Update README.md

README.md

@@ -1,3 +1,56 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+---
+
+github　repo:　[github.com/ximeiorg/ochw](https://github.com/ximeiorg/ochw)
+
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/65090e91f6a2bf0be7b61f22/JE6RBeidskDuD5Zgj-r9a.png)
+python  推理代码：
+
+```python
+def get_labels():
+    labels = []
+    with open("data/label.txt", "r", encoding="utf-8") as f:
+        for line in f:
+            # line: !	0
+            line = line.strip()
+            label = line.split("\t")[0]
+            labels.append(label)
+    return labels
+
+if __name__ == "__main__":
+    model = HandwritingTrainer.load_from_checkpoint("checkpoint-epoch=32-val_loss=0.156.ckpt")
+    model.eval()
+    model = model.to("cuda")
+    img = Image.open("./testdata/hui.png")
+    img = img.convert("RGB")
+    img = img.resize((96,96))
+    rans = transforms.Compose([
+        transforms.Resize((96, 96)),
+        transforms.ToTensor(),
+        transforms.Normalize(mean=[0.95], std=[0.2])
+    ])
+    img = trans(img)
+    img = img.unsqueeze(0)
+    img = img.to("cuda")
+    labels = get_labels()
+    with torch.no_grad():
+        output = model(img)
+        output = torch.nn.functional.softmax(output,dim=1)
+        # 获取top5的预测结果
+        top5_prob, top5_idx = torch.topk(output, 5)
+        top5_prob = top5_prob.cpu().numpy()
+        top5_idx = top5_idx.cpu().numpy()
+        for i in range(5):
+            idx = top5_idx[0][i]
+            print(f"Top {i+1} 预测标签: {labels[idx]}, 概率: {top5_prob[0][i]:.4f}")
+        
+```
+
+得到的结果如下：
+```
+Top 1 预测标签: 知, 概率: 0.9505
+Top 2 预测标签: 勉, 概率: 0.0095
+Top 3 预测标签: 贮, 概率: 0.0025
+Top 4 预测标签: 处, 概率: 0.0025
+Top 5 预测标签: ‰, 概率: 0.0025