Niger-Malitamasheq-finetuned

This model is a fine-tuned version of LIA-AvignonUniversity/IWSLT2022-Niger-Mali on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.9747
• Cer: 0.9979

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: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 60

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
8.7075	8.57	30	5.0922	0.9979
4.4557	17.14	60	3.3251	0.9979
3.5394	25.71	90	3.1326	0.9979
3.2067	34.29	120	3.0365	0.9979
3.0958	42.86	150	2.9906	0.9979
3.0147	51.43	180	2.9747	0.9979

Framework versions

• Transformers 4.32.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.4
• Tokenizers 0.13.3