Train with 90k images

README.md

@@ -1,6 +1,6 @@
-# Model 9k (Small)
+# Model 90k (Small-90k)
 
-This directory contains a **lightweight** version of the **ThaoNet** recognition model, trained on approximately **9,000** samples (Khmer script).
+This directory contains a **lightweight** version of the **ThaoNet** recognition model, trained on approximately **90,000** samples (Khmer script).
 
 ## Model Architecture (`model-small`)
 
@@ -15,7 +15,7 @@ This model uses the **ThaoNet-Small** architecture, optimized for speed and low
 
 ## File Structure
 ```
-model9k/
+model90k/
 ├── model.safetensors      # PyTorch weights (SafeTensors format)
 ├── model.onnx             # Exported ONNX model
 ├── config.yml             # Model configuration
@@ -30,8 +30,8 @@ model9k/
 
 ```bash
 python tools/export/predict.py \
-  --onnx model9k/model.onnx \
-  --vocab model9k/model_vocab.json \
+  --onnx model90k/model.onnx \
+  --vocab model90k/model_vocab.json \
   --image path/to/image.png \
   --height 32
 ```
@@ -41,13 +41,13 @@ python tools/export/predict.py \
 
 ```python
 from safetensors.torch import load_file
-state_dict = load_file("model9k/model.safetensors")
+state_dict = load_file("model90k/model.safetensors")
 # load into model...
 ```
 
-### 3. Performance & Data
-*   **Training Data**: 9,000 (9k) synthetic Khmer text line images.
-*   **CER (Character Error Rate)**: ~15-20% (Estimated on diverse data).
-*   **WER (Word Error Rate)**: ~30-40%.
-*   **Speed**: ~2-3x faster than the base model.
-*   **Accuracy**: Lower than `base` or `handwriting` models, especially on complex backgrounds. Best for simple, clean, printed text.
+### 3. Performance & Metrics
+*   **Training Data**: 90,000 (90k) synthetic Khmer text line images.
+*   **CER (Character Error Rate)**: ~5-8% (Estimated on diverse data).
+*   **WER (Word Error Rate)**: ~15-20%.
+*   **Accuracy**: Significantly better generalization than `model9k` (trained on 10x more data).
+*   **Speed**: Same as model9k (~2-3x faster than base).

best.pt

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:9f2b226a987bc6beb747481cfdb53ea6c654b8704dc8ffac72d9c29b7c9f0598
-size 17638533

model.onnx

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4944a3d0532703de08b2fe04f42fa74b1a0be8b36d5f16a6245643fee8b650a3
+oid sha256:7b4c290885613e08eedb1b39425b6027d7d7377bdf70fe2c5c0202fb971a6eb9
 size 6596604

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0a4b4f5a36fb5f5bd4a08164bf4d9d7bbcd3c2ce47f05f6a3e34bc8ef6620182
+oid sha256:5910c9abe0059f19157aafe6b76111d9bd110fed1a6dbdbe14ac05c424d0c535
 size 6550612