FlagAI-Open/Aquila2

Issues

• BAAI/AquilaChat2-7B 是否可以归纳总结文本内容？

  #144 opened 5 months ago by Yshiboo
  0

• 官方可以做一版Aquila2Chat-7B到ModelScope上吗？

  #143 opened 8 months ago by Micla-SHL
  1

• Aquila_BGE_langchain应用所需配置大概需要在什么区间？

  #138 opened 8 months ago by Micla-SHL
  1

• Aquila2 7B and Aquila2-7B LoRA How to merge parameters？

  #140 opened 8 months ago by Micla-SHL
  0

• How to review Aquila2-7B model

  #141 opened 8 months ago by Micla-SHL
  0

• Number of training tokens

  #142 opened 8 months ago by RicardoDominguez
  2

• CUDA_VISIBLE_DEVICES=4,3 python ./local_search.py时，tool中不能识别到GPU3

  #132 opened 9 months ago by tfal-yan
  1

• 指定2个以上gpu时./preprocess.sh: line 3: 2269 Bus error (core dumped) ，1个就不会，log对比在内文

  #131 opened 9 months ago by tfal-yan
  0

• How can I take the model trained using the script and apply it in Aquila_BGE_langchain?

  #130 opened 10 months ago by Micla-SHL
  1

• RuntimeError: probability tensor contains either `inf`, `nan` or element < 0

  #137 opened 10 months ago by mru4913
  0

• [Question] V100 FInetuning

  #136 opened 10 months ago by mru4913
  0

• from predict import predict

  #135 opened 10 months ago by adol001
  0

• 公开的评测榜单

  #133 opened a year ago by yucc-leon
  2

• 量化GPTQ中Aquilachat34b-4bit 代码的predict文件在哪里，Modelscope上没有上传，谢谢

  #134 opened a year ago by LukeeCR
  1

• bash finetune/34B/finetune_qlora_single_node.sh异常

  #129 opened a year ago by tfal-yan
  27

• bge有没有量化版本

  #123 opened a year ago by yanchaoguo
  1

• 流式返回怎么搞

  #117 opened a year ago by 631068264
  2

• Aquila2/examples/Aquila_BGE_langchain/BGE# CUDA_VISIBLE_DEVICES=4,3,2,1,0 ./preprocess.sh 行 3: 2866484 段错误

  #126 opened a year ago by tfal-yan
  2

• How can I save the config.json when fine-tuning Aquila 2?

  #128 opened a year ago by Micla-SHL
  3

• url 404

  #127 opened a year ago by Micla-SHL
  2

• 7b模型qlora sft出错，transformers Version: 4.35.0 ，错误信息见内

  #125 opened a year ago by tfal-yan
  9

• 34B模型用qlora做sft，设置一个进程一个主机，5个24G卡，发现脚本加载完一个卡后就提示内存不够异常退出，其他卡空闲

  #124 opened a year ago by tfal-yan
  2

• 是否支持function功能

  #116 opened a year ago by hunter-xue
  2

• 16K的上下文长度是真实的吗？

  #99 opened a year ago by DirtyKnightForVi
  8

• RAG的问答能力很差。

  #107 opened a year ago by HouYueJie
  5

• 使用NTK的情况下，推理输入超过2048 token模型无法正常回答

  #105 opened a year ago by 0-1CxH
  1

• run bash finetune/7B/finetune.sh It didn't output any useful information

  #108 opened a year ago by Micla-SHL
  13

• 请问34B模型全参数微调需要几张卡（A100、80G）？

  #110 opened a year ago by zdaaaaa
  1

• 效果没有想象当中的好

  #101 opened a year ago by lucasjinreal
  14

• 34B模型微调数据

  #106 opened a year ago by Little-rookie-ee
  2

• AutoGPTQ support

  #100 opened a year ago by lucasjinreal
  2

• aquilachat2-34b-16k这个的v1.2版本可以在哪里下载呀

  #104 opened a year ago by Edisonwei54
  3

• flagai安装报错

  #77 opened a year ago by zxzxde
  6

• AttributeError: 'AquilaForCausalLM' object has no attribute 'save_checkpoint'

  #84 opened a year ago by sincerity-being
  2

• AquilaSQL-7B 的支持最大token数是多少呢？

  #103 opened a year ago by nuaabuaa07
  3

• base-34B模型HF链接失效

  #96 opened a year ago by HalcyonLiang
  2

• finetune+chat34b 处理数据时发现问题

  #94 opened a year ago by Darrenzeng
  3

• tokenizer对aquila-legacy模板中的"### Assistant: "处理时存在解码问题

  #97 opened a year ago by Darrenzeng
  2

• ReadME中关于finetune数据的instruction字段

  #79 opened a year ago by yuiant
  1

• llama.cpp量化后的模型问答过程陷入自说自话

  #95 opened a year ago by gaord
  1

• 请问4bit量化后的模型如何保存

  #90 opened a year ago by sunyclj
  4

• hostfile的作用及修改

  #92 opened a year ago by eric971120
  2

• Token indices sequence length is longer than the specified maximum sequence length for this model (5669 > 2048). Running this sequence through the model will result in indexing errors

  #93 opened a year ago by Edisonwei54
  3

• 镜像启动容器报错

  #78 opened a year ago by Darrenzeng
  1

• No module named ‘deepspeed.accelerator‘

  #80 opened a year ago by zhangxuan602461121
  1

• 关于Flash attention相关

  #91 opened a year ago by onlyfish79
  2

• 载入tokenizer时出现ValueError，请问怎么解决？

  #75 opened a year ago by chk4991
  4

• finetune parameters

  #85 opened a year ago by onlyfish79
  2

• 4张4090能跑起来吗

  #87 opened a year ago by bird-9
  2

• 请问transformer代码示例中的predict文件在哪？

  #86 opened a year ago by Bonjour0531
  2