bert_base_lda_100_stsb

This model is a fine-tuned version of gokulsrinivasagan/bert_base_lda_100 on the GLUE STSB dataset. It achieves the following results on the evaluation set:

• Loss: 2.3354
• Pearson: nan
• Spearmanr: nan
• Combined Score: nan

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: 0.001
• 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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Pearson	Spearmanr	Combined Score
6.0379	1.0	23	2.8532	nan	nan	nan
2.286	2.0	46	2.6158	nan	nan	nan
2.1985	3.0	69	2.3354	nan	nan	nan
2.1934	4.0	92	2.4655	nan	nan	nan
2.1771	5.0	115	2.5613	nan	nan	nan
2.1903	6.0	138	2.3448	nan	nan	nan
2.2164	7.0	161	3.0915	nan	nan	nan
2.2509	8.0	184	2.3759	nan	nan	nan

Framework versions

• Transformers 4.46.3
• Pytorch 2.2.1+cu118
• Datasets 2.17.0
• Tokenizers 0.20.3