Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

mclight推理readme中的预训练模型报错KeyError: 'downsample' #48

Open
huiyiygy opened this issue Sep 24, 2024 · 1 comment
Open

Comments

@huiyiygy
Copy link

您好,我使用mclight/skydome_lighting下Readme文件中提供的STAGE_2_BEST_CKPT,对waymo数据集进行推理时会触发多个keyError
运行命令如下

python mc_to_sky/tools/infer_waymo_batch.py \
-y mc_to_sky/config/stage2/multi_view_avg.yaml \
-c ./logs/stage2_log/lightning_logs/version_0/checkpoints/epoch=269-val_loss=0.10.ckpt \
--waymo_scenes_dir ./data/waymo_skydome \
-o  ./visualization_results

在mc_to_sky/model/skymodel_enhanced.py文件中报错提示

KeyError: 'downsample'

我看了下原因是multi_view_avg.yaml配置文件中并没有这个参数。
我尝试过参照其它文件配件中的内容,在multi_view_avg.yaml添加 downsample: 4。但紧接着就会报错提示更多的key缺失。如下图
Dingtalk_20240924165603

哪怕我将缺失的参数,全部手动补齐,并注释掉self.on_train_epoch_start(),在最后读取预训练权重时,也会提示权重中存储的key与模型不匹配。

能麻烦你看下,我哪里没有配置对吗?

@yifanlu0227
Copy link
Owner

Hi! 抱歉回复晚了,这些参数不属于Stage2, 而是属于Stage 1的LDR到HDR的reconstruction。请问您设置了mc_to_sky/config/stage2/multi_view_avg.yaml 中stage 1的ckpt了吗?

sky_model:
core_method: sky_model_enhanced
ckpt_path: "YOUR CKPT HERE, ENDED WITH .ckpt"

Screenshot 2024-09-28 at 17 49 01

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants