Merge pull request #550 from amaarora/wandb

Wandb Support

timm/utils/summary.py

@@ -5,7 +5,10 @@ Hacked together by / Copyright 2020 Ross Wightman
 import csv
 import os
 from collections import OrderedDict
-
+try: 
+    import wandb
+except ImportError:
+    pass
 
 def get_outdir(path, *paths, inc=False):
     outdir = os.path.join(path, *paths)
@@ -23,10 +26,12 @@ def get_outdir(path, *paths, inc=False):
     return outdir
 
 
-def update_summary(epoch, train_metrics, eval_metrics, filename, write_header=False):
+def update_summary(epoch, train_metrics, eval_metrics, filename, write_header=False, log_wandb=False):
     rowd = OrderedDict(epoch=epoch)
     rowd.update([('train_' + k, v) for k, v in train_metrics.items()])
     rowd.update([('eval_' + k, v) for k, v in eval_metrics.items()])
+    if log_wandb:
+        wandb.log(rowd)
     with open(filename, mode='a') as cf:
         dw = csv.DictWriter(cf, fieldnames=rowd.keys())
         if write_header:  # first iteration (epoch == 1 can't be used)

train.py

@@ -52,6 +52,12 @@ try:
 except AttributeError:
     pass
 
+try:
+    import wandb
+    has_wandb = True
+except ImportError: 
+    has_wandb = False
+
 torch.backends.cudnn.benchmark = True
 _logger = logging.getLogger('train')
 
@@ -271,6 +277,8 @@ parser.add_argument('--use-multi-epochs-loader', action='store_true', default=Fa
                     help='use the multi-epochs-loader to save time at the beginning of every epoch')
 parser.add_argument('--torchscript', dest='torchscript', action='store_true',
                     help='convert model torchscript for inference')
+parser.add_argument('--log-wandb', action='store_true', default=False,
+                    help='log training and validation metrics to wandb')
 
 
 def _parse_args():
@@ -294,6 +302,13 @@ def main():
     setup_default_logging()
     args, args_text = _parse_args()
     
+    if args.log_wandb:
+        if has_wandb:
+            wandb.init(project=args.experiment, config=args)
+        else: 
+            _logger.warning("You've requested to log metrics to wandb but package not found. "
+                            "Metrics not being logged to wandb, try `pip install wandb`")
+             
     args.prefetcher = not args.no_prefetcher
     args.distributed = False
     if 'WORLD_SIZE' in os.environ:
@@ -593,7 +608,7 @@ def main():
 
             update_summary(
                 epoch, train_metrics, eval_metrics, os.path.join(output_dir, 'summary.csv'),
-                write_header=best_metric is None)
+                write_header=best_metric is None, log_wandb=args.log_wandb and has_wandb)
 
             if saver is not None:
                 # save proper checkpoint with eval metric