Metadata-Version: 2.4
Name: polarkv
Version: 1.1.0
Summary: PolarKV distributed KVCache Pool backend for LLM inference
Author: PolarKV Team, Alibaba Cloud
License: Apache-2.0
Project-URL: Homepage, https://help.aliyun.com/zh/polardb/polardb-for-mysql/user-guide/polarkvcache-inference-acceleration/
Requires-Python: >=3.10
Description-Content-Type: text/markdown

# PolarKV

Distributed KVCache Pool backend for LLM inference. Supports **SGLang** and **vLLM**.

## Installation

```bash
pip install polarkv
```

## Build From Source

Build the native SDKs from the repository root first, then assemble the wheel here:

```bash
cd ..
./build.sh client

cd python
./build_whl.sh
pip install dist/polarkv-*.whl
```

## Usage

### SGLang

```bash
--hicache-storage-backend dynamic \
--hicache-storage-backend-extra-config '{"backend_name": "polarkv", \
    "module_path": "polarkv.adapters.sglang", \
    "class_name": "PolarKVCache", \
    "polarkv_config_file":"/path/to/config.yaml", \
    "interface_v1":1 \
}'
```

### vLLM

```bash
--kv-transfer-config '{"kv_connector": "PolarKVCache", \
    "kv_role": "kv_both", \
    "kv_connector_module_path": "polarkv.adapters.vllm", \
    "kv_connector_extra_config": {"config_file":"/path/to/config.yaml"} \
}'
```
