Generate per-session LoRA adapters in <1s for inference tasks

Facingsouth1 pts0 comments

tessera-hypernetwork · PyPI

Skip to main content<br>Switch to mobile version

Warning

You are using an unsupported browser, upgrade to a newer version.

Warning

Some features may not work without JavaScript. Please try enabling it if you encounter problems.

Search PyPI

Search

tessera-hypernetwork 1.0.10

pip install tessera-hypernetwork

Copy PIP instructions

Latest release

Released:<br>Jun 15, 2026

Generate per-session LoRA adapters for inference tasks using hypernetwork synthesis

Navigation

Verified details

These details have been verified by PyPI<br>Maintainers

Terradev

Unverified details

These details have not been verified by PyPI<br>Project links

Homepage

Issues

Repository

Meta

License Expression: Apache-2.0

SPDX<br>License Expression

Author: Tessera Contributors

Tags

lora

peft

hypernetwork

machine-learning

nlp

Requires: Python >=3.8

Provides-Extra:<br>dev

Classifiers

Development Status

3 - Alpha

Intended Audience

Developers

Science/Research

Programming Language

Python :: 3

Python :: 3.8

Python :: 3.9

Python :: 3.10

Python :: 3.11

Topic

Scientific/Engineering :: Artificial Intelligence

Report project as malware

Project description

Tessera Hypernetwork

Generate per-session LoRA adapters for inference tasks using hypernetwork synthesis.

Features

Metadata-to-LoRA: Generate adapters from structured user metadata (JSON)<br>Text-to-LoRA: Generate adapters from natural language descriptions<br>Doc-to-LoRA with SHINE: Generate adapters from document content using SHINE (ICML 2026) for long-context internalization<br>FastAPI: Modern async Python web framework<br>OpenAI-compatible API: Easy integration with existing tooling<br>Installation<br>pip install tessera-hypernetwork<br>Quick Start

Commands

Generate:

From metadata (JSON string or file)

tessera generate

--from-metadata '{"task": "classification", "domain": "medical"}'

--base-model mistralai/Mistral-7B-Instruct-v0.2

--rank 16

--save ./adapter.safetensors

From text description:

tessera generate

--from-text "Medical diagnosis assistant"

--base-model mistralai/Mistral-7B-Instruct-v0.2

--rank 16

--save ./adapter.safetensors

From document:

tessera generate

--from-doc ./document.txt

--base-model mistralai/Mistral-7B-Instruct-v0.2

--rank 16

--save ./adapter.safetensors

'Generate' Sub-Commands:

--from-metadata: JSON metadata string or file path

--from-text: Natural language description

--from-doc: Document content or file path

--base-model: Base model identifier (default: mistralai/Mistral-7B-Instruct-v0.2)

--rank: LoRA rank (default: 16)

--save: Output path for safetensors file (required)

--mode: Generation mode: doc, metadata, or text (auto-inferred if not specified)

Serve

Start the hypernetwork server:

tessera serve --port 8000 --host 0.0.0.0<br>Options:

--port: Port to serve on (default: 8000)

--host: Host to bind to (default: 0.0.0.0)

--qdrant-url: Qdrant vector database URL (optional)

--workers: Number of worker processes (default: 1)

Health

Check server health status:

tessera health --url http://localhost:8000<br>Options:

--url: Server URL (default: http://localhost:8000)

List:

List available base models and their dimensions:

tessera list

LoRAX Adapter Management

Import, list, and unload adapters:

Import an adapter:

tessera lorax import-adapter

--path ./adapter.safetensors

--name my-adapter

--base-model mistralai/Mistral-7B-Instruct-v0.2

--server-url http://localhost:8000

List loaded adapters:

tessera lorax list-adapters --server-url http://localhost:8000

Unload an adapter:

tessera lorax unload --name my-adapter --server-url http://localhost:8000

API Endpoints

The hypernetwork service provides a FastAPI server with the following endpoints:

POST /v1/generate - Generate a LoRA adapter for a given prompt

GET /health - Health check endpoint

POST /v1/adapters - Import adapter safetensors

GET /v1/adapters - List loaded adapters

DELETE /v1/adapters/{name} - Unload adapter \

License

Apache-2.0

Project details

Verified details

These details have been verified by PyPI<br>Maintainers

Terradev

Unverified details

These details have not been verified by PyPI<br>Project links

Homepage

Issues

Repository

Meta

License Expression: Apache-2.0

SPDX<br>License Expression

Author: Tessera Contributors

Tags

lora

peft

hypernetwork

machine-learning

nlp

Requires: Python >=3.8

Provides-Extra:<br>dev

Classifiers

Development Status

3 - Alpha

Intended Audience

Developers

Science/Research

Programming Language

Python :: 3

Python :: 3.8

Python :: 3.9

Python :: 3.10

Python :: 3.11

Topic

Scientific/Engineering :: Artificial Intelligence

Release history

Release notifications |<br>RSS feed

This version

1.0.10

Jun 15, 2026

1.0.9

Jun 15, 2026

1.0.8

Jun 15, 2026

1.0.7

Jun 15, 2026

1.0.6

Jun 15, 2026

1.0.5

Jun 15, 2026

1.0.4

Jun 15, 2026

1.0.3

Jun 15, 2026

1.0.2

Jun 15, 2026

1.0.1

Jun 15, 2026

1.0.0

Jun 15, 2026

0.2.20

Jun 15, 2026

0.2.19

Jun 15, 2026

0.2.18

Jun 15,...

tessera generate adapters python adapter from

Related Articles