ania3000/ossbert-morph-v2.1

trainer_output

This model is a fine-tuned version of AlexeySorokin/ossbert-onc-unlab-from_multilingual-bs64-5epochs on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3048
• Accuracy: 96.0396
• Sentence accuracy: 60.7339

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Sentence accuracy
1.0056	1.0	546	0.3472	92.4552	44.0367
0.3162	2.0	1092	0.2379	94.3602	53.3945
0.2008	3.0	1638	0.2177	94.6359	54.1284
0.1436	4.0	2184	0.2147	94.9743	55.9633
0.111	5.0	2730	0.2067	95.3754	59.0826
0.0842	6.0	3276	0.2132	95.5258	59.0826
0.0643	7.0	3822	0.2264	95.4506	60.3670
0.0481	8.0	4368	0.2335	95.5007	59.8165
0.0389	9.0	4914	0.2442	95.6761	58.8991
0.0305	10.0	5460	0.2603	95.6260	59.0826
0.0197	11.0	6006	0.2640	95.6135	60.3670
0.015	12.0	6552	0.2649	95.8140	60.1835
0.0122	13.0	7098	0.2686	95.6135	59.8165
0.0093	14.0	7644	0.2790	95.5508	59.6330
0.0065	15.0	8190	0.2735	95.7764	60.3670
0.0069	16.0	8736	0.2867	95.8641	61.8349
0.005	17.0	9282	0.2903	95.8892	60.9174
0.0041	18.0	9828	0.2934	95.9519	61.4679
0.0027	19.0	10374	0.3040	95.9017	61.2844
0.0031	20.0	10920	0.3044	95.9393	61.1009
0.0026	21.0	11466	0.2994	95.9895	61.1009
0.0016	22.0	12012	0.3033	95.9769	61.1009
0.0012	23.0	12558	0.3034	96.0271	60.9174
0.0011	24.0	13104	0.3028	96.0271	61.2844
0.0012	25.0	13650	0.3048	96.0396	60.7339

Framework versions

• Transformers 4.57.3
• Pytorch 2.11.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2