38 Issues closed by 1 person

• Inference R1 with TensorRT-LLM

  #277 closed May 6, 2025

• Inquiry Regarding R1 Details: Information on R1-Zero training data / training strateties comparison / Cold-Start performance

  #282 closed May 6, 2025

• When will deepseek-reasoner (DeepSeek-R1) API accept using logprobs (soft labels)?

  #288 closed May 6, 2025

• Can you provide the deployment guide of DeepSeek R1/V3 distilled version on PCs or Mobile phones?

  #290 closed May 6, 2025

• 是否应该避免使用系统提示？

  #291 closed May 6, 2025

• [Feature Request] Providing search interfaces during thinking

  #293 closed May 6, 2025

• api能调用本地的知识库吗

  #294 closed May 5, 2025

• 服务器繁忙，是服务遭受攻击？还是有意限制？

  #296 closed May 5, 2025

• [BUG] The Server is busy

  #298 closed May 5, 2025

• 控制输出think过程的prompt案例

  #299 closed May 5, 2025

• Cannot reproduce the livecodebench result for R1

  #300 closed May 5, 2025

• DeepSeek群

  #301 closed May 4, 2025

• 本地部署如何实现联网获取最新信息回答问题

  #304 closed May 4, 2025

• Why is the CoT length shorter in DeepSeek R1 API than in the chatbot?

  #309 closed May 4, 2025

• [BUG] Receipt download failed-无法下载发票

  #310 closed May 4, 2025

• About the doubts regarding the data processing of the first phase SFT section

  #311 closed May 4, 2025

• 联网功能怎么实现？

  #312 closed May 3, 2025

• Why does selecting the deepseek-reasoner model using the deepseek api mostly return null?为什么使用deepseek api选取deepseek-reasoner模型大多返回空？

  #313 closed May 3, 2025

• R1:32B本地部署是否支持FIM补全功能？或者deepseek-coder模型后续是否升级到R1-coder ?

  #315 closed May 3, 2025

• 为什么我用transformer使用模型会有奇怪回复

  #317 closed May 3, 2025

• vllm uses wrong backend

  #318 closed May 3, 2025

• DeepSeek Train pipeline Problem

  #319 closed May 3, 2025

• Where can I download the dataset and training weights of DeepSeek - R1?

  #324 closed May 2, 2025

• VRAM required for inference

  #325 closed May 2, 2025

• 奇怪的无限输出bug

  #326 closed May 2, 2025

• 关于RAG自建知识库检索问题

  #328 closed May 2, 2025

• 并发数超过100的，一般采用什么工具部署呢

  #330 closed May 2, 2025

• 请问一下api什么时候可以恢复购买呢？最近即便是聊天也经常是服务器繁忙，感谢。

  #331 closed May 2, 2025

• ChatU第一时间集成了DeepSeek-R，欢迎企业开户体验

  #333 closed May 1, 2025

• Potential Memory Utilization Issue with vLLM Deployment of DeepSeek-R1-AWQ Model in Long Context Scenarios

  #334 closed May 1, 2025

• How does the Deepseekr1-zero model distillation work with base models like LLaMA or Qwen?

  #335 closed May 1, 2025

• How to reproduce deepseek R1-671b model performance on AIME2024

  #336 closed May 1, 2025

• 请问 R1 模型思考过程的数据结构未来会以哪种为准？

  #338 closed May 1, 2025

• Clarification on the use of `--enforce-eager` flag in vLLM

  #339 closed May 1, 2025

• chatbotui-deepseek windows本地部署一键安装包

  #344 closed Apr 30, 2025

• How to build tools call system prompt?

  #345 closed Apr 30, 2025

• r1满血版部署后推理很慢

  #346 closed Apr 30, 2025

• 请问：如何用r1模型微调或蒸馏自己的数据？还是用传统的lora? 官方好像没有介绍

  #347 closed Apr 30, 2025

1 Issue opened by 1 person

• https://api.deepseek.com/v1/embeddings 这个接口还提供服务吗？

  #652 opened May 4, 2025

5 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• Can not load DeepSeek-R1-Distill-Llama-70B with VLLM

  #351 commented on Apr 30, 2025 • 0 new comments

• Chat Template 更新后 vLLM reasoning-parser 无法正确处理

  #350 commented on Apr 30, 2025 • 0 new comments

• 怎样才能确保相同输入内容，多次生成的结果完全一致呢？

  #349 commented on Apr 30, 2025 • 0 new comments

• 探索DeepSeek等大模型的落地应用场景交流群

  #302 commented on May 4, 2025 • 0 new comments

• Table4中IF-Eval的结果有误？

  #297 commented on May 5, 2025 • 0 new comments