gpt2_mini_wiki_10M_32768_42

This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.9628
• Accuracy: 0.2499

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 40000
• training_steps: 100000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
7.8668	2.29	2000	7.5727	0.1094
6.6433	4.57	4000	6.8287	0.1424
6.1455	6.86	6000	6.4265	0.1593
5.8047	9.14	8000	6.1424	0.1703
5.5178	11.43	10000	5.9197	0.1792
5.2628	13.71	12000	5.7265	0.1893
5.0215	16.0	14000	5.5729	0.1983
4.8039	18.29	16000	5.4367	0.2067
4.6083	20.57	18000	5.3292	0.2133
4.4403	22.86	20000	5.2371	0.2207
4.2847	25.14	22000	5.1595	0.2275
4.1617	27.43	24000	5.1019	0.2331
4.0524	29.71	26000	5.0581	0.2380
3.9608	32.0	28000	5.0102	0.2415
3.8727	34.29	30000	5.0070	0.2440
3.7981	36.57	32000	4.9780	0.2466
3.7366	38.86	34000	4.9629	0.2491
3.6679	41.14	36000	4.9628	0.2499
3.6185	43.43	38000	4.9748	0.2506
3.5711	45.71	40000	4.9740	0.2516
3.5205	48.0	42000	4.9730	0.2523

Framework versions

• Transformers 4.30.2
• Pytorch 2.9.0+cu128
• Datasets 4.1.1
• Tokenizers 0.13.3