mt5-bambara-resumer-final

This model is a fine-tuned version of Fatoumataa/mt5-bambara-phase0-pro on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.3116
• Rouge1: 0.509
• Rouge2: 0.2569
• Rougel: 0.354
• Rougelsum: 0.3542

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: 1e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAFACTOR and the args are: No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 8
• label_smoothing_factor: 0.1

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge1	Rouge2	Rougel	Rougelsum
35.3040	1.0	1128	3.8356	0.4562	0.2151	0.3057	0.3056
32.9291	2.0	2256	3.6969	0.4756	0.2304	0.3237	0.3239
31.6170	3.0	3384	3.5895	0.4849	0.2372	0.3314	0.3314
30.4512	4.0	4512	3.4988	0.4863	0.2388	0.3343	0.3344
29.9236	5.0	5640	3.4431	0.4915	0.2443	0.3402	0.3403
29.3764	6.0	6768	3.3960	0.5022	0.2513	0.3459	0.3459
29.0224	7.0	7896	3.3521	0.5072	0.2553	0.3517	0.3517
28.6897	8.0	9024	3.3116	0.509	0.2569	0.354	0.3542

Framework versions

• Transformers 5.0.0
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2