exaone_CSAT_test

This model is a fine-tuned version of LGAI-EXAONE/EXAONE-3.0-7.8B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.4792
• Accuracy: 0.5628
• F1: 0.5965

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: 1e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 200
• training_steps: 2100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
24.6602	0.1121	50	24.1094	0.5477	0.5761
12.8586	0.2242	100	6.5703	0.5729	0.6062
0.3657	0.3363	150	0.4956	0.5678	0.6005
0.5527	0.4484	200	0.4880	0.5678	0.6005
0.9587	0.5605	250	0.5098	0.5729	0.6054
0.9119	0.6726	300	0.4468	0.5678	0.6016
0.0989	0.7848	350	0.4690	0.5729	0.6066
0.6981	0.8969	400	0.4612	0.5628	0.5965
0.5197	1.0090	450	0.4792	0.5628	0.5965

Framework versions

• PEFT 0.13.2
• Transformers 4.46.2
• Pytorch 2.5.1+cu121
• Datasets 3.1.0
• Tokenizers 0.20.3