Gpt2 Small ExpandedbabyLM 100M Exp3 Cnp Ratio 0p50 Mix 0p10 44

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: 256
• eval_batch_size: 256
• seed: 44
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 4000
• num_epochs: 20

Training results

Framework versions

• Transformers 4.30.2
• Pytorch 2.11.0+cu130
• Datasets 4.1.1
• Tokenizers 0.13.3

--- tags: - generated_from_trainer model-index: - name: gpt2_small_expandedbabyLM_100M_exp3_cnp_ratio_0p50_mix_0p10_44 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # gpt2_small_expandedbabyLM_100M_exp3_cnp_ratio_0p50_mix_0p10_44 This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. It achieves the following results on the evaluation set: - Loss: 3.5865 ## 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: 256 - eval_batch_size: 256 - seed: 44 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - lr_scheduler_warmup_steps: 4000 - num_epochs: 20 ### Training results | Training Loss | Epoch | Step | Validation Loss | |:-------------:|:-----:|:-----:|:---------------:| | 3.733 | 1.0 | 4729 | 4.2298 | | 3.4035 | 2.0 | 9458 | 3.8080 | | 3.2695 | 3.0 | 14187 | 3.7428 | | 3.1878 | 4.0 | 18916 | 3.6899 | | 3.1372 | 5.0 | 23645 | 3.5871 | | 3.087 | 6.0 | 28374 | 3.5865 | | 3.0527 | 7.0 | 33103 | 3.5934 | | 3.0138 | 8.0 | 37832 | 3.6094 | ### Framework versions - Transformers 4.30.2 - Pytorch 2.11.0+cu130 - Datasets 4.1.1 - Tokenizers 0.13.3