Update README.md
Browse files
README.md
CHANGED
|
@@ -60,7 +60,7 @@ Then you can load this model and run inference.
|
|
| 60 |
from setfit import SetFitModel
|
| 61 |
|
| 62 |
# Download from the 🤗 Hub
|
| 63 |
-
model = SetFitModel.from_pretrained("
|
| 64 |
# Run inference
|
| 65 |
preds = model("table. | PlotAccessor")
|
| 66 |
```
|
|
@@ -89,41 +89,6 @@ preds = model("table. | PlotAccessor")
|
|
| 89 |
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
| 90 |
-->
|
| 91 |
|
| 92 |
-
## Training Details
|
| 93 |
-
|
| 94 |
-
### Training Set Metrics
|
| 95 |
-
| Training set | Min | Median | Max |
|
| 96 |
-
|:-------------|:----|:-------|:----|
|
| 97 |
-
| Word count | 3 | 8.9868 | 28 |
|
| 98 |
-
|
| 99 |
-
### Training Hyperparameters
|
| 100 |
-
- batch_size: (32, 32)
|
| 101 |
-
- num_epochs: (10, 10)
|
| 102 |
-
- max_steps: -1
|
| 103 |
-
- sampling_strategy: oversampling
|
| 104 |
-
- num_iterations: 20
|
| 105 |
-
- body_learning_rate: (2e-05, 1e-05)
|
| 106 |
-
- head_learning_rate: 0.01
|
| 107 |
-
- loss: CosineSimilarityLoss
|
| 108 |
-
- distance_metric: cosine_distance
|
| 109 |
-
- margin: 0.25
|
| 110 |
-
- end_to_end: False
|
| 111 |
-
- use_amp: False
|
| 112 |
-
- warmup_proportion: 0.1
|
| 113 |
-
- l2_weight: 0.01
|
| 114 |
-
- seed: 42
|
| 115 |
-
- eval_max_steps: -1
|
| 116 |
-
- load_best_model_at_end: False
|
| 117 |
-
|
| 118 |
-
### Framework Versions
|
| 119 |
-
- Python: 3.10.8
|
| 120 |
-
- SetFit: 1.1.2
|
| 121 |
-
- Sentence Transformers: 5.0.0
|
| 122 |
-
- Transformers: 4.54.1
|
| 123 |
-
- PyTorch: 2.7.1+cu126
|
| 124 |
-
- Datasets: 3.6.0
|
| 125 |
-
- Tokenizers: 0.21.4
|
| 126 |
-
|
| 127 |
## Citation
|
| 128 |
|
| 129 |
### BibTeX
|
|
|
|
| 60 |
from setfit import SetFitModel
|
| 61 |
|
| 62 |
# Download from the 🤗 Hub
|
| 63 |
+
model = SetFitModel.from_pretrained("NLBSE/nlbse26_python")
|
| 64 |
# Run inference
|
| 65 |
preds = model("table. | PlotAccessor")
|
| 66 |
```
|
|
|
|
| 89 |
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
| 90 |
-->
|
| 91 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 92 |
## Citation
|
| 93 |
|
| 94 |
### BibTeX
|