You can deploy this model and start earning money today!

contemmcm/
b2665779c0f8caffb46b1124f1eafe11

Name: b2665779c0f8caffb46b1124f1eafe11
Rating: 2.5 (5 reviews)

generated_from_trainer

Views

No views yet

b2665779c0f8caffb46b1124f1eafe11

This model is a fine-tuned version of google/umt5-small on the Helsinki-NLP/opus_books [fr-ru] dataset. It achieves the following results on the evaluation set:

Loss: 2.7845
Data Size: 1.0
Epoch Runtime: 33.3940
Bleu: 4.0772

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: 5e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 32
total_eval_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: constant
num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Bleu
No log	0	0	18.5090	0	3.4572	0.1861
No log	1	204	18.2077	0.0078	4.3514	0.1860
No log	2	408	17.6318	0.0156	4.3013	0.1880
No log	3	612	16.7061	0.0312	4.9910	0.1965
No log	4	816	15.0499	0.0625	6.1947	0.2306
No log	5	1020	12.6339	0.125	7.9595	0.2300
1.3091	6	1224	8.6207	0.25	11.2590	0.2369
9.042	7	1428	5.7036	0.5	18.9859	0.5044
6.1035	8.0	1632	4.3904	1.0	33.5078	1.5602
5.3881	9.0	1836	3.9626	1.0	34.0889	0.9725
5.0087	10.0	2040	3.6828	1.0	32.6098	1.2083
4.7122	11.0	2244	3.4920	1.0	32.6205	1.5327
4.4805	12.0	2448	3.3695	1.0	33.9218	1.8678
4.3051	13.0	2652	3.2880	1.0	32.5917	2.0812
4.1942	14.0	2856	3.2364	1.0	32.2146	2.1865
4.098	15.0	3060	3.1975	1.0	32.4254	2.3421
3.9902	16.0	3264	3.1632	1.0	33.4077	2.4755
3.9219	17.0	3468	3.1334	1.0	32.8452	2.5962
3.8939	18.0	3672	3.1056	1.0	33.5164	2.6230
3.825	19.0	3876	3.0753	1.0	33.6598	2.7726
3.7533	20.0	4080	3.0556	1.0	34.1965	2.8699
3.6954	21.0	4284	3.0303	1.0	32.2774	2.9212
3.643	22.0	4488	3.0153	1.0	33.1693	3.0288
3.6197	23.0	4692	2.9958	1.0	33.2104	3.0377
3.5653	24.0	4896	2.9804	1.0	33.6497	3.1389
3.5511	25.0	5100	2.9710	1.0	32.6174	3.1599
3.4859	26.0	5304	2.9600	1.0	33.6980	3.2032
3.4148	27.0	5508	2.9398	1.0	33.2059	3.2116
3.4062	28.0	5712	2.9273	1.0	33.6179	3.2883
3.3583	29.0	5916	2.9180	1.0	32.5488	3.3525
3.334	30.0	6120	2.9069	1.0	32.4269	3.3556
3.3371	31.0	6324	2.8930	1.0	32.8480	3.3747
3.2872	32.0	6528	2.8871	1.0	33.7627	3.4744
3.2554	33.0	6732	2.8781	1.0	32.6898	3.5156
3.2344	34.0	6936	2.8670	1.0	32.5476	3.5751
3.1834	35.0	7140	2.8640	1.0	33.7895	3.5936
3.1533	36.0	7344	2.8566	1.0	32.7568	3.6558
3.1262	37.0	7548	2.8461	1.0	32.8651	3.6499
3.1695	38.0	7752	2.8323	1.0	33.5361	3.7131
3.0968	39.0	7956	2.8383	1.0	34.0012	3.7225
3.0778	40.0	8160	2.8312	1.0	33.6559	3.7515
3.0741	41.0	8364	2.8212	1.0	32.7701	3.8406
3.0614	42.0	8568	2.8151	1.0	32.7469	3.8610
3.0024	43.0	8772	2.8204	1.0	32.5930	3.8187
2.9942	44.0	8976	2.8139	1.0	33.0973	3.9242
2.9809	45.0	9180	2.8100	1.0	33.1694	3.8989
3.007	46.0	9384	2.8039	1.0	33.2063	3.9820
2.9381	47.0	9588	2.8007	1.0	32.9562	3.9867
2.9207	48.0	9792	2.7980	1.0	33.7994	4.0161
2.8969	49.0	9996	2.7896	1.0	34.3217	4.0008
2.8611	50.0	10200	2.7845	1.0	33.3940	4.0772

Framework versions

Transformers 4.57.0
Pytorch 2.8.0+cu128
Datasets 4.2.0
Tokenizers 0.22.1

contemmcm/b2665779c0f8caffb46b1124f1eafe11

b2665779c0f8caffb46b1124f1eafe11

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

contemmcm/
b2665779c0f8caffb46b1124f1eafe11