jonasknobloch/gpt2-ts_cx-en_00000-00009_50k

gpt2+ts_cx-en_00000-00009_50k

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

• Loss: 3.4121
• Accuracy: 0.3895

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: 64
• eval_batch_size: 64
• 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
4.375	0.04	10000	4.2815	0.3111
4.0754	0.08	20000	3.9984	0.3341
3.9409	0.11	30000	3.8615	0.3457
3.8554	0.15	40000	3.7798	0.3531
3.7973	0.19	50000	3.7210	0.3584
3.7421	0.23	60000	3.6750	0.3630
3.7097	0.27	70000	3.6378	0.3664
3.6741	0.3	80000	3.6061	0.3694
3.6599	0.34	90000	3.5803	0.3718
3.6356	0.38	100000	3.5584	0.3741
3.6131	0.42	110000	3.5423	0.3758
3.5991	0.46	120000	3.5254	0.3776
3.591	0.49	130000	3.5108	0.3790
3.574	0.53	140000	3.4966	0.3805
3.5606	0.57	150000	3.4866	0.3815
3.5516	0.61	160000	3.4739	0.3828
3.5423	0.64	170000	3.4650	0.3838
3.5298	0.68	180000	3.4560	0.3847
3.5287	0.72	190000	3.4479	0.3857
3.5187	0.76	200000	3.4408	0.3863
3.5157	0.8	210000	3.4339	0.3870
3.5042	0.83	220000	3.4286	0.3876
3.5033	0.87	230000	3.4229	0.3883
3.501	0.91	240000	3.4188	0.3888
3.4946	0.95	250000	3.4149	0.3892
3.4971	0.99	260000	3.4126	0.3894

Framework versions

• Transformers 4.37.1
• Pytorch 2.1.2+cu121
• Datasets 2.16.1
• Tokenizers 0.15.1