addrsRoadSjn_medium_model

This model is a fine-tuned version of openai/whisper-medium on the whsNect/addrsRoadSjn dataset. It achieves the following results on the evaluation set:

• Loss: 0.0339
• Cer: 1.7686

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Cer
0.7856	0.0349	200	0.0621	2.5624
0.0621	0.0697	400	0.0515	7.4832
0.0481	0.1046	600	0.0490	16.9947
0.0412	0.1394	800	0.0416	2.1072
0.0378	0.1743	1000	0.0379	3.2045
0.0375	0.2092	1200	0.0377	1.8046
0.0382	0.2440	1400	0.0353	2.1989
0.0351	0.2789	1600	0.0349	1.8569
0.0321	0.3138	1800	0.0341	1.8106
0.0309	0.3486	2000	0.0339	1.7686

Framework versions

• Transformers 4.41.2
• Pytorch 2.2.2+cu121
• Datasets 3.2.0
• Tokenizers 0.19.1