README.md

metadata

license: apache-2.0
base_model: google/mt5-small
tags:
  - generated_from_trainer
model-index:
  - name: text-translit-detector-ru
    results: []

text-translit-detector-ru

This model is a fine-tuned version of google/mt5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0632
• Mean Distance: 0
• Max Distance: 1

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: 0.0001
• train_batch_size: 15
• eval_batch_size: 15
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Distance	Max Distance
0.8541	1.0	2664	0.3404	0	1
0.0451	2.0	5328	0.0605	0	1
0.0112	3.0	7992	0.0411	0	1
0.0068	4.0	10656	0.0205	0	1
0.007	5.0	13320	0.0242	0	1
0.0022	6.0	15984	0.0272	0	1
0.0054	7.0	18648	0.0080	0	1
0.0036	8.0	21312	0.0252	0	1
0.0039	9.0	23976	0.0210	0	1
0.0026	10.0	26640	0.0170	0	1
0.0026	11.0	29304	0.0043	0	1
0.0029	12.0	31968	0.0135	0	1
0.0011	13.0	34632	0.0313	0	1
0.0017	14.0	37296	0.0353	0	1
0.0014	15.0	39960	0.0117	0	1
0.0014	16.0	42624	0.0140	0	1
0.0013	17.0	45288	0.0220	0	1
0.0009	18.0	47952	0.0247	0	1
0.0017	19.0	50616	0.0322	0	1
0.0022	20.0	53280	0.0314	0	1
0.0006	21.0	55944	0.0305	0	1
0.001	22.0	58608	0.0292	0	1
0.0008	23.0	61272	0.0373	0	1
0.0008	24.0	63936	0.0309	0	1
0.0008	25.0	66600	0.0385	0	1
0.0014	26.0	69264	0.0134	0	1
0.0004	27.0	71928	0.0239	0	1
0.0011	28.0	74592	0.0164	0	1
0.0002	29.0	77256	0.0186	0	1
0.0001	30.0	79920	0.0298	0	1
0.0008	31.0	82584	0.0277	0	1
0.0003	32.0	85248	0.0377	0	1
0.0003	33.0	87912	0.0354	0	1
0.0007	34.0	90576	0.0585	0	1
0.0005	35.0	93240	0.0568	0	1
0.0001	36.0	95904	0.0567	0	1
0.0009	37.0	98568	0.0605	0	1
0.0002	38.0	101232	0.0613	0	1
0.0002	39.0	103896	0.0563	0	1
0.0002	40.0	106560	0.0632	0	1

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1