This model is a fine-tuned version of nlpaueb/bert-base-greek-uncased-v1 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.3369

Model description

In this work we appropriately adapt a corpus of multiword expressions in Modern Greek, namely PolylexMG, characterised by the features detailed above formulating its spectrum of idiosyncrasy to finetune Greek BERT transformer model for masked language modelling classification and tasks. The GREEK-BERT model is pre-trained on free text corpora extracted from (a) the Greek part of Wikipedia, (b) the Greek part of the European Parliament Proceedings Parallel Corpus (Europarl), and (c) the Greek part of OSCAR (Koutsikakis et al, 2020:113), this monolingual model is based on the architecture of BERT-BASE-UNCASED. Specifically, Greek BERT has been finetuned with expressions derived from each syntactic category as they are described in PolylexMG (Fotopoulou et al., 2023) that includes 6,000 Greek lexical entries dataset entailing frozen idioms which are semantically fixed with no paradigmatic variation (Lamiroy, 2003) and light verb constructions in which the semantics are traced in the predicative noun and not the verb (Anastassiadis-Symeonidis et al., 2020).

Results, intended uses & limitations

This subsection presents the experimental evaluation results for the MWE-fine-tuned Greek BERT model with respect to classification use case. The derived setup assumes that raw text in modern Greek that may contain multiple sentences is processed by the language model and reports class with regards to whether the text segment contains multiword stereotypical expressions or not. We compared the fine-tuned BERT model with a baseline logistic regression model. The latter is using as input the same word embeddings as the MWE-fine-tuned BERT model.
Greek-MWE-Bert was trained in a masked language model setting with full-expression-subdataset. The model perplexity was measured tο 303.21 before finetuning and 3.81 after finetuning demonstrating that the model has gained domain knowledge on multiword expressions. Qualitative outcomes presented in the following tables demonstrating the model performance in the case of 16 verbal constructs. The qualitative evaluation demonstrates that the fine-tuned model in all cases generates stereotypical multiword expressions while the original Greek BERT yields incomplete free-text related parts of sentences.

The finetuned model was further finetuned using for classification-oriented architecture with the classification-task-subdataset. The Bert classifier demonstrates an accuracy equivalent to 80% with a higher precision for free text reaching 80% and lower precision of 79% for MWE. In comparison the baseline classifier yields 70% for the free text and 67% for the MWE. We can observe that the two models have only 10% difference in accuracy despite the simplicity of baseline classifier. We can explain the small difference in the trained GreekBERT tokenizer that is used by both our model and the simplistic logistic regression model. However, the MWE-finetuned-Greek-BERT model can better capture sentences that contain MWEs due to the inherent benefits that the architecture offers.

Training and evaluation data

Sample of PolylexMG full expression subdataset

text	label
αδειάζω τη γωνιά σε	1
αδειάζω πιστόλι πάνω σε	1
αλλάζω τον αδόξαστο σε	1
αλλάζω την πίστη σε	1
δεν αλλάζω ούτε κόμα σε	1
αλλάζω λόγια με	1
αλλάζω κουβέντες με	1
αλλάζω τα μυαλά σε	1
αλλάζω τα φώτα σε	1
αλλάζω τα πετρέλαια σε	1
αλλάζω τα πέταλα σε	1

Sample of PolylexMG classification subdataset

text	label
Μέσα σε λίγα λεπτά άναψαν τα αίματα και ο διαπληκτισμός άρχισε να γίνεται όλο και πιο έντονος	1
Η πρώτη έκπληξη ήρθε αμέσως μόλις άναψαν τα τέσσερα κόκκινα φανάρια και το ένα πράσινο	0
Γιατί τα κάνετε αυτά, για να γελάνε οι άλλοι μαζί μας;	0
Κάθε φορά που έμπαινε καλάθι, έβγαζαν τις ίδιες ακριβώς ιαχές για να πάει γούρι και να μην κόψει η μαγιονέζα	1
Η νέα πυρκαγιά ξεκινά από την πίσω πλευρά του Πεντελικού Όρους, σε σημείο που δεν είχε καεί	0

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-06
train_batch_size: 512
eval_batch_size: 512
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 500

Training results (Summary)

Training Loss	Epoch	Step	Validation Loss
5.2105	1.0	13	4.4870
4.4319	2.0	26	3.8456
4.0318	3.0	39	3.4164
3.7558	4.0	52	3.2849
1.1307	497.0	6461	1.3311
1.1163	498.0	6474	1.3016
1.099	499.0	6487	1.3532
1.1246	500.0	6500	1.2222

Framework versions

Transformers 4.31.0
Pytorch 2.0.1+cu118
Datasets 2.14.3
Tokenizers 0.13.3

Training results (Full)

Training Loss	Epoch	Step	Validation Loss
5.2105	1.0	13	4.4870
4.4319	2.0	26	3.8456
4.0318	3.0	39	3.4164
3.7558	4.0	52	3.2849
3.5626	5.0	65	3.3146
3.4355	6.0	78	3.1532
3.3299	7.0	91	3.0451
3.2313	8.0	104	2.9359
3.1758	9.0	117	2.8543
3.0762	10.0	130	2.8034
3.0318	11.0	143	2.7975
2.9481	12.0	156	2.6439
2.8848	13.0	169	2.6623
2.9002	14.0	182	2.6425
2.8435	15.0	195	2.6639
2.8451	16.0	208	2.6203
2.7987	17.0	221	2.5597
2.7522	18.0	234	2.5719
2.7194	19.0	247	2.6220
2.6923	20.0	260	2.5566
2.678	21.0	273	2.4172
2.6612	22.0	286	2.5726
2.6272	23.0	299	2.4478
2.6052	24.0	312	2.4366
2.5694	25.0	325	2.3694
2.593	26.0	338	2.4324
2.548	27.0	351	2.4070
2.4954	28.0	364	2.3651
2.5097	29.0	377	2.3268
2.5041	30.0	390	2.4208
2.4919	31.0	403	2.4321
2.461	32.0	416	2.3477
2.4698	33.0	429	2.4017
2.4557	34.0	442	2.3050
2.4464	35.0	455	2.3282
2.4215	36.0	468	2.3339
2.4037	37.0	481	2.2429
2.386	38.0	494	2.3452
2.3961	39.0	507	2.3312
2.3985	40.0	520	2.2921
2.3302	41.0	533	2.2711
2.3128	42.0	546	2.2344
2.3158	43.0	559	2.1982
2.2927	44.0	572	2.1473
2.3122	45.0	585	2.2317
2.2885	46.0	598	2.2060
2.2592	47.0	611	2.1943
2.2492	48.0	624	2.2361
2.2495	49.0	637	2.2059
2.2402	50.0	650	2.1461
2.241	51.0	663	2.2181
2.211	52.0	676	2.0885
2.2165	53.0	689	2.1567
2.2063	54.0	702	2.2112
2.1715	55.0	715	2.2934
2.1601	56.0	728	2.0745
2.1796	57.0	741	2.1070
2.152	58.0	754	2.0930
2.1562	59.0	767	2.1106
2.125	60.0	780	2.1529
2.1318	61.0	793	2.0296
2.1194	62.0	806	2.0323
2.1396	63.0	819	1.9835
2.1108	64.0	832	2.0066
2.0874	65.0	845	1.9062
2.0754	66.0	858	2.1728
2.0928	67.0	871	2.0197
2.0835	68.0	884	2.0767
2.0684	69.0	897	2.1482
2.0505	70.0	910	2.0667
2.0564	71.0	923	2.1489
2.0478	72.0	936	2.0015
2.0478	73.0	949	1.9215
2.0316	74.0	962	2.0238
2.0171	75.0	975	2.0014
2.0248	76.0	988	2.0775
2.0066	77.0	1001	2.0390
2.0018	78.0	1014	2.0043
1.9925	79.0	1027	2.0138
1.9614	80.0	1040	1.9499
1.9877	81.0	1053	1.9642
1.9499	82.0	1066	1.9676
1.932	83.0	1079	1.9332
1.9353	84.0	1092	1.8787
1.9672	85.0	1105	1.9720
1.9313	86.0	1118	1.9343
1.9292	87.0	1131	1.8964
1.9277	88.0	1144	1.9619
1.9158	89.0	1157	1.9608
1.921	90.0	1170	1.9171
1.9191	91.0	1183	1.8871
1.8935	92.0	1196	1.8857
1.8818	93.0	1209	1.8909
1.8782	94.0	1222	1.8951
1.9028	95.0	1235	1.9164
1.8907	96.0	1248	1.9650
1.8626	97.0	1261	1.8906
1.8413	98.0	1274	1.8957
1.854	99.0	1287	1.9644
1.8608	100.0	1300	1.8329
1.8623	101.0	1313	1.8693
1.7798	102.0	1326	1.8913
1.846	103.0	1339	1.7854
1.7972	104.0	1352	1.8611
1.8443	105.0	1365	1.8482
1.791	106.0	1378	1.7168
1.7879	107.0	1391	1.8093
1.7886	108.0	1404	1.8924
1.8192	109.0	1417	1.7715
1.7919	110.0	1430	1.7415
1.7581	111.0	1443	1.7956
1.7873	112.0	1456	1.7213
1.7873	113.0	1469	1.7340
1.7764	114.0	1482	1.8535
1.7612	115.0	1495	1.8554
1.7737	116.0	1508	1.8126
1.7416	117.0	1521	1.8327
1.7648	118.0	1534	1.6832
1.7262	119.0	1547	1.6972
1.7334	120.0	1560	1.7930
1.7172	121.0	1573	1.6962
1.7282	122.0	1586	1.8800
1.7038	123.0	1599	1.7828
1.6935	124.0	1612	1.7646
1.758	125.0	1625	1.8069
1.7018	126.0	1638	1.6958
1.6886	127.0	1651	1.6692
1.7004	128.0	1664	1.7256
1.6947	129.0	1677	1.7587
1.6897	130.0	1690	1.7484
1.7037	131.0	1703	1.8455
1.6981	132.0	1716	1.7588
1.6828	133.0	1729	1.7421
1.6596	134.0	1742	1.6933
1.6782	135.0	1755	1.7040
1.6595	136.0	1768	1.6705
1.6567	137.0	1781	1.7744
1.6588	138.0	1794	1.6545
1.6225	139.0	1807	1.7576
1.6394	140.0	1820	1.7256
1.6515	141.0	1833	1.6668
1.6331	142.0	1846	1.7884
1.6367	143.0	1859	1.7093
1.6335	144.0	1872	1.7098
1.6501	145.0	1885	1.6671
1.6192	146.0	1898	1.7073
1.6198	147.0	1911	1.6653
1.6182	148.0	1924	1.6723
1.6172	149.0	1937	1.7293
1.6129	150.0	1950	1.6545
1.6054	151.0	1963	1.6850
1.5967	152.0	1976	1.7064
1.6028	153.0	1989	1.5292
1.6156	154.0	2002	1.6477
1.5965	155.0	2015	1.6110
1.5695	156.0	2028	1.7071
1.5586	157.0	2041	1.6504
1.561	158.0	2054	1.6147
1.5643	159.0	2067	1.6941
1.5797	160.0	2080	1.7398
1.5609	161.0	2093	1.5761
1.5465	162.0	2106	1.6003
1.5467	163.0	2119	1.5839
1.5935	164.0	2132	1.6530
1.5439	165.0	2145	1.6743
1.559	166.0	2158	1.5143
1.5648	167.0	2171	1.6390
1.552	168.0	2184	1.5389
1.5164	169.0	2197	1.5879
1.5342	170.0	2210	1.6785
1.5319	171.0	2223	1.6341
1.5477	172.0	2236	1.7071
1.5364	173.0	2249	1.6268
1.5366	174.0	2262	1.7247
1.5445	175.0	2275	1.6668
1.4916	176.0	2288	1.5756
1.509	177.0	2301	1.5412
1.5316	178.0	2314	1.6270
1.5156	179.0	2327	1.6423
1.4918	180.0	2340	1.6112
1.4997	181.0	2353	1.5775
1.5187	182.0	2366	1.6248
1.5254	183.0	2379	1.5884
1.4732	184.0	2392	1.5787
1.4844	185.0	2405	1.5358
1.4882	186.0	2418	1.5144
1.478	187.0	2431	1.5223
1.5101	188.0	2444	1.5787
1.4688	189.0	2457	1.5479
1.4815	190.0	2470	1.5141
1.4925	191.0	2483	1.5939
1.467	192.0	2496	1.5471
1.4718	193.0	2509	1.6845
1.4699	194.0	2522	1.5943
1.4562	195.0	2535	1.4745
1.4451	196.0	2548	1.5922
1.4451	197.0	2561	1.5856
1.4624	198.0	2574	1.5519
1.444	199.0	2587	1.6538
1.4498	200.0	2600	1.5037
1.4285	201.0	2613	1.5539
1.4439	202.0	2626	1.5387
1.4177	203.0	2639	1.5756
1.436	204.0	2652	1.6136
1.4184	205.0	2665	1.5014
1.43	206.0	2678	1.4983
1.4347	207.0	2691	1.5896
1.39	208.0	2704	1.5506
1.4198	209.0	2717	1.5142
1.4101	210.0	2730	1.4930
1.4219	211.0	2743	1.4814
1.4039	212.0	2756	1.3750
1.4479	213.0	2769	1.5330
1.4354	214.0	2782	1.5179
1.4163	215.0	2795	1.5970
1.4459	216.0	2808	1.4755
1.3714	217.0	2821	1.4230
1.3957	218.0	2834	1.5087
1.396	219.0	2847	1.5570
1.3866	220.0	2860	1.4955
1.4122	221.0	2873	1.4272
1.371	222.0	2886	1.5209
1.3907	223.0	2899	1.4725
1.3856	224.0	2912	1.5021
1.4053	225.0	2925	1.4880
1.4074	226.0	2938	1.4988
1.3827	227.0	2951	1.5527
1.4045	228.0	2964	1.5350
1.3626	229.0	2977	1.5093
1.3795	230.0	2990	1.4497
1.3973	231.0	3003	1.5106
1.3703	232.0	3016	1.4619
1.3942	233.0	3029	1.4553
1.3447	234.0	3042	1.5061
1.3438	235.0	3055	1.5167
1.3496	236.0	3068	1.4060
1.3614	237.0	3081	1.4211
1.3618	238.0	3094	1.4624
1.359	239.0	3107	1.4450
1.3657	240.0	3120	1.4795
1.3599	241.0	3133	1.4887
1.3532	242.0	3146	1.4606
1.3528	243.0	3159	1.4225
1.3445	244.0	3172	1.3912
1.3344	245.0	3185	1.4055
1.3358	246.0	3198	1.5152
1.3591	247.0	3211	1.4825
1.3162	248.0	3224	1.4721
1.3197	249.0	3237	1.4375
1.3358	250.0	3250	1.4644
1.3374	251.0	3263	1.4449
1.3548	252.0	3276	1.4405
1.3266	253.0	3289	1.5357
1.3172	254.0	3302	1.3515
1.3089	255.0	3315	1.4408
1.3209	256.0	3328	1.3895
1.3047	257.0	3341	1.4508
1.2877	258.0	3354	1.3954
1.3409	259.0	3367	1.4417
1.31	260.0	3380	1.5124
1.3229	261.0	3393	1.4047
1.3275	262.0	3406	1.3780
1.295	263.0	3419	1.4209
1.3279	264.0	3432	1.3867
1.291	265.0	3445	1.4694
1.2839	266.0	3458	1.5100
1.3064	267.0	3471	1.3646
1.3086	268.0	3484	1.4390
1.3381	269.0	3497	1.4367
1.3333	270.0	3510	1.4078
1.2775	271.0	3523	1.5213
1.2989	272.0	3536	1.4341
1.2759	273.0	3549	1.5165
1.2796	274.0	3562	1.4705
1.3037	275.0	3575	1.3945
1.3132	276.0	3588	1.4560
1.2816	277.0	3601	1.4123
1.2934	278.0	3614	1.3742
1.2873	279.0	3627	1.3824
1.2842	280.0	3640	1.3269
1.2617	281.0	3653	1.4345
1.2661	282.0	3666	1.4682
1.3096	283.0	3679	1.3989
1.2724	284.0	3692	1.3142
1.2529	285.0	3705	1.2795
1.2611	286.0	3718	1.3844
1.2578	287.0	3731	1.3536
1.2854	288.0	3744	1.3770
1.2811	289.0	3757	1.3892
1.2189	290.0	3770	1.3767
1.283	291.0	3783	1.4034
1.2684	292.0	3796	1.3867
1.241	293.0	3809	1.3572
1.2503	294.0	3822	1.3583
1.2605	295.0	3835	1.4600
1.2697	296.0	3848	1.2754
1.2469	297.0	3861	1.4295
1.2451	298.0	3874	1.4645
1.2765	299.0	3887	1.3605
1.2482	300.0	3900	1.4915
1.2564	301.0	3913	1.3490
1.233	302.0	3926	1.3273
1.2313	303.0	3939	1.3861
1.2491	304.0	3952	1.4016
1.2607	305.0	3965	1.3714
1.2548	306.0	3978	1.3572
1.2536	307.0	3991	1.3630
1.24	308.0	4004	1.3070
1.2352	309.0	4017	1.4311
1.2643	310.0	4030	1.2794
1.2281	311.0	4043	1.3855
1.2428	312.0	4056	1.3784
1.2196	313.0	4069	1.3430
1.2116	314.0	4082	1.4230
1.2261	315.0	4095	1.4760
1.25	316.0	4108	1.3658
1.2281	317.0	4121	1.3563
1.2308	318.0	4134	1.3107
1.2247	319.0	4147	1.3554
1.2354	320.0	4160	1.3956
1.2168	321.0	4173	1.2753
1.2078	322.0	4186	1.3253
1.2481	323.0	4199	1.3025
1.2331	324.0	4212	1.3707
1.1974	325.0	4225	1.2874
1.212	326.0	4238	1.3210
1.225	327.0	4251	1.4129
1.2161	328.0	4264	1.3364
1.2304	329.0	4277	1.3822
1.1903	330.0	4290	1.4887
1.2208	331.0	4303	1.2687
1.229	332.0	4316	1.3730
1.205	333.0	4329	1.3521
1.2023	334.0	4342	1.3770
1.2151	335.0	4355	1.3095
1.2255	336.0	4368	1.3003
1.2205	337.0	4381	1.2123
1.203	338.0	4394	1.2995
1.2013	339.0	4407	1.2838
1.1997	340.0	4420	1.3023
1.2033	341.0	4433	1.3111
1.1934	342.0	4446	1.4057
1.1832	343.0	4459	1.3468
1.2405	344.0	4472	1.3362
1.1803	345.0	4485	1.4813
1.2154	346.0	4498	1.3207
1.2314	347.0	4511	1.3236
1.1927	348.0	4524	1.3428
1.2194	349.0	4537	1.3533
1.1995	350.0	4550	1.3465
1.177	351.0	4563	1.3484
1.1993	352.0	4576	1.2859
1.1687	353.0	4589	1.2699
1.2045	354.0	4602	1.3686
1.2084	355.0	4615	1.3515
1.1837	356.0	4628	1.2735
1.1937	357.0	4641	1.2835
1.2004	358.0	4654	1.2793
1.1838	359.0	4667	1.2798
1.2026	360.0	4680	1.3856
1.1669	361.0	4693	1.3719
1.1716	362.0	4706	1.2613
1.1906	363.0	4719	1.2719
1.1914	364.0	4732	1.3864
1.1874	365.0	4745	1.3255
1.1848	366.0	4758	1.2984
1.1778	367.0	4771	1.3461
1.1964	368.0	4784	1.3320
1.16	369.0	4797	1.2962
1.1873	370.0	4810	1.3035
1.1632	371.0	4823	1.3465
1.1807	372.0	4836	1.3453
1.1331	373.0	4849	1.3527
1.1694	374.0	4862	1.2928
1.1615	375.0	4875	1.3519
1.1944	376.0	4888	1.4072
1.163	377.0	4901	1.3156
1.1719	378.0	4914	1.3074
1.1721	379.0	4927	1.3121
1.1618	380.0	4940	1.3039
1.1852	381.0	4953	1.3562
1.1838	382.0	4966	1.3383
1.1616	383.0	4979	1.2922
1.1401	384.0	4992	1.2676
1.165	385.0	5005	1.2625
1.1564	386.0	5018	1.1716
1.1662	387.0	5031	1.2738
1.1761	388.0	5044	1.4011
1.1587	389.0	5057	1.3821
1.1517	390.0	5070	1.2879
1.1699	391.0	5083	1.2898
1.149	392.0	5096	1.2710
1.1541	393.0	5109	1.2612
1.1597	394.0	5122	1.2993
1.1449	395.0	5135	1.2522
1.1332	396.0	5148	1.3367
1.1537	397.0	5161	1.3018
1.1789	398.0	5174	1.3705
1.169	399.0	5187	1.3128
1.1685	400.0	5200	1.3068
1.137	401.0	5213	1.2384
1.177	402.0	5226	1.2547
1.1592	403.0	5239	1.3295
1.1477	404.0	5252	1.3415
1.1465	405.0	5265	1.2466
1.1743	406.0	5278	1.3045
1.1386	407.0	5291	1.3124
1.1379	408.0	5304	1.2826
1.1828	409.0	5317	1.2788
1.1353	410.0	5330	1.3787
1.1536	411.0	5343	1.2968
1.1495	412.0	5356	1.2920
1.1424	413.0	5369	1.3238
1.158	414.0	5382	1.3301
1.1715	415.0	5395	1.2298
1.1559	416.0	5408	1.2769
1.1399	417.0	5421	1.3263
1.186	418.0	5434	1.2924
1.1653	419.0	5447	1.3279
1.14	420.0	5460	1.2892
1.1463	421.0	5473	1.3875
1.1406	422.0	5486	1.3136
1.1705	423.0	5499	1.2579
1.1065	424.0	5512	1.2955
1.145	425.0	5525	1.2970
1.1538	426.0	5538	1.3030
1.1674	427.0	5551	1.3060
1.1283	428.0	5564	1.2325
1.1683	429.0	5577	1.3085
1.1598	430.0	5590	1.2469
1.1429	431.0	5603	1.2523
1.1552	432.0	5616	1.3124
1.1722	433.0	5629	1.2955
1.1329	434.0	5642	1.3249
1.1486	435.0	5655	1.3245
1.124	436.0	5668	1.4052
1.1092	437.0	5681	1.2399
1.135	438.0	5694	1.2788
1.1637	439.0	5707	1.2844
1.1712	440.0	5720	1.2531
1.1401	441.0	5733	1.2790
1.1195	442.0	5746	1.2876
1.1524	443.0	5759	1.2565
1.1292	444.0	5772	1.1492
1.1342	445.0	5785	1.3050
1.1628	446.0	5798	1.2911
1.1286	447.0	5811	1.3624
1.1193	448.0	5824	1.2382
1.1521	449.0	5837	1.2717
1.1128	450.0	5850	1.2865
1.1321	451.0	5863	1.2785
1.1707	452.0	5876	1.3514
1.1431	453.0	5889	1.3321
1.1413	454.0	5902	1.2886
1.0983	455.0	5915	1.3165
1.1202	456.0	5928	1.2375
1.1259	457.0	5941	1.2166
1.1353	458.0	5954	1.3579
1.1272	459.0	5967	1.2890
1.1411	460.0	5980	1.2397
1.115	461.0	5993	1.2803
1.14	462.0	6006	1.2439
1.11	463.0	6019	1.1894
1.1539	464.0	6032	1.2979
1.1052	465.0	6045	1.2281
1.1092	466.0	6058	1.2853
1.1229	467.0	6071	1.2988
1.1209	468.0	6084	1.3058
1.1147	469.0	6097	1.2705
1.1228	470.0	6110	1.2435
1.1124	471.0	6123	1.2188
1.0922	472.0	6136	1.2892
1.1228	473.0	6149	1.2250
1.1341	474.0	6162	1.2373
1.1295	475.0	6175	1.2126
1.1105	476.0	6188	1.3032
1.1223	477.0	6201	1.2190
1.1487	478.0	6214	1.2728
1.1288	479.0	6227	1.3258
1.1398	480.0	6240	1.2114
1.1127	481.0	6253	1.2695
1.135	482.0	6266	1.3376
1.106	483.0	6279	1.2860
1.0978	484.0	6292	1.3001
1.1254	485.0	6305	1.3180
1.1117	486.0	6318	1.3036
1.1249	487.0	6331	1.2380
1.1111	488.0	6344	1.3112
1.119	489.0	6357	1.2587
1.1203	490.0	6370	1.2867
1.1195	491.0	6383	1.3153
1.1304	492.0	6396	1.2762
1.1268	493.0	6409	1.2757
1.1478	494.0	6422	1.2493
1.1527	495.0	6435	1.2793
1.1252	496.0	6448	1.2435
1.1307	497.0	6461	1.3311
1.1163	498.0	6474	1.3016
1.099	499.0	6487	1.3532
1.1246	500.0	6500	1.2222

snousias/bert-base-greek-uncased-v5-finetuned-polylex-mg

bert-base-greek-uncased-v5-finetuned-polylex-mg