distilbert-base-uncased-finetuned-CEFR

This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2528
• Accuracy: 0.3350
• Precision: 0.3202
• Recall: 0.6791
• F1: 0.2925

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	1.0	50	0.2342	0.3324	0.3240	0.6540	0.2960
No log	2.0	100	0.2326	0.3330	0.3166	0.6658	0.2841
No log	3.0	150	0.2362	0.3332	0.3171	0.6680	0.2882
No log	4.0	200	0.2410	0.3335	0.3238	0.6722	0.2979
No log	5.0	250	0.2468	0.3337	0.3254	0.6657	0.2964
No log	6.0	300	0.2455	0.3341	0.3190	0.6697	0.2937
No log	7.0	350	0.2404	0.3347	0.3226	0.6795	0.2931
No log	8.0	400	0.2491	0.3341	0.3298	0.6732	0.2998
No log	9.0	450	0.2489	0.3345	0.3213	0.6763	0.2949
0.0385	10.0	500	0.2487	0.3349	0.3173	0.6780	0.2876
0.0385	11.0	550	0.2570	0.3346	0.3264	0.6754	0.2971
0.0385	12.0	600	0.2548	0.3348	0.3234	0.6746	0.2946
0.0385	13.0	650	0.2533	0.3349	0.3219	0.6806	0.2942
0.0385	14.0	700	0.2523	0.3350	0.3198	0.6801	0.2919
0.0385	15.0	750	0.2528	0.3350	0.3202	0.6791	0.2925

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1