MegEngine/InferLLM

Issues

• SG2042 服务器 make 出错

  #85 opened 3 months ago by JocelynPanPan
  0

• 无法加载模型bin或者guff模型

  #84 opened 5 months ago by velonica0
  0

• 有打算支持qwen吗

  #83 opened 7 months ago by Gooddz1
  0

• 在树莓派3b+上部署，出现无法打开模型的问题

  #82 opened 8 months ago by Dominic23331
  4

• How to build wasm file?

  #78 opened 8 months ago by njalan
  0

• 这个有windows的部署教程吗

  #80 opened 9 months ago by ASHLEYDX
  0

• 在运行llama2-13b的时候出现以下问题

  #77 opened 10 months ago by Wowoho
  0

• Linux 运行时报以下错误

  #28 opened 2 years ago by guangyuanyu
  12

• 移植问题

  #76 opened a year ago by yinghuo302
  0

• 请问可以在不支持V扩展的RISC-V CPU上运行吗

  #75 opened a year ago by TheKiteRunner24
  1

• arm 平台输出乱码

  #71 opened a year ago by tanjatang
  3

• 可以在RV64指令集的CPU上运行吗

  #74 opened a year ago by TheKiteRunner24
  0

• I got the error on centos 7: failed to tokenize string!

  #73 opened a year ago by Mignet
  2

• make报错

  #32 opened 2 years ago by hoyin2013
  3

• chatglm3有计划支持吗？

  #72 opened a year ago by sunzhe09
  3

• 感觉回答有些错乱，用的是macbook pro 推理chinese-alpaca-7b-q4

  #15 opened 2 years ago by realcarlos
  5

• mac os Big Sur 11.7.4 Linking Error , Undefined symbols

  #70 opened a year ago by junlong
  1

• 请问是否有计划支持Whisper？

  #69 opened a year ago by nxprxf
  1

• ChatGLM2 效果异常

  #66 opened a year ago by adf1178
  1

• unsupported relocation 37 on musl libc

  #58 opened a year ago by xhebox
  3

• chatglm2 GPU版本的int4、int8量化模型预测结果异常

  #67 opened a year ago by itfish1990
  1

• 能否改为GPU辅助计算

  #65 opened a year ago by bzy-080408
  6

• windows下面编译失败

  #17 opened 2 years ago by kavern1128
  8

• Thread wakening may be bottom neck for large core systems

  #61 opened a year ago by xhebox
  8

• 【new feature】通义千问有没有计划支持

  #63 opened a year ago by guangyuanyu
  2

• 请问一下，这个是不是比python的性能更好？

  #56 opened a year ago by wuent
  1

• 大模型推理中这个推理引擎如何支持 lora，ptuning等私有语料训练插件后的新模型

  #53 opened a year ago by leoluopy
  2

• make error

  #54 opened a year ago by forwiat
  13

• O3 optimization are slower on SG2042

  #60 opened a year ago by xhebox
  3

• Please support RWKV for refs and compare.

  #59 opened a year ago by xiaol
  0

• 目前不支持gpu跑吗？我看代码中有添加对gpu的支持啊

  #50 opened 2 years ago by sf9218
  2

• 请问主循环中的token指代的是什么呢？函数体内部也没有看出来是代表什么

  #51 opened a year ago by wangxu569
  1

• ChatGLM 2已经出了，什么时候可以支持一下啊

  #42 opened 2 years ago by kungrainbow
  10

• 编译错误，需要添加 -mfma 才能编译通过，但是 CPU 指令集不支持 fma，导致运行出错。

  #33 opened 2 years ago by huixiao8973
  5

• chatglm-6b下的模型格式不正确

  #41 opened 2 years ago by wildkid1024
  1

• Compile error: void inferllm::BaiChuanGraph::constuct_llm()? marked ?override?, but does not override

  #43 opened 2 years ago by cjia4
  4

• 希望可以封装openai兼容API

  #47 opened 2 years ago by nexx5
  0

• 【feature】baichuan-7b模型能不能使用baichuan-vicuna-chinese-7b模型文件

  #44 opened 2 years ago by guangyuanyu
  0

• 没有支持最新的llama.cpp的格式吗

  #38 opened 2 years ago by AceyKubbo
  0

• asserts 那个目录，建议改成 assets

  #35 opened 2 years ago by csukuangfj
  1

• 在线程数>1时，会占满cpu核心

  #34 opened 2 years ago by Yricky
  3

• decode & decode_iter 多线程会不会有问题

  #31 opened 2 years ago by guangyuanyu
  0

• 向量计算中使用CPU AVX指令，能否支持不使用AVX指令的版本

  #30 opened 2 years ago by kungrainbow
  1

• 请问如何实现的量化？

  #26 opened 2 years ago by lwo2002
  2

• 更新的太慢了，求加速

  #18 opened 2 years ago by wfan1203
  2

• isnan报错

  #24 opened 2 years ago by AugF
  1

• Support input prompt like llama.cpp

  #23 opened 2 years ago by donge
  0

• 有实现思维树方式么

  #21 opened 2 years ago by cery999
  0

• tokenizer 在哪里下载

  #19 opened 2 years ago by ljiuyuan1987
  2

• 希望能编译成so文件

  #16 opened 2 years ago by realcarlos
  3