Hugging Face's logo

AbstractPhil
/
geo-beatrix

Image Classification
PyTorch
TensorBoard
Safetensors
cifar100
geometric-learning
fractal-encoding
in-training
no-attention
no-cross-entropy
Eval Results (legacy)
Model card Files
xet
Metrics Community
AbstractPhil commited on
Commit
4e77ca7
·
verified ·
1 Parent(s): 2f7f6b9

Epoch 150: 66.19%

Browse files
Files changed (1) hide show
  1. README.md +15 -13
README.md CHANGED
@@ -36,7 +36,7 @@ model-index:
36
 
37
  🚧 **Training in Progress** 🚧
38
 
39
- Current Status: Epoch 100/200
40
 
41
  ---
42
 
@@ -46,17 +46,17 @@ Current Status: Epoch 100/200
46
  |--------|-------|
47
  | **Best Test Accuracy** | **67.43%** |
48
  | **Best Epoch** | 195 |
49
- | **Current Train Accuracy** | 73.35% |
50
- | **Current Test Accuracy** | 63.97% |
51
- | **Current α (Cantor param)** | 0.4382 |
52
  | **Total Parameters** | 46,338,737 |
53
- | **Training Time** | 0:27:25 |
54
 
55
  ### All Training Runs
56
 
57
  | Timestamp | Status | Best Epoch | Test Acc | Train Acc | α |
58
  |-----------|--------|------------|----------|-----------|---|
59
- | `20251010_025042` | 🔄 | 195 | **67.43%** | 73.35% | 0.4382 |
60
 
61
  ### Comparison to State-of-the-Art
62
 
@@ -189,23 +189,25 @@ epoch_checkpoint = hf_hub_download(
189
 
190
  ### Best Checkpoint
191
  - Epoch: 195
192
- - Train Acc: 73.35%
193
  - Test Acc: 67.43%
194
- - Alpha: 0.4382
195
  - Loss: 0.0000
196
 
197
  ### Latest 5 Epochs
198
 
199
- - **Epoch 96**: Train 74.17%, Test 0.00%, α=0.4369, Loss=0.7639
200
- - **Epoch 97**: Train 73.40%, Test 0.00%, α=0.4373, Loss=0.7384
201
- - **Epoch 98**: Train 75.51%, Test 0.00%, α=0.4369, Loss=0.7569
202
- - **Epoch 99**: Train 73.46%, Test 0.00%, α=0.4374, Loss=0.7436
203
- - **Epoch 100**: Train 73.35%, Test 63.97%, α=0.4382, Loss=0.7377
204
 
205
  ### Training Milestones
206
  - 🎯 **50% Accuracy** reached at epoch 35
207
  - 🎯 **60% Accuracy** reached at epoch 65
 
208
  - 📊 **α ≥ 0.40** reached at epoch 10
 
209
 
210
  ---
211
 
 
36
 
37
  🚧 **Training in Progress** 🚧
38
 
39
+ Current Status: Epoch 150/200
40
 
41
  ---
42
 
 
46
  |--------|-------|
47
  | **Best Test Accuracy** | **67.43%** |
48
  | **Best Epoch** | 195 |
49
+ | **Current Train Accuracy** | 73.96% |
50
+ | **Current Test Accuracy** | 66.19% |
51
+ | **Current α (Cantor param)** | 0.4444 |
52
  | **Total Parameters** | 46,338,737 |
53
+ | **Training Time** | 0:41:17 |
54
 
55
  ### All Training Runs
56
 
57
  | Timestamp | Status | Best Epoch | Test Acc | Train Acc | α |
58
  |-----------|--------|------------|----------|-----------|---|
59
+ | `20251010_025042` | 🔄 | 195 | **67.43%** | 73.96% | 0.4444 |
60
 
61
  ### Comparison to State-of-the-Art
62
 
 
189
 
190
  ### Best Checkpoint
191
  - Epoch: 195
192
+ - Train Acc: 73.96%
193
  - Test Acc: 67.43%
194
+ - Alpha: 0.4444
195
  - Loss: 0.0000
196
 
197
  ### Latest 5 Epochs
198
 
199
+ - **Epoch 146**: Train 69.94%, Test 0.00%, α=0.4428, Loss=0.5624
200
+ - **Epoch 147**: Train 74.67%, Test 0.00%, α=0.4433, Loss=0.5763
201
+ - **Epoch 148**: Train 71.69%, Test 0.00%, α=0.4430, Loss=0.5626
202
+ - **Epoch 149**: Train 75.45%, Test 0.00%, α=0.4426, Loss=0.5743
203
+ - **Epoch 150**: Train 73.96%, Test 66.19%, α=0.4444, Loss=0.5696
204
 
205
  ### Training Milestones
206
  - 🎯 **50% Accuracy** reached at epoch 35
207
  - 🎯 **60% Accuracy** reached at epoch 65
208
+ - 🏆 **Beat vit-beatrix (66.0%)** at epoch 130
209
  - 📊 **α ≥ 0.40** reached at epoch 10
210
+ - 📊 **α ≥ 0.44** (near triadic equilibrium) at epoch 105
211
 
212
  ---
213