tqdm status bar custom metrics formatting (#916)

This is a simple change that will allow custom formatting of metrics for the tqdm status bar. This is especially useful if you have small loss values and want to see it in scientific notation. Replace `metrics_format` with `'{name}: {value:e}'`

Author: lminer

tensorflow_addons/callbacks/tqdm_progress_bar.py

@@ -37,6 +37,8 @@ class TQDMProgressBar(Callback):
         update_per_second (int): Maximum number of updates in the epochs bar
             per second, this is to prevent small batches from slowing down
             training. Defaults to 10.
+        metrics_format (string): Custom format for how metrics are formatted.
+            See https://github.com/tqdm/tqdm#parameters for more detail.
         leave_epoch_progress (bool): True to leave epoch progress bars
         leave_overall_progress (bool): True to leave overall progress bar
         show_epoch_progress (bool): False to hide epoch progress bars
@@ -49,6 +51,7 @@ def __init__(self,
                  '{remaining}s,  {rate_fmt}{postfix}',
                  epoch_bar_format='{n_fmt}/{total_fmt}{bar} ETA: '
                  '{remaining}s - {desc}',
+                 metrics_format='{name}: {value:0.4f}',
                  update_per_second=10,
                  leave_epoch_progress=True,
                  leave_overall_progress=True,
@@ -75,6 +78,7 @@ def __init__(self,
         self.leave_overall_progress = leave_overall_progress
         self.show_epoch_progress = show_epoch_progress
         self.show_overall_progress = show_overall_progress
+        self.metrics_format = metrics_format
 
         # compute update interval (inverse of update per second)
         self.update_interval = 1 / update_per_second
@@ -196,7 +200,7 @@ def format_metrics(self, logs={}, factor=1):
         for metric in self.metrics:
             if metric in logs:
                 value = logs[metric] / factor
-                pair = '{name}: {value:0.4f}'.format(name=metric, value=value)
+                pair = self.metrics_format.format(name=metric, value=value)
                 metric_value_pairs.append(pair)
         metrics_string = self.metrics_separator.join(metric_value_pairs)
         return metrics_string