google_electra-small-discriminator_negsam

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: 0.6770
Accuracy: 0.8983
F1 Macro: 0.4856
Precision Destination: 0.9309
Recall Destination: 0.9642
Precision Origin: 0.6346
Recall Origin: 0.75
Precision Other: 0.6667
Recall Other: 0.2
Precision Transit: 0.0
Recall Transit: 0.0
Super Metric: 1.7142
Raw Super Metric: 1.7142

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	Precision Other	Recall Other	Super Metric	Raw Super Metric
1.115	1.0	269	0.9390	0.8695	0.3263	0.8802	0.9866	0.6	0.2727	0.0	0.0	1.2527	1.2593
0.7979	2.0	538	0.6104	0.8464	0.4048	0.9310	0.9060	0.4304	0.7727	0.2857	0.1	1.6788	1.6788
0.5499	3.0	807	0.6350	0.8599	0.3989	0.9113	0.9418	0.4717	0.5682	0.3333	0.1	1.5100	1.5100
0.3978	4.0	1076	0.4447	0.8656	0.4367	0.9386	0.9239	0.4730	0.7955	0.4286	0.15	1.7194	1.7194
0.4484	5.0	1345	0.5825	0.8887	0.4410	0.9304	0.9575	0.5789	0.75	0.5	0.1	1.7075	1.7075
0.32	6.0	1614	0.6012	0.8887	0.4423	0.9284	0.9575	0.5893	0.75	0.5	0.1	1.7075	1.7075
0.285	7.0	1883	0.6685	0.8944	0.4804	0.9306	0.9597	0.6226	0.75	0.5714	0.2	1.7097	1.7097
0.2184	8.0	2152	0.7101	0.8964	0.4825	0.9307	0.9620	0.6346	0.75	0.5714	0.2	1.7120	1.7120
0.243	9.0	2421	0.6576	0.8983	0.4856	0.9309	0.9642	0.6346	0.75	0.6667	0.2	1.7142	1.7142
0.2218	10.0	2690	0.6770	0.8983	0.4856	0.9309	0.9642	0.6346	0.75	0.6667	0.2	1.7142	1.7142

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1

Loading…

google_electra-small-discriminator_negsam

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: 0.6770

Accuracy: 0.8983

F1 Macro: 0.4856

Precision Destination: 0.9309

Recall Destination: 0.9642

Precision Origin: 0.6346

Recall Origin: 0.75

Precision Other: 0.6667

Recall Other: 0.2

Precision Transit: 0.0

Recall Transit: 0.0

Super Metric: 1.7142

Raw Super Metric: 1.7142

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	Precision Other	Recall Other	Super Metric	Raw Super Metric
1.115	1.0	269	0.9390	0.8695	0.3263	0.8802	0.9866	0.6	0.2727	0.0	0.0	1.2527	1.2593
0.7979	2.0	538	0.6104	0.8464	0.4048	0.9310	0.9060	0.4304	0.7727	0.2857	0.1	1.6788	1.6788
0.5499	3.0	807	0.6350	0.8599	0.3989	0.9113	0.9418	0.4717	0.5682	0.3333	0.1	1.5100	1.5100
0.3978	4.0	1076	0.4447	0.8656	0.4367	0.9386	0.9239	0.4730	0.7955	0.4286	0.15	1.7194	1.7194
0.4484	5.0	1345	0.5825	0.8887	0.4410	0.9304	0.9575	0.5789	0.75	0.5	0.1	1.7075	1.7075
0.32	6.0	1614	0.6012	0.8887	0.4423	0.9284	0.9575	0.5893	0.75	0.5	0.1	1.7075	1.7075
0.285	7.0	1883	0.6685	0.8944	0.4804	0.9306	0.9597	0.6226	0.75	0.5714	0.2	1.7097	1.7097
0.2184	8.0	2152	0.7101	0.8964	0.4825	0.9307	0.9620	0.6346	0.75	0.5714	0.2	1.7120	1.7120
0.243	9.0	2421	0.6576	0.8983	0.4856	0.9309	0.9642	0.6346	0.75	0.6667	0.2	1.7142	1.7142
0.2218	10.0	2690	0.6770	0.8983	0.4856	0.9309	0.9642	0.6346	0.75	0.6667	0.2	1.7142	1.7142

Framework versions

Transformers 4.51.3

Pytorch 2.6.0+cu124

Datasets 3.5.0

Tokenizers 0.21.1

Fariman/google_electra-small-discriminator_negsam

google_electra-small-discriminator_negsam

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

google_electra-small-discriminator_negsam

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions