Alkemet News
KVarN: Native vLLM backend for KV-cache quantization by Huawei (github.com)

66 points

bytheanonymousone

3 hours ago |

7 comments