ssr-SmolLM2-1.7B

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

• Loss: 0.2052
• Wer: 0.1501

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: 5e-05
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 30
• total_train_batch_size: 30
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
9.3977	2.0	6936	0.2711	0.1470
6.7256	3.0	10404	0.2451	0.1421
5.3299	4.0	13872	0.2302	0.1454
4.335	5.0	17340	0.2242	0.1360
3.856	6.0	20808	0.2217	0.1368
3.0164	7.0	24276	0.2162	0.1455
2.7697	8.0	27744	0.2197	0.1333
2.6439	9.0	31212	0.2132	0.1298
2.1508	10.0	34680	0.2134	0.1521
2.0328	11.0	38148	0.2101	0.1343
1.2622	12.0	41616	0.2131	0.1333
1.138	13.0	45084	0.2116	0.1407
0.9745	14.0	48552	0.2139	0.1486
1.0203	15.0	52020	0.2083	0.1443
0.9933	16.0	55488	0.2082	0.1572
0.6237	17.0	58956	0.2073	0.1521
1.3526	18.0	62424	0.2059	0.1464
0.8515	19.0	65892	0.2067	0.1505
0.5371	19.9945	69340	0.2052	0.1501

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1