wangyuxinwhy/uniem

Issues

• 关于s2p和s2s

  #134 opened 2 months ago by zqhi71
  0

• 微调数据中做负样本和不做负样本，哪个会更好？三种微调的数据格式，训练出来的模型会不一样吗？

  #131 opened 3 months ago by yangjianxie
  0

• 微调如何在多卡训练

  #132 opened 3 months ago by yangjianxie
  0

• 微调如何在多卡训练

  #133 opened 3 months ago by yangjianxie
  0

• M3E如何指定为GPU训练，而不是CPU训练？

  #128 opened 6 months ago by sherlock1987
  2

• 单机多卡运行时报错 has parameters that were not used in producing loss

  #89 opened a year ago by whi497
  6

• 单机多卡训练

  #101 opened a year ago by kg-nlp
  5

• Loss固定不变或者上升

  #127 opened 6 months ago by Jeru2023
  2

• 微调后模型没有保存2_Dense

  #130 opened 5 months ago by insomnia-y1999
  1

• 模型训练后生成的内容不对

  #129 opened 5 months ago by Nicklaus421
  1

• 代码跑着跑着就挂了，CUDA out of memory

  #126 opened 7 months ago by susht3
  1

• 问题

  #125 opened 7 months ago by wuguangshuo
  3

• 能不能说明一下显卡要求啊？

  #123 opened 8 months ago by pdwfree
  3

• fintuner如何使用gpu？

  #124 opened 7 months ago by susht3
  1

• checkpoint模型无法加载

  #98 opened a year ago by gctian
  3

• 微调的一些问题

  #107 opened a year ago by zhouzhou0322
  3

• 转onnx问题

  #122 opened 9 months ago by sunzx8
  1

• 实际测试 PairInBatchNegSoftmaxContrastLoss和PairInBatchNegCoSentLoss的值是一样的

  #121 opened 9 months ago by WEN6Lev57q4
  1

• 请问微调之后的模型如何支持C_MTEB数据集上的评测呢

  #120 opened 9 months ago by AugustLHHHHHH
  2

• m3e训练的时候使用的数据集是hugg上面列出的数据集，训练和测试集和验证集一起用来训练了吗？

  #119 opened 9 months ago by NLPJCL
  1

• 请教贴：文本最大长度

  #114 opened a year ago by gggdroa
  5

• CoSentLoss的一点疑惑，

  #95 opened a year ago by NLPJCL
  3

• 微调后模型保存和load的问题

  #117 opened a year ago by LeiShenVictoria
  3

• m3e-large数据集的相关问题

  #118 opened a year ago by zysNLP
  1

• 进行评测时会报错,分叉可能会导致死锁.

  #115 opened a year ago by TC9905
  1

• 求一份评测数据集

  #116 opened a year ago by dmcszero
  1

• 负采样

  #113 opened a year ago by AugustLHHHHHH
  3

• 复现不到模型效果

  #85 opened a year ago by huangjiaheng
  14

• 是否支持bge了呢

  #93 opened a year ago by steph730
  4

• sentence-transformer调用huggingface模型

  #112 opened a year ago by allendred
  1

• 关于huggingface方法调用

  #111 opened a year ago by allendred
  1

• 无法使用图嵌入

  #110 opened a year ago by cutecharmingkid
  1

• m3e模型支持openai的接口调用方式吗

  #109 opened a year ago by pingyuan2016
  1

• 支持商汤https://huggingface.co/sensenova/piccolo-large-zh的微调吗

  #104 opened a year ago by yuanjie-ai
  1

• 关于模型效果

  #108 opened a year ago by Clearloveplus7
  1

• 使用自有数据集微调m3e-base后，在文本检索任务中效果变差

  #105 opened a year ago by XuHao777
  1

• cant found the package on pypi

  #106 opened a year ago by zhouzhou0322
  1

• No sentence-transformers model found

  #100 opened a year ago by tfka
  2

• 单句分类的微调

  #103 opened a year ago by bulululu
  1

• dataset里带不带validation，loss差别非常大

  #102 opened a year ago by huohuade-blog
  1

• TripletInBatchNegSoftmaxContrastLoss的forward函数中的labels有问题

  #99 opened a year ago by zhouzj1610
  1

• 带有指令的数据集上的微调

  #97 opened a year ago by imathg
  1

• 空字符串也会输出特征，且有较高的相似度（余弦距离）

  #96 opened a year ago by zywhehe
  1

• 微调的方法是否支持仅微调部分参数

  #90 opened a year ago by Xls1994
  2

• 请教一下怎么快速构造领域数据的训练样本

  #92 opened a year ago by QQR1
  1

• 能否支持tensor并行训练，或者使用deepspeed呢？

  #91 opened a year ago by programmer-ming
  1

• 如何评测自己的模型

  #83 opened a year ago by TWJxi
  4

• 评测 M3E-base 模型 报错

  #86 opened a year ago by zhangyisheng0629
  3

• 用以下代码微调m3e-large模型后，不相关的句子相似度极高，请问需要如何处理？

  #84 opened a year ago by twwch
  6

• 生成embedding的速度比text2vec慢很多

  #80 opened a year ago by ralgond
  1