Ginkgo's AI model API client

Work in progress: this repo was just made public and we are still working on integration

A python client for Ginkgo's AI model API, to run inference on public and Ginkgo-proprietary models. Learn more in the Model API announcement.

Prerequisites

Register at https://models.ginkgobioworks.ai/ to get credits and an API KEY (of the form xxxxxxx-xxxx-xxxx-xxxx-xxxxxxxx). Store the API KEY in the GINKGOAI_API_KEY environment variable.

Installation

Install the python client with pip:

pip install ginkgo-ai-client

Usage:

Note: This is an alpha version of the client and its interface may vary in the future.

Example : masked inference with Ginkgo's AA0 model

The client requires an API key (and defaults to os.environ.get("GINKGOAI_API_KEY") if none is explicitly provided)

from ginkgo_ai_client import GinkgoAIClient, MaskedInferenceQuery

client = GinkgoAIClient()
model = "ginkgo-aa0-650M"

# SINGLE QUERY

query = MaskedInferenceQuery(sequence="MPK<mask><mask>RRL", model=model)
prediction = client.send_request(query)
# prediction.sequence == "MPKRRRRL"

# BATCH QUERY

sequences = ["MPK<mask><mask>RRL", "M<mask>RL", "MLLM<mask><mask>R"]
queries = [MaskedInferenceQuery(sequence=seq, model=model) for seq in sequences]
predictions = client.send_batch_request(queries)
# predictions[0].sequence == "MPKRRRRL"

Changing the model parameter to esm2-650M or esm2-3b in this example will perform masked inference with the ESM2 model.

Example : embedding computation with Ginkgo's 3'UTR language model

from ginkgo_ai_client import GinkgoAIClient, MeanEmbeddingQuery

client = GinkgoAIClient()
model = "ginkgo-maskedlm-3utr-v1"

# SINGLE QUERY

query = MeanEmbeddingQuery(sequence="ATTGCG", model=model)
prediction = client.send_request(query)
# prediction.embedding == [1.05, -2.34, ...]

# BATCH QUERY

sequences = ["ATTGCG", "CAATGC", "GCGCACATGT"]
queries = [MeanEmbeddingQuery(sequence=seq, model=model) for seq in sequences]
predictions = client.send_batch_request(queries)
# predictions[0].embedding == [1.05, -2.34, ...]

Available models

See the example folder and reference docs for more details on usage and parameters.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Releases

Make sure the changelog is up to date, increment the version in pyproject.toml, create a new tag, then create a release on Github (publication to PyPI is automated).