From e9dc37f19c56921df26561a4764d77e3a74e7c51 Mon Sep 17 00:00:00 2001
From: James Betker <jbetker@gmail.com>
Date: Sat, 30 Oct 2021 17:00:24 -0600
Subject: [PATCH] Mod trainer to copy config file into experiments root

---
 codes/train.py                     | 12 ++++++++----
 codes/trainer/ExtensibleTrainer.py |  6 +-----
 2 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/codes/train.py b/codes/train.py
index d2e2b973..d66e222e 100644
--- a/codes/train.py
+++ b/codes/train.py
@@ -3,6 +3,8 @@ import math
 import argparse
 import random
 import logging
+import shutil
+
 from tqdm import tqdm
 
 import torch
@@ -13,6 +15,7 @@ from utils import util, options as option
 from data import create_dataloader, create_dataset
 from trainer.ExtensibleTrainer import ExtensibleTrainer
 from time import time
+from datetime import datetime
 
 from utils.util import opt_get
 
@@ -32,7 +35,7 @@ def init_dist(backend, **kwargs):
 
 class Trainer:
 
-    def init(self, opt, launcher, all_networks={}):
+    def init(self, opt_path, opt, launcher):
         self._profile = False
         self.val_compute_psnr = opt_get(opt, ['eval', 'compute_psnr'], False)
         self.val_compute_fea = opt_get(opt, ['eval', 'compute_fea'], False)
@@ -54,6 +57,7 @@ class Trainer:
                 util.mkdirs(
                     (path for key, path in opt['path'].items() if not key == 'experiments_root' and path is not None
                      and 'pretrain_model' not in key and 'resume' not in key))
+            shutil.copy(opt_path, os.path.join(opt['path']['experiments_root'], f'{datetime.now().strftime("%d%m%Y_%H%M%S")}_{os.path.basename(opt_path)}'))
 
             # config loggers. Before it, the log will not work
             util.setup_logger('base', opt['path']['log'], 'train_' + opt['name'], level=logging.INFO,
@@ -131,7 +135,7 @@ class Trainer:
         assert self.train_loader is not None
 
         #### create model
-        self.model = ExtensibleTrainer(opt, cached_networks=all_networks)
+        self.model = ExtensibleTrainer(opt)
 
         ### Evaluators
         self.evaluators = []
@@ -280,7 +284,7 @@ class Trainer:
 
 if __name__ == '__main__':
     parser = argparse.ArgumentParser()
-    parser.add_argument('-opt', type=str, help='Path to option YAML file.', default='../options/train_noisy_audio_clips_classifier.yml')
+    parser.add_argument('-opt', type=str, help='Path to option YAML file.', default='../options/train_gpt_asr_mass_distill.yml')
     parser.add_argument('--launcher', choices=['none', 'pytorch'], default='none', help='job launcher')
     parser.add_argument('--local_rank', type=int, default=0)
     args = parser.parse_args()
@@ -306,5 +310,5 @@ if __name__ == '__main__':
         trainer.world_size = torch.distributed.get_world_size()
         trainer.rank = torch.distributed.get_rank()
 
-    trainer.init(opt, args.launcher)
+    trainer.init(args.opt, opt, args.launcher)
     trainer.do_training()
diff --git a/codes/trainer/ExtensibleTrainer.py b/codes/trainer/ExtensibleTrainer.py
index edabeab0..e2c01ea6 100644
--- a/codes/trainer/ExtensibleTrainer.py
+++ b/codes/trainer/ExtensibleTrainer.py
@@ -21,7 +21,7 @@ logger = logging.getLogger('base')
 
 
 class ExtensibleTrainer(BaseModel):
-    def __init__(self, opt, cached_networks={}):
+    def __init__(self, opt):
         super(ExtensibleTrainer, self).__init__(opt)
         if opt['dist']:
             self.rank = torch.distributed.get_rank()
@@ -54,10 +54,6 @@ class ExtensibleTrainer(BaseModel):
             if 'trainable' not in net.keys():
                 net['trainable'] = True
 
-            if name in cached_networks.keys():
-                new_net = cached_networks[name]
-            else:
-                new_net = None
             if net['type'] == 'generator':
                 if new_net is None:
                     new_net = networks.create_model(opt, net, self.netsG).to(self.device)