Replies: 2 comments 18 replies
-
|
在 issue #2175 中回复了。可以重新试试看了 |
Beta Was this translation helpful? Give feedback.
-
|
好的 |
Beta Was this translation helpful? Give feedback.
-
|
|
Beta Was this translation helpful? Give feedback.
-
|
Can you share the information by executing |
Beta Was this translation helpful? Give feedback.
-
|
|
Beta Was this translation helpful? Give feedback.
-
|
从上面的日志看，是pytorch 判断 driver 版本太低了。 pip install torch==1.12.0+cu113 torchvision==0.13.0+cu113 --extra-index-url https://download.pytorch.org/whl/cu113 |
Beta Was this translation helpful? Give feedback.
-
|
CUDA_VISIBLE_DEVICES=0,1 lmdeploy chat /sdd/model/lmdeploy/model/Qwen2-1.5B --cache-max-entry-count 0.1 --tp=2 |
Beta Was this translation helpful? Give feedback.
-
|
--tp=2 这个参数加了就报错了 不加能正常回答 但是只有RTX3060ti这张显卡在用 T4那张就没用了 |
Beta Was this translation helpful? Give feedback.
-
|
@lzhangzz do you have any idea? One T4, one 3060, different arch, emmmm |
Beta Was this translation helpful? Give feedback.
-
|
The model is in BF16 format which is not supported by the T4 GPU. |
Beta Was this translation helpful? Give feedback.
-
|
我跑glm4-9b-chat也是 俩张卡跑不起来 只有GPU1 在使用 |
Beta Was this translation helpful? Give feedback.
-
CUDA11.4 安装报错,提示404
Beta Was this translation helpful? Give feedback.
All reactions