This model is a fine-tuned version of Hartunka/distilbert_rand_100_v1 on the GLUE STSB dataset. It achieves the following results on the evaluation set:

Loss: 2.1966
Pearson: 0.2913
Spearmanr: 0.2877
Combined Score: 0.2895

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: 256
eval_batch_size: 256
seed: 10
optimizer: Use 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Pearson	Spearmanr	Combined Score
2.9205	1.0	23	2.3796	0.1291	0.1097	0.1194
1.9664	2.0	46	2.4950	0.1693	0.1372	0.1532
1.779	3.0	69	2.6714	0.2124	0.1933	0.2029
1.4527	4.0	92	2.1966	0.2913	0.2877	0.2895
1.1079	5.0	115	2.6246	0.2674	0.2614	0.2644
0.8434	6.0	138	2.5472	0.2959	0.2946	0.2953
0.6384	7.0	161	2.6544	0.2946	0.2885	0.2916
0.487	8.0	184	3.0336	0.2802	0.2731	0.2767
0.4161	9.0	207	2.9268	0.2597	0.2523	0.2560

Framework versions

Transformers 4.50.2
Pytorch 2.2.1+cu121
Datasets 2.18.0
Tokenizers 0.21.1

Hartunka/distilbert_rand_100_v1_stsb

distilbert_rand_100_v1_stsb

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Hartunka/
distilbert_rand_100_v1_stsb