universal_classifier

This model is a fine-tuned version of google/embeddinggemma-300m on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1127
• Accuracy: 0.685

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• distributed_type: multi-GPU
• num_devices: 2
• total_train_batch_size: 32
• total_eval_batch_size: 32
• 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
• training_steps: 22500

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.3060	0.0132	500	0.2985	0.175
0.2972	0.0264	1000	0.2958	0.192
0.2889	0.0396	1500	0.2951	0.193
0.2827	0.0527	2000	0.2945	0.216
0.2961	0.0659	2500	0.2883	0.208
0.2520	0.0791	3000	0.2509	0.359
0.2181	0.0923	3500	0.2166	0.449
0.1710	0.1055	4000	0.1697	0.561
0.1655	0.1187	4500	0.1563	0.588
0.1344	0.1319	5000	0.1433	0.607
0.1416	0.1451	5500	0.1378	0.631
0.1347	0.1582	6000	0.1369	0.609
0.1250	0.1714	6500	0.1358	0.63
0.1560	0.1846	7000	0.1332	0.638
0.1368	0.1978	7500	0.1322	0.648
0.1275	0.2110	8000	0.1331	0.649
0.1257	0.2242	8500	0.1297	0.654
0.1349	0.2374	9000	0.1288	0.657
0.1306	0.2506	9500	0.1262	0.655
0.1161	0.2637	10000	0.1243	0.652
0.1315	0.2769	10500	0.1249	0.666
0.1298	0.2901	11000	0.1245	0.659
0.1141	0.3033	11500	0.1221	0.664
0.1216	0.3165	12000	0.1205	0.668
0.1216	0.3297	12500	0.1204	0.67
0.1211	0.3429	13000	0.1214	0.671
0.1179	0.3561	13500	0.1204	0.666
0.1246	0.3692	14000	0.1176	0.67
0.1132	0.3824	14500	0.1170	0.669
0.1190	0.3956	15000	0.1177	0.672
0.1075	0.4088	15500	0.1173	0.688
0.1177	0.4220	16000	0.1140	0.683
0.0958	0.4352	16500	0.1150	0.678
0.1247	0.4484	17000	0.1147	0.676
0.1059	0.4615	17500	0.1138	0.687
0.1058	0.4747	18000	0.1144	0.681
0.1070	0.4879	18500	0.1146	0.69
0.1166	0.5011	19000	0.1134	0.691
0.1139	0.5143	19500	0.1128	0.684
0.1104	0.5275	20000	0.1139	0.685
0.1080	0.5407	20500	0.1144	0.677
0.1145	0.5539	21000	0.1128	0.698
0.1246	0.5670	21500	0.1126	0.688
0.1258	0.5802	22000	0.1128	0.679
0.1089	0.5934	22500	0.1127	0.685

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 5.0.0
• Tokenizers 0.22.2