gpt2_m010_tiny-stories_1024

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

• Loss: 1.2176
• Accuracy: 0.6768

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 1.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.9272	0.0524	1000	2.4654	0.4425
1.9903	0.1048	2000	1.8096	0.5657
1.7407	0.1572	3000	1.6194	0.5995
1.623	0.2096	4000	1.5257	0.6163
1.5462	0.2619	5000	1.4623	0.6285
1.4894	0.3143	6000	1.4145	0.6377
1.4533	0.3667	7000	1.3773	0.6449
1.4216	0.4191	8000	1.3503	0.6501
1.394	0.4715	9000	1.3288	0.6542
1.3762	0.5239	10000	1.3074	0.6585
1.3551	0.5763	11000	1.2899	0.6620
1.3385	0.6287	12000	1.2746	0.6650
1.323	0.6811	13000	1.2622	0.6676
1.318	0.7334	14000	1.2517	0.6698
1.3015	0.7858	15000	1.2416	0.6719
1.2967	0.8382	16000	1.2337	0.6734
1.2859	0.8906	17000	1.2269	0.6749
1.2785	0.9430	18000	1.2216	0.6760
1.2753	0.9954	19000	1.2178	0.6768

Framework versions

• Transformers 4.42.3
• Pytorch 2.2.2+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1