Hugging Face's logo

kazandaev
/
opus-mt-en-ru-finetuned

Text2Text Generation
Transformers
PyTorch
TensorBoard
Rust
marian
Generated from Trainer
Model card Files Metrics Community
2
kazandaev commited on
Commit
f08b1fe
·
2 Parent(s): 34823d6 bb7c67b

:Merge branch 'main' of https://huggingface.co/kazandaev/opus-mt-en-ru-finetuned into main

Browse files
Files changed (24) hide show
  1. .gitignore +1 -0
  2. README.md +55 -23
  3. config.json +7 -1
  4. pytorch_model.bin +2 -2
  5. runs/Feb22_18-58-37_22d008391efc/1645557026.79138/events.out.tfevents.1645557026.22d008391efc.125.1 +3 -0
  6. runs/Feb22_18-58-37_22d008391efc/events.out.tfevents.1645556822.22d008391efc.125.0 +3 -0
  7. runs/Feb23_14-20-03_035d423601e6/1645626676.839984/events.out.tfevents.1645626676.035d423601e6.127.1 +3 -0
  8. runs/Feb23_14-20-03_035d423601e6/events.out.tfevents.1645626471.035d423601e6.127.0 +3 -0
  9. runs/Feb23_21-35-59_06f9a29b9e3d/1645652856.2353837/events.out.tfevents.1645652856.06f9a29b9e3d.10505.1 +3 -0
  10. runs/Feb23_21-35-59_06f9a29b9e3d/events.out.tfevents.1645652648.06f9a29b9e3d.10505.0 +3 -0
  11. runs/Feb24_13-52-23_e7302c87fc7a/1645711475.5900414/events.out.tfevents.1645711475.e7302c87fc7a.128.1 +3 -0
  12. runs/Feb24_13-52-23_e7302c87fc7a/events.out.tfevents.1645711253.e7302c87fc7a.128.0 +3 -0
  13. runs/Feb25_10-57-53_a509e49b085a/1645787388.3308783/events.out.tfevents.1645787388.a509e49b085a.125.1 +3 -0
  14. runs/Feb25_10-57-53_a509e49b085a/events.out.tfevents.1645787176.a509e49b085a.125.0 +3 -0
  15. runs/Feb26_12-02-08_1592dd873767/1645877645.0248463/events.out.tfevents.1645877645.1592dd873767.128.1 +3 -0
  16. runs/Feb26_12-02-08_1592dd873767/events.out.tfevents.1645877424.1592dd873767.128.0 +3 -0
  17. runs/Feb27_06-04-51_db8c298e0d8d/1645942432.061396/events.out.tfevents.1645942432.db8c298e0d8d.125.1 +3 -0
  18. runs/Feb27_06-04-51_db8c298e0d8d/events.out.tfevents.1645942432.db8c298e0d8d.125.0 +3 -0
  19. runs/Feb27_13-46-28_db8c298e0d8d/1645969998.99642/events.out.tfevents.1645969998.db8c298e0d8d.125.3 +3 -0
  20. runs/Feb27_13-46-28_db8c298e0d8d/events.out.tfevents.1645969779.db8c298e0d8d.125.2 +3 -0
  21. special_tokens_map.json +1 -0
  22. tokenizer_config.json +1 -1
  23. training_args.bin +3 -0
  24. vocab.json +0 -0
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
README.md CHANGED
@@ -1,32 +1,64 @@
1
  ---
2
- tags:
3
- - translation
4
  license: apache-2.0
 
 
 
 
 
 
 
5
  ---
6
 
7
- ### opus-mt-en-ru
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
 
9
- * source languages: en
10
- * target languages: ru
11
- * OPUS readme: [en-ru](https://github.com/Helsinki-NLP/OPUS-MT-train/blob/master/models/en-ru/README.md)
12
 
13
- * dataset: opus
14
- * model: transformer-align
15
- * pre-processing: normalization + SentencePiece
16
- * download original weights: [opus-2020-02-11.zip](https://object.pouta.csc.fi/OPUS-MT-models/en-ru/opus-2020-02-11.zip)
17
- * test set translations: [opus-2020-02-11.test.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-ru/opus-2020-02-11.test.txt)
18
- * test set scores: [opus-2020-02-11.eval.txt](https://object.pouta.csc.fi/OPUS-MT-models/en-ru/opus-2020-02-11.eval.txt)
 
19
 
20
- ## Benchmarks
21
 
22
- | testset | BLEU | chr-F |
23
- |-----------------------|-------|-------|
24
- | newstest2012.en.ru | 31.1 | 0.581 |
25
- | newstest2013.en.ru | 23.5 | 0.513 |
26
- | newstest2015-enru.en.ru | 27.5 | 0.564 |
27
- | newstest2016-enru.en.ru | 26.4 | 0.548 |
28
- | newstest2017-enru.en.ru | 29.1 | 0.572 |
29
- | newstest2018-enru.en.ru | 25.4 | 0.554 |
30
- | newstest2019-enru.en.ru | 27.1 | 0.533 |
31
- | Tatoeba.en.ru | 48.4 | 0.669 |
32
 
 
 
 
 
 
1
  ---
 
 
2
  license: apache-2.0
3
+ tags:
4
+ - generated_from_trainer
5
+ metrics:
6
+ - bleu
7
+ model-index:
8
+ - name: opus-mt-en-ru-finetuned
9
+ results: []
10
  ---
11
 
12
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
13
+ should probably proofread and complete it, then remove this comment. -->
14
+
15
+ # opus-mt-en-ru-finetuned
16
+
17
+ This model is a fine-tuned version of [kazandaev/opus-mt-en-ru-finetuned](https://huggingface.co/kazandaev/opus-mt-en-ru-finetuned) on the None dataset.
18
+ It achieves the following results on the evaluation set:
19
+ - Loss: 0.7763
20
+ - Bleu: 41.0065
21
+ - Gen Len: 29.7548
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 1e-06
41
+ - train_batch_size: 49
42
+ - eval_batch_size: 24
43
+ - seed: 42
44
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
45
+ - lr_scheduler_type: linear
46
+ - num_epochs: 5
47
 
48
+ ### Training results
 
 
49
 
50
+ | Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
51
+ |:-------------:|:-----:|:------:|:---------------:|:-------:|:-------:|
52
+ | 0.6903 | 1.0 | 35147 | 0.7779 | 40.9223 | 29.7846 |
53
+ | 0.6999 | 2.0 | 70294 | 0.7776 | 40.8267 | 29.8421 |
54
+ | 0.7257 | 3.0 | 105441 | 0.7769 | 40.8549 | 29.8765 |
55
+ | 0.7238 | 4.0 | 140588 | 0.7763 | 41.0225 | 29.7129 |
56
+ | 0.7313 | 5.0 | 175735 | 0.7763 | 41.0065 | 29.7548 |
57
 
 
58
 
59
+ ### Framework versions
 
 
 
 
 
 
 
 
 
60
 
61
+ - Transformers 4.16.2
62
+ - Pytorch 1.10.0+cu111
63
+ - Datasets 1.18.3
64
+ - Tokenizers 0.11.0
config.json CHANGED
@@ -1,4 +1,5 @@
1
  {
 
2
  "_num_labels": 3,
3
  "activation_dropout": 0.0,
4
  "activation_function": "swish",
@@ -15,6 +16,7 @@
15
  ],
16
  "bos_token_id": 0,
17
  "classif_dropout": 0.0,
 
18
  "d_model": 512,
19
  "decoder_attention_heads": 8,
20
  "decoder_ffn_dim": 2048,
@@ -27,6 +29,7 @@
27
  "encoder_layerdrop": 0.0,
28
  "encoder_layers": 6,
29
  "eos_token_id": 0,
 
30
  "id2label": {
31
  "0": "LABEL_0",
32
  "1": "LABEL_1",
@@ -49,5 +52,8 @@
49
  "pad_token_id": 62517,
50
  "scale_embedding": true,
51
  "static_position_embeddings": true,
 
 
 
52
  "vocab_size": 62518
53
- }
 
1
  {
2
+ "_name_or_path": "kazandaev/opus-mt-en-ru-finetuned",
3
  "_num_labels": 3,
4
  "activation_dropout": 0.0,
5
  "activation_function": "swish",
 
16
  ],
17
  "bos_token_id": 0,
18
  "classif_dropout": 0.0,
19
+ "classifier_dropout": 0.0,
20
  "d_model": 512,
21
  "decoder_attention_heads": 8,
22
  "decoder_ffn_dim": 2048,
 
29
  "encoder_layerdrop": 0.0,
30
  "encoder_layers": 6,
31
  "eos_token_id": 0,
32
+ "forced_eos_token_id": 0,
33
  "id2label": {
34
  "0": "LABEL_0",
35
  "1": "LABEL_1",
 
52
  "pad_token_id": 62517,
53
  "scale_embedding": true,
54
  "static_position_embeddings": true,
55
+ "torch_dtype": "float32",
56
+ "transformers_version": "4.16.2",
57
+ "use_cache": true,
58
  "vocab_size": 62518
59
+ }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d15fa58c6bc3efd3629c1b6b86d9aa6d15d2751a4620aa4cdd7eed7b5cbe583b
3
- size 306991893
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95d5b73e80184cb1e491b51f865ee90375f284df7915ff0f2d75dc8abb5d3a3d
3
+ size 304935301
runs/Feb22_18-58-37_22d008391efc/1645557026.79138/events.out.tfevents.1645557026.22d008391efc.125.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:224edd778c06ded1fea8a3fb5933a516a3ed7187173aa3eb7dfbc4fac3bb5966
3
+ size 5040
runs/Feb22_18-58-37_22d008391efc/events.out.tfevents.1645556822.22d008391efc.125.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c06040d2aebac088f396ba016775c518d58e93d87609f2135fba7b2445bb52a9
3
+ size 39427
runs/Feb23_14-20-03_035d423601e6/1645626676.839984/events.out.tfevents.1645626676.035d423601e6.127.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d0567fbd62b2b50865e0958b0ab22e9209839b08b2a7eeaccdc123c9e807f9f
3
+ size 5040
runs/Feb23_14-20-03_035d423601e6/events.out.tfevents.1645626471.035d423601e6.127.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3563a2c02152a7e96801d9298433005ce4ae8d42fbf2e5391de9d77c8c630c8c
3
+ size 39455
runs/Feb23_21-35-59_06f9a29b9e3d/1645652856.2353837/events.out.tfevents.1645652856.06f9a29b9e3d.10505.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc922fcad5527c10d57d84498cc3aba19564ce685aa1199bc0f4c8a61be4a3f2
3
+ size 5040
runs/Feb23_21-35-59_06f9a29b9e3d/events.out.tfevents.1645652648.06f9a29b9e3d.10505.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:333d2eabe805d5ad70955409489ec1d4d48db7cbf4a9c020878f04ba62e08f3a
3
+ size 120815
runs/Feb24_13-52-23_e7302c87fc7a/1645711475.5900414/events.out.tfevents.1645711475.e7302c87fc7a.128.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5486b7354a8e72d53e2f26b6b6eadddd5475ade5051f563c604f477d354d27b
3
+ size 5040
runs/Feb24_13-52-23_e7302c87fc7a/events.out.tfevents.1645711253.e7302c87fc7a.128.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:087ab09d429cf5eebe085c5427bc5b65b784c8e533f5d0f583e8e48f9a17dad5
3
+ size 120815
runs/Feb25_10-57-53_a509e49b085a/1645787388.3308783/events.out.tfevents.1645787388.a509e49b085a.125.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35363a9ac10d1d12aac132e5ee32dae00b108df07d1cb12aa9bf7658496f4aa5
3
+ size 5040
runs/Feb25_10-57-53_a509e49b085a/events.out.tfevents.1645787176.a509e49b085a.125.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32e6ff57c122dde347b74bcb7602250adb6a138e298f2dc4eac5625040aea7ee
3
+ size 120815
runs/Feb26_12-02-08_1592dd873767/1645877645.0248463/events.out.tfevents.1645877645.1592dd873767.128.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2042491bd90eea9c878a867e60b8df3f2b99d45c6f105a3900698166168ea7f
3
+ size 5040
runs/Feb26_12-02-08_1592dd873767/events.out.tfevents.1645877424.1592dd873767.128.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ade63fabc6bebe441c7b86e23110b0630cd517b461f83bb83c7c3493fc65b477
3
+ size 120815
runs/Feb27_06-04-51_db8c298e0d8d/1645942432.061396/events.out.tfevents.1645942432.db8c298e0d8d.125.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f54fe5c8954493e8a17e7c5dd2f04ed61144a115b3c337737550c65fd22d83cc
3
+ size 5040
runs/Feb27_06-04-51_db8c298e0d8d/events.out.tfevents.1645942432.db8c298e0d8d.125.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:353371490ea3f9cc2c3c17bd5b6d24cd68ac518d145655b3cef66a3e3279e709
3
+ size 62467
runs/Feb27_13-46-28_db8c298e0d8d/1645969998.99642/events.out.tfevents.1645969998.db8c298e0d8d.125.3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69702151a1bc19b34c2c11f74d0e92820febb5b0643f7ef3a19db65220a6359a
3
+ size 5040
runs/Feb27_13-46-28_db8c298e0d8d/events.out.tfevents.1645969779.db8c298e0d8d.125.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a702e2754764a59630b4c718da61067815e861923cf9ebee5ba3d8c40d44511b
3
+ size 62769
special_tokens_map.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
tokenizer_config.json CHANGED
@@ -1 +1 @@
1
- {"target_lang": "ru", "source_lang": "en"}
 
1
+ {"source_lang": "en", "target_lang": "ru", "unk_token": "<unk>", "eos_token": "</s>", "pad_token": "<pad>", "model_max_length": 512, "sp_model_kwargs": {}, "special_tokens_map_file": null, "tokenizer_file": null, "name_or_path": "kazandaev/opus-mt-en-ru-finetuned", "tokenizer_class": "MarianTokenizer"}
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2729fe5fbff174543bda1285eb3ee2a213c3c6e9e6a813bb8ae8fb17c78ebfc1
3
+ size 3183
vocab.json CHANGED
The diff for this file is too large to render. See raw diff