tinyshakespeare-13m

This model is a fine-tuned version of on the tiny_shakespeare dataset. It achieves the following results on the evaluation set:

• Loss: 4.9693

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: 0.0003
• train_batch_size: 8
• eval_batch_size: 8
• 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
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 40
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	32	7.3043
7.7103	2.0	64	5.9158
7.7103	3.0	96	5.6154
5.8675	4.0	128	5.3680
5.4479	5.0	160	5.2088
5.4479	6.0	192	5.1126
5.1825	7.0	224	5.0313
4.9945	8.0	256	4.9771
4.9945	9.0	288	4.9379
4.8838	10.0	320	4.9208
4.7883	11.0	352	4.8985
4.7883	12.0	384	4.8766
4.7261	13.0	416	4.8631
4.7261	14.0	448	4.8617
4.6621	15.0	480	4.8445
4.5955	16.0	512	4.8370
4.5955	17.0	544	4.8295
4.52	18.0	576	4.8215
4.4819	19.0	608	4.8278
4.4819	20.0	640	4.8169
4.4415	21.0	672	4.8252
4.3929	22.0	704	4.8199
4.3929	23.0	736	4.8243
4.3438	24.0	768	4.8340
4.3117	25.0	800	4.8309
4.3117	26.0	832	4.8410
4.2626	27.0	864	4.8439
4.2626	28.0	896	4.8437
4.2404	29.0	928	4.8404
4.1957	30.0	960	4.8540
4.1957	31.0	992	4.8560
4.1681	32.0	1024	4.8653
4.1441	33.0	1056	4.8725
4.1441	34.0	1088	4.8770
4.1097	35.0	1120	4.8798
4.0823	36.0	1152	4.8884
4.0823	37.0	1184	4.8869
4.0783	38.0	1216	4.8925
4.0783	39.0	1248	4.8948
4.0641	40.0	1280	4.8941

Framework versions

• Transformers 4.57.1
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.22.1