universal_classifier_nouls

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: 0.5265
• Kl: 0.0138
• Brier: 0.0054
• Accuracy: 0.9270

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: 20000

Training results

Training Loss	Epoch	Step	Validation Loss	Kl	Brier	Accuracy
0.5879	0.1063	500	0.6051	0.0924	0.0382	0.7757
0.5803	0.2126	1000	0.5863	0.0735	0.0299	0.8044
0.5631	0.3189	1500	0.5734	0.0607	0.0246	0.8216
0.5740	0.4252	2000	0.5661	0.0534	0.0214	0.8395
0.5584	0.5315	2500	0.5572	0.0445	0.0179	0.8559
0.5587	0.6378	3000	0.5552	0.0425	0.0170	0.8597
0.5232	0.7440	3500	0.5520	0.0393	0.0156	0.8673
0.5395	0.8503	4000	0.5480	0.0353	0.0141	0.8715
0.5412	0.9566	4500	0.5442	0.0314	0.0124	0.8825
0.5330	1.0629	5000	0.5408	0.0281	0.0111	0.8888
0.5336	1.1692	5500	0.5392	0.0265	0.0105	0.8887
0.5509	1.2755	6000	0.5404	0.0277	0.0111	0.8900
0.5570	1.3818	6500	0.5367	0.0240	0.0096	0.8952
0.5295	1.4881	7000	0.5359	0.0232	0.0093	0.8983
0.5408	1.5944	7500	0.5370	0.0242	0.0097	0.8911
0.5273	1.7007	8000	0.5334	0.0207	0.0082	0.9086
0.5233	1.8070	8500	0.5325	0.0198	0.0079	0.9063
0.5331	1.9133	9000	0.5331	0.0204	0.0081	0.9082
0.5384	2.0196	9500	0.5316	0.0188	0.0075	0.9102
0.5284	2.1259	10000	0.5312	0.0185	0.0074	0.9124
0.5389	2.2321	10500	0.5304	0.0177	0.0070	0.9155
0.5258	2.3384	11000	0.5299	0.0172	0.0068	0.9161
0.5157	2.4447	11500	0.5301	0.0174	0.0069	0.9165
0.5175	2.5510	12000	0.5301	0.0174	0.0069	0.9152
0.5355	2.6573	12500	0.5292	0.0164	0.0065	0.9161
0.5324	2.7636	13000	0.5287	0.0160	0.0063	0.9186
0.5203	2.8699	13500	0.5288	0.0161	0.0064	0.9192
0.5237	2.9762	14000	0.5286	0.0158	0.0063	0.9194
0.5313	3.0825	14500	0.5284	0.0156	0.0062	0.9227
0.5303	3.1888	15000	0.5277	0.0150	0.0059	0.9235
0.5265	3.2951	15500	0.5279	0.0151	0.0060	0.9213
0.5177	3.4014	16000	0.5276	0.0149	0.0059	0.9236
0.5193	3.5077	16500	0.5274	0.0147	0.0058	0.9256
0.5254	3.6139	17000	0.5272	0.0144	0.0057	0.9246
0.5230	3.7202	17500	0.5271	0.0144	0.0057	0.9236
0.5137	3.8265	18000	0.5271	0.0144	0.0057	0.9243
0.5250	3.9328	18500	0.5269	0.0141	0.0056	0.9251
0.5349	4.0391	19000	0.5267	0.0140	0.0055	0.9266
0.5122	4.1454	19500	0.5266	0.0138	0.0055	0.9267
0.5121	4.2517	20000	0.5265	0.0138	0.0054	0.9270

Framework versions

• Transformers 5.17.0
• Pytorch 2.10.0+cu128
• Datasets 5.0.0
• Tokenizers 0.23.2