GroundingDINO text input only fixed 80 categoly in COCO

[FengheTan9]

Thanks for your code for finetuning GroundDINO.

I found that the input text is not a real caption but a fixed category. 😢 If so, will a new fine-tuning script be updated? 😉

Running Commend:
python tools/train.py configs/grounding_dino/grounding_dino_swin-t_finetune_16xb2_1x_coco.py

System environment:
sys.platform: linux
Python: 3.7.16 (default, Jan 17 2023, 22:20:44) [GCC 11.2.0]
CUDA available: True
numpy_random_seed: 997378063
GPU 0,1,2,3,4,5,6,7: NVIDIA GeForce RTX 3090
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 11.1, V11.1.74
GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
PyTorch: 1.13.0+cu117
PyTorch compiling details: PyTorch built with:

• GCC 9.3

• C++ Version: 201402

• Intel(R) Math Kernel Library Version 2020.0.0 Product Build 20191122 for Intel(R) 64 architecture applications

• Intel(R) MKL-DNN v2.6.0 (Git Hash 52b5f107dd9cf10910aaa19cb47f3abf9b349815)

• OpenMP 201511 (a.k.a. OpenMP 4.5)

• LAPACK is enabled (usually provided by MKL)

• NNPACK is enabled

• CPU capability usage: AVX2

• CUDA Runtime 11.7

• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86

• CuDNN 8.5

• Magma 2.6.1

• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -fopenmp -DNDEBUG -DUSE_KINETO -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -DEDGE_PROFILER_USE_KINETO -O2 -fPIC -Wno-narrowing -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=1.13.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=ON, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF,

  TorchVision: 0.14.0+cu117
  OpenCV: 4.6.0
  MMEngine: 0.8.5

Runtime environment:
cudnn_benchmark: False
mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0}
dist_cfg: {'backend': 'nccl'}
seed: 997378063
Distributed launcher: none
Distributed training: False
GPU number: 1

[ws1hope]

Do you mean that his input text here only has category information, like people, cars, etc., instead of adding further text descriptions to a category like in glip (e.g. adding suffix information)?

[hhaAndroid]

@FengheTan9 I didn't understand what you meant.

[hhaAndroid]

@FengheTan9 #11012

[FengheTan9]

@FengheTan9 I didn't understand what you meant.

The input text is only the corresponding image category (e.g. dog), and there is no additional descriptive text information (e.g. a dog in the car)

[CDchenlin]

@FengheTan9 You can define the text information in the metainfo, and it will be passed to the model during the training.

[RYHSmmc]

Same question with @FengheTan9 , @CDchenlin can you give some specific advice?