airockchip rknn-llm issues

airockchip / rknn-llm

Other

347 stars 29 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

RKNN: [22:16:52.189] failed to convert handle(1020) to fd, ret: -1, errno: 24, errstr: Too many open files

#47 17656178609 closed 3 months ago
2
LLM models have bad output

#46 kalous12 opened 4 months ago
9
将来是否会支持 moe 模型

#45 YanxingLiu opened 4 months ago
0
rkllm_init error in rkllm1.0.1

#44 eework closed 4 months ago
2
E RKNN: [00:13:05.159] failed to convert handle(1020) to fd, ret: -1, errno: 24, errstr: Too many open files

#43 wmx-github closed 4 months ago
1
Let's talk about converting and runtime 我们来谈谈转换和运行时

#42 80Builder80 opened 4 months ago
0
rkllm1.0.1 总是激活3个core

#41 AACengineer opened 4 months ago
0
是否支持自定义rkllm的模型量化？

#40 MosRat closed 4 months ago
3
phi3 load error in rkllm1.0.1

#39 eework closed 4 months ago
9
flask_server.py DEMO无法正常执行

#38 westwind027 closed 4 months ago
7
Qwen1.5-0.5B模型两次回答不一致

#37 AACengineer opened 4 months ago
10
Phi3 推理无法结束或在最后出现无效输出

#36 AndySze closed 4 months ago
2
Need a way to stop inference midway

#35 Tidus1991 closed 4 months ago
1
模型可视化、推理性能评估

#34 AACengineer opened 5 months ago
0
Open Source SDK

#33 80Builder80 closed 4 months ago
0
Phi-3 error

#32 dnhkng opened 5 months ago
10
rknn_init failed

#31 YanxingLiu opened 5 months ago
1
模型转换内存需求过大

#30 YanxingLiu opened 5 months ago
6
Is the performance bottleneck of rknn llm in the CPU？

#29 Caical opened 5 months ago
4
Phi 3 and Llama 3 compatibility?

#28 Pelochus closed 4 months ago
6
firfly rk3588 rkllmcpu占用率问题

#27 Caical opened 5 months ago
9
orangepi@orangepi5:~/rknn-llm-main/rkllm-runtime/example/build/build_linux_aarch64_Release$ taskset f0 ./llm_demo ./qwen.rkllm rkllm init start E RKNN: [07:37:38.389] failed to convert handle(1020) to fd, ret: -1, errno: 24, errstr: Too many open files Segmentation fault (core dumped)

#26 lindsayshuo opened 5 months ago
3
请问有没有支持CLIP，GLIP，DINO等模型的计划？

#25 KunxingYang opened 5 months ago
1
rk3588 ubuntu kernel版本为5.10 替换npu 代码后编译报错

#24 Caical opened 5 months ago
2
能否用在3566部署？

#23 Gooddz1 opened 5 months ago
1
rk3588s 配置和支持模型的规模

#22 AACengineer opened 5 months ago
0
Optimization failure

#21 80Builder80 opened 5 months ago
4
能否用3588s部署？

#20 deanwintrester closed 4 months ago
1
run模型时，为什么还需要指定max_context_len?

#19 AACengineer closed 5 months ago
3
Expection while converting model

#18 dic1911 opened 5 months ago
4
rkllm_init error

#17 beingjoey closed 5 months ago
4
Errors found in building kernal.

#16 sdrzmgy closed 5 months ago
6
rkllm-toolkit/examples/test.py broken?

#15 Pelochus closed 5 months ago
1
请问一下，样例中的大模型有没有具体的下载链接

#14 kaylorchen closed 5 months ago
2
Low-Level Code Request

#13 fydeos-alex opened 6 months ago
0
Need Support For INT4 Quantize

#12 fydeos-alex closed 5 months ago
2
RKLLMParam max_context_len Maximum supported length

#11 jun-zhang opened 6 months ago
0
转换MiniCPM失败

#10 jieli1990 closed 5 months ago
1
benchmark data

#9 WangFengtu1996 opened 6 months ago
3
Improve Work Need To Be Done

#8 fydeos-alex opened 6 months ago
0
What is the cause of process interruption during model conversion？

#7 Han-WG closed 6 months ago
2
编译内核时报错

#6 sdrzmgy opened 6 months ago
3
目前支持视觉大模型或者多模态大模型吗？

#5 Han-WG closed 6 months ago
1
E RKNN: [10:27:30.619] failed to allocate handle, ret: -1, errno: 14, errstr: Bad address Segmentation fault (core dumped)

#4 fydeos-alex opened 6 months ago
15
Doc Misspelled

#3 fydeos-alex closed 5 months ago
1
Phi-2 Model Load Error

#2 fydeos-alex closed 6 months ago
2
update

#1 SoulProficiency opened 6 months ago
0