metadata
library_name: transformers
license: mit
base_model: microsoft/mdeberta-v3-base
tags:
- generated_from_trainer
datasets:
- google/xtreme
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: mdeberta-v3-base-panx-wikiann-en
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: google/xtreme PAN-X.en
type: google/xtreme
args: PAN-X.en
metrics:
- name: Precision
type: precision
value: 0.8285338502007477
- name: Recall
type: recall
value: 0.8461049059804892
- name: F1
type: f1
value: 0.8372271964185787
- name: Accuracy
type: accuracy
value: 0.9318317274262442
mdeberta-v3-base-panx-wikiann-en
This model is a fine-tuned version of microsoft/mdeberta-v3-base on the google/xtreme PAN-X.en dataset. It achieves the following results on the evaluation set:
- Loss: 0.2520
- Precision: 0.8285
- Recall: 0.8461
- F1: 0.8372
- Accuracy: 0.9318
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: 2e-05
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.4565 | 1.0 | 625 | 0.2651 | 0.7942 | 0.8198 | 0.8068 | 0.9215 |
| 0.2612 | 2.0 | 1250 | 0.2490 | 0.8043 | 0.8285 | 0.8162 | 0.9257 |
| 0.2184 | 3.0 | 1875 | 0.2471 | 0.8175 | 0.8353 | 0.8263 | 0.9294 |
| 0.1636 | 4.0 | 2500 | 0.2493 | 0.8195 | 0.8434 | 0.8313 | 0.9308 |
| 0.1408 | 5.0 | 3125 | 0.2520 | 0.8285 | 0.8461 | 0.8372 | 0.9318 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.19.1