GaoPeng97/transformer-xl-chinese

Issues

• 请问训练Doupo这个模型需要多大的显存？

  #41 opened 4 years ago by lxiaodao
  2

• 重复

  #16 opened 5 years ago by Masterchenyong
  5

• inference 中的saver.restore(sess, eval_ckpt_path)模型恢復不出來

  #39 opened 4 years ago by tangyixuan01
  2

• 你好，我发了邮件给gp2015@mail.ustc.edu.cn，请查看下哦

  #31 opened 4 years ago by 1397981186
  0

• 你好，请问ppl到多少停止训练比较合适

  #32 opened 4 years ago by zychen-real
  2

• 请问tf.model中rel_multihead_attn的w是一个batch还是单条数据？

  #38 opened 4 years ago by oshindow
  0

• 请问1080Ti能训练 shi这个训练集吗

  #22 opened 5 years ago by csnever
  1

• 知乎数据

  #18 opened 5 years ago by echoyes
  3

• inference中的sess.run

  #36 opened 5 years ago by zhanbaohang
  0

• why takes only 135M gpu memory but takes 10-50G cpu memory?

  #35 opened 5 years ago by SeekPoint
  0

• inference 在哪里分词，也不知道有没有人回复

  #34 opened 5 years ago by zhitunAI
  0

• 改变输入样本的格式

  #33 opened 5 years ago by CuiYan72
  0

• inference的时候出现报错

  #30 opened 5 years ago by ScottishFold007
  0

• seed text干什么用的？

  #29 opened 5 years ago by aazjcom
  6

• AttributeError: module 'tensorflow.sparse' has no attribute 'to_dense'

  #26 opened 5 years ago by LiaoBoWen
  1

• 关于 inference

  #28 opened 5 years ago by dupeng2014
  1

• 如何用transformerXL做文本摘要任务

  #27 opened 5 years ago by Liao-Yong
  1

• 怎么在训练的时候添加样本的权重

  #25 opened 5 years ago by lomessa
  0

• 有关cache memory的问题

  #24 opened 5 years ago by echoyes
  0

• transformer-xl in 1-billion experiment with base model configuration

  #23 opened 5 years ago by cmathx
  0

• Error generating English text. How to fix?

  #8 opened 5 years ago by GenTxt
  3

• ModuleNotFoundError: No module named 'progressbar'

  #12 opened 5 years ago by ldxbaby756
  2

• inference error Assign requires shapes of both tensors to match. lhs shape= [5472,410] rhs shape= [84827,410]

  #15 opened 5 years ago by lk1983823
  7

• infernence出现的问题

  #21 opened 5 years ago by DevelMayCry-MrChen
  6

• 关于中文的预处理

  #19 opened 5 years ago by BlueFishDream
  2

• 模型层缺少encoder层

  #20 opened 5 years ago by icecream0
  0

• 训练完doupo以后, 运行inference的时候显示: ValueError: Can't load save_path when it is None.

  #14 opened 5 years ago by lk1983823
  2

• scripts/doupo_base_gpu.sh: line 175: 2335 Killed

  #17 opened 5 years ago by AaronZhangL
  1

• 您用了怎么样的设备，训练了多长时间呢

  #5 opened 5 years ago by kidsbmilk
  6

• inference restore model failed

  #13 opened 5 years ago by zhixuhao
  2

• 能加一下你的微信吗，有些问题想请教一下

  #4 opened 5 years ago by DevilEEE
  3

• 请问模型训练的终止条件是什么？

  #6 opened 5 years ago by xinyx
  5

• what is the purpose of mems_id for inference.

  #10 opened 5 years ago by o20021106
  4

• 分词之后会提升生成的质量吗

  #11 opened 5 years ago by Qcosmo
  6

• 训练，验证，推理时指定的模型文件不一样

  #7 opened 5 years ago by fengzuo97
  1

• 请问预测过程又可以优化执行时间的方法吗

  #9 opened 5 years ago by Qcosmo
  2

• AttributeError: 'list' object has no attribute 'get_shape'

  #3 opened 5 years ago by Seumi
  2

• Where is the zhihu dataset?

  #2 opened 5 years ago by Machine-Tom
  1

• hello GaoPeng97,only gpu? no cpu ?

  #1 opened 5 years ago by redcloth123
  1