QwenLM/Qwen2.5

Issues

• [Badcase]: 使用Qwen2-7b翻译中文，有额外的输出

  #1005 opened 10 days ago by cjjjy
  2

• [Badcase]: qwen2.5-72b 在昇腾910推理结果不符合预期

  #992 opened 2 months ago by tianshiyisi
  6

• [REQUEST]: working tool call configuration with localai

  #949 opened 2 months ago by codefromthecrypt
  2

• [Bug]: Qwen2 moe out of memory

  #954 opened 2 months ago by FL77N
  3

• [Bug]: vLLM got different results with PeftModelForCausalLM

  #1018 opened a month ago by chansonzhang
  3

• [Bug]: vllm 启动，openai的swarm 函数调用不正常

  #1015 opened a month ago by 18600709862
  1

• [Badcase]: 相同的数据，微调时在qwen2.5 72B预训练模型上的loss是qwen2 72B的3倍，请问2.5除了数据变多了，其他有什么不一样吗

  #935 opened 2 months ago by boundles
  17

• [Bug]: qwen2.5-72b-insruct math 自测分数和榜单分数差异较大

  #1020 opened a month ago by tianshiyisi
  5

• [Bug]: Nvidia L20推理Qwen2.5 72B GPTQ-Int8模型不符合预期

  #1006 opened 2 months ago by renne444
  5

• [Badcase]: qwen2.5 instruct 14B SFT后解码重复

  #957 opened a month ago by 520jefferson
  17

• [Badcase]: Model inference Qwen2.5-32B-Instruct-GPTQ-Int4 appears as garbled text !!!!!!!!!!!!!!!!!!

  #945 opened 2 months ago by zhanaali
  16

• [Bug]: When I use Qwen2.5-7B-Instruct-GPTQ-Int4/Int8, some error occurs

  #951 opened 2 months ago by volcano1995
  8

• [Bug]: 在 4 卡 16GB V100 机器上采用 lmdeploy 部署 qwen2.5-32b-instruct-gptq-int4 模型，最高输出速度只有 80token/s ，请问这个速度正常吗？

  #1023 opened a month ago by SolomonLeon
  0

• Qwen2.5-1.5b用LLaMA-factory lora微调后，vLLM加载模型报错，求教~

  #1022 opened a month ago by 2500035435
  0

• [Badcase]: qwen2.5一定概率生成\\n

  #1021 opened a month ago by 520jefferson
  0

• [Bug]:模型加载参数精度及显存占用不符合预期，无法用int4精度加载量化后的模型Qwen2.5-7B-Instruct-GPTQ-Int4

  #1014 opened a month ago by yanli789
  3

• [Badcase]: 相同的微调数据，Qwen1.5 14B准确率比Qwen2.5 14B高20%左右，这是什么原因

  #1016 opened a month ago by Jayc-Z
  0

• [Bug]: No heartbeat received from MQLLMEngine

  #1013 opened a month ago by hulk-zhk
  0

• [Bug]: Qwen 2.5 tool calls change function names after response

  #1012 opened a month ago by LuckLittleBoy
  3

• [REQUEST]: Qwen的性能报告能否把首Token延迟也提供一下

  #1011 opened a month ago by zhufeizzz
  1

• [Bug]: Qwen2.5-72B-instruct使用vllm部署通过函数调用输出的结果里所有汉字被转义了

  #1009 opened a month ago by ericg108
  0

• [Bug]: vllm部署Qwen2.5-72B-Instruct压测出现报错

  #1008 opened 2 months ago by WangJianQ-0118
  1

• [REQUEST]: Add finetuning scripts

  #1007 opened 2 months ago by chansonzhang
  0

• [Badcase]: Qwen2.5-72B-Instruct-GPTQ-Int4 input_size_per_partition

  #986 opened 2 months ago by hyliush
  5

• [Bug]: 文档问答会忽略部分数据，比如证书号是12345 回答的是2345

  #997 opened 2 months ago by daimashenjing
  3

• [Badcase]: 函数调用出现不正常token（iNdEx）

  #991 opened 2 months ago by abiaoa1314
  0

• [Bug]: vllm部署后，用官方的例子调用报错openai.BadRequestError: Error code: 400 - {'object': 'error', 'message': "name 'Extension' is not defined", 'type': 'BadRequestError', 'param': None, 'code': 400}

  #998 opened 2 months ago by 1gst
  2

• 关于function参数的格式问题

  #994 opened 2 months ago by XuyangHao123
  0

• [Badcase]: Qwen2.5 14B Instruct can't stop generation

  #985 opened 2 months ago by Jeremy-Hibiki
  1

• [Question]: 本地部署Qwen2后的文件上传问题

  #983 opened 2 months ago by Patrick24080735
  1

• 您好，请问Qwen2.5官方榜单中的评测数据集分别是使用多少shot评测的？

  #931 opened 2 months ago by 13416157913
  3

• [Question]: 2*A100部署qwen2.5-72B-instruct的问题

  #933 opened 2 months ago by lxb0425
  0

• [Question]: slow with xInference

  #936 opened 2 months ago by Ty2000sdu
  1

• [Question]: 如果将此模型用作翻译使用，我遇到了一些问题。

  #937 opened 2 months ago by Leowolf93
  4

• [Question]: 我使用样例代码Qwen2.5/examples/demo /web_demo.py，遇到了答非所问或者重复回答的

  #938 opened 2 months ago by fuweixia
  1

• [REQUEST]: Please share Fine tuning Qwen2-72b or Qwen2-72b-Math example code ？

  #939 opened 2 months ago by ZhuJD-China
  0

• [Question]: examples/demo/web_demo.py can't this file be enhanced more?

  #942 opened 2 months ago by vignesh1507
  0

• [Question]: How many training tokens are used for the Qwen2.5 model?

  #943 opened 2 months ago by ShaneTian
  0

• [Question]: qwen2.5没考虑发布MoE架构的模型吗？

  #944 opened 2 months ago by old-wang-95
  0

• [Question]: Suggestion to improve IFEval benchmark

  #946 opened 2 months ago by LiweiPE
  0

• [Question]: 如何使用qwen2模型进行mlm任务训练

  #947 opened 2 months ago by lisi-lisi
  1

• [Question]: 在Qwen2上finetune时（作为多模态的llm），每隔几个step会出现一次前向耗时异常（正常step的4-10倍左右），导致整体训练速度较慢

  #948 opened 2 months ago by CSammyfd
  0

• [Question]: 请问怎么得出的“can generate up to 8K tokens ”这个结论？可以描述下具体的评估方法吗？

  #950 opened 2 months ago by kakaxisisan
  0

• [Question] [AgentScope] Qwen2___5-7B-Instruct 比 Qwen2 -7B性能下降了不少啊？

  #952 opened 2 months ago by monkeywl2020
  4

• [Question]: 阿里云百炼平台发布的qwen2.5-72B和开源qwen2.5-72B能力差距大

  #955 opened 2 months ago by darvsum
  5

• [Question]: 微调

  #959 opened 2 months ago by Pysion-lin
  0

• [Question]: From which version at least, the vllm supports to infrence and serve Qwen2.5-14B-Instruct model?

  #960 opened 2 months ago by zengqingfu1442
  0

• [Question]: Can I use the QWEN2.5 model to generate sentence vectors?

  #956 opened 2 months ago by trundleyrg
  1

• [Bug]: hugging face transformers使用长文本推理时报错

  #953 opened 2 months ago by allen20200111
  1

• [Bug]: <|endoftext|> , <|endoftext|>Human: endless loop asking itself questions . . .

  #932 opened 2 months ago by vishnunuk
  6