Eimhin03/output_model_shunyalabs_data_only_20000_steps

outout_model_shunyalabs_data_only

This model is a fine-tuned version of Eimhin03/outout_model_shunyalabs_data_only on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6551
• Wer: 32.8164

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 200
• training_steps: 20000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
0.4197	0.4883	1250	1.1524	59.0164
0.4407	0.9766	2500	0.9607	55.6196
0.1682	1.4648	3750	0.9998	54.7334
0.2280	1.9531	5000	0.9385	49.5052
0.0718	2.4414	6250	0.8866	45.9607
0.1222	2.9297	7500	0.8641	50.2437
0.0554	3.4180	8750	0.8247	43.1546
0.0874	3.9062	10000	0.7700	42.4605
0.0374	4.3945	11250	0.7748	42.0913
0.0219	4.8828	12500	0.7469	38.2514
0.0053	5.3711	13750	0.7218	37.0846
0.0065	5.8594	15000	0.6925	34.5444
0.0005	6.3477	16250	0.6818	34.7068
0.0007	6.8359	17500	0.6666	33.2890
0.0001	7.3242	18750	0.6584	32.7278
0.0005	7.8125	20000	0.6551	32.8164

Framework versions

• Transformers 5.0.1.dev0
• Pytorch 2.9.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.2