JittorLLMs

JittorLLMs copied to clipboard

如何支持多显卡

9

你好配置文件支持多显卡么？如何设置谢谢~~

ChinaGPT

从huggingface 4.28.1中提取`LlamaForCasualLM` 等方法试图加载stable-vicuna-13b时出现问题。修改以下几处后能加载模型但因没有`stream_chat`功能报错 修改内容： ``` python def _make_causal_mask( #input_ids_shape: torch.Size, dtype: torch.dtype, device: torch.device, past_key_values_length: int = 0 input_ids_shape: Any, dtype: torch.dtype, device: torch.device, past_key_values_length: int = 0 ): ```...

sablin39

修复注释中的大小写混乱, 与代码意图不符的问题

1

luo-zi

支持visualglm-6b吗

1

如题

TEENN2233

请问部署成功环境中gcc和g++的版本是多少？

2

我首先使用系统自带的gcc 11.3.0 报错  我尝试了 conda install gcc=8 gxx=8, 报错为GOMP的链接错误  尝试了conda install gcc=11 gxx=11 报错为C++ type traits 语法错误 

2catycm

使用torchrun命令时报错

1

报错信息为 ``` Traceback (most recent call last): File "", line 1, in ModuleNotFoundError: No module named 'torch.distributed.run'; 'torch.distributed' is not a package ```

liangnn17

在运行cli_demo.py时出现： python cli_demo.py chatglm [i 0516 03:37:17.233181 80 compiler.py:955] Jittor(1.3.7.16) src: /home/zs/anaconda3/lib/python3.9/site-packages/jittor [i 0516 03:37:17.241668 80 compiler.py:956] g++ at /usr/bin/g++(7.5.0) [i 0516 03:37:17.241838 80 compiler.py:957] cache_path: /home/zs/.cache/jittor/jt1.3.7/g++7.5.0/py3.9.7/Linux-5.4.0-12x80/IntelRXeonRSilx76/default [i 0516 03:37:17.277480...

LianouXianJZ

python cli_demo.py报jittor 编译错误

1

`JittorLLMs]# python cli_demo.py chatglm [i 0513 13:38:43.964825 64 compiler.py:955] Jittor(1.3.7.16) src: /root/JittorLLMs/myenv/lib/python3.10/site-packages/jittor [i 0513 13:38:43.968178 64 compiler.py:956] g++ at /usr/bin/g++(8.5.0) [i 0513 13:38:43.968239 64 compiler.py:957] cache_path: /root/.cache/jittor/jt1.3.7/g++8.5.0/py3.10.9/Linux-5.4.203-xa2/IntelRXeonRPlax40/main [i 0513 13:38:43.971742...

IUgly

执行`python cli_demo.py chatglm`时，卡住不动

2

执行`python cli_demo.py chatglm`时，一直卡在这里不动，如图所示：  我的python版本是3.11.3

jianzi0307

我看了下mac上好像还是调用cpu的。

DevilMayCry2333