Skip to main content

Zylora

zylora

Serverless GPU functions — deploy ML models with a decorator, scale to zero, pay per GPU-second.

PyPI License: MIT Python 3.10+

Quickstart

pip install zylora

Define a GPU function

import zylora

@zylora.fn(gpu="H100")
def embed(text: str) -> list[float]:
    from transformers import AutoModel
    model = AutoModel.from_pretrained("BAAI/bge-large-en-v1.5")
    return model.encode(text).tolist()

Deploy

zy deploy

Invoke

# Local (development)
result = embed("hello world")

# Remote (deployed)
result = embed.remote("hello world")

# Batch
results = embed.map(["hello", "world", "foo"])

# Streaming (for LLMs)
for token in generate.stream("Tell me a joke"):
    print(token, end="", flush=True)

# Async
job = embed.remote_async("hello")
result = job.result()

Client API (without decorator)

from zylora import Zylora

zy = Zylora()
result = zy.invoke("embed", {"text": "hello"})

Authentication

Set your API key via environment variable:

export ZYLORA_API_KEY="zy_live_..."

Or use the CLI:

zy login

Configuration

@zylora.fn() options

Parameter Type Default Description
gpu str required GPU type: t4, l4, rtx4090, a100_40gb, a100_80gb, l40s, h100, h200, b200, mi300x
name str function name Custom function name
packages list[str] [] pip packages to install
model str None Model to download during build
timeout int 300 Execution timeout (seconds)
min_instances int 0 Minimum warm instances
max_instances int 10 Maximum concurrent instances
concurrency int 1 Requests per instance
image str None Custom base Docker image
runtime str "python312" Python runtime version
routing str "cost_optimized" Routing strategy
visibility str "private" Function visibility

Documentation

Full documentation: docs.zylora.dev

License

MIT

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

zylora-0.1.0.tar.gz (24.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

zylora-0.1.0-py3-none-any.whl (15.2 kB view details)

Uploaded Python 3

File details

Details for the file zylora-0.1.0.tar.gz.

File metadata

  • Download URL: zylora-0.1.0.tar.gz
  • Upload date:
  • Size: 24.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for zylora-0.1.0.tar.gz
Algorithm Hash digest
SHA256 3efc0c9b9357403f8fcc4dd9694676e2c80e7996d69fa3ea1c8b3613adced601
MD5 e9753e552bce7e818e2c23652c0e07a3
BLAKE2b-256 25edf31bafbffef4930d8a6c2a87df0b924cd83a544c274279f02980ae6e7b52

See more details on using hashes here.

Provenance

The following attestation bundles were made for zylora-0.1.0.tar.gz:

Publisher: sdk-publish-python.yml on zylora-dev/zylora-python

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file zylora-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: zylora-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 15.2 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for zylora-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 3c0ee1f79eb1abb5165dc31f99b970f153798fdb5e201bbf2aa8e6ee62748cfd
MD5 4da9030dbe58e740832e6cc062138ef9
BLAKE2b-256 f836b09b17999cc4f4f7518cd5520b20ee27c963a2a935768fe264447a694a72

See more details on using hashes here.

Provenance

The following attestation bundles were made for zylora-0.1.0-py3-none-any.whl:

Publisher: sdk-publish-python.yml on zylora-dev/zylora-python

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

This release

0.1.0 This release

2 files

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page