Skip to content

Commit

Permalink
Update more notebooks to be available on the website (microsoft#1890)
Browse files Browse the repository at this point in the history 
* Update more notebooks to be available on the website

* fix notebook

* update link
  • Loading branch information
@jackgerrits
jackgerrits committed Mar 10, 2024
1 parent 54ad6d1 commit 3622aa3
Show file tree
Hide file tree
Showing 7 changed files with 147 additions and 305 deletions.
40 changes: 27 additions & 13 deletions notebook/agentchat_capability_long_context_handling.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4,19 +4,20 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"### Handling A Long Context via `TransformChatHistory`\n",
"# Handling A Long Context via `TransformChatHistory`\n",
"\n",
"This notebook illustrates how you can use the `TransformChatHistory` capability to give any `Conversable` agent an ability to handle a long context. "
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"## Uncomment to install pyautogen if you don't have it already\n",
"#! pip install pyautogen"
"This notebook illustrates how you can use the `TransformChatHistory` capability to give any `Conversable` agent an ability to handle a long context. \n",
"\n",
"````{=mdx}\n",
":::info Requirements\n",
"Install `pyautogen`:\n",
"```bash\n",
"pip install pyautogen\n",
"```\n",
"\n",
"For more information, please refer to the [installation guide](/docs/installation/).\n",
":::\n",
"````"
]
},
{
Expand Down Expand Up @@ -45,6 +46,12 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"````{=mdx}\n",
":::tip\n",
"Learn more about configuring LLMs for agents [here](/docs/topics/llm_configuration).\n",
":::\n",
"````\n",
"\n",
"To add this ability to any agent, define the capability and then use `add_to_agent`."
]
},
Expand Down Expand Up @@ -652,6 +659,13 @@
}
],
"metadata": {
"front_matter": {
"description": "Use the TransformChatHistory capability to handle long contexts",
"tags": [
"long context handling",
"capability"
]
},
"kernelspec": {
"display_name": "Python 3",
"language": "python",
Expand All @@ -667,7 +681,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.10.13"
"version": "3.11.7"
}
},
"nbformat": 4,
Expand Down
79 changes: 22 additions & 57 deletions notebook/agentchat_chess.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -5,15 +5,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"<a href=\"https://colab.research.google.com/github/microsoft/autogen/blob/main/notebook/agentchat_chess.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"# Auto Generated Agent Chat: Chess Game Playing While Chitchatting by GPT-4 Agents\n",
"# Chess Game Playing While Chitchatting by GPT-4 Agents\n",
"\n",
"AutoGen offers conversable agents powered by LLM, tool, or human, which can be used to perform tasks collectively via automated chat. This framework allows tool use and human participation through multi-agent conversation.\n",
"Please find documentation about this feature [here](https://microsoft.github.io/autogen/docs/Use-Cases/agent_chat).\n",
Expand All @@ -22,10 +14,17 @@
"\n",
"## Requirements\n",
"\n",
"AutoGen requires `Python>=3.8`. To run this notebook example, please install:\n",
"````{=mdx}\n",
":::info Requirements\n",
"Some extra dependencies are needed for this notebook, which can be installed via pip:\n",
"\n",
"```bash\n",
"pip install pyautogen\n",
"```"
"pip install pyautogen chess\n",
"```\n",
"\n",
"For more information, please refer to the [installation guide](/docs/installation/).\n",
":::\n",
"````"
]
},
{
Expand All @@ -35,16 +34,13 @@
"outputs": [],
"source": [
"%%capture --no-stderr\n",
"# %pip install \"pyautogen>=0.2.3\"\n",
"from collections import defaultdict\n",
"from typing import Any, Dict, List, Optional, Union\n",
"\n",
"import chess\n",
"import chess.svg\n",
"\n",
"import autogen\n",
"\n",
"%pip install chess -U"
"import autogen"
]
},
{
Expand All @@ -68,54 +64,19 @@
" filter_dict={\n",
" \"model\": [\"gpt-4\", \"gpt4\", \"gpt-4-32k\", \"gpt-4-32k-0314\", \"gpt-4-32k-v0314\"],\n",
" },\n",
")\n",
"# config_list_gpt35 = autogen.config_list_from_json(\n",
"# \"OAI_CONFIG_LIST\",\n",
"# filter_dict={\n",
"# \"model\": {\n",
"# \"gpt-3.5-turbo\",\n",
"# \"gpt-3.5-turbo-16k\",\n",
"# \"gpt-3.5-turbo-16k-0613\",\n",
"# \"gpt-3.5-turbo-0301\",\n",
"# \"chatgpt-35-turbo-0301\",\n",
"# \"gpt-35-turbo-v0301\",\n",
"# },\n",
"# },\n",
"# )"
")"
]
},
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"It first looks for environment variable \"OAI_CONFIG_LIST\" which needs to be a valid json string. If that variable is not found, it then looks for a json file named \"OAI_CONFIG_LIST\". It filters the configs by models (you can filter by other keys as well). Only the gpt-4 models are kept in the list based on the filter condition.\n",
"\n",
"The config list looks like the following:\n",
"```python\n",
"config_list = [\n",
" {\n",
" 'model': 'gpt-4',\n",
" 'api_key': '<your OpenAI API key here>',\n",
" },\n",
" {\n",
" 'model': 'gpt-4',\n",
" 'api_key': '<your Azure OpenAI API key here>',\n",
" 'base_url': '<your Azure OpenAI API base here>',\n",
" 'api_type': 'azure',\n",
" 'api_version': '2024-02-15-preview',\n",
" },\n",
" {\n",
" 'model': 'gpt-4-32k',\n",
" 'api_key': '<your Azure OpenAI API key here>',\n",
" 'base_url': '<your Azure OpenAI API base here>',\n",
" 'api_type': 'azure',\n",
" 'api_version': '2024-02-15-preview',\n",
" },\n",
"]\n",
"```\n",
"\n",
"You can set the value of config_list in any way you prefer. Please refer to this [notebook](https://github.com/microsoft/autogen/blob/main/notebook/oai_openai_utils.ipynb) for full code examples of the different methods."
"````{=mdx}\n",
":::tip\n",
"Learn more about configuring LLMs for agents [here](/docs/topics/llm_configuration).\n",
":::\n",
"````"
]
},
{
Expand Down Expand Up @@ -996,6 +957,10 @@
}
],
"metadata": {
"front_matter": {
"tags": ["chess"],
"description": "Use AutoGen to create two agents that are able to play chess"
},
"kernelspec": {
"display_name": "flaml",
"language": "python",
Expand Down
73 changes: 23 additions & 50 deletions notebook/agentchat_compression.ipynb
View file
Original file line number Diff line number Diff line change
Expand Up @@ -4,14 +4,7 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"<a href=\"https://colab.research.google.com/github/microsoft/autogen/blob/main/notebook/agentchat_compression.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Auto Generated Agent Chat: Conversations with Chat History Compression Enabled\n",
"# Conversations with Chat History Compression Enabled\n",
"\n",
"**CompressibleAgent will be deprecated.** \n",
"\n",
Expand All @@ -22,9 +15,11 @@
"In this notebook, we demonstrate how to enable compression of history messages using the `CompressibleAgent`. While this agent retains all the default functionalities of the `AssistantAgent`, it also provides the added feature of compression when activated through the `compress_config` setting.\n",
"\n",
"Different compression modes are supported:\n",
"\n",
"1. `compress_config=False` (Default): `CompressibleAgent` is equivalent to `AssistantAgent`.\n",
"2. `compress_config=True` or `compress_config={\"mode\": \"TERMINATE\"}`: no compression will be performed. However, we will count token usage before sending requests to the OpenAI model. The conversation will be terminated directly if the total token usage exceeds the maximum token usage allowed by the model (to avoid the token limit error from OpenAI API).\n",
"3. `compress_config={\"mode\": \"COMPRESS\", \"trigger_count\": <your pre-set number>}, \"leave_last_n\": <your pre-set number>`: compression is enabled.\n",
"3. `compress_config={\"mode\": \"COMPRESS\", \"trigger_count\": <your pre-set number>, \"leave_last_n\": <your pre-set number>}`: compression is enabled.\n",
"\n",
" ```python\n",
" # default compress_config\n",
" compress_config = {\n",
Expand All @@ -38,12 +33,13 @@
" \"verbose\": False, # if True, print out the content to be compressed and the compressed content\n",
" }\n",
" ```\n",
"\n",
" Currently, our compression logic is as follows:\n",
" 1. We will always leave the first user message (as well as system prompts) and compress the rest of the history messages.\n",
" 2. You can choose to not compress the last n messages in the history with \"leave_last_n\".\n",
" 2. The summary is performed on a per-message basis, with the role of the messages (See compressed content in the example below).\n",
"\n",
"4. `compress_config={\"mode\": \"CUSTOMIZED\", \"compress_function\": <A customized function for compression>}`: the `compress_function` function will be called on trigger count. The function should accept a list of messages as input and return a tuple of (is_success: bool, compressed_messages: List[Dict]). The whole message history (except system prompt) will be passed.\n",
"4. `compress_config={\"mode\": \"CUSTOMIZED\", \"compress_function\": <A customized function for compression>}t`: the `compress_function` function will be called on trigger count. The function should accept a list of messages as input and return a tuple of (is_success: bool, compressed_messages: List[Dict]). The whole message history (except system prompt) will be passed.\n",
"\n",
"\n",
"By adjusting `trigger_count`, you can decide when to compress the history messages based on existing tokens. If this is a float number between 0 and 1, it is interpreted as a ratio of max tokens allowed by the model. For example, the AssistantAgent uses gpt-4 with max tokens 8192, the trigger_count = 0.7 * 8192 = 5734.4 -> 5734. Do not set `trigger_count` to the max tokens allowed by the model, since the same LLM is employed for compression and it needs tokens to generate the compressed content. \n",
Expand All @@ -56,19 +52,16 @@
"\n",
"## Requirements\n",
"\n",
"AutoGen requires `Python>=3.8`. To run this notebook example, please install:\n",
"````{=mdx}\n",
":::info Requirements\n",
"Install `pyautogen`:\n",
"```bash\n",
"pip install pyautogen\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"# %pip install pyautogen~=0.1.0"
"```\n",
"\n",
"For more information, please refer to the [installation guide](/docs/installation/).\n",
":::\n",
"````"
]
},
{
Expand Down Expand Up @@ -105,35 +98,11 @@
"cell_type": "markdown",
"metadata": {},
"source": [
"It first looks for environment variable \"OAI_CONFIG_LIST\" which needs to be a valid json string. If that variable is not found, it then looks for a json file named \"OAI_CONFIG_LIST\". It filters the configs by models (you can filter by other keys as well).\n",
"\n",
"The config list looks like the following:\n",
"```python\n",
"config_list = [\n",
" {\n",
" 'model': 'gpt-4',\n",
" 'api_key': '<your OpenAI API key here>',\n",
" },\n",
" {\n",
" 'model': 'gpt-4',\n",
" 'api_key': '<your Azure OpenAI API key here>',\n",
" 'base_url': '<your Azure OpenAI API base here>',\n",
" 'api_type': 'azure',\n",
" 'api_version': '2024-02-15-preview',\n",
" },\n",
" {\n",
" 'model': 'gpt-4-32k',\n",
" 'api_key': '<your Azure OpenAI API key here>',\n",
" 'base_url': '<your Azure OpenAI API base here>',\n",
" 'api_type': 'azure',\n",
" 'api_version': '2024-02-15-preview',\n",
" },\n",
"]\n",
"```\n",
"\n",
"If you open this notebook in colab, you can upload your files by clicking the file icon on the left panel and then choose \"upload file\" icon.\n",
"\n",
"You can set the value of config_list in other ways you prefer, e.g., loading from a YAML file."
"````{=mdx}\n",
":::tip\n",
"Learn more about configuring LLMs for agents [here](/docs/topics/llm_configuration).\n",
":::\n",
"````"
]
},
{
Expand Down Expand Up @@ -884,6 +853,10 @@
}
],
"metadata": {
"front_matter": {
"description": "Learn about the CompressibleAgent",
"tags": []
},
"kernelspec": {
"display_name": "msft",
"language": "python",
Expand Down
41 changes: 15 additions & 26 deletions notebook/agentchat_custom_model.ipynb
View file
Original file line number Diff line number Diff line change
@@ -1,13 +1,5 @@
{
"cells": [
{
"attachments": {},
"cell_type": "markdown",
"metadata": {},
"source": [
"<a href=\"https://colab.research.google.com/github/microsoft/autogen/blob/main/notebook/agentchat_custom_model.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
]
},
{
"attachments": {},
"cell_type": "markdown",
Expand All @@ -25,26 +17,17 @@
"\n",
"## Requirements\n",
"\n",
"AutoGen requires `Python>=3.8`. To run this notebook example, please install:\n",
"````{=mdx}\n",
":::info Requirements\n",
"Some extra dependencies are needed for this notebook, which can be installed via pip:\n",
"\n",
"```bash\n",
"pip install pyautogen torch transformers sentencepiece\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"execution": {
"iopub.execute_input": "2023-02-13T23:40:52.317406Z",
"iopub.status.busy": "2023-02-13T23:40:52.316561Z",
"iopub.status.idle": "2023-02-13T23:40:52.321193Z",
"shell.execute_reply": "2023-02-13T23:40:52.320628Z"
}
},
"outputs": [],
"source": [
"# %pip install pyautogen~=0.2.0b4 torch git+https://github.com/huggingface/transformers sentencepiece"
"```\n",
"\n",
"For more information, please refer to the [installation guide](/docs/installation/).\n",
":::\n",
"````"
]
},
{
Expand Down Expand Up @@ -455,6 +438,12 @@
}
],
"metadata": {
"front_matter": {
"description": "Define and laod a custom model",
"tags": [
"custom model"
]
},
"kernelspec": {
"display_name": "Python 3",
"language": "python",
Expand Down
Loading

0 comments on commit 3622aa3

Please sign in to comment.