# Multi LLM
VerifAI Implementation of invoking multiple large language models concurrently and ranking results
******
## Quick-Start
For further information look to [Running the Multi_LLM Application](#running-the-application) inside of [Usage](#usage)
### Installation
`pip3 install multillm`
Edit the [config file](#config-json) to point to your google and openai API keys.
Edit and add your "credentials" of each *llm* to the [config.json](config.json) file:
```json
"llms": [
{
"file": "bard.py",
"class_name": "BARD",
"model": "chat-bison@001",
* "credentials": "/path/to/google/key.json" *
},
```
<details>
<summary><strong><em> Example google-app-credentials.json</em></strong></summary>
```json
{
"client_id": "123489-6qr4p.apps.googleusercontent.com",
"client_secret": "fx-d3456-tryf0g9f9",
"quota_project_id": "my-llm-training",
"refresh_token": "1-34GFH89KLwe-eft",
"type": "authorized_user"
}
```
</details>
<details>
<summary><strong><em> Example openai-credentials.json</em></strong></summary>
```json
{
"organization" : "org-jc8901FDLI0267",
"api_key" : "rt-067FGDiTL834"
}
```
</details>
### Basic Example
`multillm -c config.json -prompt "write a python function to find a root of the function f using Newton's method"`
The above example will run the [config.json](config.json) with the provided prompt.
<details>
<summary><strong>Output of the above multillm command</strong></summary>
```bash
loading module bard...
finished loading module bard
registered model BARD <bard.BARD object at 0x10e4e5b70>
loading module GPT...
finished loading module GPT
registered model GPT <GPT.GPT object at 0x10e4e5d20>
loaded llms: {'bard': <module 'bard' from '/Users/Verifai/models/bard.py'>, 'GPT': <module 'GPT' from '/Users/Verifai/models/GPT.py'>}
calling model: BARD
calling model: GPT
model chat-bison@001
```
*BARD Response:*
```python
def newton(f, df, x0, tol=1e-6, maxiter=100):
"""
Find a root of the function f using Newton's method.
Args:
f: The function to find a root of.
df: The derivative of f.
x0: The initial guess for the root.
tol: The tolerance for convergence.
maxiter: The maximum number of iterations to perform.
Returns:
The root of f, or None if no root was found.
"""
for i in range(maxiter):
x1 = x0 - f(x0) / df(x0)
if abs(x1 - x0) < tol:
return x1
x0 = x1
return None
```
*GPT Response:*
```python
def newton_method(f, f_prime, initial_guess, tol=1e-6, max_iter=100):
"""
Newton's method for finding the root of a function.
Parameters:
f (function): The function for which the root is to be found.
f_prime (function): The derivative of f.
initial_guess (float): The initial guess for the root.
tol (float): The desired tolerance (default 1e-6).
max_iter (int): The maximum number of iterations (default 100).
Returns:
float: The root found by Newton's method, or None if no root is found.
"""
x = initial_guess
for _ in range(max_iter):
fx = f(x)
if abs(fx) < tol:
return x
fpx = f_prime(x)
if fpx == 0:
return None
x -= fx / fpx
return None
```
### **MultiLLM** Explanation and Ranking for :
* *"The **GPT** implementation of the Newton's method function (newton_method) is well-documented with clear parameter explanations and return format. It also includes appropriate input validation checks and handles the case when the derivative is 0. Overall, it is a comprehensive and robust implementation."*
* *"The **BARD** implementation of the Newton's method function (newton) is also well-documented and accepts the necessary parameters. However, it lacks input validation checks, such as when the derivative is 0, and does not provide an explicit return value when no root is found. It can be improved by addressing these limitations."*
*"Based on these factors, I would rank GPT higher than BARD in terms of the clarity, completeness, and robustness of the implementation."*
</details>
******
### Development
Please follow below instructions to install and run the repo locally for development and testing:
```bash
# Clone the repo to local machine
git clone https://github.com/verifai/mllm.git
cd <your-repo-directory>
# Create a new python virtual environment
python3 -m venv myenv
source myenv/bin/activate
# install the dependencies
pip install -r multillm/requirements.txt
# prompt the multillm with the repo (not the package / multillm command) in macOS:
python run_multillm.py -c config.json -prompt "How to learn python? Answer me within 100 words." -debug -macos
# prompt the multillm with the repo (not the package / multillm command) in Linux:
python run_multillm.py -c config.json -prompt "How to learn python? Answer me within 100 words." -debug
# start the flask server (But for running the command, we don't need to start the flask server)
python serve.py
```
## Usage
The Multi_LLM application provides a powerful and efficient solution for invoking multiple large language models (LLMs) concurrently and effectively managing their outputs. This section guides you through the steps required to run the application using various command-line options and configuration files.
### Prerequisites
Before running the Multi_LLM application, ensure that you have the following prerequisites in place:
- Python >= 3.8 installed on your system.
### Command-Line Usage
The Multi_LLM application can be executed from the command line using the following command structure:
```
multillm -c <config_file> -prompt "<prompt_text>"
```
- `<config_file>`: The path to the configuration file containing LLM details.
- `<prompt_text>`: The prompt text you want to provide to the language models.
### Config Files
The configuration file [config.json](config.json) defines the language models and their associated parameters. See [Config JSON](#config-json) for more information.
It follows the structure outlined below:
<details>
<summary><strong>Example Config.json file</strong></summary>
```json
{
"Config": {
"Multi_LLM": {
"llms": [
{
"file": "bard.py",
"class_name": "BARD",
"model": "chat-bison@001",
"credentials": "/path/to/google/key.json"
},
{
"file": "GPT.py",
"class_name": "GPT",
"model": "gpt-3.5-turbo",
"credentials": "/path/to/openai/key.json"
}
]
}
}
}
```
</details>
### Running the Application
To run the Multi_LLM application, follow these steps:
1. pip install multillm
2. Download the [config.json](config.json) file with the desired language model configurations.
3. Execute the following command, replacing `<config_file>` with the actual path to your configuration file and `<prompt_text>` with the desired prompt text:
```bash
multillm -c <config_file> -prompt "<prompt_text>"
```
4. The application will run the specified language models concurrently, process their responses using the provided prompt, and display the results.
******
## Adding a new LLM interface
This section will guide you through the process of adding a new LLM by extending the `BaseLLM` class and customizing it to fit the requirements of your specific language model.
### Follow these steps to add an new LLM interface
1. **Start by creating a new Python file in your project directory '<NewLLM.py>'** , or within the appropriate package, where you'll define a new class that inherits from `BaseLLM`. Implement the required methods: get_response() and get_content(). The get_response() method should execute your language model with the provided prompt, and the get_content() method should extract relevant content from the response.
- **See Example Below:**
<details> <summary>Example NewLLM.py</summary>
```python
import os,sys
from multillm.BaseLLM import BaseLLM
from multillm.Prompt import Prompt
# <add additional imports here>
# NewLLM interface
"""
The NewLLM class extends the BaseModel class and overrides the get_response() method, providing an implementation.
"""
class NewLLM(BaseLLM):
# ... (attributes and __init__ method)
def __init__ (self, **kwargs):
# add values here directly or if kwargs are specified they are taken from the config file
defaults = {
"class_name" : "NewLLM",
"model" : "newLLM-bison@06",
"credentials" : "/path/to-my/key.json"
}
# ... Call API and get response from NewLLM
def get_response(self, prompt):
# Implement your language model interaction here
# access credentials file from *self.credentials*
# access model from *self.model*
# access class_name from *self.class_name*
response = <"Generated response from NewLLM model based on prompt">
return response
# ... Parse and Filter raw response from NewLLM and return text/code content
def get_content(self, response):
# Implement content extraction from the response
content = "Extracted content from response"
return content
```
</details>
2. **Add NewLLM in config.json file, in the 'llms' section**
<details>
<summary><strong>Add NewLLM to the config.json file</strong></summary>
```json
{
"Config": {
"Multi_LLM": {
"llms": [
{
"file": "/full-path/NewLLM.py",
"class_name": "NEWLLM",
"model": "chat-bison@001",
"credentials": "/path/to/google/key.json"
}, .... ]
}
}
```
</details>
3. **Call multillm to run your new LLM or embedd it in your code**:
- `multillm -c config.json -prompt "wite a function to sort a billion integers"`
4. **(Optional) Embedd NewLLM.py in your code**
- You can now use your custom `NewLLM` class in your application code. Instantiate it, call its methods, and integrate it into your application's workflow.
```python
custom_llm = NewLLM(model="custom_model", credentials="your_credentials")
prompt = "Generate something amazing."
response = custom_llm.get_response(prompt)
content = custom_llm.get_content(response)
print(content)
```
By extending the provided `BaseLLM` class, you can easily create custom language model implementations tailored to your project's needs. This structured approach ensures consistency and modularity in your codebase, allowing you to focus on the unique aspects of your language model while leveraging the foundational structure provided by `BaseLLM`.
******
## Example of the GPT interface (included with multillm release)
<details>
<summary><strong>Example Model GPT.py</strong></summary>
```python
import os,sys
import openai
import json
from multillm.BaseLLM import BaseLLM
from multillm.Prompt import Prompt
# Openai gpt interface
"""
The GPT class extends the BaseModel class, implements the required methods: get_response() and get_content().
The get_response() method takes a response parameter and returns the content of the first response in the given response object.
"""
class GPT(BaseLLM):
#implement here
def __init__ (self, **kwargs):
# add values here directly or if kwargs are specified they are taken from the config file
defaults = {
"class_name" : "GPT",
"model" : "gpt-3.5-turbo",
"credentials" : "key.json"
}
# Get Content -- Required Method
def get_content(self, response):
""" Get the text from the response of an LLM
e.g.: openai returns the following response, this method should return the 'content'.
{ "choices": [{{"message":
{"content": "def binary_sort(arr):"}}}]}
"""
return response["choices"][0]["message"]["content"]
# Get Response -- Required Method, Call openai API
def get_response(self, prompt):
# setup prompt for API call
messages=[]
messages.append( {"role": prompt.get_role(), "content" : prompt.get_string()})
if prompt.context:
messages.append({"role": prompt.get_role(), "content" : prompt.get_context()})
# Read Credentials file specified in the config.json, setup for openai
if not os.path.exists(self.credentials):
print('error (multi_llm): could not find openai_credentials: {0}' .format(self.credentials))
return
# Open the file for reading
try:
with open(self.credentials, 'r') as file:
# Load the JSON data from the file
data = json.load(file)
openai.organization = data['organization']
openai.api_key = data['api_key']
except Exception as e:
print('(multi_llm) error: could not load credentials {0} : {1}' .format(self.credentials,str(e)))
return
# do API call
response = openai.ChatCompletion.create(
model = self.model,
messages=messages
)
if response:
return(self.get_content(response))
else:
return response
```
</details>
******
## Config JSON
### Using the `config.json` File
The `config.json` file offers a convenient way to configure and load multiple language models using the "Multi_LLM" framework. This section will guide you through the process of creating and utilizing a `config.json` file to load and use specific language models in your application.
### Configuration Setup
To get started, follow these steps to configure your `config.json` file:
1. **Create a Configuration File**: Create a new file named `config.json` in your project directory or the desired location.
2. **Configure LLMs**: Define the language models you want to use within the `"llms"` array. Each model configuration includes details such as the Python file containing the model implementation, class name, model name, and credentials file path.
```json
{
"Config": {
"Multi_LLM": {
"llms": [
{
"file": "bard.py",
"class_name": "BARD",
"model": "chat-bison@001",
"credentials": "/path/to/google/key.json"
},
{
"file": "GPT.py",
"class_name": "GPT",
"model": "gpt-3.5-turbo",
"credentials": "/path/to/openai/key.json"
}
]
}
}
}
```
******
### Loading Models Using MultiLLM
After creating the `config.json` file, you can use the "Multi_LLM" framework to load and utilize the configured models in your application. Follow these steps:
1. **Instantiate Multi_LLM**: Create an instance of the `Multi_LLM` class and provide the path to your `config.json` file.
```python
from Multi_LLM import Multi_LLM
# Specify the path to your config.json file
config_path = "path/to/config.json"
# Instantiate Multi_LLM
multi_llm = Multi_LLM(config=config_path)
```
2. **Run Models**: Use the `run` method to run the loaded models and process their responses. Provide a prompt and an optional action chain if needed. The responses from each model will be returned in a dictionary.
```python
prompt = "Translate this sentence."
action_chain = None # You can define an action chain if required
responses = multi_llm.run(prompt, action_chain)
print(responses)
```
## Action Class
This is the interface class which we use to operate on the output while still in parallel. Action class instances are define by the user and can be chained indefinitely with other Actiion instances.
There are two methods that are used to interface with the class: `apply()` and `then()`.
The first, `apply()` is what is used internally to call the methods we register. Registration is as follows:
```python
# Write an interfacing function, a simple in/out
def capitalize(data):
data = data.upper()
return data
def lower(data):
data = data.lower()
return data
# Create the respective objects
action1 = Action(operation=capitalize)
action2 = Action(operation=lower)
# Create the chain of actions
action_chain = action1.then(action2)
```
What we did was create some actions that are simple I/O operations on data. We can then chain them together using the `then()` method mentioned above. The order of operations for then is left to right, in this case action1 will go first then action2. These actions can be tailored to your own specification, since the actions are serial, you can specify the information that is going to be passed into the function and what information will be returned.
After you have created the action chain you can pass this into the `MultiLLM.run()` method and run.
### Action Chain
The `Action Chain` is a core component of processesing LLM output. In the provided [example.py](examples/example.py) we see the first action defined is for processing the LLM response and extracting the code alone from the response. This could be a first of many steps, what could follow could be saving the code to the file or extracting information through the `ast` module. The actions are meant to be ran on each of the model's outputs so they should be generalized for use.
## Rank Class
The `Rank` class is identical in functionality to the `Action` class, though the use is different. While the `Action` class is used to accept and modify data in each of the models, the `Rank` class' purpose is to modify the final combined output of the LLMs.
This output is stored in a dictionary keyed by the model's names. The `Rank` class's methods should be written to take this data and operate on it.
```python
#Write an interfacing function, a simple print
def print_llms(dictionary):
for key,val in dictionary.items():
print(key, val)
# Create Rank instances
rank_object = Rank(operation=print_llms)
# Running the LLM, assuming an instance of
# MultiLLM named mLLM
results = mLLLM.run(prompt, action_chain, rank_object)
```
In the above code we are doing very similar actions as those seen above. The methodoly is the same. In this case we only have a single object, so instead of creating a `rank_chain` we simply pass in our single instantiated object.
## Architecture
### High Level Overview
### MultiLLM Class
This is the highest level, here we can instantiate MultiLLM objects using either config files (see [config files](###Config-Files)) or manually instantiating a custom or hosted LLM.
In this class we call multiple LLMs concurrently and then we can operate on the results of each in parallel using the [Action Class](###Action-Class).
The "MultiLLM" Python code provides a versatile framework for managing and orchestrating multiple language models (LLMs) within a single application. This code is designed to enhance the efficiency of working with language models by enabling concurrent execution, response processing, and model loading from configuration files. The key features and components of the code include:
- **Concurrent Model Execution:** The code allows you to run multiple language models concurrently, facilitating efficient utilization of computational resources. This is particularly useful for scenarios that involve processing multiple prompts or tasks simultaneously.
- **Action Chain Processing:** The framework supports the concept of "action chains," which enable you to preprocess model responses using a sequence of predefined actions. This empowers you to refine and enhance the output generated by the language models.
- **Model Loading from Configuration:** You can load LLMs from a configuration file. This JSON-based configuration includes essential details about each model, such as the model class, associated credentials, and file paths. The code can dynamically load these models, making it easy to add new models or modify existing ones without altering the core codebase.
- **Redis Integration:** The code features optional integration with a Redis instance. It checks for a successful Redis connection and adjusts its behavior accordingly. If Redis connectivity is not established, the code gracefully handles the situation by setting a flag that reflects the absence of a Redis connection.
- **Simplified Model Registration:** The code includes a straightforward method to register models within the framework. This allows for the inclusion of custom LLM implementations while ensuring that the models are appropriately organized and accessible for execution.
By using the "MultiLLM" Python code, developers can streamline their interactions with multiple language models, seamlessly integrating them into various applications or projects. The code promotes modularity, reusability, and parallelism in working with language models, ultimately enhancing the user experience and productivity.
### BaseLLM
`BaseLLM` is designed to serve as the basis for implementing various language model classes. This `BaseLLM` class encapsulates essential attributes and methods necessary for interfacing with language models. The code establishes a structured foundation for building specific language model implementations and harmonizes their interaction within a larger application context.
Key features and components of the code include:
- **Attributes for Language Models:** The `BaseLLM` class declares a set of attributes, such as `model`, `roles`, `messages`, `temp`, `api_key`, `max_tokens`, and `args`, that are pertinent to language models. These attributes are meant to be customized and adapted as needed for specific model implementations.
- **Customizable Initialization:** The `__init__()` method in the `BaseLLM` class facilitates flexible instantiation of model instances by allowing the specification of custom values via keyword arguments. It enables the convenient configuration of attributes like `name`, `credentials`, `model`, and `class_name`.
- **Placeholder Methods:** The `BaseLLM` class defines two placeholder methods: `get_response()` and `get_content()`. The `get_response()` method is designed to receive a `Prompt` object and run the associated language model with the provided prompt. The actual implementation of this method is expected to be customized in derived classes to perform the model-specific interactions. Similarly, the `get_content()` method is meant to be implemented by deriving classes, providing an interface to extract relevant content from model responses.
- **Structured Framework:** The code encapsulates a structured framework that abstracts common functionalities of language models. By inheriting from the `BaseLLM` class, developers can focus on implementing model-specific interactions while leveraging the established structure for attribute handling and method placeholders.
`BaseLLM` aims to streamline the development of specific language model implementations by providing a consistent structure and standardized attributes. Developers can extend this base class to create custom language model classes that seamlessly integrate into the broader application ecosystem. This modular approach promotes reusability, maintainability, and consistent design patterns when working with various language models.
******
## Contribution
You can contribute by extending the models located in [models](models). See the [BaseLLM](#baselmm) section for more information on the necessary mechanisms for extending the BaseLLM class.
For further information contact [ethansaurusrex](https://github.com/ethansaurusrex)
Raw data
{
"_id": null,
"home_page": null,
"name": "multillm",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": null,
"author": null,
"author_email": "VerifAI Inc <hello@verifai.ai>",
"download_url": "https://files.pythonhosted.org/packages/0f/5e/3044c292ef9df9710d85344cb6c10d44e987ed522b191faebc6a44494e09/multillm-0.1019.tar.gz",
"platform": null,
"description": "# Multi LLM\n\nVerifAI Implementation of invoking multiple large language models concurrently and ranking results\n\n******\n\n## Quick-Start\n\nFor further information look to [Running the Multi_LLM Application](#running-the-application) inside of [Usage](#usage)\n\n### Installation\n\n`pip3 install multillm`\n\nEdit the [config file](#config-json) to point to your google and openai API keys. \nEdit and add your \"credentials\" of each *llm* to the [config.json](config.json) file:\n\n```json\n\"llms\": [\n {\n \"file\": \"bard.py\",\n \"class_name\": \"BARD\",\n \"model\": \"chat-bison@001\",\n * \"credentials\": \"/path/to/google/key.json\" *\n },\n```\n\n<details> \n <summary><strong><em> Example google-app-credentials.json</em></strong></summary>\n\n```json\n {\n \"client_id\": \"123489-6qr4p.apps.googleusercontent.com\",\n \"client_secret\": \"fx-d3456-tryf0g9f9\",\n \"quota_project_id\": \"my-llm-training\",\n \"refresh_token\": \"1-34GFH89KLwe-eft\",\n \"type\": \"authorized_user\"\n }\n```\n</details>\n\n\n<details> \n <summary><strong><em> Example openai-credentials.json</em></strong></summary>\n\n```json\n {\n \"organization\" : \"org-jc8901FDLI0267\",\n \"api_key\" : \"rt-067FGDiTL834\"\n}\n \n```\n</details>\n\n### Basic Example\n`multillm -c config.json -prompt \"write a python function to find a root of the function f using Newton's method\"`\n\nThe above example will run the [config.json](config.json) with the provided prompt. \n\n<details>\n\t<summary><strong>Output of the above multillm command</strong></summary>\n \n```bash\nloading module bard...\nfinished loading module bard\nregistered model BARD <bard.BARD object at 0x10e4e5b70>\nloading module GPT...\nfinished loading module GPT\nregistered model GPT <GPT.GPT object at 0x10e4e5d20>\nloaded llms: {'bard': <module 'bard' from '/Users/Verifai/models/bard.py'>, 'GPT': <module 'GPT' from '/Users/Verifai/models/GPT.py'>}\ncalling model: BARD\ncalling model: GPT\nmodel chat-bison@001\n```\n*BARD Response:*\n```python\ndef newton(f, df, x0, tol=1e-6, maxiter=100):\n \"\"\"\n Find a root of the function f using Newton's method.\n\n Args:\n f: The function to find a root of.\n df: The derivative of f.\n x0: The initial guess for the root.\n tol: The tolerance for convergence.\n maxiter: The maximum number of iterations to perform.\n \n\n Returns:\n The root of f, or None if no root was found.\n \"\"\"\n\n for i in range(maxiter):\n x1 = x0 - f(x0) / df(x0)\n if abs(x1 - x0) < tol:\n return x1\n x0 = x1\n return None\n```\n\n*GPT Response:*\n```python \ndef newton_method(f, f_prime, initial_guess, tol=1e-6, max_iter=100):\n \"\"\"\n Newton's method for finding the root of a function.\n\n Parameters:\n f (function): The function for which the root is to be found.\n f_prime (function): The derivative of f.\n initial_guess (float): The initial guess for the root.\n tol (float): The desired tolerance (default 1e-6).\n max_iter (int): The maximum number of iterations (default 100).\n\n Returns:\n float: The root found by Newton's method, or None if no root is found.\n \"\"\"\n x = initial_guess\n for _ in range(max_iter):\n fx = f(x)\n if abs(fx) < tol:\n return x\n fpx = f_prime(x)\n if fpx == 0:\n return None\n x -= fx / fpx\n return None\n```\n\t\n### **MultiLLM** Explanation and Ranking for :\n \n* *\"The **GPT** implementation of the Newton's method function (newton_method) is well-documented with clear parameter explanations and return format. It also includes appropriate input validation checks and handles the case when the derivative is 0. Overall, it is a comprehensive and robust implementation.\"*\n \n* *\"The **BARD** implementation of the Newton's method function (newton) is also well-documented and accepts the necessary parameters. However, it lacks input validation checks, such as when the derivative is 0, and does not provide an explicit return value when no root is found. It can be improved by addressing these limitations.\"*\n \n*\"Based on these factors, I would rank GPT higher than BARD in terms of the clarity, completeness, and robustness of the implementation.\"*\n</details>\n\n******\n\n### Development\n\nPlease follow below instructions to install and run the repo locally for development and testing:\n\n```bash\n# Clone the repo to local machine\ngit clone https://github.com/verifai/mllm.git\n\ncd <your-repo-directory>\n\n# Create a new python virtual environment\npython3 -m venv myenv\nsource myenv/bin/activate\n\n# install the dependencies\npip install -r multillm/requirements.txt\n\n# prompt the multillm with the repo (not the package / multillm command) in macOS:\npython run_multillm.py -c config.json -prompt \"How to learn python? Answer me within 100 words.\" -debug -macos\n\n# prompt the multillm with the repo (not the package / multillm command) in Linux:\npython run_multillm.py -c config.json -prompt \"How to learn python? Answer me within 100 words.\" -debug\n\n# start the flask server (But for running the command, we don't need to start the flask server)\npython serve.py\n```\n\n\n## Usage\n\nThe Multi_LLM application provides a powerful and efficient solution for invoking multiple large language models (LLMs) concurrently and effectively managing their outputs. This section guides you through the steps required to run the application using various command-line options and configuration files.\n\n### Prerequisites\n\nBefore running the Multi_LLM application, ensure that you have the following prerequisites in place:\n\n- Python >= 3.8 installed on your system.\n\n\n### Command-Line Usage\n\nThe Multi_LLM application can be executed from the command line using the following command structure:\n\n```\nmultillm -c <config_file> -prompt \"<prompt_text>\"\n```\n\n- `<config_file>`: The path to the configuration file containing LLM details.\n- `<prompt_text>`: The prompt text you want to provide to the language models.\n\n### Config Files\n\nThe configuration file [config.json](config.json) defines the language models and their associated parameters. See [Config JSON](#config-json) for more information.\nIt follows the structure outlined below: \n\n<details>\n\t<summary><strong>Example Config.json file</strong></summary>\n\n```json\n{\n \"Config\": {\n \"Multi_LLM\": {\n \"llms\": [\n {\n \"file\": \"bard.py\",\n \"class_name\": \"BARD\",\n \"model\": \"chat-bison@001\",\n \"credentials\": \"/path/to/google/key.json\"\n },\n {\n \"file\": \"GPT.py\",\n \"class_name\": \"GPT\",\n \"model\": \"gpt-3.5-turbo\",\n \"credentials\": \"/path/to/openai/key.json\"\n }\n ]\n }\n }\n}\n```\n</details>\n\n### Running the Application\n\nTo run the Multi_LLM application, follow these steps:\n\n1. pip install multillm\n \n2. Download the [config.json](config.json) file with the desired language model configurations.\n\n3. Execute the following command, replacing `<config_file>` with the actual path to your configuration file and `<prompt_text>` with the desired prompt text:\n ```bash\n multillm -c <config_file> -prompt \"<prompt_text>\"\n ```\n\n4. The application will run the specified language models concurrently, process their responses using the provided prompt, and display the results.\n\n******\n\n## Adding a new LLM interface\n\nThis section will guide you through the process of adding a new LLM by extending the `BaseLLM` class and customizing it to fit the requirements of your specific language model.\n\n### Follow these steps to add an new LLM interface\n\n1. **Start by creating a new Python file in your project directory '<NewLLM.py>'** , or within the appropriate package, where you'll define a new class that inherits from `BaseLLM`. Implement the required methods: get_response() and get_content(). The get_response() method should execute your language model with the provided prompt, and the get_content() method should extract relevant content from the response.\n - **See Example Below:** \n\n\t<details> <summary>Example NewLLM.py</summary>\n\t\n\t```python\n\timport os,sys\n\tfrom multillm.BaseLLM import BaseLLM\n\tfrom multillm.Prompt import Prompt\n\t# <add additional imports here>\n\t\n\t\n\t# NewLLM interface \n\t\"\"\" \n\tThe NewLLM class extends the BaseModel class and overrides the get_response() method, providing an implementation. \n\t\"\"\"\n\tclass NewLLM(BaseLLM):\n\t # ... (attributes and __init__ method)\n\t def __init__ (self, **kwargs):\n\t\t # add values here directly or if kwargs are specified they are taken from the config file\n\t defaults = {\n\t \"class_name\" : \"NewLLM\",\n\t \"model\" : \"newLLM-bison@06\",\n\t \"credentials\" : \"/path/to-my/key.json\"\n\t }\n\t # ... Call API and get response from NewLLM\n\t def get_response(self, prompt):\n\t # Implement your language model interaction here\n\t # access credentials file from *self.credentials*\n\t # access model from *self.model*\n\t \t# access class_name from *self.class_name*\n\t response = <\"Generated response from NewLLM model based on prompt\">\n\t return response\n\t\n\t # ... Parse and Filter raw response from NewLLM and return text/code content\n\t def get_content(self, response):\n\t # Implement content extraction from the response\n\t content = \"Extracted content from response\"\n\t return content\n\t```\n \n \t</details>\n\n2. **Add NewLLM in config.json file, in the 'llms' section**\n\n\t<details>\n\t <summary><strong>Add NewLLM to the config.json file</strong></summary>\n\n\t```json\n\t{\n\t \"Config\": {\n\t \"Multi_LLM\": {\n\t \"llms\": [\n\t {\n\t \"file\": \"/full-path/NewLLM.py\",\n\t \"class_name\": \"NEWLLM\",\n\t \"model\": \"chat-bison@001\",\n\t \"credentials\": \"/path/to/google/key.json\"\n\t }, .... ]\n\t\t}\n\t}\n\t```\n\t\n\t</details>\n\n\n3. **Call multillm to run your new LLM or embedd it in your code**:\n - `multillm -c config.json -prompt \"wite a function to sort a billion integers\"`\n \n4. **(Optional) Embedd NewLLM.py in your code**\n - You can now use your custom `NewLLM` class in your application code. Instantiate it, call its methods, and integrate it into your application's workflow.\n\n```python\ncustom_llm = NewLLM(model=\"custom_model\", credentials=\"your_credentials\")\nprompt = \"Generate something amazing.\"\nresponse = custom_llm.get_response(prompt)\ncontent = custom_llm.get_content(response)\nprint(content)\n```\n\nBy extending the provided `BaseLLM` class, you can easily create custom language model implementations tailored to your project's needs. This structured approach ensures consistency and modularity in your codebase, allowing you to focus on the unique aspects of your language model while leveraging the foundational structure provided by `BaseLLM`.\n\n******\n\n## Example of the GPT interface (included with multillm release)\n<details>\n <summary><strong>Example Model GPT.py</strong></summary>\n\t\n```python\n\t\nimport os,sys\nimport openai\nimport json\nfrom multillm.BaseLLM import BaseLLM\nfrom multillm.Prompt import Prompt\n\n\n# Openai gpt interface\n\"\"\"\nThe GPT class extends the BaseModel class, implements the required methods: get_response() and get_content().\nThe get_response() method takes a response parameter and returns the content of the first response in the given response object.\n\"\"\"\nclass GPT(BaseLLM):\n\n #implement here\n def __init__ (self, **kwargs):\n # add values here directly or if kwargs are specified they are taken from the config file\n defaults = {\n \"class_name\" : \"GPT\",\n \"model\" : \"gpt-3.5-turbo\",\n \"credentials\" : \"key.json\"\n }\n\n # Get Content -- Required Method\n def get_content(self, response):\n \n \"\"\" Get the text from the response of an LLM\n e.g.: openai returns the following response, this method should return the 'content'.\n { \"choices\": [{{\"message\": \n\t\t\t{\"content\": \"def binary_sort(arr):\"}}}]}\n \"\"\"\n return response[\"choices\"][0][\"message\"][\"content\"]\n\n # Get Response -- Required Method, Call openai API\n def get_response(self, prompt):\n # setup prompt for API call\n messages=[]\n \n messages.append( {\"role\": prompt.get_role(), \"content\" : prompt.get_string()})\n if prompt.context:\n messages.append({\"role\": prompt.get_role(), \"content\" : prompt.get_context()})\n \n # Read Credentials file specified in the config.json, setup for openai\n if not os.path.exists(self.credentials):\n print('error (multi_llm): could not find openai_credentials: {0}' .format(self.credentials))\n return \n\n # Open the file for reading\n try:\n with open(self.credentials, 'r') as file:\n # Load the JSON data from the file\n data = json.load(file)\n openai.organization = data['organization']\n openai.api_key = data['api_key'] \n\n except Exception as e:\n print('(multi_llm) error: could not load credentials {0} : {1}' .format(self.credentials,str(e)))\n return\n \n # do API call\n response = openai.ChatCompletion.create(\n model = self.model,\n messages=messages\n )\n if response:\n return(self.get_content(response))\n else:\n return response\n```\n</details>\n\n******\n\n## Config JSON\n\n### Using the `config.json` File\n\nThe `config.json` file offers a convenient way to configure and load multiple language models using the \"Multi_LLM\" framework. This section will guide you through the process of creating and utilizing a `config.json` file to load and use specific language models in your application.\n\n### Configuration Setup\n\nTo get started, follow these steps to configure your `config.json` file:\n\n1. **Create a Configuration File**: Create a new file named `config.json` in your project directory or the desired location.\n\n2. **Configure LLMs**: Define the language models you want to use within the `\"llms\"` array. Each model configuration includes details such as the Python file containing the model implementation, class name, model name, and credentials file path.\n\n```json\n{\n \"Config\": {\n \"Multi_LLM\": {\n \"llms\": [\n {\n \"file\": \"bard.py\",\n \"class_name\": \"BARD\",\n \"model\": \"chat-bison@001\",\n \"credentials\": \"/path/to/google/key.json\"\n },\n {\n \"file\": \"GPT.py\",\n \"class_name\": \"GPT\",\n \"model\": \"gpt-3.5-turbo\",\n \"credentials\": \"/path/to/openai/key.json\"\n }\n ]\n }\n }\n}\n```\n******\n\n### Loading Models Using MultiLLM\n\nAfter creating the `config.json` file, you can use the \"Multi_LLM\" framework to load and utilize the configured models in your application. Follow these steps:\n\n1. **Instantiate Multi_LLM**: Create an instance of the `Multi_LLM` class and provide the path to your `config.json` file.\n\n```python\nfrom Multi_LLM import Multi_LLM\n\n# Specify the path to your config.json file\nconfig_path = \"path/to/config.json\"\n\n# Instantiate Multi_LLM\nmulti_llm = Multi_LLM(config=config_path)\n```\n\n2. **Run Models**: Use the `run` method to run the loaded models and process their responses. Provide a prompt and an optional action chain if needed. The responses from each model will be returned in a dictionary.\n\n```python\nprompt = \"Translate this sentence.\"\naction_chain = None # You can define an action chain if required\n\nresponses = multi_llm.run(prompt, action_chain)\nprint(responses)\n```\n\n## Action Class\nThis is the interface class which we use to operate on the output while still in parallel. Action class instances are define by the user and can be chained indefinitely with other Actiion instances.\nThere are two methods that are used to interface with the class: `apply()` and `then()`.\nThe first, `apply()` is what is used internally to call the methods we register. Registration is as follows:\n```python\n# Write an interfacing function, a simple in/out\ndef capitalize(data):\n\tdata = data.upper()\n\treturn data\n\t\ndef lower(data):\n\tdata = data.lower()\n\treturn data\n\t\n# Create the respective objects\naction1 = Action(operation=capitalize)\naction2 = Action(operation=lower)\n\n# Create the chain of actions\naction_chain = action1.then(action2)\n```\nWhat we did was create some actions that are simple I/O operations on data. We can then chain them together using the `then()` method mentioned above. The order of operations for then is left to right, in this case action1 will go first then action2. These actions can be tailored to your own specification, since the actions are serial, you can specify the information that is going to be passed into the function and what information will be returned.\n\nAfter you have created the action chain you can pass this into the `MultiLLM.run()` method and run.\n\n### Action Chain\n\nThe `Action Chain` is a core component of processesing LLM output. In the provided [example.py](examples/example.py) we see the first action defined is for processing the LLM response and extracting the code alone from the response. This could be a first of many steps, what could follow could be saving the code to the file or extracting information through the `ast` module. The actions are meant to be ran on each of the model's outputs so they should be generalized for use.\n\n## Rank Class\nThe `Rank` class is identical in functionality to the `Action` class, though the use is different. While the `Action` class is used to accept and modify data in each of the models, the `Rank` class' purpose is to modify the final combined output of the LLMs. \nThis output is stored in a dictionary keyed by the model's names. The `Rank` class's methods should be written to take this data and operate on it. \n```python\n#Write an interfacing function, a simple print\ndef print_llms(dictionary):\n\tfor key,val in dictionary.items():\n\t\tprint(key, val)\n\n# Create Rank instances\nrank_object = Rank(operation=print_llms)\n\n# Running the LLM, assuming an instance of \n# MultiLLM named mLLM\nresults = mLLLM.run(prompt, action_chain, rank_object)\n```\nIn the above code we are doing very similar actions as those seen above. The methodoly is the same. In this case we only have a single object, so instead of creating a `rank_chain` we simply pass in our single instantiated object.\n## Architecture\n### High Level Overview\n\n### MultiLLM Class\nThis is the highest level, here we can instantiate MultiLLM objects using either config files (see [config files](###Config-Files)) or manually instantiating a custom or hosted LLM. \nIn this class we call multiple LLMs concurrently and then we can operate on the results of each in parallel using the [Action Class](###Action-Class).\n\nThe \"MultiLLM\" Python code provides a versatile framework for managing and orchestrating multiple language models (LLMs) within a single application. This code is designed to enhance the efficiency of working with language models by enabling concurrent execution, response processing, and model loading from configuration files. The key features and components of the code include:\n\n- **Concurrent Model Execution:** The code allows you to run multiple language models concurrently, facilitating efficient utilization of computational resources. This is particularly useful for scenarios that involve processing multiple prompts or tasks simultaneously.\n\n- **Action Chain Processing:** The framework supports the concept of \"action chains,\" which enable you to preprocess model responses using a sequence of predefined actions. This empowers you to refine and enhance the output generated by the language models.\n\n- **Model Loading from Configuration:** You can load LLMs from a configuration file. This JSON-based configuration includes essential details about each model, such as the model class, associated credentials, and file paths. The code can dynamically load these models, making it easy to add new models or modify existing ones without altering the core codebase.\n\n- **Redis Integration:** The code features optional integration with a Redis instance. It checks for a successful Redis connection and adjusts its behavior accordingly. If Redis connectivity is not established, the code gracefully handles the situation by setting a flag that reflects the absence of a Redis connection.\n\n- **Simplified Model Registration:** The code includes a straightforward method to register models within the framework. This allows for the inclusion of custom LLM implementations while ensuring that the models are appropriately organized and accessible for execution.\n\nBy using the \"MultiLLM\" Python code, developers can streamline their interactions with multiple language models, seamlessly integrating them into various applications or projects. The code promotes modularity, reusability, and parallelism in working with language models, ultimately enhancing the user experience and productivity.\n\n\n### BaseLLM\n\n`BaseLLM` is designed to serve as the basis for implementing various language model classes. This `BaseLLM` class encapsulates essential attributes and methods necessary for interfacing with language models. The code establishes a structured foundation for building specific language model implementations and harmonizes their interaction within a larger application context.\n\nKey features and components of the code include:\n\n- **Attributes for Language Models:** The `BaseLLM` class declares a set of attributes, such as `model`, `roles`, `messages`, `temp`, `api_key`, `max_tokens`, and `args`, that are pertinent to language models. These attributes are meant to be customized and adapted as needed for specific model implementations.\n\n- **Customizable Initialization:** The `__init__()` method in the `BaseLLM` class facilitates flexible instantiation of model instances by allowing the specification of custom values via keyword arguments. It enables the convenient configuration of attributes like `name`, `credentials`, `model`, and `class_name`.\n\n- **Placeholder Methods:** The `BaseLLM` class defines two placeholder methods: `get_response()` and `get_content()`. The `get_response()` method is designed to receive a `Prompt` object and run the associated language model with the provided prompt. The actual implementation of this method is expected to be customized in derived classes to perform the model-specific interactions. Similarly, the `get_content()` method is meant to be implemented by deriving classes, providing an interface to extract relevant content from model responses.\n\n- **Structured Framework:** The code encapsulates a structured framework that abstracts common functionalities of language models. By inheriting from the `BaseLLM` class, developers can focus on implementing model-specific interactions while leveraging the established structure for attribute handling and method placeholders.\n\n`BaseLLM` aims to streamline the development of specific language model implementations by providing a consistent structure and standardized attributes. Developers can extend this base class to create custom language model classes that seamlessly integrate into the broader application ecosystem. This modular approach promotes reusability, maintainability, and consistent design patterns when working with various language models.\n\n******\n\n## Contribution\n\nYou can contribute by extending the models located in [models](models). See the [BaseLLM](#baselmm) section for more information on the necessary mechanisms for extending the BaseLLM class.\nFor further information contact [ethansaurusrex](https://github.com/ethansaurusrex)\n",
"bugtrack_url": null,
"license": null,
"summary": "VerifAI MultiLLM: A module to invoke multiple LLMs concurrently and rank their results",
"version": "0.1019",
"project_urls": {
"Bug Tracker": "https://github.com/verifai/multiLLM",
"Homepage": "https://github.com/verifai/multiLLM"
},
"split_keywords": [],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "094cafd8bc323c776a505af8a2e4070f7814243053edb9d9322a8e71f46f2dd5",
"md5": "865edb4b40e609a35dffef7f5d6e0d6f",
"sha256": "4fb664bce9c9fa36a98a2a25077164ecd33aa6180a4a7da853e05d1c9fafe8f4"
},
"downloads": -1,
"filename": "multillm-0.1019-py3-none-any.whl",
"has_sig": false,
"md5_digest": "865edb4b40e609a35dffef7f5d6e0d6f",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 59274,
"upload_time": "2024-07-24T19:09:19",
"upload_time_iso_8601": "2024-07-24T19:09:19.292474Z",
"url": "https://files.pythonhosted.org/packages/09/4c/afd8bc323c776a505af8a2e4070f7814243053edb9d9322a8e71f46f2dd5/multillm-0.1019-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "0f5e3044c292ef9df9710d85344cb6c10d44e987ed522b191faebc6a44494e09",
"md5": "0efb3041c0793f9476daad9d906507ef",
"sha256": "2bc12b3f27d0d4d19f90e642c7c7d5ad35602245a9ee27bbfe415925857f20ac"
},
"downloads": -1,
"filename": "multillm-0.1019.tar.gz",
"has_sig": false,
"md5_digest": "0efb3041c0793f9476daad9d906507ef",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 42761,
"upload_time": "2024-07-24T19:09:22",
"upload_time_iso_8601": "2024-07-24T19:09:22.088555Z",
"url": "https://files.pythonhosted.org/packages/0f/5e/3044c292ef9df9710d85344cb6c10d44e987ed522b191faebc6a44494e09/multillm-0.1019.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2024-07-24 19:09:22",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "verifai",
"github_project": "multiLLM",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [
{
"name": "transformers",
"specs": []
},
{
"name": "redis",
"specs": []
},
{
"name": "openai",
"specs": []
},
{
"name": "vertexai",
"specs": []
},
{
"name": "google-cloud-aiplatform",
"specs": []
}
],
"lcname": "multillm"
}
JSON
