tinyshakespeare-42m

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

• Loss: 4.8737

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: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss
No log	1.0	16	6.9032
No log	2.0	32	5.9671
No log	3.0	48	5.7496
6.6567	4.0	64	5.6037
6.6567	5.0	80	5.4368
6.6567	6.0	96	5.2910
5.4744	7.0	112	5.1749
5.4744	8.0	128	5.0940
5.4744	9.0	144	5.0452
5.0133	10.0	160	5.0255
5.0133	11.0	176	5.0034
5.0133	12.0	192	4.9969
4.756	13.0	208	4.9831
4.756	14.0	224	4.9877
4.756	15.0	240	4.9961
4.5668	16.0	256	5.0005
4.5668	17.0	272	5.0135
4.5668	18.0	288	5.0311
4.3775	19.0	304	5.0535
4.3775	20.0	320	5.0667
4.3775	21.0	336	5.0927
4.1854	22.0	352	5.1175
4.1854	23.0	368	5.1487
4.1854	24.0	384	5.1702
3.9762	25.0	400	5.1887
3.9762	26.0	416	5.2351
3.9762	27.0	432	5.2616
3.9762	28.0	448	5.2936
3.793	29.0	464	5.3212
3.793	30.0	480	5.3417
3.793	31.0	496	5.3652
3.639	32.0	512	5.3838
3.639	33.0	528	5.4027
3.639	34.0	544	5.4140
3.5439	35.0	560	5.4247
3.5439	36.0	576	5.4353
3.5439	37.0	592	5.4410
3.4797	38.0	608	5.4424
3.4797	39.0	624	5.4424
3.4797	40.0	640	5.4422

Framework versions

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