robbert-v2-dutch-base-finetuned-emotion Open-source Model - Free for Dutch Text Emotion Classification

Robbert V2 Dutch Base Finetuned Emotion

Developed by antalvdb

This model is a fine-tuned version of the RobBERT v2 Dutch base model for sentiment analysis tasks, primarily used for sentiment classification of Dutch texts.

Text Classification

Transformers

Open Source License:MIT #Dutch Sentiment Analysis #High-Round Fine-Tuning #F1 Optimization

Downloads 67

Release Time : 7/25/2024

Model Overview

A sentiment analysis model fine-tuned on the RobBERT v2 Dutch base model, capable of classifying sentiment in Dutch texts.

Model Features

Dutch Sentiment Analysis

A sentiment analysis model specifically optimized for Dutch texts.

Based on RobBERT v2

Uses the RobBERT v2 Dutch base model as the pre-training foundation.

Fine-Tuning Optimization

Fine-tuned for 50 rounds on specific sentiment analysis tasks.

Model Capabilities

Dutch Text Classification

Sentiment Analysis

Use Cases

Sentiment Analysis

Social Media Sentiment Monitoring

Analyze the sentiment tendencies of Dutch social media posts.

Accuracy 52%, F1 score 0.5123

Customer Feedback Analysis

Perform sentiment classification on Dutch customer reviews.

🚀 robbert-v2-dutch-base-finetuned-emotion

This model is a fine - tuned version of pdelobelle/robbert-v2-dutch-base. It is trained on the None dataset and achieves the following results on the evaluation set:

Loss: 3.3545
Accuracy: 0.52
F1: 0.5123

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

🔧 Technical Details

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 2e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.5586	1.0	25	1.4429	0.42	0.2485
1.4425	2.0	50	1.3576	0.46	0.3533
1.2834	3.0	75	1.3207	0.5	0.4369
1.1051	4.0	100	1.3228	0.48	0.4217
0.9053	5.0	125	1.3705	0.49	0.4302
0.7326	6.0	150	1.4522	0.53	0.5019
0.5724	7.0	175	1.5445	0.53	0.5064
0.4411	8.0	200	1.6560	0.54	0.5120
0.3476	9.0	225	1.7233	0.51	0.4845
0.2324	10.0	250	1.9150	0.52	0.5056
0.1866	11.0	275	2.0207	0.52	0.4975
0.165	12.0	300	2.0863	0.52	0.5094
0.1291	13.0	325	2.1584	0.5	0.4833
0.0762	14.0	350	2.2296	0.55	0.5332
0.0577	15.0	375	2.3171	0.5	0.4986
0.0424	16.0	400	2.4509	0.5	0.4795
0.0253	17.0	425	2.5444	0.49	0.4917
0.0191	18.0	450	2.5894	0.51	0.5031
0.0123	19.0	475	2.7144	0.5	0.4995
0.01	20.0	500	2.7358	0.53	0.5231
0.0086	21.0	525	2.8282	0.48	0.4825
0.0064	22.0	550	2.8421	0.52	0.5244
0.0059	23.0	575	2.9267	0.53	0.5200
0.005	24.0	600	2.9568	0.52	0.5074
0.0044	25.0	625	3.0420	0.47	0.4755
0.0066	26.0	650	3.0421	0.48	0.4881
0.0039	27.0	675	3.1039	0.51	0.4960
0.0033	28.0	700	3.1226	0.51	0.4955
0.0033	29.0	725	3.1215	0.51	0.4999
0.003	30.0	750	3.1649	0.51	0.4980
0.0025	31.0	775	3.1716	0.5	0.4921
0.0028	32.0	800	3.2371	0.5	0.4956
0.0028	33.0	825	3.1730	0.52	0.5154
0.0055	34.0	850	3.1842	0.49	0.4884
0.0022	35.0	875	3.2324	0.51	0.4955
0.0023	36.0	900	3.2221	0.52	0.5089
0.002	37.0	925	3.2756	0.51	0.4981
0.0021	38.0	950	3.2866	0.51	0.5010
0.0019	39.0	975	3.2882	0.51	0.5010
0.0018	40.0	1000	3.2864	0.51	0.4967
0.0017	41.0	1025	3.3101	0.51	0.4967
0.0017	42.0	1050	3.3215	0.52	0.5089
0.0016	43.0	1075	3.3253	0.51	0.5043
0.0056	44.0	1100	3.3118	0.51	0.5043
0.0016	45.0	1125	3.3566	0.51	0.4981
0.0016	46.0	1150	3.3593	0.51	0.4981
0.0016	47.0	1175	3.3638	0.51	0.4981
0.0017	48.0	1200	3.3605	0.52	0.5089
0.0017	49.0	1225	3.3526	0.52	0.5123
0.0016	50.0	1250	3.3545	0.52	0.5123

Framework versions

Transformers 4.42.4
Pytorch 2.3.1+cu121
Datasets 2.20.0
Tokenizers 0.19.1

📄 License

This model is released under the MIT license.

Property	Details
Model Type	robbert-v2-dutch-base-finetuned-emotion
Training Data	None
Metrics	accuracy, f1

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご