【求助帖】2406期-林希老师-课程20(LLM->多模态理论+案例讲解),MiniGPT-V2执行时demo_v2.py报错 #563
Labels
No Label
bug
duplicate
enhancement
help wanted
invalid
question
wontfix
No Milestone
No project
No Assignees
2 Participants
Notifications
Due Date
No due date set.
Dependencies
No dependencies set.
Reference: HswOAuth/llm_course#563
Loading…
Reference in New Issue
Block a user
No description provided.
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
demo_v2.py执行时会出错ImportError: cannot import name 'driver' from 'triton.runtime' .重装了triton2.0.0，问题仍然存在;如果换为安装available的较新版本trition2.1.0，虽然这个问题会避免,但安装时就会提示pytorch2.0.0和trition2.1.0有兼容问题，并且安装完成后也会出其它兼容性问题。其它python,cuda版本确信都与课件一致。
(minigptv) root@autodl-container-101f43b926-1f402c61:~/autodl-tmp/MiniGPT-4# python demo_v2.py --cfg-path eval_configs/minigptv2_eval.yaml --gpu-id 0
Traceback (most recent call last):
File "/root/autodl-tmp/MiniGPT-4/demo_v2.py", line 18, in
from minigpt4.common.config import Config
File "/root/autodl-tmp/MiniGPT-4/minigpt4/init.py", line 16, in
from minigpt4.models import *
File "/root/autodl-tmp/MiniGPT-4/minigpt4/models/init.py", line 13, in
from minigpt4.models.base_model import BaseModel
File "/root/autodl-tmp/MiniGPT-4/minigpt4/models/base_model.py", line 17, in
from peft import (
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/peft/init.py", line 22, in
from .mapping import MODEL_TYPE_TO_PEFT_MODEL_MAPPING, PEFT_TYPE_TO_CONFIG_MAPPING, get_peft_config, get_peft_model
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/peft/mapping.py", line 16, in
from .peft_model import (
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/peft/peft_model.py", line 31, in
from .tuners import LoraModel, PrefixEncoder, PromptEmbedding, PromptEncoder
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/peft/tuners/init.py", line 20, in
from .lora import LoraConfig, LoraModel
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/peft/tuners/lora.py", line 36, in
import bitsandbytes as bnb
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/bitsandbytes/init.py", line 15, in
from .nn import modules
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/bitsandbytes/nn/init.py", line 21, in
from .triton_based_modules import (
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/bitsandbytes/nn/triton_based_modules.py", line 7, in
from bitsandbytes.triton.int8_matmul_mixed_dequantize import (
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/bitsandbytes/triton/int8_matmul_mixed_dequantize.py", line 13, in
from .matmul_perf_model import early_config_prune, estimate_matmul_time
File "/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/bitsandbytes/triton/matmul_perf_model.py", line 10, in
from triton.runtime import driver
ImportError: cannot import name 'driver' from 'triton.runtime' (/root/miniconda3/envs/minigptv/lib/python3.9/site-packages/triton/runtime/init.py)
【求助帖】2406期-Ray老师-课程20(LLM->多模态理论+案例讲解),MiniGPT-V2执行时demo_v2.py报错to 【求助帖】2406期-林希老师-课程20(LLM->多模态理论+案例讲解),MiniGPT-V2执行时demo_v2.py报错https://github.com/Vision-CAIR/MiniGPT-4/blob/main/environment.yml 可以基于官方代码库的配置文件排查下软件包的版本问题
感谢老师指点。确实是bitsandbytes的版本与environment.yaml里的不一致。老师课件里script用的是pip install -U bitsandbytes,这会安装到0.45.2版本，但其实应该使用0.37.0版本。