Manifest

How to make prompt programming with Foundation Models a little easier.

Table of Contents

• Install
• Getting Started
• Manifest
• Local HuggingFace Models
• Development

Install

Download the code:

git clone [email protected]:HazyResearch/manifest.git
cd manifest

Install:

pip install -e .

Dev Install:

make dev

Getting Started

Running is simple to get started. If using OpenAI, set export OPENAI_API_KEY=<OPENAIKEY> then run

from manifest import Manifest

# Start a manifest session to OpenAI - default `engine=text-davinci-002`
manifest = Manifest(
    client_name = "openai",
)
manifest.run("Why is the grass green?")

Manifest Components

Manifest is meant to be a very light weight package to help with prompt design and iteration. Three key design decisions of Manifest are

• Prompt are functional -- they can take an input example and dynamically change
• All models are behind APIs
• Supports caching of model inputs/outputs for iteration, reproducibility, and cost saving

Models

Manifest provides model clients for OpenAI, AI21, OPT (assuming model is loaded locally), and HuggingFace (see below for how to use locally hosted HuggingFace models). You can toggle between the models by changing client_name and client_connection. For example, if a HuggingFace model is loaded locally, run

manifest = Manifest(
    client_name = "huggingface",
    client_connection = "http://127.0.0.1:5000",
)

You can see the model details and possible model inputs to run() via

print(manifest.client.get_model_params())
print(manifest.client.get_model_inputs())

Prompts

A Manifest prompt is a function that accepts a single input to generate a string prompt to send to a model.

from manifest import Prompt
prompt = Prompt(lambda x: f"Hello, my name is {x}")
print(prompt("Laurel"))
>>> "Hello, my name is Laurel"

Running

result = manifest.run(prompt, "Laurel")

will send ``Hello, my name is Laurel'' to the model.

As you saw above, if you don't want your prompt to change, we also support static strings

result = manifest.run("Hello, my name is static")

Global Cache

We support having queries and results stored in a global cache that can be shared across users. We treat inputs and outputs as key value pairs and support SQLite or Redis backends. To start with global caching using SQLite, run

manifest = Manifest(
    client_name = "openai",
    cache_name = "sqlite",
    cache_connection = "mycache.sqlite",
)

The cache will be saved in mycache.sqlite.

We also support Redis backend.

manifest = Manifest(
    client_name = "openai",
    cache_name = "redis",
    cache_connection = "localhost:6379"
)

As a hint, if you want to get Redis running, see the docker run command below under development.

Sessions

Each Manifest run supports a session that, in addition to a global cache, connects to a local SQLite DB to store user query history.

manifest = Manifest(
    client_name = "openai",
    cache_name = "sqlite",
    cache_connection = "mycache.sqlite",
    session_id = "grass_color",
)

will start a Manifest session with the session name grass_color. This can be helpful for a user to logically keep track of sessions, see interaction history, and resume sessions if desired. If the session id provided is _default, we generate a random id for the user.

After a few queries, the user can explore their history

manifest.get_last_queries(4)

will retrieve the last 4 model queries and responses.

Running Queries

Once you have a session open, you can write and develop prompts.

prompt = Prompt(lambda x: "Hello, my name is {x}")
result = manifest.run(prompt, "Laurel")

You can also run over multiple examples.

results = manifest.batch_run(prompt, ["Laurel", "Avanika"])

If something doesn't go right, you can also ask to get a raw manifest Response.

result_object = manifest.batch_run(prompt, ["Laurel", "Avanika"], return_response=True)
print(result_object.get_request())
print(result_object.is_cached())
print(result_object.get_json_response())

By default, we do not truncate results based on a stop token. You can change this by either passing a new stop token to a Manifest session or to a run or batch_run.

result = manifest.run(prompt, "Laurel", stop_token="and")

If you want to change default parameters to a model, we pass those as kwargs to the client.

result = manifest.run(prompt, "Laurel", max_tokens=50)

Local Huggingface Models

To use a HuggingFace generative model, in manifest/api we have a Falsk application that hosts the models for you.

In a separate terminal or Tmux/Screen session, to load 6B parameters models, run

python3 manifest/api/app.py \
    --model_type huggingface \
    --model_name_or_path EleutherAI/gpt-j-6B \
    --device 0

You will see the Flask session start and output a URL http://127.0.0.1:5000. Pass this in to Manifest. If you want to use a different port, set the FLASK_PORT environment variable.

manifest = Manifest(
    client_name = "huggingface",
    client_connection = "http://127.0.0.1:5000",
)

If you have a custom model you trained, pass the model path to --model_name_or_path.

To help load larger models, we also support using parallelize() from HF, accelerate, and bitsandbytes. You will need to install these packages first. We list the commands to load larger models below.

• T0pp

python3 manifest/api/app.py \
    --model_type huggingface \
    --model_name_or_path bigscience/T0pp \
    --use_hf_parallelize

• NeoX 20B (requires at least 60GB of GPU memory)

python3 manifest/api/app.py \
    --model_type huggingface \
    --model_name_or_path EleutherAI/gpt-neox-20b \
    --use_accelerate_multigpu \
    --percent_max_gpu_mem_reduction 0.75

• Boom 175B (requires at least 240GB of GPU memory)

python3 manifest/api/app.py \
    --model_type huggingface \
    --model_name_or_path bigscience/bloom \
    --use_bitsandbytes \
    --percent_max_gpu_mem_reduction 0.85

Development

Before submitting a PR, run

export REDIS_PORT="6380"  # or whatever PORT local redis is running for those tests
cd <REDIS_PATH>
docker run -d -p 127.0.0.1:${REDIS_PORT}:6379 -v `pwd`:`pwd` -w `pwd` --name manifest_redis_test redis
make test