mt5-base-finetuned-novel-chinese-to-spanish-v1

This model is a fine-tuned version of quickman/mt5-base-finetuned-chinese-to-spanish on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.2288
• Score: 0.0063
• Counts: [609, 331, 205, 120]
• Totals: [838, 774, 710, 646]
• Precisions: [72.67303102625299, 42.76485788113695, 28.87323943661972, 18.575851393188856]
• Bp: 0.0002
• Sys Len: 838
• Ref Len: 8089
• Bleu: 0.0063
• Gen Len: 19.0

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 40
• training_steps: 20000

Training results

Training Loss	Epoch	Step	Validation Loss	Score	Counts	Totals	Precisions	Bp	Sys Len	Ref Len	Bleu	Gen Len
2.7093	0.28	500	1.9080	0.0035	[510, 185, 91, 37]	[848, 784, 720, 656]	[60.14150943396226, 23.596938775510203, 12.63888888888889, 5.640243902439025]	0.0002	848	8089	0.0035	19.0
2.4994	0.55	1000	1.7520	0.0036	[524, 199, 100, 46]	[842, 778, 714, 650]	[62.23277909738717, 25.57840616966581, 14.005602240896359, 7.076923076923077]	0.0002	842	8089	0.0036	19.0
2.3427	0.83	1500	1.6632	0.0040	[530, 212, 109, 53]	[844, 780, 716, 652]	[62.796208530805686, 27.17948717948718, 15.223463687150838, 8.128834355828221]	0.0002	844	8089	0.0040	19.0
2.211	1.1	2000	1.5980	0.0050	[548, 230, 123, 66]	[855, 791, 727, 663]	[64.09356725146199, 29.077117572692792, 16.91884456671252, 9.95475113122172]	0.0002	855	8089	0.0050	19.0
2.1536	1.38	2500	1.5442	0.0053	[552, 239, 137, 77]	[852, 788, 724, 660]	[64.78873239436619, 30.32994923857868, 18.92265193370166, 11.666666666666666]	0.0002	852	8089	0.0053	19.0
2.079	1.66	3000	1.5088	0.0055	[551, 244, 142, 84]	[854, 790, 726, 662]	[64.51990632318501, 30.88607594936709, 19.55922865013774, 12.688821752265861]	0.0002	854	8089	0.0055	19.0
2.0374	1.93	3500	1.4768	0.0054	[557, 259, 149, 83]	[849, 785, 721, 657]	[65.60659599528857, 32.99363057324841, 20.665742024965326, 12.633181126331811]	0.0002	849	8089	0.0054	19.0
2.0064	2.21	4000	1.4418	0.0054	[559, 266, 157, 91]	[844, 780, 716, 652]	[66.23222748815166, 34.1025641025641, 21.92737430167598, 13.957055214723926]	0.0002	844	8089	0.0054	19.0
1.9536	2.48	4500	1.4194	0.0056	[557, 260, 157, 87]	[849, 785, 721, 657]	[65.60659599528857, 33.12101910828026, 21.7753120665742, 13.242009132420092]	0.0002	849	8089	0.0056	19.0
1.9436	2.76	5000	1.4030	0.0051	[561, 262, 151, 85]	[841, 777, 713, 649]	[66.70630202140309, 33.71943371943372, 21.1781206171108, 13.097072419106317]	0.0002	841	8089	0.0051	19.0
1.8939	3.04	5500	1.3826	0.0059	[568, 277, 169, 99]	[848, 784, 720, 656]	[66.98113207547169, 35.33163265306123, 23.47222222222222, 15.091463414634147]	0.0002	848	8089	0.0059	19.0
1.8497	3.31	6000	1.3649	0.0059	[576, 288, 180, 107]	[843, 779, 715, 651]	[68.32740213523131, 36.97047496790757, 25.174825174825173, 16.43625192012289]	0.0002	843	8089	0.0059	19.0
1.8177	3.59	6500	1.3575	0.0060	[585, 285, 173, 98]	[847, 783, 719, 655]	[69.06729634002362, 36.39846743295019, 24.061196105702365, 14.961832061068701]	0.0002	847	8089	0.0060	19.0
1.8368	3.86	7000	1.3428	0.0061	[583, 285, 171, 95]	[851, 787, 723, 659]	[68.50763807285547, 36.213468869123254, 23.651452282157678, 14.41578148710167]	0.0002	851	8089	0.0061	19.0
1.7906	4.14	7500	1.3295	0.0059	[581, 284, 167, 88]	[850, 786, 722, 658]	[68.3529411764706, 36.1323155216285, 23.130193905817176, 13.373860182370821]	0.0002	850	8089	0.0059	19.0
1.766	4.42	8000	1.3204	0.0057	[575, 279, 161, 89]	[848, 784, 720, 656]	[67.80660377358491, 35.58673469387755, 22.36111111111111, 13.567073170731707]	0.0002	848	8089	0.0057	19.0
1.7615	4.69	8500	1.3124	0.0061	[590, 293, 176, 100]	[848, 784, 720, 656]	[69.5754716981132, 37.37244897959184, 24.444444444444443, 15.24390243902439]	0.0002	848	8089	0.0061	19.0
1.7741	4.97	9000	1.3057	0.0062	[590, 298, 180, 105]	[846, 782, 718, 654]	[69.73995271867612, 38.107416879795394, 25.069637883008358, 16.05504587155963]	0.0002	846	8089	0.0062	19.0
1.7266	5.24	9500	1.2969	0.0062	[592, 304, 182, 104]	[846, 782, 718, 654]	[69.97635933806147, 38.87468030690537, 25.348189415041784, 15.902140672782874]	0.0002	846	8089	0.0062	19.0
1.7309	5.52	10000	1.2904	0.0054	[580, 287, 166, 88]	[840, 776, 712, 648]	[69.04761904761905, 36.98453608247423, 23.314606741573034, 13.580246913580247]	0.0002	840	8089	0.0054	19.0
1.6973	5.79	10500	1.2818	0.0059	[591, 302, 179, 100]	[842, 778, 714, 650]	[70.19002375296913, 38.81748071979435, 25.07002801120448, 15.384615384615385]	0.0002	842	8089	0.0059	19.0
1.6613	6.07	11000	1.2757	0.0058	[596, 302, 185, 102]	[840, 776, 712, 648]	[70.95238095238095, 38.91752577319588, 25.98314606741573, 15.74074074074074]	0.0002	840	8089	0.0058	19.0
1.6699	6.35	11500	1.2689	0.0063	[600, 316, 197, 113]	[842, 778, 714, 650]	[71.25890736342043, 40.616966580976865, 27.591036414565828, 17.384615384615383]	0.0002	842	8089	0.0063	19.0
1.6566	6.62	12000	1.2630	0.0064	[610, 320, 194, 109]	[844, 780, 716, 652]	[72.27488151658768, 41.02564102564103, 27.094972067039105, 16.717791411042946]	0.0002	844	8089	0.0064	19.0
1.6417	6.9	12500	1.2592	0.0065	[606, 325, 201, 116]	[843, 779, 715, 651]	[71.88612099644128, 41.7201540436457, 28.111888111888113, 17.81874039938556]	0.0002	843	8089	0.0065	19.0
1.6703	7.17	13000	1.2531	0.0072	[616, 325, 198, 113]	[855, 791, 727, 663]	[72.046783625731, 41.08723135271808, 27.235213204951858, 17.043740573152338]	0.0002	855	8089	0.0072	19.0
1.6283	7.45	13500	1.2508	0.0069	[614, 334, 209, 122]	[846, 782, 718, 654]	[72.57683215130024, 42.710997442455245, 29.108635097493035, 18.654434250764528]	0.0002	846	8089	0.0069	19.0
1.6139	7.73	14000	1.2485	0.0056	[595, 315, 192, 111]	[833, 769, 705, 641]	[71.42857142857143, 40.96228868660598, 27.23404255319149, 17.316692667706707]	0.0002	833	8089	0.0056	19.0
1.6203	8.0	14500	1.2425	0.0067	[613, 329, 203, 119]	[845, 781, 717, 653]	[72.54437869822485, 42.12548015364917, 28.312412831241282, 18.223583460949463]	0.0002	845	8089	0.0067	19.0
1.6289	8.28	15000	1.2414	0.0061	[603, 322, 200, 119]	[837, 773, 709, 645]	[72.04301075268818, 41.65588615782665, 28.208744710860366, 18.449612403100776]	0.0002	837	8089	0.0061	19.0
1.6301	8.55	15500	1.2386	0.0063	[610, 328, 205, 123]	[838, 774, 710, 646]	[72.79236276849642, 42.377260981912144, 28.87323943661972, 19.040247678018577]	0.0002	838	8089	0.0063	19.0
1.5992	8.83	16000	1.2379	0.0061	[603, 323, 200, 119]	[837, 773, 709, 645]	[72.04301075268818, 41.785252263906855, 28.208744710860366, 18.449612403100776]	0.0002	837	8089	0.0061	19.0
1.5984	9.11	16500	1.2367	0.0060	[597, 317, 195, 116]	[837, 773, 709, 645]	[71.32616487455198, 41.00905562742562, 27.50352609308886, 17.984496124031008]	0.0002	837	8089	0.0060	19.0
1.6026	9.38	17000	1.2336	0.0063	[606, 326, 204, 124]	[838, 774, 710, 646]	[72.31503579952268, 42.11886304909561, 28.732394366197184, 19.195046439628484]	0.0002	838	8089	0.0063	19.0
1.6059	9.66	17500	1.2319	0.0061	[606, 330, 206, 123]	[835, 771, 707, 643]	[72.57485029940119, 42.80155642023346, 29.13719943422914, 19.12908242612753]	0.0002	835	8089	0.0061	19.0
1.6227	9.93	18000	1.2294	0.0063	[609, 334, 209, 122]	[837, 773, 709, 645]	[72.75985663082437, 43.20827943078913, 29.478138222849083, 18.914728682170544]	0.0002	837	8089	0.0063	19.0
1.6031	10.21	18500	1.2300	0.0060	[605, 328, 203, 120]	[835, 771, 707, 643]	[72.45508982035928, 42.54215304798962, 28.712871287128714, 18.662519440124417]	0.0002	835	8089	0.0060	19.0
1.5746	10.49	19000	1.2301	0.0064	[612, 335, 209, 123]	[838, 774, 710, 646]	[73.0310262529833, 43.281653746770026, 29.43661971830986, 19.040247678018577]	0.0002	838	8089	0.0064	19.0
1.5689	10.76	19500	1.2288	0.0063	[609, 331, 205, 120]	[838, 774, 710, 646]	[72.67303102625299, 42.76485788113695, 28.87323943661972, 18.575851393188856]	0.0002	838	8089	0.0063	19.0
1.5928	11.04	20000	1.2288	0.0063	[609, 331, 205, 120]	[838, 774, 710, 646]	[72.67303102625299, 42.76485788113695, 28.87323943661972, 18.575851393188856]	0.0002	838	8089	0.0063	19.0

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3