👨🔥💢ViSoBERT human finetune word (segmented)
Collection
ViSoBERT finetune for HSD - with human-reference annotated data, in word-level (using vncorenlp segmentation). Numbers denote different seeds • 5 items • Updated
# Load model directly
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained("trhgquan/visobert-human-finetune-seg-seed-6969")
model = AutoModelForSequenceClassification.from_pretrained("trhgquan/visobert-human-finetune-seg-seed-6969")Quick Links
visobert-human-finetune-seg-seed-6969
This model is a fine-tuned version of uitnlp/visobert on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3589
- Accuracy: 0.8825
- Precision: 0.7296
- Recall: 0.6876
- F1: 0.7068
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: 0.0001
- train_batch_size: 64
- eval_batch_size: 64
- 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_steps: 500
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 346 | 0.3514 | 0.8656 | 0.6970 | 0.6739 | 0.6527 |
| 0.3551 | 2.0 | 692 | 0.3589 | 0.8825 | 0.7296 | 0.6876 | 0.7068 |
| 0.1986 | 3.0 | 1038 | 0.4328 | 0.8787 | 0.7072 | 0.7021 | 0.7043 |
| 0.1986 | 4.0 | 1384 | 0.5486 | 0.8829 | 0.7486 | 0.6557 | 0.6838 |
| 0.1016 | 5.0 | 1730 | 0.6443 | 0.8784 | 0.7173 | 0.6937 | 0.7013 |
| 0.0687 | 6.0 | 2076 | 0.6960 | 0.8709 | 0.6958 | 0.6813 | 0.6861 |
| 0.0687 | 7.0 | 2422 | 0.6090 | 0.875 | 0.7156 | 0.6963 | 0.7026 |
Framework versions
- Transformers 4.51.1
- Pytorch 2.5.1+cu124
- Datasets 3.5.0
- Tokenizers 0.21.0
- Downloads last month
- 2
Model tree for trhgquan/visobert-human-finetune-seg-seed-6969
Base model
uitnlp/visobert
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="trhgquan/visobert-human-finetune-seg-seed-6969")