Can not train

[chengdianJiang]

Hi may I ask how to do with below, I want to train this model without pretrained model.

Using tensorboardX debug is begin Fix size testing. training chunk_sizes: [8] The output will be saved to C:\Deep Blue\MCMOT-master\src\lib....\exp\mot\ Setting up data... Dataset root: /data/yfzhang/MOT/JDE Total 0 image files in visdrone dataset. dataset summary OrderedDict([('visdrone', defaultdict(<class 'int'>, {}))]) Total 23 multi-scales: [[640, 320], [672, 352], [704, 384], [736, 416], [736, 416], [768, 448], [768, 416], [800, 480], [832, 512], [832, 448], [864, 544], [864, 480], [896, 576], [928, 608], [928, 512], [960, 640], [992, 672], [992, 544], [1056, 576], [1088, 608], [1088, 608], [1152, 640], [1216, 672]] heads: {'hm': 10, 'wh': 2, 'id': 128, 'reg': 2} opt: Namespace(K=200, arch='hrnet_32', batch_size=8, cat_spec_wh=False, chunk_sizes=[8], conf_thres=0.4, data_cfg='../src/lib/cfg/data.json', data_dir='C:/Deep Blue/Autopilot_practice/MCMOT-master/src/VisDrone2019-MOT-train', dataset='jde', debug_dir='C:\Deep Blue\MCMOT-master\src\lib\..\..\exp\mot\debug', dense_wh=False, det_thres=0.3, down_ratio=4, exp_dir='C:\Deep Blue\MCMOT-master\src\lib\..\..\exp\mot', exp_id='', fix_res=True, gen_scale=True, gpus=[0], gpus_str='0', head_conv=256, heads={'hm': 10, 'wh': 2, 'id': 128, 'reg': 2}, hide_data_time=False, hm_weight=1, id_loss='ce', id_weight=1, input_h=416, input_img='/users/duanyou/c5/all_pretrain/test.txt', input_mode='video', input_res=416, input_video='../videos/uav_339.mp4', input_w=416, input_wh=(1088, 608), is_debug=False, keep_res=False, load_model='', lr=7e-05, lr_step=[10, 20], master_batch_size=8, mean=None, metric='loss', min_box_area=100, mse_loss=False, multi_scale=True, nID_dict=defaultdict(<class 'int'>, {}), nms_thres=0.4, norm_wh=False, not_cuda_benchmark=False, not_prefetch_test=False, not_reg_offset=False, num_classes=10, num_epochs=40, num_iters=-1, num_stacks=1, num_workers=0, off_weight=1, output_format='video', output_h=104, output_res=104, output_root='../results', output_w=104, pad=31, print_iter=0, reg_loss='l1', reg_offset=True, reid_cls_ids='0,1,2,3,4,5,6,7,8,9', reid_dim=128, resume=False, root_dir='C:\Deep Blue\MCMOT-master\src\lib\..\..', save_all=False, save_dir='C:\Deep Blue\MCMOT-master\src\lib\..\..\exp\mot\', seed=317, std=None, task='mot', test=False, test_mot15=False, test_mot16=False, test_mot17=False, test_mot20=False, track_buffer=30, trainval=False, val_intervals=10, val_mot15=False, val_mot16=False, val_mot17=False, val_mot20=False, vis_thresh=0.5, wh_weight=0.1) Creating model... 'cp' 不是内部或外部命令，也不是可运行的程序 或批处理文件。 Non pretrained model specified! Starting training... Drop LR to 7e-06 Drop LR to 7.000000000000001e-07

[chengdianJiang]

The train.py can work, but tensorboard shows that the loss is also zero, as shown in below.... Can you do me a favour? thanks

[YagnikThanki]

Hello, @chengdianJiang I think you need to modify your save_dir. I think it will solve your error. Your model is not saved on the place where it should be