docs(mkdocs): 更新版本号0.1.0->0.2.0; 更新训练日志

Author: zjZSTU

docs/index.md

@@ -1,6 +1,6 @@
 # YOLO_v1
 
-实现`YOLO_v1`算法
+实现`YOLO_v1`目标检测算法
 
 ## 实现流程
 
@@ -10,7 +10,7 @@
 4. 训练模型
 5. 计算`mAP`
 
-`50`轮训练完成后能够实现`99.01%`的`mAP`
+`50`轮训练完成后能够实现`97.31%`的`mAP`
 
 ## 相关链接
 

docs/log.md

@@ -8,259 +8,254 @@
 * 优化器：`SGD`，学习率`1e-3`，动量大小`0.9`
 * 衰减器：每隔`4`轮衰减`4%`，学习因子`0.96`
 
+## 检测结果
+
+```
+compute mAP
+{'cucumber': 63, 'mushroom': 61, 'eggplant': 62}
+98.16% = cucumber AP 
+93.77% = eggplant AP 
+100.00% = mushroom AP 
+mAP = 97.31%
+```
+
 ## 训练日志
 
 ```
 $ python train.py 
+24
 Epoch 0/49
 ----------
-train Loss: 4.3550
+train Loss: 4.6631
 save model
 
 Epoch 1/49
 ----------
-train Loss: 3.4803
+train Loss: 3.9457
 save model
 
 Epoch 2/49
 ----------
-train Loss: 3.3921
+train Loss: 3.5757
 save model
 
 Epoch 3/49
 ----------
-train Loss: 3.0650
+train Loss: 3.4125
 save model
 
 Epoch 4/49
 ----------
-train Loss: 2.9081
+train Loss: 3.1608
 save model
 
 Epoch 5/49
 ----------
-train Loss: 2.5893
-save model
+train Loss: 3.2524
 
 Epoch 6/49
 ----------
-train Loss: 2.5640
+train Loss: 2.8278
 save model
 
 Epoch 7/49
 ----------
-train Loss: 2.4905
+train Loss: 2.7577
 save model
 
 Epoch 8/49
 ----------
-train Loss: 2.1913
+train Loss: 2.6739
 save model
 
 Epoch 9/49
 ----------
-train Loss: 2.1152
+train Loss: 2.4874
 save model
 
 Epoch 10/49
 ----------
-train Loss: 1.9428
-save model
+train Loss: 2.5400
 
 Epoch 11/49
 ----------
-train Loss: 1.7271
-save model
+train Loss: 2.7458
 
 Epoch 12/49
 ----------
-train Loss: 1.4372
-save model
+train Loss: 2.5519
 
 Epoch 13/49
 ----------
-train Loss: 1.8185
+train Loss: 2.5486
 
 Epoch 14/49
 ----------
-train Loss: 1.5460
+train Loss: 2.4983
 
 Epoch 15/49
 ----------
-train Loss: 1.1692
+train Loss: 2.4122
 save model
 
 Epoch 16/49
 ----------
-train Loss: 1.0540
+train Loss: 2.3499
 save model
 
 Epoch 17/49
 ----------
-train Loss: 0.9028
+train Loss: 2.3360
 save model
 
 Epoch 18/49
 ----------
-train Loss: 0.7702
+train Loss: 2.1346
 save model
 
 Epoch 19/49
 ----------
-train Loss: 0.7176
+train Loss: 1.6496
 save model
 
 Epoch 20/49
 ----------
-train Loss: 0.7485
+train Loss: 1.4959
+save model
 
 Epoch 21/49
 ----------
-train Loss: 0.6307
+train Loss: 1.2388
 save model
 
 Epoch 22/49
 ----------
-train Loss: 0.5581
+train Loss: 0.9731
 save model
 
 Epoch 23/49
 ----------
-train Loss: 0.5320
+train Loss: 0.8746
 save model
 
 Epoch 24/49
 ----------
-train Loss: 0.5893
+train Loss: 0.8926
 
 Epoch 25/49
 ----------
-train Loss: 0.5185
+train Loss: 0.7697
 save model
 
 Epoch 26/49
 ----------
-train Loss: 0.6156
+train Loss: 0.7731
 
 Epoch 27/49
 ----------
-train Loss: 0.5096
+train Loss: 0.6818
 save model
 
 Epoch 28/49
 ----------
-train Loss: 0.5403
+train Loss: 0.6873
 
 Epoch 29/49
 ----------
-train Loss: 0.4653
+train Loss: 0.6238
 save model
 
 Epoch 30/49
 ----------
-train Loss: 0.3850
+train Loss: 0.5284
 save model
 
 Epoch 31/49
 ----------
-train Loss: 0.3609
+train Loss: 0.4824
 save model
 
 Epoch 32/49
 ----------
-train Loss: 0.4063
+train Loss: 0.4355
+save model
 
 Epoch 33/49
 ----------
-train Loss: 0.3349
+train Loss: 0.4300
 save model
 
 Epoch 34/49
 ----------
-train Loss: 0.2629
+train Loss: 0.4019
 save model
 
 Epoch 35/49
 ----------
-train Loss: 0.3319
+train Loss: 0.3622
+save model
 
 Epoch 36/49
 ----------
-train Loss: 0.2790
+train Loss: 0.4424
 
 Epoch 37/49
 ----------
-train Loss: 0.2487
+train Loss: 0.3394
 save model
 
 Epoch 38/49
 ----------
-train Loss: 0.2325
+train Loss: 0.3256
 save model
 
 Epoch 39/49
 ----------
-train Loss: 0.2146
+train Loss: 0.2458
 save model
 
 Epoch 40/49
 ----------
-train Loss: 0.2087
-save model
+train Loss: 0.2592
 
 Epoch 41/49
 ----------
-train Loss: 0.1626
-save model
+train Loss: 0.2518
 
 Epoch 42/49
 ----------
-train Loss: 0.1446
+train Loss: 0.2172
 save model
 
 Epoch 43/49
 ----------
-train Loss: 0.1372
-save model
+train Loss: 0.2442
 
 Epoch 44/49
 ----------
-train Loss: 0.1260
+train Loss: 0.1925
 save model
 
 Epoch 45/49
 ----------
-train Loss: 0.1231
+train Loss: 0.1607
 save model
 
 Epoch 46/49
 ----------
-train Loss: 0.1232
+train Loss: 0.1828
 
 Epoch 47/49
 ----------
-train Loss: 0.1475
+train Loss: 0.1770
 
 Epoch 48/49
 ----------
-train Loss: 0.1226
-save model
+train Loss: 0.1690
 
 Epoch 49/49
 ----------
-train Loss: 0.1000
-save model
+train Loss: 0.1730
 
-Training complete in 6m 28s
-```
-
-## 检测结果
-
-```
-compute mAP
-{'cucumber': 63, 'mushroom': 61, 'eggplant': 62}
-99.43% = cucumber AP 
-98.39% = eggplant AP 
-99.22% = mushroom AP 
-mAP = 99.01%
+Training complete in 21m 39s
 ```

mkdocs.yml

@@ -13,7 +13,7 @@ site_dir: 'site'
 # 额外信息
 extra:
     # 版本号
-    version: 0.1.0
+    version: 0.2.0
 # 主题
 theme: 
     # name: 'readthedocs'