kvcache-ai/ktransformers

Issues

• 复现InternLM2.5-7B-Chat-1M报错

  #114 opened 6 days ago by Cherishyt
  0

• 硬件配置支持

  #112 opened 13 days ago by Cherishyt
  1

• RuntimeError CUDA error when running Infinite Bench

  #113 opened 9 days ago by Flitieter
  0

• Detailed specification of the computer hardware to run 236B DeepSeek-Coder-V2

  #108 opened 17 days ago by atomlayer
  1

• Support for New SOTA MoE: tencent/Tencent-Hunyuan-Large

  #111 opened 14 days ago by ThomasBaruzier
  0

• install error on windows, need help

  #109 opened 16 days ago by gaowayne
  0

• how to implement new algorithm in this repo?

  #105 opened 19 days ago by lumiere-ml
  1

• feature request: support internvl2

  #107 opened 19 days ago by kolinfluence
  0

• Long prompt with DeepSeek crashing with tensor size mismatch

  #101 opened a month ago by bitbottrap
  11

• Error loading model: token_embd.weight not found in GGUF file

  #102 opened a month ago by antonovkz
  1

• Attempting to increase output to 16k results in crash during output

  #104 opened a month ago by bitbottrap
  1

• Error Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

  #96 opened 2 months ago by drrros
  5

• How to infer quantized models on CPU&GPU

  #103 opened a month ago by shuzhang-pku
  1

• Does ktransformers support deepseek V2.5？

  #100 opened a month ago by huliangbing
  2

• Specify MAX_NEW_TOKENS for ktransformers server

  #92 opened 2 months ago by arthurv
  2

• Busy loop in cpu_backend/task_queue.cpp keeps 1 thread at 100% CPU when queue is empty

  #80 opened a month ago by sayap
  5

• are marline and q4k totally equivalent?

  #87 opened 2 months ago by Eutenacity
  5

• How can I use opencompass benchmark tools to test ktransformers in long context?

  #91 opened 2 months ago by AsVoider
  1

• Getting reasonable performance on dual RTX 3090 and 128gb

  #85 opened 2 months ago by trilog-inc
  7

• Deepseekv2推理速度很慢，看样子似乎在cpu上做推理，gpu利用率很低

  #93 opened 2 months ago by Chain-Mao
  3

• ImportError: DLL load failed while importing KTransformersOps: The specified module was not found.

  #94 opened 2 months ago by SCP12rs
  7

• Suggestion to add DeepSeek v2.5 support

  #95 opened 2 months ago by arisau
  4

• Installation Problem

  #90 opened 2 months ago by Chain-Mao
  1

• Installation requirements

  #89 opened 2 months ago by arthurv
  4

• 可以给出详细的硬件配置清单吗?

  #84 opened 2 months ago by qixing-ai
  2

• Seg Fault on long replies

  #82 opened 2 months ago by matthusby
  2

• What is the maximum input token size supported for DeepSeek V2?

  #68 opened 2 months ago by fengyang95
  1

• 8-GPU configuration on L40 OOM

  #76 opened 2 months ago by fengyang95
  8

• Is deepseek-ai/DeepSeek-V2.5 supported?

  #79 opened 2 months ago by AshD
  9

• docker container fails to start due to missing package 'uvicorn'

  #66 opened 2 months ago by sammcj
  1

• How can i run internlm2_5-7b-chat-1m in ktransformers?

  #74 opened 2 months ago by Ma1oneZhang
  4

• Missing pip packages flash_attn and wheel

  #69 opened 3 months ago by bitbottrap
  2

• When the input token exceeds 4096, an error will occur.

  #73 opened 3 months ago by fengyang95
  4

• UnboundLocalError: cannot access local variable 'chunck_mask' where it is not associated with a value

  #70 opened 3 months ago by fengyang95
  2

• Would you support glm4-chat-1m

  #65 opened 3 months ago by choyakawa
  1

• docker builds and pip install broken - No module named 'cpufeature'

  #61 opened 3 months ago by sammcj
  5

• More Efficient Layer Distribution for DeepSeek Coder v2 on Multiple GPUs and CPUs

  #49 opened 3 months ago by BGFGB
  4

• Support for Mistral-Large-Instruct-2407-GGUF ？

  #53 opened 3 months ago by LIUKAI0815
  2

• Add a instruction for configuring CUDA_HOME and CUDA_PATH to the install section of README.md.

  #54 opened 3 months ago by hyx1999
  2

• How to properly disable offloading MoE layers to CPU?

  #50 opened 3 months ago by molamooo
  5

• Cannot run DeepSeek V2 Chat in server mode on 2 GPUs

  #45 opened 3 months ago by ELigoP
  1

• Mixtral-8x7B-v0.1 GGUF file error

  #42 opened 3 months ago by RealLittleXian
  1

• CUDA error: No kernel image is available for execution on the device

  #44 opened 3 months ago by Forsworns
  2

• Ubuntu 24.04 GLIBCXX version fail

  #37 opened 3 months ago by ELigoP
  3

• Can I run llama3.1 70b with rtx4090+64g ddr5 ram?

  #47 opened 3 months ago by codeMonkey-shin
  1

• [ENHANCEMENT] improve GPU utilization for multi-GPU

  #46 opened 3 months ago by ELigoP
  1

• ollama chat not realised

  #32 opened 3 months ago by xldistance
  2

• using docker start api server can't set max_new_tokens

  #31 opened 3 months ago by goldenquant
  1

• Unable to use the web interface

  #33 opened 3 months ago by xldistance
  0

• using docker got errors

  #28 opened 3 months ago by goldenquant
  3