You can deploy this model and start earning money today!

cshim-cmu/
k40

Name: k40
Availability: InStock

pytorch

marian

generated_from_trainer

apache-2.0

Views

No views yet

k40

This model is a fine-tuned version of Helsinki-NLP/opus-mt-es-fi on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.5030
Bleu: 1.3285
Chrf: 29.1408
Gen Len: 36.2445

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: 16
eval_batch_size: 16
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Chrf	Gen Len
1.0912	0.09	1000	0.9278	0.7853	11.5201	17.5151
0.8925	0.18	2000	0.8232	0.2399	11.5642	39.34
0.8025	0.27	3000	0.7735	0.1884	13.2798	42.3893
0.7558	0.36	4000	0.7433	0.289	14.5377	45.5101
0.7231	0.45	5000	0.7226	0.3316	14.794	51.2535
0.683	0.54	6000	0.7070	0.3806	15.2357	52.8823
0.6658	0.63	7000	0.6937	0.3154	15.616	57.4125
0.6379	0.72	8000	0.6826	0.402	16.5046	52.1821
0.6283	0.81	9000	0.6725	0.411	16.6021	55.4467
0.6101	0.9	10000	0.6639	0.3015	17.1943	56.5302
0.5936	0.99	11000	0.6569	0.4211	17.8252	54.0332
0.579	1.08	12000	0.6490	0.4441	18.3971	52.6831
0.5753	1.18	13000	0.6420	0.4407	18.8088	51.8511
0.5644	1.27	14000	0.6362	0.5369	19.2326	47.7676
0.5495	1.36	15000	0.6302	0.512	19.7579	47.4537
0.5423	1.45	16000	0.6239	0.5639	20.1563	47.1348
0.5368	1.54	17000	0.6191	0.5211	20.7364	46.9195
0.5247	1.63	18000	0.6135	0.7775	21.3953	44.162
0.5215	1.72	19000	0.6087	0.696	21.2283	46.6419
0.5075	1.81	20000	0.6044	0.732	21.5368	45.6972
0.5062	1.9	21000	0.5995	0.7427	21.7062	46.6952
0.4886	1.99	22000	0.5954	0.7231	21.7736	47.4225
0.492	2.08	23000	0.5906	0.8538	22.24	45.2978
0.4785	2.17	24000	0.5864	0.8839	22.4173	45.4507
0.4712	2.26	25000	0.5824	0.8281	22.9807	44.7052
0.4681	2.35	26000	0.5786	0.8944	22.8454	44.4185
0.4625	2.44	27000	0.5754	0.5707	23.0566	44.9034
0.4519	2.53	28000	0.5714	0.9313	23.44	43.9356
0.4524	2.62	29000	0.5687	0.9375	23.8348	44.7334
0.4436	2.71	30000	0.5652	0.8937	23.8125	43.6308
0.4443	2.8	31000	0.5619	0.611	23.4966	44.6278
0.4368	2.89	32000	0.5584	0.7807	24.2701	44.007
0.4305	2.98	33000	0.5554	0.6147	24.275	43.506
0.4294	3.07	34000	0.5529	0.6857	24.8292	42.2354
0.4216	3.16	35000	0.5495	1.048	24.7773	41.2535
0.412	3.25	36000	0.5479	0.7311	25.0649	39.7364
0.4059	3.34	37000	0.5456	1.26	25.7873	39.7022
0.4024	3.43	38000	0.5433	1.1248	25.6285	41.4497
0.4026	3.53	39000	0.5412	0.9491	25.9591	39.9829
0.4007	3.62	40000	0.5392	1.1765	25.851	40.9618
0.399	3.71	41000	0.5366	1.0397	26.2359	40.2354
0.3948	3.8	42000	0.5350	1.2626	26.5781	40.5473
0.3906	3.89	43000	0.5330	0.897	26.6871	39.5463
0.3902	3.98	44000	0.5317	1.1734	26.9085	39.2777
0.3895	4.07	45000	0.5303	1.2225	26.9647	40.9366
0.3825	4.16	46000	0.5285	1.3356	27.1977	38.1368
0.3825	4.25	47000	0.5269	1.3431	27.3653	38.8954
0.3766	4.34	48000	0.5253	1.3826	27.2921	38.6107
0.3735	4.43	49000	0.5244	1.385	27.4263	39.165
0.3725	4.52	50000	0.5226	1.5999	27.5186	38.6841
0.3724	4.61	51000	0.5215	1.4658	27.6238	38.0875
0.3657	4.7	52000	0.5203	1.4563	27.691	37.4386
0.3627	4.79	53000	0.5189	1.572	28.1626	37.9588
0.3608	4.88	54000	0.5176	1.2823	28.0151	37.1942
0.3613	4.97	55000	0.5164	1.5966	27.9882	36.7666
0.3642	5.06	56000	0.5154	1.3163	28.2748	37.9095
0.3575	5.15	57000	0.5148	1.3452	28.4505	38.0221
0.3523	5.24	58000	0.5137	1.4571	28.4722	38.1751
0.3571	5.33	59000	0.5128	1.5662	28.6911	37.1227
0.3586	5.42	60000	0.5117	1.3369	28.4032	36.4396
0.3513	5.51	61000	0.5115	1.5459	28.7866	36.6298
0.3506	5.6	62000	0.5103	1.4093	29.0212	36.8712
0.3494	5.69	63000	0.5095	1.4116	28.7563	37.3048
0.3443	5.79	64000	0.5086	1.5624	28.9575	36.8169
0.3399	5.88	65000	0.5077	1.6834	28.9275	36.5392
0.3386	5.97	66000	0.5070	1.7754	29.019	36.1569
0.3483	6.06	67000	0.5059	1.6649	29.2083	36.4598
0.3439	6.15	68000	0.5058	1.5521	29.1137	36.7032
0.3397	6.24	69000	0.5050	1.7231	29.0362	35.7736
0.3416	6.33	70000	0.5049	1.6662	29.0505	36.4588
0.342	6.42	71000	0.5039	1.583	29.0309	36.3783
0.3372	6.51	72000	0.5030	1.3285	29.1408	36.2445

Framework versions

Transformers 4.28.1
Pytorch 2.0.0+cu117
Datasets 2.11.0
Tokenizers 0.13.3

cshim-cmu/k40

k40

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

cshim-cmu/
k40