From c8e95b0ddc9eb2417c88a51b83e869d606c6fe0b Mon Sep 17 00:00:00 2001
From: D-X-Y <280835372@qq.com>
Date: Sat, 22 May 2021 09:43:48 +0000
Subject: [PATCH] Update LFNA

---
 exps/LFNA/lfna.py | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/exps/LFNA/lfna.py b/exps/LFNA/lfna.py
index d916cc4..3264c3e 100644
--- a/exps/LFNA/lfna.py
+++ b/exps/LFNA/lfna.py
@@ -100,9 +100,15 @@ def pretrain(base_model, meta_model, criterion, xenv, args, logger):
         weight_decay=args.weight_decay,
         amsgrad=True,
     )
+    logger.log("Pre-train the meta-model")
+    logger.log("Using the optimizer: {:}".format(optimizer))
 
     meta_model.set_best_dir(logger.path(None) / "checkpoint-pretrain")
+    per_epoch_time, start_time = AverageMeter(), time.time()
     for iepoch in range(args.epochs):
+        left_time = "Time Left: {:}".format(
+            convert_secs2time(per_epoch_time.avg * (args.epochs - iepoch), True)
+        )
         total_meta_losses, total_match_losses = [], []
         for ibatch in range(args.meta_batch):
             rand_index = random.randint(0, meta_model.meta_length - xenv.seq_length - 1)
@@ -151,7 +157,11 @@ def pretrain(base_model, meta_model, criterion, xenv, args, logger):
                 final_match_loss.item(),
             )
             + ", batch={:}".format(len(total_meta_losses))
+            + ", success={:}, best_score={:.4f}".format(success, -best_score)
+            + " {:}".format(left_time)
         )
+        per_epoch_time.update(time.time() - start_time)
+        start_time = time.time()
 
 
 def main(args):