Update README.md
Browse files
README.md
CHANGED
|
@@ -13,6 +13,30 @@ source: "AutoTrain"
|
|
| 13 |
training_type: "fine-tuning"
|
| 14 |
geographical_location: "US"
|
| 15 |
hardware_used: "A100"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 16 |
---
|
| 17 |
# Fine-tuned TrOCR Model for Portuguese
|
| 18 |
|
|
|
|
| 13 |
training_type: "fine-tuning"
|
| 14 |
geographical_location: "US"
|
| 15 |
hardware_used: "A100"
|
| 16 |
+
tags:
|
| 17 |
+
- ocr
|
| 18 |
+
- optical-character-recognition
|
| 19 |
+
- text-recognition
|
| 20 |
+
- trocr
|
| 21 |
+
- vision
|
| 22 |
+
- vision-and-language
|
| 23 |
+
model-index:
|
| 24 |
+
- name: trocr-finetuned-portugues
|
| 25 |
+
results:
|
| 26 |
+
- task:
|
| 27 |
+
type: optical-character-recognition
|
| 28 |
+
name: Optical Character Recognition
|
| 29 |
+
dataset:
|
| 30 |
+
type: mazafard/portugues_ocr_dataset_full
|
| 31 |
+
name: portugues_ocr_dataset_full
|
| 32 |
+
args: default
|
| 33 |
+
metrics:
|
| 34 |
+
- type: cer
|
| 35 |
+
value: 0.01
|
| 36 |
+
name: Character Error Rate
|
| 37 |
+
- type: wer
|
| 38 |
+
value: 0.05
|
| 39 |
+
name: Word Error Rate
|
| 40 |
---
|
| 41 |
# Fine-tuned TrOCR Model for Portuguese
|
| 42 |
|