simple_fin_run_bs16_lr5e-06_layers2_reg0

This model is a fine-tuned version of Zamza/XLM-roberta-large-ftit-emb-lr01 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.5237
• Precision Type: 0.8765
• Recall Type: 0.6676
• F1 Type: 0.7423
• Accuracy Type: 0.6676
• Precision Class: 0.8504
• Recall Class: 0.4460
• F1 Class: 0.5573
• Accuracy Class: 0.4460
• Precision Rel: 0.9641
• Recall Rel: 0.5818
• F1 Rel: 0.7157
• Accuracy Rel: 0.5818

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-06
• train_batch_size: 16
• eval_batch_size: 16
• 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: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision Type	Recall Type	F1 Type	Accuracy Type	Precision Class	Recall Class	F1 Class	Accuracy Class	Precision Rel	Recall Rel	F1 Rel	Accuracy Rel
9.243	0.2181	1000	9.0652	0.8067	0.6465	0.7026	0.6465	0.7449	0.1339	0.1822	0.1339	0.9567	0.1610	0.2419	0.1610
8.5682	0.4362	2000	8.0975	0.8497	0.6325	0.7076	0.6325	0.7865	0.4520	0.5443	0.4520	0.9595	0.4999	0.6422	0.4999
7.7782	0.6543	3000	7.5428	0.8565	0.6538	0.7250	0.6538	0.8021	0.4995	0.5869	0.4995	0.9626	0.4858	0.6319	0.4858
7.5106	0.8724	4000	7.1057	0.8638	0.6694	0.7398	0.6694	0.8252	0.4818	0.5793	0.4818	0.9614	0.5721	0.7059	0.5721
6.8248	1.0905	5000	6.8260	0.8632	0.6121	0.6939	0.6121	0.8286	0.4536	0.5539	0.4536	0.9626	0.4806	0.6265	0.4806
6.4373	1.3086	6000	6.5821	0.8712	0.6362	0.7178	0.6362	0.8330	0.4558	0.5586	0.4558	0.9623	0.5714	0.7050	0.5714
6.158	1.5267	7000	6.4428	0.8674	0.6563	0.7312	0.6563	0.8349	0.5228	0.6186	0.5228	0.9617	0.5857	0.7161	0.5857
6.4975	1.7448	8000	6.2574	0.8697	0.6594	0.7347	0.6594	0.8342	0.4824	0.5846	0.4824	0.9632	0.5956	0.7258	0.5956
6.4181	1.9629	9000	6.1495	0.8693	0.6626	0.7354	0.6626	0.8400	0.4272	0.5366	0.4272	0.9626	0.5700	0.7039	0.5700
5.8706	2.1810	10000	6.0631	0.8693	0.6620	0.7360	0.6620	0.8391	0.4872	0.5916	0.4872	0.9635	0.5384	0.6783	0.5384
5.8946	2.3991	11000	5.9728	0.8735	0.6322	0.7144	0.6322	0.8425	0.4496	0.5585	0.4496	0.9636	0.5112	0.6549	0.5112
5.7131	2.6172	12000	5.9089	0.8689	0.6869	0.7539	0.6869	0.8426	0.4598	0.5686	0.4598	0.9644	0.5904	0.7223	0.5904
5.6304	2.8353	13000	5.8487	0.8746	0.6338	0.7165	0.6338	0.8429	0.4437	0.5531	0.4437	0.9642	0.5558	0.6941	0.5558
5.6237	3.0534	14000	5.8082	0.8738	0.6335	0.7153	0.6335	0.8462	0.4237	0.5342	0.4237	0.9642	0.5396	0.6799	0.5396
5.638	3.2715	15000	5.7676	0.8686	0.6687	0.7391	0.6687	0.8459	0.4373	0.5475	0.4373	0.9644	0.5699	0.7061	0.5699
5.7302	3.4896	16000	5.7459	0.8699	0.6623	0.7349	0.6623	0.8471	0.4462	0.5568	0.4462	0.9644	0.5571	0.6963	0.5571
5.2957	3.7077	17000	5.7062	0.8669	0.6648	0.7347	0.6648	0.8453	0.4521	0.5618	0.4521	0.9633	0.5793	0.7126	0.5793
5.7661	3.9258	18000	5.6539	0.8766	0.6546	0.7334	0.6546	0.8478	0.4228	0.5350	0.4228	0.9635	0.5494	0.6881	0.5494
5.2341	4.1439	19000	5.6492	0.8715	0.6713	0.7434	0.6713	0.8477	0.4383	0.5483	0.4383	0.9635	0.5775	0.7120	0.5775
5.3054	4.3621	20000	5.6480	0.8763	0.6781	0.7504	0.6781	0.8473	0.4628	0.5731	0.4628	0.9643	0.5563	0.6951	0.5563
5.5869	4.5802	21000	5.6087	0.8748	0.6468	0.7256	0.6468	0.8489	0.4395	0.5482	0.4395	0.9638	0.5463	0.6859	0.5463
5.3553	4.7983	22000	5.5900	0.8715	0.6676	0.7402	0.6676	0.8483	0.4371	0.5478	0.4371	0.9647	0.5641	0.7020	0.5641
5.0912	5.0164	23000	5.6066	0.8761	0.6458	0.7258	0.6458	0.8482	0.4421	0.5542	0.4421	0.9639	0.5429	0.6832	0.5429
5.1303	5.2345	24000	5.5864	0.8707	0.6736	0.7437	0.6736	0.8484	0.4492	0.5602	0.4492	0.9638	0.5816	0.7152	0.5816
5.3516	5.4526	25000	5.5852	0.8765	0.6690	0.7438	0.6690	0.8470	0.4518	0.5618	0.4518	0.9643	0.5860	0.7189	0.5860
4.9028	5.6707	26000	5.5735	0.8712	0.6718	0.7425	0.6718	0.8464	0.4414	0.5509	0.4414	0.9639	0.5784	0.7135	0.5784
4.9972	5.8888	27000	5.5550	0.8718	0.6724	0.7436	0.6724	0.8470	0.4458	0.5560	0.4458	0.9633	0.5913	0.7227	0.5913
4.9198	6.1069	28000	5.5740	0.8756	0.6605	0.7365	0.6605	0.8474	0.4448	0.5554	0.4448	0.9638	0.5796	0.7135	0.5796
4.7808	6.3250	29000	5.5434	0.8753	0.6628	0.7383	0.6628	0.8471	0.4360	0.5466	0.4360	0.9635	0.5824	0.7162	0.5824
4.9946	6.5431	30000	5.5566	0.8768	0.6694	0.7438	0.6694	0.8498	0.4451	0.5570	0.4451	0.9639	0.5758	0.7109	0.5758
5.0043	6.7612	31000	5.5347	0.8760	0.6745	0.7476	0.6745	0.8500	0.4420	0.5542	0.4420	0.9643	0.5666	0.7035	0.5666
5.2214	6.9793	32000	5.5465	0.8765	0.6765	0.7489	0.6765	0.8498	0.4507	0.5623	0.4507	0.9642	0.5666	0.7033	0.5666
4.9683	7.1974	33000	5.5190	0.8775	0.6734	0.7474	0.6734	0.8497	0.4625	0.5729	0.4625	0.9645	0.5783	0.7133	0.5783
4.854	7.4155	34000	5.5495	0.8757	0.6712	0.7448	0.6712	0.8494	0.4515	0.5628	0.4515	0.9649	0.5718	0.7082	0.5718
4.8145	7.6336	35000	5.5441	0.8753	0.6789	0.7500	0.6789	0.8505	0.4486	0.5601	0.4486	0.9638	0.5880	0.7204	0.5880
5.0554	7.8517	36000	5.5305	0.8740	0.6709	0.7435	0.6709	0.8486	0.4473	0.5577	0.4473	0.9641	0.5848	0.7183	0.5848
4.4384	8.0698	37000	5.5391	0.8760	0.6826	0.7532	0.6826	0.8497	0.4592	0.5697	0.4592	0.9639	0.5869	0.7197	0.5869
4.786	8.2879	38000	5.5325	0.8750	0.6822	0.7522	0.6822	0.8487	0.4576	0.5676	0.4576	0.9640	0.5949	0.7264	0.5949
4.6294	8.5060	39000	5.5384	0.8763	0.6692	0.7434	0.6692	0.8493	0.4511	0.5622	0.4511	0.9641	0.5848	0.7180	0.5848
4.7249	8.7241	40000	5.5392	0.8768	0.6671	0.7422	0.6671	0.8507	0.4437	0.5550	0.4437	0.9644	0.5785	0.7134	0.5785
4.6873	8.9422	41000	5.5245	0.8764	0.6638	0.7393	0.6638	0.8511	0.4374	0.5495	0.4374	0.9643	0.5669	0.7040	0.5669
4.8909	9.1603	42000	5.5267	0.8768	0.6660	0.7413	0.6660	0.8507	0.4383	0.5499	0.4383	0.9644	0.5785	0.7133	0.5785
4.5894	9.3784	43000	5.5281	0.8758	0.6766	0.7484	0.6766	0.8506	0.4546	0.5657	0.4546	0.9640	0.5879	0.7204	0.5879
4.7697	9.5965	44000	5.5261	0.8764	0.6684	0.7427	0.6684	0.8505	0.4472	0.5587	0.4472	0.9640	0.5809	0.7147	0.5809
4.6071	9.8146	45000	5.5237	0.8765	0.6676	0.7423	0.6676	0.8504	0.4460	0.5573	0.4460	0.9641	0.5818	0.7157	0.5818

Framework versions

• Transformers 4.49.0
• Pytorch 2.6.0+cu124
• Datasets 3.4.1
• Tokenizers 0.21.1