdinger\", \"Werner Heisenberg\"],\n",
+ "\"correct_answer\": \"Erwin SchrΓΆdinger\",\n",
+ "\"explanation\": \"Erwin SchrΓΆdinger is credited with coining the term 'entanglement' and developing the concept of quantum entanglement in the 1930s.\"\n",
+ "}\n",
+ "]\n",
+ "}\n"
+ ]
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [
+ "# Parse the JSON response into an MCQList object\n",
+ "mcq_list = MCQList.parse_raw(completion.choices[0].message.content)\n",
+ "\n",
+ "mcq_list"
+ ],
+ "metadata": {
+ "colab": {
+ "base_uri": "https://localhost:8080/"
+ },
+ "id": "3pYf4ooplvtQ",
+ "outputId": "6dc37320-bbab-4b05-d7f1-32fb370564e7"
+ },
+ "execution_count": null,
+ "outputs": [
+ {
+ "output_type": "execute_result",
+ "data": {
+ "text/plain": [
+ "MCQList(questions=[MCQ(question='What is quantum entanglement?', options=['A phenomenon where two particles are connected by a spring', 'A process where two particles become correlated in such a way that the state of one particle cannot be described independently of the others', 'A type of nuclear reaction', 'A property of black holes'], correct_answer='A process where two particles become correlated in such a way that the state of one particle cannot be described independently of the others', explanation='Quantum entanglement is a phenomenon in which two or more particles become correlated in such a way that the state of one particle cannot be described independently of the others, even when they are separated by large distances.'), MCQ(question='What is the key feature of entangled particles?', options=['They are always in the same location', 'They are always moving at the same speed', 'They are connected by a physical medium', 'Their properties are correlated, regardless of the distance between them'], correct_answer='Their properties are correlated, regardless of the distance between them', explanation='The key feature of entangled particles is that their properties, such as spin or polarization, are correlated, regardless of the distance between them.'), MCQ(question='What is the concept that explains how entangled particles can affect each other instantaneously?', options=['Quantum teleportation', 'Wormholes', 'Spooky action at a distance', 'Quantum superposition'], correct_answer='Spooky action at a distance', explanation=\"The concept that explains how entangled particles can affect each other instantaneously is often referred to as 'spooky action at a distance', a term coined by Albert Einstein.\"), MCQ(question='What is the potential application of quantum entanglement?', options=['Developing more efficient solar panels', 'Creating more powerful magnets', 'Enabling secure quantum communication', 'Building faster-than-light spacecraft'], correct_answer='Enabling secure quantum communication', explanation='One of the potential applications of quantum entanglement is enabling secure quantum communication, as any attempt to measure or eavesdrop on the communication would disturb the entanglement and be detectable.'), MCQ(question='Who is credited with the concept of quantum entanglement?', options=['Albert Einstein', 'Niels Bohr', 'Erwin SchrΓΆdinger', 'Werner Heisenberg'], correct_answer='Erwin SchrΓΆdinger', explanation=\"Erwin SchrΓΆdinger is credited with coining the term 'entanglement' and developing the concept of quantum entanglement in the 1930s.\")])"
+ ]
+ },
+ "metadata": {},
+ "execution_count": 15
+ }
+ ]
+ },
+ {
+ "cell_type": "code",
+ "source": [],
+ "metadata": {
+ "id": "80TLM8gVmB6g"
+ },
+ "execution_count": null,
+ "outputs": []
+ }
+ ]
+}
\ No newline at end of file
![\"portkey\"](\"https://analyticsindiamag.com/wp-content/uploads/2023/08/Logo-on-white-background.png\"){kind=logo}
\n",
+ " \n",
+ "