electra-finetuned-combo-textbook

This model is a fine-tuned version of google/electra-small-generator on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.0564

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss
3.1589	1.0	2444	2.8061
2.8979	2.0	4888	2.6797
2.7362	3.0	7332	2.5658
2.6336	4.0	9776	2.5164
2.5261	5.0	12220	2.4465
2.4427	6.0	14664	2.4096
2.3747	7.0	17108	2.3456
2.2923	8.0	19552	2.3135
2.2594	9.0	21996	2.2672
2.2265	10.0	24440	2.2565
2.2075	11.0	26884	2.2290
2.1548	12.0	29328	2.1606
2.0962	13.0	31772	2.2014
2.0714	14.0	34216	2.1514
2.0423	15.0	36660	2.1457
2.0263	16.0	39104	2.0935
2.0029	17.0	41548	2.1112
1.9786	18.0	43992	2.0855
1.9352	19.0	46436	2.0737
1.9182	20.0	48880	2.0564

Framework versions

• Transformers 4.38.2
• Pytorch 1.13.1+cu116
• Datasets 2.18.0
• Tokenizers 0.15.2