QwenLM/Qwen1.5

Issues

• 【您好，请求帮助】魔搭下载的qwen1_5-14b-chat-q2_k.gguf 使用llama.cpp中的./server部署运行后，对话或者postman接口访问服务偶尔不返回生成结果。

  #432 opened 9 days ago by hanshishang
  0

• MOD训练的模型能否用vllm推理

  #431 opened 9 days ago by LiuChen19960902
  0

• 有没有function_call的功能实现啊？

  #428 opened 9 days ago by ucasiggcas
  2

• qwen1.5 fail load lora

  #417 opened 12 days ago by kynow2
  1

• 【求助】关于算法备案的问题

  #430 opened 9 days ago by bh4ffu
  0

• 我无法按照常规文本格式提供超过2048字符的详细回复

  #429 opened 10 days ago by wwjCMP
  1

• Qwen1.5 应用 langchain agent 问题

  #426 opened 11 days ago by bzr1
  0

• 多轮对话训练数据格式问题

  #427 opened 11 days ago by yangxin60-tal
  2

• 请问作者有尝试过neftune实验吗？

  #418 opened 14 days ago by fanbooo
  3

• vllm After calling LLM.generate, there is no output

  #425 opened 11 days ago by shuaiFLYINTHESKY
  1

• 32B-AWQ预测慢而且输出乱码

  #421 opened 12 days ago by wingjson
  2

• 💡 [REQUEST] - qwen1.5 110b能提供一个8bit量化的版本么

  #383 opened 12 days ago by starsliao
  1

• swift微调qwen1.5-32b，然后用autoawq量化，用vllm部署有时会出现感叹号

  #381 opened 12 days ago by CNXDZS
  5

• [BUG] <ModuleNotFoundError: No module named 'transformers_modules.Qwen1'>

  #384 opened 12 days ago by sugarandgugu
  1

• 多机多卡训练配置问题

  #386 opened 12 days ago by summerrain321
  1

• 💡 [REQUEST] - <有没有vLLM 部署Qwen1.5-110b-chat版本的？>

  #392 opened 12 days ago by ucas010
  2

• AWQ 模型的精度和使用

  #401 opened 12 days ago by chenchunhui97
  1

• 关于device_map

  #402 opened 12 days ago by Tvrco
  1

• 为什么同样的脚本（Qwen1.5/examples/web_demo.py）去执行Qwen1.5-32B-Chat-GPTQ-Int4的推理时，4090 24G 比 V100 32G 回答速度快了5倍，这是什么原因，是显卡性能的问题，还是代码还有哪个配置没有打开，导致V100的计算能力没有发挥出来？

  #405 opened 12 days ago by lbl1120
  2

• Qwen1.5-7B-Chat推理，结果不一致

  #409 opened 12 days ago by yecphaha
  4

• Readme inconsistent with small model licences

  #411 opened 12 days ago by ozppupbg
  1

• RuntimeError: Failed to import transformers.models.qwen2.modeling_qwen2 because of the following error (look up to see its traceback): Failed to import transformers.integrations.peft because of the following error (look up to see its traceback): No module named 'torch.distributed.checkpoint.format_utils'

  #413 opened 12 days ago by imrankh46
  3

• why the base model can chat?

  #419 opened 12 days ago by smallflyingpig
  1

• Qwen-7B量化4bit的预测结果问题

  #420 opened 12 days ago by yecphaha
  1

• 💡 [REQUEST] - <title> finetuning data format question

  #424 opened 13 days ago by lzl-mt
  1

• 是否可提供与model.py类似的模型原代码

  #422 opened 13 days ago by WzjCoder
  3

• 无

  #423 opened 13 days ago by gg22mm
  0

• 求问Dify模型配置参数

  #415 opened 13 days ago by wencan
  1

• 4B的alpacaeval leaderboard分数

  #414 opened 16 days ago by LiuChen19960902
  0

• why padding_side right?

  #416 opened 13 days ago by smallflyingpig
  1

• unlimited output，vllm 0.4.0 post ，0.5b-chat

  #410 opened 17 days ago by Jun-Howie
  1

• 14b模型为啥总偷懒？

  #412 opened 17 days ago by bwb0101
  1

• 评测脚本问题

  #408 opened 18 days ago by chk4991
  1

• FlashAttention only support fp16 and bf16 data type

  #406 opened 18 days ago by umbraclet16
  2

• how to reproduce QWEN1.5-7B-CHAT results

  #382 opened a month ago by chunniunai220ml
  2

• Ollama库中，不同的标签后缀含义是什么？

  #403 opened 19 days ago by RuikangSun
  1

• 请问有没有qwen-110b-chat的代码能力的测试结果，HumanEval，MBPP或者是LiveCodeBench等等

  #400 opened 19 days ago by White-Friday
  1

• Qwen1.5-72B-Chat在CEval val数据集上Zero shot评测为76.76正常吗，我看文章里都是84左右

  #399 opened 19 days ago by boundles
  5

• Qwen1.5-32B-Chat-AWQ 加载模型报错No module named 'transformers.models.starcoder2'

  #404 opened 20 days ago by xin-fight
  1

• qwen1 和 qwen1.5 finetune脚本为什么数据预处理会不一样？

  #391 opened 20 days ago by spider-man-jiaye
  1

• qwen1.5的sft数据处理部分使用了预训练的数据处理方式

  #396 opened 20 days ago by bjr-yql
  1

• A question about the structure of the model

  #397 opened 22 days ago by WzjCoder
  0

• 用2080TI 22G单张显卡跑 qwen 1.5 32B-INT4模型没有成功

  #389 opened 24 days ago by zhuyang2356
  5

• LlamaIndex能否直接使用modelscode下载的千问模型进行RAG检索？

  #395 opened 22 days ago by BigFishDreamWater
  0

• Qwen的代码开源了吗？

  #393 opened 23 days ago by WzjCoder
  4

• ValueError: Unable to create tensor, you should probably activate truncation and/or padding with 'padding=True' 'truncation=True' to have batched tensors with the same length. Perhaps your features (`labels` in this case) have excessive nesting (inputs type `list` where type `int` is expected).

  #388 opened 24 days ago by sjlmg
  4

• Not able to setup Qwen1.5-72B-Chat with Taskweaver

  #390 opened 24 days ago by Haxeebraja
  1

• Qwen1.5-MoE-A2.7B-Chat-GPTQ-Int4 demo encount: RuntimeError: CUDA error: invalid configuration argument

  #385 opened 24 days ago by MasterYi1024
  6

• How to batch inference for lora

  #380 opened a month ago by may012345
  1

• vllm本地部署Qwen1.5-72B-chat，在一次性生成长文本的任务中，出现超出字数限制的提示

  #387 opened 25 days ago by JSLW
  1