luchangli03/export_llama_to_onnx

Issues

• 有没有llama 的onnx inference 脚本

  #7 opened a year ago by hujuntao123
  3

• 请问一下对应llama3的transformers版本是多少

  #19 opened 4 months ago by sihouzi21c
  1

• 执行export_llama3.py时为什么会导出一大堆中间文件（参数）？

  #16 opened 4 months ago by Lingzzyy
  3

• 请问Qwen转换出错问题:RuntimeError: Sizes of tensors must match except in dimension 2. Expected size 28 but got size 4 for tensor number 1 in the list.

  #17 opened 5 months ago by yanxiao1930
  4

• Please uninstall/disable FlashAttention (and maybe xformers) before model conversion。请问这句话的意思是在模型转换前必须重新训练一个不使用flashattention的模型吗？

  #18 opened 4 months ago by PeterXingke
  1

• Can you give an example code of how to deduce onnx model after qwen switched to onnx

  #15 opened 7 months ago by Pengjie-W
  1

• 有导出Qwen-VL的7B模型转onnx的程序吗？

  #14 opened 9 months ago by chantjhang
  0

• 使用onnx库读取转换好的onnx模型报错

  #13 opened 10 months ago by L1-M1ng
  7

• 用3090 导出7b 和13b llama2 报oom

  #6 opened 10 months ago by hujuntao123
  1

• 显存占用增加

  #5 opened 10 months ago by cdxzyc
  1

• convert Qwen question

  #1 opened 10 months ago by OneStepAndTwoSteps
  4

• 转换QWen-7B错误

  #12 opened 10 months ago by L1-M1ng
  2

• 转换qwen模型的时候，提示atten_mask:5 error.

  #10 opened a year ago by louwangzhiyuY
  2

• 转换llama也提示错误。 AttributeError: 'tuple' object has no attribute 'get_usable_length'

  #11 opened a year ago by louwangzhiyuY
  1

• 使用3090导出 QWen-7b，报OOM问题。

  #9 opened a year ago by linthy94
  3

• 单卡a6000 50g会oom

  #8 opened a year ago by 77281900000
  1

• 请问如何正确推理使用cuda导出的fp16 onnx chatglm2-6b-32k模型？

  #4 opened a year ago by yuunnn-w
  1

• 适用范围

  #3 opened a year ago by hardlipay
  1