pipeline-ai 0.5.0b18

Pipelines for machine learning workloads.

Pipeline

powered by mystic

Table of Contents

• About
• Version roadmap
  • v0.4.0
  • v0.5.0
  • Beyond
• Quickstart
  • Basic maths
  • Transformers (GPT-Neo 125M)
• Installation instructions
  • Linux, Mac (intel)
  • Mac (arm/M1)
• Development
• License

About

Pipeline is a python library that provides a simple way to construct computational graphs for AI/ML. The library is suitable for both development and production environments supporting inference and training/finetuning. This library is also a direct interface to Pipeline.ai which provides a compute engine to run pipelines at scale and on enterprise GPUs.

The syntax used for defining AI/ML pipelines shares some similarities in syntax to sessions in Tensorflow v1, and Flows found in Prefect. In future releases we will be moving away from this syntax to a C based graph compiler which interprets python directly (and other languages) allowing users of the API to compose graphs in a more native way to the chosen language.

Version roadmap

v0.4.0 (Jan 2023)

• Custom environments on PipelineCloud (remote compute services)
• Kwarg inputs to runs
• Extended IO inputs to pipeline_function objects

v0.5.0 (Jan/Feb 2023)

• Pipeline chaining
• if statements & while/for loops

Beyond

• Run log streaming
• Run progress tracking
• Resource dedication
• Pipeline scecific remote load balancer (10% of traffic to one pipeline 80% to another)
• Usage capping
• Run result streaming
• Progromatic autoscaling
• Alerts
• Events
• Different python versions on remote compute services

Quickstart

:warning: Uploading pipelines to Pipeline Cloud works best in Python 3.9. We strongly recommend you use Python 3.9 when uploading pipelines because the pipeline-ai library is still in beta and is known to cause opaque errors when pipelines are serialised from a non-3.9 environment.

Basic maths

from pipeline import Pipeline, Variable, pipeline_function


@pipeline_function
def square(a: float) -> float:
    return a**2

@pipeline_function
def multiply(a: float, b: float) -> float:
    return a * b

with Pipeline("maths") as pipeline:
    flt_1 = Variable(type_class=float, is_input=True)
    flt_2 = Variable(type_class=float, is_input=True)
    pipeline.add_variables(flt_1, flt_2)

    sq_1 = square(flt_1)
    res_1 = multiply(flt_2, sq_1)
    pipeline.output(res_1)

output_pipeline = Pipeline.get_pipeline("maths")
print(output_pipeline.run(5.0, 6.0))

Transformers (GPT-Neo 125M)

Note: requires torch and transformers as dependencies.

from pipeline import Pipeline, Variable
from pipeline.objects.huggingface.TransformersModelForCausalLM import (
    TransformersModelForCausalLM,
)

with Pipeline("hf-pipeline") as builder:
    input_str = Variable(str, is_input=True)
    model_kwargs = Variable(dict, is_input=True)

    builder.add_variables(input_str, model_kwargs)

    hf_model = TransformersModelForCausalLM(
        model_path="EleutherAI/gpt-neo-125M",
        tokenizer_path="EleutherAI/gpt-neo-125M",
    )
    hf_model.load()
    output_str = hf_model.predict(input_str, model_kwargs)

    builder.output(output_str)

output_pipeline = Pipeline.get_pipeline("hf-pipeline")

print(
    output_pipeline.run(
        "Hello my name is", {"min_length": 100, "max_length": 150, "temperature": 0.5}
    )
)

Installation instructions

Linux, Mac (intel)

pip install -U pipeline-ai

Mac (arm/M1)

Due to the ARM architecture of the M1 core it is necessary to take additional steps to install Pipeline, mostly due to the transformers library. We recoomend running inside of a conda environment as shown below.

1. Make sure Rosetta2 is disabled.
2. From terminal run:

xcode-select --install

3. Install Miniforge, instructions here: https://github.com/conda-forge/miniforge or follow the below:
   1. Download the Miniforge install script here: https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-MacOSX-arm64.sh
   2. Make the shell executable and run

   sudo chmod 775 Miniforge3-MacOSX-arm64.sh
   ./Miniforge3-MacOSX-arm64.sh
   

4. Create a conda based virtual env and activate:

conda create --name pipeline-env python=3.9
conda activate pipeline-env

5. Install tensorflow

conda install -c apple tensorflow-deps
python -m pip install -U pip
python -m pip install -U tensorflow-macos
python -m pip install -U tensorflow-metal

6. Install transformers

conda install -c huggingface transformers -y

7. Install pipeline

python -m pip install -U pipeline-ai

Development

This project is made with poetry, so firstly setup poetry on your machine.

Once that is done, please run

sh setup.sh

With this you should be good to go. This sets up dependencies, pre-commit hooks and pre-push hooks.

You can manually run pre commit hooks with

pre-commit run --all-files

To run tests manually please run

pytest

License

Pipeline is licensed under Apache Software License Version 2.0.