math_question_grade_detection_v12

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

• Loss: 0.8913
• Accuracy: 0.7397
• Precision: 0.7255
• Recall: 0.7397
• F1: 0.7178

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• training_steps: 1500

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
No log	0.0855	50	3.0343	0.1710	0.1287	0.1710	0.0917
No log	0.1709	100	2.4792	0.3151	0.2979	0.3151	0.2439
No log	0.2564	150	2.1397	0.4428	0.4101	0.4428	0.3752
No log	0.3419	200	1.9824	0.4419	0.3872	0.4419	0.3638
No log	0.4274	250	1.7765	0.5072	0.4772	0.5072	0.4666
No log	0.5128	300	1.6510	0.5524	0.5347	0.5524	0.5147
No log	0.5983	350	1.5160	0.5793	0.5501	0.5793	0.5319
No log	0.6838	400	1.4481	0.5898	0.5608	0.5898	0.5437
No log	0.7692	450	1.3791	0.6148	0.5758	0.6148	0.5678
1.9748	0.8547	500	1.3154	0.6196	0.6123	0.6196	0.5779
1.9748	0.9402	550	1.2399	0.6484	0.6168	0.6484	0.6119
1.9748	1.0256	600	1.1968	0.6340	0.6181	0.6340	0.5889
1.9748	1.1111	650	1.2477	0.6215	0.6014	0.6215	0.5825
1.9748	1.1966	700	1.2098	0.6340	0.6285	0.6340	0.5884
1.9748	1.2821	750	1.1316	0.6619	0.6442	0.6619	0.6385
1.9748	1.3675	800	1.0783	0.6744	0.6644	0.6744	0.6462
1.9748	1.4530	850	1.0512	0.6907	0.6728	0.6907	0.6583
1.9748	1.5385	900	1.0388	0.6945	0.6909	0.6945	0.6697
1.9748	1.6239	950	0.9954	0.6974	0.6748	0.6974	0.6707
1.0265	1.7094	1000	0.9812	0.7128	0.6888	0.7128	0.6874
1.0265	1.7949	1050	0.9717	0.7099	0.6907	0.7099	0.6852
1.0265	1.8803	1100	0.9437	0.7099	0.6823	0.7099	0.6866
1.0265	1.9658	1150	0.9724	0.7061	0.7096	0.7061	0.6800
1.0265	2.0513	1200	0.9168	0.7224	0.7099	0.7224	0.6976
1.0265	2.1368	1250	0.9097	0.7243	0.7109	0.7243	0.6996
1.0265	2.2222	1300	0.9072	0.7329	0.7336	0.7329	0.7083
1.0265	2.3077	1350	0.9028	0.7262	0.7114	0.7262	0.7033
1.0265	2.3932	1400	0.8951	0.7301	0.7145	0.7301	0.7068
1.0265	2.4786	1450	0.8949	0.7378	0.7339	0.7378	0.7154
0.687	2.5641	1500	0.8913	0.7397	0.7255	0.7397	0.7178

Framework versions

• Transformers 4.46.3
• Pytorch 2.4.0
• Datasets 3.1.0
• Tokenizers 0.20.3