Jin Yao

icon indicating a website link https://yaojin17.github.io/ icon indicating an email [email protected]

icon company University of Virginia

Results 13 comments of Jin Yao

Are the ground-truth poses correct?

@flamehaze1115 Super Thanks! Your response has saved me a lot of time!

Question of ValueError in Annotation2Mesh.py

Same question! Have you find the solution? If so, could you share it with me? @Wspiron Thank you so much!

cannot import name 'data_dir' from 'helpers'

same issue

could you provide the rendering scripts to evalutate ?

Same problem. Have you figured out the correct fov and camera distance for rendering? @Costwen

Cannot import name 'UnencryptedCookieSessionFactoryConfig' from 'pyramid.session' (unknown location) when run “run_unlearn.py --target_model_name_or_path ../../models/Yi-6B --per_device_train_batch_size 1 --do_unlearn --output_dir ./output --overwrite_output_dir --num_train_epochs 1 --logging_steps 1 --learning_rate 2e-5 --warmup_ratio 0.03 --overwrite_cache --save_total_limit 1 --fsdp "full_shard auto_wrap" --fsdp_transformer_layer_cls_to_wrap 'LlamaDecoderLayer' --bf16 True --tf32 True --weight_decay 0. --lr_scheduler_type "cosine" --domain github --gradient_accumulation_steps 85 --unlearn_method gradient_ascent “

I haven't encountered this issue before. Please ensure you're using `transformers==4.36.0`. On my end, the pyramid version I'm using is 2.0.2. You might want to check if you're using the...

Cannot import name 'UnencryptedCookieSessionFactoryConfig' from 'pyramid.session' (unknown location) when run “run_unlearn.py --target_model_name_or_path ../../models/Yi-6B --per_device_train_batch_size 1 --do_unlearn --output_dir ./output --overwrite_output_dir --num_train_epochs 1 --logging_steps 1 --learning_rate 2e-5 --warmup_ratio 0.03 --overwrite_cache --save_total_limit 1 --fsdp "full_shard auto_wrap" --fsdp_transformer_layer_cls_to_wrap 'LlamaDecoderLayer' --bf16 True --tf32 True --weight_decay 0. --lr_scheduler_type "cosine" --domain github --gradient_accumulation_steps 85 --unlearn_method gradient_ascent “

``` name: base channels: - conda-forge - https://repo.anaconda.com/pkgs/main - defaults dependencies: - _libgcc_mutex=0.1=main - _openmp_mutex=5.1=1_gnu - asttokens=2.4.1=pyhd8ed1ab_0 - brotlipy=0.7.0=py310h7f8727e_1002 - bzip2=1.0.8=h7b6447c_0 - ca-certificates=2024.2.2=hbcca054_0 - certifi=2024.2.2=pyhd8ed1ab_0 - cffi=1.15.1=py310h5eee18b_3 - charset-normalizer=2.0.4=pyhd3eb1b0_0 -...

Cannot import name 'UnencryptedCookieSessionFactoryConfig' from 'pyramid.session' (unknown location) when run “run_unlearn.py --target_model_name_or_path ../../models/Yi-6B --per_device_train_batch_size 1 --do_unlearn --output_dir ./output --overwrite_output_dir --num_train_epochs 1 --logging_steps 1 --learning_rate 2e-5 --warmup_ratio 0.03 --overwrite_cache --save_total_limit 1 --fsdp "full_shard auto_wrap" --fsdp_transformer_layer_cls_to_wrap 'LlamaDecoderLayer' --bf16 True --tf32 True --weight_decay 0. --lr_scheduler_type "cosine" --domain github --gradient_accumulation_steps 85 --unlearn_method gradient_ascent “

I updated the above package list since I found the records of my experiments. I used CUDA 11.8 and A800 GPU.

The new version of Pytorch needs to specify the tensor location

Apologies for the late response. Has the issue been resolved? I'm using `torch==2.1.2+cu118 `on my end.

The new version of Pytorch needs to specify the tensor location

``` name: base channels: - conda-forge - https://repo.anaconda.com/pkgs/main - defaults dependencies: - _libgcc_mutex=0.1=main - _openmp_mutex=5.1=1_gnu - asttokens=2.4.1=pyhd8ed1ab_0 - brotlipy=0.7.0=py310h7f8727e_1002 - bzip2=1.0.8=h7b6447c_0 - ca-certificates=2024.2.2=hbcca054_0 - certifi=2024.2.2=pyhd8ed1ab_0 - cffi=1.15.1=py310h5eee18b_3 - charset-normalizer=2.0.4=pyhd3eb1b0_0 -...

How much video memory is needed for this experiment？

Hi, I use 8 A100 GPUs with 80GB of memory each to fine-tune the model. For your case, I suggest using FP16 training and reducing the number of LoRA trainable...