gpt2_m100_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: 1.1723
• Accuracy: 0.6875

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
2.8323	0.0506	1000	2.4024	0.4558
1.9289	0.1012	2000	1.7509	0.5790
1.6823	0.1518	3000	1.5725	0.6104
1.5643	0.2024	4000	1.4746	0.6285
1.4919	0.2530	5000	1.4118	0.6402
1.443	0.3036	6000	1.3691	0.6484
1.4054	0.3543	7000	1.3326	0.6552
1.3716	0.4049	8000	1.3062	0.6604
1.3477	0.4555	9000	1.2838	0.6649
1.3298	0.5061	10000	1.2637	0.6687
1.3088	0.5567	11000	1.2462	0.6723
1.2944	0.6073	12000	1.2335	0.6748
1.278	0.6579	13000	1.2207	0.6773
1.2658	0.7085	14000	1.2098	0.6796
1.2567	0.7591	15000	1.2005	0.6816
1.2506	0.8097	16000	1.1921	0.6832
1.241	0.8603	17000	1.1847	0.6847
1.2338	0.9109	18000	1.1789	0.6860
1.2306	0.9615	19000	1.1743	0.6871

Framework versions

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