categorization-finetuned-20220721-164940-distilled-20220810-123313

This model is a fine-tuned version of carted-nlp/categorization-finetuned-20220721-164940 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0787
• Accuracy: 0.8416
• F1: 0.8396

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: 7e-06
• train_batch_size: 64
• eval_batch_size: 64
• seed: 314
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1500
• num_epochs: 15.0

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
0.2976	0.56	2500	0.1441	0.7219	0.7071
0.1417	1.12	5000	0.1180	0.7719	0.7653
0.1236	1.69	7500	0.1076	0.7901	0.7854
0.1148	2.25	10000	0.1014	0.8015	0.7977
0.1092	2.81	12500	0.0972	0.8089	0.8052
0.1043	3.37	15000	0.0942	0.8135	0.8102
0.1013	3.94	17500	0.0916	0.8181	0.8147
0.0985	4.5	20000	0.0897	0.8219	0.8190
0.0962	5.06	22500	0.0881	0.8241	0.8215
0.0945	5.62	25000	0.0866	0.8270	0.8246
0.0928	6.19	27500	0.0857	0.8286	0.8262
0.0912	6.75	30000	0.0843	0.8310	0.8286
0.0901	7.31	32500	0.0836	0.8321	0.8299
0.0887	7.87	35000	0.0827	0.8339	0.8315
0.0879	8.43	37500	0.0821	0.8350	0.8329
0.0875	9.0	40000	0.0814	0.8362	0.8342
0.0865	9.56	42500	0.0811	0.8370	0.8348
0.0855	10.12	45000	0.0806	0.8375	0.8355
0.0853	10.68	47500	0.0798	0.8386	0.8367
0.0845	11.25	50000	0.0799	0.8392	0.8372
0.0844	11.81	52500	0.0793	0.8401	0.8383
0.0838	12.37	55000	0.0793	0.8402	0.8381
0.0834	12.93	57500	0.0790	0.8410	0.8390
0.0832	13.5	60000	0.0788	0.8414	0.8394
0.083	14.06	62500	0.0787	0.8415	0.8395
0.0828	14.62	65000	0.0787	0.8416	0.8396

Framework versions

• Transformers 4.17.0
• Pytorch 1.11.0+cu113
• Datasets 2.3.2
• Tokenizers 0.11.6