xlm-roberta-large-bs-16-lr-0.0001-ep-1-wp-0.1-gacc-8-gnm-1.0-FP16-mx-512-v0.1

This model is a fine-tuned version of FacebookAI/xlm-roberta-large on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 2.2438

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: 0.0001
train_batch_size: 16
eval_batch_size: 16
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 128
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_ratio: 0.1
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
17.501	0.0055	50	4.8573
16.4238	0.0109	100	4.2333
15.0223	0.0164	150	4.1599
14.6734	0.0219	200	4.0074
14.8891	0.0273	250	nan
14.0058	0.0328	300	3.5820
13.7471	0.0382	350	3.4834
14.0411	0.0437	400	3.4724
13.7614	0.0492	450	3.4450
13.728	0.0546	500	3.3631
13.6001	0.0601	550	3.3878
12.943	0.0656	600	3.3878
14.0021	0.0710	650	3.1696
13.4041	0.0765	700	3.2144
13.2302	0.0819	750	3.1456
13.3945	0.0874	800	3.1081
13.3763	0.0929	850	3.0475
13.3499	0.0983	900	3.2461
13.5559	0.1038	950	3.0163
13.839	0.1093	1000	3.0701
13.3534	0.1147	1050	2.9885
13.2552	0.1202	1100	3.0023
13.6676	0.1256	1150	nan
13.1216	0.1311	1200	3.0053
12.6853	0.1366	1250	2.8969
12.9434	0.1420	1300	2.9016
12.2164	0.1475	1350	2.8974
12.825	0.1530	1400	2.9705
12.7314	0.1584	1450	2.8804
12.7405	0.1639	1500	2.8514
12.5693	0.1694	1550	2.8858
12.2698	0.1748	1600	2.8437
12.19	0.1803	1650	2.9199
12.2267	0.1857	1700	2.7915
12.1787	0.1912	1750	2.9066
12.1286	0.1967	1800	2.8383
12.3344	0.2021	1850	nan
13.0251	0.2076	1900	2.8345
12.4427	0.2131	1950	2.7413
12.6127	0.2185	2000	2.7285
12.6358	0.2240	2050	2.7807
12.2132	0.2294	2100	2.7657
12.5298	0.2349	2150	2.7935
12.156	0.2404	2200	2.6942
12.2265	0.2458	2250	2.7374
12.0772	0.2513	2300	2.6400
11.7906	0.2568	2350	2.6862
11.5912	0.2622	2400	2.6664
12.242	0.2677	2450	2.7530
11.3089	0.2731	2500	2.7606
11.2301	0.2786	2550	2.6787
11.9706	0.2841	2600	2.7440
11.5268	0.2895	2650	2.6760
11.8031	0.2950	2700	2.6846
11.6836	0.3005	2750	nan
11.4748	0.3059	2800	2.6796
11.9102	0.3114	2850	2.7101
11.4223	0.3169	2900	2.7066
12.0939	0.3223	2950	2.5908
11.5229	0.3278	3000	nan
10.8909	0.3332	3050	2.5104
11.2679	0.3387	3100	2.6391
11.6102	0.3442	3150	2.6375
11.1783	0.3496	3200	2.5392
11.5862	0.3551	3250	2.6254
11.0802	0.3606	3300	2.4951
11.2194	0.3660	3350	2.5535
10.8891	0.3715	3400	2.4888
11.1372	0.3769	3450	2.6514
11.1702	0.3824	3500	nan
11.1283	0.3879	3550	2.4935
11.858	0.3933	3600	2.6377
10.6952	0.3988	3650	2.5486
11.1094	0.4043	3700	2.5827
10.5929	0.4097	3750	2.5155
10.9796	0.4152	3800	2.6333
11.4408	0.4207	3850	2.4885
11.3756	0.4261	3900	2.6248
10.6489	0.4316	3950	2.5080
11.2278	0.4370	4000	2.6829
10.9081	0.4425	4050	nan
10.3177	0.4480	4100	2.5467
11.1393	0.4534	4150	2.4981
11.109	0.4589	4200	2.5696
10.5874	0.4644	4250	2.5346
10.2922	0.4698	4300	2.5247
11.1379	0.4753	4350	2.5050
10.9258	0.4807	4400	2.4393
10.7622	0.4862	4450	2.5386
10.5537	0.4917	4500	2.4742
10.6157	0.4971	4550	2.5183
10.5721	0.5026	4600	2.4624
10.448	0.5081	4650	nan
10.9621	0.5135	4700	2.4363
10.5947	0.5190	4750	2.4489
10.4982	0.5244	4800	nan
10.241	0.5299	4850	2.4834
10.8498	0.5354	4900	nan
10.291	0.5408	4950	2.4880
10.032	0.5463	5000	2.4780
10.6992	0.5518	5050	2.4536
10.3189	0.5572	5100	2.5406
10.36	0.5627	5150	2.5421
10.1413	0.5682	5200	2.5299
10.4146	0.5736	5250	2.4525
10.0561	0.5791	5300	2.5126
10.3447	0.5845	5350	2.4347
10.2634	0.5900	5400	2.3891
10.067	0.5955	5450	2.4418
10.479	0.6009	5500	2.4801
9.8486	0.6064	5550	2.4651
10.2608	0.6119	5600	2.3497
10.0271	0.6173	5650	2.5478
9.8674	0.6228	5700	2.3528
10.1599	0.6282	5750	2.4087
9.9866	0.6337	5800	2.3972
10.5326	0.6392	5850	2.4910
10.2033	0.6446	5900	2.3823
9.8695	0.6501	5950	2.3799
10.0466	0.6556	6000	2.4245
9.5177	0.6610	6050	2.4596
10.4291	0.6665	6100	2.4178
10.0009	0.6719	6150	2.3328
10.0692	0.6774	6200	2.3533
9.6967	0.6829	6250	2.4248
9.9892	0.6883	6300	2.3493
10.1783	0.6938	6350	2.3389
10.019	0.6993	6400	2.4507
9.8618	0.7047	6450	2.2831
10.3984	0.7102	6500	2.3761
9.919	0.7157	6550	2.5036
9.2917	0.7211	6600	2.3926
9.6774	0.7266	6650	2.3494
10.0028	0.7320	6700	2.3653
9.6192	0.7375	6750	2.3574
9.9689	0.7430	6800	2.4544
10.0934	0.7484	6850	2.4070
10.0145	0.7539	6900	2.3699
9.559	0.7594	6950	nan
10.5713	0.7648	7000	2.3410
9.7507	0.7703	7050	nan
9.9102	0.7757	7100	2.4138
9.4241	0.7812	7150	2.2941
9.6202	0.7867	7200	2.3024
9.5112	0.7921	7250	2.3756
9.4726	0.7976	7300	2.3240
9.5841	0.8031	7350	2.4397
9.1056	0.8085	7400	nan
9.0733	0.8140	7450	2.3982
9.9461	0.8194	7500	2.3694
9.1871	0.8249	7550	2.3681
9.723	0.8304	7600	2.3977
9.7697	0.8358	7650	2.4167
9.2425	0.8413	7700	2.2994
9.5511	0.8468	7750	2.3465
9.8158	0.8522	7800	2.3081
9.4219	0.8577	7850	2.2640
9.4233	0.8632	7900	2.3290
9.3864	0.8686	7950	2.2964
9.4981	0.8741	8000	2.2984
9.1101	0.8795	8050	2.3284
9.1299	0.8850	8100	2.3426
8.9554	0.8905	8150	2.3206
9.5779	0.8959	8200	2.2987
9.1416	0.9014	8250	2.3276
9.4434	0.9069	8300	2.2201
9.1004	0.9123	8350	2.2855
9.3678	0.9178	8400	2.3188
9.2545	0.9232	8450	2.3988
9.3835	0.9287	8500	2.2233
9.7359	0.9342	8550	2.2780
9.2803	0.9396	8600	2.3142
8.9966	0.9451	8650	2.2083
9.2548	0.9506	8700	2.4125
10.0036	0.9560	8750	2.1931
9.4264	0.9615	8800	2.1629
9.102	0.9669	8850	2.3306
9.3087	0.9724	8900	2.2894
8.9155	0.9779	8950	2.2347
9.1586	0.9833	9000	2.3156
9.2523	0.9888	9050	nan
9.541	0.9943	9100	2.2957
9.4701	0.9997	9150	2.2438

Framework versions

Transformers 4.47.1
Pytorch 2.5.1+cu124
Datasets 3.1.0
Tokenizers 0.21.0

BounharAbdelaziz/XLM-RoBERTa-Morocco

xlm-roberta-large-bs-16-lr-0.0001-ep-1-wp-0.1-gacc-8-gnm-1.0-FP16-mx-512-v0.1

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions