Three-Stage Fine-Tuning Experiment For Transfer Learning

I wrote an app to classify an object using transfer learning, and the model was trained on CIFAR-10 dataset, all by myself.

Validating the Hypothesis: Does Stage-3 Fine-Tuning Actually Help?

Revisiting the Prediction

Earlier, we argued—based purely on learning signals and capacity analysis—that Stage-3 fine-tuning was unlikely to deliver meaningful gains. Importantly, that conclusion was formed before running the experiment. Now we can evaluate that hypothesis against real data.

---

Two-Stage Fine-Tuning: Reference Performance

The best results from the two-stage setup were:

Stage-1 (fc only)

Train Loss: 1.1005 | Val Loss: 1.3094
Train Acc: 85.70% | Val Acc: 72.50%

Stage-2 (layer4 + fc)

Train Loss: 0.7060 | Val Loss: 1.0932
Train Acc: 100.00% | Val Acc: 86.20%

This establishes a strong and stable baseline. Most of the transferable signal is clearly captured during Stage-2, with validation accuracy reaching 86.2%.

---

Three-Stage Fine-Tuning: Empirical Results

Now consider the corresponding three-stage experiment.

Stage-1 (fc only)

Train Loss: 1.2423 | Val Loss: 1.3520
Train Acc: 78.70% | Val Acc: 71.80%

Stage-2 (layer4 + fc)

Train Loss: 0.7114 | Val Loss: 1.0743
Train Acc: 100.00% | Val Acc: 86.30%

Stage-3 (layer3 + layer4 + fc)

Train Loss: 0.7084 | Val Loss: 1.0758
Train Acc: 100.00% | Val Acc: 86.80%

The final validation accuracy improves by only 0.6% relative to the two-stage setup.

---

Interpreting the Marginal Gain from Stage-3

A 0.6% improvement is not meaningless, but it is structurally small given the amount of additional capacity and risk introduced by unfreezing layer3. Several key observations stand out:

• Training accuracy reaches 100% as early as Stage-2
• Validation loss remains essentially unchanged across Stage-2 and Stage-3
• The train–validation gap does not meaningfully shrink

This pattern strongly suggests that Stage-3 is not addressing the dominant source of error.

---

Is It True That 100% Train Accuracy Limits Validation Gains?

The intuition is directionally correct, and it can be stated more precisely. When training accuracy reaches 100%:

• The model has enough capacity to perfectly fit the training distribution
• Additional parameter updates mostly reshape decision boundaries around training samples
• Generalization error becomes dominated by data limitations, not model expressiveness

In this regime, increasing capacity or unfreezing deeper layers rarely improves validation accuracy unless:

• The dataset is substantially larger
• Label noise is reduced
• Stronger invariances are introduced via data augmentation

Without these, deeper fine-tuning tends to yield diminishing returns.

---

Why Stage-3 Only Helps Marginally Here

Unfreezing layer3 allows the network to modify mid-level representations, but the data indicates that:

• Mid-level features were already sufficient
• High-level semantic alignment (Stage-2) did most of the work
• The remaining errors are not representational

In other words, the model is no longer “missing the right features” — it is limited by the amount of information available to learn robust invariances.

---

Why This Outcome Is Still a Success

The goal of Stage-3 was not to chase numbers blindly, but to validate a hypothesis. The experiment confirms:

• The earlier analysis was directionally correct
• Two-stage fine-tuning is near-optimal for this dataset
• Further gains require changing the data regime, not the backbone depth

This is exactly what a well-designed ablation study should do: reduce uncertainty.

---

Key Takeaway

Stage-3 fine-tuning provides only marginal gains once training accuracy saturates, because the bottleneck shifts from representation learning to data-driven generalization. When train accuracy reaches 100%, deeper fine-tuning rarely yields large validation improvements unless the data itself changes. The results validate a disciplined, data-first approach to transfer learning — and demonstrate why stopping at two stages is often the most rational choice.

Raw Config & Logs

2-stage fine-tuning:

========== CONFIG ==========
NUM_CLASSES = 10
TRAINING_SAMPLE_PER_CLASS = 100
VALIDATION_SAMPLE_PER_CLASS = 100
BATCH_SIZE = 256
EPOCHS = 100
TRAINABLE_LAYERS_STAGE1 = 1
TRAINABLE_LAYERS_STAGE2 = 2
EARLY_STOP_PATIENCE = 100
USE_COSINE_LR = True
COSINE_T_MAX = 100
COSINE_ETA_MIN = 1e-06
============================

===== Stage-1: Train fc only =====
[01/100] Train Loss: 2.4508 | Val Loss: 2.3878 | Train Acc: 8.80% | Val Acc: 15.10%
[02/100] Train Loss: 2.2751 | Val Loss: 2.1953 | Train Acc: 18.10% | Val Acc: 22.70%
[03/100] Train Loss: 2.0997 | Val Loss: 2.0263 | Train Acc: 30.60% | Val Acc: 31.90%
[04/100] Train Loss: 1.9801 | Val Loss: 1.8800 | Train Acc: 37.50% | Val Acc: 40.60%
[05/100] Train Loss: 1.8626 | Val Loss: 1.7790 | Train Acc: 46.20% | Val Acc: 49.10%
[06/100] Train Loss: 1.7675 | Val Loss: 1.6993 | Train Acc: 53.50% | Val Acc: 56.20%
[07/100] Train Loss: 1.6714 | Val Loss: 1.6340 | Train Acc: 61.90% | Val Acc: 58.90%
[08/100] Train Loss: 1.5955 | Val Loss: 1.5817 | Train Acc: 64.90% | Val Acc: 62.80%
[09/100] Train Loss: 1.5356 | Val Loss: 1.5417 | Train Acc: 66.90% | Val Acc: 64.10%
[10/100] Train Loss: 1.4891 | Val Loss: 1.5128 | Train Acc: 70.70% | Val Acc: 65.10%
[11/100] Train Loss: 1.4518 | Val Loss: 1.4882 | Train Acc: 72.60% | Val Acc: 65.90%
[12/100] Train Loss: 1.4148 | Val Loss: 1.4558 | Train Acc: 72.60% | Val Acc: 67.80%
[13/100] Train Loss: 1.3866 | Val Loss: 1.4393 | Train Acc: 72.90% | Val Acc: 69.40%
[14/100] Train Loss: 1.3460 | Val Loss: 1.4256 | Train Acc: 76.40% | Val Acc: 69.60%
[15/100] Train Loss: 1.3213 | Val Loss: 1.4103 | Train Acc: 76.90% | Val Acc: 69.00%
[16/100] Train Loss: 1.3125 | Val Loss: 1.3959 | Train Acc: 75.40% | Val Acc: 70.10%
[17/100] Train Loss: 1.2905 | Val Loss: 1.3851 | Train Acc: 76.40% | Val Acc: 70.10%
[18/100] Train Loss: 1.2784 | Val Loss: 1.3807 | Train Acc: 78.20% | Val Acc: 70.60%
[19/100] Train Loss: 1.2698 | Val Loss: 1.3745 | Train Acc: 77.10% | Val Acc: 70.40%
[20/100] Train Loss: 1.2544 | Val Loss: 1.3661 | Train Acc: 79.20% | Val Acc: 70.00%
[21/100] Train Loss: 1.2566 | Val Loss: 1.3573 | Train Acc: 78.50% | Val Acc: 70.80%
[22/100] Train Loss: 1.2472 | Val Loss: 1.3563 | Train Acc: 79.00% | Val Acc: 71.30%
[23/100] Train Loss: 1.2398 | Val Loss: 1.3511 | Train Acc: 79.00% | Val Acc: 71.00%
[24/100] Train Loss: 1.2123 | Val Loss: 1.3458 | Train Acc: 81.00% | Val Acc: 71.10%
[25/100] Train Loss: 1.2169 | Val Loss: 1.3359 | Train Acc: 78.80% | Val Acc: 71.90%
[26/100] Train Loss: 1.2062 | Val Loss: 1.3362 | Train Acc: 80.30% | Val Acc: 71.80%
[27/100] Train Loss: 1.1997 | Val Loss: 1.3420 | Train Acc: 80.80% | Val Acc: 70.60%
[28/100] Train Loss: 1.1865 | Val Loss: 1.3348 | Train Acc: 83.00% | Val Acc: 71.60%
[29/100] Train Loss: 1.1775 | Val Loss: 1.3280 | Train Acc: 82.40% | Val Acc: 71.60%
[30/100] Train Loss: 1.1846 | Val Loss: 1.3287 | Train Acc: 80.40% | Val Acc: 72.00%
[31/100] Train Loss: 1.1783 | Val Loss: 1.3267 | Train Acc: 81.20% | Val Acc: 71.60%
[32/100] Train Loss: 1.1767 | Val Loss: 1.3296 | Train Acc: 82.20% | Val Acc: 71.30%
[33/100] Train Loss: 1.1755 | Val Loss: 1.3268 | Train Acc: 82.50% | Val Acc: 70.90%
[34/100] Train Loss: 1.1769 | Val Loss: 1.3229 | Train Acc: 81.20% | Val Acc: 71.70%
[35/100] Train Loss: 1.1740 | Val Loss: 1.3204 | Train Acc: 83.60% | Val Acc: 72.00%
[36/100] Train Loss: 1.1427 | Val Loss: 1.3186 | Train Acc: 83.60% | Val Acc: 71.80%
[37/100] Train Loss: 1.1553 | Val Loss: 1.3218 | Train Acc: 81.80% | Val Acc: 70.90%
[38/100] Train Loss: 1.1466 | Val Loss: 1.3254 | Train Acc: 83.30% | Val Acc: 71.40%
[39/100] Train Loss: 1.1515 | Val Loss: 1.3191 | Train Acc: 83.40% | Val Acc: 71.40%
[40/100] Train Loss: 1.1280 | Val Loss: 1.3183 | Train Acc: 84.80% | Val Acc: 71.50%
[41/100] Train Loss: 1.1382 | Val Loss: 1.3212 | Train Acc: 84.30% | Val Acc: 70.70%
[42/100] Train Loss: 1.1277 | Val Loss: 1.3164 | Train Acc: 84.20% | Val Acc: 71.80%
[43/100] Train Loss: 1.1515 | Val Loss: 1.3144 | Train Acc: 82.90% | Val Acc: 72.00%
[44/100] Train Loss: 1.1391 | Val Loss: 1.3151 | Train Acc: 83.00% | Val Acc: 71.30%
[45/100] Train Loss: 1.1320 | Val Loss: 1.3171 | Train Acc: 83.40% | Val Acc: 70.90%
[46/100] Train Loss: 1.1386 | Val Loss: 1.3156 | Train Acc: 82.40% | Val Acc: 71.10%
[47/100] Train Loss: 1.1206 | Val Loss: 1.3122 | Train Acc: 85.50% | Val Acc: 72.30%
[48/100] Train Loss: 1.1282 | Val Loss: 1.3127 | Train Acc: 84.60% | Val Acc: 71.60%
[49/100] Train Loss: 1.1247 | Val Loss: 1.3120 | Train Acc: 84.00% | Val Acc: 71.00%
[50/100] Train Loss: 1.1171 | Val Loss: 1.3127 | Train Acc: 84.60% | Val Acc: 71.20%
[51/100] Train Loss: 1.1130 | Val Loss: 1.3160 | Train Acc: 85.70% | Val Acc: 71.50%
[52/100] Train Loss: 1.1169 | Val Loss: 1.3134 | Train Acc: 83.60% | Val Acc: 72.30%
[53/100] Train Loss: 1.1080 | Val Loss: 1.3143 | Train Acc: 85.40% | Val Acc: 71.60%
[54/100] Train Loss: 1.1040 | Val Loss: 1.3131 | Train Acc: 85.80% | Val Acc: 71.20%
[55/100] Train Loss: 1.1029 | Val Loss: 1.3125 | Train Acc: 85.00% | Val Acc: 71.40%
[56/100] Train Loss: 1.1153 | Val Loss: 1.3133 | Train Acc: 84.50% | Val Acc: 71.40%
[57/100] Train Loss: 1.1137 | Val Loss: 1.3117 | Train Acc: 86.00% | Val Acc: 71.50%
[58/100] Train Loss: 1.0955 | Val Loss: 1.3099 | Train Acc: 86.10% | Val Acc: 72.40%
[59/100] Train Loss: 1.1120 | Val Loss: 1.3102 | Train Acc: 85.00% | Val Acc: 71.70%
[60/100] Train Loss: 1.1045 | Val Loss: 1.3101 | Train Acc: 86.20% | Val Acc: 71.80%
[61/100] Train Loss: 1.1071 | Val Loss: 1.3122 | Train Acc: 85.80% | Val Acc: 72.20%
[62/100] Train Loss: 1.1005 | Val Loss: 1.3094 | Train Acc: 85.70% | Val Acc: 72.50%
[63/100] Train Loss: 1.1030 | Val Loss: 1.3072 | Train Acc: 86.20% | Val Acc: 71.40%
[64/100] Train Loss: 1.1120 | Val Loss: 1.3066 | Train Acc: 84.60% | Val Acc: 71.70%
[65/100] Train Loss: 1.0953 | Val Loss: 1.3072 | Train Acc: 86.40% | Val Acc: 71.30%
[66/100] Train Loss: 1.1009 | Val Loss: 1.3085 | Train Acc: 85.10% | Val Acc: 71.60%
[67/100] Train Loss: 1.0936 | Val Loss: 1.3081 | Train Acc: 87.00% | Val Acc: 71.60%
[68/100] Train Loss: 1.0929 | Val Loss: 1.3064 | Train Acc: 86.30% | Val Acc: 71.60%
[69/100] Train Loss: 1.0919 | Val Loss: 1.3084 | Train Acc: 86.30% | Val Acc: 71.50%
[70/100] Train Loss: 1.0904 | Val Loss: 1.3099 | Train Acc: 86.40% | Val Acc: 71.80%
[71/100] Train Loss: 1.0885 | Val Loss: 1.3077 | Train Acc: 87.00% | Val Acc: 72.00%
[72/100] Train Loss: 1.0880 | Val Loss: 1.3059 | Train Acc: 86.40% | Val Acc: 71.70%
[73/100] Train Loss: 1.1060 | Val Loss: 1.3062 | Train Acc: 84.50% | Val Acc: 71.80%
[74/100] Train Loss: 1.1173 | Val Loss: 1.3056 | Train Acc: 84.60% | Val Acc: 72.00%
[75/100] Train Loss: 1.0906 | Val Loss: 1.3038 | Train Acc: 88.20% | Val Acc: 72.00%
[76/100] Train Loss: 1.0939 | Val Loss: 1.3049 | Train Acc: 86.90% | Val Acc: 72.20%
[77/100] Train Loss: 1.0885 | Val Loss: 1.3087 | Train Acc: 87.50% | Val Acc: 72.10%
[78/100] Train Loss: 1.0907 | Val Loss: 1.3092 | Train Acc: 87.10% | Val Acc: 71.90%
[79/100] Train Loss: 1.1004 | Val Loss: 1.3116 | Train Acc: 84.40% | Val Acc: 71.30%
[80/100] Train Loss: 1.0904 | Val Loss: 1.3104 | Train Acc: 86.20% | Val Acc: 71.50%
[81/100] Train Loss: 1.0877 | Val Loss: 1.3106 | Train Acc: 87.20% | Val Acc: 71.80%
[82/100] Train Loss: 1.0919 | Val Loss: 1.3098 | Train Acc: 86.10% | Val Acc: 71.80%
[83/100] Train Loss: 1.0964 | Val Loss: 1.3068 | Train Acc: 85.90% | Val Acc: 71.80%
[84/100] Train Loss: 1.0896 | Val Loss: 1.3064 | Train Acc: 86.90% | Val Acc: 71.60%
[85/100] Train Loss: 1.1009 | Val Loss: 1.3088 | Train Acc: 85.60% | Val Acc: 71.70%
[86/100] Train Loss: 1.0908 | Val Loss: 1.3086 | Train Acc: 86.10% | Val Acc: 71.60%
[87/100] Train Loss: 1.0846 | Val Loss: 1.3057 | Train Acc: 86.70% | Val Acc: 71.90%
[88/100] Train Loss: 1.0979 | Val Loss: 1.3065 | Train Acc: 85.30% | Val Acc: 72.00%
[89/100] Train Loss: 1.1118 | Val Loss: 1.3069 | Train Acc: 84.50% | Val Acc: 72.10%
[90/100] Train Loss: 1.1001 | Val Loss: 1.3068 | Train Acc: 85.70% | Val Acc: 72.20%
[91/100] Train Loss: 1.0903 | Val Loss: 1.3067 | Train Acc: 86.40% | Val Acc: 72.30%
[92/100] Train Loss: 1.0863 | Val Loss: 1.3054 | Train Acc: 87.00% | Val Acc: 72.20%
[93/100] Train Loss: 1.0797 | Val Loss: 1.3066 | Train Acc: 87.90% | Val Acc: 71.90%
[94/100] Train Loss: 1.0980 | Val Loss: 1.3085 | Train Acc: 85.40% | Val Acc: 72.10%
[95/100] Train Loss: 1.0910 | Val Loss: 1.3069 | Train Acc: 86.20% | Val Acc: 71.80%
[96/100] Train Loss: 1.0819 | Val Loss: 1.3068 | Train Acc: 86.60% | Val Acc: 71.60%
[97/100] Train Loss: 1.0825 | Val Loss: 1.3077 | Train Acc: 87.40% | Val Acc: 71.90%
[98/100] Train Loss: 1.0918 | Val Loss: 1.3063 | Train Acc: 86.40% | Val Acc: 72.00%
[99/100] Train Loss: 1.0879 | Val Loss: 1.3071 | Train Acc: 86.20% | Val Acc: 71.80%
[100/100] Train Loss: 1.0795 | Val Loss: 1.3069 | Train Acc: 86.10% | Val Acc: 71.90%

Stage-1 Best results:
Train Loss: 1.1005 | Val Loss: 1.3094 | Train Acc: 85.70% | Val Acc: 72.50%
Stage-1 Training Time: 357.59 seconds

Loaded Stage-1 best-val model for Stage-2 fine-tuning

===== Stage-2: Unfreeze layer4 + fc =====
[01/100] Train Loss: 1.1014 | Val Loss: 1.2606 | Train Acc: 85.30% | Val Acc: 76.90%
[02/100] Train Loss: 0.9456 | Val Loss: 1.2211 | Train Acc: 94.40% | Val Acc: 79.30%
[03/100] Train Loss: 0.8781 | Val Loss: 1.2030 | Train Acc: 97.10% | Val Acc: 79.90%
[04/100] Train Loss: 0.8319 | Val Loss: 1.1859 | Train Acc: 99.10% | Val Acc: 80.40%
[05/100] Train Loss: 0.8099 | Val Loss: 1.1695 | Train Acc: 99.60% | Val Acc: 79.50%
[06/100] Train Loss: 0.7852 | Val Loss: 1.1605 | Train Acc: 99.80% | Val Acc: 79.80%
[07/100] Train Loss: 0.7728 | Val Loss: 1.1526 | Train Acc: 99.90% | Val Acc: 81.00%
[08/100] Train Loss: 0.7626 | Val Loss: 1.1463 | Train Acc: 99.90% | Val Acc: 81.30%
[09/100] Train Loss: 0.7574 | Val Loss: 1.1410 | Train Acc: 100.00% | Val Acc: 81.40%
[10/100] Train Loss: 0.7465 | Val Loss: 1.1364 | Train Acc: 100.00% | Val Acc: 81.40%
[11/100] Train Loss: 0.7452 | Val Loss: 1.1256 | Train Acc: 100.00% | Val Acc: 82.20%
[12/100] Train Loss: 0.7389 | Val Loss: 1.1229 | Train Acc: 100.00% | Val Acc: 82.60%
[13/100] Train Loss: 0.7353 | Val Loss: 1.1266 | Train Acc: 100.00% | Val Acc: 82.20%
[14/100] Train Loss: 0.7313 | Val Loss: 1.1261 | Train Acc: 100.00% | Val Acc: 81.40%
[15/100] Train Loss: 0.7310 | Val Loss: 1.1156 | Train Acc: 100.00% | Val Acc: 81.80%
[16/100] Train Loss: 0.7271 | Val Loss: 1.1067 | Train Acc: 100.00% | Val Acc: 83.10%
[17/100] Train Loss: 0.7241 | Val Loss: 1.1071 | Train Acc: 100.00% | Val Acc: 82.80%
[18/100] Train Loss: 0.7236 | Val Loss: 1.1076 | Train Acc: 100.00% | Val Acc: 82.90%
[19/100] Train Loss: 0.7223 | Val Loss: 1.1079 | Train Acc: 100.00% | Val Acc: 83.40%
[20/100] Train Loss: 0.7217 | Val Loss: 1.1039 | Train Acc: 100.00% | Val Acc: 83.90%
[21/100] Train Loss: 0.7194 | Val Loss: 1.1026 | Train Acc: 100.00% | Val Acc: 84.10%
[22/100] Train Loss: 0.7175 | Val Loss: 1.1039 | Train Acc: 100.00% | Val Acc: 83.20%
[23/100] Train Loss: 0.7186 | Val Loss: 1.1018 | Train Acc: 100.00% | Val Acc: 83.40%
[24/100] Train Loss: 0.7176 | Val Loss: 1.1090 | Train Acc: 100.00% | Val Acc: 82.70%
[25/100] Train Loss: 0.7173 | Val Loss: 1.1082 | Train Acc: 100.00% | Val Acc: 82.80%
[26/100] Train Loss: 0.7164 | Val Loss: 1.1020 | Train Acc: 100.00% | Val Acc: 84.10%
[27/100] Train Loss: 0.7152 | Val Loss: 1.0983 | Train Acc: 100.00% | Val Acc: 84.00%
[28/100] Train Loss: 0.7151 | Val Loss: 1.1011 | Train Acc: 100.00% | Val Acc: 84.80%
[29/100] Train Loss: 0.7146 | Val Loss: 1.1012 | Train Acc: 100.00% | Val Acc: 84.50%
[30/100] Train Loss: 0.7156 | Val Loss: 1.1006 | Train Acc: 100.00% | Val Acc: 84.70%
[31/100] Train Loss: 0.7131 | Val Loss: 1.0972 | Train Acc: 100.00% | Val Acc: 84.90%
[32/100] Train Loss: 0.7119 | Val Loss: 1.0933 | Train Acc: 100.00% | Val Acc: 84.80%
[33/100] Train Loss: 0.7113 | Val Loss: 1.0972 | Train Acc: 100.00% | Val Acc: 84.50%
[34/100] Train Loss: 0.7116 | Val Loss: 1.0966 | Train Acc: 100.00% | Val Acc: 85.20%
[35/100] Train Loss: 0.7117 | Val Loss: 1.0912 | Train Acc: 100.00% | Val Acc: 84.70%
[36/100] Train Loss: 0.7107 | Val Loss: 1.0941 | Train Acc: 100.00% | Val Acc: 84.50%
[37/100] Train Loss: 0.7109 | Val Loss: 1.0971 | Train Acc: 100.00% | Val Acc: 84.60%
[38/100] Train Loss: 0.7098 | Val Loss: 1.0979 | Train Acc: 100.00% | Val Acc: 84.90%
[39/100] Train Loss: 0.7098 | Val Loss: 1.1001 | Train Acc: 100.00% | Val Acc: 84.90%
[40/100] Train Loss: 0.7096 | Val Loss: 1.0962 | Train Acc: 100.00% | Val Acc: 84.40%
[41/100] Train Loss: 0.7090 | Val Loss: 1.0928 | Train Acc: 100.00% | Val Acc: 85.00%
[42/100] Train Loss: 0.7072 | Val Loss: 1.0905 | Train Acc: 100.00% | Val Acc: 85.10%
[43/100] Train Loss: 0.7073 | Val Loss: 1.0896 | Train Acc: 100.00% | Val Acc: 85.30%
[44/100] Train Loss: 0.7093 | Val Loss: 1.0881 | Train Acc: 100.00% | Val Acc: 85.60%
[45/100] Train Loss: 0.7069 | Val Loss: 1.0909 | Train Acc: 100.00% | Val Acc: 85.50%
[46/100] Train Loss: 0.7075 | Val Loss: 1.0952 | Train Acc: 100.00% | Val Acc: 85.50%
[47/100] Train Loss: 0.7062 | Val Loss: 1.0945 | Train Acc: 100.00% | Val Acc: 85.20%
[48/100] Train Loss: 0.7062 | Val Loss: 1.0910 | Train Acc: 100.00% | Val Acc: 85.50%
[49/100] Train Loss: 0.7064 | Val Loss: 1.0867 | Train Acc: 100.00% | Val Acc: 85.70%
[50/100] Train Loss: 0.7062 | Val Loss: 1.0891 | Train Acc: 100.00% | Val Acc: 85.70%
[51/100] Train Loss: 0.7060 | Val Loss: 1.0932 | Train Acc: 100.00% | Val Acc: 86.20%
[52/100] Train Loss: 0.7061 | Val Loss: 1.0933 | Train Acc: 100.00% | Val Acc: 85.80%
[53/100] Train Loss: 0.7064 | Val Loss: 1.0913 | Train Acc: 100.00% | Val Acc: 85.70%
[54/100] Train Loss: 0.7056 | Val Loss: 1.0887 | Train Acc: 100.00% | Val Acc: 85.40%
[55/100] Train Loss: 0.7056 | Val Loss: 1.0910 | Train Acc: 100.00% | Val Acc: 85.30%
[56/100] Train Loss: 0.7065 | Val Loss: 1.0905 | Train Acc: 100.00% | Val Acc: 85.40%
[57/100] Train Loss: 0.7050 | Val Loss: 1.0941 | Train Acc: 100.00% | Val Acc: 85.50%
[58/100] Train Loss: 0.7049 | Val Loss: 1.0940 | Train Acc: 100.00% | Val Acc: 85.60%
[59/100] Train Loss: 0.7044 | Val Loss: 1.0915 | Train Acc: 100.00% | Val Acc: 85.70%
[60/100] Train Loss: 0.7051 | Val Loss: 1.0893 | Train Acc: 100.00% | Val Acc: 86.00%
[61/100] Train Loss: 0.7048 | Val Loss: 1.0884 | Train Acc: 100.00% | Val Acc: 85.70%
[62/100] Train Loss: 0.7049 | Val Loss: 1.0920 | Train Acc: 100.00% | Val Acc: 85.40%
[63/100] Train Loss: 0.7041 | Val Loss: 1.0922 | Train Acc: 100.00% | Val Acc: 85.30%
[64/100] Train Loss: 0.7039 | Val Loss: 1.0938 | Train Acc: 100.00% | Val Acc: 85.30%
[65/100] Train Loss: 0.7036 | Val Loss: 1.0945 | Train Acc: 100.00% | Val Acc: 85.40%
[66/100] Train Loss: 0.7035 | Val Loss: 1.0937 | Train Acc: 100.00% | Val Acc: 85.40%
[67/100] Train Loss: 0.7047 | Val Loss: 1.0891 | Train Acc: 100.00% | Val Acc: 85.60%
[68/100] Train Loss: 0.7033 | Val Loss: 1.0885 | Train Acc: 100.00% | Val Acc: 86.00%
[69/100] Train Loss: 0.7044 | Val Loss: 1.0900 | Train Acc: 100.00% | Val Acc: 85.60%
[70/100] Train Loss: 0.7038 | Val Loss: 1.0905 | Train Acc: 100.00% | Val Acc: 85.70%
[71/100] Train Loss: 0.7031 | Val Loss: 1.0914 | Train Acc: 100.00% | Val Acc: 85.80%
[72/100] Train Loss: 0.7030 | Val Loss: 1.0931 | Train Acc: 100.00% | Val Acc: 85.40%
[73/100] Train Loss: 0.7030 | Val Loss: 1.0937 | Train Acc: 100.00% | Val Acc: 85.70%
[74/100] Train Loss: 0.7040 | Val Loss: 1.0926 | Train Acc: 100.00% | Val Acc: 85.80%
[75/100] Train Loss: 0.7034 | Val Loss: 1.0905 | Train Acc: 100.00% | Val Acc: 86.00%
[76/100] Train Loss: 0.7027 | Val Loss: 1.0898 | Train Acc: 100.00% | Val Acc: 85.90%
[77/100] Train Loss: 0.7033 | Val Loss: 1.0887 | Train Acc: 100.00% | Val Acc: 85.80%
[78/100] Train Loss: 0.7030 | Val Loss: 1.0894 | Train Acc: 100.00% | Val Acc: 85.80%
[79/100] Train Loss: 0.7024 | Val Loss: 1.0889 | Train Acc: 100.00% | Val Acc: 85.90%
[80/100] Train Loss: 0.7030 | Val Loss: 1.0903 | Train Acc: 100.00% | Val Acc: 85.20%
[81/100] Train Loss: 0.7036 | Val Loss: 1.0907 | Train Acc: 100.00% | Val Acc: 85.10%
[82/100] Train Loss: 0.7033 | Val Loss: 1.0908 | Train Acc: 100.00% | Val Acc: 84.90%
[83/100] Train Loss: 0.7024 | Val Loss: 1.0898 | Train Acc: 100.00% | Val Acc: 85.10%
[84/100] Train Loss: 0.7032 | Val Loss: 1.0893 | Train Acc: 100.00% | Val Acc: 85.90%
[85/100] Train Loss: 0.7033 | Val Loss: 1.0886 | Train Acc: 100.00% | Val Acc: 85.50%
[86/100] Train Loss: 0.7034 | Val Loss: 1.0886 | Train Acc: 100.00% | Val Acc: 85.80%
[87/100] Train Loss: 0.7032 | Val Loss: 1.0885 | Train Acc: 100.00% | Val Acc: 85.70%
[88/100] Train Loss: 0.7024 | Val Loss: 1.0897 | Train Acc: 100.00% | Val Acc: 85.40%
[89/100] Train Loss: 0.7024 | Val Loss: 1.0901 | Train Acc: 100.00% | Val Acc: 85.40%
[90/100] Train Loss: 0.7033 | Val Loss: 1.0914 | Train Acc: 100.00% | Val Acc: 85.30%
[91/100] Train Loss: 0.7022 | Val Loss: 1.0915 | Train Acc: 100.00% | Val Acc: 85.30%
[92/100] Train Loss: 0.7022 | Val Loss: 1.0913 | Train Acc: 100.00% | Val Acc: 85.30%
[93/100] Train Loss: 0.7024 | Val Loss: 1.0904 | Train Acc: 100.00% | Val Acc: 85.70%
[94/100] Train Loss: 0.7024 | Val Loss: 1.0912 | Train Acc: 100.00% | Val Acc: 85.60%
[95/100] Train Loss: 0.7030 | Val Loss: 1.0906 | Train Acc: 100.00% | Val Acc: 85.60%
[96/100] Train Loss: 0.7026 | Val Loss: 1.0908 | Train Acc: 100.00% | Val Acc: 85.50%
[97/100] Train Loss: 0.7025 | Val Loss: 1.0921 | Train Acc: 100.00% | Val Acc: 85.40%
[98/100] Train Loss: 0.7019 | Val Loss: 1.0923 | Train Acc: 100.00% | Val Acc: 85.50%
[99/100] Train Loss: 0.7014 | Val Loss: 1.0924 | Train Acc: 100.00% | Val Acc: 85.30%
[100/100] Train Loss: 0.7032 | Val Loss: 1.0914 | Train Acc: 100.00% | Val Acc: 85.70%

Stage-2 Best results:
Train Loss: 0.7060 | Val Loss: 1.0932 | Train Acc: 100.00% | Val Acc: 86.20%
Stage-2 Training Time: 464.50 seconds

3-stage fine-tuning:

========== CONFIG ==========
NUM_CLASSES = 10
TRAINING_SAMPLE_PER_CLASS = 100
VALIDATION_SAMPLE_PER_CLASS = 100
BATCH_SIZE = 256
EPOCHS per stage = 50
TRAINABLE_LAYERS_STAGE1 = 1
TRAINABLE_LAYERS_STAGE2 = 2
TRAINABLE_LAYERS_STAGE3 = 3
LR_STAGE1 = 0.001
LR_STAGE2 = 0.0001
LR_STAGE3 = 1e-05
EARLY_STOP_PATIENCE = 100
USE_COSINE_LR = True
COSINE_T_MAX = 50
COSINE_ETA_MIN = 1e-06
============================

===== Stage-1: Train fc only =====
[01/50] Train Loss: 2.4508 | Val Loss: 2.3878 | Train Acc: 8.80% | Val Acc: 15.10%
[02/50] Train Loss: 2.2751 | Val Loss: 2.1954 | Train Acc: 18.10% | Val Acc: 22.70%
[03/50] Train Loss: 2.0999 | Val Loss: 2.0268 | Train Acc: 30.60% | Val Acc: 31.90%
[04/50] Train Loss: 1.9808 | Val Loss: 1.8812 | Train Acc: 37.50% | Val Acc: 40.60%
[05/50] Train Loss: 1.8641 | Val Loss: 1.7809 | Train Acc: 45.90% | Val Acc: 48.90%
[06/50] Train Loss: 1.7699 | Val Loss: 1.7025 | Train Acc: 53.50% | Val Acc: 56.10%
[07/50] Train Loss: 1.6754 | Val Loss: 1.6384 | Train Acc: 61.80% | Val Acc: 58.50%
[08/50] Train Loss: 1.6006 | Val Loss: 1.5873 | Train Acc: 64.80% | Val Acc: 62.30%
[09/50] Train Loss: 1.5422 | Val Loss: 1.5480 | Train Acc: 66.80% | Val Acc: 63.70%
[10/50] Train Loss: 1.4972 | Val Loss: 1.5200 | Train Acc: 70.40% | Val Acc: 65.00%
[11/50] Train Loss: 1.4611 | Val Loss: 1.4970 | Train Acc: 72.20% | Val Acc: 65.50%
[12/50] Train Loss: 1.4256 | Val Loss: 1.4653 | Train Acc: 71.80% | Val Acc: 67.70%
[13/50] Train Loss: 1.3984 | Val Loss: 1.4491 | Train Acc: 72.80% | Val Acc: 69.10%
[14/50] Train Loss: 1.3595 | Val Loss: 1.4353 | Train Acc: 75.60% | Val Acc: 69.70%
[15/50] Train Loss: 1.3363 | Val Loss: 1.4223 | Train Acc: 75.80% | Val Acc: 68.80%
[16/50] Train Loss: 1.3279 | Val Loss: 1.4084 | Train Acc: 75.10% | Val Acc: 69.70%
[17/50] Train Loss: 1.3078 | Val Loss: 1.3968 | Train Acc: 76.00% | Val Acc: 70.10%
[18/50] Train Loss: 1.2967 | Val Loss: 1.3914 | Train Acc: 77.90% | Val Acc: 70.50%
[19/50] Train Loss: 1.2893 | Val Loss: 1.3879 | Train Acc: 77.00% | Val Acc: 70.60%
[20/50] Train Loss: 1.2759 | Val Loss: 1.3802 | Train Acc: 78.10% | Val Acc: 70.20%
[21/50] Train Loss: 1.2780 | Val Loss: 1.3711 | Train Acc: 77.20% | Val Acc: 70.50%
[22/50] Train Loss: 1.2689 | Val Loss: 1.3676 | Train Acc: 77.40% | Val Acc: 70.50%
[23/50] Train Loss: 1.2637 | Val Loss: 1.3649 | Train Acc: 77.80% | Val Acc: 70.40%
[24/50] Train Loss: 1.2378 | Val Loss: 1.3621 | Train Acc: 80.70% | Val Acc: 71.00%
[25/50] Train Loss: 1.2423 | Val Loss: 1.3520 | Train Acc: 78.70% | Val Acc: 71.80%
[26/50] Train Loss: 1.2339 | Val Loss: 1.3485 | Train Acc: 79.50% | Val Acc: 71.60%
[27/50] Train Loss: 1.2273 | Val Loss: 1.3524 | Train Acc: 79.80% | Val Acc: 70.40%
[28/50] Train Loss: 1.2168 | Val Loss: 1.3500 | Train Acc: 81.70% | Val Acc: 71.20%
[29/50] Train Loss: 1.2104 | Val Loss: 1.3471 | Train Acc: 80.90% | Val Acc: 71.20%
[30/50] Train Loss: 1.2171 | Val Loss: 1.3449 | Train Acc: 79.30% | Val Acc: 71.50%
[31/50] Train Loss: 1.2137 | Val Loss: 1.3424 | Train Acc: 79.70% | Val Acc: 71.20%
[32/50] Train Loss: 1.2130 | Val Loss: 1.3414 | Train Acc: 80.90% | Val Acc: 71.40%
[33/50] Train Loss: 1.2136 | Val Loss: 1.3421 | Train Acc: 80.10% | Val Acc: 71.30%
[34/50] Train Loss: 1.2141 | Val Loss: 1.3397 | Train Acc: 80.10% | Val Acc: 71.30%
[35/50] Train Loss: 1.2138 | Val Loss: 1.3384 | Train Acc: 81.40% | Val Acc: 71.20%
[36/50] Train Loss: 1.1868 | Val Loss: 1.3343 | Train Acc: 81.60% | Val Acc: 71.60%
[37/50] Train Loss: 1.1987 | Val Loss: 1.3371 | Train Acc: 80.10% | Val Acc: 71.10%
[38/50] Train Loss: 1.1935 | Val Loss: 1.3394 | Train Acc: 81.30% | Val Acc: 70.90%
[39/50] Train Loss: 1.1979 | Val Loss: 1.3374 | Train Acc: 80.60% | Val Acc: 71.50%
[40/50] Train Loss: 1.1783 | Val Loss: 1.3379 | Train Acc: 82.70% | Val Acc: 71.20%
[41/50] Train Loss: 1.1894 | Val Loss: 1.3372 | Train Acc: 81.70% | Val Acc: 71.20%
[42/50] Train Loss: 1.1849 | Val Loss: 1.3359 | Train Acc: 81.10% | Val Acc: 71.30%
[43/50] Train Loss: 1.2071 | Val Loss: 1.3361 | Train Acc: 80.70% | Val Acc: 71.30%
[44/50] Train Loss: 1.1973 | Val Loss: 1.3351 | Train Acc: 80.70% | Val Acc: 71.30%
[45/50] Train Loss: 1.1918 | Val Loss: 1.3352 | Train Acc: 81.60% | Val Acc: 71.20%
[46/50] Train Loss: 1.2008 | Val Loss: 1.3365 | Train Acc: 78.30% | Val Acc: 71.10%
[47/50] Train Loss: 1.1870 | Val Loss: 1.3371 | Train Acc: 82.60% | Val Acc: 71.10%
[48/50] Train Loss: 1.1956 | Val Loss: 1.3376 | Train Acc: 81.40% | Val Acc: 71.20%
[49/50] Train Loss: 1.1941 | Val Loss: 1.3360 | Train Acc: 80.50% | Val Acc: 71.50%
[50/50] Train Loss: 1.1868 | Val Loss: 1.3348 | Train Acc: 80.90% | Val Acc: 71.60%

Stage-1 Best results:
Train Loss: 1.2423 | Val Loss: 1.3520 | Train Acc: 78.70% | Val Acc: 71.80%
Stage-1 Training Time: 190.20 seconds

Loaded Stage-1 best-val model for Stage-2 fine-tuning


===== Stage-2: Unfreeze layer4 + fc =====
[01/50] Train Loss: 1.1782 | Val Loss: 1.2806 | Train Acc: 81.50% | Val Acc: 75.20%
[02/50] Train Loss: 1.0057 | Val Loss: 1.2432 | Train Acc: 90.30% | Val Acc: 77.50%
[03/50] Train Loss: 0.9084 | Val Loss: 1.2032 | Train Acc: 96.50% | Val Acc: 79.40%
[04/50] Train Loss: 0.8585 | Val Loss: 1.1689 | Train Acc: 98.00% | Val Acc: 81.50%
[05/50] Train Loss: 0.8178 | Val Loss: 1.1442 | Train Acc: 99.50% | Val Acc: 81.40%
[06/50] Train Loss: 0.8023 | Val Loss: 1.1325 | Train Acc: 99.70% | Val Acc: 81.80%
[07/50] Train Loss: 0.7780 | Val Loss: 1.1186 | Train Acc: 100.00% | Val Acc: 82.50%
[08/50] Train Loss: 0.7615 | Val Loss: 1.1147 | Train Acc: 100.00% | Val Acc: 82.90%
[09/50] Train Loss: 0.7604 | Val Loss: 1.1119 | Train Acc: 100.00% | Val Acc: 82.70%
[10/50] Train Loss: 0.7497 | Val Loss: 1.1024 | Train Acc: 99.80% | Val Acc: 83.60%
[11/50] Train Loss: 0.7456 | Val Loss: 1.0977 | Train Acc: 100.00% | Val Acc: 84.50%
[12/50] Train Loss: 0.7357 | Val Loss: 1.0980 | Train Acc: 100.00% | Val Acc: 85.00%
[13/50] Train Loss: 0.7328 | Val Loss: 1.1003 | Train Acc: 100.00% | Val Acc: 84.90%
[14/50] Train Loss: 0.7325 | Val Loss: 1.0936 | Train Acc: 100.00% | Val Acc: 84.20%
[15/50] Train Loss: 0.7293 | Val Loss: 1.0925 | Train Acc: 100.00% | Val Acc: 84.30%
[16/50] Train Loss: 0.7254 | Val Loss: 1.0919 | Train Acc: 100.00% | Val Acc: 84.60%
[17/50] Train Loss: 0.7243 | Val Loss: 1.0887 | Train Acc: 100.00% | Val Acc: 85.00%
[18/50] Train Loss: 0.7241 | Val Loss: 1.0841 | Train Acc: 100.00% | Val Acc: 85.00%
[19/50] Train Loss: 0.7226 | Val Loss: 1.0824 | Train Acc: 100.00% | Val Acc: 85.50%
[20/50] Train Loss: 0.7202 | Val Loss: 1.0832 | Train Acc: 100.00% | Val Acc: 85.00%
[21/50] Train Loss: 0.7189 | Val Loss: 1.0844 | Train Acc: 100.00% | Val Acc: 85.40%
[22/50] Train Loss: 0.7180 | Val Loss: 1.0837 | Train Acc: 100.00% | Val Acc: 85.30%
[23/50] Train Loss: 0.7198 | Val Loss: 1.0820 | Train Acc: 100.00% | Val Acc: 85.10%
[24/50] Train Loss: 0.7188 | Val Loss: 1.0747 | Train Acc: 100.00% | Val Acc: 85.50%
[25/50] Train Loss: 0.7174 | Val Loss: 1.0714 | Train Acc: 100.00% | Val Acc: 85.70%
[26/50] Train Loss: 0.7169 | Val Loss: 1.0751 | Train Acc: 100.00% | Val Acc: 85.90%
[27/50] Train Loss: 0.7158 | Val Loss: 1.0762 | Train Acc: 100.00% | Val Acc: 85.90%
[28/50] Train Loss: 0.7157 | Val Loss: 1.0778 | Train Acc: 100.00% | Val Acc: 85.80%
[29/50] Train Loss: 0.7149 | Val Loss: 1.0783 | Train Acc: 100.00% | Val Acc: 86.00%
[30/50] Train Loss: 0.7149 | Val Loss: 1.0788 | Train Acc: 100.00% | Val Acc: 85.90%
[31/50] Train Loss: 0.7136 | Val Loss: 1.0769 | Train Acc: 100.00% | Val Acc: 85.70%
[32/50] Train Loss: 0.7137 | Val Loss: 1.0752 | Train Acc: 100.00% | Val Acc: 85.80%
[33/50] Train Loss: 0.7139 | Val Loss: 1.0762 | Train Acc: 100.00% | Val Acc: 85.90%
[34/50] Train Loss: 0.7128 | Val Loss: 1.0761 | Train Acc: 100.00% | Val Acc: 86.00%
[35/50] Train Loss: 0.7122 | Val Loss: 1.0775 | Train Acc: 100.00% | Val Acc: 86.20%
[36/50] Train Loss: 0.7132 | Val Loss: 1.0761 | Train Acc: 100.00% | Val Acc: 85.80%
[37/50] Train Loss: 0.7124 | Val Loss: 1.0749 | Train Acc: 100.00% | Val Acc: 86.00%
[38/50] Train Loss: 0.7115 | Val Loss: 1.0738 | Train Acc: 100.00% | Val Acc: 85.90%
[39/50] Train Loss: 0.7125 | Val Loss: 1.0743 | Train Acc: 100.00% | Val Acc: 86.10%
[40/50] Train Loss: 0.7108 | Val Loss: 1.0739 | Train Acc: 100.00% | Val Acc: 86.00%
[41/50] Train Loss: 0.7108 | Val Loss: 1.0739 | Train Acc: 100.00% | Val Acc: 86.10%
[42/50] Train Loss: 0.7109 | Val Loss: 1.0727 | Train Acc: 100.00% | Val Acc: 86.20%
[43/50] Train Loss: 0.7111 | Val Loss: 1.0736 | Train Acc: 100.00% | Val Acc: 85.90%
[44/50] Train Loss: 0.7114 | Val Loss: 1.0743 | Train Acc: 100.00% | Val Acc: 86.30%
[45/50] Train Loss: 0.7099 | Val Loss: 1.0735 | Train Acc: 100.00% | Val Acc: 86.10%
[46/50] Train Loss: 0.7096 | Val Loss: 1.0728 | Train Acc: 100.00% | Val Acc: 86.20%
[47/50] Train Loss: 0.7095 | Val Loss: 1.0730 | Train Acc: 100.00% | Val Acc: 85.80%
[48/50] Train Loss: 0.7108 | Val Loss: 1.0733 | Train Acc: 100.00% | Val Acc: 86.30%
[49/50] Train Loss: 0.7122 | Val Loss: 1.0732 | Train Acc: 100.00% | Val Acc: 86.20%
[50/50] Train Loss: 0.7096 | Val Loss: 1.0734 | Train Acc: 100.00% | Val Acc: 86.30%

Stage-2 Best results:
Train Loss: 0.7114 | Val Loss: 1.0743 | Train Acc: 100.00% | Val Acc: 86.30%
Stage-2 Training Time: 250.83 seconds

Loaded Stage-2 best-val model for Stage-3 fine-tuning


===== Stage-3: Unfreeze layer3 + layer4 + fc =====
[01/50] Train Loss: 0.7131 | Val Loss: 1.0790 | Train Acc: 100.00% | Val Acc: 86.20%
[02/50] Train Loss: 0.7116 | Val Loss: 1.0775 | Train Acc: 100.00% | Val Acc: 86.20%
[03/50] Train Loss: 0.7097 | Val Loss: 1.0794 | Train Acc: 100.00% | Val Acc: 86.00%
[04/50] Train Loss: 0.7102 | Val Loss: 1.0778 | Train Acc: 100.00% | Val Acc: 86.00%
[05/50] Train Loss: 0.7115 | Val Loss: 1.0768 | Train Acc: 100.00% | Val Acc: 85.90%
[06/50] Train Loss: 0.7091 | Val Loss: 1.0770 | Train Acc: 100.00% | Val Acc: 86.10%
[07/50] Train Loss: 0.7092 | Val Loss: 1.0771 | Train Acc: 100.00% | Val Acc: 86.70%
[08/50] Train Loss: 0.7092 | Val Loss: 1.0766 | Train Acc: 100.00% | Val Acc: 86.70%
[09/50] Train Loss: 0.7104 | Val Loss: 1.0757 | Train Acc: 100.00% | Val Acc: 86.50%
[10/50] Train Loss: 0.7085 | Val Loss: 1.0772 | Train Acc: 100.00% | Val Acc: 86.50%
[11/50] Train Loss: 0.7103 | Val Loss: 1.0759 | Train Acc: 100.00% | Val Acc: 86.30%
[12/50] Train Loss: 0.7081 | Val Loss: 1.0753 | Train Acc: 100.00% | Val Acc: 86.30%
[13/50] Train Loss: 0.7090 | Val Loss: 1.0762 | Train Acc: 100.00% | Val Acc: 86.10%
[14/50] Train Loss: 0.7086 | Val Loss: 1.0767 | Train Acc: 100.00% | Val Acc: 86.60%
[15/50] Train Loss: 0.7084 | Val Loss: 1.0758 | Train Acc: 100.00% | Val Acc: 86.80%
[16/50] Train Loss: 0.7076 | Val Loss: 1.0747 | Train Acc: 100.00% | Val Acc: 86.20%
[17/50] Train Loss: 0.7075 | Val Loss: 1.0740 | Train Acc: 100.00% | Val Acc: 86.30%
[18/50] Train Loss: 0.7084 | Val Loss: 1.0743 | Train Acc: 100.00% | Val Acc: 86.20%
[19/50] Train Loss: 0.7082 | Val Loss: 1.0740 | Train Acc: 100.00% | Val Acc: 86.10%
[20/50] Train Loss: 0.7083 | Val Loss: 1.0740 | Train Acc: 100.00% | Val Acc: 86.20%
[21/50] Train Loss: 0.7076 | Val Loss: 1.0745 | Train Acc: 100.00% | Val Acc: 86.20%
[22/50] Train Loss: 0.7071 | Val Loss: 1.0762 | Train Acc: 100.00% | Val Acc: 86.00%
[23/50] Train Loss: 0.7086 | Val Loss: 1.0762 | Train Acc: 100.00% | Val Acc: 86.10%
[24/50] Train Loss: 0.7075 | Val Loss: 1.0760 | Train Acc: 100.00% | Val Acc: 86.30%
[25/50] Train Loss: 0.7076 | Val Loss: 1.0764 | Train Acc: 100.00% | Val Acc: 86.30%
[26/50] Train Loss: 0.7078 | Val Loss: 1.0766 | Train Acc: 100.00% | Val Acc: 86.60%
[27/50] Train Loss: 0.7078 | Val Loss: 1.0751 | Train Acc: 100.00% | Val Acc: 86.40%
[28/50] Train Loss: 0.7079 | Val Loss: 1.0739 | Train Acc: 100.00% | Val Acc: 86.00%
[29/50] Train Loss: 0.7078 | Val Loss: 1.0735 | Train Acc: 100.00% | Val Acc: 86.00%
[30/50] Train Loss: 0.7086 | Val Loss: 1.0736 | Train Acc: 100.00% | Val Acc: 86.30%
[31/50] Train Loss: 0.7070 | Val Loss: 1.0745 | Train Acc: 100.00% | Val Acc: 86.40%
[32/50] Train Loss: 0.7061 | Val Loss: 1.0736 | Train Acc: 100.00% | Val Acc: 86.40%
[33/50] Train Loss: 0.7063 | Val Loss: 1.0743 | Train Acc: 100.00% | Val Acc: 86.10%
[34/50] Train Loss: 0.7072 | Val Loss: 1.0752 | Train Acc: 100.00% | Val Acc: 86.10%
[35/50] Train Loss: 0.7082 | Val Loss: 1.0763 | Train Acc: 100.00% | Val Acc: 86.00%
[36/50] Train Loss: 0.7077 | Val Loss: 1.0759 | Train Acc: 100.00% | Val Acc: 86.00%
[37/50] Train Loss: 0.7080 | Val Loss: 1.0758 | Train Acc: 100.00% | Val Acc: 86.00%
[38/50] Train Loss: 0.7067 | Val Loss: 1.0742 | Train Acc: 100.00% | Val Acc: 86.00%
[39/50] Train Loss: 0.7069 | Val Loss: 1.0742 | Train Acc: 100.00% | Val Acc: 86.10%
[40/50] Train Loss: 0.7068 | Val Loss: 1.0736 | Train Acc: 100.00% | Val Acc: 86.30%
[41/50] Train Loss: 0.7062 | Val Loss: 1.0743 | Train Acc: 100.00% | Val Acc: 86.10%
[42/50] Train Loss: 0.7062 | Val Loss: 1.0747 | Train Acc: 100.00% | Val Acc: 86.00%
[43/50] Train Loss: 0.7063 | Val Loss: 1.0757 | Train Acc: 100.00% | Val Acc: 86.00%
[44/50] Train Loss: 0.7088 | Val Loss: 1.0754 | Train Acc: 100.00% | Val Acc: 86.20%
[45/50] Train Loss: 0.7059 | Val Loss: 1.0751 | Train Acc: 100.00% | Val Acc: 86.20%
[46/50] Train Loss: 0.7069 | Val Loss: 1.0762 | Train Acc: 100.00% | Val Acc: 85.90%
[47/50] Train Loss: 0.7058 | Val Loss: 1.0755 | Train Acc: 100.00% | Val Acc: 86.20%
[48/50] Train Loss: 0.7064 | Val Loss: 1.0758 | Train Acc: 100.00% | Val Acc: 86.20%
[49/50] Train Loss: 0.7061 | Val Loss: 1.0756 | Train Acc: 100.00% | Val Acc: 86.10%
[50/50] Train Loss: 0.7058 | Val Loss: 1.0753 | Train Acc: 100.00% | Val Acc: 86.10%

Stage-3 Best results:
Train Loss: 0.7084 | Val Loss: 1.0758 | Train Acc: 100.00% | Val Acc: 86.80%
Stage-3 Training Time: 250.40 seconds

Any comments? Feel free to participate below in the Facebook comment section.