Odd output in eval.json after training

[yaseryacoob]

I ran the training (locally), and got this in the train folder (after about 6 hours or so)., see the eval.json

I tried to run the training again and got

rastervision run local -e idb.experiment Ensuring input files exist [####################################] 100% Checking for existing output [####################################] 100% make: Nothing to be done for 'all'.

-rwxrwxrwx 1 root root 214 May 13 16:24 checkpoint -rwxrwxrwx 1 root root 2339 May 13 16:23 command-config.json -rwxrwxrwx 1 root root 6580271 May 13 16:24 events.out.tfevents.1589387049.473fba7942a9 -rwxrwxrwx 1 root root 6040812 May 13 16:24 graph.pbtxt -rwxrwxrwx 1 root root 9396911 May 13 16:24 model -rwxrwxrwx 1 root root 18273184 May 13 16:24 model.ckpt-0.data-00000-of-00001 -rwxrwxrwx 1 root root 19819 May 13 16:24 model.ckpt-0.index -rwxrwxrwx 1 root root 3047854 May 13 16:24 model.ckpt-0.meta -rwxrwxrwx 1 root root 18273184 May 13 16:24 model.ckpt-1.data-00000-of-00001 -rwxrwxrwx 1 root root 19819 May 13 16:24 model.ckpt-1.index -rwxrwxrwx 1 root root 3047854 May 13 16:24 model.ckpt-1.meta -rwxrwxrwx 1 root root 25766277 May 13 16:24 multi-city.tar.gz

and the eval.json is odd

{ "per_scene": { "paramaribo_test": [ { "class_id": 1, "recall": 0.003331903803648599, "count_error": 86282.08435984561, "f1": 0.006492213956561171, "precision": 0.14688042062753656, "gt_count": 151565, "class_name": "Building" }, { "class_id": 2, "recall": 0.8396761416426244, "count_error": 140247.57092749514, "f1": 0.8733388716556868, "precision": 0.9171078731191155, "gt_count": 1431490, "class_name": "Background" }, { "class_id": null, "recall": 0.7596027933331441, "count_error": 135080.80226081848, "f1": 0.7903451578086202, "precision": 0.8433646842615671, "gt_count": 1583055, "class_name": "average" } ] }, "overall": [ { "class_id": 1, "recall": 0.003331903803648599, "count_error": 86282.08435984561, "f1": 0.006492213956561171, "precision": 0.14688042062753656, "gt_count": 151565, "class_name": "Building" }, { "class_id": 2, "recall": 0.8396761416426244, "count_error": 140247.57092749514, "f1": 0.8733388716556868, "precision": 0.9171078731191155, "gt_count": 1431490, "class_name": "Background" }, { "class_id": null, "recall": 0.7596027933331441, "count_error": 135080.80226081848, "f1": 0.7903451578086202, "precision": 0.8433646842615671, "gt_count": 1583055, "class_name": "average" } ] }