Upload folder using huggingface_hub

.gitattributes

@@ -33,3 +33,11 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_134854/vis_data/wandb/run-20240623_134900-1tj91a85/run-1tj91a85.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_135747/vis_data/wandb/run-20240623_135753-haxnz4ri/run-haxnz4ri.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_141103/vis_data/wandb/run-20240623_141111-j57x5qxb/run-j57x5qxb.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_150212/vis_data/wandb/run-20240623_150217-mlt4pykm/run-mlt4pykm.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_151455/vis_data/wandb/run-20240623_151501-nv52dul7/run-nv52dul7.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_153346/vis_data/wandb/run-20240623_153353-q9ka8o44/run-q9ka8o44.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_191953/vis_data/wandb/run-20240623_191959-ljo9rur3/run-ljo9rur3.wandb filter=lfs diff=lfs merge=lfs -text
+llava_ft/20240623_193616/vis_data/wandb/run-20240623_193622-jigkwuih/run-jigkwuih.wandb filter=lfs diff=lfs merge=lfs -text

internvl_ft_run_10_filter/iter_4000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:83c4428c4241b5c7f332d93048ca22713c8b704dab05f6f747be468ed4a705e6
+size 300000

internvl_ft_run_10_filter/last_checkpoint

@@ -0,0 +1 @@
+/root/wangqun/work_dirs/internvl_ft_run_10_filter/iter_4608.pth

internvl_ft_run_12_filter/20250304_104043/20250304_104043.log

@@ -0,0 +1,195 @@
+2025/03/04 10:40:44 - mmengine - DEBUG - An `DeepSpeedStrategy` instance is built from registry, and its implementation can be found in xtuner.engine._strategy.deepspeed
+2025/03/04 10:40:44 - mmengine - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.10.13 (main, Sep 11 2023, 13:44:35) [GCC 11.2.0]
+    CUDA available: True
+    MUSA available: False
+    numpy_random_seed: 1685007727
+    GPU 0: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /usr/local/cuda
+    NVCC: Cuda compilation tools, release 12.2, V12.2.140
+    GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
+    PyTorch: 2.4.1+cu121
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.4.2 (Git Hash 1137e04ec0b5251ca2b4400a4fd3c667ce843d67)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX512
+  - CUDA Runtime 12.1
+  - NVCC architecture flags: -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;-gencode;arch=compute_90,code=sm_90
+  - CuDNN 90.1  (built against CUDA 12.4)
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=9.1.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.4.1, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, 
+
+    TorchVision: 0.19.1+cu121
+    OpenCV: 4.9.0
+    MMEngine: 0.10.6
+
+Runtime environment:
+    launcher: none
+    randomness: {'seed': None, 'deterministic': False}
+    cudnn_benchmark: False
+    mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0}
+    dist_cfg: {'backend': 'nccl'}
+    seed: None
+    deterministic: False
+    Distributed launcher: none
+    Distributed training: False
+    GPU number: 1
+------------------------------------------------------------
+
+2025/03/04 10:40:44 - mmengine - INFO - Config:
+accumulative_counts = 2
+batch_size = 1
+betas = (
+    0.9,
+    0.999,
+)
+custom_hooks = [
+    dict(
+        tokenizer=dict(
+            pretrained_model_name_or_path='/root/models/InternVL2_2B',
+            trust_remote_code=True,
+            type='transformers.AutoTokenizer.from_pretrained'),
+        type='xtuner.engine.hooks.DatasetInfoHook'),
+]
+data_path = '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json'
+data_root = '/root/data/tempData/screenshot_od'
+dataloader_num_workers = 4
+default_hooks = dict(
+    checkpoint=dict(
+        by_epoch=False,
+        interval=1000,
+        max_keep_ckpts=-1,
+        save_optimizer=False,
+        type='mmengine.hooks.CheckpointHook'),
+    logger=dict(
+        interval=10,
+        log_metric_by_epoch=False,
+        type='mmengine.hooks.LoggerHook'),
+    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
+    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
+    timer=dict(type='mmengine.hooks.IterTimerHook'))
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+image_folder = '/root/data/tempData/screenshot_odimages'
+launcher = 'none'
+llava_dataset = dict(
+    data_paths='/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+    image_folders='/root/data/tempData/screenshot_odimages',
+    max_length=8192,
+    model_path='/root/models/InternVL2_2B',
+    template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+    type='xtuner.dataset.InternVL_V1_5_Dataset')
+load_from = None
+log_level = 'DEBUG'
+log_processor = dict(by_epoch=False)
+lr = 2e-05
+max_epochs = 4
+max_length = 8192
+max_norm = 1
+model = dict(
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm_lora=dict(
+        lora_alpha=256,
+        lora_dropout=0.05,
+        r=128,
+        target_modules=None,
+        task_type='CAUSAL_LM',
+        type='peft.LoraConfig'),
+    model_path='/root/models/InternVL2_2B',
+    quantization_llm=True,
+    quantization_vit=False,
+    type='xtuner.model.InternVL_V1_5')
+optim_type = 'torch.optim.AdamW'
+optim_wrapper = dict(
+    optimizer=dict(
+        betas=(
+            0.9,
+            0.999,
+        ),
+        lr=2e-05,
+        type='torch.optim.AdamW',
+        weight_decay=0.05),
+    type='DeepSpeedOptimWrapper')
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=0.12,
+        start_factor=1e-05,
+        type='mmengine.optim.LinearLR'),
+    dict(
+        begin=0.12,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=4,
+        eta_min=0.0,
+        type='mmengine.optim.CosineAnnealingLR'),
+]
+path = '/root/models/InternVL2_2B'
+prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.internlm2_chat'
+randomness = dict(deterministic=False, seed=None)
+resume = False
+runner_type = 'FlexibleRunner'
+save_steps = 1000
+save_total_limit = -1
+strategy = dict(
+    config=dict(
+        bf16=dict(enabled=True),
+        fp16=dict(enabled=False, initial_scale_power=16),
+        gradient_accumulation_steps='auto',
+        gradient_clipping='auto',
+        train_micro_batch_size_per_gpu='auto',
+        zero_allow_untested_optimizer=True,
+        zero_force_ds_cpu_optimizer=False,
+        zero_optimization=dict(overlap_comm=True, stage=2)),
+    exclude_frozen_parameters=True,
+    gradient_accumulation_steps=2,
+    gradient_clipping=1,
+    sequence_parallel_size=1,
+    train_micro_batch_size_per_gpu=1,
+    type='xtuner.engine.DeepSpeedStrategy')
+tokenizer = dict(
+    pretrained_model_name_or_path='/root/models/InternVL2_2B',
+    trust_remote_code=True,
+    type='transformers.AutoTokenizer.from_pretrained')
+train_cfg = dict(max_epochs=4, type='xtuner.engine.runner.TrainLoop')
+train_dataloader = dict(
+    batch_size=1,
+    collate_fn=dict(type='xtuner.dataset.collate_fns.default_collate_fn'),
+    dataset=dict(
+        data_paths=
+        '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+        image_folders='/root/data/tempData/screenshot_odimages',
+        max_length=8192,
+        model_path='/root/models/InternVL2_2B',
+        template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+        type='xtuner.dataset.InternVL_V1_5_Dataset'),
+    num_workers=4,
+    sampler=dict(
+        length_property='modality_length',
+        per_device_batch_size=2,
+        type='xtuner.dataset.samplers.LengthGroupedSampler'))
+visualizer = dict(
+    type='mmengine.visualization.Visualizer',
+    vis_backends=[
+        dict(type='mmengine.visualization.TensorboardVisBackend'),
+    ])
+warmup_ratio = 0.03
+weight_decay = 0.05
+work_dir = '/root/wangqun/work_dirs/internvl_ft_run_12_filter'
+
+2025/03/04 10:40:44 - mmengine - DEBUG - An `TensorboardVisBackend` instance is built from registry, and its implementation can be found in mmengine.visualization.vis_backend
+2025/03/04 10:40:44 - mmengine - DEBUG - An `Visualizer` instance is built from registry, and its implementation can be found in mmengine.visualization.visualizer
+2025/03/04 10:40:44 - mmengine - DEBUG - Attribute `_env_initialized` is not defined in <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'> or `<class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized is False, `_init_env` will be called and <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized will be set to True

internvl_ft_run_12_filter/20250304_104403/20250304_104403.log

@@ -0,0 +1,203 @@
+2025/03/04 10:44:03 - mmengine - DEBUG - An `DeepSpeedStrategy` instance is built from registry, and its implementation can be found in xtuner.engine._strategy.deepspeed
+2025/03/04 10:44:03 - mmengine - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.10.13 (main, Sep 11 2023, 13:44:35) [GCC 11.2.0]
+    CUDA available: True
+    MUSA available: False
+    numpy_random_seed: 1125505633
+    GPU 0: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /usr/local/cuda
+    NVCC: Cuda compilation tools, release 12.2, V12.2.140
+    GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
+    PyTorch: 2.4.1+cu121
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.4.2 (Git Hash 1137e04ec0b5251ca2b4400a4fd3c667ce843d67)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX512
+  - CUDA Runtime 12.1
+  - NVCC architecture flags: -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;-gencode;arch=compute_90,code=sm_90
+  - CuDNN 90.1  (built against CUDA 12.4)
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=9.1.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.4.1, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, 
+
+    TorchVision: 0.19.1+cu121
+    OpenCV: 4.9.0
+    MMEngine: 0.10.6
+
+Runtime environment:
+    launcher: none
+    randomness: {'seed': None, 'deterministic': False}
+    cudnn_benchmark: False
+    mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0}
+    dist_cfg: {'backend': 'nccl'}
+    seed: None
+    deterministic: False
+    Distributed launcher: none
+    Distributed training: False
+    GPU number: 1
+------------------------------------------------------------
+
+2025/03/04 10:44:04 - mmengine - INFO - Config:
+accumulative_counts = 2
+batch_size = 1
+betas = (
+    0.9,
+    0.999,
+)
+custom_hooks = [
+    dict(
+        tokenizer=dict(
+            pretrained_model_name_or_path='/root/models/InternVL2_2B',
+            trust_remote_code=True,
+            type='transformers.AutoTokenizer.from_pretrained'),
+        type='xtuner.engine.hooks.DatasetInfoHook'),
+]
+data_path = '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json'
+data_root = '/root/data/tempData/screenshot_od'
+dataloader_num_workers = 4
+default_hooks = dict(
+    checkpoint=dict(
+        by_epoch=False,
+        interval=1000,
+        max_keep_ckpts=-1,
+        save_optimizer=False,
+        type='mmengine.hooks.CheckpointHook'),
+    logger=dict(
+        interval=10,
+        log_metric_by_epoch=False,
+        type='mmengine.hooks.LoggerHook'),
+    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
+    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
+    timer=dict(type='mmengine.hooks.IterTimerHook'))
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+image_folder = '/root/data/tempData/screenshot_odimages'
+launcher = 'none'
+llava_dataset = dict(
+    data_paths='/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+    image_folders='/root/data/tempData/screenshot_odimages',
+    max_length=8192,
+    model_path='/root/models/InternVL2_2B',
+    template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+    type='xtuner.dataset.InternVL_V1_5_Dataset')
+load_from = None
+log_level = 'DEBUG'
+log_processor = dict(by_epoch=False)
+lr = 2e-05
+max_epochs = 4
+max_length = 8192
+max_norm = 1
+model = dict(
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm_lora=dict(
+        lora_alpha=256,
+        lora_dropout=0.05,
+        r=128,
+        target_modules=None,
+        task_type='CAUSAL_LM',
+        type='peft.LoraConfig'),
+    model_path='/root/models/InternVL2_2B',
+    quantization_llm=True,
+    quantization_vit=False,
+    type='xtuner.model.InternVL_V1_5')
+optim_type = 'torch.optim.AdamW'
+optim_wrapper = dict(
+    optimizer=dict(
+        betas=(
+            0.9,
+            0.999,
+        ),
+        lr=2e-05,
+        type='torch.optim.AdamW',
+        weight_decay=0.05),
+    type='DeepSpeedOptimWrapper')
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=0.12,
+        start_factor=1e-05,
+        type='mmengine.optim.LinearLR'),
+    dict(
+        begin=0.12,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=4,
+        eta_min=0.0,
+        type='mmengine.optim.CosineAnnealingLR'),
+]
+path = '/root/models/InternVL2_2B'
+prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.internlm2_chat'
+randomness = dict(deterministic=False, seed=None)
+resume = False
+runner_type = 'FlexibleRunner'
+save_steps = 1000
+save_total_limit = -1
+strategy = dict(
+    config=dict(
+        bf16=dict(enabled=True),
+        fp16=dict(enabled=False, initial_scale_power=16),
+        gradient_accumulation_steps='auto',
+        gradient_clipping='auto',
+        train_micro_batch_size_per_gpu='auto',
+        zero_allow_untested_optimizer=True,
+        zero_force_ds_cpu_optimizer=False,
+        zero_optimization=dict(overlap_comm=True, stage=2)),
+    exclude_frozen_parameters=True,
+    gradient_accumulation_steps=2,
+    gradient_clipping=1,
+    sequence_parallel_size=1,
+    train_micro_batch_size_per_gpu=1,
+    type='xtuner.engine.DeepSpeedStrategy')
+tokenizer = dict(
+    pretrained_model_name_or_path='/root/models/InternVL2_2B',
+    trust_remote_code=True,
+    type='transformers.AutoTokenizer.from_pretrained')
+train_cfg = dict(max_epochs=4, type='xtuner.engine.runner.TrainLoop')
+train_dataloader = dict(
+    batch_size=1,
+    collate_fn=dict(type='xtuner.dataset.collate_fns.default_collate_fn'),
+    dataset=dict(
+        data_paths=
+        '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+        image_folders='/root/data/tempData/screenshot_odimages',
+        max_length=8192,
+        model_path='/root/models/InternVL2_2B',
+        template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+        type='xtuner.dataset.InternVL_V1_5_Dataset'),
+    num_workers=4,
+    sampler=dict(
+        length_property='modality_length',
+        per_device_batch_size=2,
+        type='xtuner.dataset.samplers.LengthGroupedSampler'))
+visualizer = dict(
+    type='mmengine.visualization.Visualizer',
+    vis_backends=[
+        dict(type='mmengine.visualization.TensorboardVisBackend'),
+    ])
+warmup_ratio = 0.03
+weight_decay = 0.05
+work_dir = '/root/wangqun/work_dirs/internvl_ft_run_12_filter'
+
+2025/03/04 10:44:04 - mmengine - DEBUG - An `TensorboardVisBackend` instance is built from registry, and its implementation can be found in mmengine.visualization.vis_backend
+2025/03/04 10:44:04 - mmengine - DEBUG - An `Visualizer` instance is built from registry, and its implementation can be found in mmengine.visualization.visualizer
+2025/03/04 10:44:04 - mmengine - DEBUG - Attribute `_env_initialized` is not defined in <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'> or `<class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized is False, `_init_env` will be called and <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized will be set to True
+2025/03/04 10:44:05 - mmengine - DEBUG - Get class `RuntimeInfoHook` from "hook" registry in "mmengine"
+2025/03/04 10:44:05 - mmengine - DEBUG - An `RuntimeInfoHook` instance is built from registry, and its implementation can be found in mmengine.hooks.runtime_info_hook
+2025/03/04 10:44:05 - mmengine - DEBUG - An `IterTimerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.iter_timer_hook
+2025/03/04 10:44:05 - mmengine - DEBUG - An `DistSamplerSeedHook` instance is built from registry, and its implementation can be found in mmengine.hooks.sampler_seed_hook
+2025/03/04 10:44:05 - mmengine - DEBUG - An `LoggerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.logger_hook
+2025/03/04 10:44:05 - mmengine - DEBUG - An `ParamSchedulerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.param_scheduler_hook
+2025/03/04 10:44:05 - mmengine - DEBUG - An `CheckpointHook` instance is built from registry, and its implementation can be found in mmengine.hooks.checkpoint_hook
+2025/03/04 10:44:05 - mmengine - WARNING - Failed to search registry with scope "mmengine" in the "builder" registry tree. As a workaround, the current "builder" registry in "xtuner" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmengine" is a correct scope, or whether the registry is initialized.

internvl_ft_run_12_filter/20250304_104403/vis_data/events.out.tfevents.1741056245.intern-studio-40019814.17053.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a1c288ad221462a4a8a981e1feef2141f61fb02c92b0d81c47078d4d277a3fa4
+size 4845

internvl_ft_run_12_filter/20250304_111639/vis_data/events.out.tfevents.1741058201.intern-studio-40019814.23140.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6c197cbc5fe46cd6ee94a44546a7ff96eed17a9f96594b200c09a174483bdf28
+size 4983

internvl_ft_run_12_filter/20250304_112305/vis_data/events.out.tfevents.1741058586.intern-studio-40019814.25156.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a3ecc6d8f1eb6882b13ebc2091216762d530789a91dd5d211545f065c08500c1
+size 4983

internvl_ft_run_12_filter/20250304_112538/vis_data/events.out.tfevents.1741058739.intern-studio-40019814.26649.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be6291cbd77f2207ec598e9e75e13fab608b52277bea94cf9dda0e5d99ff87c1
+size 4987

internvl_ft_run_12_filter/20250304_113017/20250304_113017.log

@@ -0,0 +1,633 @@
+2025/03/04 11:30:18 - mmengine - DEBUG - An `DeepSpeedStrategy` instance is built from registry, and its implementation can be found in xtuner.engine._strategy.deepspeed
+2025/03/04 11:30:18 - mmengine - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.10.13 (main, Sep 11 2023, 13:44:35) [GCC 11.2.0]
+    CUDA available: True
+    MUSA available: False
+    numpy_random_seed: 1251330375
+    GPU 0: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /usr/local/cuda
+    NVCC: Cuda compilation tools, release 12.2, V12.2.140
+    GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
+    PyTorch: 2.4.1+cu121
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.4.2 (Git Hash 1137e04ec0b5251ca2b4400a4fd3c667ce843d67)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX512
+  - CUDA Runtime 12.1
+  - NVCC architecture flags: -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;-gencode;arch=compute_90,code=sm_90
+  - CuDNN 90.1  (built against CUDA 12.4)
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=9.1.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.4.1, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, 
+
+    TorchVision: 0.19.1+cu121
+    OpenCV: 4.9.0
+    MMEngine: 0.10.6
+
+Runtime environment:
+    launcher: none
+    randomness: {'seed': None, 'deterministic': False}
+    cudnn_benchmark: False
+    mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0}
+    dist_cfg: {'backend': 'nccl'}
+    seed: None
+    deterministic: False
+    Distributed launcher: none
+    Distributed training: False
+    GPU number: 1
+------------------------------------------------------------
+
+2025/03/04 11:30:18 - mmengine - INFO - Config:
+accumulative_counts = 2
+batch_size = 1
+betas = (
+    0.9,
+    0.999,
+)
+custom_hooks = [
+    dict(
+        tokenizer=dict(
+            pretrained_model_name_or_path=
+            '/root/share/new_models/OpenGVLab/InternVL2-2B',
+            trust_remote_code=True,
+            type='transformers.AutoTokenizer.from_pretrained'),
+        type='xtuner.engine.hooks.DatasetInfoHook'),
+]
+data_path = '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json'
+dataloader_num_workers = 4
+default_hooks = dict(
+    checkpoint=dict(
+        by_epoch=False,
+        interval=1000,
+        max_keep_ckpts=-1,
+        save_optimizer=False,
+        type='mmengine.hooks.CheckpointHook'),
+    logger=dict(
+        interval=10,
+        log_metric_by_epoch=False,
+        type='mmengine.hooks.LoggerHook'),
+    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
+    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
+    timer=dict(type='mmengine.hooks.IterTimerHook'))
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+image_folder = '/'
+launcher = 'none'
+llava_dataset = dict(
+    data_paths='/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+    image_folders='/',
+    max_length=8192,
+    model_path='/root/share/new_models/OpenGVLab/InternVL2-2B',
+    template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+    type='xtuner.dataset.InternVL_V1_5_Dataset')
+load_from = None
+log_level = 'DEBUG'
+log_processor = dict(by_epoch=False)
+lr = 2e-05
+max_epochs = 4
+max_length = 8192
+max_norm = 1
+model = dict(
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm_lora=dict(
+        lora_alpha=256,
+        lora_dropout=0.05,
+        r=128,
+        target_modules=None,
+        task_type='CAUSAL_LM',
+        type='peft.LoraConfig'),
+    model_path='/root/share/new_models/OpenGVLab/InternVL2-2B',
+    quantization_llm=True,
+    quantization_vit=False,
+    type='xtuner.model.InternVL_V1_5')
+optim_type = 'torch.optim.AdamW'
+optim_wrapper = dict(
+    optimizer=dict(
+        betas=(
+            0.9,
+            0.999,
+        ),
+        lr=2e-05,
+        type='torch.optim.AdamW',
+        weight_decay=0.05),
+    type='DeepSpeedOptimWrapper')
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=0.12,
+        start_factor=1e-05,
+        type='mmengine.optim.LinearLR'),
+    dict(
+        begin=0.12,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=4,
+        eta_min=0.0,
+        type='mmengine.optim.CosineAnnealingLR'),
+]
+path = '/root/share/new_models/OpenGVLab/InternVL2-2B'
+prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.internlm2_chat'
+randomness = dict(deterministic=False, seed=None)
+resume = False
+runner_type = 'FlexibleRunner'
+save_steps = 1000
+save_total_limit = -1
+strategy = dict(
+    config=dict(
+        bf16=dict(enabled=True),
+        fp16=dict(enabled=False, initial_scale_power=16),
+        gradient_accumulation_steps='auto',
+        gradient_clipping='auto',
+        train_micro_batch_size_per_gpu='auto',
+        zero_allow_untested_optimizer=True,
+        zero_force_ds_cpu_optimizer=False,
+        zero_optimization=dict(overlap_comm=True, stage=2)),
+    exclude_frozen_parameters=True,
+    gradient_accumulation_steps=2,
+    gradient_clipping=1,
+    sequence_parallel_size=1,
+    train_micro_batch_size_per_gpu=1,
+    type='xtuner.engine.DeepSpeedStrategy')
+tokenizer = dict(
+    pretrained_model_name_or_path=
+    '/root/share/new_models/OpenGVLab/InternVL2-2B',
+    trust_remote_code=True,
+    type='transformers.AutoTokenizer.from_pretrained')
+train_cfg = dict(max_epochs=4, type='xtuner.engine.runner.TrainLoop')
+train_dataloader = dict(
+    batch_size=1,
+    collate_fn=dict(type='xtuner.dataset.collate_fns.default_collate_fn'),
+    dataset=dict(
+        data_paths=
+        '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+        image_folders='/',
+        max_length=8192,
+        model_path='/root/share/new_models/OpenGVLab/InternVL2-2B',
+        template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+        type='xtuner.dataset.InternVL_V1_5_Dataset'),
+    num_workers=4,
+    sampler=dict(
+        length_property='modality_length',
+        per_device_batch_size=2,
+        type='xtuner.dataset.samplers.LengthGroupedSampler'))
+visualizer = dict(
+    type='mmengine.visualization.Visualizer',
+    vis_backends=[
+        dict(type='mmengine.visualization.TensorboardVisBackend'),
+    ])
+warmup_ratio = 0.03
+weight_decay = 0.05
+work_dir = '/root/wangqun/work_dirs/internvl_ft_run_12_filter'
+
+2025/03/04 11:30:18 - mmengine - DEBUG - An `TensorboardVisBackend` instance is built from registry, and its implementation can be found in mmengine.visualization.vis_backend
+2025/03/04 11:30:18 - mmengine - DEBUG - An `Visualizer` instance is built from registry, and its implementation can be found in mmengine.visualization.visualizer
+2025/03/04 11:30:18 - mmengine - DEBUG - Attribute `_env_initialized` is not defined in <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'> or `<class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized is False, `_init_env` will be called and <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized will be set to True
+2025/03/04 11:30:19 - mmengine - DEBUG - Get class `RuntimeInfoHook` from "hook" registry in "mmengine"
+2025/03/04 11:30:19 - mmengine - DEBUG - An `RuntimeInfoHook` instance is built from registry, and its implementation can be found in mmengine.hooks.runtime_info_hook
+2025/03/04 11:30:19 - mmengine - DEBUG - An `IterTimerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.iter_timer_hook
+2025/03/04 11:30:19 - mmengine - DEBUG - An `DistSamplerSeedHook` instance is built from registry, and its implementation can be found in mmengine.hooks.sampler_seed_hook
+2025/03/04 11:30:19 - mmengine - DEBUG - An `LoggerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.logger_hook
+2025/03/04 11:30:19 - mmengine - DEBUG - An `ParamSchedulerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.param_scheduler_hook
+2025/03/04 11:30:19 - mmengine - DEBUG - An `CheckpointHook` instance is built from registry, and its implementation can be found in mmengine.hooks.checkpoint_hook
+2025/03/04 11:30:19 - mmengine - WARNING - Failed to search registry with scope "mmengine" in the "builder" registry tree. As a workaround, the current "builder" registry in "xtuner" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmengine" is a correct scope, or whether the registry is initialized.
+2025/03/04 11:30:19 - mmengine - DEBUG - An `from_pretrained` instance is built from registry, and its implementation can be found in transformers.models.auto.tokenization_auto
+2025/03/04 11:30:19 - mmengine - DEBUG - An `DatasetInfoHook` instance is built from registry, and its implementation can be found in xtuner.engine.hooks.dataset_info_hook
+2025/03/04 11:30:19 - mmengine - INFO - Hooks will be executed in the following order:
+before_run:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+before_train:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(NORMAL      ) DatasetInfoHook                    
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+before_train_epoch:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(NORMAL      ) DistSamplerSeedHook                
+ -------------------- 
+before_train_iter:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+after_train_iter:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+(LOW         ) ParamSchedulerHook                 
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+after_train_epoch:
+(NORMAL      ) IterTimerHook                      
+(LOW         ) ParamSchedulerHook                 
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+before_val:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) DatasetInfoHook                    
+ -------------------- 
+before_val_epoch:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+before_val_iter:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+after_val_iter:
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+after_val_epoch:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+(LOW         ) ParamSchedulerHook                 
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+after_val:
+(VERY_HIGH   ) RuntimeInfoHook                    
+ -------------------- 
+after_train:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+before_test:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) DatasetInfoHook                    
+ -------------------- 
+before_test_epoch:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+before_test_iter:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+after_test_iter:
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+after_test_epoch:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+after_test:
+(VERY_HIGH   ) RuntimeInfoHook                    
+ -------------------- 
+after_run:
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+2025/03/04 11:30:19 - mmengine - DEBUG - An `FlexibleRunner` instance is built from registry, its implementation can be found inmmengine.runner._flexible_runner
+2025/03/04 11:30:19 - mmengine - INFO - Starting to loading data and calc length
+2025/03/04 11:30:19 - mmengine - INFO - =======Starting to process /root/data/tempData/screenshot_od/layout_ocr_multi_scale.json =======
+2025/03/04 11:30:26 - mmengine - INFO - =======total 4806 samples of /root/data/tempData/screenshot_od/layout_ocr_multi_scale.json=======
+2025/03/04 11:30:26 - mmengine - INFO - end loading data and calc length
+2025/03/04 11:30:26 - mmengine - INFO - =======total 4806 samples=======
+2025/03/04 11:30:26 - mmengine - DEBUG - An `InternVL_V1_5_Dataset` instance is built from registry, and its implementation can be found in xtuner.dataset.internvl_dataset
+2025/03/04 11:30:26 - mmengine - INFO - LengthGroupedSampler is used.
+2025/03/04 11:30:26 - mmengine - INFO - LengthGroupedSampler construction is complete, and the selected attribute is modality_length
+2025/03/04 11:30:26 - mmengine - DEBUG - An `LengthGroupedSampler` instance is built from registry, and its implementation can be found in xtuner.dataset.samplers.length_grouped
+2025/03/04 11:30:26 - mmengine - WARNING - Dataset InternVL_V1_5_Dataset has no metainfo. ``dataset_meta`` in visualizer will be None.
+2025/03/04 11:30:26 - mmengine - DEBUG - An `TrainLoop` instance is built from registry, and its implementation can be found in xtuner.engine.runner.loops
+2025/03/04 11:30:26 - mmengine - INFO - Start to load InternVL_V1_5 model.
+2025/03/04 11:30:26 - mmengine - DEBUG - Get class `BaseDataPreprocessor` from "model" registry in "mmengine"
+2025/03/04 11:30:26 - mmengine - DEBUG - An `BaseDataPreprocessor` instance is built from registry, and its implementation can be found in mmengine.model.base_model.data_preprocessor
+2025/03/04 11:30:32 - mmengine - DEBUG - An `LoraConfig` instance is built from registry, and its implementation can be found in peft.tuners.lora.config
+2025/03/04 11:30:34 - mmengine - INFO - InternVL_V1_5(
+  (data_preprocessor): BaseDataPreprocessor()
+  (model): InternVLChatModel(
+    (vision_model): InternVisionModel(
+      (embeddings): InternVisionEmbeddings(
+        (patch_embedding): Conv2d(3, 1024, kernel_size=(14, 14), stride=(14, 14))
+      )
+      (encoder): InternVisionEncoder(
+        (layers): ModuleList(
+          (0-23): 24 x InternVisionEncoderLayer(
+            (attn): InternAttention(
+              (qkv): Linear(in_features=1024, out_features=3072, bias=True)
+              (attn_drop): Dropout(p=0.0, inplace=False)
+              (proj_drop): Dropout(p=0.0, inplace=False)
+              (proj): Linear(in_features=1024, out_features=1024, bias=True)
+            )
+            (mlp): InternMLP(
+              (act): GELUActivation()
+              (fc1): Linear(in_features=1024, out_features=4096, bias=True)
+              (fc2): Linear(in_features=4096, out_features=1024, bias=True)
+            )
+            (norm1): LayerNorm((1024,), eps=1e-06, elementwise_affine=True)
+            (norm2): LayerNorm((1024,), eps=1e-06, elementwise_affine=True)
+            (drop_path1): Identity()
+            (drop_path2): Identity()
+          )
+        )
+      )
+    )
+    (language_model): PeftModelForCausalLM(
+      (base_model): LoraModel(
+        (model): InternLM2ForCausalLM(
+          (model): InternLM2Model(
+            (tok_embeddings): Embedding(92553, 2048, padding_idx=2)
+            (layers): ModuleList(
+              (0-23): 24 x InternLM2DecoderLayer(
+                (attention): InternLM2Attention(
+                  (wqkv): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=4096, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=4096, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (wo): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=2048, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=2048, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (rotary_emb): InternLM2DynamicNTKScalingRotaryEmbedding()
+                )
+                (feed_forward): InternLM2MLP(
+                  (w1): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=8192, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=8192, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (w3): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=8192, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=8192, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (w2): lora.Linear(
+                    (base_layer): Linear4bit(in_features=8192, out_features=2048, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=8192, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=2048, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (act_fn): SiLU()
+                )
+                (attention_norm): InternLM2RMSNorm()
+                (ffn_norm): InternLM2RMSNorm()
+              )
+            )
+            (norm): InternLM2RMSNorm()
+          )
+          (output): lora.Linear(
+            (base_layer): Linear4bit(in_features=2048, out_features=92553, bias=False)
+            (lora_dropout): ModuleDict(
+              (default): Dropout(p=0.05, inplace=False)
+            )
+            (lora_A): ModuleDict(
+              (default): Linear(in_features=2048, out_features=128, bias=False)
+            )
+            (lora_B): ModuleDict(
+              (default): Linear(in_features=128, out_features=92553, bias=False)
+            )
+            (lora_embedding_A): ParameterDict()
+            (lora_embedding_B): ParameterDict()
+            (lora_magnitude_vector): ModuleDict()
+          )
+        )
+      )
+    )
+    (mlp1): Sequential(
+      (0): LayerNorm((4096,), eps=1e-05, elementwise_affine=True)
+      (1): Linear(in_features=4096, out_features=2048, bias=True)
+      (2): GELU(approximate='none')
+      (3): Linear(in_features=2048, out_features=2048, bias=True)
+    )
+  )
+)
+2025/03/04 11:30:34 - mmengine - INFO - InternVL_V1_5 construction is complete
+2025/03/04 11:30:34 - mmengine - DEBUG - An `InternVL_V1_5` instance is built from registry, and its implementation can be found in xtuner.model.internvl
+2025/03/04 11:30:34 - mmengine - DEBUG - Get class `DefaultOptimWrapperConstructor` from "optimizer wrapper constructor" registry in "mmengine"
+2025/03/04 11:30:34 - mmengine - DEBUG - An `DefaultOptimWrapperConstructor` instance is built from registry, and its implementation can be found in mmengine.optim.optimizer.default_constructor
+2025/03/04 11:30:34 - mmengine - DEBUG - An `AdamW` instance is built from registry, and its implementation can be found in torch.optim.adamw
+2025/03/04 11:30:34 - mmengine - DEBUG - Get class `DeepSpeedOptimWrapper` from "optim_wrapper" registry in "mmengine"
+2025/03/04 11:30:34 - mmengine - DEBUG - An `DeepSpeedOptimWrapper` instance is built from registry, and its implementation can be found in mmengine._strategy.deepspeed
+2025/03/04 11:30:36 - mmengine - DEBUG - The `end` of <class 'mmengine.optim.scheduler.lr_scheduler.LinearLR'> is not set. Use the max epochs/iters of train loop as default.
+2025/03/04 11:30:36 - mmengine - DEBUG - The `end` of <class 'mmengine.optim.scheduler.lr_scheduler.CosineAnnealingLR'> is not set. Use the max epochs/iters of train loop as default.
+2025/03/04 11:30:36 - mmengine - INFO - Num train samples 4806
+2025/03/04 11:30:36 - mmengine - INFO - train example:
+2025/03/04 11:30:36 - mmengine - INFO -  <s><|im_start|> system
+You are an AI assistant whose name is InternLM (书生·浦语).<|im_end|><|im_start|>user
+ <img> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> <IMG_CONTEXT> </img>  
+请从这张聊天截图中提取结构化信息<|im_end|><|im_start|> assistant
+{
+  "dialog_name": "<对方正在输入...",
+  "conversation": [
+    {
+      "timestamp": "",
+      "speaker": "<对方正在输入...",
+      "content": "不是",
+      "message_bbox": {
+        "min_x": 855,
+        "max_x": 923,
+        "min_y": 236,
+        "max_y": 269
+      },
+      "image": "",
+      "transfer": [],
+      "file": []
+    },
+    {
+      "timestamp": "",
+      "speaker": "<对方正在输入...",
+      "content": "在淘宝里",
+      "message_bbox": {
+        "min_x": 783,
+        "max_x": 921,
+        "min_y": 345,
+        "max_y": 377
+      },
+      "image": "",
+      "transfer": [],
+      "file": []
+    },
+    {
+      "timestamp": "",
+      "speaker": "<对方正在输入...",
+      "content": "不能发微信",
+      "message_bbox": {
+        "min_x": 747,
+        "max_x": 923,
+        "min_y": 452,
+        "max_y": 486
+      },
+      "image": "",
+      "transfer": [],
+      "file": []
+    },
+    {
+      "timestamp": "",
+      "speaker": "<对方正在输入...",
+      "content": "两字",
+      "message_bbox": {
+        "min_x": 854,
+        "max_x": 922,
+        "min_y": 560,
+        "max_y": 594
+      },
+      "image": "",
+      "transfer": [],
+      "file": []
+    },
+    {
+      "timestamp": "",
+      "speaker": "<对方正在输入...",
+      "content": "微信",
+      "message_bbox": {
+        "min_x": 854,
+        "max_x": 924,
+        "min_y": 670,
+        "max_y": 702
+      },
+      "image": "",
+      "transfer": [],
+      "file": []
+    },
+    {
+      "timestamp": "",
+      "speaker": "<对方正在输入...",
+      "content": "①微信",
+      "message_bbox": {
+        "min_x": 788,
+        "max_x": 922,
+        "min_y": 777,
+        "max_y": 811
+      },
+      "image": "",
+      "transfer": [],
+      "file": []
+    }
+  ]
+}<|im_end|>
+2025/03/04 11:30:36 - mmengine - WARNING - "FileClient" will be deprecated in future. Please use io functions in https://mmengine.readthedocs.io/en/latest/api/fileio.html#file-io
+2025/03/04 11:30:36 - mmengine - WARNING - "HardDiskBackend" is the alias of "LocalBackend" and the former will be deprecated in future.
+2025/03/04 11:30:36 - mmengine - INFO - Checkpoints will be saved to /root/wangqun/work_dirs/internvl_ft_run_12_filter.
+2025/03/04 11:31:22 - mmengine - INFO - Iter(train) [   10/19224]  lr: 3.1324e-07  eta: 1 day, 0:14:22  time: 4.5416  data_time: 0.0107  memory: 17874  loss: 0.4848
+2025/03/04 11:31:37 - mmengine - INFO - Iter(train) [   20/19224]  lr: 6.6106e-07  eta: 16:10:09  time: 1.5206  data_time: 0.0143  memory: 11809  loss: 0.5176
+2025/03/04 11:31:52 - mmengine - INFO - Iter(train) [   30/19224]  lr: 1.0089e-06  eta: 13:23:24  time: 1.4721  data_time: 0.0109  memory: 11450  loss: 0.5703
+2025/03/04 11:32:06 - mmengine - INFO - Iter(train) [   40/19224]  lr: 1.3567e-06  eta: 11:59:30  time: 1.4671  data_time: 0.0111  memory: 11364  loss: 0.4955
+2025/03/04 11:32:20 - mmengine - INFO - Iter(train) [   50/19224]  lr: 1.7045e-06  eta: 11:03:04  time: 1.3731  data_time: 0.0113  memory: 11364  loss: 0.6113
+2025/03/04 11:32:34 - mmengine - INFO - Iter(train) [   60/19224]  lr: 2.0524e-06  eta: 10:26:18  time: 1.3907  data_time: 0.0112  memory: 11117  loss: 0.6183
+2025/03/04 11:32:47 - mmengine - INFO - Iter(train) [   70/19224]  lr: 2.4002e-06  eta: 9:57:44  time: 1.3416  data_time: 0.0109  memory: 11027  loss: 0.6026
+2025/03/04 11:33:00 - mmengine - INFO - Iter(train) [   80/19224]  lr: 2.7480e-06  eta: 9:33:38  time: 1.2760  data_time: 0.0107  memory: 11055  loss: 0.6811
+2025/03/04 11:33:10 - mmengine - INFO - Iter(train) [   90/19224]  lr: 3.0958e-06  eta: 9:05:42  time: 1.0182  data_time: 0.0101  memory: 10558  loss: 0.5202
+2025/03/04 11:33:18 - mmengine - INFO - Iter(train) [  100/19224]  lr: 3.4436e-06  eta: 8:34:44  time: 0.7483  data_time: 0.0091  memory: 9661  loss: 0.5684
+2025/03/04 11:33:36 - mmengine - INFO - Iter(train) [  110/19224]  lr: 3.7915e-06  eta: 8:39:45  time: 1.7979  data_time: 0.0105  memory: 15104  loss: 0.3638
+2025/03/04 11:33:52 - mmengine - INFO - Iter(train) [  120/19224]  lr: 4.1393e-06  eta: 8:39:11  time: 1.6204  data_time: 0.0104  memory: 12016  loss: 0.3501
+2025/03/04 11:34:07 - mmengine - INFO - Iter(train) [  130/19224]  lr: 4.4871e-06  eta: 8:36:04  time: 1.5144  data_time: 0.0105  memory: 11578  loss: 0.3903
+2025/03/04 11:34:22 - mmengine - INFO - Iter(train) [  140/19224]  lr: 4.8349e-06  eta: 8:32:49  time: 1.4901  data_time: 0.0109  memory: 11441  loss: 0.3476
+2025/03/04 11:34:36 - mmengine - INFO - Iter(train) [  150/19224]  lr: 5.1828e-06  eta: 8:29:01  time: 1.4462  data_time: 0.0114  memory: 11303  loss: 0.3756
+2025/03/04 11:34:50 - mmengine - INFO - Iter(train) [  160/19224]  lr: 5.5306e-06  eta: 8:23:51  time: 1.3546  data_time: 0.0105  memory: 11123  loss: 0.3872
+2025/03/04 11:35:02 - mmengine - INFO - Iter(train) [  170/19224]  lr: 5.8784e-06  eta: 8:17:20  time: 1.2510  data_time: 0.0105  memory: 10922  loss: 0.3984
+2025/03/04 11:35:14 - mmengine - INFO - Iter(train) [  180/19224]  lr: 6.2262e-06  eta: 8:09:41  time: 1.1468  data_time: 0.0114  memory: 10672  loss: 0.4183
+2025/03/04 11:35:24 - mmengine - INFO - Iter(train) [  190/19224]  lr: 6.5740e-06  eta: 8:00:03  time: 0.9815  data_time: 0.0096  memory: 10248  loss: 0.3544
+2025/03/04 11:35:31 - mmengine - INFO - Iter(train) [  200/19224]  lr: 6.9219e-06  eta: 7:46:45  time: 0.6902  data_time: 0.0091  memory: 9804  loss: 0.3184
+2025/03/04 11:35:51 - mmengine - INFO - Iter(train) [  210/19224]  lr: 7.2697e-06  eta: 7:55:34  time: 2.0721  data_time: 0.0104  memory: 17788  loss: 0.2691
+2025/03/04 11:36:07 - mmengine - INFO - Iter(train) [  220/19224]  lr: 7.6175e-06  eta: 7:56:38  time: 1.5929  data_time: 0.0118  memory: 11865  loss: 0.3032
+2025/03/04 11:36:22 - mmengine - INFO - Iter(train) [  230/19224]  lr: 7.9653e-06  eta: 7:56:16  time: 1.4955  data_time: 0.0109  memory: 11605  loss: 0.3026
+2025/03/04 11:36:37 - mmengine - INFO - Iter(train) [  240/19224]  lr: 8.3132e-06  eta: 7:55:03  time: 1.4318  data_time: 0.0110  memory: 11348  loss: 0.2916
+2025/03/04 11:36:51 - mmengine - INFO - Iter(train) [  250/19224]  lr: 8.6610e-06  eta: 7:53:24  time: 1.3910  data_time: 0.0106  memory: 11212  loss: 0.3301
+2025/03/04 11:37:04 - mmengine - INFO - Iter(train) [  260/19224]  lr: 9.0088e-06  eta: 7:51:32  time: 1.3642  data_time: 0.0106  memory: 11092  loss: 0.3150
+2025/03/04 11:37:16 - mmengine - INFO - Iter(train) [  270/19224]  lr: 9.3566e-06  eta: 7:47:31  time: 1.1699  data_time: 0.0109  memory: 10895  loss: 0.4017
+2025/03/04 11:37:26 - mmengine - INFO - Iter(train) [  280/19224]  lr: 9.7045e-06  eta: 7:41:37  time: 0.9782  data_time: 0.0094  memory: 10390  loss: 0.2858
+2025/03/04 11:37:34 - mmengine - INFO - Iter(train) [  290/19224]  lr: 1.0052e-05  eta: 7:35:05  time: 0.8845  data_time: 0.0096  memory: 9940  loss: 0.3397
+2025/03/04 11:37:41 - mmengine - INFO - Iter(train) [  300/19224]  lr: 1.0400e-05  eta: 7:26:50  time: 0.6804  data_time: 0.0089  memory: 9679  loss: 0.3477
+2025/03/04 11:37:58 - mmengine - INFO - Iter(train) [  310/19224]  lr: 1.0748e-05  eta: 7:29:22  time: 1.6891  data_time: 0.0105  memory: 13096  loss: 0.2749
+2025/03/04 11:38:14 - mmengine - INFO - Iter(train) [  320/19224]  lr: 1.1096e-05  eta: 7:30:40  time: 1.5811  data_time: 0.0108  memory: 11972  loss: 0.2802
+2025/03/04 11:38:29 - mmengine - INFO - Iter(train) [  330/19224]  lr: 1.1444e-05  eta: 7:31:27  time: 1.5370  data_time: 0.0105  memory: 11557  loss: 0.2905
+2025/03/04 11:38:44 - mmengine - INFO - Iter(train) [  340/19224]  lr: 1.1791e-05  eta: 7:31:20  time: 1.4478  data_time: 0.0110  memory: 11408  loss: 0.2810
+2025/03/04 11:38:58 - mmengine - INFO - Iter(train) [  350/19224]  lr: 1.2139e-05  eta: 7:30:44  time: 1.3936  data_time: 0.0107  memory: 11262  loss: 0.3028
+2025/03/04 11:39:11 - mmengine - INFO - Iter(train) [  360/19224]  lr: 1.2487e-05  eta: 7:29:38  time: 1.3336  data_time: 0.0114  memory: 11153  loss: 0.2933
+2025/03/04 11:39:24 - mmengine - INFO - Iter(train) [  370/19224]  lr: 1.2835e-05  eta: 7:28:05  time: 1.2765  data_time: 0.0108  memory: 11027  loss: 0.3327
+2025/03/04 11:39:35 - mmengine - INFO - Iter(train) [  380/19224]  lr: 1.3183e-05  eta: 7:25:06  time: 1.0944  data_time: 0.0100  memory: 10580  loss: 0.3867
+2025/03/04 11:39:45 - mmengine - INFO - Iter(train) [  390/19224]  lr: 1.3530e-05  eta: 7:21:20  time: 0.9780  data_time: 0.0097  memory: 10354  loss: 0.3154
+2025/03/04 11:39:52 - mmengine - INFO - Iter(train) [  400/19224]  lr: 1.3878e-05  eta: 7:16:13  time: 0.7822  data_time: 0.0089  memory: 9683  loss: 0.3445
+2025/03/04 11:40:09 - mmengine - INFO - Iter(train) [  410/19224]  lr: 1.4226e-05  eta: 7:17:55  time: 1.6434  data_time: 0.0107  memory: 12505  loss: 0.2493
+2025/03/04 11:40:25 - mmengine - INFO - Iter(train) [  420/19224]  lr: 1.4574e-05  eta: 7:19:01  time: 1.5764  data_time: 0.0106  memory: 11912  loss: 0.2502
+2025/03/04 11:40:40 - mmengine - INFO - Iter(train) [  430/19224]  lr: 1.4922e-05  eta: 7:19:47  time: 1.5375  data_time: 0.0108  memory: 11529  loss: 0.2680
+2025/03/04 11:40:54 - mmengine - INFO - Iter(train) [  440/19224]  lr: 1.5270e-05  eta: 7:19:43  time: 1.4275  data_time: 0.0106  memory: 11389  loss: 0.2787
+2025/03/04 11:41:08 - mmengine - INFO - Iter(train) [  450/19224]  lr: 1.5617e-05  eta: 7:19:33  time: 1.4134  data_time: 0.0110  memory: 11266  loss: 0.3042
+2025/03/04 11:41:21 - mmengine - INFO - Iter(train) [  460/19224]  lr: 1.5965e-05  eta: 7:18:12  time: 1.2417  data_time: 0.0103  memory: 11039  loss: 0.3354
+2025/03/04 11:41:34 - mmengine - INFO - Iter(train) [  470/19224]  lr: 1.6313e-05  eta: 7:17:27  time: 1.3231  data_time: 0.0105  memory: 10907  loss: 0.3018
+2025/03/04 11:41:46 - mmengine - INFO - Iter(train) [  480/19224]  lr: 1.6661e-05  eta: 7:15:48  time: 1.1828  data_time: 0.0098  memory: 10818  loss: 0.3590
+2025/03/04 11:41:56 - mmengine - INFO - Iter(train) [  490/19224]  lr: 1.7009e-05  eta: 7:13:20  time: 1.0446  data_time: 0.0096  memory: 10395  loss: 0.3291
+2025/03/04 11:42:04 - mmengine - INFO - Iter(train) [  500/19224]  lr: 1.7357e-05  eta: 7:09:04  time: 0.7395  data_time: 0.0089  memory: 9992  loss: 0.2996
+2025/03/04 11:42:21 - mmengine - INFO - Iter(train) [  510/19224]  lr: 1.7704e-05  eta: 7:11:10  time: 1.7567  data_time: 0.0102  memory: 14188  loss: 0.2741
+2025/03/04 11:42:37 - mmengine - INFO - Iter(train) [  520/19224]  lr: 1.8052e-05  eta: 7:12:15  time: 1.6026  data_time: 0.0107  memory: 12266  loss: 0.2345
+2025/03/04 11:42:52 - mmengine - INFO - Iter(train) [  530/19224]  lr: 1.8400e-05  eta: 7:12:25  time: 1.4521  data_time: 0.0105  memory: 11810  loss: 0.2394
+2025/03/04 11:43:08 - mmengine - INFO - Iter(train) [  540/19224]  lr: 1.8748e-05  eta: 7:13:25  time: 1.6034  data_time: 0.0117  memory: 11765  loss: 0.2383
+2025/03/04 11:43:22 - mmengine - INFO - Iter(train) [  550/19224]  lr: 1.9096e-05  eta: 7:13:17  time: 1.4077  data_time: 0.0110  memory: 11367  loss: 0.2818
+2025/03/04 11:43:36 - mmengine - INFO - Iter(train) [  560/19224]  lr: 1.9443e-05  eta: 7:13:12  time: 1.4201  data_time: 0.0108  memory: 11259  loss: 0.2765
+2025/03/04 11:43:50 - mmengine - INFO - Iter(train) [  570/19224]  lr: 1.9791e-05  eta: 7:12:48  time: 1.3616  data_time: 0.0130  memory: 11167  loss: 0.2869
+2025/03/04 11:44:03 - mmengine - INFO - Iter(train) [  580/19224]  lr: 2.0000e-05  eta: 7:12:11  time: 1.3208  data_time: 0.0104  memory: 10925  loss: 0.3095
+2025/03/04 11:44:14 - mmengine - INFO - Iter(train) [  590/19224]  lr: 2.0000e-05  eta: 7:10:19  time: 1.0787  data_time: 0.0098  memory: 10596  loss: 0.2788
+2025/03/04 11:44:23 - mmengine - INFO - Iter(train) [  600/19224]  lr: 2.0000e-05  eta: 7:07:43  time: 0.9279  data_time: 0.0097  memory: 10005  loss: 0.2877
+2025/03/04 11:44:41 - mmengine - INFO - Iter(train) [  610/19224]  lr: 2.0000e-05  eta: 7:09:22  time: 1.7489  data_time: 0.0104  memory: 13855  loss: 0.2535
+2025/03/04 11:44:56 - mmengine - INFO - Iter(train) [  620/19224]  lr: 2.0000e-05  eta: 7:09:57  time: 1.5451  data_time: 0.0106  memory: 12281  loss: 0.2177
+2025/03/04 11:45:11 - mmengine - INFO - Iter(train) [  630/19224]  lr: 2.0000e-05  eta: 7:10:25  time: 1.5314  data_time: 0.0107  memory: 11697  loss: 0.2348
+2025/03/04 11:45:26 - mmengine - INFO - Iter(train) [  640/19224]  lr: 1.9999e-05  eta: 7:10:31  time: 1.4570  data_time: 0.0107  memory: 11459  loss: 0.2709
+2025/03/04 11:45:40 - mmengine - INFO - Iter(train) [  650/19224]  lr: 1.9999e-05  eta: 7:10:22  time: 1.4061  data_time: 0.0109  memory: 11390  loss: 0.2576
+2025/03/04 11:45:53 - mmengine - INFO - Iter(train) [  660/19224]  lr: 1.9999e-05  eta: 7:09:48  time: 1.3197  data_time: 0.0109  memory: 11218  loss: 0.3110
+2025/03/04 11:46:06 - mmengine - INFO - Iter(train) [  670/19224]  lr: 1.9999e-05  eta: 7:09:04  time: 1.2780  data_time: 0.0106  memory: 10981  loss: 0.3224
+2025/03/04 11:46:17 - mmengine - INFO - Iter(train) [  680/19224]  lr: 1.9998e-05  eta: 7:07:38  time: 1.1260  data_time: 0.0104  memory: 10725  loss: 0.3063
+2025/03/04 11:46:27 - mmengine - INFO - Iter(train) [  690/19224]  lr: 1.9998e-05  eta: 7:05:30  time: 0.9567  data_time: 0.0097  memory: 10172  loss: 0.2703
+2025/03/04 11:46:35 - mmengine - INFO - Iter(train) [  700/19224]  lr: 1.9998e-05  eta: 7:02:45  time: 0.8067  data_time: 0.0091  memory: 9872  loss: 0.3140

internvl_ft_run_12_filter/20250304_113017/vis_data/events.out.tfevents.1741059019.intern-studio-40019814.28433.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:fd425b2aca9e0b39d18b30a09b663885a921b56d30f9da0e450be82102f9a02a
+size 23159

internvl_ft_run_12_filter/20250304_114757/vis_data/events.out.tfevents.1741060079.intern-studio-40019814.34025.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:79ff3861cce2bb33eeb67d1b5be27e2cd23b93bed47b5cfbe8aaced8a16400cb
+size 257492

internvl_ft_run_12_filter/internvl_v2_internlm2_2b_qlora_finetune_copy.py

@@ -0,0 +1,145 @@
+accumulative_counts = 1
+batch_size = 2
+betas = (
+    0.9,
+    0.999,
+)
+custom_hooks = [
+    dict(
+        tokenizer=dict(
+            pretrained_model_name_or_path=
+            '/root/wangqun/models/internvl2-2B',
+            trust_remote_code=True,
+            type='transformers.AutoTokenizer.from_pretrained'),
+        type='xtuner.engine.hooks.DatasetInfoHook'),
+]
+data_path = '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json'
+dataloader_num_workers = 4
+default_hooks = dict(
+    checkpoint=dict(
+        by_epoch=False,
+        interval=1000,
+        max_keep_ckpts=-1,
+        save_optimizer=False,
+        type='mmengine.hooks.CheckpointHook'),
+    logger=dict(
+        interval=10,
+        log_metric_by_epoch=False,
+        type='mmengine.hooks.LoggerHook'),
+    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
+    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
+    timer=dict(type='mmengine.hooks.IterTimerHook'))
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+image_folder = '/'
+launcher = 'none'
+llava_dataset = dict(
+    data_paths='/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+    image_folders='/',
+    max_length=8192,
+    model_path='/root/wangqun/models/internvl2-2B',
+    template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+    type='xtuner.dataset.InternVL_V1_5_Dataset')
+load_from = None
+log_level = 'DEBUG'
+log_processor = dict(by_epoch=False)
+lr = 2e-05
+max_epochs = 4
+max_length = 8192
+max_norm = 1
+model = dict(
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm_lora=dict(
+        lora_alpha=256,
+        lora_dropout=0.05,
+        r=128,
+        target_modules=None,
+        task_type='CAUSAL_LM',
+        type='peft.LoraConfig'),
+    model_path='/root/wangqun/models/internvl2-2B',
+    quantization_llm=True,
+    quantization_vit=False,
+    type='xtuner.model.InternVL_V1_5')
+optim_type = 'torch.optim.AdamW'
+optim_wrapper = dict(
+    optimizer=dict(
+        betas=(
+            0.9,
+            0.999,
+        ),
+        lr=2e-05,
+        type='torch.optim.AdamW',
+        weight_decay=0.05),
+    type='DeepSpeedOptimWrapper')
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=0.12,
+        start_factor=1e-05,
+        type='mmengine.optim.LinearLR'),
+    dict(
+        begin=0.12,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=4,
+        eta_min=0.0,
+        type='mmengine.optim.CosineAnnealingLR'),
+]
+path = '/root/wangqun/models/internvl2-2B'
+prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.internlm2_chat'
+randomness = dict(deterministic=False, seed=None)
+resume = False
+runner_type = 'FlexibleRunner'
+save_steps = 1000
+save_total_limit = -1
+strategy = dict(
+    config=dict(
+        bf16=dict(enabled=True),
+        fp16=dict(enabled=False, initial_scale_power=16),
+        gradient_accumulation_steps='auto',
+        gradient_clipping='auto',
+        train_micro_batch_size_per_gpu='auto',
+        zero_allow_untested_optimizer=True,
+        zero_force_ds_cpu_optimizer=False,
+        zero_optimization=dict(overlap_comm=True, stage=2)),
+    exclude_frozen_parameters=True,
+    gradient_accumulation_steps=1,
+    gradient_clipping=1,
+    sequence_parallel_size=1,
+    train_micro_batch_size_per_gpu=2,
+    type='xtuner.engine.DeepSpeedStrategy')
+tokenizer = dict(
+    pretrained_model_name_or_path=
+    '/root/wangqun/models/internvl2-2B',
+    trust_remote_code=True,
+    type='transformers.AutoTokenizer.from_pretrained')
+train_cfg = dict(max_epochs=4, type='xtuner.engine.runner.TrainLoop')
+train_dataloader = dict(
+    batch_size=2,
+    collate_fn=dict(type='xtuner.dataset.collate_fns.default_collate_fn'),
+    dataset=dict(
+        data_paths=
+        '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+        image_folders='/',
+        max_length=8192,
+        model_path='/root/wangqun/models/internvl2-2B',
+        template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+        type='xtuner.dataset.InternVL_V1_5_Dataset'),
+    num_workers=4,
+    sampler=dict(
+        length_property='modality_length',
+        per_device_batch_size=2,
+        type='xtuner.dataset.samplers.LengthGroupedSampler'))
+visualizer = dict(
+    type='mmengine.visualization.Visualizer',
+    vis_backends=[
+        dict(type='mmengine.visualization.TensorboardVisBackend'),
+    ])
+warmup_ratio = 0.03
+weight_decay = 0.05
+work_dir = '/root/wangqun/work_dirs/internvl_ft_run_12_filter'

internvl_ft_run_12_filter/iter_1000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ff54e611914bd3584430380a53da9940c23015b16917b0eb7de9fb652280f0d8
+size 301244162

internvl_ft_run_12_filter/iter_2000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d42f971e9c3b56fa86cc64566873987c3b8654ac9495f287f4f8bf2e1c191472
+size 301318338

internvl_ft_run_12_filter/iter_3000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb25e46fa242cc70a082db9a47c87593ff9e15290cad1e6d2bde7e4d6e58f18e
+size 301392386

internvl_ft_run_12_filter/iter_4000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a2666d6ebb5c3f4476d5118e25a858ce6c628983fffbdddc5bfda13b262d60a6
+size 301466626

internvl_ft_run_12_filter/iter_5000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5320f8061e8cd2d95a58555e0717bc49ba5aa06195808a762994fc2f7cf40986
+size 301540866

internvl_ft_run_12_filter/iter_6000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:18a21687fe64d7dd3943dbea5cec4c0f46cc8c09d7e0e0a24cec1710e6f2e91a
+size 301615042

internvl_ft_run_12_filter/iter_7000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0b9c67595b834aaabc5cb02b9faa6cecdb478f622152668fe6493675b94378f5
+size 301689218

internvl_ft_run_12_filter/iter_8000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d957225a068b5fbd164698068c5df34c796ee6f58ee76438a6e00804d864447f
+size 301763330

internvl_ft_run_12_filter/iter_9000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a7a49c32e1fb80179256e644dcaf8ab29f19cd1c90edf14d10ff136016f8412a
+size 301837506

internvl_ft_run_12_filter/iter_9612.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:167ac77f8c658792a016e1a6fe123f62036edb2cc67cf0fb40248b3feeb0c651
+size 301882434

internvl_ft_run_12_filter/last_checkpoint

@@ -0,0 +1 @@
+/root/wangqun/work_dirs/internvl_ft_run_12_filter/iter_9612.pth

internvl_ft_run_13_filter/20250304_121519/20250304_121519.log

@@ -0,0 +1,464 @@
+2025/03/04 12:15:19 - mmengine - DEBUG - An `DeepSpeedStrategy` instance is built from registry, and its implementation can be found in xtuner.engine._strategy.deepspeed
+2025/03/04 12:15:19 - mmengine - INFO - 
+------------------------------------------------------------
+System environment:
+    sys.platform: linux
+    Python: 3.10.13 (main, Sep 11 2023, 13:44:35) [GCC 11.2.0]
+    CUDA available: True
+    MUSA available: False
+    numpy_random_seed: 49487546
+    GPU 0: NVIDIA A100-SXM4-80GB
+    CUDA_HOME: /usr/local/cuda
+    NVCC: Cuda compilation tools, release 12.2, V12.2.140
+    GCC: gcc (Ubuntu 9.4.0-1ubuntu1~20.04.2) 9.4.0
+    PyTorch: 2.4.1+cu121
+    PyTorch compiling details: PyTorch built with:
+  - GCC 9.3
+  - C++ Version: 201703
+  - Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
+  - Intel(R) MKL-DNN v3.4.2 (Git Hash 1137e04ec0b5251ca2b4400a4fd3c667ce843d67)
+  - OpenMP 201511 (a.k.a. OpenMP 4.5)
+  - LAPACK is enabled (usually provided by MKL)
+  - NNPACK is enabled
+  - CPU capability usage: AVX512
+  - CUDA Runtime 12.1
+  - NVCC architecture flags: -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;-gencode;arch=compute_90,code=sm_90
+  - CuDNN 90.1  (built against CUDA 12.4)
+  - Magma 2.6.1
+  - Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=12.1, CUDNN_VERSION=9.1.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-stringop-overflow -Wsuggest-override -Wno-psabi -Wno-error=pedantic -Wno-error=old-style-cast -Wno-missing-braces -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_VERSION=2.4.1, USE_CUDA=ON, USE_CUDNN=ON, USE_CUSPARSELT=1, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_GLOO=ON, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF, USE_ROCM_KERNEL_ASSERT=OFF, 
+
+    TorchVision: 0.19.1+cu121
+    OpenCV: 4.9.0
+    MMEngine: 0.10.6
+
+Runtime environment:
+    launcher: none
+    randomness: {'seed': None, 'deterministic': False}
+    cudnn_benchmark: False
+    mp_cfg: {'mp_start_method': 'fork', 'opencv_num_threads': 0}
+    dist_cfg: {'backend': 'nccl'}
+    seed: None
+    deterministic: False
+    Distributed launcher: none
+    Distributed training: False
+    GPU number: 1
+------------------------------------------------------------
+
+2025/03/04 12:15:20 - mmengine - INFO - Config:
+accumulative_counts = 2
+batch_size = 1
+betas = (
+    0.9,
+    0.999,
+)
+custom_hooks = [
+    dict(
+        tokenizer=dict(
+            pretrained_model_name_or_path=
+            '/root/share/new_models/OpenGVLab/InternVL2_5/InternVL2_5-2B',
+            trust_remote_code=True,
+            type='transformers.AutoTokenizer.from_pretrained'),
+        type='xtuner.engine.hooks.DatasetInfoHook'),
+]
+data_path = '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json'
+dataloader_num_workers = 4
+default_hooks = dict(
+    checkpoint=dict(
+        by_epoch=False,
+        interval=1000,
+        max_keep_ckpts=-1,
+        save_optimizer=False,
+        type='mmengine.hooks.CheckpointHook'),
+    logger=dict(
+        interval=10,
+        log_metric_by_epoch=False,
+        type='mmengine.hooks.LoggerHook'),
+    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
+    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
+    timer=dict(type='mmengine.hooks.IterTimerHook'))
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+image_folder = '/'
+launcher = 'none'
+llava_dataset = dict(
+    data_paths='/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+    image_folders='/',
+    max_length=8192,
+    model_path='/root/share/new_models/OpenGVLab/InternVL2_5/InternVL2_5-2B',
+    template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+    type='xtuner.dataset.InternVL_V1_5_Dataset')
+load_from = None
+log_level = 'DEBUG'
+log_processor = dict(by_epoch=False)
+lr = 2e-05
+max_epochs = 4
+max_length = 8192
+max_norm = 1
+model = dict(
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm_lora=dict(
+        lora_alpha=256,
+        lora_dropout=0.05,
+        r=128,
+        target_modules=None,
+        task_type='CAUSAL_LM',
+        type='peft.LoraConfig'),
+    model_path='/root/share/new_models/OpenGVLab/InternVL2_5/InternVL2_5-2B',
+    quantization_llm=True,
+    quantization_vit=False,
+    type='xtuner.model.InternVL_V1_5')
+optim_type = 'torch.optim.AdamW'
+optim_wrapper = dict(
+    optimizer=dict(
+        betas=(
+            0.9,
+            0.999,
+        ),
+        lr=2e-05,
+        type='torch.optim.AdamW',
+        weight_decay=0.05),
+    type='DeepSpeedOptimWrapper')
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=0.12,
+        start_factor=1e-05,
+        type='mmengine.optim.LinearLR'),
+    dict(
+        begin=0.12,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=4,
+        eta_min=0.0,
+        type='mmengine.optim.CosineAnnealingLR'),
+]
+path = '/root/share/new_models/OpenGVLab/InternVL2_5/InternVL2_5-2B'
+prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.internlm2_chat'
+randomness = dict(deterministic=False, seed=None)
+resume = False
+runner_type = 'FlexibleRunner'
+save_steps = 1000
+save_total_limit = -1
+strategy = dict(
+    config=dict(
+        bf16=dict(enabled=True),
+        fp16=dict(enabled=False, initial_scale_power=16),
+        gradient_accumulation_steps='auto',
+        gradient_clipping='auto',
+        train_micro_batch_size_per_gpu='auto',
+        zero_allow_untested_optimizer=True,
+        zero_force_ds_cpu_optimizer=False,
+        zero_optimization=dict(overlap_comm=True, stage=2)),
+    exclude_frozen_parameters=True,
+    gradient_accumulation_steps=2,
+    gradient_clipping=1,
+    sequence_parallel_size=1,
+    train_micro_batch_size_per_gpu=1,
+    type='xtuner.engine.DeepSpeedStrategy')
+tokenizer = dict(
+    pretrained_model_name_or_path=
+    '/root/share/new_models/OpenGVLab/InternVL2_5/InternVL2_5-2B',
+    trust_remote_code=True,
+    type='transformers.AutoTokenizer.from_pretrained')
+train_cfg = dict(max_epochs=4, type='xtuner.engine.runner.TrainLoop')
+train_dataloader = dict(
+    batch_size=1,
+    collate_fn=dict(type='xtuner.dataset.collate_fns.default_collate_fn'),
+    dataset=dict(
+        data_paths=
+        '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+        image_folders='/',
+        max_length=8192,
+        model_path=
+        '/root/share/new_models/OpenGVLab/InternVL2_5/InternVL2_5-2B',
+        template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+        type='xtuner.dataset.InternVL_V1_5_Dataset'),
+    num_workers=4,
+    sampler=dict(
+        length_property='modality_length',
+        per_device_batch_size=2,
+        type='xtuner.dataset.samplers.LengthGroupedSampler'))
+visualizer = dict(
+    type='mmengine.visualization.Visualizer',
+    vis_backends=[
+        dict(type='mmengine.visualization.TensorboardVisBackend'),
+    ])
+warmup_ratio = 0.03
+weight_decay = 0.05
+work_dir = '/root/wangqun/work_dirs/internvl_ft_run_13_filter'
+
+2025/03/04 12:15:20 - mmengine - DEBUG - An `TensorboardVisBackend` instance is built from registry, and its implementation can be found in mmengine.visualization.vis_backend
+2025/03/04 12:15:20 - mmengine - DEBUG - An `Visualizer` instance is built from registry, and its implementation can be found in mmengine.visualization.visualizer
+2025/03/04 12:15:20 - mmengine - DEBUG - Attribute `_env_initialized` is not defined in <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'> or `<class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized is False, `_init_env` will be called and <class 'mmengine.visualization.vis_backend.TensorboardVisBackend'>._env_initialized will be set to True
+2025/03/04 12:15:20 - mmengine - DEBUG - Get class `RuntimeInfoHook` from "hook" registry in "mmengine"
+2025/03/04 12:15:20 - mmengine - DEBUG - An `RuntimeInfoHook` instance is built from registry, and its implementation can be found in mmengine.hooks.runtime_info_hook
+2025/03/04 12:15:20 - mmengine - DEBUG - An `IterTimerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.iter_timer_hook
+2025/03/04 12:15:20 - mmengine - DEBUG - An `DistSamplerSeedHook` instance is built from registry, and its implementation can be found in mmengine.hooks.sampler_seed_hook
+2025/03/04 12:15:20 - mmengine - DEBUG - An `LoggerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.logger_hook
+2025/03/04 12:15:20 - mmengine - DEBUG - An `ParamSchedulerHook` instance is built from registry, and its implementation can be found in mmengine.hooks.param_scheduler_hook
+2025/03/04 12:15:20 - mmengine - DEBUG - An `CheckpointHook` instance is built from registry, and its implementation can be found in mmengine.hooks.checkpoint_hook
+2025/03/04 12:15:20 - mmengine - WARNING - Failed to search registry with scope "mmengine" in the "builder" registry tree. As a workaround, the current "builder" registry in "xtuner" is used to build instance. This may cause unexpected failure when running the built modules. Please check whether "mmengine" is a correct scope, or whether the registry is initialized.
+2025/03/04 12:15:21 - mmengine - DEBUG - An `from_pretrained` instance is built from registry, and its implementation can be found in transformers.models.auto.tokenization_auto
+2025/03/04 12:15:21 - mmengine - DEBUG - An `DatasetInfoHook` instance is built from registry, and its implementation can be found in xtuner.engine.hooks.dataset_info_hook
+2025/03/04 12:15:21 - mmengine - INFO - Hooks will be executed in the following order:
+before_run:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+before_train:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(NORMAL      ) DatasetInfoHook                    
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+before_train_epoch:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(NORMAL      ) DistSamplerSeedHook                
+ -------------------- 
+before_train_iter:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+after_train_iter:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+(LOW         ) ParamSchedulerHook                 
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+after_train_epoch:
+(NORMAL      ) IterTimerHook                      
+(LOW         ) ParamSchedulerHook                 
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+before_val:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) DatasetInfoHook                    
+ -------------------- 
+before_val_epoch:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+before_val_iter:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+after_val_iter:
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+after_val_epoch:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+(LOW         ) ParamSchedulerHook                 
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+after_val:
+(VERY_HIGH   ) RuntimeInfoHook                    
+ -------------------- 
+after_train:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(VERY_LOW    ) CheckpointHook                     
+ -------------------- 
+before_test:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) DatasetInfoHook                    
+ -------------------- 
+before_test_epoch:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+before_test_iter:
+(NORMAL      ) IterTimerHook                      
+ -------------------- 
+after_test_iter:
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+after_test_epoch:
+(VERY_HIGH   ) RuntimeInfoHook                    
+(NORMAL      ) IterTimerHook                      
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+after_test:
+(VERY_HIGH   ) RuntimeInfoHook                    
+ -------------------- 
+after_run:
+(BELOW_NORMAL) LoggerHook                         
+ -------------------- 
+2025/03/04 12:15:21 - mmengine - DEBUG - An `FlexibleRunner` instance is built from registry, its implementation can be found inmmengine.runner._flexible_runner
+2025/03/04 12:15:21 - mmengine - INFO - Starting to loading data and calc length
+2025/03/04 12:15:21 - mmengine - INFO - =======Starting to process /root/data/tempData/screenshot_od/layout_ocr_multi_scale.json =======
+2025/03/04 12:15:28 - mmengine - INFO - =======total 4806 samples of /root/data/tempData/screenshot_od/layout_ocr_multi_scale.json=======
+2025/03/04 12:15:28 - mmengine - INFO - end loading data and calc length
+2025/03/04 12:15:28 - mmengine - INFO - =======total 4806 samples=======
+2025/03/04 12:15:28 - mmengine - DEBUG - An `InternVL_V1_5_Dataset` instance is built from registry, and its implementation can be found in xtuner.dataset.internvl_dataset
+2025/03/04 12:15:28 - mmengine - INFO - LengthGroupedSampler is used.
+2025/03/04 12:15:28 - mmengine - INFO - LengthGroupedSampler construction is complete, and the selected attribute is modality_length
+2025/03/04 12:15:28 - mmengine - DEBUG - An `LengthGroupedSampler` instance is built from registry, and its implementation can be found in xtuner.dataset.samplers.length_grouped
+2025/03/04 12:15:28 - mmengine - WARNING - Dataset InternVL_V1_5_Dataset has no metainfo. ``dataset_meta`` in visualizer will be None.
+2025/03/04 12:15:28 - mmengine - DEBUG - An `TrainLoop` instance is built from registry, and its implementation can be found in xtuner.engine.runner.loops
+2025/03/04 12:15:28 - mmengine - INFO - Start to load InternVL_V1_5 model.
+2025/03/04 12:15:28 - mmengine - DEBUG - Get class `BaseDataPreprocessor` from "model" registry in "mmengine"
+2025/03/04 12:15:28 - mmengine - DEBUG - An `BaseDataPreprocessor` instance is built from registry, and its implementation can be found in mmengine.model.base_model.data_preprocessor
+2025/03/04 12:15:55 - mmengine - DEBUG - An `LoraConfig` instance is built from registry, and its implementation can be found in peft.tuners.lora.config
+2025/03/04 12:15:58 - mmengine - INFO - InternVL_V1_5(
+  (data_preprocessor): BaseDataPreprocessor()
+  (model): InternVLChatModel(
+    (vision_model): InternVisionModel(
+      (embeddings): InternVisionEmbeddings(
+        (patch_embedding): Conv2d(3, 1024, kernel_size=(14, 14), stride=(14, 14))
+      )
+      (encoder): InternVisionEncoder(
+        (layers): ModuleList(
+          (0-23): 24 x InternVisionEncoderLayer(
+            (attn): InternAttention(
+              (qkv): Linear(in_features=1024, out_features=3072, bias=True)
+              (attn_drop): Dropout(p=0.0, inplace=False)
+              (proj_drop): Dropout(p=0.0, inplace=False)
+              (proj): Linear(in_features=1024, out_features=1024, bias=True)
+            )
+            (mlp): InternMLP(
+              (act): GELUActivation()
+              (fc1): Linear(in_features=1024, out_features=4096, bias=True)
+              (fc2): Linear(in_features=4096, out_features=1024, bias=True)
+            )
+            (norm1): LayerNorm((1024,), eps=1e-06, elementwise_affine=True)
+            (norm2): LayerNorm((1024,), eps=1e-06, elementwise_affine=True)
+            (drop_path1): Identity()
+            (drop_path2): Identity()
+          )
+        )
+      )
+    )
+    (language_model): PeftModelForCausalLM(
+      (base_model): LoraModel(
+        (model): InternLM2ForCausalLM(
+          (model): InternLM2Model(
+            (tok_embeddings): Embedding(92553, 2048, padding_idx=2)
+            (layers): ModuleList(
+              (0-23): 24 x InternLM2DecoderLayer(
+                (attention): InternLM2Attention(
+                  (wqkv): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=4096, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=4096, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (wo): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=2048, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=2048, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (rotary_emb): InternLM2DynamicNTKScalingRotaryEmbedding()
+                )
+                (feed_forward): InternLM2MLP(
+                  (w1): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=8192, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=8192, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (w3): lora.Linear(
+                    (base_layer): Linear4bit(in_features=2048, out_features=8192, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=2048, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=8192, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (w2): lora.Linear(
+                    (base_layer): Linear4bit(in_features=8192, out_features=2048, bias=False)
+                    (lora_dropout): ModuleDict(
+                      (default): Dropout(p=0.05, inplace=False)
+                    )
+                    (lora_A): ModuleDict(
+                      (default): Linear(in_features=8192, out_features=128, bias=False)
+                    )
+                    (lora_B): ModuleDict(
+                      (default): Linear(in_features=128, out_features=2048, bias=False)
+                    )
+                    (lora_embedding_A): ParameterDict()
+                    (lora_embedding_B): ParameterDict()
+                    (lora_magnitude_vector): ModuleDict()
+                  )
+                  (act_fn): SiLU()
+                )
+                (attention_norm): InternLM2RMSNorm()
+                (ffn_norm): InternLM2RMSNorm()
+              )
+            )
+            (norm): InternLM2RMSNorm()
+          )
+          (output): lora.Linear(
+            (base_layer): Linear4bit(in_features=2048, out_features=92553, bias=False)
+            (lora_dropout): ModuleDict(
+              (default): Dropout(p=0.05, inplace=False)
+            )
+            (lora_A): ModuleDict(
+              (default): Linear(in_features=2048, out_features=128, bias=False)
+            )
+            (lora_B): ModuleDict(
+              (default): Linear(in_features=128, out_features=92553, bias=False)
+            )
+            (lora_embedding_A): ParameterDict()
+            (lora_embedding_B): ParameterDict()
+            (lora_magnitude_vector): ModuleDict()
+          )
+        )
+      )
+    )
+    (mlp1): Sequential(
+      (0): LayerNorm((4096,), eps=1e-05, elementwise_affine=True)
+      (1): Linear(in_features=4096, out_features=2048, bias=True)
+      (2): GELU(approximate='none')
+      (3): Linear(in_features=2048, out_features=2048, bias=True)
+    )
+  )
+)
+2025/03/04 12:15:58 - mmengine - INFO - InternVL_V1_5 construction is complete
+2025/03/04 12:15:58 - mmengine - DEBUG - An `InternVL_V1_5` instance is built from registry, and its implementation can be found in xtuner.model.internvl
+2025/03/04 12:15:58 - mmengine - DEBUG - Get class `DefaultOptimWrapperConstructor` from "optimizer wrapper constructor" registry in "mmengine"
+2025/03/04 12:15:58 - mmengine - DEBUG - An `DefaultOptimWrapperConstructor` instance is built from registry, and its implementation can be found in mmengine.optim.optimizer.default_constructor
+2025/03/04 12:15:58 - mmengine - DEBUG - An `AdamW` instance is built from registry, and its implementation can be found in torch.optim.adamw
+2025/03/04 12:15:58 - mmengine - DEBUG - Get class `DeepSpeedOptimWrapper` from "optim_wrapper" registry in "mmengine"
+2025/03/04 12:15:58 - mmengine - DEBUG - An `DeepSpeedOptimWrapper` instance is built from registry, and its implementation can be found in mmengine._strategy.deepspeed

internvl_ft_run_13_filter/20250304_121519/vis_data/events.out.tfevents.1741061720.intern-studio-40019814.41268.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:46db7041eca09f33cd2c42f058a3f2665a3e1928517a93cb3132aa388a2accdb
+size 4914

internvl_ft_run_13_filter/20250304_213711/vis_data/events.out.tfevents.1741095432.intern-studio-40019814.159243.0

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1d38718bec644b61e1e53b07ceca6b2c0adb59aedff598ab38d1fa9244ac23d0
+size 512032

internvl_ft_run_13_filter/internvl_v2_internlm2_2b_qlora_finetune_copy.py

@@ -0,0 +1,146 @@
+accumulative_counts = 2
+batch_size = 1
+betas = (
+    0.9,
+    0.999,
+)
+custom_hooks = [
+    dict(
+        tokenizer=dict(
+            pretrained_model_name_or_path=
+            '/data/wangqun/models/InternVL2_5-2B',
+            trust_remote_code=True,
+            type='transformers.AutoTokenizer.from_pretrained'),
+        type='xtuner.engine.hooks.DatasetInfoHook'),
+]
+data_path = '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json'
+dataloader_num_workers = 4
+default_hooks = dict(
+    checkpoint=dict(
+        by_epoch=False,
+        interval=1000,
+        max_keep_ckpts=-1,
+        save_optimizer=False,
+        type='mmengine.hooks.CheckpointHook'),
+    logger=dict(
+        interval=10,
+        log_metric_by_epoch=False,
+        type='mmengine.hooks.LoggerHook'),
+    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
+    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
+    timer=dict(type='mmengine.hooks.IterTimerHook'))
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+image_folder = '/'
+launcher = 'none'
+llava_dataset = dict(
+    data_paths='/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+    image_folders='/',
+    max_length=8192,
+    model_path='/data/wangqun/models/InternVL2_5-2B',
+    template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+    type='xtuner.dataset.InternVL_V1_5_Dataset')
+load_from = None
+log_level = 'DEBUG'
+log_processor = dict(by_epoch=False)
+lr = 2e-05
+max_epochs = 4
+max_length = 8192
+max_norm = 1
+model = dict(
+    freeze_llm=True,
+    freeze_visual_encoder=True,
+    llm_lora=dict(
+        lora_alpha=256,
+        lora_dropout=0.05,
+        r=128,
+        target_modules=None,
+        task_type='CAUSAL_LM',
+        type='peft.LoraConfig'),
+    model_path='/data/wangqun/models/InternVL2_5-2B',
+    quantization_llm=True,
+    quantization_vit=False,
+    type='xtuner.model.InternVL_V1_5')
+optim_type = 'torch.optim.AdamW'
+optim_wrapper = dict(
+    optimizer=dict(
+        betas=(
+            0.9,
+            0.999,
+        ),
+        lr=2e-05,
+        type='torch.optim.AdamW',
+        weight_decay=0.05),
+    type='DeepSpeedOptimWrapper')
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=0.12,
+        start_factor=1e-05,
+        type='mmengine.optim.LinearLR'),
+    dict(
+        begin=0.12,
+        by_epoch=True,
+        convert_to_iter_based=True,
+        end=4,
+        eta_min=0.0,
+        type='mmengine.optim.CosineAnnealingLR'),
+]
+path = '/data/wangqun/models/InternVL2_5-2B'
+prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.internlm2_chat'
+randomness = dict(deterministic=False, seed=None)
+resume = False
+runner_type = 'FlexibleRunner'
+save_steps = 1000
+save_total_limit = -1
+strategy = dict(
+    config=dict(
+        bf16=dict(enabled=True),
+        fp16=dict(enabled=False, initial_scale_power=16),
+        gradient_accumulation_steps='auto',
+        gradient_clipping='auto',
+        train_micro_batch_size_per_gpu='auto',
+        zero_allow_untested_optimizer=True,
+        zero_force_ds_cpu_optimizer=False,
+        zero_optimization=dict(overlap_comm=True, stage=2)),
+    exclude_frozen_parameters=True,
+    gradient_accumulation_steps=2,
+    gradient_clipping=1,
+    sequence_parallel_size=1,
+    train_micro_batch_size_per_gpu=1,
+    type='xtuner.engine.DeepSpeedStrategy')
+tokenizer = dict(
+    pretrained_model_name_or_path=
+    '/data/wangqun/models/InternVL2_5-2B',
+    trust_remote_code=True,
+    type='transformers.AutoTokenizer.from_pretrained')
+train_cfg = dict(max_epochs=4, type='xtuner.engine.runner.TrainLoop')
+train_dataloader = dict(
+    batch_size=1,
+    collate_fn=dict(type='xtuner.dataset.collate_fns.default_collate_fn'),
+    dataset=dict(
+        data_paths=
+        '/root/data/tempData/screenshot_od/layout_ocr_multi_scale.json',
+        image_folders='/',
+        max_length=8192,
+        model_path=
+        '/data/wangqun/models/InternVL2_5-2B',
+        template='xtuner.utils.PROMPT_TEMPLATE.internlm2_chat',
+        type='xtuner.dataset.InternVL_V1_5_Dataset'),
+    num_workers=4,
+    sampler=dict(
+        length_property='modality_length',
+        per_device_batch_size=2,
+        type='xtuner.dataset.samplers.LengthGroupedSampler'))
+visualizer = dict(
+    type='mmengine.visualization.Visualizer',
+    vis_backends=[
+        dict(type='mmengine.visualization.TensorboardVisBackend'),
+    ])
+warmup_ratio = 0.03
+weight_decay = 0.05
+work_dir = '/root/wangqun/work_dirs/internvl_ft_run_13_filter'

internvl_ft_run_13_filter/iter_1000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b24f6c6bd3851bce9ce45f4bba96fdbcee632c7c622ebe837ec5b973085bbb73
+size 301244994

internvl_ft_run_13_filter/iter_10000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1c692ec736bd23df366f5c00daad06614571f7fca440f7964cff27fbfd8bb3e
+size 301920578

internvl_ft_run_13_filter/iter_11000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ea8fcd7154e0d6ded70bc9e6b67d7a1a8cf59c57c566eeb6bea962cf721f1591
+size 301995330

internvl_ft_run_13_filter/iter_12000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:43f6034da940b55224f9205875cc2740bfccc058a3f145d0b7bcaccd115536f5
+size 302070338

internvl_ft_run_13_filter/iter_13000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:92bfbc4a13bfd7b0eea33d12fc9615ec744297673ed571cd140f40871481f46d
+size 302145474

internvl_ft_run_13_filter/iter_14000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b0eb4fa01c3378ea8eb3acf2b084fa954a72d29fcf63b2bfe27553d61e79fb8b
+size 302220482

internvl_ft_run_13_filter/iter_15000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1bd81bba2007b30f19aa53382d801968c7f349487162031fe2dc8ebd823e4c27
+size 302295490

internvl_ft_run_13_filter/iter_16000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8603f7e271fcefb72f50bf1397f05f03b55505c3272e4680fdfa27c4de9dc157
+size 302370562

internvl_ft_run_13_filter/iter_17000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3d9041144ef0a48ad437e9ce94558444dbbc6c485d9a779048f6fe08a44e324a
+size 302445570

internvl_ft_run_13_filter/iter_18000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1032dda4ee201c9f851825745b001b30864d7c84cdb11ace42b1d9ffa70be77b
+size 302520514

internvl_ft_run_13_filter/iter_19000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f17fa3c628a60d737567747eb440d606b0ee0fa89bf4c8f5acbcb48e13e3089f
+size 302594882

internvl_ft_run_13_filter/iter_19224.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:75eca927c93ffe7129d4f1fc84bdeb42a6563e6bc97c1f66d98a17dd1f479ae6
+size 302611458

internvl_ft_run_13_filter/iter_2000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:039ecebe9872764ca4c96ff04029e8921a938cad5ad026009a090270788aad4b
+size 301320450

internvl_ft_run_13_filter/iter_3000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:779c990e71dfbac4e0c18606852eb8f564bb232947404a0342a4c06c2792e2e9
+size 301395330

internvl_ft_run_13_filter/iter_4000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:85bd648ac928cf3e76058af3404fb87ce098e5e05dc6e126707f1a5318f74119
+size 301470594

internvl_ft_run_13_filter/iter_5000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a9378786acb4c41e9f719d475acb3806e467e198d0ca929ae82880e8b826ff06
+size 301545538

internvl_ft_run_13_filter/iter_6000.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f5341528784c8762d29de4b7d16e32266b87ee3aade8daede3996d4a6fc6989e
+size 301620546