Alikhan Urumov
commited on
Commit
•
92c81c0
1
Parent(s):
dafe4b7
update model card README.md
Browse files
README.md
CHANGED
|
@@ -1,8 +1,6 @@
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
| 4 |
-
metrics:
|
| 5 |
-
- rouge
|
| 6 |
model-index:
|
| 7 |
- name: t5-russian-spell
|
| 8 |
results: []
|
|
@@ -13,20 +11,19 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 13 |
|
| 14 |
# t5-russian-spell
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
ывсем привет выныканалетоп армии и это двадцать пятый день спец операций на украине ет самый главной новости российские военные ракетами кинжалы калибр уничтожили крупную военную топливную базу украины ракетным ударом по населенному пункту поджетамиром уничтжены более стаукраинских военных в две тысячи двадцать втором году
|
| 19 |
-
|
| 20 |
-
Всем привет! Вы канале от армии. И это 25 день спецопераций на Украине. Есть самые главные новости. Российские военные ракетами «Кинжалы» и «Кинжалы» калибра уничтожили крупную военную топливную базу Украины ракетным ударом по населенному пункту Поджандаром уничтожены более ста украинских военных в 2022 году.
|
| 21 |
-
|
| 22 |
-
This model is a fine-tuned version of [sberbank-ai/ruT5-base](https://huggingface.co/sberbank-ai/ruT5-base) on the None dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
-
-
|
| 25 |
-
-
|
| 26 |
-
-
|
| 27 |
-
-
|
| 28 |
-
-
|
| 29 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
|
| 31 |
## Model description
|
| 32 |
|
|
@@ -45,40 +42,15 @@ More information needed
|
|
| 45 |
### Training hyperparameters
|
| 46 |
|
| 47 |
The following hyperparameters were used during training:
|
| 48 |
-
- learning_rate:
|
| 49 |
- train_batch_size: 8
|
| 50 |
- eval_batch_size: 8
|
| 51 |
- seed: 42
|
| 52 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 53 |
- lr_scheduler_type: linear
|
| 54 |
-
- num_epochs:
|
| 55 |
- mixed_precision_training: Native AMP
|
| 56 |
|
| 57 |
-
### Training results
|
| 58 |
-
|
| 59 |
-
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|
| 60 |
-
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
|
| 61 |
-
| 0.1709 | 0.2 | 2500 | 0.4521 | 44.2702 | 21.5514 | 44.2338 | 44.0689 | 61.032 |
|
| 62 |
-
| 0.1557 | 0.4 | 5000 | 0.4639 | 44.2613 | 21.8757 | 44.2442 | 44.0816 | 60.914 |
|
| 63 |
-
| 0.4276 | 0.6 | 7500 | 0.4139 | 45.2125 | 22.268 | 45.1434 | 44.9713 | 60.494 |
|
| 64 |
-
| 0.4675 | 0.8 | 10000 | 0.4016 | 44.2872 | 22.2143 | 44.27 | 44.0862 | 61.018 |
|
| 65 |
-
| 0.5048 | 1.0 | 12500 | 0.3923 | 44.9732 | 22.2551 | 45.0251 | 44.822 | 60.952 |
|
| 66 |
-
| 0.4362 | 1.21 | 15000 | 0.3920 | 44.8982 | 21.9817 | 44.8949 | 44.7051 | 61.29 |
|
| 67 |
-
| 0.426 | 1.41 | 17500 | 0.3879 | 45.4473 | 22.5263 | 45.4284 | 45.2483 | 60.674 |
|
| 68 |
-
| 0.4174 | 1.61 | 20000 | 0.3832 | 45.4006 | 22.2695 | 45.382 | 45.2161 | 60.92 |
|
| 69 |
-
| 0.4229 | 1.81 | 22500 | 0.3774 | 45.2545 | 22.2894 | 45.2335 | 45.065 | 60.722 |
|
| 70 |
-
| 0.4071 | 2.01 | 25000 | 0.3782 | 45.2875 | 22.4234 | 45.2902 | 45.1445 | 61.138 |
|
| 71 |
-
| 0.3966 | 2.21 | 27500 | 0.3782 | 45.1692 | 22.197 | 45.2311 | 45.0222 | 60.68 |
|
| 72 |
-
| 0.389 | 2.41 | 30000 | 0.3744 | 45.6209 | 22.5031 | 45.6023 | 45.4973 | 60.878 |
|
| 73 |
-
| 0.3896 | 2.61 | 32500 | 0.3718 | 45.2454 | 22.4507 | 45.2479 | 45.1446 | 60.76 |
|
| 74 |
-
| 0.3961 | 2.81 | 35000 | 0.3711 | 45.2779 | 22.4165 | 45.2661 | 45.1617 | 60.984 |
|
| 75 |
-
| 0.3765 | 3.01 | 37500 | 0.3705 | 45.1666 | 22.6603 | 45.0916 | 44.9798 | 60.994 |
|
| 76 |
-
| 0.3757 | 3.22 | 40000 | 0.3709 | 45.1587 | 22.4539 | 45.1129 | 45.0461 | 60.828 |
|
| 77 |
-
| 0.3776 | 3.42 | 42500 | 0.3688 | 45.247 | 22.6266 | 45.2351 | 45.1111 | 60.93 |
|
| 78 |
-
| 0.3691 | 3.62 | 45000 | 0.3693 | 45.3799 | 22.5152 | 45.3839 | 45.2705 | 60.846 |
|
| 79 |
-
| 0.3786 | 3.82 | 47500 | 0.3676 | 45.1151 | 22.4675 | 45.0866 | 44.9917 | 60.922 |
|
| 80 |
-
|
| 81 |
-
|
| 82 |
### Framework versions
|
| 83 |
|
| 84 |
- Transformers 4.17.0
|
|
|
|
| 1 |
---
|
| 2 |
tags:
|
| 3 |
- generated_from_trainer
|
|
|
|
|
|
|
| 4 |
model-index:
|
| 5 |
- name: t5-russian-spell
|
| 6 |
results: []
|
|
|
|
| 11 |
|
| 12 |
# t5-russian-spell
|
| 13 |
|
| 14 |
+
This model is a fine-tuned version of [UrukHan/t5-russian-spell](https://huggingface.co/UrukHan/t5-russian-spell) on the None dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
It achieves the following results on the evaluation set:
|
| 16 |
+
- eval_loss: 0.3178
|
| 17 |
+
- eval_rouge1: 48.5901
|
| 18 |
+
- eval_rouge2: 26.2479
|
| 19 |
+
- eval_rougeL: 48.645
|
| 20 |
+
- eval_rougeLsum: 48.5704
|
| 21 |
+
- eval_gen_len: 61.0637
|
| 22 |
+
- eval_runtime: 438.689
|
| 23 |
+
- eval_samples_per_second: 3.077
|
| 24 |
+
- eval_steps_per_second: 0.385
|
| 25 |
+
- epoch: 0.41
|
| 26 |
+
- step: 65000
|
| 27 |
|
| 28 |
## Model description
|
| 29 |
|
|
|
|
| 42 |
### Training hyperparameters
|
| 43 |
|
| 44 |
The following hyperparameters were used during training:
|
| 45 |
+
- learning_rate: 5e-05
|
| 46 |
- train_batch_size: 8
|
| 47 |
- eval_batch_size: 8
|
| 48 |
- seed: 42
|
| 49 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 50 |
- lr_scheduler_type: linear
|
| 51 |
+
- num_epochs: 1
|
| 52 |
- mixed_precision_training: Native AMP
|
| 53 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 54 |
### Framework versions
|
| 55 |
|
| 56 |
- Transformers 4.17.0
|