Instructions to use dandankim/distilbert-token-classifier with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use dandankim/distilbert-token-classifier with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("token-classification", model="dandankim/distilbert-token-classifier")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForTokenClassification tokenizer = AutoTokenizer.from_pretrained("dandankim/distilbert-token-classifier") model = AutoModelForTokenClassification.from_pretrained("dandankim/distilbert-token-classifier", device_map="auto") - Notebooks
- Google Colab
- Kaggle
|
Download README.md from dandankim/distilbert-token-classifier: direct link, hf CLI and curl.
- Browser
- Download file 2.43 kB
-
https://huggingface.co/dandankim/distilbert-token-classifier/resolve/main/README.md
- Command line
-
hf download hf://dandankim/distilbert-token-classifier/README.md
-
curl -L -o README.md https://huggingface.co/dandankim/distilbert-token-classifier/resolve/main/README.md
2.43 kB
metadata
library_name: transformers
license: mit
base_model: FacebookAI/roberta-base
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: distilbert-token-classifier
results: []
distilbert-token-classifier
This model is a fine-tuned version of FacebookAI/roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0728
- Precision: 0.9694
- Recall: 0.9767
- F1: 0.9730
- Accuracy: 0.9846
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 20
- eval_batch_size: 20
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.2
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 1.0798 | 1.0 | 119 | 0.5221 | 0.5989 | 0.3881 | 0.4710 | 0.8427 |
| 0.2561 | 2.0 | 238 | 0.1148 | 0.9162 | 0.9214 | 0.9188 | 0.9716 |
| 0.0901 | 3.0 | 357 | 0.0863 | 0.9729 | 0.9584 | 0.9656 | 0.9799 |
| 0.0735 | 4.0 | 476 | 0.0699 | 0.9658 | 0.9701 | 0.9680 | 0.9827 |
| 0.0528 | 5.0 | 595 | 0.0674 | 0.9545 | 0.9761 | 0.9652 | 0.9831 |
| 0.0505 | 6.0 | 714 | 0.0659 | 0.9689 | 0.9757 | 0.9723 | 0.9841 |
| 0.0394 | 7.0 | 833 | 0.0696 | 0.9633 | 0.9771 | 0.9701 | 0.9839 |
| 0.0278 | 8.0 | 952 | 0.0728 | 0.9640 | 0.9772 | 0.9706 | 0.9837 |
| 0.0241 | 9.0 | 1071 | 0.0728 | 0.9694 | 0.9767 | 0.9730 | 0.9846 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.0