Search results

Feature Request / 功能需求： 增加并发支持

Status: Open.

#661 In kvcache-ai/ktransformers;

· yeungtuzi opened on Feb 25, 2025

DeepSeek-V3-GGUF-Q4_K_M blk8层缺少blk.8.ffn_up_exps.weight

Status: Open.

#656 In kvcache-ai/ktransformers;

· Moe-Shiro opened on Feb 25, 2025

ktranformers编译报错：iqk_mul_mat.inc:2573:15: error: ‘block_q8_0_x4’ does not name a type; did you mean ‘block_q8_0’?

Status: Open.

#655 In kvcache-ai/ktransformers;

· coelien opened on Feb 25, 2025

能否提供一个守护进程将ktransformers以系统服务的方式运行？

Status: Open.

#652 In kvcache-ai/ktransformers;

· yeungtuzi opened on Feb 25, 2025

出现了不兼容情况

Status: Open.

#648 In kvcache-ai/ktransformers;

· yuyangtian1000 opened on Feb 24, 2025

LLM Produces Unprompted Responses After Initial Message

Status: Open.

#647 In kvcache-ai/ktransformers;

· Husamx opened on Feb 24, 2025

Deepseek R1 671B with ollama on AWS Graviton4 bare metal intance performance

Status: Open.

#646 In kvcache-ai/ktransformers;

· westlakequant opened on Feb 24, 2025

OSError: /mnt/ai/models/unsloth/DeepSeek-R1-GGUF/DeepSeek-R1-UD-Q2_K_XL/ does not appear to have a file named config.json. Checkout 'https://huggingface.co//mnt/ai/models/unsloth/DeepSeek-R1-GGUF/DeepSeek-R1-UD-Q2_K_XL//tree/None' for available files.

Status: Open.

#642 In kvcache-ai/ktransformers;

· goooodggy opened on Feb 24, 2025

使用的是A800的服务器，CPU占用很高，但是内存占用才几个G，如下

Status: Open.

#641 In kvcache-ai/ktransformers;

· White-Friday opened on Feb 24, 2025

物理内存低占用，推理时载入内存，速度极慢

Status: Open.

#639 In kvcache-ai/ktransformers;

· Sk3W3d opened on Feb 24, 2025

Killed while loading weight

Status: Open.

#637 In kvcache-ai/ktransformers;

· NicholasIrving opened on Feb 24, 2025

dequantize_q6_k(): incompatible function arguments

Status: Open.

#636 In kvcache-ai/ktransformers;

· PurpleLikeNero opened on Feb 24, 2025