Issues

• [DEV] Priority-aware request scheduling

  InfiniTensor/InfiniLM#570 · Filter by author xiaoba17 opened on Sep 13, 2026

• [BUG] 昇腾平台图模式开启输出异常

  InfiniTensor/InfiniLM#525 · Filter by author Jingbo-gao opened on Aug 4, 2026

• [DEV] 推理服务添加chunked prefill支持

  InfiniTensor/InfiniLM#520 · Filter by author ma-hang opened on Jul 31, 2026

• [DEV] feat: support partial cache hits for multimodal visual embeddings

  InfiniTensor/InfiniLM#477 · Filter by author ma-hang opened on Jul 6, 2026

• [BUG] Model loading logic fails to filter weight shards correctly for models with non-standard naming (e.g., Mistral-Large-Instruct)

  InfiniTensor/InfiniLM#467 · Filter by author rubik-hua opened on Jun 30, 2026

• [DEV] Support Qwen3.5

  InfiniTensor/InfiniLM#445 · Filter by author PanZezhong1725 opened on Jun 18, 2026

• [DEV] Add MoE inference and expert parallel support

  InfiniTensor/InfiniLM#443 · Filter by author qinyiqun opened on Jun 18, 2026

• [DEV]feat: support AWQ/GPTQ Marlin w4a16 quantization

  InfiniTensor/InfiniLM#435 · Filter by author qinyiqun opened on Jun 16, 2026

• [DEV] support qwen3 vl

  InfiniTensor/InfiniLM#434 · Filter by author wooway777 opened on Jun 15, 2026

• [BUG] fix ascend support

  InfiniTensor/InfiniLM#431 · Filter by author wooway777 opened on Jun 12, 2026

• [DEV] InfiniLM warmup 逻辑完善

  InfiniTensor/InfiniLM#429 · Filter by author spike-zhu opened on Jun 11, 2026

• python 3.10 doesn't have override in typings

  InfiniTensor/InfiniLM#420 · Filter by author wooway777 opened on Jun 9, 2026

• [DEV] 释放GIL锁后，优化推理过程中显存占用

  InfiniTensor/InfiniLM#407 · Filter by author pengcheng888 opened on Jun 3, 2026

• [DEV] InfiniLM 支持 GPT2

  InfiniTensor/InfiniLM#406 · Filter by author spike-zhu opened on Jun 1, 2026

• [Refactor] Decouple RoPE Scaling via Factory Pattern and Introduce TP-Safe RoPE Cache

  InfiniTensor/InfiniLM#401 · Filter by author rubik-hua opened on May 28, 2026

• [BUG] Missing spaces between English words during streaming inference with Mistral/internlm3 models

  InfiniTensor/InfiniLM#398 · Filter by author rubik-hua opened on May 25, 2026

• [DEV] InfiniLM 接入摩尔 flash-attn

  InfiniTensor/InfiniLM#394 · Filter by author spike-zhu opened on May 21, 2026

• [Feature]Implement Llama 3.1 Frequency-Aware RoPE Scaling

  InfiniTensor/InfiniLM#392 · Filter by author rubik-hua opened on May 20, 2026

• [DEV] build with cmake

  InfiniTensor/InfiniLM#390 · Filter by author wooway777 opened on May 20, 2026

• [DEV] MiniCPM-V服务支持连续批次

  InfiniTensor/InfiniLM#369 · Filter by author PanZezhong1725 opened on May 12, 2026

• refactor: use processor in infer backup

  InfiniTensor/InfiniLM#367 · Filter by author wooway777 opened on May 12, 2026

• [DEV] 多模态模型前缀匹配基建

  InfiniTensor/InfiniLM#344 · Filter by author PanZezhong1725 opened on May 6, 2026

• [DEV] 支持PD分离

  InfiniTensor/InfiniLM#340 · Filter by author ma-hang opened on Apr 30, 2026

• [DEV] 模型推理参数预处理weight_preprocess及gptq_qy的示例

  InfiniTensor/InfiniLM#336 · Filter by author qinyiqun opened on Apr 29, 2026

• [DEV] 服务部分添加connector

  InfiniTensor/InfiniLM#323 · Filter by author ma-hang opened on Apr 24, 2026