Update README.md

README.md

@@ -10,36 +10,36 @@ training_args = TrainingArguments(
 )
 
 Step	Training Loss
-500	3.134100
-1000	2.987100
-1500	2.847400
-2000	2.730400
-2500	2.664100
-3000	2.555100
-3500	2.507200
-4000	2.468100
-4500	2.346600
-5000	2.333200
-5500	2.300300
-6000	2.230000
-6500	2.224100
-7000	2.150800
-7500	2.117000
-8000	2.097300
-8500	2.017600
-9000	2.000500
-9500	1.963300
-10000	1.927300
-10500	1.916700
-11000	1.862900
-11500	1.847900
-12000	1.833500
-12500	1.794200
-13000	1.780200
-13500	1.760300
-14000	1.741900
-14500	1.732500
-15000	1.710700
-15500	1.703900
-16000	1.698000
-TrainOutput(global_step=16060, training_loss=2.154029521253308, metrics={'train_runtime': 5497.0051, 'train_samples_per_second': 373.698, 'train_steps_per_second': 2.922, 'total_flos': 0.0, 'train_loss': 2.154029521253308, 'epoch': 20.0})
+500	5.111300
+1000	3.064000
+1500	2.899200
+2000	2.779200
+2500	2.710700
+3000	2.608300
+3500	2.546900
+4000	2.491100
+4500	2.404400
+5000	2.374700
+5500	2.324800
+6000	2.257300
+6500	2.239900
+7000	2.173400
+7500	2.134500
+8000	2.115700
+8500	2.046100
+9000	2.025600
+9500	1.989900
+10000	1.953900
+10500	1.940900
+11000	1.894000
+11500	1.872400
+12000	1.854300
+12500	1.823300
+13000	1.811900
+13500	1.789700
+14000	1.764800
+14500	1.753300
+15000	1.735000
+15500	1.727400
+16000	1.719900
+TrainOutput(global_step=16060, training_loss=2.2460614238848278, metrics={'train_runtime': 5541.9227, 'train_samples_per_second': 370.669, 'train_steps_per_second': 2.898, 'total_flos': 0.0, 'train_loss': 2.2460614238848278, 'epoch': 20.0})