Sam4rano/nollysenti_afro-xlmr-base

results_afro-xlmr-base

This model is a fine-tuned version of Davlan/afro-xlmr-base on the Davlan/nollysenti dataset. It achieves the following results on the evaluation set:

• Loss: 0.3771
• Accuracy: 0.9129
• F1: 0.9149
• Precision: 0.9080
• Recall: 0.9219
• Pred Classes: 2
• Label Classes: 2

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: 3e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• 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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 4

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Pred Classes	Label Classes
0.576	0.3546	100	0.4729	0.8258	0.8121	0.8996	0.7401	2	2
0.4317	0.7092	200	0.4224	0.8802	0.8795	0.9006	0.8593	2	2
0.4079	1.0638	300	0.3737	0.8942	0.8967	0.8912	0.9021	2	2
0.3399	1.4184	400	0.3896	0.8974	0.9009	0.8850	0.9174	2	2
0.3357	1.7730	500	0.3707	0.9082	0.9102	0.9061	0.9144	2	2
0.3256	2.1277	600	0.3689	0.9176	0.9188	0.9202	0.9174	2	2
0.2899	2.4823	700	0.3789	0.9253	0.9252	0.9429	0.9083	2	2
0.2793	2.8369	800	0.3703	0.9160	0.9189	0.9027	0.9358	2	2
0.2712	3.1915	900	0.3737	0.9253	0.9273	0.9189	0.9358	2	2
0.2523	3.5461	1000	0.4002	0.9114	0.9143	0.8994	0.9297	2	2
0.2669	3.9007	1100	0.3847	0.9191	0.9210	0.9154	0.9266	2	2

Framework versions

• Transformers 4.56.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.0