alankar-classifier-v1

This model is a fine-tuned version of ai4bharat/IndicBERTv2-MLM-only on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.2309
• F1 Micro: 0.3
• Roc Auc Micro: 0.8965

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: 3e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	F1 Micro	Roc Auc Micro
No log	1.0	9	0.5472	0.0	0.4839
No log	2.0	18	0.3904	0.0	0.4565
No log	3.0	27	0.3580	0.0	0.4888
No log	4.0	36	0.3461	0.0	0.5664
No log	5.0	45	0.3480	0.0	0.5303
No log	6.0	54	0.3479	0.0	0.5542
No log	7.0	63	0.3390	0.0	0.6699
No log	8.0	72	0.3341	0.0	0.7271
No log	9.0	81	0.3236	0.0	0.7471
No log	10.0	90	0.3166	0.0	0.7808
No log	11.0	99	0.3066	0.0	0.8193
No log	12.0	108	0.2969	0.0	0.8252
No log	13.0	117	0.2921	0.0	0.8428
No log	14.0	126	0.2883	0.0	0.8462
No log	15.0	135	0.2798	0.0	0.8521
No log	16.0	144	0.2732	0.1111	0.8555
No log	17.0	153	0.2732	0.1111	0.8647
No log	18.0	162	0.2678	0.1111	0.8662
No log	19.0	171	0.2697	0.1111	0.8691
No log	20.0	180	0.2629	0.1111	0.8818
No log	21.0	189	0.2560	0.2105	0.8833
No log	22.0	198	0.2478	0.2105	0.8848
No log	23.0	207	0.2515	0.2105	0.8843
No log	24.0	216	0.2458	0.3	0.8867
No log	25.0	225	0.2423	0.2105	0.8857
No log	26.0	234	0.2418	0.3	0.8853
No log	27.0	243	0.2402	0.3	0.8955
No log	28.0	252	0.2410	0.3	0.8916
No log	29.0	261	0.2352	0.3	0.8940
No log	30.0	270	0.2331	0.3	0.8940
No log	31.0	279	0.2312	0.3	0.8970
No log	32.0	288	0.2338	0.3	0.8979
No log	33.0	297	0.2338	0.3	0.8975
No log	34.0	306	0.2327	0.3	0.8945
No log	35.0	315	0.2309	0.3	0.8950
No log	36.0	324	0.2309	0.3	0.8970
No log	37.0	333	0.2304	0.3	0.8975
No log	38.0	342	0.2308	0.3	0.8965
No log	39.0	351	0.2309	0.3	0.8965
No log	40.0	360	0.2309	0.3	0.8965

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 2.14.4
• Tokenizers 0.21.1