JayYip/m3tl

Issues

• Performance issue in m3tl/read_write_tfrecord.py

  #115 opened 7 months ago by DLPerf
  1

• 讨论：大佬，这个框架怎么适配未训练大模型推理

  #120 opened 8 months ago by EdwardChan5000
  1

• 疑惑：没有看到哪里执行了 text 2 tokens 的过程

  #119 opened a year ago by EdwardChan5000
  1

• 日志：大佬支持 tfevent 输出效果？

  #118 opened a year ago by EdwardChan5000
  5

• 使用rdrop, it shows a tf.split error. 大佬帮忙看看

  #112 opened 2 years ago by EdwardChan5000
  1

• 使用multi worker后效率很慢，GPU利用率也大大下降

  #73 opened 4 years ago by vanillar7
  1

• Unused attribute in object (root).body.bert.embedding_layer.token_type_embeddings: ['VARIABLE_VALUE']

  #111 opened 2 years ago by EdwardChan5000
  1

• ModuleAttributeError: 'Embedding' object has no attribute 'weights'

  #104 opened 2 years ago by xingq98
  1

• Problem_info.txt not generated.

  #110 opened 2 years ago by jmschnei
  2

• [cls] or pooled embedding?

  #107 opened 2 years ago by EdwardChan5000
  1

• 请教复现问题

  #99 opened 2 years ago by 18140663659
  8

• NameError: name 'decode_predictions' is not defined

  #103 opened 2 years ago by hoangthangta
  1

• How to use early stopping for saving the best train accuracy?

  #102 opened 2 years ago by hoangthangta
  1

• 同一任务不同数据集

  #100 opened 2 years ago by xingq98
  1

• 文档打不开

  #97 opened 2 years ago by xxllp
  1

• 关于创建数据处理函数问题

  #93 opened 3 years ago by AlexYoung757
  3

• 'TFBertEmbeddings' object has no attribute 'word_embeddings'

  #87 opened 3 years ago by wwewwt
  4

• 如何进行推理？

  #91 opened 3 years ago by GinkgoX
  1

• error： Invalid argument: cycle_length must be > 0

  #90 opened 3 years ago by AlexYoung757
  0

• 该项目是否支持文本摘要生成任务

  #85 opened 3 years ago by jollybupt
  1

• notebook is not working

  #78 opened 4 years ago by wolyosung
  3

• 想請問未來會支援Multi-modal嗎?

  #39 opened 4 years ago by QAQOAO
  4

• 请问您的项目支持tf1.15吗？

  #56 opened 4 years ago by kFoodie
  10

• When I use the new version, there's some problems.

  #68 opened 4 years ago by AIikai
  11

• How to slove Loss 0 Error

  #66 opened 4 years ago by dayL-W
  1

• No OpKernel was registered to support Op 'NcclAllReduce'

  #61 opened 4 years ago by KATEhuang920909
  1

• question

  #54 opened 4 years ago by AIikai
  1

• TopLayer for Regression

  #51 opened 4 years ago by claudiu-mihaila
  1

• 请教一些问题

  #49 opened 4 years ago by wenzitracyliu
  0

• ValueError: The two structures don't have the same nested structure

  #44 opened 4 years ago by nefujiangping
  7

• make_hidden_model()调用时参数传递错误？

  #43 opened 4 years ago by nefujiangping
  1

• Out-of-memory issue

  #40 opened 4 years ago by autapomorphy
  1

• 想了解一下更具体的训练过程

  #42 opened 4 years ago by nefujiangping
  4

• 多gpu训练步数问题

  #29 opened 4 years ago by zhongbin1
  1

• Bad recall f1 precision

  #28 opened 4 years ago by cszer
  2

• Example for Pre-training

  #27 opened 4 years ago by 1e0ng
  1

• 多GPU训练bert问题

  #31 opened 4 years ago by GaoJianchao
  1

• problem

  #41 opened 4 years ago by ZTurboX
  6

• 想了解一下怎么可以把 输入数据形式不同的两种任务 放在一起训练

  #38 opened 4 years ago by 652994331
  1

• How to prepare two sequences as input for bert-multitask-learning?

  #33 opened 5 years ago by rudra0713
  7

• bug of label set in classification task

  #34 opened 5 years ago by chiyuzhang94
  4

• Shape Mismatch error for new data set

  #32 opened 5 years ago by rudra0713
  6

• 为什么baseline.md里multitask的performance基本上都要比single task更差

  #30 opened 5 years ago by OYE93
  4

• bert model use the way of collectiveallreduce to train in multi-gpu

  #22 opened 5 years ago by weizhifei12345
  2

• 关于训练时间

  #21 opened 5 years ago by Fengwills
  7

• 分类问题

  #24 opened 5 years ago by ZTurboX
  11

• question for training process

  #20 opened 5 years ago by chiyuzhang94
  9

• How to add vocab

  #26 opened 5 years ago by cszer
  3

• 你好，请教nan的问题

  #25 opened 5 years ago by keaideii
  6

• Code error ?

  #23 opened 5 years ago by Single430
  1