visobert-emotion-optimized-v3

tunakite03/visobert-emotion-optimized-v3

visobert-emotion-optimized-v3 is a machine learning model from tunakite03. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

3% of all-time downloads

All-time downloads

121

Public

Parameters

101M

10.1 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors406 MB · 100%

At a glance

Library: transformers
Model type: xlm-roberta
Access: Public
Created: Jan 24, 2026
Updated: Jan 24, 2026
SHA: 0fc6ee67

Model card

From the Hugging Face model README

visobert-emotion-optimized-v3

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.0063
Accuracy: 0.6537
F1 Macro: 0.6278
F1 Weighted: 0.6492
Precision: 0.6267
Recall: 0.6426

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: 8e-06
train_batch_size: 48
eval_batch_size: 48
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 96
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine_with_restarts
lr_scheduler_warmup_ratio: 0.1
num_epochs: 20
mixed_precision_training: Native AMP
label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	F1 Weighted	Precision	Recall
3.8954	0.3268	50	1.9468	0.1589	0.1059	0.1486	0.1493	0.1362
3.8656	0.6536	100	1.9339	0.2434	0.1753	0.2302	0.2236	0.1903
3.807	0.9804	150	1.8963	0.3309	0.2213	0.2697	0.2507	0.2630
3.656	1.3072	200	1.8048	0.3673	0.2202	0.2850	0.3656	0.2802
3.4504	1.6340	250	1.6582	0.4184	0.2977	0.3458	0.4413	0.3606
3.1892	1.9608	300	1.4728	0.4694	0.3699	0.3939	0.5098	0.4426
2.8037	2.2876	350	1.3209	0.5	0.4299	0.4321	0.5478	0.5030
2.5357	2.6144	400	1.1946	0.5408	0.4811	0.4914	0.5343	0.5661
2.2335	2.9412	450	1.1124	0.5496	0.5030	0.5120	0.5277	0.6036
2.0153	3.2680	500	1.0267	0.5831	0.5395	0.5580	0.5494	0.6186
1.981	3.5948	550	0.9960	0.6035	0.5650	0.5899	0.5513	0.6194
1.7674	3.9216	600	0.9611	0.5991	0.5515	0.5715	0.5888	0.6012
1.6524	4.2484	650	0.9391	0.6050	0.5675	0.5923	0.5601	0.6163
1.4879	4.5752	700	0.9185	0.6195	0.5750	0.5989	0.5828	0.6162
1.5042	4.9020	750	0.9301	0.6210	0.5825	0.6117	0.5820	0.6251
1.268	5.2288	800	0.9292	0.6093	0.5741	0.5949	0.5847	0.6082
1.319	5.5556	850	0.9423	0.6283	0.5960	0.6224	0.5884	0.6367
1.2712	5.8824	900	0.9488	0.6297	0.5917	0.6259	0.5820	0.6389
1.1265	6.2092	950	0.9460	0.6370	0.6042	0.6338	0.5893	0.6361
1.0817	6.5359	1000	0.9640	0.6356	0.6087	0.6331	0.5922	0.6559
1.0775	6.8627	1050	0.9468	0.6356	0.6074	0.6274	0.6057	0.6254
0.8771	7.1895	1100	0.9628	0.6501	0.6208	0.6492	0.6025	0.6577
0.9378	7.5163	1150	0.9718	0.6385	0.6060	0.6353	0.5931	0.6390
0.9276	7.8431	1200	0.9912	0.6443	0.6149	0.6444	0.5998	0.6407
0.8178	8.1699	1250	0.9782	0.6487	0.6187	0.6452	0.6097	0.6382
0.7594	8.4967	1300	0.9691	0.6516	0.6164	0.6464	0.6100	0.6348
0.7828	8.8235	1350	0.9795	0.6487	0.6146	0.6402	0.6132	0.6275

Framework versions

Transformers 4.57.6
Pytorch 2.9.0+cu126
Datasets 4.0.0
Tokenizers 0.22.2

Library: transformers
Type: xlm-roberta
Created: Jan 24, 2026
Updated: Jan 24, 2026