Sequence length: How to limit that to 30, it is increasing automatically as the no. of epochs is increasing

[pranavraikote]

I'm custom training vid2vid for Pose-to-body generations, and given below is an extract of the logs used for my custom training. It says this despite specifying n_frame_total 30 in my parameters used for training ->> --------- Updating training sequence length to 120 ---------

Any way to limit this to 30, or this is the way it has to train for getting good results? Can anyone clarify this?

----------------Parameters used---------------- TTUR: False add_face_disc: False basic_point_only: False batchSize: 8 beta1: 0.5 checkpoints_dir: ./checkpoints continue_train: True dataroot: /mnt/FS/datasets dataset_mode: pose debug: False densepose_only: False display_freq: 100 display_id: 0 display_winsize: 512 feat_num: 3 fg: False fg_labels: [26] fineSize: 256 fp16: False gan_mode: ls gpu_ids: [0, 1, 2, 3, 4, 5, 6, 7] input_nc: 6 isTrain: True label_feat: False label_nc: 0 lambda_F: 10.0 lambda_T: 10.0 lambda_feat: 10.0 loadSize: 384 load_features: False load_pretrain: local_rank: 0 lr: 0.0002 max_dataset_size: inf max_frames_backpropagate: 1 max_frames_per_gpu: 5 max_t_step: 1 model: vid2vid nThreads: 2 n_blocks: 9 n_blocks_local: 3 n_downsample_E: 3 n_downsample_G: 3 n_frames_D: 3 n_frames_G: 3 n_frames_total: 30 n_gpus_gen: 8 n_layers_D: 3 n_local_enhancers: 1 n_scales_spatial: 1 n_scales_temporal: 2 name: /mnt/FS/datasets/vid2vid/test ndf: 64 nef: 32 netE: simple netG: composite ngf: 64 niter: 10 niter_decay: 10 niter_fix_global: 0 niter_step: 5 no_canny_edge: False no_dist_map: False no_first_img: False no_flip: False no_flow: False no_ganFeat: False no_html: False no_vgg: False norm: batch num_D: 2 openpose_only: False output_nc: 3 phase: train pool_size: 1 print_freq: 100 random_drop_prob: 0.05 random_scale_points: False remove_face_labels: False resize_or_crop: Scaleheight_and_scaledCrop save_epoch_freq: 1 save_latest_freq: 1000 serial_batches: False sparse_D: False tf_log: False use_instance: False use_single_G: False which_epoch: latest -------------- End ---------------- CustomDatasetDataLoader dataset [PoseDataset] was created

training videos = 5070

vid2vid ---------- Networks initialized -------------

---------- Networks initialized -------------

Resuming from epoch 14 at iteration 144 update learning rate: 0.000200 -> 0.000140 update learning rate: 0.000200 -> 0.000140 --------- Updating training sequence length to 120 --------- -------- Updating number of backpropagated frames to 1 ----------