tinybert_base_train_book_ent_15p_s_init_kd_complete

This model is a fine-tuned version of google/bert_uncased_L-4_H-512_A-8 on the gokulsrinivasagan/processed_wikitext-103-raw-v1-ld dataset. It achieves the following results on the evaluation set:

• Loss: 102.6946
• Accuracy: 0.0678

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: 0.0001
• train_batch_size: 96
• eval_batch_size: 96
• 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
• lr_scheduler_warmup_steps: 10000
• num_epochs: 24

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
476.1804	4.1982	10000	487.3371	0.0019
187.4831	8.3963	20000	177.2452	0.0604
138.1048	12.5945	30000	127.5909	0.0661
120.8144	16.7926	40000	110.4267	0.0687
113.2977	20.9908	50000	102.6946	0.0678

Framework versions

• Transformers 4.51.2
• Pytorch 2.6.0+cu126
• Datasets 3.5.0
• Tokenizers 0.21.1