google_electra-small-discriminator

This model is a fine-tuned version of google/electra-small-discriminator on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.5391
Accuracy: 0.8906
F1 Macro: 0.4073
Precision Destination: 0.9266
Recall Destination: 0.9597
Precision Origin: 0.6034
Recall Origin: 0.7955
Precision Other: 0.0
Recall Other: 0.0
Precision Transit: 0.0
Recall Transit: 0.0
Super Metric: 1.7552
Raw Super Metric: 1.7552

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
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
num_epochs: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Precision Destination	Recall Destination	Precision Origin	Recall Origin	Super Metric	Raw Super Metric
0.9832	1.0	185	0.8369	0.8580	0.2309	0.8580	1.0	0.0	0.0	0.98	1.0
0.6996	2.0	370	0.5563	0.8349	0.3584	0.9281	0.8949	0.3889	0.7955	1.6903	1.6903
0.5078	3.0	555	0.4766	0.8599	0.3710	0.9305	0.9284	0.44	0.75	1.6784	1.6784
0.4209	4.0	740	0.4103	0.8714	0.3847	0.9332	0.9374	0.4861	0.7955	1.7328	1.7328
0.3891	5.0	925	0.4396	0.8810	0.3957	0.9338	0.9463	0.5294	0.8182	1.7645	1.7645
0.2696	6.0	1110	0.5042	0.8887	0.4026	0.9304	0.9575	0.5738	0.7955	1.7529	1.7529
0.3568	7.0	1295	0.5009	0.8925	0.4093	0.9267	0.9620	0.6140	0.7955	1.7574	1.7574
0.2611	8.0	1480	0.5011	0.8772	0.3907	0.9316	0.9441	0.5147	0.7955	1.7395	1.7395
0.2724	9.0	1665	0.5354	0.8925	0.4093	0.9267	0.9620	0.6140	0.7955	1.7574	1.7574
0.3317	10.0	1850	0.5391	0.8906	0.4073	0.9266	0.9597	0.6034	0.7955	1.7552	1.7552

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1

Loading…

google_electra-small-discriminator

This model is a fine-tuned version of google/electra-small-discriminator on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.5391

Accuracy: 0.8906

F1 Macro: 0.4073

Precision Destination: 0.9266

Recall Destination: 0.9597

Precision Origin: 0.6034

Recall Origin: 0.7955

Precision Other: 0.0

Recall Other: 0.0

Precision Transit: 0.0

Recall Transit: 0.0

Super Metric: 1.7552

Raw Super Metric: 1.7552

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

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

num_epochs: 10

mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Precision Destination	Recall Destination	Precision Origin	Recall Origin	Super Metric	Raw Super Metric
0.9832	1.0	185	0.8369	0.8580	0.2309	0.8580	1.0	0.0	0.0	0.98	1.0
0.6996	2.0	370	0.5563	0.8349	0.3584	0.9281	0.8949	0.3889	0.7955	1.6903	1.6903
0.5078	3.0	555	0.4766	0.8599	0.3710	0.9305	0.9284	0.44	0.75	1.6784	1.6784
0.4209	4.0	740	0.4103	0.8714	0.3847	0.9332	0.9374	0.4861	0.7955	1.7328	1.7328
0.3891	5.0	925	0.4396	0.8810	0.3957	0.9338	0.9463	0.5294	0.8182	1.7645	1.7645
0.2696	6.0	1110	0.5042	0.8887	0.4026	0.9304	0.9575	0.5738	0.7955	1.7529	1.7529
0.3568	7.0	1295	0.5009	0.8925	0.4093	0.9267	0.9620	0.6140	0.7955	1.7574	1.7574
0.2611	8.0	1480	0.5011	0.8772	0.3907	0.9316	0.9441	0.5147	0.7955	1.7395	1.7395
0.2724	9.0	1665	0.5354	0.8925	0.4093	0.9267	0.9620	0.6140	0.7955	1.7574	1.7574
0.3317	10.0	1850	0.5391	0.8906	0.4073	0.9266	0.9597	0.6034	0.7955	1.7552	1.7552

Framework versions

Transformers 4.51.3

Pytorch 2.6.0+cu124

Datasets 3.5.0

Tokenizers 0.21.1

Fariman/google_electra-small-discriminator

google_electra-small-discriminator

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

google_electra-small-discriminator

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions