Tele-AI/TeleChat2

Issues

• TeleChat2-3B的单机多卡训练微调报错Exception: Current loss scale already at minimum - cannot decrease scale anymore. Exiting run.

  #35 opened 8 days ago
  0

• TeleChat2-3B推理示例报错：AttributeError: 'Telechat2Tokenizer' object has no attribute 'apply_chat_template'

  #29 opened 13 days ago
  1

• 910B推理35B模型

  #26 opened 16 days ago
  0

• telechat和telechat2系列有什么区别

  #25 opened 17 days ago
  0

• pip install -r requirements 报错

  #24 opened 13 days ago
  4

• 35b推理的prompt和7b的不一样吗，35b推理出来的答案有问题

  #20 opened 24 days ago
  2

• 35b

  #19 opened 24 days ago
  0

• vllm多卡推理问题

  #18 opened a month ago
  4

• telechat7b的vllm推理，max_model_len和max_tokens值问题

  #16 opened a month ago
  3

• vllm 0.6.4.post1支持

  #15 opened a month ago
  1

• 910b推理

  #13 opened a month ago
  2

• 关于system字段适配

  #12 opened a month ago
  1

• 关于chat_template

  #11 opened a month ago
  3

• sft_single_node.sh:line3: $'\r': command not found 问题解决

  #10 opened a month ago
  1

• Telechat-115B推理显存要求

  #5 opened a month ago
  1

• 能不能开放int4量化权重

  #4 opened 3 months ago
  1

• 量化模型权重

  #3 opened 3 months ago
  0

• 权重没上传上来

  #2 opened a month ago
  1