GitLab

use marlin for lm_head, lm_head only calc last token for prefill
extend context window to 19K for DeepSeek-V3/R1 within 24GB VRAM

fix server and add prefix cache for server

Modify and add any incorrect or missing content in the `install.md`

🔖 release v0.2.1.post1

Fix cmake error caused by lack of environment variables in Windows environment

Update README_ZH.md

Fix precision mla

Add files via upload

Fix typo in DeepseekR1_V3_tutorial.md Line 174

Fix `GLIBCXX_3.4.30' not found in Anaconda

Merge branch 'fix_precision_MLA' of https://github.com/kvcache-ai/ktransformers into server-prefix-cache

Merge branch 'fix_precision_MLA' of https://github.com/kvcache-ai/ktransformers into server-prefix-cache

Update install.md

Update SUMMARY.md