Issues

• warning with lr_step

  Berkeley-Data/hpt#80 · Filter by author taeil opened on Apr 13, 2021

• check what sen12ms_loader.py is doing

  Berkeley-Data/hpt#79 · Filter by author taeil opened on Apr 11, 2021
  

• evaluate fusion pretraining on s1 or s2 by zero padding.

  Berkeley-Data/hpt#77 · Filter by author taeil opened on Apr 10, 2021
  1

  

• pretraining with different augmentation set

  Berkeley-Data/hpt#76 · Filter by author tchken opened on Apr 9, 2021
  2

  

• Document, term, table (set place holder) from 0409 discussion -- full, partial, optional, and sensor(original).

  Berkeley-Data/hpt#74 · Filter by author tchken opened on Apr 9, 2021
  3

  

• pretrained model eval on BigEarthNet classification

  Berkeley-Data/hpt#72 · Filter by author taeil opened on Apr 8, 2021

• investigate why 1K supervised evaluation shows as high accuracy as full training set

  Berkeley-Data/hpt#71 · Filter by author taeil opened on Apr 5, 2021
  6

  

• bigearthnet compute s1 mean + std

  Berkeley-Data/hpt#70 · Filter by author tchken opened on Apr 5, 2021
  5

  

• pytorch warning

  Berkeley-Data/hpt#69 · Filter by author taeil opened on Apr 3, 2021

• pretraining with s1/s2 fusion data

  Berkeley-Data/hpt#68 · Filter by author taeil opened on Apr 3, 2021
  9

• combine extract and evaluation into single step

  Berkeley-Data/hpt#67 · Filter by author taeil opened on Apr 2, 2021

• save only the best model to wandb.dir

  Berkeley-Data/hpt#66 · Filter by author taeil opened on Apr 2, 2021

• Load input module 

  Berkeley-Data/hpt#65 · Filter by author taeil opened on Mar 28, 2021
  1

• baseline training 1x1 conv layer and weight

  Berkeley-Data/hpt#62 · Filter by author taeil opened on Mar 27, 2021
  3

  

• full train set and test set

  Berkeley-Data/hpt#61 · Filter by author taeil opened on Mar 27, 2021
  1

• collect samples based on prediction results.

  Berkeley-Data/hpt#60 · Filter by author taeil opened on Mar 23, 2021
  1

  

• Pixel Label Training

  Berkeley-Data/hpt#57 · Filter by author oropezaev opened on Mar 21, 2021 · next milestone
  3

• collecting images from multiple workers

  Berkeley-Data/hpt#56 · Filter by author taeil opened on Mar 21, 2021 · next milestone
  1

• add two other repos as submodules

  Berkeley-Data/hpt#55 · Filter by author taeil opened on Mar 21, 2021

• UndefinedMetricWarnings while running classification/main_train.py on SEN12MS

  Berkeley-Data/hpt#52 · Filter by author suryagutta opened on Mar 21, 2021 · next milestone
  3

• t-SNE quantitative evaluation on pre-train MoCo V2 for SEN12MS

  Berkeley-Data/hpt#51 · Filter by author oropezaev opened on Mar 20, 2021 · next milestone

• sen12ms evaluation deep dive and documentation

  Berkeley-Data/hpt#50 · Filter by author tchken opened on Mar 20, 2021
  3

  

• draft the section similar to 4.1.1 from geo-aware paper

  Berkeley-Data/hpt#49 · Filter by author suryagutta opened on Mar 20, 2021 · next milestone

• convert moco pretraining to sen12ms evaluation pretraining model

  Berkeley-Data/hpt#48 · Filter by author taeil opened on Mar 19, 2021 · pre-training sen12ms with min variation
  7

  

• fine-tune / sen12 simple evaluation on sen12ms

  Berkeley-Data/hpt#44 · Filter by author taeil opened on Mar 14, 2021
  9