Add training info
Browse files
README.md
CHANGED
|
@@ -28,9 +28,52 @@ You can then run inference with this model like so:
|
|
| 28 |
from span_marker import SpanMarkerModel
|
| 29 |
|
| 30 |
# Download from the 🤗 Hub
|
| 31 |
-
model = SpanMarkerModel.from_pretrained("
|
| 32 |
# Run inference
|
| 33 |
entities = model.predict("Amelia Earhart flew her single engine Lockheed Vega 5B across the Atlantic to Paris.")
|
| 34 |
```
|
| 35 |
|
| 36 |
-
See the [SpanMarker](https://github.com/tomaarsen/SpanMarkerNER) repository for documentation and additional information on this library.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
from span_marker import SpanMarkerModel
|
| 29 |
|
| 30 |
# Download from the 🤗 Hub
|
| 31 |
+
model = SpanMarkerModel.from_pretrained("tomaarsen/span-marker-xlm-roberta-large-verbs")
|
| 32 |
# Run inference
|
| 33 |
entities = model.predict("Amelia Earhart flew her single engine Lockheed Vega 5B across the Atlantic to Paris.")
|
| 34 |
```
|
| 35 |
|
| 36 |
+
See the [SpanMarker](https://github.com/tomaarsen/SpanMarkerNER) repository for documentation and additional information on this library.
|
| 37 |
+
|
| 38 |
+
### Performance
|
| 39 |
+
|
| 40 |
+
It achieves the following results on the evaluation set:
|
| 41 |
+
- Loss: 0.0152
|
| 42 |
+
- Overall Precision: 0.9845
|
| 43 |
+
- Overall Recall: 0.9849
|
| 44 |
+
- Overall F1: 0.9847
|
| 45 |
+
- Overall Accuracy: 0.9962
|
| 46 |
+
|
| 47 |
+
## Training procedure
|
| 48 |
+
|
| 49 |
+
### Training hyperparameters
|
| 50 |
+
|
| 51 |
+
The following hyperparameters were used during training:
|
| 52 |
+
- learning_rate: 1e-05
|
| 53 |
+
- train_batch_size: 4
|
| 54 |
+
- eval_batch_size: 4
|
| 55 |
+
- seed: 42
|
| 56 |
+
- gradient_accumulation_steps: 2
|
| 57 |
+
- total_train_batch_size: 8
|
| 58 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 59 |
+
- lr_scheduler_type: linear
|
| 60 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 61 |
+
- num_epochs: 3
|
| 62 |
+
|
| 63 |
+
### Training results
|
| 64 |
+
|
| 65 |
+
| Training Loss | Epoch | Step | Validation Loss | Overall Precision | Overall Recall | Overall F1 | Overall Accuracy |
|
| 66 |
+
|:-------------:|:-----:|:----:|:---------------:|:-----------------:|:--------------:|:----------:|:----------------:|
|
| 67 |
+
| 0.036 | 0.61 | 1000 | 0.0151 | 0.9911 | 0.9733 | 0.9821 | 0.9956 |
|
| 68 |
+
| 0.0126 | 1.22 | 2000 | 0.0131 | 0.9856 | 0.9864 | 0.9860 | 0.9965 |
|
| 69 |
+
| 0.0175 | 1.83 | 3000 | 0.0154 | 0.9735 | 0.9894 | 0.9814 | 0.9953 |
|
| 70 |
+
| 0.0115 | 2.45 | 4000 | 0.0172 | 0.9821 | 0.9871 | 0.9845 | 0.9962 |
|
| 71 |
+
|
| 72 |
+
|
| 73 |
+
### Framework versions
|
| 74 |
+
|
| 75 |
+
- Transformers 4.30.2
|
| 76 |
+
- Pytorch 2.0.1+cu118
|
| 77 |
+
- Datasets 2.13.1
|
| 78 |
+
- Tokenizers 0.13.3
|
| 79 |
+
- SpanMarker 1.2.3
|