sophgo/LLM-TPU

Issues

• bm1684x的芯片是否在seq_length增加时，推理时间就会变得很长的问题？

  #55 opened 4 days ago by iwantofun
  0

• FTL非常高，这属于正常吗？

  #54 opened 6 days ago by iwantofun
  0

• 请问后续有支持InternVl 2-8B的计划嘛？

  #53 opened 6 days ago by Li-Lai
  0

• InternVL2的C++代码疑问？

  #52 opened 9 days ago by w8501
  1

• Llama2 demo下make出错

  #51 opened 13 days ago by lishaung99
  0

• How to support qwen2.5-7b-instruction ?

  #49 opened a month ago by wuhuzi
  1

• IMPORTANT: LICENSE file required

  #46 opened 21 days ago by lazyparser
  1

• 有商业需求，请问一下各模型的性能怎么样。

  #44 opened 22 days ago by Arthassssss
  1

• ImportError: dynamic module does not define module export function

  #33 opened 22 days ago by muzi0111
  2

• Qwen2可以转onnx,转bmodel的时候出现以下问题

  #30 opened 22 days ago by tzhang2014
  2

• 运行模型出错

  #23 opened 22 days ago by wzf2020
  2

• ChatGLM3的web demo无法运行成功

  #18 opened 22 days ago by S0uLHun43r
  2

• 导出onnx出现warning

  #16 opened 22 days ago by githubzjj1
  2

• Qwen1.5 1b8和Qwen2 7b推理到最后出现重复性回答

  #35 opened 22 days ago by loredunk
  3

• 基于LLAVA的多模态大模型也是主流模型，请问有支持计划吗？

  #47 opened a month ago by Li-Lai
  1

• glm4v 啥时候提供已转好的bmodel下载链接呢

  #48 opened a month ago by tang799319844
  1

• 导出MiniCPM-V-2_6的ONNX模型出错

  #45 opened a month ago by thunder95
  3

• 请问后续有支撑qwen2-vl的计划嘛？

  #43 opened 2 months ago by kong1414
  1

• 常见问题的Q11，[a53lite runtimellerror] get function send api error, ret2，有没有其他的排错思路？

  #41 opened 2 months ago by loredunk
  1

• 请问支持Langchain-Chatchat吗？

  #40 opened 3 months ago by PeiwenWu
  2

• Support for Llama 3.1 model

  #36 opened 3 months ago by sbmandava
  2

• 在跑github下载已经转好的qwen-vl-chat-combine.bmodel模型时，会提示内存不足

  #37 opened 3 months ago by xuyang1102
  3

• 请问有支持iFlytekSpark的模型计划吗

  #38 opened 3 months ago by tzhang2014
  2

• llma3 is not available after conversion

  #13 opened 3 months ago by Bao0ne
  3

• 问题太长导致回复到一半就终止了

  #12 opened 3 months ago by githubzjj1
  2

• ChatGLM3-6B转onnx报错：torch.onnx.errors.CheckerError: The model does not have an ir_version set properly.

  #11 opened 3 months ago by kurosakiharachan
  1

• 运行llama3输出为乱序

  #29 opened 3 months ago by jayzou3773
  6

• chat.cpp:141: void Qwen::init(const std::vector<int>&, std::string): Assertion `true == ret' failed.

  #34 opened 3 months ago by loredunk
  0

• 标准提问格式，请大家按照这个方式进行提问~（重要）（非常重要）

  #31 opened 4 months ago by chuxiaoyi2023
  0

• It is the Llama3 compile part support Llama3-chinese?

  #15 opened 4 months ago by zifeng-radxa
  1

• Waiting for Qwen2 gradio web demo

  #27 opened 4 months ago by zifeng-radxa
  0

• Qwen2-7B-Instruct 导出 onnx 报错

  #25 opened 4 months ago by zifeng-radxa
  1

• 请问我想将列表里没有的大模型转成bmodel,应该怎么做?

  #24 opened 4 months ago by xinyinan9527
  1

• 转换qwen1.5出现的问题

  #3 opened 8 months ago by yuyun2000
  17

• Llama3 web_demo code is too old and output without end

  #21 opened 6 months ago by zifeng-radxa
  0

• Llama3 pipeline output � error decode

  #19 opened 6 months ago by zifeng-radxa
  0

• 万人血书MiniCPM-2B！

  #1 opened 8 months ago by xiabo0816
  2

• Web client not working

  #10 opened 7 months ago by Bao0ne
  2

• Unable to run llama2-7b according to readme

  #7 opened 7 months ago by Bao0ne
  6

• untils.h存在错误，卸载tensor进行查看时，只能看到1/4或1/2的tensor值，其他值都为0

  #9 opened 7 months ago by szxysdt
  1

• GPU memory allocation failure

  #8 opened 7 months ago by szxysdt
  2

• 刷机包下载不成功，报错：No available servers found

  #6 opened 7 months ago by shanchenjie
  1

• 能不能对不同的模型（尤其是差异很大的模型，比如 SD），分别写下教程？

  #2 opened 8 months ago by raw34
  5

• 关于CV180x的适配问题

  #4 opened 8 months ago by xpww
  2