learning_rate: 5e-05
train_batch_size: 128
eval_batch_size: 128
seed: 42
gradient_accumulation_steps: 8
total_train_batch_size: 1024
optimizer: Use OptimizerNames.ADAMW_HF with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 200
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
4.3759	0.5735	100	4.2809
3.8527	1.1434	200	3.7609
3.3289	1.7168	300	3.2264
2.9251	2.2867	400	2.8520
2.7058	2.8602	500	2.6128
2.5376	3.4301	600	2.4893
2.3017	4.0	700	2.3807
2.3375	4.5735	800	2.3030
2.2953	5.1434	900	2.2198
2.2365	5.7168	1000	2.1678
2.1709	6.2867	1100	2.1113
2.0904	6.8602	1200	2.0296
2.0443	7.4301	1300	1.9785
1.8951	8.0	1400	1.9323
1.9251	8.5735	1500	1.8842
1.8639	9.1434	1600	1.8533
1.8359	9.7168	1700	1.8028
1.8146	10.2867	1800	1.7665
1.8021	10.8602	1900	1.7489
1.7819	11.4301	2000	1.7204
1.6114	12.0	2100	1.6807
1.7171	12.5735	2200	1.6776
1.6929	13.1434	2300	1.6672
1.6784	13.7168	2400	1.6448
1.6724	14.2867	2500	1.5998
1.6372	14.8602	2600	1.5929
1.6156	15.4301	2700	1.5809
1.5284	16.0	2800	1.5574
1.5787	16.5735	2900	1.5392
1.5807	17.1434	3000	1.5195
1.5732	17.7168	3100	1.5140
1.5191	18.2867	3200	1.5144
1.5057	18.8602	3300	1.4768
1.5029	19.4301	3400	1.4749
1.3981	20.0	3500	1.4581
1.4977	20.5735	3600	1.4443
1.4729	21.1434	3700	1.4306
1.4637	21.7168	3800	1.4297
1.4432	22.2867	3900	1.4210
1.4343	22.8602	4000	1.4103
1.4179	23.4301	4100	1.3859
1.3212	24.0	4200	1.4049
1.4155	24.5735	4300	1.3742
1.3983	25.1434	4400	1.3840
1.3998	25.7168	4500	1.3647
1.3565	26.2867	4600	1.3581
1.3805	26.8602	4700	1.3466
1.3731	27.4301	4800	1.3374
1.2775	28.0	4900	1.3320
1.3217	28.5735	5000	1.3215
1.3404	29.1434	5100	1.3185
1.3235	29.7168	5200	1.3119
1.3445	30.2867	5300	1.3268
1.3243	30.8602	5400	1.3003
1.2853	31.4301	5500	1.2758
1.1974	32.0	5600	1.2880
1.2663	32.5735	5700	1.2987
1.2617	33.1434	5800	1.2877
1.2717	33.7168	5900	1.2772
1.2652	34.2867	6000	1.2786
1.2668	34.8602	6100	1.2532
1.2628	35.4301	6200	1.2404
1.1627	36.0	6300	1.2596
1.2476	36.5735	6400	1.2492
1.2457	37.1434	6500	1.2402
1.2393	37.7168	6600	1.2334
1.2378	38.2867	6700	1.2394
1.2272	38.8602	6800	1.2213
1.2204	39.4301	6900	1.2100
1.1453	40.0	7000	1.2134
1.1963	40.5735	7100	1.2253
1.2227	41.1434	7200	1.2188
1.2166	41.7168	7300	1.2008
1.2091	42.2867	7400	1.2055
1.2127	42.8602	7500	1.2055
1.1997	43.4301	7600	1.1971
1.0956	44.0	7700	1.1974
1.1672	44.5735	7800	1.1813
1.1746	45.1434	7900	1.1858
1.1569	45.7168	8000	1.1858
1.1644	46.2867	8100	1.1759
1.1629	46.8602	8200	1.1726
1.1744	47.4301	8300	1.1646
1.0728	48.0	8400	1.1666
1.1436	48.5735	8500	1.1801
1.126	49.1434	8600	1.1663
1.1428	49.7168	8700	1.1510
1.1298	50.2867	8800	1.1669
1.1083	50.8602	8900	1.1516
1.1404	51.4301	9000	1.1462
1.0507	52.0	9100	1.1323
1.1203	52.5735	9200	1.1381
1.1176	53.1434	9300	1.1368
1.1228	53.7168	9400	1.1321
1.1341	54.2867	9500	1.1388
1.1133	54.8602	9600	1.1385
1.0973	55.4301	9700	1.1329
1.0327	56.0	9800	1.1310
1.0939	56.5735	9900	1.1411
1.0856	57.1434	10000	1.1149
1.1024	57.7168	10100	1.1196
1.0865	58.2867	10200	1.1196
1.0842	58.8602	10300	1.1250
1.0786	59.4301	10400	1.1089
0.9986	60.0	10500	1.1002
1.0667	60.5735	10600	1.1080
1.064	61.1434	10700	1.0939
1.0784	61.7168	10800	1.1257
1.0663	62.2867	10900	1.0998
1.0794	62.8602	11000	1.0964
1.0457	63.4301	11100	1.0893
0.9931	64.0	11200	1.1005
1.0574	64.5735	11300	1.0909
1.0619	65.1434	11400	1.0873
1.0669	65.7168	11500	1.0914
1.0273	66.2867	11600	1.0926
1.0467	66.8602	11700	1.0774
1.0418	67.4301	11800	1.0900
0.9703	68.0	11900	1.0799
1.0118	68.5735	12000	1.0846
1.0261	69.1434	12100	1.0652
1.028	69.7168	12200	1.0750
1.0204	70.2867	12300	1.0674
1.0448	70.8602	12400	1.0690
0.9965	71.4301	12500	1.0509
0.9445	72.0	12600	1.0611
1.0125	72.5735	12700	1.0688
0.9895	73.1434	12800	1.0629
1.0051	73.7168	12900	1.0446
0.9876	74.2867	13000	1.0576
0.9867	74.8602	13100	1.0576
0.9902	75.4301	13200	1.0572
0.9248	76.0	13300	1.0636
0.9988	76.5735	13400	1.0623
0.998	77.1434	13500	1.0668
0.9777	77.7168	13600	1.0523
0.9915	78.2867	13700	1.0476
0.9914	78.8602	13800	1.0419
0.992	79.4301	13900	1.0407
0.922	80.0	14000	1.0459
0.9738	80.5735	14100	1.0424
0.9996	81.1434	14200	1.0459
0.9856	81.7168	14300	1.0328
0.9839	82.2867	14400	1.0587
0.9748	82.8602	14500	1.0369
0.9601	83.4301	14600	1.0350
0.8863	84.0	14700	1.0475
0.9639	84.5735	14800	1.0441
0.9671	85.1434	14900	1.0326
0.9515	85.7168	15000	1.0256
0.9309	86.2867	15100	1.0351
0.9479	86.8602	15200	1.0349
0.9738	87.4301	15300	1.0221
0.8831	88.0	15400	1.0311
0.9384	88.5735	15500	1.0231
0.9404	89.1434	15600	1.0369
0.9354	89.7168	15700	1.0220
0.9228	90.2867	15800	1.0354
0.9312	90.8602	15900	1.0232
0.9466	91.4301	16000	1.0174
0.8721	92.0	16100	1.0168
0.9328	92.5735	16200	1.0216
0.9331	93.1434	16300	1.0145
0.9435	93.7168	16400	1.0208
0.9253	94.2867	16500	1.0163
0.9231	94.8602	16600	1.0159
0.9273	95.4301	16700	1.0044
0.8571	96.0	16800	1.0047
0.9293	96.5735	16900	1.0111
0.9264	97.1434	17000	1.0088
0.9221	97.7168	17100	1.0074
0.9143	98.2867	17200	1.0123
0.9119	98.8602	17300	1.0049
0.9036	99.4301	17400	0.9973
0.869	100.0	17500	1.0031
0.9024	100.5735	17600	1.0054
0.896	101.1434	17700	0.9966
0.8883	101.7168	17800	1.0010
0.9053	102.2867	17900	1.0160
0.9263	102.8602	18000	1.0012
0.9042	103.4301	18100	1.0145
0.8341	104.0	18200	0.9961
0.9028	104.5735	18300	1.0040
0.9026	105.1434	18400	0.9844
0.8801	105.7168	18500	0.9907
0.8956	106.2867	18600	0.9983
0.9023	106.8602	18700	0.9840
0.8712	107.4301	18800	0.9946
0.8374	108.0	18900	0.9944
0.9102	108.5735	19000	0.9928
0.8914	109.1434	19100	0.9778
0.8852	109.7168	19200	0.9928
0.8826	110.2867	19300	0.9829
0.881	110.8602	19400	0.9747
0.878	111.4301	19500	0.9809
0.8219	112.0	19600	0.9707
0.8797	112.5735	19700	0.9727
0.8911	113.1434	19800	0.9759
0.8706	113.7168	19900	0.9809
0.8952	114.2867	20000	0.9871
0.8585	114.8602	20100	0.9777
0.8679	115.4301	20200	0.9745
0.8289	116.0	20300	0.9793
0.8647	116.5735	20400	0.9791
0.8711	117.1434	20500	0.9806
0.8734	117.7168	20600	0.9831
0.8427	118.2867	20700	0.9862
0.8678	118.8602	20800	0.9786
0.8687	119.4301	20900	0.9738
0.8072	120.0	21000	0.9758
0.854	120.5735	21100	0.9647
0.8679	121.1434	21200	0.9648
0.8652	121.7168	21300	0.9747
0.8621	122.2867	21400	0.9818
0.8446	122.8602	21500	0.9861
0.8618	123.4301	21600	0.9597
0.8102	124.0	21700	0.9710
0.8605	124.5735	21800	0.9672
0.845	125.1434	21900	0.9760
0.8452	125.7168	22000	0.9756
0.848	126.2867	22100	0.9624
0.8522	126.8602	22200	0.9559
0.8481	127.4301	22300	0.9682
0.7941	128.0	22400	0.9693
0.8381	128.5735	22500	0.9690
0.8498	129.1434	22600	0.9656
0.8558	129.7168	22700	0.9619
0.8241	130.2867	22800	0.9568
0.8422	130.8602	22900	0.9662
0.8446	131.4301	23000	0.9598
0.7789	132.0	23100	0.9545
0.8287	132.5735	23200	0.9655
0.8413	133.1434	23300	0.9517
0.8344	133.7168	23400	0.9638
0.8287	134.2867	23500	0.9715
0.8253	134.8602	23600	0.9619
0.8232	135.4301	23700	0.9617
0.7756	136.0	23800	0.9559
0.8306	136.5735	23900	0.9612
0.8383	137.1434	24000	0.9481
0.8176	137.7168	24100	0.9537
0.8292	138.2867	24200	0.9640
0.8332	138.8602	24300	0.9477
0.8296	139.4301	24400	0.9636
0.7812	140.0	24500	0.9598
0.8303	140.5735	24600	0.9559
0.8207	141.1434	24700	0.9397
0.8163	141.7168	24800	0.9461
0.8177	142.2867	24900	0.9570
0.8185	142.8602	25000	0.9555
0.821	143.4301	25100	0.9620
0.7833	144.0	25200	0.9386
0.8218	144.5735	25300	0.9523
0.8156	145.1434	25400	0.9508
0.8205	145.7168	25500	0.9559
0.8204	146.2867	25600	0.9569
0.8113	146.8602	25700	0.9489
0.8093	147.4301	25800	0.9483
0.7531	148.0	25900	0.9512
0.8023	148.5735	26000	0.9543
0.8088	149.1434	26100	0.9640
0.8054	149.7168	26200	0.9508
0.8164	150.2867	26300	0.9372
0.8307	150.8602	26400	0.9495
0.8084	151.4301	26500	0.9408
0.7723	152.0	26600	0.9561
0.7929	152.5735	26700	0.9448
0.7836	153.1434	26800	0.9495
0.8016	153.7168	26900	0.9471
0.8073	154.2867	27000	0.9490
0.811	154.8602	27100	0.9445
0.808	155.4301	27200	0.9497
0.7446	156.0	27300	0.9624
0.8005	156.5735	27400	0.9527
0.8003	157.1434	27500	0.9406
0.7985	157.7168	27600	0.9430
0.8044	158.2867	27700	0.9457
0.817	158.8602	27800	0.9419
0.8118	159.4301	27900	0.9360
0.7421	160.0	28000	0.9401
0.8	160.5735	28100	0.9437
0.7981	161.1434	28200	0.9223
0.7917	161.7168	28300	0.9423
0.8029	162.2867	28400	0.9357
0.7934	162.8602	28500	0.9504
0.7907	163.4301	28600	0.9382
0.7451	164.0	28700	0.9348
0.7871	164.5735	28800	0.9435
0.7918	165.1434	28900	0.9399
0.8033	165.7168	29000	0.9382
0.7913	166.2867	29100	0.9279
0.7902	166.8602	29200	0.9461
0.781	167.4301	29300	0.9376
0.7472	168.0	29400	0.9337
0.7756	168.5735	29500	0.9419
0.7944	169.1434	29600	0.9313
0.7868	169.7168	29700	0.9308
0.7914	170.2867	29800	0.9339
0.7904	170.8602	29900	0.9410
0.7923	171.4301	30000	0.9442
0.7428	172.0	30100	0.9383
0.7822	172.5735	30200	0.9369
0.7956	173.1434	30300	0.9367
0.7888	173.7168	30400	0.9332
0.79	174.2867	30500	0.9264
0.7922	174.8602	30600	0.9313
0.7817	175.4301	30700	0.9363
0.7282	176.0	30800	0.9198
0.7819	176.5735	30900	0.9448
0.7692	177.1434	31000	0.9359
0.7725	177.7168	31100	0.9350
0.7819	178.2867	31200	0.9376
0.7951	178.8602	31300	0.9235
0.7793	179.4301	31400	0.9309
0.7439	180.0	31500	0.9385
0.7843	180.5735	31600	0.9364
0.781	181.1434	31700	0.9339
0.7859	181.7168	31800	0.9422
0.7684	182.2867	31900	0.9404
0.7722	182.8602	32000	0.9209
0.7799	183.4301	32100	0.9293
0.7344	184.0	32200	0.9242
0.7847	184.5735	32300	0.9372
0.7825	185.1434	32400	0.9271
0.7733	185.7168	32500	0.9302
0.7603	186.2867	32600	0.9334
0.7825	186.8602	32700	0.9228
0.7806	187.4301	32800	0.9348
0.7184	188.0	32900	0.9262
0.7866	188.5735	33000	0.9271
0.7649	189.1434	33100	0.9236
0.7825	189.7168	33200	0.9378
0.7788	190.2867	33300	0.9328
0.7771	190.8602	33400	0.9311
0.7771	191.4301	33500	0.9356
0.722	192.0	33600	0.9329
0.7683	192.5735	33700	0.9266
0.7724	193.1434	33800	0.9198
0.7491	193.7168	33900	0.9304
0.7591	194.2867	34000	0.9245
0.7609	194.8602	34100	0.9380
0.7672	195.4301	34200	0.9298
0.7194	196.0	34300	0.9120
0.7618	196.5735	34400	0.9269
0.7847	197.1434	34500	0.9285
0.7746	197.7168	34600	0.9113
0.7885	198.2867	34700	0.9460
0.7768	198.8602	34800	0.9382

Framework versions

Transformers 4.49.0
Pytorch 2.6.0+cu124
Datasets 3.5.0
Tokenizers 0.21.1

Sphincz/monolang_roberta

monolang_roberta

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Sphincz/
monolang_roberta