repo_id
stringlengths 4
122
| author
stringlengths 2
38
⌀ | model_type
stringlengths 2
33
⌀ | files_per_repo
int64 2
39k
| downloads_30d
int64 0
33.7M
| library
stringlengths 2
37
⌀ | likes
int64 0
4.87k
| pipeline
stringlengths 5
30
⌀ | pytorch
bool 2
classes | tensorflow
bool 2
classes | jax
bool 2
classes | license
stringlengths 2
33
⌀ | languages
stringlengths 2
1.63k
⌀ | datasets
stringlengths 2
2.58k
⌀ | co2
stringlengths 6
258
⌀ | prs_count
int64 0
125
| prs_open
int64 0
120
| prs_merged
int64 0
46
| prs_closed
int64 0
34
| discussions_count
int64 0
218
| discussions_open
int64 0
148
| discussions_closed
int64 0
70
| tags
stringlengths 2
513
| has_model_index
bool 2
classes | has_metadata
bool 2
classes | has_text
bool 1
class | text_length
int64 201
598k
| readme
stringlengths 0
598k
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Helsinki-NLP/opus-tatoeba-es-zh
|
Helsinki-NLP
|
marian
| 12 | 19 |
transformers
| 1 |
translation
| true | true | false |
apache-2.0
|
['es', 'zh']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['translation']
| false | true | true | 2,624 |
### es-zh
* source group: Spanish
* target group: Chinese
* OPUS readme: [spa-zho](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/spa-zho/README.md)
* model: transformer
* source language(s): spa
* target language(s): cjy_Hans cjy_Hant cmn cmn_Hans cmn_Hant hsn hsn_Hani lzh nan wuu yue_Hans yue_Hant
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* a sentence initial language token is required in the form of `>>id<<` (id = valid target language ID)
* download original weights: [opus-2021-01-04.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/spa-zho/opus-2021-01-04.zip)
* test set translations: [opus-2021-01-04.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/spa-zho/opus-2021-01-04.test.txt)
* test set scores: [opus-2021-01-04.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/spa-zho/opus-2021-01-04.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.spa.zho | 38.8 | 0.324 |
### System Info:
- hf_name: es-zh
- source_languages: spa
- target_languages: zho
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/spa-zho/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['es', 'zh']
- src_constituents: ('Spanish', {'spa'})
- tgt_constituents: ('Chinese', {'wuu_Bopo', 'wuu', 'cmn_Hang', 'lzh_Kana', 'lzh', 'wuu_Hani', 'lzh_Yiii', 'yue_Hans', 'cmn_Hani', 'cjy_Hans', 'cmn_Hans', 'cmn_Kana', 'zho_Hans', 'zho_Hant', 'yue', 'cmn_Bopo', 'yue_Hang', 'lzh_Hans', 'wuu_Latn', 'yue_Hant', 'hak_Hani', 'lzh_Bopo', 'cmn_Hant', 'lzh_Hani', 'lzh_Hang', 'cmn', 'lzh_Hira', 'yue_Bopo', 'yue_Hani', 'gan', 'zho', 'cmn_Yiii', 'yue_Hira', 'cmn_Latn', 'yue_Kana', 'cjy_Hant', 'cmn_Hira', 'nan_Hani', 'nan'})
- src_multilingual: False
- tgt_multilingual: False
- long_pair: spa-zho
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/spa-zho/opus-2021-01-04.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/spa-zho/opus-2021-01-04.test.txt
- src_alpha3: spa
- tgt_alpha3: zho
- chrF2_score: 0.324
- bleu: 38.8
- brevity_penalty: 0.878
- ref_len: 22762.0
- src_name: Spanish
- tgt_name: Chinese
- train_date: 2021-01-04 00:00:00
- src_alpha2: es
- tgt_alpha2: zh
- prefer_old: False
- short_pair: es-zh
- helsinki_git_sha: dfdcef114ffb8a8dbb7a3fcf84bde5af50309500
- transformers_git_sha: 1310e1a758edc8e89ec363db76863c771fbeb1de
- port_machine: LM0-400-22516.local
- port_time: 2021-01-04-18:53
|
Helsinki-NLP/opus-tatoeba-fi-en
|
Helsinki-NLP
|
marian
| 12 | 42 |
transformers
| 2 |
translation
| true | true | false |
apache-2.0
|
['fi', 'en']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['translation']
| false | true | true | 2,776 |
### fi-en
* source group: Finnish
* target group: English
* OPUS readme: [fin-eng](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/fin-eng/README.md)
* model: transformer-align
* source language(s): fin
* target language(s): eng
* model: transformer-align
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opusTCv20210807+bt-2021-08-25.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/fin-eng/opusTCv20210807+bt-2021-08-25.zip)
* test set translations: [opusTCv20210807+bt-2021-08-25.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/fin-eng/opusTCv20210807+bt-2021-08-25.test.txt)
* test set scores: [opusTCv20210807+bt-2021-08-25.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/fin-eng/opusTCv20210807+bt-2021-08-25.eval.txt)
## Benchmarks
| testset | BLEU | chr-F | #sent | #words | BP |
|---------|-------|-------|-------|--------|----|
| newsdev2015-enfi.fin-eng | 27.1 | 0.550 | 1500 | 32104 | 0.988 |
| newstest2015-enfi.fin-eng | 28.5 | 0.560 | 1370 | 27356 | 0.980 |
| newstest2016-enfi.fin-eng | 31.7 | 0.586 | 3000 | 63043 | 1.000 |
| newstest2017-enfi.fin-eng | 34.6 | 0.610 | 3002 | 61936 | 0.988 |
| newstest2018-enfi.fin-eng | 25.4 | 0.530 | 3000 | 62325 | 0.981 |
| newstest2019-fien.fin-eng | 30.6 | 0.577 | 1996 | 36227 | 0.994 |
| newstestB2016-enfi.fin-eng | 25.8 | 0.538 | 3000 | 63043 | 0.987 |
| newstestB2017-enfi.fin-eng | 29.6 | 0.572 | 3002 | 61936 | 0.999 |
| newstestB2017-fien.fin-eng | 29.6 | 0.572 | 3002 | 61936 | 0.999 |
| Tatoeba-test-v2021-08-07.fin-eng | 54.1 | 0.700 | 10000 | 75212 | 0.988 |
### System Info:
- hf_name: fi-en
- source_languages: fin
- target_languages: eng
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/fin-eng/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['fi', 'en']
- src_constituents: ('Finnish', {'fin'})
- tgt_constituents: ('English', {'eng'})
- src_multilingual: False
- tgt_multilingual: False
- long_pair: fin-eng
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/fin-eng/opusTCv20210807+bt-2021-08-25.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/fin-eng/opusTCv20210807+bt-2021-08-25.test.txt
- src_alpha3: fin
- tgt_alpha3: eng
- chrF2_score: 0.7
- bleu: 54.1
- src_name: Finnish
- tgt_name: English
- train_date: 2021-08-25 00:00:00
- src_alpha2: fi
- tgt_alpha2: en
- prefer_old: False
- short_pair: fi-en
- helsinki_git_sha: 2ef219d5b67f0afb0c6b732cd07001d84181f002
- transformers_git_sha: 12b4d66a80419db30a15e7b9d4208ceb9887c03b
- port_machine: LM0-400-22516.local
- port_time: 2021-11-04-21:36
|
Helsinki-NLP/opus-tatoeba-fr-it
|
Helsinki-NLP
|
marian
| 12 | 68 |
transformers
| 0 |
translation
| true | true | false |
apache-2.0
|
['fr', 'it']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['translation']
| false | true | true | 2,152 |
### fr-it
* source group: French
* target group: Italian
* OPUS readme: [fra-ita](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/fra-ita/README.md)
* model: transformer-align
* source language(s): fra
* target language(s): ita
* raw source language(s): fra
* raw target language(s): ita
* model: transformer-align
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opusTCv20210807-2021-11-11.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/fra-ita/opusTCv20210807-2021-11-11.zip)
* test set translations: [opusTCv20210807-2021-11-11.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/fra-ita/opusTCv20210807-2021-11-11.test.txt)
* test set scores: [opusTCv20210807-2021-11-11.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/fra-ita/opusTCv20210807-2021-11-11.eval.txt)
## Benchmarks
| testset | BLEU | chr-F | #sent | #words | BP |
|---------|-------|-------|-------|--------|----|
| Tatoeba-test-v2021-08-07.fra-ita | 54.8 | 0.737 | 10000 | 61517 | 0.953 |
### System Info:
- hf_name: fr-it
- source_languages: fra
- target_languages: ita
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/fra-ita/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['fr', 'it']
- src_constituents: ('French', {'fra'})
- tgt_constituents: ('Italian', {'ita'})
- src_multilingual: False
- tgt_multilingual: False
- long_pair: fra-ita
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/fra-ita/opusTCv20210807-2021-11-11.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/fra-ita/opusTCv20210807-2021-11-11.test.txt
- src_alpha3: fra
- tgt_alpha3: ita
- chrF2_score: 0.737
- bleu: 54.8
- src_name: French
- tgt_name: Italian
- train_date: 2021-11-11 00:00:00
- src_alpha2: fr
- tgt_alpha2: it
- prefer_old: False
- short_pair: fr-it
- helsinki_git_sha: 7ab0c987850187e0b10342bfc616cd47c027ba18
- transformers_git_sha: df1f94eb4a18b1a27d27e32040b60a17410d516e
- port_machine: LM0-400-22516.local
- port_time: 2021-11-11-19:40
|
Helsinki-NLP/opus-tatoeba-he-fr
|
Helsinki-NLP
|
marian
| 12 | 7 |
transformers
| 0 |
translation
| true | true | false |
apache-2.0
|
['he', 'fr']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['translation']
| false | true | true | 2,022 |
### he-fr
* source group: Hebrew
* target group: French
* OPUS readme: [heb-fra](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/heb-fra/README.md)
* model: transformer
* source language(s): heb
* target language(s): fra
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opus-2020-12-10.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/heb-fra/opus-2020-12-10.zip)
* test set translations: [opus-2020-12-10.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/heb-fra/opus-2020-12-10.test.txt)
* test set scores: [opus-2020-12-10.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/heb-fra/opus-2020-12-10.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.heb.fra | 47.3 | 0.644 |
### System Info:
- hf_name: he-fr
- source_languages: heb
- target_languages: fra
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/heb-fra/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['he', 'fr']
- src_constituents: ('Hebrew', {'heb'})
- tgt_constituents: ('French', {'fra'})
- src_multilingual: False
- tgt_multilingual: False
- long_pair: heb-fra
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/heb-fra/opus-2020-12-10.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/heb-fra/opus-2020-12-10.test.txt
- src_alpha3: heb
- tgt_alpha3: fra
- chrF2_score: 0.644
- bleu: 47.3
- brevity_penalty: 0.9740000000000001
- ref_len: 26123.0
- src_name: Hebrew
- tgt_name: French
- train_date: 2020-12-10 00:00:00
- src_alpha2: he
- tgt_alpha2: fr
- prefer_old: False
- short_pair: he-fr
- helsinki_git_sha: b317f78a3ec8a556a481b6a53dc70dc11769ca96
- transformers_git_sha: 1310e1a758edc8e89ec363db76863c771fbeb1de
- port_machine: LM0-400-22516.local
- port_time: 2020-12-11-16:03
|
Helsinki-NLP/opus-tatoeba-he-it
|
Helsinki-NLP
|
marian
| 11 | 8 |
transformers
| 0 |
translation
| true | false | false |
apache-2.0
|
['he', 'it']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['translation']
| false | true | true | 2,012 |
### he-it
* source group: Hebrew
* target group: Italian
* OPUS readme: [heb-ita](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/heb-ita/README.md)
* model: transformer
* source language(s): heb
* target language(s): ita
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opus-2020-12-10.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/heb-ita/opus-2020-12-10.zip)
* test set translations: [opus-2020-12-10.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/heb-ita/opus-2020-12-10.test.txt)
* test set scores: [opus-2020-12-10.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/heb-ita/opus-2020-12-10.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.heb.ita | 41.1 | 0.643 |
### System Info:
- hf_name: he-it
- source_languages: heb
- target_languages: ita
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/heb-ita/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['he', 'it']
- src_constituents: ('Hebrew', {'heb'})
- tgt_constituents: ('Italian', {'ita'})
- src_multilingual: False
- tgt_multilingual: False
- long_pair: heb-ita
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/heb-ita/opus-2020-12-10.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/heb-ita/opus-2020-12-10.test.txt
- src_alpha3: heb
- tgt_alpha3: ita
- chrF2_score: 0.643
- bleu: 41.1
- brevity_penalty: 0.997
- ref_len: 11464.0
- src_name: Hebrew
- tgt_name: Italian
- train_date: 2020-12-10 00:00:00
- src_alpha2: he
- tgt_alpha2: it
- prefer_old: False
- short_pair: he-it
- helsinki_git_sha: b317f78a3ec8a556a481b6a53dc70dc11769ca96
- transformers_git_sha: 1310e1a758edc8e89ec363db76863c771fbeb1de
- port_machine: LM0-400-22516.local
- port_time: 2020-12-11-16:01
|
Helsinki-NLP/opus-tatoeba-it-he
|
Helsinki-NLP
|
marian
| 12 | 7 |
transformers
| 0 |
translation
| true | true | false |
apache-2.0
|
['it', 'he']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['translation']
| false | true | true | 2,011 |
### it-he
* source group: Italian
* target group: Hebrew
* OPUS readme: [ita-heb](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ita-heb/README.md)
* model: transformer
* source language(s): ita
* target language(s): heb
* model: transformer
* pre-processing: normalization + SentencePiece (spm32k,spm32k)
* download original weights: [opus-2020-12-10.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/ita-heb/opus-2020-12-10.zip)
* test set translations: [opus-2020-12-10.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ita-heb/opus-2020-12-10.test.txt)
* test set scores: [opus-2020-12-10.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ita-heb/opus-2020-12-10.eval.txt)
## Benchmarks
| testset | BLEU | chr-F |
|-----------------------|-------|-------|
| Tatoeba-test.ita.heb | 38.5 | 0.593 |
### System Info:
- hf_name: it-he
- source_languages: ita
- target_languages: heb
- opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ita-heb/README.md
- original_repo: Tatoeba-Challenge
- tags: ['translation']
- languages: ['it', 'he']
- src_constituents: ('Italian', {'ita'})
- tgt_constituents: ('Hebrew', {'heb'})
- src_multilingual: False
- tgt_multilingual: False
- long_pair: ita-heb
- prepro: normalization + SentencePiece (spm32k,spm32k)
- url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/ita-heb/opus-2020-12-10.zip
- url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/ita-heb/opus-2020-12-10.test.txt
- src_alpha3: ita
- tgt_alpha3: heb
- chrF2_score: 0.593
- bleu: 38.5
- brevity_penalty: 0.985
- ref_len: 9796.0
- src_name: Italian
- tgt_name: Hebrew
- train_date: 2020-12-10 00:00:00
- src_alpha2: it
- tgt_alpha2: he
- prefer_old: False
- short_pair: it-he
- helsinki_git_sha: b317f78a3ec8a556a481b6a53dc70dc11769ca96
- transformers_git_sha: 1310e1a758edc8e89ec363db76863c771fbeb1de
- port_machine: LM0-400-22516.local
- port_time: 2020-12-11-16:02
|
HenryHXR/t5-base-finetuned-scitldr
|
HenryHXR
|
t5
| 21 | 3 |
transformers
| 0 |
text2text-generation
| true | false | false |
apache-2.0
| null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 1,523 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# t5-base-finetuned-scitldr
This model is a fine-tuned version of [t5-base](https://huggingface.co/t5-base) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 2.0232
- Rouge1: 35.2134
- Rouge2: 16.8919
- Rougel: 30.8442
- Rougelsum: 30.9316
- Gen Len: 18.7981
## 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: 3e-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Rouge1 | Rouge2 | Rougel | Rougelsum | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|:-------:|:---------:|:-------:|
| 2.0533 | 1.0 | 996 | 2.0285 | 34.9774 | 16.6163 | 30.6177 | 30.7038 | 18.7981 |
| 2.0994 | 2.0 | 1992 | 2.0232 | 35.2134 | 16.8919 | 30.8442 | 30.9316 | 18.7981 |
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.11.0
|
Herais/pred_genre
|
Herais
|
bert
| 8 | 4 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
|
['zh']
|
['Custom']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['classification']
| false | true | true | 1,718 |
This model predicts the time period given a synopsis of about 200 Chinese characters.
The model is trained on TV and Movie datasets and takes simplified Chinese as input.
We trained the model from the "hfl/chinese-bert-wwm-ext" checkpoint.
#### Sample Usage
from transformers import BertTokenizer, BertForSequenceClassification
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
checkpoint = "Herais/pred_genre"
tokenizer = BertTokenizer.from_pretrained(checkpoint,
problem_type="single_label_classification")
model = BertForSequenceClassification.from_pretrained(checkpoint).to(device)
label2id_genre = {'涉案': 7, '都市': 10, '革命': 12, '农村': 4, '传奇': 0,
'其它': 2, '传记': 1, '青少': 11, '军旅': 3, '武打': 6,
'科幻': 9, '神话': 8, '宫廷': 5}
id2label_genre = {7: '涉案', 10: '都市', 12: '革命', 4: '农村', 0: '传奇',
2: '其它', 1: '传记', 11: '青少', 3: '军旅', 6: '武打',
9: '科幻', 8: '神话', 5: '宫廷'}
synopsis = """加油吧！检察官。鲤州市安平区检察院检察官助理蔡晓与徐美津是两个刚入职场的“菜鸟”。\
他们在老检察官冯昆的指导与鼓励下，凭借着自己的一腔热血与对检察事业的执著追求，克服工作上的种种困难，\
成功办理电竞赌博、虚假诉讼、水产市场涉黑等一系列复杂案件，惩治了犯罪分子，维护了人民群众的合法权益，\
为社会主义法治建设贡献了自己的一份力量。在这个过程中，蔡晓与徐美津不仅得到了业务能力上的提升，\
也领悟了人生的真谛，学会真诚地面对家人与朋友，收获了亲情与友谊，成长为合格的员额检察官，\
继续为检察事业贡献自己的青春。 """
inputs = tokenizer(synopsis, truncation=True, max_length=512, return_tensors='pt')
model.eval()
outputs = model(**input)
label_ids_pred = torch.argmax(outputs.logits, dim=1).to('cpu').numpy()
labels_pred = [id2label_timeperiod[label] for label in labels_pred]
print(labels_pred)
# ['涉案']
Citation
TBA
|
Herais/pred_timeperiod
|
Herais
|
bert
| 8 | 4 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
|
['zh']
|
['Custom']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['classification']
| false | true | true | 1,487 |
This model predicts the time period given a synopsis of about 200 Chinese characters.
The model is trained on TV and Movie datasets and takes simplified Chinese as input.
We trained the model from the "hfl/chinese-bert-wwm-ext" checkpoint.
#### Sample Usage
from transformers import BertTokenizer, BertForSequenceClassification
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
checkpoint = "Herais/pred_timeperiod"
tokenizer = BertTokenizer.from_pretrained(checkpoint,
problem_type="single_label_classification")
model = BertForSequenceClassification.from_pretrained(checkpoint).to(device)
label2id_timeperiod = {'古代': 0, '当代': 1, '现代': 2, '近代': 3, '重大': 4}
id2label_timeperiod = {0: '古代', 1: '当代', 2: '现代', 3: '近代', 4: '重大'}
synopsis = """加油吧！检察官。鲤州市安平区检察院检察官助理蔡晓与徐美津是两个刚入职场的“菜鸟”。\
他们在老检察官冯昆的指导与鼓励下，凭借着自己的一腔热血与对检察事业的执著追求，克服工作上的种种困难，\
成功办理电竞赌博、虚假诉讼、水产市场涉黑等一系列复杂案件，惩治了犯罪分子，维护了人民群众的合法权益，\
为社会主义法治建设贡献了自己的一份力量。在这个过程中，蔡晓与徐美津不仅得到了业务能力上的提升，\
也领悟了人生的真谛，学会真诚地面对家人与朋友，收获了亲情与友谊，成长为合格的员额检察官，\
继续为检察事业贡献自己的青春。 """
inputs = tokenizer(synopsis, truncation=True, max_length=512, return_tensors='pt')
model.eval()
outputs = model(**input)
label_ids_pred = torch.argmax(outputs.logits, dim=1).to('cpu').numpy()
labels_pred = [id2label_timeperiod[label] for label in labels_pred]
print(labels_pred)
# ['当代']
Citation
{}
|
Hetarth/marian-finetuned-hi-hinglish
|
Hetarth
|
marian
| 9 | 3 |
transformers
| 0 |
text2text-generation
| false | true | false |
apache-2.0
| null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_keras_callback']
| true | true | true | 1,361 |
<!-- This model card has been generated automatically according to the information Keras had access to. You should
probably proofread and complete it, then remove this comment. -->
# marian-finetuned-hi-hinglish
This model is a fine-tuned version of [Helsinki-NLP/opus-mt-hi-en](https://huggingface.co/Helsinki-NLP/opus-mt-hi-en) on an unknown dataset.
It achieves the following results on the evaluation set:
- Train Loss: 4.1869
- Validation Loss: 4.0607
- Epoch: 0
## 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:
- optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 5e-05, 'decay_steps': 279, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
- training_precision: float32
### Training results
| Train Loss | Validation Loss | Epoch |
|:----------:|:---------------:|:-----:|
| 4.1869 | 4.0607 | 0 |
### Framework versions
- Transformers 4.16.2
- TensorFlow 2.7.0
- Datasets 1.18.3
- Tokenizers 0.11.0
|
HeyLucasLeao/byt5-base-pt-product-reviews
|
HeyLucasLeao
|
t5
| 6 | 6 |
transformers
| 2 |
text2text-generation
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 2,307 |
Create README.md
## ByT5 Base Portuguese Product Reviews
#### Model Description
This is a finetuned version from ByT5 Base by Google for Sentimental Analysis from Product Reviews in Portuguese.
##### Paper: https://arxiv.org/abs/2105.13626
#### Training data
It was trained from products reviews from a Americanas.com. You can found the data here: https://github.com/HeyLucasLeao/finetuning-byt5-model.
#### Training Procedure
It was finetuned using the Trainer Class available on the Hugging Face library. For evaluation it was used accuracy, precision, recall and f1 score.
##### Learning Rate: **1e-4**
##### Epochs: **1**
##### Colab for Finetuning: https://drive.google.com/file/d/17TcaN52moq7i7TE2EbcVbwQEQuAIQU63/view?usp=sharing
##### Colab for Metrics: https://colab.research.google.com/drive/1wbTDfOsE45UL8Q3ZD1_FTUmdVOKCcJFf#scrollTo=S4nuLkAFrlZ6
#### Score:
```python
Training Set:
'accuracy': 0.9019706922688226,
'f1': 0.9305820610687022,
'precision': 0.9596555965559656,
'recall': 0.9032183375781431
Test Set:
'accuracy': 0.9019409684035312,
'f1': 0.9303758732034697,
'precision': 0.9006660401258529,
'recall': 0.9621126145787866
Validation Set:
'accuracy': 0.9044948078526491,
'f1': 0.9321924443009364,
'precision': 0.9024426549173129,
'recall': 0.9639705531617191
```
#### Goals
My true intention was totally educational, thus making available a this version of the model as a example for future proposes.
How to use
``` python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
import torch
if torch.cuda.is_available():
device = torch.device('cuda')
else:
device = torch.device('cpu')
print(device)
tokenizer = AutoTokenizer.from_pretrained("HeyLucasLeao/byt5-base-pt-product-reviews")
model = AutoModelForSeq2SeqLM.from_pretrained("HeyLucasLeao/byt5-base-pt-product-reviews")
model.to(device)
def classificar_review(review):
inputs = tokenizer([review], padding='max_length', truncation=True, max_length=512, return_tensors='pt')
input_ids = inputs.input_ids.to(device)
attention_mask = inputs.attention_mask.to(device)
output = model.generate(input_ids, attention_mask=attention_mask)
pred = np.argmax(output.cpu(), axis=1)
dici = {0: 'Review Negativo', 1: 'Review Positivo'}
return dici[pred.item()]
classificar_review(review)
```
|
HeyLucasLeao/byt5-small-pt-product-reviews
|
HeyLucasLeao
|
t5
| 6 | 3 |
transformers
| 1 |
text2text-generation
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 2,301 |
Create README.md
## ByT5 Small Portuguese Product Reviews
#### Model Description
This is a finetuned version from ByT5 Small by Google for Sentimental Analysis from Product Reviews in Portuguese.
##### Paper: https://arxiv.org/abs/2105.13626
#### Training data
It was trained from products reviews from a Americanas.com. You can found the data here: https://github.com/HeyLucasLeao/finetuning-byt5-model.
#### Training Procedure
It was finetuned using the Trainer Class available on the Hugging Face library. For evaluation it was used accuracy, precision, recall and f1 score.
##### Learning Rate: **1e-4**
##### Epochs: **1**
##### Colab for Finetuning: https://colab.research.google.com/drive/1EChTeQkGeXi_52lClBNazHVuSNKEHN2f
##### Colab for Metrics: https://colab.research.google.com/drive/1o4tcsP3lpr1TobtE3Txhp9fllxPWXxlw#scrollTo=PXAoog5vQaTn
#### Score:
```python
Training Set:
'accuracy': 0.8974239585927603,
'f1': 0.927229848590765,
'precision': 0.9580290812115055,
'recall': 0.8983492356469835
Test Set:
'accuracy': 0.8957881282882026,
'f1': 0.9261366030421776,
'precision': 0.9559431131213848,
'recall': 0.8981326359661668
Validation Set:
'accuracy': 0.8925383190163382,
'f1': 0.9239208204149773,
'precision': 0.9525448733710351,
'recall': 0.8969668904839083
```
#### Goals
My true intention was totally educational, thus making available a this version of the model as a example for future proposes.
How to use
``` python
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
import torch
if torch.cuda.is_available():
device = torch.device('cuda')
else:
device = torch.device('cpu')
print(device)
tokenizer = AutoTokenizer.from_pretrained("HeyLucasLeao/byt5-small-pt-product-reviews")
model = AutoModelForSeq2SeqLM.from_pretrained("HeyLucasLeao/byt5-small-pt-product-reviews")
model.to(device)
def classificar_review(review):
inputs = tokenizer([review], padding='max_length', truncation=True, max_length=512, return_tensors='pt')
input_ids = inputs.input_ids.to(device)
attention_mask = inputs.attention_mask.to(device)
output = model.generate(input_ids, attention_mask=attention_mask)
pred = np.argmax(output.cpu(), axis=1)
dici = {0: 'Review Negativo', 1: 'Review Positivo'}
return dici[pred.item()]
classificar_review(review)
```
|
HeyLucasLeao/gpt-neo-small-emo-lyrics
|
HeyLucasLeao
|
gpt_neo
| 10 | 5 |
transformers
| 0 |
text-generation
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 4,424 |
Create README.md
## Emo Bot
#### Model Description
This is a finetuned version from GPT-Neo-125M for Generating Music Lyrics by Emo Genre.
#### Training data
It was trained with 2381 songs by 15 bands that were important to emo culture in the early 2000s, not necessary directly playing on the genre.
#### Training Procedure
It was finetuned using the Trainer Class available on the Hugging Face library.
##### Learning Rate: **2e-4**
##### Epochs: **40**
##### Colab for Finetuning: https://colab.research.google.com/drive/1jwTYI1AygQf7FV9vCHTWA4Gf5i--sjsD?usp=sharing
##### Colab for Testing: https://colab.research.google.com/drive/1wSP4Wyr1-DTTNQbQps_RCO3ThhH-eeZc?usp=sharing
#### Goals
My true intention was totally educational, thus making available a this version of the model as a example for future proposes.
How to use
``` python
from transformers import AutoTokenizer, AutoModelForCausalLM
import re
if torch.cuda.is_available():
device = torch.device('cuda')
else:
device = torch.device('cpu')
print(device)
tokenizer = AutoTokenizer.from_pretrained("HeyLucasLeao/gpt-neo-small-emo-lyrics")
model = AutoModelForCausalLM.from_pretrained("HeyLucasLeao/gpt-neo-small-emo-lyrics")
model.to('cuda')
generated = tokenizer('I miss you',return_tensors='pt').input_ids.cuda()
#Generating texts
sample_outputs = model.generate(generated,
# Use sampling instead of greedy decoding
do_sample=True,
# Keep only top 3 token with the highest probability
top_k=10,
# Maximum sequence length
max_length=200,
# Keep only the most probable tokens with cumulative probability of 95%
top_p=0.95,
# Changes randomness of generated sequences
temperature=2.,
# Number of sequences to generate
num_return_sequences=3)
# Decoding and printing sequences
for i, sample_output in enumerate(sample_outputs):
texto = tokenizer.decode(sample_output.tolist())
regex_padding = re.sub('<|pad|>', '', texto)
regex_barra = re.sub('[|+]', '', regex_padding)
espaço = re.sub('[ +]', ' ', regex_barra)
resultado = re.sub('[\n](2, )', '\n', espaço)
print(">> Text {}: {}".format(i+1, resultado + '\n'))
""">> Texto 1: I miss you
I miss you more than anything
And if you change your mind
I do it like a change of mind
I always do it like theeah
Everybody wants a surprise
Everybody needs to stay collected
I keep your locked and numbered
Use this instead: Run like the wind
Use this instead: Run like the sun
And come back down: You've been replaced
Don't want to be the same
Tomorrow
I don't even need your name
The message is on the way
make it while you're holding on
It's better than it is
Everything more security than a parade
Im getting security
angs the world like a damned soul
We're hanging on a queue
and the truth is on the way
Are you listening?
We're getting security
Send me your soldiers
We're getting blood on"""
""">> Texto 2: I miss you
And I could forget your name
All the words we'd hear
You miss me
I need you
And I need you
You were all by my side
When we'd talk to no one
And I
Just to talk to you
It's easier than it has to be
Except for you
You missed my know-all
You meant to hug me
And I
Just want to feel you touch me
We'll work up
Something wild, just from the inside
Just get closer to me
I need you
You were all by my side
When we*d talk to you
, you better admit
That I'm too broken to be small
You're part of me
And I need you
But I
Don't know how
But I know I need you
Must"""
""">> Texto 3: I miss you
And I can't lie
Inside my head
All the hours you've been through
If I could change your mind
I would give it all away
And I'd give it all away
Just to give it away
To you
Now I wish that I could change
Just to you
I miss you so much
If I could change
So much
I'm looking down
At the road
The one that's already been
Searching for a better way to go
So much I need to see it clear
topk wish me an ehive
I wish I wish I wish I knew
I can give well
In this lonely night
The lonely night
I miss you
I wish it well
If I could change
So much
I need you"""
```
|
HeyLucasLeao/gpt-neo-small-portuguese
|
HeyLucasLeao
|
gpt_neo
| 10 | 149 |
transformers
| 1 |
text-generation
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 1 | 1 | 0 |
[]
| false | false | true | 3,771 |
## GPT-Neo Small Portuguese
#### Model Description
This is a finetuned version from GPT-Neo 125M by EletheurAI to Portuguese language.
#### Training data
It was trained from 227,382 selected texts from a PTWiki Dump. You can found all the data from here: https://archive.org/details/ptwiki-dump-20210520
#### Training Procedure
Every text was passed through a GPT2-Tokenizer with bos and eos tokens to separate them, with max sequence length that the GPT-Neo could support. It was finetuned using the default metrics of the Trainer Class, available on the Hugging Face library.
##### Learning Rate: **2e-4**
##### Epochs: **1**
#### Goals
My true intention was totally educational, thus making available a Portuguese version of this model.
How to use
``` python
from transformers import AutoTokenizer, AutoModelForCausalLM
tokenizer = AutoTokenizer.from_pretrained("HeyLucasLeao/gpt-neo-small-portuguese")
model = AutoModelForCausalLM.from_pretrained("HeyLucasLeao/gpt-neo-small-portuguese")
text = 'eu amo o brasil.'
generated = tokenizer(f'<|startoftext|> {text}',
return_tensors='pt').input_ids.cuda()
#Generating texts
sample_outputs = model.generate(generated,
# Use sampling instead of greedy decoding
do_sample=True,
# Keep only top 3 token with the highest probability
top_k=3,
# Maximum sequence length
max_length=200,
# Keep only the most probable tokens with cumulative probability of 95%
top_p=0.95,
# Changes randomness of generated sequences
temperature=1.9,
# Number of sequences to generate
num_return_sequences=3)
# Decoding and printing sequences
for i, sample_output in enumerate(sample_outputs):
print(">> Generated text {}\\\\
\\\\
{}".format(i+1, tokenizer.decode(sample_output.tolist())))
# >> Generated text
#Setting `pad_token_id` to `eos_token_id`:50256 for open-end generation.
#>> Generated text 1
#<|startoftext|> eu amo o brasil. O termo foi usado por alguns autores como uma forma de designar a formação do poder político do Brasil. A partir da década de 1960, o termo passou a ser usado para designar a formação política do Brasil. A partir de meados da década de 1970 e até o inicio dos anos 2000, o termo foi aplicado à formação político-administrativo do país, sendo utilizado por alguns autores como uma expressão de "política de direita". História Antecedentes O termo "político-administrário" foi usado pela primeira vez em 1891 por um gru
#>> Generated text 2
#<|startoftext|> eu amo o brasil. É uma das muitas pessoas do mundo, ao contrário da maioria das pessoas, que são chamados de "pessoas do Brasil", que são chamados de "brincos do país" e que têm uma carreira de mais de um século. O termo "brincal de ouro" é usado em referências às pessoas que vivem no Brasil, e que são chamados "brincos do país", que são "cidade" e que vivem na cidade de Nova York e que vive em um país onde a maior parte das pessoas são chamados de "cidades". Hist
#>> Generated text 3
#<|startoftext|> eu amo o brasil. É uma expressão que se refere ao uso de um instrumento musical em particular para se referir à qualidade musical, o que é uma expressão da qualidade da qualidade musical de uma pessoa. A expressão "amor" (em inglês, amo), é a expressão que pode ser usada com o intuito empregado em qualquer situação em que a vontade de uma pessoa de se sentir amado ou amoroso é mais do que um desejo de uma vontade. Em geral, a expressão "amoro" (do inglês, amo) pode também se referir tanto a uma pessoa como um instrumento de cordas ou de uma
```
|
HfSpeechUtils/convert_wav2vec2_to_hf
|
HfSpeechUtils
| null | 4 | 0 | null | 2 | null | false | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 821 |
# Convert Fairseq Wav2Vec2 to HF
This repo has two scripts that can show how to convert a fairseq checkpoint to HF Transformers.
It's important to always check in a forward pass that the two checkpoints are the same. The procedure should be as follows:
1. Download original model
2. Create HF version of the model:
```
huggingface-cli repo create <name_of_model> --organization <org_of_model>
git clone https://huggingface.co/<org_of_model>/<name_of_model>
```
3. Convert the model
```
./run_convert.sh <name_of_model> <path/to/orig/checkpoint/> 0
```
The "0" means that checkpoint is **not** a fine-tuned one.
4. Verify that models are equal:
```
./run_forward.py <name_of_model> <path/to/orig/checkpoint/> 0
```
Check the scripts to better understand how they work or contact https://huggingface.co/patrickvonplaten
|
Hinova/distilbert-base-uncased-finetuned-cola
|
Hinova
|
distilbert
| 10 | 3 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
| null |
['glue']
| null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| false | true | true | 1,564 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-cola
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8301
- Matthews Correlation: 0.5481
## 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: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|
| 0.5252 | 1.0 | 535 | 0.5094 | 0.4268 |
| 0.3515 | 2.0 | 1070 | 0.5040 | 0.4948 |
| 0.2403 | 3.0 | 1605 | 0.5869 | 0.5449 |
| 0.1731 | 4.0 | 2140 | 0.7338 | 0.5474 |
| 0.1219 | 5.0 | 2675 | 0.8301 | 0.5481 |
### Framework versions
- Transformers 4.9.2
- Pytorch 1.9.0
- Datasets 1.11.0
- Tokenizers 0.10.3
|
Hoang/distilbert-base-uncased-finetuned-squad
|
Hoang
|
distilbert
| 18 | 6 |
transformers
| 0 |
question-answering
| true | false | false |
apache-2.0
| null |
['squad']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| false | true | true | 1,284 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-squad
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the squad dataset.
It achieves the following results on the evaluation set:
- Loss: 1.1582
## 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: 3
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:-----:|:---------------:|
| 1.2176 | 1.0 | 5533 | 1.1429 |
| 0.9425 | 2.0 | 11066 | 1.1196 |
| 0.7586 | 3.0 | 16599 | 1.1582 |
### Framework versions
- Transformers 4.10.0
- Pytorch 1.9.0+cu102
- Datasets 1.11.0
- Tokenizers 0.10.3
|
Holako/NER_model_holako
|
Holako
|
xlm-roberta
| 8 | 7 |
transformers
| 0 |
token-classification
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 998 |
#### How to use
You can use this model with Transformers *pipeline* for NER.
```python
from transformers import AutoTokenizer, AutoModelForTokenClassification
from transformers import pipeline
tokenizer = AutoTokenizer.from_pretrained("Holako/NER_model_holako")
model = AutoModelForTokenClassification.from_pretrained("Holako/NER_model_holako")
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
example = "اسمي احمد"
ner_results = nlp(example)
print(ner_results)
```
#### Limitations and bias
This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.
=======
#### Limitations and bias
This model is limited by its training dataset of entity-annotated news articles from a specific span of time. This may not generalize well for all use cases in different domains.
## Training data
Language|Dataset
-|-
Arabic | [ANERcorp](https://camel.abudhabi.nyu.edu/anercorp/)
|
HooshvareLab/albert-fa-zwnj-base-v2-ner
|
HooshvareLab
|
albert
| 9 | 18 |
transformers
| 0 |
token-classification
| true | true | false | null |
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,671 |
# AlbertNER
This model fine-tuned for the Named Entity Recognition (NER) task on a mixed NER dataset collected from [ARMAN](https://github.com/HaniehP/PersianNER), [PEYMA](http://nsurl.org/2019-2/tasks/task-7-named-entity-recognition-ner-for-farsi/), and [WikiANN](https://elisa-ie.github.io/wikiann/) that covered ten types of entities:
- Date (DAT)
- Event (EVE)
- Facility (FAC)
- Location (LOC)
- Money (MON)
- Organization (ORG)
- Percent (PCT)
- Person (PER)
- Product (PRO)
- Time (TIM)
## Dataset Information
| | Records | B-DAT | B-EVE | B-FAC | B-LOC | B-MON | B-ORG | B-PCT | B-PER | B-PRO | B-TIM | I-DAT | I-EVE | I-FAC | I-LOC | I-MON | I-ORG | I-PCT | I-PER | I-PRO | I-TIM |
|:------|----------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|
| Train | 29133 | 1423 | 1487 | 1400 | 13919 | 417 | 15926 | 355 | 12347 | 1855 | 150 | 1947 | 5018 | 2421 | 4118 | 1059 | 19579 | 573 | 7699 | 1914 | 332 |
| Valid | 5142 | 267 | 253 | 250 | 2362 | 100 | 2651 | 64 | 2173 | 317 | 19 | 373 | 799 | 387 | 717 | 270 | 3260 | 101 | 1382 | 303 | 35 |
| Test | 6049 | 407 | 256 | 248 | 2886 | 98 | 3216 | 94 | 2646 | 318 | 43 | 568 | 888 | 408 | 858 | 263 | 3967 | 141 | 1707 | 296 | 78 |
## Evaluation
The following tables summarize the scores obtained by model overall and per each class.
**Overall**
| Model | accuracy | precision | recall | f1 |
|:----------:|:--------:|:---------:|:--------:|:--------:|
| Albert | 0.993405 | 0.938907 | 0.943966 | 0.941429 |
**Per entities**
| | number | precision | recall | f1 |
|:---: |:------: |:---------: |:--------: |:--------: |
| DAT | 407 | 0.820639 | 0.820639 | 0.820639 |
| EVE | 256 | 0.936803 | 0.984375 | 0.960000 |
| FAC | 248 | 0.925373 | 1.000000 | 0.961240 |
| LOC | 2884 | 0.960818 | 0.960818 | 0.960818 |
| MON | 98 | 0.913978 | 0.867347 | 0.890052 |
| ORG | 3216 | 0.920892 | 0.937500 | 0.929122 |
| PCT | 94 | 0.946809 | 0.946809 | 0.946809 |
| PER | 2644 | 0.960000 | 0.944024 | 0.951945 |
| PRO | 318 | 0.942943 | 0.987421 | 0.964670 |
| TIM | 43 | 0.780488 | 0.744186 | 0.761905 |
## How To Use
You use this model with Transformers pipeline for NER.
### Installing requirements
```bash
pip install sentencepiece
pip install transformers
```
### How to predict using pipeline
```python
from transformers import AutoTokenizer
from transformers import AutoModelForTokenClassification # for pytorch
from transformers import TFAutoModelForTokenClassification # for tensorflow
from transformers import pipeline
model_name_or_path = "HooshvareLab/albert-fa-zwnj-base-v2-ner" # Albert
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
model = AutoModelForTokenClassification.from_pretrained(model_name_or_path) # Pytorch
# model = TFAutoModelForTokenClassification.from_pretrained(model_name_or_path) # Tensorflow
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
example = "در سال ۲۰۱۳ درگذشت و آندرتیکر و کین برای او مراسم یادبود گرفتند."
ner_results = nlp(example)
print(ner_results)
```
## Questions?
Post a Github issue on the [ParsNER Issues](https://github.com/hooshvare/parsner/issues) repo.
|
HooshvareLab/albert-fa-zwnj-base-v2
|
HooshvareLab
|
albert
| 11 | 374 |
transformers
| 0 |
fill-mask
| true | true | false |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 698 |
# ALBERT-Persian
A Lite BERT for Self-supervised Learning of Language Representations for the Persian Language
> میتونی بهش بگی برت_کوچولو
> Call it little_berty
### BibTeX entry and citation info
Please cite in your publication as the following:
```bibtex
@misc{ALBERTPersian,
author = {Hooshvare Team},
title = {ALBERT-Persian: A Lite BERT for Self-supervised Learning of Language Representations for the Persian Language},
year = {2021},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/m3hrdadfi/albert-persian}},
}
```
## Questions?
Post a Github issue on the [ALBERT-Persian](https://github.com/m3hrdadfi/albert-persian) repo.
|
HooshvareLab/bert-base-parsbert-armanner-uncased
|
HooshvareLab
|
bert
| 13 | 1,067 |
transformers
| 0 |
token-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 4,673 |
## ParsBERT: Transformer-based Model for Persian Language Understanding
ParsBERT is a monolingual language model based on Google’s BERT architecture with the same configurations as BERT-Base.
Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515)
All the models (downstream tasks) are uncased and trained with whole word masking. (coming soon stay tuned)
## Persian NER [ARMAN, PEYMA, ARMAN+PEYMA]
This task aims to extract named entities in the text, such as names and label with appropriate `NER` classes such as locations, organizations, etc. The datasets used for this task contain sentences that are marked with `IOB` format. In this format, tokens that are not part of an entity are tagged as `”O”` the `”B”`tag corresponds to the first word of an object, and the `”I”` tag corresponds to the rest of the terms of the same entity. Both `”B”` and `”I”` tags are followed by a hyphen (or underscore), followed by the entity category. Therefore, the NER task is a multi-class token classification problem that labels the tokens upon being fed a raw text. There are two primary datasets used in Persian NER, `ARMAN`, and `PEYMA`. In ParsBERT, we prepared ner for both datasets as well as a combination of both datasets.
### ARMAN
ARMAN dataset holds 7,682 sentences with 250,015 sentences tagged over six different classes.
1. Organization
2. Location
3. Facility
4. Event
5. Product
6. Person
| Label | # |
|:------------:|:-----:|
| Organization | 30108 |
| Location | 12924 |
| Facility | 4458 |
| Event | 7557 |
| Product | 4389 |
| Person | 15645 |
**Download**
You can download the dataset from [here](https://github.com/HaniehP/PersianNER)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|---------|----------|------------|--------------|----------|----------------|------------|
| ARMAN | 93.10* | 89.9 | 84.03 | 86.55 | - | 77.45 |
## How to use :hugs:
| Notebook | Description | |
|:----------|:-------------|------:|
| [How to use Pipelines](https://github.com/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) | Simple and efficient way to use State-of-the-Art models on downstream tasks through transformers | [](https://colab.research.google.com/github/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) |
## Cite
Please cite the following paper in your publication if you are using [ParsBERT](https://arxiv.org/abs/2005.12515) in your research:
```markdown
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Acknowledgments
We hereby, express our gratitude to the [Tensorflow Research Cloud (TFRC) program](https://tensorflow.org/tfrc) for providing us with the necessary computation resources. We also thank [Hooshvare](https://hooshvare.com) Research Group for facilitating dataset gathering and scraping online text resources.
## Contributors
- Mehrdad Farahani: [Linkedin](https://www.linkedin.com/in/m3hrdadfi/), [Twitter](https://twitter.com/m3hrdadfi), [Github](https://github.com/m3hrdadfi)
- Mohammad Gharachorloo: [Linkedin](https://www.linkedin.com/in/mohammad-gharachorloo/), [Twitter](https://twitter.com/MGharachorloo), [Github](https://github.com/baarsaam)
- Marzieh Farahani: [Linkedin](https://www.linkedin.com/in/marziehphi/), [Twitter](https://twitter.com/marziehphi), [Github](https://github.com/marziehphi)
- Mohammad Manthouri: [Linkedin](https://www.linkedin.com/in/mohammad-manthouri-aka-mansouri-07030766/), [Twitter](https://twitter.com/mmanthouri), [Github](https://github.com/mmanthouri)
- Hooshvare Team: [Official Website](https://hooshvare.com/), [Linkedin](https://www.linkedin.com/company/hooshvare), [Twitter](https://twitter.com/hooshvare), [Github](https://github.com/hooshvare), [Instagram](https://www.instagram.com/hooshvare/)
+ And a special thanks to Sara Tabrizi for her fantastic poster design. Follow her on: [Linkedin](https://www.linkedin.com/in/sara-tabrizi-64548b79/), [Behance](https://www.behance.net/saratabrizi), [Instagram](https://www.instagram.com/sara_b_tabrizi/)
## Releases
### Release v0.1 (May 29, 2019)
This is the first version of our ParsBERT NER!
|
HooshvareLab/bert-base-parsbert-ner-uncased
|
HooshvareLab
|
bert
| 13 | 756 |
transformers
| 0 |
token-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 5,521 |
## ParsBERT: Transformer-based Model for Persian Language Understanding
ParsBERT is a monolingual language model based on Google’s BERT architecture with the same configurations as BERT-Base.
Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515)
All the models (downstream tasks) are uncased and trained with whole word masking. (coming soon stay tuned)
## Persian NER [ARMAN, PEYMA, ARMAN+PEYMA]
This task aims to extract named entities in the text, such as names and label with appropriate `NER` classes such as locations, organizations, etc. The datasets used for this task contain sentences that are marked with `IOB` format. In this format, tokens that are not part of an entity are tagged as `”O”` the `”B”`tag corresponds to the first word of an object, and the `”I”` tag corresponds to the rest of the terms of the same entity. Both `”B”` and `”I”` tags are followed by a hyphen (or underscore), followed by the entity category. Therefore, the NER task is a multi-class token classification problem that labels the tokens upon being fed a raw text. There are two primary datasets used in Persian NER, `ARMAN`, and `PEYMA`. In ParsBERT, we prepared ner for both datasets as well as a combination of both datasets.
### PEYMA
PEYMA dataset includes 7,145 sentences with a total of 302,530 tokens from which 41,148 tokens are tagged with seven different classes.
1. Organization
2. Money
3. Location
4. Date
5. Time
6. Person
7. Percent
| Label | # |
|:------------:|:-----:|
| Organization | 16964 |
| Money | 2037 |
| Location | 8782 |
| Date | 4259 |
| Time | 732 |
| Person | 7675 |
| Percent | 699 |
**Download**
You can download the dataset from [here](http://nsurl.org/tasks/task-7-named-entity-recognition-ner-for-farsi/)
---
### ARMAN
ARMAN dataset holds 7,682 sentences with 250,015 sentences tagged over six different classes.
1. Organization
2. Location
3. Facility
4. Event
5. Product
6. Person
| Label | # |
|:------------:|:-----:|
| Organization | 30108 |
| Location | 12924 |
| Facility | 4458 |
| Event | 7557 |
| Product | 4389 |
| Person | 15645 |
**Download**
You can download the dataset from [here](https://github.com/HaniehP/PersianNER)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|:---------------:|:--------:|:----------:|:--------------:|:----------:|:----------------:|:------------:|
| ARMAN + PEYMA | 95.13* | - | - | - | - | - |
| PEYMA | 98.79* | - | 90.59 | - | 84.00 | - |
| ARMAN | 93.10* | 89.9 | 84.03 | 86.55 | - | 77.45 |
## How to use :hugs:
| Notebook | Description | |
|:----------|:-------------|------:|
| [How to use Pipelines](https://github.com/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) | Simple and efficient way to use State-of-the-Art models on downstream tasks through transformers | [](https://colab.research.google.com/github/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) |
## Cite
Please cite the following paper in your publication if you are using [ParsBERT](https://arxiv.org/abs/2005.12515) in your research:
```markdown
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Acknowledgments
We hereby, express our gratitude to the [Tensorflow Research Cloud (TFRC) program](https://tensorflow.org/tfrc) for providing us with the necessary computation resources. We also thank [Hooshvare](https://hooshvare.com) Research Group for facilitating dataset gathering and scraping online text resources.
## Contributors
- Mehrdad Farahani: [Linkedin](https://www.linkedin.com/in/m3hrdadfi/), [Twitter](https://twitter.com/m3hrdadfi), [Github](https://github.com/m3hrdadfi)
- Mohammad Gharachorloo: [Linkedin](https://www.linkedin.com/in/mohammad-gharachorloo/), [Twitter](https://twitter.com/MGharachorloo), [Github](https://github.com/baarsaam)
- Marzieh Farahani: [Linkedin](https://www.linkedin.com/in/marziehphi/), [Twitter](https://twitter.com/marziehphi), [Github](https://github.com/marziehphi)
- Mohammad Manthouri: [Linkedin](https://www.linkedin.com/in/mohammad-manthouri-aka-mansouri-07030766/), [Twitter](https://twitter.com/mmanthouri), [Github](https://github.com/mmanthouri)
- Hooshvare Team: [Official Website](https://hooshvare.com/), [Linkedin](https://www.linkedin.com/company/hooshvare), [Twitter](https://twitter.com/hooshvare), [Github](https://github.com/hooshvare), [Instagram](https://www.instagram.com/hooshvare/)
+ And a special thanks to Sara Tabrizi for her fantastic poster design. Follow her on: [Linkedin](https://www.linkedin.com/in/sara-tabrizi-64548b79/), [Behance](https://www.behance.net/saratabrizi), [Instagram](https://www.instagram.com/sara_b_tabrizi/)
## Releases
### Release v0.1 (May 29, 2019)
This is the first version of our ParsBERT NER!
|
HooshvareLab/bert-base-parsbert-peymaner-uncased
|
HooshvareLab
|
bert
| 13 | 46 |
transformers
| 0 |
token-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 4,776 |
## ParsBERT: Transformer-based Model for Persian Language Understanding
ParsBERT is a monolingual language model based on Google’s BERT architecture with the same configurations as BERT-Base.
Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515)
All the models (downstream tasks) are uncased and trained with whole word masking. (coming soon stay tuned)
## Persian NER [ARMAN, PEYMA, ARMAN+PEYMA]
This task aims to extract named entities in the text, such as names and label with appropriate `NER` classes such as locations, organizations, etc. The datasets used for this task contain sentences that are marked with `IOB` format. In this format, tokens that are not part of an entity are tagged as `”O”` the `”B”`tag corresponds to the first word of an object, and the `”I”` tag corresponds to the rest of the terms of the same entity. Both `”B”` and `”I”` tags are followed by a hyphen (or underscore), followed by the entity category. Therefore, the NER task is a multi-class token classification problem that labels the tokens upon being fed a raw text. There are two primary datasets used in Persian NER, `ARMAN`, and `PEYMA`. In ParsBERT, we prepared ner for both datasets as well as a combination of both datasets.
### PEYMA
PEYMA dataset includes 7,145 sentences with a total of 302,530 tokens from which 41,148 tokens are tagged with seven different classes.
1. Organization
2. Money
3. Location
4. Date
5. Time
6. Person
7. Percent
| Label | # |
|:------------:|:-----:|
| Organization | 16964 |
| Money | 2037 |
| Location | 8782 |
| Date | 4259 |
| Time | 732 |
| Person | 7675 |
| Percent | 699 |
**Download**
You can download the dataset from [here](http://nsurl.org/tasks/task-7-named-entity-recognition-ner-for-farsi/)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|---------|----------|------------|--------------|----------|----------------|------------|
| PEYMA | 98.79* | - | 90.59 | - | 84.00 | - |
## How to use :hugs:
| Notebook | Description | |
|:----------|:-------------|------:|
| [How to use Pipelines](https://github.com/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) | Simple and efficient way to use State-of-the-Art models on downstream tasks through transformers | [](https://colab.research.google.com/github/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) |
## Cite
Please cite the following paper in your publication if you are using [ParsBERT](https://arxiv.org/abs/2005.12515) in your research:
```markdown
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Acknowledgments
We hereby, express our gratitude to the [Tensorflow Research Cloud (TFRC) program](https://tensorflow.org/tfrc) for providing us with the necessary computation resources. We also thank [Hooshvare](https://hooshvare.com) Research Group for facilitating dataset gathering and scraping online text resources.
## Contributors
- Mehrdad Farahani: [Linkedin](https://www.linkedin.com/in/m3hrdadfi/), [Twitter](https://twitter.com/m3hrdadfi), [Github](https://github.com/m3hrdadfi)
- Mohammad Gharachorloo: [Linkedin](https://www.linkedin.com/in/mohammad-gharachorloo/), [Twitter](https://twitter.com/MGharachorloo), [Github](https://github.com/baarsaam)
- Marzieh Farahani: [Linkedin](https://www.linkedin.com/in/marziehphi/), [Twitter](https://twitter.com/marziehphi), [Github](https://github.com/marziehphi)
- Mohammad Manthouri: [Linkedin](https://www.linkedin.com/in/mohammad-manthouri-aka-mansouri-07030766/), [Twitter](https://twitter.com/mmanthouri), [Github](https://github.com/mmanthouri)
- Hooshvare Team: [Official Website](https://hooshvare.com/), [Linkedin](https://www.linkedin.com/company/hooshvare), [Twitter](https://twitter.com/hooshvare), [Github](https://github.com/hooshvare), [Instagram](https://www.instagram.com/hooshvare/)
+ And a special thanks to Sara Tabrizi for her fantastic poster design. Follow her on: [Linkedin](https://www.linkedin.com/in/sara-tabrizi-64548b79/), [Behance](https://www.behance.net/saratabrizi), [Instagram](https://www.instagram.com/sara_b_tabrizi/)
## Releases
### Release v0.1 (May 29, 2019)
This is the first version of our ParsBERT NER!
|
HooshvareLab/bert-base-parsbert-uncased
|
HooshvareLab
|
bert
| 7 | 76,301 |
transformers
| 11 |
fill-mask
| true | true | true | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 5,818 |
## ParsBERT: Transformer-based Model for Persian Language Understanding
ParsBERT is a monolingual language model based on Google’s BERT architecture with the same configurations as BERT-Base.
Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515)
All the models (downstream tasks) are uncased and trained with whole word masking. (coming soon stay tuned)
---
## Introduction
This model is pre-trained on a large Persian corpus with various writing styles from numerous subjects (e.g., scientific, novels, news) with more than 2M documents. A large subset of this corpus was crawled manually.
As a part of ParsBERT methodology, an extensive pre-processing combining POS tagging and WordPiece segmentation was carried out to bring the corpus into a proper format. This process produces more than 40M true sentences.
## Evaluation
ParsBERT is evaluated on three NLP downstream tasks: Sentiment Analysis (SA), Text Classification, and Named Entity Recognition (NER). For this matter and due to insufficient resources, two large datasets for SA and two for text classification were manually composed, which are available for public use and benchmarking. ParsBERT outperformed all other language models, including multilingual BERT and other hybrid deep learning models for all tasks, improving the state-of-the-art performance in Persian language modeling.
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
### Sentiment Analysis (SA) task
| Dataset | ParsBERT | mBERT | DeepSentiPers |
|:--------------------------:|:---------:|:-----:|:-------------:|
| Digikala User Comments | 81.74* | 80.74 | - |
| SnappFood User Comments | 88.12* | 87.87 | - |
| SentiPers (Multi Class) | 71.11* | - | 69.33 |
| SentiPers (Binary Class) | 92.13* | - | 91.98 |
### Text Classification (TC) task
| Dataset | ParsBERT | mBERT |
|:-----------------:|:--------:|:-----:|
| Digikala Magazine | 93.59* | 90.72 |
| Persian News | 97.19* | 95.79 |
### Named Entity Recognition (NER) task
| Dataset | ParsBERT | mBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|:-------:|:--------:|:--------:|:----------:|:--------------:|:----------:|:----------------:|:------------:|
| PEYMA | 93.10* | 86.64 | - | 90.59 | - | 84.00 | - |
| ARMAN | 98.79* | 95.89 | 89.9 | 84.03 | 86.55 | - | 77.45 |
**If you tested ParsBERT on a public dataset and you want to add your results to the table above, open a pull request or contact us. Also make sure to have your code available online so we can add it as a reference**
## How to use
### TensorFlow 2.0
```python
from transformers import AutoConfig, AutoTokenizer, TFAutoModel
config = AutoConfig.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
tokenizer = AutoTokenizer.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
model = AutoModel.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
text = "ما در هوشواره معتقدیم با انتقال صحیح دانش و آگاهی، همه افراد میتوانند از ابزارهای هوشمند استفاده کنند. شعار ما هوش مصنوعی برای همه است."
tokenizer.tokenize(text)
>>> ['ما', 'در', 'هوش', '##واره', 'معتقدیم', 'با', 'انتقال', 'صحیح', 'دانش', 'و', 'اگاهی', '،', 'همه', 'افراد', 'میتوانند', 'از', 'ابزارهای', 'هوشمند', 'استفاده', 'کنند', '.', 'شعار', 'ما', 'هوش', 'مصنوعی', 'برای', 'همه', 'است', '.']
```
### Pytorch
```python
from transformers import AutoConfig, AutoTokenizer, AutoModel
config = AutoConfig.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
tokenizer = AutoTokenizer.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
model = AutoModel.from_pretrained("HooshvareLab/bert-base-parsbert-uncased")
```
## NLP Tasks Tutorial
Coming soon stay tuned
## Cite
Please cite the following paper in your publication if you are using [ParsBERT](https://arxiv.org/abs/2005.12515) in your research:
```markdown
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Acknowledgments
We hereby, express our gratitude to the [Tensorflow Research Cloud (TFRC) program](https://tensorflow.org/tfrc) for providing us with the necessary computation resources. We also thank [Hooshvare](https://hooshvare.com) Research Group for facilitating dataset gathering and scraping online text resources.
## Contributors
- Mehrdad Farahani: [Linkedin](https://www.linkedin.com/in/m3hrdadfi/), [Twitter](https://twitter.com/m3hrdadfi), [Github](https://github.com/m3hrdadfi)
- Mohammad Gharachorloo: [Linkedin](https://www.linkedin.com/in/mohammad-gharachorloo/), [Twitter](https://twitter.com/MGharachorloo), [Github](https://github.com/baarsaam)
- Marzieh Farahani: [Linkedin](https://www.linkedin.com/in/marziehphi/), [Twitter](https://twitter.com/marziehphi), [Github](https://github.com/marziehphi)
- Mohammad Manthouri: [Linkedin](https://www.linkedin.com/in/mohammad-manthouri-aka-mansouri-07030766/), [Twitter](https://twitter.com/mmanthouri), [Github](https://github.com/mmanthouri)
- Hooshvare Team: [Official Website](https://hooshvare.com/), [Linkedin](https://www.linkedin.com/company/hooshvare), [Twitter](https://twitter.com/hooshvare), [Github](https://github.com/hooshvare), [Instagram](https://www.instagram.com/hooshvare/)
## Releases
### Release v0.1 (May 27, 2019)
This is the first version of our ParsBERT based on BERT<sub>BASE</sub>
|
HooshvareLab/bert-fa-base-uncased-clf-digimag
|
HooshvareLab
|
bert
| 12 | 120 |
transformers
| 0 |
text-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,687 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian Text Classification [DigiMag, Persian News]
The task target is labeling texts in a supervised manner in both existing datasets `DigiMag` and `Persian News`.
### DigiMag
A total of 8,515 articles scraped from [Digikala Online Magazine](https://www.digikala.com/mag/). This dataset includes seven different classes.
1. Video Games
2. Shopping Guide
3. Health Beauty
4. Science Technology
5. General
6. Art Cinema
7. Books Literature
| Label | # |
|:------------------:|:----:|
| Video Games | 1967 |
| Shopping Guide | 125 |
| Health Beauty | 1610 |
| Science Technology | 2772 |
| General | 120 |
| Art Cinema | 1667 |
| Books Literature | 254 |
**Download**
You can download the dataset from [here](https://drive.google.com/uc?id=1YgrCYY-Z0h2z0-PfWVfOGt1Tv0JDI-qz)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT |
|:-----------------:|:-----------:|:-----------:|:-----:|
| Digikala Magazine | 93.65* | 93.59 | 90.72 |
## How to use :hugs:
| Task | Notebook |
|---------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Text Classification | [](https://colab.research.google.com/github/hooshvare/parsbert/blob/master/notebooks/Taaghche_Sentiment_Analysis.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-clf-persiannews
|
HooshvareLab
|
bert
| 12 | 98 |
transformers
| 2 |
text-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,726 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian Text Classification [DigiMag, Persian News]
The task target is labeling texts in a supervised manner in both existing datasets `DigiMag` and `Persian News`.
### Persian News
A dataset of various news articles scraped from different online news agencies' websites. The total number of articles is 16,438, spread over eight different classes.
1. Economic
2. International
3. Political
4. Science Technology
5. Cultural Art
6. Sport
7. Medical
| Label | # |
|:------------------:|:----:|
| Social | 2170 |
| Economic | 1564 |
| International | 1975 |
| Political | 2269 |
| Science Technology | 2436 |
| Cultural Art | 2558 |
| Sport | 1381 |
| Medical | 2085 |
**Download**
You can download the dataset from [here](https://drive.google.com/uc?id=1B6xotfXCcW9xS1mYSBQos7OCg0ratzKC)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT |
|:-----------------:|:-----------:|:-----------:|:-----:|
| Persian News | 97.44* | 97.19 | 95.79 |
## How to use :hugs:
| Task | Notebook |
|---------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Text Classification | [](https://colab.research.google.com/github/hooshvare/parsbert/blob/master/notebooks/Taaghche_Sentiment_Analysis.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-ner-arman
|
HooshvareLab
|
bert
| 12 | 15 |
transformers
| 0 |
token-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,073 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian NER [ARMAN, PEYMA]
This task aims to extract named entities in the text, such as names and label with appropriate `NER` classes such as locations, organizations, etc. The datasets used for this task contain sentences that are marked with `IOB` format. In this format, tokens that are not part of an entity are tagged as `”O”` the `”B”`tag corresponds to the first word of an object, and the `”I”` tag corresponds to the rest of the terms of the same entity. Both `”B”` and `”I”` tags are followed by a hyphen (or underscore), followed by the entity category. Therefore, the NER task is a multi-class token classification problem that labels the tokens upon being fed a raw text. There are two primary datasets used in Persian NER, `ARMAN`, and `PEYMA`.
### ARMAN
ARMAN dataset holds 7,682 sentences with 250,015 sentences tagged over six different classes.
1. Organization
2. Location
3. Facility
4. Event
5. Product
6. Person
| Label | # |
|:------------:|:-----:|
| Organization | 30108 |
| Location | 12924 |
| Facility | 4458 |
| Event | 7557 |
| Product | 4389 |
| Person | 15645 |
**Download**
You can download the dataset from [here](https://github.com/HaniehP/PersianNER)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|---------|-------------|-------------|-------|------------|--------------|----------|----------------|------------|
| ARMAN | 99.84* | 98.79 | 95.89 | 89.9 | 84.03 | 86.55 | - | 77.45 |
## How to use :hugs:
| Notebook | Description | |
|:----------|:-------------|------:|
| [How to use Pipelines](https://github.com/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) | Simple and efficient way to use State-of-the-Art models on downstream tasks through transformers | [](https://colab.research.google.com/github/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-ner-peyma
|
HooshvareLab
|
bert
| 12 | 182 |
transformers
| 1 |
token-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,176 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian NER [ARMAN, PEYMA]
This task aims to extract named entities in the text, such as names and label with appropriate `NER` classes such as locations, organizations, etc. The datasets used for this task contain sentences that are marked with `IOB` format. In this format, tokens that are not part of an entity are tagged as `”O”` the `”B”`tag corresponds to the first word of an object, and the `”I”` tag corresponds to the rest of the terms of the same entity. Both `”B”` and `”I”` tags are followed by a hyphen (or underscore), followed by the entity category. Therefore, the NER task is a multi-class token classification problem that labels the tokens upon being fed a raw text. There are two primary datasets used in Persian NER, `ARMAN`, and `PEYMA`.
### PEYMA
PEYMA dataset includes 7,145 sentences with a total of 302,530 tokens from which 41,148 tokens are tagged with seven different classes.
1. Organization
2. Money
3. Location
4. Date
5. Time
6. Person
7. Percent
| Label | # |
|:------------:|:-----:|
| Organization | 16964 |
| Money | 2037 |
| Location | 8782 |
| Date | 4259 |
| Time | 732 |
| Person | 7675 |
| Percent | 699 |
**Download**
You can download the dataset from [here](http://nsurl.org/tasks/task-7-named-entity-recognition-ner-for-farsi/)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|---------|-------------|-------------|-------|------------|--------------|----------|----------------|------------|
| PEYMA | 93.40* | 93.10 | 86.64 | - | 90.59 | - | 84.00 | - |
## How to use :hugs:
| Notebook | Description | |
|:----------|:-------------|------:|
| [How to use Pipelines](https://github.com/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) | Simple and efficient way to use State-of-the-Art models on downstream tasks through transformers | [](https://colab.research.google.com/github/hooshvare/parsbert-ner/blob/master/persian-ner-pipeline.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-sentiment-deepsentipers-binary
|
HooshvareLab
|
bert
| 12 | 38 |
transformers
| 2 |
text-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,226 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian Sentiment [Digikala, SnappFood, DeepSentiPers]
It aims to classify text, such as comments, based on their emotional bias. We tested three well-known datasets for this task: `Digikala` user comments, `SnappFood` user comments, and `DeepSentiPers` in two binary-form and multi-form types.
### DeepSentiPers
which is a balanced and augmented version of SentiPers, contains 12,138 user opinions about digital products labeled with five different classes; two positives (i.e., happy and delighted), two negatives (i.e., furious and angry) and one neutral class. Therefore, this dataset can be utilized for both multi-class and binary classification. In the case of binary classification, the neutral class and its corresponding sentences are removed from the dataset.
**Binary:**
1. Negative (Furious + Angry)
2. Positive (Happy + Delighted)
**Multi**
1. Furious
2. Angry
3. Neutral
4. Happy
5. Delighted
| Label | # |
|:---------:|:----:|
| Furious | 236 |
| Angry | 1357 |
| Neutral | 2874 |
| Happy | 2848 |
| Delighted | 2516 |
**Download**
You can download the dataset from:
- [SentiPers](https://github.com/phosseini/sentipers)
- [DeepSentiPers](https://github.com/JoyeBright/DeepSentiPers)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | DeepSentiPers |
|:------------------------:|:-----------:|:-----------:|:-----:|:-------------:|
| SentiPers (Multi Class) | 71.31* | 71.11 | - | 69.33 |
| SentiPers (Binary Class) | 92.42* | 92.13 | - | 91.98 |
## How to use :hugs:
| Task | Notebook |
|---------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Sentiment Analysis | [](https://colab.research.google.com/github/hooshvare/parsbert/blob/master/notebooks/Taaghche_Sentiment_Analysis.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-sentiment-deepsentipers-multi
|
HooshvareLab
|
bert
| 12 | 307 |
transformers
| 0 |
text-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,227 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian Sentiment [Digikala, SnappFood, DeepSentiPers]
It aims to classify text, such as comments, based on their emotional bias. We tested three well-known datasets for this task: `Digikala` user comments, `SnappFood` user comments, and `DeepSentiPers` in two binary-form and multi-form types.
### DeepSentiPers
which is a balanced and augmented version of SentiPers, contains 12,138 user opinions about digital products labeled with five different classes; two positives (i.e., happy and delighted), two negatives (i.e., furious and angry) and one neutral class. Therefore, this dataset can be utilized for both multi-class and binary classification. In the case of binary classification, the neutral class and its corresponding sentences are removed from the dataset.
**Binary:**
1. Negative (Furious + Angry)
2. Positive (Happy + Delighted)
**Multi**
1. Furious
2. Angry
3. Neutral
4. Happy
5. Delighted
| Label | # |
|:---------:|:----:|
| Furious | 236 |
| Angry | 1357 |
| Neutral | 2874 |
| Happy | 2848 |
| Delighted | 2516 |
**Download**
You can download the dataset from:
- [SentiPers](https://github.com/phosseini/sentipers)
- [DeepSentiPers](https://github.com/JoyeBright/DeepSentiPers)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | DeepSentiPers |
|:------------------------:|:-----------:|:-----------:|:-----:|:-------------:|
| SentiPers (Multi Class) | 71.31* | 71.11 | - | 69.33 |
| SentiPers (Binary Class) | 92.42* | 92.13 | - | 91.98 |
## How to use :hugs:
| Task | Notebook |
|---------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Sentiment Analysis | [](https://colab.research.google.com/github/hooshvare/parsbert/blob/master/notebooks/Taaghche_Sentiment_Analysis.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-sentiment-digikala
|
HooshvareLab
|
bert
| 12 | 181 |
transformers
| 0 |
text-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,633 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian Sentiment [Digikala, SnappFood, DeepSentiPers]
It aims to classify text, such as comments, based on their emotional bias. We tested three well-known datasets for this task: `Digikala` user comments, `SnappFood` user comments, and `DeepSentiPers` in two binary-form and multi-form types.
### Digikala
Digikala user comments provided by [Open Data Mining Program (ODMP)](https://www.digikala.com/opendata/). This dataset contains 62,321 user comments with three labels:
| Label | # |
|:---------------:|:------:|
| no_idea | 10394 |
| not_recommended | 15885 |
| recommended | 36042 |
**Download**
You can download the dataset from [here](https://www.digikala.com/opendata/)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | DeepSentiPers |
|:------------------------:|:-----------:|:-----------:|:-----:|:-------------:|
| Digikala User Comments | 81.72 | 81.74* | 80.74 | - |
## How to use :hugs:
| Task | Notebook |
|---------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Sentiment Analysis | [](https://colab.research.google.com/github/hooshvare/parsbert/blob/master/notebooks/Taaghche_Sentiment_Analysis.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased-sentiment-snappfood
|
HooshvareLab
|
bert
| 12 | 112 |
transformers
| 0 |
text-classification
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,609 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Persian Sentiment [Digikala, SnappFood, DeepSentiPers]
It aims to classify text, such as comments, based on their emotional bias. We tested three well-known datasets for this task: `Digikala` user comments, `SnappFood` user comments, and `DeepSentiPers` in two binary-form and multi-form types.
### SnappFood
[Snappfood](https://snappfood.ir/) (an online food delivery company) user comments containing 70,000 comments with two labels (i.e. polarity classification):
1. Happy
2. Sad
| Label | # |
|:--------:|:-----:|
| Negative | 35000 |
| Positive | 35000 |
**Download**
You can download the dataset from [here](https://drive.google.com/uc?id=15J4zPN1BD7Q_ZIQ39VeFquwSoW8qTxgu)
## Results
The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures.
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | DeepSentiPers |
|:------------------------:|:-----------:|:-----------:|:-----:|:-------------:|
| SnappFood User Comments | 87.98 | 88.12* | 87.87 | - |
## How to use :hugs:
| Task | Notebook |
|---------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Sentiment Analysis | [](https://colab.research.google.com/github/hooshvare/parsbert/blob/master/notebooks/Taaghche_Sentiment_Analysis.ipynb) |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-base-uncased
|
HooshvareLab
|
bert
| 7 | 7,806 |
transformers
| 2 |
fill-mask
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['bert-fa', 'bert-persian', 'persian-lm']
| false | true | true | 6,869 |
# ParsBERT (v2.0)
A Transformer-based Model for Persian Language Understanding
We reconstructed the vocabulary and fine-tuned the ParsBERT v1.1 on the new Persian corpora in order to provide some functionalities for using ParsBERT in other scopes!
Please follow the [ParsBERT](https://github.com/hooshvare/parsbert) repo for the latest information about previous and current models.
## Introduction
ParsBERT is a monolingual language model based on Google’s BERT architecture. This model is pre-trained on large Persian corpora with various writing styles from numerous subjects (e.g., scientific, novels, news) with more than `3.9M` documents, `73M` sentences, and `1.3B` words.
Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515)
## Intended uses & limitations
You can use the raw model for either masked language modeling or next sentence prediction, but it's mostly intended to
be fine-tuned on a downstream task. See the [model hub](https://huggingface.co/models?search=bert-fa) to look for
fine-tuned versions on a task that interests you.
### How to use
#### TensorFlow 2.0
```python
from transformers import AutoConfig, AutoTokenizer, TFAutoModel
config = AutoConfig.from_pretrained("HooshvareLab/bert-fa-base-uncased")
tokenizer = AutoTokenizer.from_pretrained("HooshvareLab/bert-fa-base-uncased")
model = TFAutoModel.from_pretrained("HooshvareLab/bert-fa-base-uncased")
text = "ما در هوشواره معتقدیم با انتقال صحیح دانش و آگاهی، همه افراد میتوانند از ابزارهای هوشمند استفاده کنند. شعار ما هوش مصنوعی برای همه است."
tokenizer.tokenize(text)
>>> ['ما', 'در', 'هوش', '##واره', 'معتقدیم', 'با', 'انتقال', 'صحیح', 'دانش', 'و', 'اگاهی', '،', 'همه', 'افراد', 'میتوانند', 'از', 'ابزارهای', 'هوشمند', 'استفاده', 'کنند', '.', 'شعار', 'ما', 'هوش', 'مصنوعی', 'برای', 'همه', 'است', '.']
```
#### Pytorch
```python
from transformers import AutoConfig, AutoTokenizer, AutoModel
config = AutoConfig.from_pretrained("HooshvareLab/bert-fa-base-uncased")
tokenizer = AutoTokenizer.from_pretrained("HooshvareLab/bert-fa-base-uncased")
model = AutoModel.from_pretrained("HooshvareLab/bert-fa-base-uncased")
```
## Training
ParsBERT trained on a massive amount of public corpora ([Persian Wikidumps](https://dumps.wikimedia.org/fawiki/), [MirasText](https://github.com/miras-tech/MirasText)) and six other manually crawled text data from a various type of websites ([BigBang Page](https://bigbangpage.com/) `scientific`, [Chetor](https://www.chetor.com/) `lifestyle`, [Eligasht](https://www.eligasht.com/Blog/) `itinerary`, [Digikala](https://www.digikala.com/mag/) `digital magazine`, [Ted Talks](https://www.ted.com/talks) `general conversational`, Books `novels, storybooks, short stories from old to the contemporary era`).
As a part of ParsBERT methodology, an extensive pre-processing combining POS tagging and WordPiece segmentation was carried out to bring the corpora into a proper format.
## Goals
Objective goals during training are as below (after 300k steps).
``` bash
***** Eval results *****
global_step = 300000
loss = 1.4392426
masked_lm_accuracy = 0.6865794
masked_lm_loss = 1.4469004
next_sentence_accuracy = 1.0
next_sentence_loss = 6.534152e-05
```
## Derivative models
### Base Config
#### ParsBERT v2.0 Model
- [HooshvareLab/bert-fa-base-uncased](https://huggingface.co/HooshvareLab/bert-fa-base-uncased)
#### ParsBERT v2.0 Sentiment Analysis
- [HooshvareLab/bert-fa-base-uncased-sentiment-digikala](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-sentiment-digikala)
- [HooshvareLab/bert-fa-base-uncased-sentiment-snappfood](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-sentiment-snappfood)
- [HooshvareLab/bert-fa-base-uncased-sentiment-deepsentipers-binary](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-sentiment-deepsentipers-binary)
- [HooshvareLab/bert-fa-base-uncased-sentiment-deepsentipers-multi](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-sentiment-deepsentipers-multi)
#### ParsBERT v2.0 Text Classification
- [HooshvareLab/bert-fa-base-uncased-clf-digimag](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-clf-digimag)
- [HooshvareLab/bert-fa-base-uncased-clf-persiannews](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-clf-persiannews)
#### ParsBERT v2.0 NER
- [HooshvareLab/bert-fa-base-uncased-ner-peyma](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-ner-peyma)
- [HooshvareLab/bert-fa-base-uncased-ner-arman](https://huggingface.co/HooshvareLab/bert-fa-base-uncased-ner-arman)
## Eval results
ParsBERT is evaluated on three NLP downstream tasks: Sentiment Analysis (SA), Text Classification, and Named Entity Recognition (NER). For this matter and due to insufficient resources, two large datasets for SA and two for text classification were manually composed, which are available for public use and benchmarking. ParsBERT outperformed all other language models, including multilingual BERT and other hybrid deep learning models for all tasks, improving the state-of-the-art performance in Persian language modeling.
### Sentiment Analysis (SA) Task
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | DeepSentiPers |
|:------------------------:|:-----------:|:-----------:|:-----:|:-------------:|
| Digikala User Comments | 81.72 | 81.74* | 80.74 | - |
| SnappFood User Comments | 87.98 | 88.12* | 87.87 | - |
| SentiPers (Multi Class) | 71.31* | 71.11 | - | 69.33 |
| SentiPers (Binary Class) | 92.42* | 92.13 | - | 91.98 |
### Text Classification (TC) Task
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT |
|:-----------------:|:-----------:|:-----------:|:-----:|
| Digikala Magazine | 93.65* | 93.59 | 90.72 |
| Persian News | 97.44* | 97.19 | 95.79 |
### Named Entity Recognition (NER) Task
| Dataset | ParsBERT v2 | ParsBERT v1 | mBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF |
|:-------:|:-----------:|:-----------:|:-----:|:----------:|:------------:|:--------:|:--------------:|:----------:|
| PEYMA | 93.40* | 93.10 | 86.64 | - | 90.59 | - | 84.00 | - |
| ARMAN | 99.84* | 98.79 | 95.89 | 89.9 | 84.03 | 86.55 | - | 77.45 |
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/bert-fa-zwnj-base-ner
|
HooshvareLab
|
bert
| 10 | 65 |
transformers
| 3 |
token-classification
| true | true | true | null |
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,629 |
# BertNER
This model fine-tuned for the Named Entity Recognition (NER) task on a mixed NER dataset collected from [ARMAN](https://github.com/HaniehP/PersianNER), [PEYMA](http://nsurl.org/2019-2/tasks/task-7-named-entity-recognition-ner-for-farsi/), and [WikiANN](https://elisa-ie.github.io/wikiann/) that covered ten types of entities:
- Date (DAT)
- Event (EVE)
- Facility (FAC)
- Location (LOC)
- Money (MON)
- Organization (ORG)
- Percent (PCT)
- Person (PER)
- Product (PRO)
- Time (TIM)
## Dataset Information
| | Records | B-DAT | B-EVE | B-FAC | B-LOC | B-MON | B-ORG | B-PCT | B-PER | B-PRO | B-TIM | I-DAT | I-EVE | I-FAC | I-LOC | I-MON | I-ORG | I-PCT | I-PER | I-PRO | I-TIM |
|:------|----------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|
| Train | 29133 | 1423 | 1487 | 1400 | 13919 | 417 | 15926 | 355 | 12347 | 1855 | 150 | 1947 | 5018 | 2421 | 4118 | 1059 | 19579 | 573 | 7699 | 1914 | 332 |
| Valid | 5142 | 267 | 253 | 250 | 2362 | 100 | 2651 | 64 | 2173 | 317 | 19 | 373 | 799 | 387 | 717 | 270 | 3260 | 101 | 1382 | 303 | 35 |
| Test | 6049 | 407 | 256 | 248 | 2886 | 98 | 3216 | 94 | 2646 | 318 | 43 | 568 | 888 | 408 | 858 | 263 | 3967 | 141 | 1707 | 296 | 78 |
## Evaluation
The following tables summarize the scores obtained by model overall and per each class.
**Overall**
| Model | accuracy | precision | recall | f1 |
|:----------:|:--------:|:---------:|:--------:|:--------:|
| Bert | 0.995086 | 0.953454 | 0.961113 | 0.957268 |
**Per entities**
| | number | precision | recall | f1 |
|:---: |:------: |:---------: |:--------: |:--------: |
| DAT | 407 | 0.860636 | 0.864865 | 0.862745 |
| EVE | 256 | 0.969582 | 0.996094 | 0.982659 |
| FAC | 248 | 0.976190 | 0.991935 | 0.984000 |
| LOC | 2884 | 0.970232 | 0.971914 | 0.971072 |
| MON | 98 | 0.905263 | 0.877551 | 0.891192 |
| ORG | 3216 | 0.939125 | 0.954602 | 0.946800 |
| PCT | 94 | 1.000000 | 0.968085 | 0.983784 |
| PER | 2645 | 0.965244 | 0.965974 | 0.965608 |
| PRO | 318 | 0.981481 | 1.000000 | 0.990654 |
| TIM | 43 | 0.692308 | 0.837209 | 0.757895 |
## How To Use
You use this model with Transformers pipeline for NER.
### Installing requirements
```bash
pip install transformers
```
### How to predict using pipeline
```python
from transformers import AutoTokenizer
from transformers import AutoModelForTokenClassification # for pytorch
from transformers import TFAutoModelForTokenClassification # for tensorflow
from transformers import pipeline
model_name_or_path = "HooshvareLab/bert-fa-zwnj-base-ner"
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
model = AutoModelForTokenClassification.from_pretrained(model_name_or_path) # Pytorch
# model = TFAutoModelForTokenClassification.from_pretrained(model_name_or_path) # Tensorflow
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
example = "در سال ۲۰۱۳ درگذشت و آندرتیکر و کین برای او مراسم یادبود گرفتند."
ner_results = nlp(example)
print(ner_results)
```
## Questions?
Post a Github issue on the [ParsNER Issues](https://github.com/hooshvare/parsner/issues) repo.
|
HooshvareLab/bert-fa-zwnj-base
|
HooshvareLab
|
bert
| 10 | 9,765 |
transformers
| 3 |
fill-mask
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 1,083 |
# ParsBERT (v3.0)
A Transformer-based Model for Persian Language Understanding
The new version of BERT v3.0 for Persian is available today and can tackle the zero-width non-joiner character for Persian writing. Also, the model was trained on new multi-types corpora with a new set of vocabulary.
## Introduction
ParsBERT is a monolingual language model based on Google’s BERT architecture. This model is pre-trained on large Persian corpora with various writing styles from numerous subjects (e.g., scientific, novels, news).
Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515)
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@article{ParsBERT,
title={ParsBERT: Transformer-based Model for Persian Language Understanding},
author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri},
journal={ArXiv},
year={2020},
volume={abs/2005.12515}
}
```
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/distilbert-fa-zwnj-base-ner
|
HooshvareLab
|
distilbert
| 9 | 14,712 |
transformers
| 1 |
token-classification
| true | true | false | null |
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,641 |
# DistilbertNER
This model fine-tuned for the Named Entity Recognition (NER) task on a mixed NER dataset collected from [ARMAN](https://github.com/HaniehP/PersianNER), [PEYMA](http://nsurl.org/2019-2/tasks/task-7-named-entity-recognition-ner-for-farsi/), and [WikiANN](https://elisa-ie.github.io/wikiann/) that covered ten types of entities:
- Date (DAT)
- Event (EVE)
- Facility (FAC)
- Location (LOC)
- Money (MON)
- Organization (ORG)
- Percent (PCT)
- Person (PER)
- Product (PRO)
- Time (TIM)
## Dataset Information
| | Records | B-DAT | B-EVE | B-FAC | B-LOC | B-MON | B-ORG | B-PCT | B-PER | B-PRO | B-TIM | I-DAT | I-EVE | I-FAC | I-LOC | I-MON | I-ORG | I-PCT | I-PER | I-PRO | I-TIM |
|:------|----------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|
| Train | 29133 | 1423 | 1487 | 1400 | 13919 | 417 | 15926 | 355 | 12347 | 1855 | 150 | 1947 | 5018 | 2421 | 4118 | 1059 | 19579 | 573 | 7699 | 1914 | 332 |
| Valid | 5142 | 267 | 253 | 250 | 2362 | 100 | 2651 | 64 | 2173 | 317 | 19 | 373 | 799 | 387 | 717 | 270 | 3260 | 101 | 1382 | 303 | 35 |
| Test | 6049 | 407 | 256 | 248 | 2886 | 98 | 3216 | 94 | 2646 | 318 | 43 | 568 | 888 | 408 | 858 | 263 | 3967 | 141 | 1707 | 296 | 78 |
## Evaluation
The following tables summarize the scores obtained by model overall and per each class.
**Overall**
| Model | accuracy | precision | recall | f1 |
|:----------:|:--------:|:---------:|:--------:|:--------:|
| Distilbert | 0.994534 | 0.946326 | 0.95504 | 0.950663 |
**Per entities**
| | number | precision | recall | f1 |
|:---: |:------: |:---------: |:--------: |:--------: |
| DAT | 407 | 0.812048 | 0.828010 | 0.819951 |
| EVE | 256 | 0.955056 | 0.996094 | 0.975143 |
| FAC | 248 | 0.972549 | 1.000000 | 0.986083 |
| LOC | 2884 | 0.968403 | 0.967060 | 0.967731 |
| MON | 98 | 0.925532 | 0.887755 | 0.906250 |
| ORG | 3216 | 0.932095 | 0.951803 | 0.941846 |
| PCT | 94 | 0.936842 | 0.946809 | 0.941799 |
| PER | 2645 | 0.959818 | 0.957278 | 0.958546 |
| PRO | 318 | 0.963526 | 0.996855 | 0.979907 |
| TIM | 43 | 0.760870 | 0.813953 | 0.786517 |
## How To Use
You use this model with Transformers pipeline for NER.
### Installing requirements
```bash
pip install transformers
```
### How to predict using pipeline
```python
from transformers import AutoTokenizer
from transformers import AutoModelForTokenClassification # for pytorch
from transformers import TFAutoModelForTokenClassification # for tensorflow
from transformers import pipeline
model_name_or_path = "HooshvareLab/distilbert-fa-zwnj-base-ner"
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
model = AutoModelForTokenClassification.from_pretrained(model_name_or_path) # Pytorch
# model = TFAutoModelForTokenClassification.from_pretrained(model_name_or_path) # Tensorflow
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
example = "در سال ۲۰۱۳ درگذشت و آندرتیکر و کین برای او مراسم یادبود گرفتند."
ner_results = nlp(example)
print(ner_results)
```
## Questions?
Post a Github issue on the [ParsNER Issues](https://github.com/hooshvare/parsner/issues) repo.
|
HooshvareLab/distilbert-fa-zwnj-base
|
HooshvareLab
|
distilbert
| 9 | 593 |
transformers
| 0 |
fill-mask
| true | true | false |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 291 |
# DistilBERT
This model can tackle the zero-width non-joiner character for Persian writing. Also, the model was trained on new multi-types corpora with a new set of vocabulary.
## Questions?
Post a Github issue on the [ParsBERT Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
HooshvareLab/gpt2-fa-comment
|
HooshvareLab
|
gpt2
| 12 | 6 |
transformers
| 0 |
text-generation
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 16,752 |
# Persian Comment Generator
The model can generate comments based on your aspects, and the model was fine-tuned on [persiannlp/parsinlu](https://github.com/persiannlp/parsinlu). Currently, the model only supports aspects in the food and movie scope. You can see the whole aspects in the following section.
## Comments Aspects
```text
<s>نمونه دیدگاه هم خوب هم بد به طور کلی <sep>
<s>نمونه دیدگاه خوب به طور کلی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر ارزش غذایی و ارزش خرید <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم و بسته بندی <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و کیفیت <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت <sep>
<s>نمونه دیدگاه منفی از نظر کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر طعم <sep>
<s>نمونه دیدگاه خیلی خوب به طور کلی <sep>
<s>نمونه دیدگاه خوب از نظر بسته بندی <sep>
<s>نمونه دیدگاه منفی از نظر کیفیت و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارسال و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و طعم <sep>
<s>نمونه دیدگاه منفی به طور کلی <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر ارسال <sep>
<s>نمونه دیدگاه منفی از نظر طعم <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و ارزش خرید <sep>
<s>نمونه دیدگاه نظری ندارم به طور کلی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم <sep>
<s>نمونه دیدگاه خیلی منفی به طور کلی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش خرید و کیفیت و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت <sep>
<s>نمونه دیدگاه منفی از نظر طعم و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر طعم و کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارسال <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش خرید و طعم <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بسته بندی و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و بسته بندی و کیفیت <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و بسته بندی و ارسال <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت <sep>
<s>نمونه دیدگاه منفی از نظر بسته بندی <sep>
<s>نمونه دیدگاه خوب از نظر طعم و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر طعم و کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر ارسال و کیفیت <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و بسته بندی و ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه منفی از نظر ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و طعم <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و طعم <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم و ارسال <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر طعم و ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر طعم و ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بسته بندی و ارسال و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و بسته بندی و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بسته بندی و کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر بسته بندی و کیفیت و طعم <sep>
<s>نمونه دیدگاه خوب از نظر ارسال <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و بسته بندی و ارزش غذایی و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش غذایی و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر طعم و ارسال <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خوب از نظر بسته بندی و ارزش خرید <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش غذایی و طعم <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت و ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش غذایی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش غذایی و ارزش خرید <sep>
<s>نمونه دیدگاه منفی از نظر طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و ارسال <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و طعم <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش غذایی و بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر طعم و کیفیت و ارسال <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و بسته بندی و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر بسته بندی و طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر ارسال و طعم <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و ارسال <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش غذایی و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و طعم و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بسته بندی و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و کیفیت و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه منفی از نظر بسته بندی و کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و کیفیت و ارزش خرید و بسته بندی <sep>
<s>نمونه دیدگاه خوب از نظر ارزش غذایی و ارسال <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و طعم و ارزش خرید و ارسال <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارسال و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش خرید و بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارسال و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و ارزش خرید و ارسال <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و ارزش خرید و طعم <sep>
<s>نمونه دیدگاه خوب از نظر بسته بندی و کیفیت <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بسته بندی و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و بسته بندی و ارسال <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بسته بندی و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه نظری ندارم از نظر بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت و بسته بندی و طعم <sep>
<s>نمونه دیدگاه خوب از نظر طعم و بسته بندی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و ارزش خرید و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و بسته بندی <sep>
<s>نمونه دیدگاه خوب از نظر ارزش خرید و ارزش غذایی <sep>
<s>نمونه دیدگاه منفی از نظر طعم و بسته بندی <sep>
<s>نمونه دیدگاه منفی از نظر کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و ارزش غذایی و بسته بندی <sep>
<s>نمونه دیدگاه خوب از نظر ارسال و بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارسال <sep>
<s>نمونه دیدگاه نظری ندارم از نظر طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه منفی از نظر ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارسال و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و کیفیت و بسته بندی و ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر طعم و بسته بندی و ارزش خرید <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت و ارسال <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و کیفیت و ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و ارسال و ارزش خرید <sep>
<s>نمونه دیدگاه نظری ندارم از نظر ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارسال و ارزش خرید و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بسته بندی و طعم و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کیفیت و ارسال و بسته بندی <sep>
<s>نمونه دیدگاه منفی از نظر بسته بندی و طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بسته بندی و ارسال <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارسال و کیفیت <sep>
<s>نمونه دیدگاه خوب از نظر کیفیت و ارسال <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش خرید و ارزش غذایی <sep>
<s>نمونه دیدگاه خوب از نظر ارزش غذایی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و ارزش غذایی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارسال و بسته بندی و کیفیت <sep>
<s>نمونه دیدگاه منفی از نظر بسته بندی و طعم <sep>
<s>نمونه دیدگاه منفی از نظر بسته بندی و ارزش غذایی <sep>
<s>نمونه دیدگاه منفی از نظر طعم و کیفیت و ارزش خرید <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و ارزش غذایی و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش غذایی و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش خرید و طعم و بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کیفیت و بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارزش خرید و کیفیت و طعم <sep>
<s>نمونه دیدگاه منفی از نظر ارزش خرید و کیفیت و طعم <sep>
<s>نمونه دیدگاه منفی از نظر کیفیت و طعم و ارزش غذایی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارسال و کیفیت و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش غذایی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر طعم و بسته بندی و ارسال <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و بسته بندی و طعم <sep>
<s>نمونه دیدگاه خیلی خوب از نظر ارزش غذایی و طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش غذایی و کیفیت <sep>
<s>نمونه دیدگاه منفی از نظر ارزش خرید و طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کیفیت و طعم و بسته بندی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر ارسال و ارزش خرید <sep>
<s>نمونه دیدگاه خیلی منفی از نظر ارزش خرید و طعم و کیفیت <sep>
<s>نمونه دیدگاه خیلی منفی از نظر طعم و ارسال <sep>
<s>نمونه دیدگاه منفی از نظر موسیقی و بازی <sep>
<s>نمونه دیدگاه منفی از نظر داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر صدا <sep>
<s>نمونه دیدگاه خیلی منفی از نظر داستان <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و فیلمبرداری و کارگردانی و بازی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بازی <sep>
<s>نمونه دیدگاه منفی از نظر داستان و بازی <sep>
<s>نمونه دیدگاه منفی از نظر بازی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر داستان و کارگردانی و بازی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر داستان و بازی <sep>
<s>نمونه دیدگاه خوب از نظر بازی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بازی و داستان و کارگردانی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی <sep>
<s>نمونه دیدگاه خوب از نظر بازی و داستان <sep>
<s>نمونه دیدگاه خوب از نظر داستان و بازی <sep>
<s>نمونه دیدگاه خوب از نظر داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر داستان و بازی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و داستان <sep>
<s>نمونه دیدگاه خیلی منفی از نظر داستان و کارگردانی و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بازی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کارگردانی <sep>
<s>نمونه دیدگاه منفی از نظر کارگردانی و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی و بازی <sep>
<s>نمونه دیدگاه خوب از نظر کارگردانی و بازی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر صحنه و کارگردانی <sep>
<s>نمونه دیدگاه منفی از نظر بازی و کارگردانی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و داستان و کارگردانی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و کارگردانی و فیلمبرداری و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی و بازی و موسیقی <sep>
<s>نمونه دیدگاه خوب از نظر صحنه و بازی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و موسیقی و کارگردانی <sep>
<s>نمونه دیدگاه خوب از نظر داستان و کارگردانی <sep>
<s>نمونه دیدگاه خوب از نظر بازی و کارگردانی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بازی و کارگردانی <sep>
<s>نمونه دیدگاه منفی از نظر کارگردانی و موسیقی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بازی و داستان <sep>
<s>نمونه دیدگاه خوب از نظر کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر بازی و کارگردانی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی و داستان <sep>
<s>نمونه دیدگاه خیلی منفی از نظر داستان و کارگردانی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر داستان و کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان <sep>
<s>نمونه دیدگاه خوب از نظر بازی و داستان و موسیقی و کارگردانی و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی منفی از نظر داستان و بازی و کارگردانی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بازی و داستان <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و بازی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و بازی و کارگردانی <sep>
<s>نمونه دیدگاه منفی از نظر بازی و داستان <sep>
<s>نمونه دیدگاه خوب از نظر فیلمبرداری و صحنه و موسیقی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و کارگردانی و بازی <sep>
<s>نمونه دیدگاه نظری ندارم از نظر بازی <sep>
<s>نمونه دیدگاه منفی از نظر داستان و کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و بازی و صحنه <sep>
<s>نمونه دیدگاه خوب از نظر کارگردانی و داستان و بازی و فیلمبرداری <sep>
<s>نمونه دیدگاه خوب از نظر بازی و صحنه و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و صحنه و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و موسیقی و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی و صحنه <sep>
<s>نمونه دیدگاه خیلی خوب از نظر فیلمبرداری و صحنه و داستان و کارگردانی <sep>
<s>نمونه دیدگاه منفی از نظر کارگردانی و بازی <sep>
<s>نمونه دیدگاه منفی از نظر کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر داستان و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی و بازی و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر فیلمبرداری و بازی و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر کارگردانی و بازی و داستان و صحنه <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر موسیقی و کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کارگردانی و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر موسیقی و صحنه <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر صحنه و فیلمبرداری و داستان و بازی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و داستان و موسیقی و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کارگردانی و صدا و صحنه و داستان <sep>
<s>نمونه دیدگاه خوب از نظر داستان و کارگردانی و بازی <sep>
<s>نمونه دیدگاه منفی از نظر داستان و بازی و کارگردانی <sep>
<s>نمونه دیدگاه خوب از نظر داستان و بازی و موسیقی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و کارگردانی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کارگردانی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر کارگردانی و بازی و صحنه <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر کارگردانی و بازی <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر صحنه و فیلمبرداری و داستان <sep>
<s>نمونه دیدگاه خوب از نظر موسیقی و داستان <sep>
<s>نمونه دیدگاه منفی از نظر موسیقی و بازی و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر صدا و بازی <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و صحنه و فیلمبرداری <sep>
<s>نمونه دیدگاه خیلی منفی از نظر بازی و فیلمبرداری و داستان و کارگردانی <sep>
<s>نمونه دیدگاه خیلی منفی از نظر صحنه <sep>
<s>نمونه دیدگاه منفی از نظر داستان و صحنه <sep>
<s>نمونه دیدگاه منفی از نظر بازی و صحنه و صدا <sep>
<s>نمونه دیدگاه خیلی منفی از نظر فیلمبرداری و صدا <sep>
<s>نمونه دیدگاه خیلی خوب از نظر موسیقی <sep>
<s>نمونه دیدگاه خوب از نظر بازی و کارگردانی و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و فیلمبرداری و موسیقی و کارگردانی و داستان <sep>
<s>نمونه دیدگاه هم خوب هم بد از نظر فیلمبرداری و داستان و بازی <sep>
<s>نمونه دیدگاه منفی از نظر صحنه و فیلمبرداری و داستان <sep>
<s>نمونه دیدگاه خیلی خوب از نظر بازی و کارگردانی و داستان <sep>
```
## Questions?
Post a Github issue on the [ParsGPT2 Issues](https://github.com/hooshvare/parsgpt/issues) repo.
|
HooshvareLab/gpt2-fa-poetry
|
HooshvareLab
|
gpt2
| 12 | 11 |
transformers
| 0 |
text-generation
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,446 |
# Persian Poet GPT2
## Poets
The model can generate poetry based on your favorite poet, and you need to add one of the following lines as the input the box on the right side or follow the [fine-tuning notebook](https://colab.research.google.com/github/hooshvare/parsgpt/blob/master/notebooks/Persian_Poetry_FineTuning.ipynb).
```text
<s>رودکی<|startoftext|>
<s>فردوسی<|startoftext|>
<s>کسایی<|startoftext|>
<s>ناصرخسرو<|startoftext|>
<s>منوچهری<|startoftext|>
<s>فرخی سیستانی<|startoftext|>
<s>مسعود سعد سلمان<|startoftext|>
<s>ابوسعید ابوالخیر<|startoftext|>
<s>باباطاهر<|startoftext|>
<s>فخرالدین اسعد گرگانی<|startoftext|>
<s>اسدی توسی<|startoftext|>
<s>هجویری<|startoftext|>
<s>خیام<|startoftext|>
<s>نظامی<|startoftext|>
<s>عطار<|startoftext|>
<s>سنایی<|startoftext|>
<s>خاقانی<|startoftext|>
<s>انوری<|startoftext|>
<s>عبدالواسع جبلی<|startoftext|>
<s>نصرالله منشی<|startoftext|>
<s>مهستی گنجوی<|startoftext|>
<s>باباافضل کاشانی<|startoftext|>
<s>مولوی<|startoftext|>
<s>سعدی<|startoftext|>
<s>خواجوی کرمانی<|startoftext|>
<s>عراقی<|startoftext|>
<s>سیف فرغانی<|startoftext|>
<s>حافظ<|startoftext|>
<s>اوحدی<|startoftext|>
<s>شیخ محمود شبستری<|startoftext|>
<s>عبید زاکانی<|startoftext|>
<s>امیرخسرو دهلوی<|startoftext|>
<s>سلمان ساوجی<|startoftext|>
<s>شاه نعمتالله ولی<|startoftext|>
<s>جامی<|startoftext|>
<s>هلالی جغتایی<|startoftext|>
<s>وحشی<|startoftext|>
<s>محتشم کاشانی<|startoftext|>
<s>شیخ بهایی<|startoftext|>
<s>عرفی<|startoftext|>
<s>رضیالدین آرتیمانی<|startoftext|>
<s>صائب تبریزی<|startoftext|>
<s>فیض کاشانی<|startoftext|>
<s>بیدل دهلوی<|startoftext|>
<s>هاتف اصفهانی<|startoftext|>
<s>فروغی بسطامی<|startoftext|>
<s>قاآنی<|startoftext|>
<s>ملا هادی سبزواری<|startoftext|>
<s>پروین اعتصامی<|startoftext|>
<s>ملکالشعرای بهار<|startoftext|>
<s>شهریار<|startoftext|>
<s>رهی معیری<|startoftext|>
<s>اقبال لاهوری<|startoftext|>
<s>خلیلالله خلیلی<|startoftext|>
<s>شاطرعباس صبوحی<|startoftext|>
<s>نیما یوشیج ( آوای آزاد )<|startoftext|>
<s>احمد شاملو<|startoftext|>
<s>سهراب سپهری<|startoftext|>
<s>فروغ فرخزاد<|startoftext|>
<s>سیمین بهبهانی<|startoftext|>
<s>مهدی اخوان ثالث<|startoftext|>
<s>محمدحسن بارق شفیعی<|startoftext|>
<s>شیون فومنی<|startoftext|>
<s>کامبیز صدیقی کسمایی<|startoftext|>
<s>بهرام سالکی<|startoftext|>
<s>عبدالقهّار عاصی<|startoftext|>
<s>اِ لیـــار (جبار محمدی )<|startoftext|>
```
## Questions?
Post a Github issue on the [ParsGPT2 Issues](https://github.com/hooshvare/parsgpt/issues) repo.
|
HooshvareLab/gpt2-fa
|
HooshvareLab
|
gpt2
| 12 | 486 |
transformers
| 2 |
text-generation
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 455 |
# ParsGPT2
### BibTeX entry and citation info
Please cite in publications as the following:
```bibtex
@misc{ParsGPT2,
author = {Hooshvare Team},
title = {ParsGPT2 the Persian version of GPT2},
year = {2021},
publisher = {GitHub},
journal = {GitHub repository},
howpublished = {\url{https://github.com/hooshvare/parsgpt}},
}
```
## Questions?
Post a Github issue on the [ParsGPT2 Issues](https://github.com/hooshvare/parsgpt/issues) repo.
|
HooshvareLab/roberta-fa-zwnj-base-ner
|
HooshvareLab
|
roberta
| 11 | 30 |
transformers
| 0 |
token-classification
| true | true | true | null |
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 3,635 |
# RobertaNER
This model fine-tuned for the Named Entity Recognition (NER) task on a mixed NER dataset collected from [ARMAN](https://github.com/HaniehP/PersianNER), [PEYMA](http://nsurl.org/2019-2/tasks/task-7-named-entity-recognition-ner-for-farsi/), and [WikiANN](https://elisa-ie.github.io/wikiann/) that covered ten types of entities:
- Date (DAT)
- Event (EVE)
- Facility (FAC)
- Location (LOC)
- Money (MON)
- Organization (ORG)
- Percent (PCT)
- Person (PER)
- Product (PRO)
- Time (TIM)
## Dataset Information
| | Records | B-DAT | B-EVE | B-FAC | B-LOC | B-MON | B-ORG | B-PCT | B-PER | B-PRO | B-TIM | I-DAT | I-EVE | I-FAC | I-LOC | I-MON | I-ORG | I-PCT | I-PER | I-PRO | I-TIM |
|:------|----------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|--------:|
| Train | 29133 | 1423 | 1487 | 1400 | 13919 | 417 | 15926 | 355 | 12347 | 1855 | 150 | 1947 | 5018 | 2421 | 4118 | 1059 | 19579 | 573 | 7699 | 1914 | 332 |
| Valid | 5142 | 267 | 253 | 250 | 2362 | 100 | 2651 | 64 | 2173 | 317 | 19 | 373 | 799 | 387 | 717 | 270 | 3260 | 101 | 1382 | 303 | 35 |
| Test | 6049 | 407 | 256 | 248 | 2886 | 98 | 3216 | 94 | 2646 | 318 | 43 | 568 | 888 | 408 | 858 | 263 | 3967 | 141 | 1707 | 296 | 78 |
## Evaluation
The following tables summarize the scores obtained by model overall and per each class.
**Overall**
| Model | accuracy | precision | recall | f1 |
|:----------:|:--------:|:---------:|:--------:|:--------:|
| Roberta | 0.994849 | 0.949816 | 0.960235 | 0.954997 |
**Per entities**
| | number | precision | recall | f1 |
|:---: |:------: |:---------: |:--------: |:--------: |
| DAT | 407 | 0.844869 | 0.869779 | 0.857143 |
| EVE | 256 | 0.948148 | 1.000000 | 0.973384 |
| FAC | 248 | 0.957529 | 1.000000 | 0.978304 |
| LOC | 2884 | 0.965422 | 0.968100 | 0.966759 |
| MON | 98 | 0.937500 | 0.918367 | 0.927835 |
| ORG | 3216 | 0.943662 | 0.958333 | 0.950941 |
| PCT | 94 | 1.000000 | 0.968085 | 0.983784 |
| PER | 2646 | 0.957030 | 0.959562 | 0.958294 |
| PRO | 318 | 0.963636 | 1.000000 | 0.981481 |
| TIM | 43 | 0.739130 | 0.790698 | 0.764045 |
## How To Use
You use this model with Transformers pipeline for NER.
### Installing requirements
```bash
pip install transformers
```
### How to predict using pipeline
```python
from transformers import AutoTokenizer
from transformers import AutoModelForTokenClassification # for pytorch
from transformers import TFAutoModelForTokenClassification # for tensorflow
from transformers import pipeline
model_name_or_path = "HooshvareLab/roberta-fa-zwnj-base-ner"
tokenizer = AutoTokenizer.from_pretrained(model_name_or_path)
model = AutoModelForTokenClassification.from_pretrained(model_name_or_path) # Pytorch
# model = TFAutoModelForTokenClassification.from_pretrained(model_name_or_path) # Tensorflow
nlp = pipeline("ner", model=model, tokenizer=tokenizer)
example = "در سال ۲۰۱۳ درگذشت و آندرتیکر و کین برای او مراسم یادبود گرفتند."
ner_results = nlp(example)
print(ner_results)
```
## Questions?
Post a Github issue on the [ParsNER Issues](https://github.com/hooshvare/parsner/issues) repo.
|
HooshvareLab/roberta-fa-zwnj-base
|
HooshvareLab
|
roberta
| 11 | 1,591 |
transformers
| 0 |
fill-mask
| true | true | true |
apache-2.0
|
['fa']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 290 |
# Roberta
This model can tackle the zero-width non-joiner character for Persian writing. Also, the model was trained on new multi-types corpora with a new set of vocabulary.
## Questions?
Post a Github issue on the [ParsRoBERTa Issues](https://github.com/hooshvare/parsbert/issues) repo.
|
Hormigo/roberta-base-bne-finetuned-amazon_reviews_multi
|
Hormigo
|
roberta
| 13 | 3 |
transformers
| 0 |
text-classification
| true | false | false |
cc-by-4.0
| null |
['amazon_reviews_multi']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| false | true | true | 1,317 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# roberta-base-bne-finetuned-amazon_reviews_multi
This model is a fine-tuned version of [BSC-TeMU/roberta-base-bne](https://huggingface.co/BSC-TeMU/roberta-base-bne) on the amazon_reviews_multi dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2275
- Accuracy: 0.9335
## 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: 2
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| 0.1909 | 1.0 | 1250 | 0.1717 | 0.9333 |
| 0.0932 | 2.0 | 2500 | 0.2275 | 0.9335 |
### Framework versions
- Transformers 4.9.2
- Pytorch 1.9.0+cu102
- Datasets 1.11.0
- Tokenizers 0.10.3
|
Huertas97/en_roberta_base_leetspeak_ner
|
Huertas97
| null | 17 | 4 |
spacy
| 1 |
token-classification
| false | false | false |
apache-2.0
|
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['spacy', 'token-classification']
| false | true | true | 1,159 |
| Feature | Description |
| --- | --- |
| **Name** | `en_roberta_base_leetspeak_ner` |
| **Version** | `0.0.0` |
| **spaCy** | `>=3.2.1,<3.3.0` |
| **Default Pipeline** | `transformer`, `ner` |
| **Components** | `transformer`, `ner` |
| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
| **Sources** | [roberta-base](https://huggingface.co/roberta-base) pre-trained model on English language using a masked language modeling (MLM) objective by Yinhan Liu et al. <br> [LeetSpeak-NER](https://huggingface.co/spaces/Huertas97/LeetSpeak-NER) app where this model is in production for countering information disorders|
| **License** | Apache 2.0 |
| **Author** | [Álvaro Huertas García](https://www.linkedin.com/in/alvaro-huertas-garcia/) at [AI+DA](http://aida.etsisi.upm.es/) |
### Label Scheme
<details>
<summary>View label scheme (4 labels for 1 components)</summary>
| Component | Labels |
| --- | --- |
| **`ner`** | `INV_CAMO`, `LEETSPEAK`, `MIX`, `PUNCT_CAMO` |
</details>
### Accuracy
| Type | Score |
| --- | --- |
| `ENTS_F` | 82.80 |
| `ENTS_P` | 79.66 |
| `ENTS_R` | 86.20 |
| `TRANSFORMER_LOSS` | 177808.42 |
| `NER_LOSS` | 608427.31 |
|
Huertas97/es_roberta_base_bne_leetspeak_ner
|
Huertas97
| null | 18 | 12 |
spacy
| 1 |
token-classification
| false | false | false |
apache-2.0
|
['es']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['spacy', 'token-classification']
| false | true | true | 1,337 |
| Feature | Description |
| --- | --- |
| **Name** | `es_roberta_base_bne_leetspeak_ner` |
| **Version** | `0.0.0` |
| **spaCy** | `>=3.2.1,<3.3.0` |
| **Default Pipeline** | `transformer`, `ner` |
| **Components** | `transformer`, `ner` |
| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
| **Sources** | [PlanTL-GOB-ES/roberta-base-bne](https://huggingface.co/PlanTL-GOB-ES/roberta-base-bne) model a transformer-based masked language model for the Spanish language pre-trained with a total of 570GB of clean and deduplicated text compiled from the web crawlings performed by the National Library of Spain (Biblioteca Nacional de España) <br> [LeetSpeak-NER](https://huggingface.co/spaces/Huertas97/LeetSpeak-NER) app where this model is in production for countering information disorders|
| **License** | Apache 2.0 |
| **Author** | [Álvaro Huertas García](https://www.linkedin.com/in/alvaro-huertas-garcia/) at [AI+DA](http://aida.etsisi.upm.es/) |
### Label Scheme
<details>
<summary>View label scheme (4 labels for 1 components)</summary>
| Component | Labels |
| --- | --- |
| **`ner`** | `INV_CAMO`, `LEETSPEAK`, `MIX`, `PUNCT_CAMO` |
</details>
### Accuracy
| Type | Score |
| --- | --- |
| `ENTS_F` | 91.82 |
| `ENTS_P` | 89.79 |
| `ENTS_R` | 93.94 |
| `TRANSFORMER_LOSS` | 166484.92 |
| `NER_LOSS` | 318457.35 |
|
HueyNemud/das22-10-camembert_pretrained
|
HueyNemud
|
camembert
| 13 | 394 |
transformers
| 0 |
fill-mask
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 2,525 |
# CamemBERT pretrained on french trade directories from the XIXth century
This mdoel is part of the material of the paper
> Abadie, N., Carlinet, E., Chazalon, J., Duménieu, B. (2022). A
> Benchmark of Named Entity Recognition Approaches in Historical
> Documents Application to 19𝑡ℎ Century French Directories. In: Uchida,
> S., Barney, E., Eglin, V. (eds) Document Analysis Systems. DAS 2022.
> Lecture Notes in Computer Science, vol 13237. Springer, Cham.
> https://doi.org/10.1007/978-3-031-06555-2_30
The source code to train this model is available on the [GitHub repository](https://github.com/soduco/paper-ner-bench-das22) of the paper as a Jupyter notebook in `src/ner/10-camembert_pretraining.ipynb`.
## Model description
This model pre-train the model [Jean-Baptiste/camembert-ner](https://huggingface.co/Jean-Baptiste/camembert-ner) on a set of ~845k entries from Paris trade directories from the XIXth century extracted with OCR.
Trade directory entries are short and strongly structured texts that giving the name, activity and location of a person or business, e.g:
```
Peynaud, R. de la Vieille Bouclerie, 18. Richard, Joullain et comp., (commission- —Phéâtre Français. naire, (entrepôt), au port de la Rapée-
```
## Intended uses & limitations
This model is intended for reproducibility of the NER evaluation published in the DAS2022 paper.
Several derived models trained for NER on trade directories are available on HuggingFace, each trained on a different dataset :
- [das22-10-camembert_pretrained_finetuned_ref](): trained for NER on ~6000 directory entries manually corrected.
- [das22-10-camembert_pretrained_finetuned_pero](): trained for NER on ~6000 directory entries extracted with PERO-OCR.
- [das22-10-camembert_pretrained_finetuned_tess](): trained for NER on ~6000 directory entries extracted with Tesseract.
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3.0
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:------:|:---------------:|
| 1.9603 | 1.0 | 100346 | 1.8005 |
| 1.7032 | 2.0 | 200692 | 1.6460 |
| 1.5879 | 3.0 | 301038 | 1.5570 |
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.1+cu102
- Datasets 1.17.0
- Tokenizers 0.10.3
|
Huffon/sentence-klue-roberta-base
|
Huffon
|
roberta
| 8 | 1,104 |
sentence-transformers
| 5 | null | true | false | false | null |
['ko']
|
['klue']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['roberta', 'sentence-transformers']
| false | true | true | 1,943 |
# KLUE RoBERTa base model for Sentence Embeddings
This is the `sentence-klue-roberta-base` model. The sentence-transformers repository allows to train and use Transformer models for generating sentence and text embeddings.
The model is described in the paper [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://arxiv.org/abs/1908.10084)
## Usage (Sentence-Transformers)
Using this model becomes more convenient when you have [sentence-transformers](https://github.com/UKPLab/sentence-transformers) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
import torch
from sentence_transformers import SentenceTransformer, util
model = SentenceTransformer("Huffon/sentence-klue-roberta-base")
docs = [
"1992년 7월 8일 손흥민은 강원도 춘천시 후평동에서 아버지 손웅정과 어머니 길은자의 차남으로 태어나 그곳에서 자랐다.",
"형은 손흥윤이다.",
"춘천 부안초등학교를 졸업했고, 춘천 후평중학교에 입학한 후 2학년때 원주 육민관중학교 축구부에 들어가기 위해 전학하여 졸업하였으며, 2008년 당시 FC 서울의 U-18팀이었던 동북고등학교 축구부에서 선수 활동 중 대한축구협회 우수선수 해외유학 프로젝트에 선발되어 2008년 8월 독일 분데스리가의 함부르크 유소년팀에 입단하였다.",
"함부르크 유스팀 주전 공격수로 2008년 6월 네덜란드에서 열린 4개국 경기에서 4게임에 출전, 3골을 터뜨렸다.",
"1년간의 유학 후 2009년 8월 한국으로 돌아온 후 10월에 개막한 FIFA U-17 월드컵에 출전하여 3골을 터트리며 한국을 8강으로 이끌었다.",
"그해 11월 함부르크의 정식 유소년팀 선수 계약을 체결하였으며 독일 U-19 리그 4경기 2골을 넣고 2군 리그에 출전을 시작했다.",
"독일 U-19 리그에서 손흥민은 11경기 6골, 2부 리그에서는 6경기 1골을 넣으며 재능을 인정받아 2010년 6월 17세의 나이로 함부르크의 1군 팀 훈련에 참가, 프리시즌 활약으로 함부르크와 정식 계약을 한 후 10월 18세에 함부르크 1군 소속으로 독일 분데스리가에 데뷔하였다.",
]
document_embeddings = model.encode(docs)
query = "손흥민은 어린 나이에 유럽에 진출하였다."
query_embedding = model.encode(query)
top_k = min(5, len(docs))
cos_scores = util.pytorch_cos_sim(query_embedding, document_embeddings)[0]
top_results = torch.topk(cos_scores, k=top_k)
print(f"입력 문장: {query}")
print(f"<입력 문장과 유사한 {top_k} 개의 문장>")
for i, (score, idx) in enumerate(zip(top_results[0], top_results[1])):
print(f"{i+1}: {docs[idx]} {'(유사도: {:.4f})'.format(score)}")
```
|
Humair/all-mpnet-base-v2-finetuned-v2
|
Humair
|
mpnet
| 12 | 6 |
sentence-transformers
| 0 |
sentence-similarity
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['sentence-transformers', 'feature-extraction', 'sentence-similarity', 'transformers']
| false | true | true | 3,792 |
# Humair/all-mpnet-base-v2-finetuned-v2
This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.
<!--- Describe your model here -->
## Usage (Sentence-Transformers)
Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed:
```
pip install -U sentence-transformers
```
Then you can use the model like this:
```python
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('Humair/all-mpnet-base-v2-finetuned-v2')
embeddings = model.encode(sentences)
print(embeddings)
```
## Usage (HuggingFace Transformers)
Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
```python
from transformers import AutoTokenizer, AutoModel
import torch
#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0] #First element of model_output contains all token embeddings
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']
# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('Humair/all-mpnet-base-v2-finetuned-v2')
model = AutoModel.from_pretrained('Humair/all-mpnet-base-v2-finetuned-v2')
# Tokenize sentences
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
# Compute token embeddings
with torch.no_grad():
model_output = model(**encoded_input)
# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
```
## Evaluation Results
<!--- Describe how your model was evaluated -->
For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name=Humair/all-mpnet-base-v2-finetuned-v2)
## Training
The model was trained with the parameters:
**DataLoader**:
`torch.utils.data.dataloader.DataLoader` of length 1 with parameters:
```
{'batch_size': 128, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'}
```
**Loss**:
`sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss` with parameters:
```
{'scale': 20.0, 'similarity_fct': 'cos_sim'}
```
Parameters of the fit()-Method:
```
{
"epochs": 2,
"evaluation_steps": 0,
"evaluator": "NoneType",
"max_grad_norm": 1,
"optimizer_class": "<class 'transformers.optimization.AdamW'>",
"optimizer_params": {
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 10000,
"weight_decay": 0.01
}
```
## Full Model Architecture
```
SentenceTransformer(
(0): Transformer({'max_seq_length': 32, 'do_lower_case': False}) with Transformer model: MPNetModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
```
## Citing & Authors
<!--- Describe where people can find more information -->
|
Hyeon/distilbert-base-uncased-finetuned-cola
|
Hyeon
|
distilbert
| 13 | 3 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
| null |
['glue']
| null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 1,572 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-cola
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8575
- Matthews Correlation: 0.5443
## 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: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|
| 0.5242 | 1.0 | 535 | 0.5258 | 0.4391 |
| 0.346 | 2.0 | 1070 | 0.5264 | 0.5074 |
| 0.2334 | 3.0 | 1605 | 0.6808 | 0.5074 |
| 0.1711 | 4.0 | 2140 | 0.7737 | 0.5373 |
| 0.1205 | 5.0 | 2675 | 0.8575 | 0.5443 |
### Framework versions
- Transformers 4.15.0
- Pytorch 1.10.0+cu111
- Datasets 1.17.0
- Tokenizers 0.10.3
|
IDEA-CCNL/Erlangshen-MegatronBert-1.3B
|
IDEA-CCNL
|
megatron-bert
| 5 | 390 |
transformers
| 7 | null | true | false | false |
apache-2.0
|
['zh']
| null | null | 0 | 0 | 0 | 0 | 1 | 1 | 0 |
['bert', 'NLU', 'FewCLUE', 'ZeroCLUE']
| false | true | true | 4,569 |
# Erlangshen-MegatronBert-1.3B
- Github: [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)
- Docs: [Fengshenbang-Docs](https://fengshenbang-doc.readthedocs.io/)
## 简介 Brief Introduction
2021登顶FewCLUE和ZeroCLUE，处理NLU任务，开源时最大的中文BERT模型
It topped FewCLUE and ZeroCLUE benchmarks in 2021, solving NLU tasks, was the largest BERT when publicly released.
## 模型分类 Model Taxonomy
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
| :----: | :----: | :----: | :----: | :----: | :----: |
| 通用 General | 自然语言理解 NLU | 二郎神 Erlangshen | MegatronBERT | 1.3B | 中文 Chinese |
## 模型信息 Model Information
Encoder结构为主的双向语言模型，专注于解决各种自然语言理解任务。
我们跟进了[Megatron-LM](https://github.com/NVIDIA/Megatron-LM)的工作，使用了32张A100，总共耗时14天在悟道语料库（180 GB版本）上训练了十亿级别参数量的BERT。同时，鉴于中文语法和大规模训练的难度，我们使用四种预训练策略来改进BERT：1) 整词掩码, 2) 知识动态遮掩, 3) 句子顺序预测, 4) 层前归一化.
A bidirectional language model based on the Encoder structure, focusing on solving various NLU tasks.
We follow [Megatron-LM](https://github.com/NVIDIA/Megatron-LM), using 32 A100s and spending 14 days training a billion-level BERT on WuDao Corpora (180 GB version). Given Chinese grammar and the difficulty of large-scale training, we use four pre-training procedures to improve BERT: 1) Whole Word Masking (WWM), 2) Knowledge-based Dynamic Masking (KDM), 3) Sentence Order Prediction (SOP), 4) Pre-layer Normalization (Pre-LN).
### 成就 Achievements
1.2021年11月10日，Erlangshen-MegatronBert-1.3B在FewCLUE上取得第一。其中，它在CHIDF(成语填空)和TNEWS(新闻分类)子任务中的表现优于人类表现。此外，它在CHIDF(成语填空), CSLDCP(学科文献分类), OCNLI(自然语言推理)任务中均名列前茅。
2.2022年1月24日，Erlangshen-MegatronBert-1.3B在CLUE基准测试中的ZeroCLUE中取得第一。具体到子任务，我们在CSLDCP(主题文献分类), TNEWS(新闻分类), IFLYTEK(应用描述分类), CSL(抽象关键字识别)和CLUEWSC(参考消歧)任务中取得第一。
3.在2022年7月10日，Erlangshen-MegatronBert-1.3B在CLUE基准的语义匹配任务中取得第一。
1.On November 10, 2021, Erlangshen-MegatronBert-1.3B topped the FewCLUE benchmark. Among them, our Erlangshen outperformed human performance in CHIDF (idiom fill-in-the-blank) and TNEWS (news classification) subtasks. In addition, our Erlangshen ranked the top in CHIDF (idiom fill-in-the-blank), CSLDCP (subject literature classification), and OCNLI (natural language inference) tasks.
2.On January 24, 2022, Erlangshen-MegatronBert-1.3B topped the ZeroCLUE benchmark. For each of these tasks, we rank the top ones in CSLDCP (Subject Literature Classification), TNEWS (News Classification), IFLYTEK (Application Description Classification), CSL (Abstract Keyword Recognition), and CLUEWSC (Referential Disambiguation) tasks.
3.Erlangshen-MegatronBert-1.3B topped the CLUE benchmark semantic matching task on July 10, 2022.
### 下游效果 Performance
| 模型 | afqmc | tnews | iflytek | ocnli | cmnli | wsc | csl |
| :--------: | :-----: | :----: | :-----: | :----: | :----: | :----: | :----: |
| roberta-wwm-ext-large | 0.7514 | 0.5872 | 0.6152 | 0.777 | 0.814 | 0.8914 | 0.86 |
| Erlangshen-MegatronBert-1.3B | 0.7608 | 0.5996 | 0.6234 | 0.7917 | 0.81 | 0.9243 | 0.872 |
## 使用 Usage
``` python
from transformers import MegatronBertConfig, MegatronBertModel
from transformers import BertTokenizer
tokenizer = BertTokenizer.from_pretrained("IDEA-CCNL/Erlangshen-MegatronBert-1.3B")
config = MegatronBertConfig.from_pretrained("IDEA-CCNL/Erlangshen-MegatronBert-1.3B")
model = MegatronBertModel.from_pretrained("IDEA-CCNL/Erlangshen-MegatronBert-1.3B")
```
## 引用 Citation
如果您在您的工作中使用了我们的模型，可以引用我们的[论文](https://arxiv.org/abs/2209.02970)：
If you are using the resource for your work, please cite the our [paper](https://arxiv.org/abs/2209.02970):
```text
@article{fengshenbang,
author = {Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen and Ruyi Gan and Jiaxing Zhang},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}
```
也可以引用我们的[网站](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
You can also cite our [website](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
```text
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}
```
|
IDEA-CCNL/Randeng-MegatronT5-770M
|
IDEA-CCNL
|
t5
| 5 | 5 |
transformers
| 5 |
text2text-generation
| true | false | false |
apache-2.0
|
['zh']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,968 |
# Randeng-MegatronT5-770M
- Github: [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)
- Docs: [Fengshenbang-Docs](https://fengshenbang-doc.readthedocs.io/)
## 简介 Brief Introduction
善于处理NLT任务，中文版的T5-large。
Good at solving NLT tasks, Chinese T5-large.
## 模型分类 Model Taxonomy
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
| :----: | :----: | :----: | :----: | :----: | :----: |
| 通用 General | 自然语言转换 NLT | 燃灯 Randeng | MegatronT5 | 770M | 中文-Chinese |
## 模型信息 Model Information
为了得到一个大规模的中文版的T5，我们使用了Megatron-LM的方法和悟道语料库(180G版本)用于预训练。具体地，我们在预训练阶段中使用了[Megatron-LM](https://github.com/NVIDIA/Megatron-LM) 大概花费了16张A100约14天。
To get a large-scale Chinese T5, we use of [Megatron-LM](https://github.com/NVIDIA/Megatron-LM) and WuDao Corpora (180 GB version) for pre-training. Specifically, in the pre-training phase which cost about 14 days with 16 A100 GPUs.
## 使用 Usage
因为[transformers](https://github.com/huggingface/transformers)库中是没有Randeng-MegatronT5-770M相关的模型结构的，所以你可以在我们的[Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)中找到并且运行代码。
Since there is no structure of Randeng-MegatronT5-770M in [transformers library](https://github.com/huggingface/transformers), you can find the structure of Randeng-MegatronT5-770M and run the codes in [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM).
```shell
git clone https://github.com/IDEA-CCNL/Fengshenbang-LM.git
```
### 加载模型 Loading Models
```python
from fengshen import T5ForConditionalGeneration
from fengshen import T5Config
from fengshen import T5Tokenizer
tokenizer = T5Tokenizer.from_pretrained('IDEA-CCNL/Randeng-MegatronT5-770M')
config = T5Config.from_pretrained('IDEA-CCNL/Randeng-MegatronT5-770M')
model = T5ForConditionalGeneration.from_pretrained('IDEA-CCNL/Randeng-MegatronT5-770M')
```
## 引用 Citation
如果您在您的工作中使用了我们的模型，可以引用我们的[论文](https://arxiv.org/abs/2209.02970)：
If you are using the resource for your work, please cite the our [paper](https://arxiv.org/abs/2209.02970):
```text
@article{fengshenbang,
author = {Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen and Ruyi Gan and Jiaxing Zhang},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}
```
也可以引用我们的[网站](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
You can also cite our [website](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
```text
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}
```
|
IDEA-CCNL/Wenzhong-GPT2-3.5B
|
IDEA-CCNL
|
gpt2
| 9 | 214 |
transformers
| 4 |
text-generation
| true | false | false |
apache-2.0
|
['zh']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,853 |
# Wenzhong-GPT2-3.5B
- Github: [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)
- Docs: [Fengshenbang-Docs](https://fengshenbang-doc.readthedocs.io/)
## 简介 Brief Introduction
善于处理NLG任务，目前最大的，中文版的GPT2
Focused on handling NLG tasks, the current largest, Chinese GPT2.
## 模型分类 Model Taxonomy
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
| :----: | :----: | :----: | :----: | :----: | :----: |
| 通用 General | 自然语言生成 NLG| 闻仲 Wenzhong | GPT2 | 3.5B | 中文 Chinese |
## 模型信息 Model Information
为了可以获得一个强大的单向语言模型，我们采用GPT模型结构，并且应用于中文语料上。具体地，这个模型拥有30层解码器和35亿参数，这比原本的GPT2-XL还要大。我们在100G的中文语料上预训练，这消耗了32个NVIDIA A100显卡大约28小时。据我们所知，它是目前最大的中文的GPT模型。
To obtain a robust unidirectional language model, we adopt the GPT model structure and apply it to the Chinese corpus. Specifically, this model has 30 decoder layers and 3.5 billion parameters, which is larger than the original GPT2-XL. We pre-train it on 100G of Chinese corpus, which consumes 32 NVIDIA A100 GPUs for about 28 hours. To the best of our knowledge, it is the largest Chinese GPT model currently available.
## 使用 Usage
### 加载模型 Loading Models
```python
from transformers import GPT2Tokenizer, GPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('IDEA-CCNL/Wenzhong-GPT2-3.5B')
model = GPT2Model.from_pretrained('IDEA-CCNL/Wenzhong-GPT2-3.5B')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
### 使用示例 Usage Examples
```python
from transformers import pipeline, set_seed
set_seed(55)
generator = pipeline('text-generation', model='IDEA-CCNL/Wenzhong-GPT2-3.5B')
generator("北京位于", max_length=30, num_return_sequences=1)
```
## 引用 Citation
如果您在您的工作中使用了我们的模型，可以引用我们的[论文](https://arxiv.org/abs/2209.02970)：
If you are using the resource for your work, please cite the our [paper](https://arxiv.org/abs/2209.02970):
```text
@article{fengshenbang,
author = {Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen and Ruyi Gan and Jiaxing Zhang},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}
```
也可以引用我们的[网站](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
You can also cite our [website](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
```text
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}
```
|
IDEA-CCNL/Yuyuan-GPT2-3.5B
|
IDEA-CCNL
|
gpt2
| 10 | 54 |
transformers
| 2 |
text-generation
| true | false | false |
apache-2.0
|
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 2,891 |
# Yuyuan-GPT2-3.5B
- Github: [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)
- Docs: [Fengshenbang-Docs](https://fengshenbang-doc.readthedocs.io/)
## 简介 Brief Introduction
目前最大的，医疗领域的生成语言模型GPT2。
The currently largest, generative language model GPT2 in the medical domain.
## 模型分类 Model Taxonomy
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
| :----: | :----: | :----: | :----: | :----: | :----: |
| 特殊 Special | 领域 Domain | 余元 Yuyuan | GPT2 | 3.5B | - |
## 模型信息 Model Information
我们采用与Wenzhong-GPT2-3.5B相同的架构，在50GB的医学(PubMed)语料库上进行预训练。我们使用了32个NVIDIA A100显卡大约7天。我们的Yuyuan-GPT2-3.5B是医疗领域最大的开源的GPT2模型。进一步地，模型可以通过计算困惑度（PPL）来判断事实。为了完成问答功能，我们将短语模式从疑问的形式转换为了陈述句。
We adopt the same architecture as Wenzhong-GPT2-3.5B to be pre-trained on 50 GB medical (PubMed) corpus. We use 32 NVIDIA A100 GPUs for about 7 days. Our Yuyuan-GPT2-3.5B is the largest open-source GPT2 model in the medical domain. We further allow the model to judge facts by computing perplexity (PPL). To accomplish question-and-answer functionality, we transform the phrase pattern from interrogative to declarative.
## 使用 Usage
### 加载模型 Loading Models
```python
from transformers import GPT2Tokenizer, GPT2Model
tokenizer = GPT2Tokenizer.from_pretrained('IDEA-CCNL/Yuyuan-GPT2-3.5B')
model = GPT2Model.from_pretrained('IDEA-CCNL/Yuyuan-GPT2-3.5B')
text = "Replace me by any text you'd like."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)
```
### 使用示例 Usage Examples
```python
from transformers import pipeline, set_seed
set_seed(55)
generator = pipeline('text-generation', model='IDEA-CCNL/Yuyuan-GPT2-3.5B')
generator("Diabetics should not eat", max_length=30, num_return_sequences=1)
```
## 引用 Citation
如果您在您的工作中使用了我们的模型，可以引用我们的[论文](https://arxiv.org/abs/2209.02970)：
If you are using the resource for your work, please cite the our [paper](https://arxiv.org/abs/2209.02970):
```text
@article{fengshenbang,
author = {Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen and Ruyi Gan and Jiaxing Zhang},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}
```
也可以引用我们的[网站](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
You can also cite our [website](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
```text
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}
```
|
IDEA-CCNL/Zhouwenwang-Unified-1.3B
|
IDEA-CCNL
|
megatron-bert
| 5 | 49 |
transformers
| 1 | null | true | false | false |
apache-2.0
|
['zh']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 4,952 |
# Zhouwenwang-Unified-1.3B
- Github: [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)
- Docs: [Fengshenbang-Docs](https://fengshenbang-doc.readthedocs.io/)
## 简介 Brief Introduction
与追一科技合作探索的中文统一模型，13亿参数的编码器结构模型。
The Chinese unified model explored in cooperation with Zhuiyi Technology, the encoder structure model with 1.3B parameters.
## 模型分类 Model Taxonomy
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
| :----: | :----: | :----: | :----: | :----: | :----: |
| 特殊 Special | 探索 Exploration | 周文王 Zhouwenwang | 待定 TBD | 1.3B | 中文 Chinese |
## 模型信息 Model Information
IDEA研究院认知计算中心联合追一科技有限公司提出的具有新结构的大模型。该模型在预训练阶段时考虑统一LM和MLM的任务，这让其同时具备生成和理解的能力，并且增加了旋转位置编码技术。目前已有13亿参数的Zhouwenwang-Unified-1.3B大模型，是中文领域中可以同时做LM和MLM任务的最大的模型。我们后续会持续在模型规模、知识融入、监督辅助任务等方向不断优化。
A large-scale model (Zhouwenwang-Unified-1.3B) with a new structure proposed by IDEA CCNL and Zhuiyi Technology. The model considers the task of unifying LM (Language Modeling) and MLM (Masked Language Modeling) during the pre-training phase, which gives it both generative and comprehension capabilities, and applys rotational position encoding. At present, Zhouwenwang-Unified-1.3B with 13B parameters is the largest Chinese model that can do both LM and MLM tasks. In the future, we will continue to optimize it in the direction of model size, knowledge incorporation, and supervisory assistance tasks.
### 下游任务 Performance
下游中文任务的得分（没有做任何数据增强）。
Scores on downstream chinese tasks (without any data augmentation)
| 模型 Model | afqmc | tnews | iflytek | ocnli | cmnli | wsc | csl |
| :--------: | :-----: | :----: | :-----: | :----: | :----: | :----: | :----: |
| roberta-wwm-ext-large | 0.7514 | 0.5872 | 0.6152 | 0.7770 | 0.8140 | 0.8914 | 0.8600 |
| Zhouwenwang-Unified-1.3B | 0.7463 | 0.6036 | 0.6288 | 0.7654 | 0.7741 | 0.8849 | 0. 8777 |
## 使用 Usage
因为[transformers](https://github.com/huggingface/transformers)库中是没有 Zhouwenwang-Unified-1.3B相关的模型结构的，所以你可以在我们的[Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)中找到并且运行代码。
Since there is no structure of Zhouwenwang-Unified-1.3B in [transformers library](https://github.com/huggingface/transformers), you can find the structure of Zhouwenwang-Unified-1.3B and run the codes in [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM).
```shell
git clone https://github.com/IDEA-CCNL/Fengshenbang-LM.git
```
### 加载模型 Loading Models
```python
from fengshen import RoFormerModel
from fengshen import RoFormerConfig
from transformers import BertTokenizer
tokenizer = BertTokenizer.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-1.3B")
config = RoFormerConfig.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-1.3B")
model = RoFormerModel.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-1.3B")
```
### 使用示例 Usage Examples
你可以使用该模型进行续写任务。
You can use the model for continuation writing tasks.
```python
from fengshen import RoFormerModel
from transformers import AutoTokenizer
import torch
import numpy as np
sentence = '清华大学位于'
max_length = 32
tokenizer = AutoTokenizer.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-1.3B")
model = RoFormerModel.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-1.3B")
for i in range(max_length):
encode = torch.tensor(
[[tokenizer.cls_token_id]+tokenizer.encode(sentence, add_special_tokens=False)]).long()
logits = model(encode)[0]
logits = torch.nn.functional.linear(
logits, model.embeddings.word_embeddings.weight)
logits = torch.nn.functional.softmax(
logits, dim=-1).cpu().detach().numpy()[0]
sentence = sentence + \
tokenizer.decode(int(np.random.choice(logits.shape[1], p=logits[-1])))
if sentence[-1] == '。':
break
print(sentence)
```
## 引用 Citation
如果您在您的工作中使用了我们的模型，可以引用我们的[论文](https://arxiv.org/abs/2209.02970)：
If you are using the resource for your work, please cite the our [paper](https://arxiv.org/abs/2209.02970):
```text
@article{fengshenbang,
author = {Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen and Ruyi Gan and Jiaxing Zhang},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}
```
也可以引用我们的[网站](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
You can also cite our [website](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
```text
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}
```
|
IDEA-CCNL/Zhouwenwang-Unified-110M
|
IDEA-CCNL
|
megatron-bert
| 5 | 16 |
transformers
| 3 | null | true | false | false |
apache-2.0
|
['zh']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 4,250 |
# Zhouwenwang-Unified-110M
- Github: [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)
- Docs: [Fengshenbang-Docs](https://fengshenbang-doc.readthedocs.io/)
## 简介 Brief Introduction
与追一科技合作探索的中文统一模型，1.1亿参数的编码器结构模型。
The Chinese unified model explored in cooperation with Zhuiyi Technology, the encoder structure model with 110M parameters.
## 模型分类 Model Taxonomy
| 需求 Demand | 任务 Task | 系列 Series | 模型 Model | 参数 Parameter | 额外 Extra |
| :----: | :----: | :----: | :----: | :----: | :----: |
| 特殊 Special | 探索 Exploration | 周文王 Zhouwenwang | 待定 TBD | 110M | 中文 Chinese |
## 模型信息 Model Information
IDEA研究院认知计算中心联合追一科技有限公司提出的具有新结构的大模型。该模型在预训练阶段时考虑统一LM和MLM的任务，这让其同时具备生成和理解的能力，并且增加了旋转位置编码技术。我们后续会持续在模型规模、知识融入、监督辅助任务等方向不断优化。
A large-scale model (Zhouwenwang-Unified-1.3B) with a new structure proposed by IDEA CCNL and Zhuiyi Technology. The model considers the task of unifying LM (Language Modeling) and MLM (Masked Language Modeling) during the pre-training phase, which gives it both generative and comprehension capabilities, and applys rotational position encoding. In the future, we will continue to optimize it in the direction of model size, knowledge incorporation, and supervisory assistance tasks.
## 使用 Usage
因为[transformers](https://github.com/huggingface/transformers)库中是没有 Zhouwenwang-Unified-110M相关的模型结构的，所以你可以在我们的[Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM)中找到并且运行代码。
Since there is no structure of Zhouwenwang-Unified-110M in [transformers library](https://github.com/huggingface/transformers), you can find the structure of Zhouwenwang-Unified-110M and run the codes in [Fengshenbang-LM](https://github.com/IDEA-CCNL/Fengshenbang-LM).
```shell
git clone https://github.com/IDEA-CCNL/Fengshenbang-LM.git
```
### 加载模型 Loading Models
```python
from fengshen import RoFormerModel
from fengshen import RoFormerConfig
from transformers import BertTokenizer
tokenizer = BertTokenizer.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-110M")
config = RoFormerConfig.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-110M")
model = RoFormerModel.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-110M")
```
### 使用示例 Usage Examples
你可以使用该模型进行续写任务。
You can use the model for continuation writing tasks.
```python
from fengshen import RoFormerModel
from transformers import AutoTokenizer
import torch
import numpy as np
sentence = '清华大学位于'
max_length = 32
tokenizer = AutoTokenizer.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-110M")
model = RoFormerModel.from_pretrained("IDEA-CCNL/Zhouwenwang-Unified-110M")
for i in range(max_length):
encode = torch.tensor(
[[tokenizer.cls_token_id]+tokenizer.encode(sentence, add_special_tokens=False)]).long()
logits = model(encode)[0]
logits = torch.nn.functional.linear(
logits, model.embeddings.word_embeddings.weight)
logits = torch.nn.functional.softmax(
logits, dim=-1).cpu().detach().numpy()[0]
sentence = sentence + \
tokenizer.decode(int(np.random.choice(logits.shape[1], p=logits[-1])))
if sentence[-1] == '。':
break
print(sentence)
```
## 引用 Citation
如果您在您的工作中使用了我们的模型，可以引用我们的[论文](https://arxiv.org/abs/2209.02970)：
If you are using the resource for your work, please cite the our [paper](https://arxiv.org/abs/2209.02970):
```text
@article{fengshenbang,
author = {Junjie Wang and Yuxiang Zhang and Lin Zhang and Ping Yang and Xinyu Gao and Ziwei Wu and Xiaoqun Dong and Junqing He and Jianheng Zhuo and Qi Yang and Yongfeng Huang and Xiayu Li and Yanghan Wu and Junyu Lu and Xinyu Zhu and Weifeng Chen and Ting Han and Kunhao Pan and Rui Wang and Hao Wang and Xiaojun Wu and Zhongshen Zeng and Chongpei Chen and Ruyi Gan and Jiaxing Zhang},
title = {Fengshenbang 1.0: Being the Foundation of Chinese Cognitive Intelligence},
journal = {CoRR},
volume = {abs/2209.02970},
year = {2022}
}
```
也可以引用我们的[网站](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
You can also cite our [website](https://github.com/IDEA-CCNL/Fengshenbang-LM/):
```text
@misc{Fengshenbang-LM,
title={Fengshenbang-LM},
author={IDEA-CCNL},
year={2021},
howpublished={\url{https://github.com/IDEA-CCNL/Fengshenbang-LM}},
}
```
|
IMJONEZZ/SlovenBERTcina
|
IMJONEZZ
|
roberta
| 21 | 3 |
transformers
| 1 |
fill-mask
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 1,147 |
#Slovak RoBERTA Masked Language Model
###83Mil Parameters in small model
Medium and Large models coming soon!
RoBERTA pretrained tokenizer vocab and merges included.
---
##Training params:
- **Dataset**:
8GB Slovak Monolingual dataset including ParaCrawl (monolingual), OSCAR, and several gigs of my own findings and cleaning.
- **Preprocessing**:
Tokenized with a pretrained ByteLevelBPETokenizer trained on the same dataset. Uncased, with s, pad, /s, unk, and mask special tokens.
- **Evaluation results**:
- Mnoho ľudí tu<mask>
* žije.
* žijú.
* je.
* trpí.
- Ako sa<mask>
* máte
* máš
* má
* hovorí
- Plážová sezóna pod Zoborom patrí medzi<mask> obdobia.
* ročné
* najkrajšie
* najobľúbenejšie
* najnáročnejšie
- **Limitations**:
The current model is fairly small, although it works very well. This model is meant to be finetuned on downstream tasks e.g. Part-of-Speech tagging, Question Answering, anything in GLUE or SUPERGLUE.
- **Credit**:
If you use this or any of my models in research or professional work, please credit me - Christopher Brousseau in said work.
|
IMSyPP/hate_speech_en
|
IMSyPP
|
bert
| 7 | 1,747 |
transformers
| 5 |
text-classification
| true | false | false |
mit
|
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 652 |
# Hate Speech Classifier for Social Media Content in English Language
A monolingual model for hate speech classification of social media content in English language. The model was trained on 103190 YouTube comments and tested on an independent test set of 20554 YouTube comments. It is based on English BERT base pre-trained language model.
## Tokenizer
During training the text was preprocessed using the original English BERT base tokenizer. We suggest the same tokenizer is used for inference.
## Model output
The model classifies each input into one of four distinct classes:
* 0 - acceptable
* 1 - inappropriate
* 2 - offensive
* 3 - violent
|
IMSyPP/hate_speech_it
|
IMSyPP
|
bert
| 6 | 83 |
transformers
| 0 |
text-classification
| true | false | false |
mit
|
['it']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 654 |
# Hate Speech Classifier for Social Media Content in Italian Language
A monolingual model for hate speech classification of social media content in Italian language. The model was trained on 119,670 YouTube comments and tested on an independent test set of 21,072 YouTube comments. It is based on Italian ALBERTO pre-trained language model.
## Tokenizer
During training the text was preprocessed using the original Italian ALBERTO tokenizer. We suggest the same tokenizer is used for inference.
## Model output
The model classifies each input into one of four distinct classes:
* 0 - acceptable
* 1 - inappropriate
* 2 - offensive
* 3 - violent
|
IMSyPP/hate_speech_nl
|
IMSyPP
|
bert
| 7 | 45 |
transformers
| 1 |
text-classification
| true | false | false |
mit
|
['nl']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 677 |
# Hate Speech Classifier for Social Media Content in Dutch
A monolingual model for hate speech classification of social media content in Dutch. The model was trained on 20000 social media posts (youtube, twitter, facebook) and tested on an independent test set of 2000 posts. It is based on thepre-trained language model [BERTje](https://huggingface.co/wietsedv/bert-base-dutch-cased).
## Tokenizer
During training the text was preprocessed using the BERTje tokenizer. We suggest the same tokenizer is used for inference.
## Model output
The model classifies each input into one of four distinct classes:
* 0 - acceptable
* 1 - inappropriate
* 2 - offensive
* 3 - violent
|
IMSyPP/hate_speech_slo
|
IMSyPP
|
bert
| 7 | 9 |
transformers
| 0 |
text-classification
| true | false | false |
mit
|
['sl']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 670 |
# Hate Speech Classifier for Social Media Content in Slovenian Language
A monolingual model for hate speech classification of social media content in Slovenian language. The model was trained on 50,000 Twitter comments and tested on an independent test set of 10,000 Twitter comments. It is based on multilingual CroSloEngual BERT pre-trained language model.
## Tokenizer
During training the text was preprocessed using the original CroSloEngual BERT tokenizer. We suggest the same tokenizer is used for inference.
## Model output
The model classifies each input into one of four distinct classes:
* 0 - acceptable
* 1 - inappropriate
* 2 - offensive
* 3 - violent
|
Iacopo/Shakespear-GPT2
|
Iacopo
|
gpt2
| 13 | 4 |
transformers
| 0 |
text-generation
| true | false | false |
mit
| null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 1,095 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# output
This model is a fine-tuned version of [gpt2](https://huggingface.co/gpt2) on a dataset of Shakespeare's plays.
## Model description
The model is the original gpt-2 model fine-tuned on a custom dataset.
## Intended uses & limitations
The model can be used to generate Shakespearean-like text. Consider that because it comes from plays, such a typographical structure might be reproduced.
## Training and evaluation data
Trained with Shakespeare's plays corpus.
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 1
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 2.0
### Training results
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.0+cu111
- Datasets 1.18.0
- Tokenizers 0.11.0
|
Ifromspace/GRIEFSOFT
|
Ifromspace
|
gpt2
| 6 | 6 |
transformers
| 1 |
text-generation
| true | false | false | null |
['ru']
| null | null | 0 | 0 | 0 | 0 | 1 | 1 | 0 |
['PyTorch', 'Transformers', '4ulan']
| false | true | true | 265 |
**Fork of https://huggingface.co/sberbank-ai/rugpt3large_based_on_gpt2**
Забавное для дискордика))00))
ROADMAP:
- Собираю датасетик из книжек про попаданцев. <------------------------- Сейчас тут.
- Дообучаю.
- Выбрасываю в дискордик.
https://discord.gg/HpeadKH
|
IlyaGusev/mbart_ru_sum_gazeta
|
IlyaGusev
|
mbart
| 7 | 1,318 |
transformers
| 18 |
summarization
| true | false | false |
apache-2.0
|
['ru']
|
['IlyaGusev/gazeta']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['summarization', 'mbart']
| false | true | true | 5,295 |
# MBARTRuSumGazeta
## Model description
This is a ported version of [fairseq model](https://www.dropbox.com/s/fijtntnifbt9h0k/gazeta_mbart_v2_fairseq.tar.gz).
For more details, please see [Dataset for Automatic Summarization of Russian News](https://arxiv.org/abs/2006.11063).
## Intended uses & limitations
#### How to use
Colab: [link](https://colab.research.google.com/drive/1wdo_nPZPk6dWAn1J8nGx4Z5Ef82jCCob)
```python
from transformers import MBartTokenizer, MBartForConditionalGeneration
model_name = "IlyaGusev/mbart_ru_sum_gazeta"
tokenizer = MBartTokenizer.from_pretrained(model_name)
model = MBartForConditionalGeneration.from_pretrained(model_name)
article_text = "..."
input_ids = tokenizer(
[article_text],
max_length=600,
padding="max_length",
truncation=True,
return_tensors="pt",
)["input_ids"]
output_ids = model.generate(
input_ids=input_ids,
no_repeat_ngram_size=4
)[0]
summary = tokenizer.decode(output_ids, skip_special_tokens=True)
print(summary)
```
#### Limitations and bias
- The model should work well with Gazeta.ru articles, but for any other agencies it can suffer from domain shift
## Training data
- Dataset: [Gazeta](https://huggingface.co/datasets/IlyaGusev/gazeta)
## Training procedure
- Fairseq training script: [train.sh](https://github.com/IlyaGusev/summarus/blob/master/external/bart_scripts/train.sh)
- Porting: [Colab link](https://colab.research.google.com/drive/13jXOlCpArV-lm4jZQ0VgOpj6nFBYrLAr)
## Eval results
* Train dataset: **Gazeta v1 train**
* Test dataset: **Gazeta v1 test**
* Source max_length: **600**
* Target max_length: **200**
* no_repeat_ngram_size: **4**
* num_beams: **5**
| Model | R-1-f | R-2-f | R-L-f | chrF | METEOR | BLEU | Avg char length |
|:--------------------------|:------|:------|:------|:-------|:-------|:-----|:-----|
| [mbart_ru_sum_gazeta](https://huggingface.co/IlyaGusev/mbart_ru_sum_gazeta) | **32.4** | 14.3 | 28.0 | 39.7 | **26.4** | 12.1 | 371 |
| [rut5_base_sum_gazeta](https://huggingface.co/IlyaGusev/rut5_base_sum_gazeta) | 32.2 | **14.4** | **28.1** | **39.8** | 25.7 | **12.3** | 330 |
| [rugpt3medium_sum_gazeta](https://huggingface.co/IlyaGusev/rugpt3medium_sum_gazeta) | 26.2 | 7.7 | 21.7 | 33.8 | 18.2 | 4.3 | 244 |
* Train dataset: **Gazeta v1 train**
* Test dataset: **Gazeta v2 test**
* Source max_length: **600**
* Target max_length: **200**
* no_repeat_ngram_size: **4**
* num_beams: **5**
| Model | R-1-f | R-2-f | R-L-f | chrF | METEOR | BLEU | Avg char length |
|:--------------------------|:------|:------|:------|:-------|:-------|:-----|:-----|
| [mbart_ru_sum_gazeta](https://huggingface.co/IlyaGusev/mbart_ru_sum_gazeta) | **28.7** | **11.1** | 24.4 | **37.3** | **22.7** | **9.4** | 373 |
| [rut5_base_sum_gazeta](https://huggingface.co/IlyaGusev/rut5_base_sum_gazeta) | 28.6 | **11.1** | **24.5** | 37.2 | 22.0 | **9.4** | 331 |
| [rugpt3medium_sum_gazeta](https://huggingface.co/IlyaGusev/rugpt3medium_sum_gazeta) | 24.1 | 6.5 | 19.8 | 32.1 | 16.3 | 3.6 | 242 |
Predicting all summaries:
```python
import json
import torch
from transformers import MBartTokenizer, MBartForConditionalGeneration
from datasets import load_dataset
def gen_batch(inputs, batch_size):
batch_start = 0
while batch_start < len(inputs):
yield inputs[batch_start: batch_start + batch_size]
batch_start += batch_size
def predict(
model_name,
input_records,
output_file,
max_source_tokens_count=600,
batch_size=4
):
device = "cuda" if torch.cuda.is_available() else "cpu"
tokenizer = MBartTokenizer.from_pretrained(model_name)
model = MBartForConditionalGeneration.from_pretrained(model_name).to(device)
predictions = []
for batch in gen_batch(inputs, batch_size):
texts = [r["text"] for r in batch]
input_ids = tokenizer(
batch,
return_tensors="pt",
padding="max_length",
truncation=True,
max_length=max_source_tokens_count
)["input_ids"].to(device)
output_ids = model.generate(
input_ids=input_ids,
no_repeat_ngram_size=4
)
summaries = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
for s in summaries:
print(s)
predictions.extend(summaries)
with open(output_file, "w") as w:
for p in predictions:
w.write(p.strip().replace("\n", " ") + "\n")
gazeta_test = load_dataset('IlyaGusev/gazeta', script_version="v1.0")["test"]
predict("IlyaGusev/mbart_ru_sum_gazeta", list(gazeta_test), "mbart_predictions.txt")
```
Evaluation: https://github.com/IlyaGusev/summarus/blob/master/evaluate.py
Flags: --language ru --tokenize-after --lower
### BibTeX entry and citation info
```bibtex
@InProceedings{10.1007/978-3-030-59082-6_9,
author="Gusev, Ilya",
editor="Filchenkov, Andrey and Kauttonen, Janne and Pivovarova, Lidia",
title="Dataset for Automatic Summarization of Russian News",
booktitle="Artificial Intelligence and Natural Language",
year="2020",
publisher="Springer International Publishing",
address="Cham",
pages="122--134",
isbn="978-3-030-59082-6"
}
```
|
IlyaGusev/rubert_ext_sum_gazeta
|
IlyaGusev
|
bert
| 8 | 346 |
transformers
| 0 |
token-classification
| true | false | false |
apache-2.0
|
['ru']
|
['IlyaGusev/gazeta']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['summarization', 'token-classification', 't5']
| false | true | true | 2,014 |
# RuBERTExtSumGazeta
## Model description
Model for extractive summarization based on [rubert-base-cased](DeepPavlov/rubert-base-cased)
## Intended uses & limitations
#### How to use
Colab: [link](https://colab.research.google.com/drive/1Q8_v3H-kxdJhZIiyLYat7Kj02qDq7M1L)
```python
import razdel
from transformers import AutoTokenizer, BertForTokenClassification
model_name = "IlyaGusev/rubert_ext_sum_gazeta"
tokenizer = AutoTokenizer.from_pretrained(model_name)
sep_token = tokenizer.sep_token
sep_token_id = tokenizer.sep_token_id
model = BertForTokenClassification.from_pretrained(model_name)
article_text = "..."
sentences = [s.text for s in razdel.sentenize(article_text)]
article_text = sep_token.join(sentences)
inputs = tokenizer(
[article_text],
max_length=500,
padding="max_length",
truncation=True,
return_tensors="pt",
)
sep_mask = inputs["input_ids"][0] == sep_token_id
# Fix token_type_ids
current_token_type_id = 0
for pos, input_id in enumerate(inputs["input_ids"][0]):
inputs["token_type_ids"][0][pos] = current_token_type_id
if input_id == sep_token_id:
current_token_type_id = 1 - current_token_type_id
# Infer model
with torch.no_grad():
outputs = model(**inputs)
logits = outputs.logits[0, :, 1]
# Choose sentences
logits = logits[sep_mask]
logits, indices = logits.sort(descending=True)
logits, indices = logits.cpu().tolist(), indices.cpu().tolist()
pairs = list(zip(logits, indices))
pairs = pairs[:3]
indices = list(sorted([idx for _, idx in pairs]))
summary = " ".join([sentences[idx] for idx in indices])
print(summary)
```
#### Limitations and bias
- The model should work well with Gazeta.ru articles, but for any other agencies it can suffer from domain shift
## Training data
- Dataset: [Gazeta](https://huggingface.co/datasets/IlyaGusev/gazeta)
## Training procedure
TBD
## Eval results
TBD
Evaluation: https://github.com/IlyaGusev/summarus/blob/master/evaluate.py
Flags: --language ru --tokenize-after --lower
|
IlyaGusev/rubert_telegram_headlines
|
IlyaGusev
|
encoder-decoder
| 6 | 135 |
transformers
| 3 |
summarization
| true | false | false |
apache-2.0
|
['ru']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['summarization']
| false | true | true | 5,514 |
# RuBertTelegramHeadlines
## Model description
Example model for [Headline generation competition](https://competitions.codalab.org/competitions/29905)
Based on [RuBERT](http://docs.deeppavlov.ai/en/master/features/models/bert.html) model
## Intended uses & limitations
#### How to use
```python
from transformers import AutoTokenizer, EncoderDecoderModel
model_name = "IlyaGusev/rubert_telegram_headlines"
tokenizer = AutoTokenizer.from_pretrained(model_name, do_lower_case=False, do_basic_tokenize=False, strip_accents=False)
model = EncoderDecoderModel.from_pretrained(model_name)
article_text = "..."
input_ids = tokenizer(
[article_text],
add_special_tokens=True,
max_length=256,
padding="max_length",
truncation=True,
return_tensors="pt",
)["input_ids"]
output_ids = model.generate(
input_ids=input_ids,
max_length=64,
no_repeat_ngram_size=3,
num_beams=10,
top_p=0.95
)[0]
headline = tokenizer.decode(output_ids, skip_special_tokens=True, clean_up_tokenization_spaces=True)
print(headline)
```
## Training data
- Dataset: [ru_all_split.tar.gz](https://www.dropbox.com/s/ykqk49a8avlmnaf/ru_all_split.tar.gz)
## Training procedure
```python
import random
import torch
from torch.utils.data import Dataset
from tqdm.notebook import tqdm
from transformers import BertTokenizer, EncoderDecoderModel, Trainer, TrainingArguments, logging
def convert_to_tensors(
tokenizer,
text,
max_text_tokens_count,
max_title_tokens_count = None,
title = None
):
inputs = tokenizer(
text,
add_special_tokens=True,
max_length=max_text_tokens_count,
padding="max_length",
truncation=True
)
result = {
"input_ids": torch.tensor(inputs["input_ids"]),
"attention_mask": torch.tensor(inputs["attention_mask"]),
}
if title is not None:
outputs = tokenizer(
title,
add_special_tokens=True,
max_length=max_title_tokens_count,
padding="max_length",
truncation=True
)
decoder_input_ids = torch.tensor(outputs["input_ids"])
decoder_attention_mask = torch.tensor(outputs["attention_mask"])
labels = decoder_input_ids.clone()
labels[decoder_attention_mask == 0] = -100
result.update({
"labels": labels,
"decoder_input_ids": decoder_input_ids,
"decoder_attention_mask": decoder_attention_mask
})
return result
class GetTitleDataset(Dataset):
def __init__(
self,
original_records,
sample_rate,
tokenizer,
max_text_tokens_count,
max_title_tokens_count
):
self.original_records = original_records
self.sample_rate = sample_rate
self.tokenizer = tokenizer
self.max_text_tokens_count = max_text_tokens_count
self.max_title_tokens_count = max_title_tokens_count
self.records = []
for record in tqdm(original_records):
if random.random() > self.sample_rate:
continue
tensors = convert_to_tensors(
tokenizer=tokenizer,
title=record["title"],
text=record["text"],
max_title_tokens_count=self.max_title_tokens_count,
max_text_tokens_count=self.max_text_tokens_count
)
self.records.append(tensors)
def __len__(self):
return len(self.records)
def __getitem__(self, index):
return self.records[index]
def train(
train_records,
val_records,
pretrained_model_path,
train_sample_rate=1.0,
val_sample_rate=1.0,
output_model_path="models",
checkpoint=None,
max_text_tokens_count=256,
max_title_tokens_count=64,
batch_size=8,
logging_steps=1000,
eval_steps=10000,
save_steps=10000,
learning_rate=0.00003,
warmup_steps=2000,
num_train_epochs=3
):
logging.set_verbosity_info()
tokenizer = BertTokenizer.from_pretrained(
pretrained_model_path,
do_lower_case=False,
do_basic_tokenize=False,
strip_accents=False
)
train_dataset = GetTitleDataset(
train_records,
train_sample_rate,
tokenizer,
max_text_tokens_count=max_text_tokens_count,
max_title_tokens_count=max_title_tokens_count
)
val_dataset = GetTitleDataset(
val_records,
val_sample_rate,
tokenizer,
max_text_tokens_count=max_text_tokens_count,
max_title_tokens_count=max_title_tokens_count
)
model = EncoderDecoderModel.from_encoder_decoder_pretrained(pretrained_model_path, pretrained_model_path)
training_args = TrainingArguments(
output_dir=output_model_path,
per_device_train_batch_size=batch_size,
per_device_eval_batch_size=batch_size,
do_train=True,
do_eval=True,
overwrite_output_dir=False,
logging_steps=logging_steps,
eval_steps=eval_steps,
evaluation_strategy="steps",
save_steps=save_steps,
learning_rate=learning_rate,
warmup_steps=warmup_steps,
num_train_epochs=num_train_epochs,
max_steps=-1,
save_total_limit=1,
)
trainer = Trainer(
model=model,
args=training_args,
train_dataset=train_dataset,
eval_dataset=val_dataset
)
trainer.train(checkpoint)
model.save_pretrained(output_model_path)
```
|
IlyaGusev/rubertconv_toxic_clf
|
IlyaGusev
|
bert
| 8 | 302 |
transformers
| 1 |
text-classification
| true | false | false |
apache-2.0
|
['ru']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['text-classification']
| false | true | true | 1,239 |
# RuBERTConv Toxic Classifier
## Model description
Based on [rubert-base-cased-conversational](https://huggingface.co/DeepPavlov/rubert-base-cased-conversational) model
## Intended uses & limitations
#### How to use
Colab: [link](https://colab.research.google.com/drive/1veKO9hke7myxKigZtZho_F-UM2fD9kp8)
```python
from transformers import pipeline
model_name = "IlyaGusev/rubertconv_toxic_clf"
pipe = pipeline("text-classification", model=model_name, tokenizer=model_name, framework="pt")
text = "Ты придурок из интернета"
pipe([text])
```
## Training data
Datasets:
- [2ch]( https://www.kaggle.com/blackmoon/russian-language-toxic-comments)
- [Odnoklassniki](https://www.kaggle.com/alexandersemiletov/toxic-russian-comments)
- [Toloka Persona Chat Rus](https://toloka.ai/ru/datasets)
- [Koziev's Conversations](https://github.com/Koziev/NLP_Datasets/blob/master/Conversations/Data) with [toxic words vocabulary](https://www.dropbox.com/s/ou6lx03b10yhrfl/bad_vocab.txt.tar.gz)
Augmentations:
- ё -> е
- Remove or add "?" or "!"
- Fix CAPS
- Concatenate toxic and non-toxic texts
- Concatenate two non-toxic texts
- Add toxic words from vocabulary
- Add typos
- Mask toxic words with "*", "@", "$"
## Training procedure
TBA
|
IlyaGusev/rubertconv_toxic_editor
|
IlyaGusev
|
bert
| 8 | 110 |
transformers
| 4 |
token-classification
| true | false | false |
apache-2.0
|
['ru']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['token-classification']
| false | true | true | 1,615 |
# RuBERTConv Toxic Editor
## Model description
Tagging model for detoxification based on [rubert-base-cased-conversational](https://huggingface.co/DeepPavlov/rubert-base-cased-conversational).
4 possible classes:
- Equal = save tokens
- Replace = replace tokens with mask
- Delete = remove tokens
- Insert = insert mask before tokens
Use in pair with [mask filler](https://huggingface.co/IlyaGusev/sber_rut5_filler).
## Intended uses & limitations
#### How to use
Colab: [link](https://colab.research.google.com/drive/1NUSO1QGlDgD-IWXa2SpeND089eVxrCJW)
```python
import torch
from transformers import AutoTokenizer, pipeline
tagger_model_name = "IlyaGusev/rubertconv_toxic_editor"
device = "cuda" if torch.cuda.is_available() else "cpu"
device_num = 0 if device == "cuda" else -1
tagger_pipe = pipeline(
"token-classification",
model=tagger_model_name,
tokenizer=tagger_model_name,
framework="pt",
device=device_num,
aggregation_strategy="max"
)
text = "..."
tagger_predictions = tagger_pipe([text], batch_size=1)
sample_predictions = tagger_predictions[0]
print(sample_predictions)
```
## Training data
- Dataset: [russe_detox_2022](https://github.com/skoltech-nlp/russe_detox_2022/tree/main/data)
## Training procedure
- Parallel corpus convertion: [compute_tags.py](https://github.com/IlyaGusev/rudetox/blob/main/rudetox/marker/compute_tags.py)
- Training script: [train.py](https://github.com/IlyaGusev/rudetox/blob/main/rudetox/marker/train.py)
- Pipeline step: [dvc.yaml, train_marker](https://github.com/IlyaGusev/rudetox/blob/main/dvc.yaml#L367)
## Eval results
TBA
|
IlyaGusev/rugpt3medium_sum_gazeta
|
IlyaGusev
|
gpt2
| 10 | 283 |
transformers
| 2 |
summarization
| true | false | false |
['apache-2.0']
|
['ru']
|
['IlyaGusev/gazeta']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['causal-lm', 'summarization']
| false | true | true | 3,248 |
# RuGPT3MediumSumGazeta
## Model description
This is the model for abstractive summarization for Russian based on [rugpt3medium_based_on_gpt2](https://huggingface.co/sberbank-ai/rugpt3medium_based_on_gpt2).
## Intended uses & limitations
#### How to use
Colab: [link](https://colab.research.google.com/drive/1eR-ev0Y5ISWIwGnzYYoHyGMaSIUz8GTN)
```python
import torch
from transformers import AutoTokenizer, AutoModelForCausalLM
model_name = "IlyaGusev/rugpt3medium_sum_gazeta"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)
article_text = "..."
text_tokens = tokenizer(
article_text,
max_length=600,
add_special_tokens=False,
padding=False,
truncation=True
)["input_ids"]
input_ids = text_tokens + [tokenizer.sep_token_id]
input_ids = torch.LongTensor([input_ids])
output_ids = model.generate(
input_ids=input_ids,
no_repeat_ngram_size=4
)
summary = tokenizer.decode(output_ids[0], skip_special_tokens=False)
summary = summary.split(tokenizer.sep_token)[1]
summary = summary.split(tokenizer.eos_token)[0]
print(summary)
```
## Training data
- Dataset: [Gazeta](https://huggingface.co/datasets/IlyaGusev/gazeta)
## Training procedure
- Training script: [train.py](https://github.com/IlyaGusev/summarus/blob/master/external/hf_scripts/train.py)
- Config: [gpt_training_config.json](https://github.com/IlyaGusev/summarus/blob/master/external/hf_scripts/configs/gpt_training_config.json)
## Eval results
* Train dataset: **Gazeta v1 train**
* Test dataset: **Gazeta v1 test**
* Source max_length: **600**
* Target max_length: **200**
* no_repeat_ngram_size: **4**
* num_beams: **5**
| Model | R-1-f | R-2-f | R-L-f | chrF | METEOR | BLEU | Avg char length |
|:--------------------------|:------|:------|:------|:-------|:-------|:-----|:-----|
| [mbart_ru_sum_gazeta](https://huggingface.co/IlyaGusev/mbart_ru_sum_gazeta) | **32.4** | 14.3 | 28.0 | 39.7 | **26.4** | 12.1 | 371 |
| [rut5_base_sum_gazeta](https://huggingface.co/IlyaGusev/rut5_base_sum_gazeta) | 32.2 | **14.4** | **28.1** | **39.8** | 25.7 | **12.3** | 330 |
| [rugpt3medium_sum_gazeta](https://huggingface.co/IlyaGusev/rugpt3medium_sum_gazeta) | 26.2 | 7.7 | 21.7 | 33.8 | 18.2 | 4.3 | 244 |
* Train dataset: **Gazeta v1 train**
* Test dataset: **Gazeta v2 test**
* Source max_length: **600**
* Target max_length: **200**
* no_repeat_ngram_size: **4**
* num_beams: **5**
| Model | R-1-f | R-2-f | R-L-f | chrF | METEOR | BLEU | Avg char length |
|:--------------------------|:------|:------|:------|:-------|:-------|:-----|:-----|
| [mbart_ru_sum_gazeta](https://huggingface.co/IlyaGusev/mbart_ru_sum_gazeta) | **28.7** | **11.1** | 24.4 | **37.3** | **22.7** | **9.4** | 373 |
| [rut5_base_sum_gazeta](https://huggingface.co/IlyaGusev/rut5_base_sum_gazeta) | 28.6 | **11.1** | **24.5** | 37.2 | 22.0 | **9.4** | 331 |
| [rugpt3medium_sum_gazeta](https://huggingface.co/IlyaGusev/rugpt3medium_sum_gazeta) | 24.1 | 6.5 | 19.8 | 32.1 | 16.3 | 3.6 | 242 |
Evaluation script: [evaluate.py](https://github.com/IlyaGusev/summarus/blob/master/evaluate.py)
Flags: --language ru --tokenize-after --lower
|
IlyaGusev/rut5_base_headline_gen_telegram
|
IlyaGusev
|
t5
| 8 | 2,266 |
transformers
| 1 |
summarization
| true | false | false |
apache-2.0
|
['ru']
| null | null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['summarization']
| false | true | true | 1,019 |
# RuT5TelegramHeadlines
## Model description
Based on [rut5-base](https://huggingface.co/cointegrated/rut5-base) model
## Intended uses & limitations
#### How to use
```python
from transformers import AutoTokenizer, T5ForConditionalGeneration
model_name = "IlyaGusev/rut5_base_headline_gen_telegram"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = T5ForConditionalGeneration.from_pretrained(model_name)
article_text = "..."
input_ids = tokenizer(
[article_text],
max_length=600,
add_special_tokens=True,
padding="max_length",
truncation=True,
return_tensors="pt"
)["input_ids"]
output_ids = model.generate(
input_ids=input_ids
)[0]
headline = tokenizer.decode(output_ids, skip_special_tokens=True)
print(headline)
```
## Training data
- Dataset: [ru_all_split.tar.gz](https://www.dropbox.com/s/ykqk49a8avlmnaf/ru_all_split.tar.gz)
## Training procedure
- Training script: [train.py](https://github.com/IlyaGusev/summarus/blob/master/external/hf_scripts/train.py)
|
IlyaGusev/rut5_base_sum_gazeta
|
IlyaGusev
|
t5
| 8 | 900 |
transformers
| 0 |
summarization
| true | false | false |
['apache-2.0']
|
['ru']
|
['IlyaGusev/gazeta']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['summarization', 't5']
| false | true | true | 4,781 |
# RuT5SumGazeta
## Model description
This is the model for abstractive summarization for Russian based on [rut5-base](https://huggingface.co/cointegrated/rut5-base).
## Intended uses & limitations
#### How to use
Colab: [link](https://colab.research.google.com/drive/1re5E26ZIDUpAx1gOCZkbF3hcwjozmgG0)
```python
from transformers import AutoTokenizer, T5ForConditionalGeneration
model_name = "IlyaGusev/rut5_base_sum_gazeta"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = T5ForConditionalGeneration.from_pretrained(model_name)
article_text = "..."
input_ids = tokenizer(
[article_text],
max_length=600,
add_special_tokens=True,
padding="max_length",
truncation=True,
return_tensors="pt"
)["input_ids"]
output_ids = model.generate(
input_ids=input_ids,
no_repeat_ngram_size=4
)[0]
summary = tokenizer.decode(output_ids, skip_special_tokens=True)
print(summary)
```
## Training data
- Dataset: [Gazeta](https://huggingface.co/datasets/IlyaGusev/gazeta)
## Training procedure
- Training script: [train.py](https://github.com/IlyaGusev/summarus/blob/master/external/hf_scripts/train.py)
- Config: [t5_training_config.json](https://github.com/IlyaGusev/summarus/blob/master/external/hf_scripts/configs/t5_training_config.json)
## Eval results
* Train dataset: **Gazeta v1 train**
* Test dataset: **Gazeta v1 test**
* Source max_length: **600**
* Target max_length: **200**
* no_repeat_ngram_size: **4**
* num_beams: **5**
| Model | R-1-f | R-2-f | R-L-f | chrF | METEOR | BLEU | Avg char length |
|:--------------------------|:------|:------|:------|:-------|:-------|:-----|:-----|
| [mbart_ru_sum_gazeta](https://huggingface.co/IlyaGusev/mbart_ru_sum_gazeta) | **32.4** | 14.3 | 28.0 | 39.7 | **26.4** | 12.1 | 371 |
| [rut5_base_sum_gazeta](https://huggingface.co/IlyaGusev/rut5_base_sum_gazeta) | 32.2 | **14.4** | **28.1** | **39.8** | 25.7 | **12.3** | 330 |
| [rugpt3medium_sum_gazeta](https://huggingface.co/IlyaGusev/rugpt3medium_sum_gazeta) | 26.2 | 7.7 | 21.7 | 33.8 | 18.2 | 4.3 | 244 |
* Train dataset: **Gazeta v1 train**
* Test dataset: **Gazeta v2 test**
* Source max_length: **600**
* Target max_length: **200**
* no_repeat_ngram_size: **4**
* num_beams: **5**
| Model | R-1-f | R-2-f | R-L-f | chrF | METEOR | BLEU | Avg char length |
|:--------------------------|:------|:------|:------|:-------|:-------|:-----|:-----|
| [mbart_ru_sum_gazeta](https://huggingface.co/IlyaGusev/mbart_ru_sum_gazeta) | **28.7** | **11.1** | 24.4 | **37.3** | **22.7** | **9.4** | 373 |
| [rut5_base_sum_gazeta](https://huggingface.co/IlyaGusev/rut5_base_sum_gazeta) | 28.6 | **11.1** | **24.5** | 37.2 | 22.0 | **9.4** | 331 |
| [rugpt3medium_sum_gazeta](https://huggingface.co/IlyaGusev/rugpt3medium_sum_gazeta) | 24.1 | 6.5 | 19.8 | 32.1 | 16.3 | 3.6 | 242 |
Predicting all summaries:
```python
import json
import torch
from transformers import AutoTokenizer, T5ForConditionalGeneration
from datasets import load_dataset
def gen_batch(inputs, batch_size):
batch_start = 0
while batch_start < len(inputs):
yield inputs[batch_start: batch_start + batch_size]
batch_start += batch_size
def predict(
model_name,
input_records,
output_file,
max_source_tokens_count=600,
batch_size=8
):
device = "cuda" if torch.cuda.is_available() else "cpu"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = T5ForConditionalGeneration.from_pretrained(model_name).to(device)
predictions = []
for batch in gen_batch(input_records, batch_size):
texts = [r["text"] for r in batch]
input_ids = tokenizer(
texts,
add_special_tokens=True,
max_length=max_source_tokens_count,
padding="max_length",
truncation=True,
return_tensors="pt"
)["input_ids"].to(device)
output_ids = model.generate(
input_ids=input_ids,
no_repeat_ngram_size=4
)
summaries = tokenizer.batch_decode(output_ids, skip_special_tokens=True)
for s in summaries:
print(s)
predictions.extend(summaries)
with open(output_file, "w") as w:
for p in predictions:
w.write(p.strip().replace("\n", " ") + "\n")
gazeta_test = load_dataset('IlyaGusev/gazeta', script_version="v1.0")["test"]
predict("IlyaGusev/rut5_base_sum_gazeta", list(gazeta_test), "t5_predictions.txt")
```
Evaluation script: [evaluate.py](https://github.com/IlyaGusev/summarus/blob/master/evaluate.py)
Flags: --language ru --tokenize-after --lower
|
IlyaGusev/xlm_roberta_large_headline_cause_full
|
IlyaGusev
|
xlm-roberta
| 7 | 595 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
|
['ru', 'en']
|
['IlyaGusev/headline_cause']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['xlm-roberta-large']
| false | true | true | 3,037 |
# XLM-RoBERTa HeadlineCause Full
## Model description
This model was trained to predict the presence of causal relations between two headlines. This model is for the Full task with 7 possible labels: titles are almost the same, A causes B, B causes A, A refutes B, B refutes A, A linked with B in another way, A is not linked to B. English and Russian languages are supported.
You can use hosted inference API to infer a label for a headline pair. To do this, you shoud seperate headlines with ```</s>``` token.
For example:
```
Песков опроверг свой перевод на удаленку</s>Дмитрий Песков перешел на удаленку
```
## Intended uses & limitations
#### How to use
```python
from tqdm.notebook import tqdm
from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
def get_batch(data, batch_size):
start_index = 0
while start_index < len(data):
end_index = start_index + batch_size
batch = data[start_index:end_index]
yield batch
start_index = end_index
def pipe_predict(data, pipe, batch_size=64):
raw_preds = []
for batch in tqdm(get_batch(data, batch_size)):
raw_preds += pipe(batch)
return raw_preds
MODEL_NAME = TOKENIZER_NAME = "IlyaGusev/xlm_roberta_large_headline_cause_full"
tokenizer = AutoTokenizer.from_pretrained(TOKENIZER_NAME, do_lower_case=False)
model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
model.eval()
pipe = pipeline("text-classification", model=model, tokenizer=tokenizer, framework="pt", return_all_scores=True)
texts = [
(
"Judge issues order to allow indoor worship in NC churches",
"Some local churches resume indoor services after judge lifted NC governor’s restriction"
),
(
"Gov. Kevin Stitt defends $2 million purchase of malaria drug touted by Trump",
"Oklahoma spent $2 million on malaria drug touted by Trump"
),
(
"Песков опроверг свой перевод на удаленку",
"Дмитрий Песков перешел на удаленку"
)
]
pipe_predict(texts, pipe)
```
#### Limitations and bias
The models are intended to be used on news headlines. No other limitations are known.
## Training data
* HuggingFace dataset: [IlyaGusev/headline_cause](https://huggingface.co/datasets/IlyaGusev/headline_cause)
* GitHub: [IlyaGusev/HeadlineCause](https://github.com/IlyaGusev/HeadlineCause)
## Training procedure
* Notebook: [HeadlineCause](https://colab.research.google.com/drive/1NAnD0OJ0TnYCJRsHpYUyYkjr_yi8ObcA)
* Stand-alone script: [train.py](https://github.com/IlyaGusev/HeadlineCause/blob/main/headline_cause/train.py)
## Eval results
Evaluation results can be found in the [arxiv paper](https://arxiv.org/pdf/2108.12626.pdf).
### BibTeX entry and citation info
```bibtex
@misc{gusev2021headlinecause,
title={HeadlineCause: A Dataset of News Headlines for Detecting Causalities},
author={Ilya Gusev and Alexey Tikhonov},
year={2021},
eprint={2108.12626},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
IlyaGusev/xlm_roberta_large_headline_cause_simple
|
IlyaGusev
|
xlm-roberta
| 7 | 13 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
|
['ru', 'en']
|
['IlyaGusev/headline_cause']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['xlm-roberta-large']
| false | true | true | 2,954 |
# XLM-RoBERTa HeadlineCause Simple
## Model description
This model was trained to predict the presence of causal relations between two headlines. This model is for the Simple task with 3 possible labels: A causes B, B causes A, no causal relation. English and Russian languages are supported.
You can use hosted inference API to infer a label for a headline pair. To do this, you shoud seperate headlines with ```</s>``` token.
For example:
```
Песков опроверг свой перевод на удаленку</s>Дмитрий Песков перешел на удаленку
```
## Intended uses & limitations
#### How to use
```python
from tqdm.notebook import tqdm
from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
def get_batch(data, batch_size):
start_index = 0
while start_index < len(data):
end_index = start_index + batch_size
batch = data[start_index:end_index]
yield batch
start_index = end_index
def pipe_predict(data, pipe, batch_size=64):
raw_preds = []
for batch in tqdm(get_batch(data, batch_size)):
raw_preds += pipe(batch)
return raw_preds
MODEL_NAME = TOKENIZER_NAME = "IlyaGusev/xlm_roberta_large_headline_cause_simple"
tokenizer = AutoTokenizer.from_pretrained(TOKENIZER_NAME, do_lower_case=False)
model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
model.eval()
pipe = pipeline("text-classification", model=model, tokenizer=tokenizer, framework="pt", return_all_scores=True)
texts = [
(
"Judge issues order to allow indoor worship in NC churches",
"Some local churches resume indoor services after judge lifted NC governor’s restriction"
),
(
"Gov. Kevin Stitt defends $2 million purchase of malaria drug touted by Trump",
"Oklahoma spent $2 million on malaria drug touted by Trump"
),
(
"Песков опроверг свой перевод на удаленку",
"Дмитрий Песков перешел на удаленку"
)
]
pipe_predict(texts, pipe)
```
#### Limitations and bias
The models are intended to be used on news headlines. No other limitations are known.
## Training data
* HuggingFace dataset: [IlyaGusev/headline_cause](https://huggingface.co/datasets/IlyaGusev/headline_cause)
* GitHub: [IlyaGusev/HeadlineCause](https://github.com/IlyaGusev/HeadlineCause)
## Training procedure
* Notebook: [HeadlineCause](https://colab.research.google.com/drive/1NAnD0OJ0TnYCJRsHpYUyYkjr_yi8ObcA)
* Stand-alone script: [train.py](https://github.com/IlyaGusev/HeadlineCause/blob/main/headline_cause/train.py)
## Eval results
Evaluation results can be found in the [arxiv paper](https://arxiv.org/pdf/2108.12626.pdf).
### BibTeX entry and citation info
```bibtex
@misc{gusev2021headlinecause,
title={HeadlineCause: A Dataset of News Headlines for Detecting Causalities},
author={Ilya Gusev and Alexey Tikhonov},
year={2021},
eprint={2108.12626},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
Ilyes/wav2vec2-large-xlsr-53-french
|
Ilyes
|
wav2vec2
| 11 | 1,033 |
transformers
| 2 |
automatic-speech-recognition
| true | false | false |
apache-2.0
|
['fr']
|
['common_voice']
| null | 1 | 0 | 1 | 0 | 0 | 0 | 0 |
['audio', 'automatic-speech-recognition', 'speech', 'xlsr-fine-tuning-week']
| true | true | true | 1,876 |
## Evaluation on Common Voice FR Test
The script used for training and evaluation can be found here: https://github.com/irebai/wav2vec2
```python
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import (
Wav2Vec2ForCTC,
Wav2Vec2Processor,
)
import re
model_name = "Ilyes/wav2vec2-large-xlsr-53-french"
device = "cpu" # "cuda"
model = Wav2Vec2ForCTC.from_pretrained(model_name).to(device)
processor = Wav2Vec2Processor.from_pretrained(model_name)
ds = load_dataset("common_voice", "fr", split="test", cache_dir="./data/fr")
chars_to_ignore_regex = '[\,\?\.\!\;\:\"\“\%\‘\”\�\‘\’\’\’\‘\…\·\!\ǃ\?\«\‹\»\›“\”\\ʿ\ʾ\„\∞\\|\.\,\;\:\*\—\–\─\―\_\/\:\ː\;\,\=\«\»\→]'
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower().replace("’", "'")
return batch
resampler = torchaudio.transforms.Resample(48_000, 16_000)
ds = ds.map(map_to_array)
def map_to_pred(batch):
features = processor(batch["speech"], sampling_rate=batch["sampling_rate"][0], padding=True, return_tensors="pt")
input_values = features.input_values.to(device)
attention_mask = features.attention_mask.to(device)
with torch.no_grad():
logits = model(input_values, attention_mask=attention_mask).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["predicted"] = processor.batch_decode(pred_ids)
batch["target"] = batch["sentence"]
return batch
result = ds.map(map_to_pred, batched=True, batch_size=16, remove_columns=list(ds.features.keys()))
wer = load_metric("wer")
print(wer.compute(predictions=result["predicted"], references=result["target"]))
```
## Results
WER=12.82%
CER=4.40%
|
Ilyes/wav2vec2-large-xlsr-53-french_punctuation
|
Ilyes
|
wav2vec2
| 10 | 9 |
transformers
| 0 |
automatic-speech-recognition
| true | false | true |
apache-2.0
|
['fr']
|
['common_voice']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['audio', 'automatic-speech-recognition', 'speech', 'xlsr-fine-tuning']
| false | true | true | 3,278 |
## Evaluation on Common Voice FR Test
```python
import re
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import (
Wav2Vec2ForCTC,
Wav2Vec2Processor,
)
model_name = "Ilyes/wav2vec2-large-xlsr-53-french_punctuation"
model = Wav2Vec2ForCTC.from_pretrained(model_name).to('cuda')
processor = Wav2Vec2Processor.from_pretrained(model_name)
ds = load_dataset("common_voice", "fr", split="test")
chars_to_ignore_regex = '[\;\:\"\“\%\‘\”\�\‘\’\’\’\‘\…\·\ǃ\«\‹\»\›“\”\\ʿ\ʾ\„\∞\\|\;\:\*\—\–\─\―\_\/\:\ː\;\=\«\»\→]'
def normalize_text(text):
text = text.lower().strip()
text = re.sub('œ', 'oe', text)
text = re.sub('æ', 'ae', text)
text = re.sub("’|´|′|ʼ|‘|ʻ|`", "'", text)
text = re.sub("'+ ", " ", text)
text = re.sub(" '+", " ", text)
text = re.sub("'$", " ", text)
text = re.sub("' ", " ", text)
text = re.sub("−|‐", "-", text)
text = re.sub(" -", "", text)
text = re.sub("- ", "", text)
text = re.sub(chars_to_ignore_regex, '', text)
return text
def map_to_array(batch):
speech, _ = torchaudio.load(batch["path"])
batch["speech"] = resampler.forward(speech.squeeze(0)).numpy()
batch["sampling_rate"] = resampler.new_freq
batch["sentence"] = normalize_text(batch["sentence"])
return batch
ds = ds.map(map_to_array)
resampler = torchaudio.transforms.Resample(48_000, 16_000)
def map_to_pred(batch):
features = processor(batch["speech"], sampling_rate=batch["sampling_rate"][0], padding=True, return_tensors="pt")
input_values = features.input_values.to(device)
attention_mask = features.attention_mask.to(device)
with torch.no_grad():
logits = model(input_values, attention_mask=attention_mask).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["predicted"] = processor.batch_decode(pred_ids)
batch["target"] = batch["sentence"]
# remove duplicates
batch["target"] = re.sub('\.+', '.', batch["target"])
batch["target"] = re.sub('\?+', '?', batch["target"])
batch["target"] = re.sub('!+', '!', batch["target"])
batch["target"] = re.sub(',+', ',', batch["target"])
return batch
result = ds.map(map_to_pred, batched=True, batch_size=16, remove_columns=list(ds.features.keys()))
wer = load_metric("wer")
print(wer.compute(predictions=result["predicted"], references=result["target"]))
```
## Some results
| Reference | Prediction |
| ------------- | ------------- |
| il vécut à new york et y enseigna une grande partie de sa vie. | il a vécu à new york et y enseigna une grande partie de sa vie. |
| au classement par nations, l'allemagne est la tenante du titre. | au classement der nation l'allemagne est la tenante du titre. |
| voici un petit calcul pour fixer les idées. | voici un petit calcul pour fixer les idées. |
| oh! tu dois être beau avec | oh! tu dois être beau avec. |
| babochet vous le voulez? | baboche, vous le voulez? |
| la commission est, par conséquent, défavorable à cet amendement. | la commission est, par conséquent, défavorable à cet amendement. |
All the references and predictions of the test corpus are already available in this repository.
## Results
text + punctuation
WER=21.47% CER=7.21%
text (without punctuation)
WER=19.71% CER=6.91%
|
InfoCoV/Cro-CoV-cseBERT
|
InfoCoV
|
bert
| 12 | 5 |
transformers
| 0 |
fill-mask
| true | false | false | null | null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | false | true | 704 |
## Usage:
```
from sentence_transformers import models
from sentence_transformers import SentenceTransformer
word_embedding_model = models.Transformer('Cro-CoV-cseBERT')
pooling_model = models.Pooling(word_embedding_model.get_word_embedding_dimension(),
pooling_mode_mean_tokens=True,
pooling_mode_cls_token=False,
pooling_mode_max_tokens=False)
model = SentenceTransformer(modules=[word_embedding_model, pooling_model], device='') ## device = 'gpu' or 'cpu'
texts_emb = model.encode(texts)
```
## Datasets:
https://github.com/InfoCoV/InfoCoV
## Paper:
Please cite https://www.mdpi.com/2076-3417/11/21/10442
|
Intel/bert-base-uncased-mnli-sparse-70-unstructured-no-classifier
|
Intel
|
bert
| 7 | 10 |
transformers
| 0 |
fill-mask
| true | false | false | null |
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 1,063 |
# Sparse BERT base model fine tuned to MNLI without classifier layer (uncased)
Fine tuned sparse BERT base to MNLI (GLUE Benchmark) task from [bert-base-uncased-sparse-70-unstructured](https://huggingface.co/Intel/bert-base-uncased-sparse-70-unstructured).
<br>
This model doesn't have a classifier layer to enable easier loading of the model for training to other downstream tasks.
In all the other layers this model is similar to [bert-base-uncased-mnli-sparse-70-unstructured](https://huggingface.co/Intel/bert-base-uncased-mnli-sparse-70-unstructured).
<br><br>
Note: This model requires `transformers==2.10.0`
## Evaluation Results
Matched: 82.5%
Mismatched: 83.3%
This model can be further fine-tuned to other tasks and achieve the following evaluation results:
| Task | QQP (Acc/F1) | QNLI (Acc) | SST-2 (Acc) | STS-B (Pears/Spear) | SQuADv1.1 (Acc/F1) |
|------|--------------|------------|-------------|---------------------|--------------------|
| | 90.2/86.7 | 90.3 | 91.5 | 88.9/88.6 | 80.5/88.2 |
|
Intel/bert-base-uncased-mnli-sparse-70-unstructured
|
Intel
|
bert
| 7 | 10 |
transformers
| 0 |
text-classification
| true | false | false | null |
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 738 |
# Sparse BERT base model fine tuned to MNLI (uncased)
Fine tuned sparse BERT base to MNLI (GLUE Benchmark) task from [bert-base-uncased-sparse-70-unstructured](https://huggingface.co/Intel/bert-base-uncased-sparse-70-unstructured).
<br><br>
Note: This model requires `transformers==2.10.0`
## Evaluation Results
Matched: 82.5%
Mismatched: 83.3%
This model can be further fine-tuned to other tasks and achieve the following evaluation results:
| Task | QQP (Acc/F1) | QNLI (Acc) | SST-2 (Acc) | STS-B (Pears/Spear) | SQuADv1.1 (Acc/F1) |
|------|--------------|------------|-------------|---------------------|--------------------|
| | 90.2/86.7 | 90.3 | 91.5 | 88.9/88.6 | 80.5/88.2 |
|
Intel/bert-base-uncased-sparse-1_2
|
Intel
|
bert
| 8 | 8 |
transformers
| 0 | null | true | false | false | null |
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 929 |
# Sparse BERT base model (uncased)
Pretrained model pruned to 1:2 structured sparsity.
The model is a pruned version of the [BERT base model](https://huggingface.co/bert-base-uncased).
## Intended Use
The model can be used for fine-tuning to downstream tasks with sparsity already embeded to the model.
To keep the sparsity a mask should be added to each sparse weight blocking the optimizer from updating the zeros.
## Evaluation Results
We get the following results on the tasks development set, all results are mean of 5 different seeded models:
| Task | MNLI-m (Acc) | MNLI-mm (Acc) | QQP (Acc/F1) | QNLI (Acc) | SST-2 (Acc) | STS-B (Pears/Spear) | SQuADv1.1 (Acc/F1) |
|------|--------------|---------------|--------------|------------|-------------|---------------------|--------------------|
| | 83.3 | 83.9 | 90.8/87.6 | 90.4 | 91.3 | 88.8/88.3 | 80.5/88.2 |
|
Intel/bert-base-uncased-sparse-70-unstructured
|
Intel
|
bert
| 7 | 20 |
transformers
| 0 |
fill-mask
| true | false | false | null |
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 409 |
# Sparse BERT base model (uncased)
Pretrained model pruned to 70% sparsity.
The model is a pruned version of the [BERT base model](https://huggingface.co/bert-base-uncased).
## Intended Use
The model can be used for fine-tuning to downstream tasks with sparsity already embeded to the model.
To keep the sparsity a mask should be added to each sparse weight blocking the optimizer from updating the zeros.
|
Intel/bert-base-uncased-sparse-85-unstructured-pruneofa
|
Intel
|
bert
| 9 | 22 |
transformers
| 0 |
fill-mask
| true | true | false |
apache-2.0
|
['en']
|
['wikipedia', 'bookcorpus']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['fill-mask']
| false | true | true | 422 |
# 85% Sparse BERT-Large (uncased) Prune OFA
This model is a result from our paper [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754) presented in ENLSP NeurIPS Workshop 2021.
For further details on the model and its result, see our paper and our implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/bert-base-uncased-sparse-90-unstructured-pruneofa
|
Intel
|
bert
| 9 | 397 |
transformers
| 0 |
fill-mask
| true | true | false |
apache-2.0
|
['en']
|
['wikipedia', 'bookcorpus']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['fill-mask']
| false | true | true | 421 |
# 90% Sparse BERT-Base (uncased) Prune OFA
This model is a result from our paper [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754) presented in ENLSP NeurIPS Workshop 2021.
For further details on the model and its result, see our paper and our implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/bert-base-uncased-squadv1.1-sparse-80-1x4-block-pruneofa
|
Intel
|
bert
| 8 | 23 |
transformers
| 0 |
question-answering
| true | false | false |
apache-2.0
|
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 598 |
# 80% 1x4 Block Sparse BERT-Base (uncased) Fine Tuned on SQuADv1.1
This model is a result of fine-tuning a Prune OFA 80% 1x4 block sparse pre-trained BERT-Base combined with knowledge distillation.
This model yields the following results on SQuADv1.1 development set:<br>
`{"exact_match": 81.2867, "f1": 88.4735}`
For further details see our paper, [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754), and our open source implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/bert-large-uncased-sparse-90-unstructured-pruneofa
|
Intel
|
bert
| 9 | 8 |
transformers
| 0 |
fill-mask
| true | true | false |
apache-2.0
|
['en']
|
['wikipedia', 'bookcorpus']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['fill-mask']
| false | true | true | 422 |
# 90% Sparse BERT-Large (uncased) Prune OFA
This model is a result from our paper [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754) presented in ENLSP NeurIPS Workshop 2021.
For further details on the model and its result, see our paper and our implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/bert-large-uncased-squadv1.1-sparse-90-unstructured
|
Intel
|
bert
| 9 | 6 |
transformers
| 0 |
question-answering
| true | true | false | null |
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 600 |
# 90% Sparse BERT-Large (uncased) Fine Tuned on SQuADv1.1
This model is a result of fine-tuning a Prune OFA 90% sparse pre-trained BERT-Large combined with knowledge distillation.
This model yields the following results on SQuADv1.1 development set:<br>
`{"exact_match": 83.56669820245979, "f1": 90.20829352733487}`
For further details see our paper, [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754), and our open source implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/distilbert-base-uncased-sparse-85-unstructured-pruneofa
|
Intel
|
distilbert
| 9 | 8 |
transformers
| 0 |
fill-mask
| true | true | false |
apache-2.0
|
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 427 |
# 85% Sparse DistilBERT-Base (uncased) Prune OFA
This model is a result from our paper [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754) presented in ENLSP NeurIPS Workshop 2021.
For further details on the model and its result, see our paper and our implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/distilbert-base-uncased-sparse-90-unstructured-pruneofa
|
Intel
|
distilbert
| 9 | 30 |
transformers
| 2 |
fill-mask
| true | true | false |
apache-2.0
|
['en']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| false | true | true | 427 |
# 90% Sparse DistilBERT-Base (uncased) Prune OFA
This model is a result from our paper [Prune Once for All: Sparse Pre-Trained Language Models](https://arxiv.org/abs/2111.05754) presented in ENLSP NeurIPS Workshop 2021.
For further details on the model and its result, see our paper and our implementation available [here](https://github.com/IntelLabs/Model-Compression-Research-Package/tree/main/research/prune-once-for-all).
|
Intel/dynamic_tinybert
|
Intel
|
bert
| 9 | 1,330 |
transformers
| 4 |
question-answering
| true | false | false | null | null | null | null | 1 | 0 | 1 | 0 | 0 | 0 | 0 |
['question-answering', 'bert']
| false | true | true | 5,375 |
# Model Card for dynamic_tinybert
# Model Details
## Model Description
Dynamic-TinyBERT: Boost TinyBERT’s Inference Efficiency by Dynamic Sequence Length
- **Developed by:** Intel
- **Shared by [Optional]:** Intel
- **Model type:** Question Answering
- **Language(s) (NLP):** More information needed
- **License:** More information needed
- **Parent Model:** BERT
- **Resources for more information:**
- [Associated Paper](https://neurips2021-nlp.github.io/papers/16/CameraReady/Dynamic_TinyBERT_NLSP2021_camera_ready.pdf)
# Uses
## Direct Use
This model can be used for the task of question answering.
## Downstream Use [Optional]
More information needed.
## Out-of-Scope Use
The model should not be used to intentionally create hostile or alienating environments for people.
# Bias, Risks, and Limitations
Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups.
## Recommendations
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
# Training Details
## Training Data
The model authors note in the [associated paper](https://neurips2021-nlp.github.io/papers/16/CameraReady/Dynamic_TinyBERT_NLSP2021_camera_ready.pdf):
> All our experiments are evaluated on the challenging question-answering benchmark SQuAD1.1 [11].
## Training Procedure
### Preprocessing
The model authors note in the [associated paper](https://neurips2021-nlp.github.io/papers/16/CameraReady/Dynamic_TinyBERT_NLSP2021_camera_ready.pdf):
> We start with a pre-trained general-TinyBERT student, which was trained to learn the general knowledge of BERT using the general-distillation method presented by TinyBERT. We perform transformer distillation from a fine- tuned BERT teacher to the student, following the same training steps used in the original TinyBERT: (1) **intermediate-layer distillation (ID)** — learning the knowledge residing in the hidden states and attentions matrices, and (2) **prediction-layer distillation (PD)** — fitting the predictions of the teacher.
### Speeds, Sizes, Times
The model authors note in the [associated paper](https://neurips2021-nlp.github.io/papers/16/CameraReady/Dynamic_TinyBERT_NLSP2021_camera_ready.pdf):
>For our Dynamic-TinyBERT model we use the architecture of TinyBERT6L: a small BERT model with 6 layers, a hidden size of 768, a feed forward size of 3072 and 12 heads.
# Evaluation
## Testing Data, Factors & Metrics
### Testing Data
More information needed
### Factors
More information needed
### Metrics
More information needed
## Results
The model authors note in the [associated paper](https://neurips2021-nlp.github.io/papers/16/CameraReady/Dynamic_TinyBERT_NLSP2021_camera_ready.pdf):
| Model | Max F1 (full model) | Best Speedup within BERT-1% |
|------------------|---------------------|-----------------------------|
| Dynamic-TinyBERT | 88.71 | 3.3x |
# Model Examination
More information needed
# Environmental Impact
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
- **Hardware Type:** Titan GPU
- **Hours used:** More information needed
- **Cloud Provider:** More information needed
- **Compute Region:** More information needed
- **Carbon Emitted:** More information needed
# Technical Specifications [optional]
## Model Architecture and Objective
More information needed
## Compute Infrastructure
More information needed
### Hardware
More information needed
### Software
More information needed.
# Citation
**BibTeX:**
```bibtex
@misc{https://doi.org/10.48550/arxiv.2111.09645,
doi = {10.48550/ARXIV.2111.09645},
url = {https://arxiv.org/abs/2111.09645},
author = {Guskin, Shira and Wasserblat, Moshe and Ding, Ke and Kim, Gyuwan},
keywords = {Computation and Language (cs.CL), Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences},
title = {Dynamic-TinyBERT: Boost TinyBERT's Inference Efficiency by Dynamic Sequence Length},
publisher = {arXiv},
year = {2021},
```
**APA:**
More information needed
# Glossary [optional]
More information needed
# More Information [optional]
More information needed
# Model Card Authors [optional]
Intel in collaboration with Ezi Ozoani and the Hugging Face team
# Model Card Contact
More information needed
# How to Get Started with the Model
Use the code below to get started with the model.
<details>
<summary> Click to expand </summary>
```python
from transformers import AutoTokenizer, AutoModelForQuestionAnswering
tokenizer = AutoTokenizer.from_pretrained("Intel/dynamic_tinybert")
model = AutoModelForQuestionAnswering.from_pretrained("Intel/dynamic_tinybert")
```
</details>
|
IsaacBot/bert-base-uncased-finetuned-GP-Sentiment
|
IsaacBot
|
bert
| 13 | 3 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
| null | null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 1,939 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-base-uncased-finetuned-GP-Sentiment
This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on an unknown dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7815
- F1: 0.6808
- Accuracy: 0.7390
## 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: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:------:|:--------:|
| No log | 1.0 | 313 | 0.7492 | 0.6448 | 0.6845 |
| 0.7419 | 2.0 | 626 | 0.7281 | 0.6800 | 0.7350 |
| 0.7419 | 3.0 | 939 | 0.7815 | 0.6808 | 0.7390 |
| 0.5309 | 4.0 | 1252 | 0.8782 | 0.6799 | 0.7422 |
| 0.336 | 5.0 | 1565 | 1.1222 | 0.6792 | 0.7390 |
| 0.336 | 6.0 | 1878 | 1.1544 | 0.6671 | 0.7174 |
| 0.219 | 7.0 | 2191 | 1.3721 | 0.6627 | 0.7246 |
| 0.1541 | 8.0 | 2504 | 1.4864 | 0.6652 | 0.7326 |
| 0.1541 | 9.0 | 2817 | 1.6475 | 0.6660 | 0.7446 |
| 0.1094 | 10.0 | 3130 | 1.6749 | 0.6700 | 0.7446 |
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.11.0
|
IsabellaKarabasz/roberta-base-bne-finetuned-amazon_reviews_multi
|
IsabellaKarabasz
|
roberta
| 11 | 3 |
transformers
| 0 |
text-classification
| true | false | false |
cc-by-4.0
| null |
['amazon_reviews_multi']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| false | true | true | 954 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# roberta-base-bne-finetuned-amazon_reviews_multi
This model is a fine-tuned version of [BSC-TeMU/roberta-base-bne](https://huggingface.co/BSC-TeMU/roberta-base-bne) on the amazon_reviews_multi dataset.
## 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: 2
### Framework versions
- Transformers 4.9.2
- Pytorch 1.9.0+cu102
- Datasets 1.11.0
- Tokenizers 0.10.3
|
Iskaj/hf-challenge-test
|
Iskaj
|
wav2vec2
| 20 | 12 |
transformers
| 0 |
automatic-speech-recognition
| true | false | false | null |
['ab']
|
['common_voice']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['automatic-speech-recognition', 'mozilla-foundation/common_voice_7_0', 'generated_from_trainer']
| true | true | true | 1,082 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
#
This model is a fine-tuned version of [hf-test/xls-r-dummy](https://huggingface.co/hf-test/xls-r-dummy) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - AB dataset.
It achieves the following results on the evaluation set:
- Loss: 156.8789
- Wer: 1.3456
## 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: 0.0003
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 10
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.1+cu102
- Datasets 1.18.1.dev0
- Tokenizers 0.11.0
|
Iskaj/newnew
|
Iskaj
|
wav2vec2
| 11 | 7 |
transformers
| 0 |
automatic-speech-recognition
| true | false | false |
apache-2.0
|
['nl']
|
['common_voice']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['automatic-speech-recognition', 'mozilla-foundation/common_voice_8_0', 'generated_from_trainer']
| true | true | true | 1,195 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# newnew
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - NL dataset.
It achieves the following results on the evaluation set:
- Loss: 11.4375
- Wer: 1.0
## 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
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 4000
- num_epochs: 50.0
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.17.0.dev0
- Pytorch 1.10.2+cu102
- Datasets 1.18.3.dev0
- Tokenizers 0.11.0
|
Iskaj/xlsr300m_cv_8.0_nl
|
Iskaj
|
wav2vec2
| 14 | 11 |
transformers
| 0 |
automatic-speech-recognition
| true | false | false |
apache-2.0
|
['nl']
|
['mozilla-foundation/common_voice_8_0']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['automatic-speech-recognition', 'mozilla-foundation/common_voice_8_0', 'mozilla-foundation/common_voice_7_0', 'nl', 'robust-speech-event', 'model_for_talk', 'hf-asr-leaderboard']
| true | true | true | 1,381 |
# xlsr300m_cv_8.0_nl
#### Evaluation Commands
1. To evaluate on `mozilla-foundation/common_voice_8_0` with split `test`
```bash
python eval.py --model_id Iskaj/xlsr300m_cv_8.0_nl --dataset mozilla-foundation/common_voice_8_0 --config nl --split test
```
2. To evaluate on `speech-recognition-community-v2/dev_data`
```bash
python eval.py --model_id Iskaj/xlsr300m_cv_8.0_nl --dataset speech-recognition-community-v2/dev_data --config nl --split validation --chunk_length_s 5.0 --stride_length_s 1.0
```
### Inference
```python
import torch
from datasets import load_dataset
from transformers import AutoModelForCTC, AutoProcessor
import torchaudio.functional as F
model_id = "Iskaj/xlsr300m_cv_8.0_nl"
sample_iter = iter(load_dataset("mozilla-foundation/common_voice_8_0", "nl", split="test", streaming=True, use_auth_token=True))
sample = next(sample_iter)
resampled_audio = F.resample(torch.tensor(sample["audio"]["array"]), 48_000, 16_000).numpy()
model = AutoModelForCTC.from_pretrained(model_id)
processor = AutoProcessor.from_pretrained(model_id)
inputs = processor(resampled_audio, sampling_rate=16_000, return_tensors="pt")
with torch.no_grad():
logits = model(**inputs).logits
predicted_ids = torch.argmax(logits, dim=-1)
transcription = processor.batch_decode(predicted_ids)
transcription[0].lower()
#'het kontine schip lag aangemeert in de aven'
```
|
JAlexis/Bertv1_fine
|
JAlexis
|
bert
| 8 | 7 |
transformers
| 0 |
question-answering
| true | false | false | null |
['en']
|
['squad2', 'cord19']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['pytorch', 'question-answering']
| false | true | true | 1,446 |
## Model description
This model was obtained by fine-tuning deepset/bert-base-cased-squad2 on Cord19 Dataset.
## How to use
```python
from transformers.pipelines import pipeline
model_name = "JAlexis/PruebaBert"
nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
inputs = {
'question': 'How can I protect myself against covid-19?',
'context': 'Preventative measures consist of recommendations to wear a mask in public, maintain social distancing of at least six feet, wash hands regularly, and use hand sanitizer. To facilitate this aim, we adapt the conceptual model and measures of Liao et al. [6] to the current context of the COVID-19 pandemic and the culture of the USA. Applying this model in a different time and context provides an opportunity to make comparisons of reactions to information sources across a decade of evolving attitudes toward media and government, between two cultures (Hong Kong vs. the USA), and between two considerably different global pandemics (H1N1 vs. COVID-19). ',
'question': 'How can I protect myself against covid-19?',
'context': ' ',
}
nlp(inputs)
```
## Overview
```
Language model: deepset/bert-base-cased-squad2
Language: English
Downstream-task: Q&A
Datasets: CORD-19 from 31rd January 2022
Code: Haystack and FARM
Infrastructure: Tesla T4
```
## Hyperparameters
```
batch_size = 8
n_epochs = 7
max_seq_len = max_length
learning_rate = AdamW: 2e-5
```
|
JAlexis/PruebaBert
|
JAlexis
|
bert
| 8 | 7 |
transformers
| 0 |
question-answering
| true | false | false | null |
['en']
|
['squad2', 'cord19']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['pytorch', 'question-answering']
| false | true | true | 1,446 |
## Model description
This model was obtained by fine-tuning deepset/bert-base-cased-squad2 on Cord19 Dataset.
## How to use
```python
from transformers.pipelines import pipeline
model_name = "JAlexis/PruebaBert"
nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
inputs = {
'question': 'How can I protect myself against covid-19?',
'context': 'Preventative measures consist of recommendations to wear a mask in public, maintain social distancing of at least six feet, wash hands regularly, and use hand sanitizer. To facilitate this aim, we adapt the conceptual model and measures of Liao et al. [6] to the current context of the COVID-19 pandemic and the culture of the USA. Applying this model in a different time and context provides an opportunity to make comparisons of reactions to information sources across a decade of evolving attitudes toward media and government, between two cultures (Hong Kong vs. the USA), and between two considerably different global pandemics (H1N1 vs. COVID-19). ',
'question': 'How can I protect myself against covid-19?',
'context': ' ',
}
nlp(inputs)
```
## Overview
```
Language model: deepset/bert-base-cased-squad2
Language: English
Downstream-task: Q&A
Datasets: CORD-19 from 31rd January 2022
Code: Haystack and FARM
Infrastructure: Tesla T4
```
## Hyperparameters
```
batch_size = 8
n_epochs = 9
max_seq_len = max_length
learning_rate = AdamW: 1e-5
```
|
JBNLRY/distilbert-base-uncased-finetuned-cola
|
JBNLRY
|
distilbert
| 13 | 3 |
transformers
| 0 |
text-classification
| true | false | false |
apache-2.0
| null |
['glue']
| null | 1 | 1 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 1,572 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# distilbert-base-uncased-finetuned-cola
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the glue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.8366
- Matthews Correlation: 0.5472
## 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: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Matthews Correlation |
|:-------------:|:-----:|:----:|:---------------:|:--------------------:|
| 0.5224 | 1.0 | 535 | 0.5432 | 0.4243 |
| 0.3447 | 2.0 | 1070 | 0.4968 | 0.5187 |
| 0.2347 | 3.0 | 1605 | 0.6540 | 0.5280 |
| 0.1747 | 4.0 | 2140 | 0.7547 | 0.5367 |
| 0.1255 | 5.0 | 2675 | 0.8366 | 0.5472 |
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.11.0
|
JDBN/t5-base-fr-qg-fquad
|
JDBN
|
t5
| 11 | 247 |
transformers
| 3 |
text2text-generation
| true | false | true | null |
['fr']
|
['fquad', 'piaf']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['pytorch', 't5', 'question-generation', 'seq2seq']
| false | true | true | 3,256 |
# T5 Question Generation and Question Answering
## Model description
This model is a T5 Transformers model (airklizz/t5-base-multi-fr-wiki-news) that was fine-tuned in french on 3 different tasks
* question generation
* question answering
* answer extraction
It obtains quite good results on FQuAD validation dataset.
## Intended uses & limitations
This model functions for the 3 tasks mentionned earlier and was not tested on other tasks.
```python
from transformers import T5ForConditionalGeneration, T5Tokenizer
model = T5ForConditionalGeneration.from_pretrained("JDBN/t5-base-fr-qg-fquad")
tokenizer = T5Tokenizer.from_pretrained("JDBN/t5-base-fr-qg-fquad")
```
## Training data
The initial model used was https://huggingface.co/airKlizz/t5-base-multi-fr-wiki-news. This model was finetuned on a dataset composed of FQuAD and PIAF on the 3 tasks mentioned previously.
The data were preprocessed like this
* question generation: "generate question: Barack Hussein Obama, né le 4 aout 1961, est un homme politique américain et avocat. Il a été élu <hl> en 2009 <hl> pour devenir le 44ème président des Etats-Unis d'Amérique."
* question answering: "question: Quand Barack Hussein Obamaa-t-il été élu président des Etats-Unis d’Amérique? context: Barack Hussein Obama, né le 4 aout 1961, est un homme politique américain et avocat. Il a été élu en 2009 pour devenir le 44ème président des Etats-Unis d’Amérique."
* answer extraction: "extract_answers: Barack Hussein Obama, né le 4 aout 1961, est un homme politique américain et avocat. <hl> Il a été élu en 2009 pour devenir le 44ème président des Etats-Unis d’Amérique <hl>."
The preprocessing we used was implemented in https://github.com/patil-suraj/question_generation
## Eval results
#### On FQuAD validation set
| BLEU_1 | BLEU_2 | BLEU_3 | BLEU_4 | METEOR | ROUGE_L | CIDEr |
|--------|--------|--------|--------|--------|---------|-------|
| 0.290 | 0.203 | 0.149 | 0.111 | 0.197 | 0.284 | 1.038 |
#### Question Answering metrics
For these metrics, the performance of this question answering model (https://huggingface.co/illuin/camembert-base-fquad) on FQuAD original question and on T5 generated questions are compared.
| Questions | Exact Match | F1 Score |
|------------------|--------|--------|
|Original FQuAD | 54.015 | 77.466 |
|Generated | 45.765 | 67.306 |
### BibTeX entry and citation info
```bibtex
@misc{githubPatil,
author = {Patil Suraj},
title = {question generation GitHub repository},
year = {2020},
howpublished={\url{https://github.com/patil-suraj/question_generation}}
}
@article{T5,
title={Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer},
author={Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu},
year={2019},
eprint={1910.10683},
archivePrefix={arXiv},
primaryClass={cs.LG}
}
@misc{dhoffschmidt2020fquad,
title={FQuAD: French Question Answering Dataset},
author={Martin d'Hoffschmidt and Wacim Belblidia and Tom Brendlé and Quentin Heinrich and Maxime Vidal},
year={2020},
eprint={2002.06071},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
```
|
JIWON/bert-base-finetuned-nli
|
JIWON
|
bert
| 10 | 3 |
transformers
| 0 |
text-classification
| true | false | false | null | null |
['klue']
| null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 1,444 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# bert-base-finetuned-nli
This model is a fine-tuned version of [klue/bert-base](https://huggingface.co/klue/bert-base) on the klue dataset.
It achieves the following results on the evaluation set:
- Loss: 0.6210
- Accuracy: 0.085
## 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: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 5
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:--------:|
| No log | 1.0 | 196 | 0.6210 | 0.085 |
| No log | 2.0 | 392 | 0.5421 | 0.0643 |
| 0.5048 | 3.0 | 588 | 0.5523 | 0.062 |
| 0.5048 | 4.0 | 784 | 0.5769 | 0.0533 |
| 0.5048 | 5.0 | 980 | 0.5959 | 0.052 |
### Framework versions
- Transformers 4.16.2
- Pytorch 1.10.0+cu111
- Datasets 1.18.3
- Tokenizers 0.11.0
|
Jacobo/aristoBERTo
|
Jacobo
|
bert
| 10 | 10 |
transformers
| 1 |
fill-mask
| true | false | false | null |
['grc']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
[]
| true | true | true | 1,841 |
# aristoBERTo
aristoBERTo is a transformer model for ancient Greek, a low resource language. We initialized the pre-training with weights from [GreekBERT](https://huggingface.co/nlpaueb/bert-base-greek-uncased-v1), a Greek version of BERT which was trained on a large corpus of modern Greek (~ 30 GB of texts). We continued the pre-training with an ancient Greek corpus of about 900 MB, which was scrapped from the web and post-processed. Duplicate texts and editorial punctuation were removed.
Applied to the processing of ancient Greek, aristoBERTo outperforms xlm-roberta-base and mdeberta in most downstream tasks like the labeling of POS, MORPH, DEP and LEMMA.
aristoBERTo is provided by the [Diogenet project](https://diogenet.ucsd.edu) of the University of California, San Diego.
## Intended uses
This model was created for fine-tuning with spaCy and the ancient Greek Universal Dependency datasets as well as a NER corpus produced by the [Diogenet project](https://diogenet.ucsd.edu). As a fill-mask model, AristoBERTo can also be used in the restoration of damaged Greek papyri, inscriptions, and manuscripts.
It achieves the following results on the evaluation set:
- Loss: 1.6323
## Training procedure
### Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-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: 20.0
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:-------:|:---------------:|
| 1.377 | 20.0 | 3414220 | 1.6314 |
### Framework versions
- Transformers 4.14.0.dev0
- Pytorch 1.10.0+cu102
- Datasets 1.16.1
- Tokenizers 0.10.3
|
Jacobo/axiothea
|
Jacobo
|
roberta
| 10 | 5 |
transformers
| 0 |
fill-mask
| true | false | false | null |
['grc']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['generated_from_trainer']
| true | true | true | 2,046 |
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# axiothea
This is an experimental roberta model trained with an ancient Greek corpus of about 900 MB, which was scrapped from the web and post-processed. Duplicate texts and editorial punctuation were removed. The training dataset will be soon available in the Huggingface datasets hub. Training a model of ancient Greek is challenging given that it is a low resource language from which 50% of the register has only survived in fragmentary texts. The model is provided by the Diogenet project at the University of California, San Diego.
It achieves the following results on the evaluation set:
- Loss: 3.3351
## 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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:-------:|:---------------:|
| 4.7013 | 1.0 | 341422 | 4.8813 |
| 4.2866 | 2.0 | 682844 | 4.4422 |
| 4.0496 | 3.0 | 1024266 | 4.2132 |
| 3.8503 | 4.0 | 1365688 | 4.0246 |
| 3.6917 | 5.0 | 1707110 | 3.8756 |
| 3.4917 | 6.0 | 2048532 | 3.7381 |
| 3.3907 | 7.0 | 2389954 | 3.6107 |
| 3.2876 | 8.0 | 2731376 | 3.5044 |
| 3.1994 | 9.0 | 3072798 | 3.3980 |
| 3.0806 | 10.0 | 3414220 | 3.3095 |
### Framework versions
- Transformers 4.13.0.dev0
- Pytorch 1.10.0+cu102
- Datasets 1.14.0
- Tokenizers 0.10.3
|
Jacobo/grc_ud_perseus_dioberto
|
Jacobo
| null | 23 | 2 |
spacy
| 0 |
token-classification
| false | false | false | null |
['grc']
| null | null | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
['spacy', 'token-classification']
| false | true | true | 56,921 |
| Feature | Description |
| --- | --- |
| **Name** | `grc_ud_perseus_dioberto` |
| **Version** | `3.2.0` |
| **spaCy** | `>=3.2.0,<3.3.0` |
| **Default Pipeline** | `transformer`, `morphologizer`, `lemmatizer`, `tagger`, `parser`, `senter` |
| **Components** | `transformer`, `morphologizer`, `lemmatizer`, `tagger`, `parser`, `senter` |
| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
| **Sources** | n/a |
| **License** | n/a |
| **Author** | [n/a]() |
### Label Scheme
<details>
<summary>View label scheme (822 labels for 4 components)</summary>
| Component | Labels |
| --- | --- |
| **`morphologizer`** | `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `POS=ADV`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=NOUN`, `POS=VERB\|Tense=Past\|VerbForm=Inf\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=NOUN`, `POS=PUNCT`, `POS=CCONJ`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=NOUN`, `POS=ADP`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=DET`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=NOUN`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=NOUN`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=NOUN`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=NOUN`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=NOUN`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=NOUN`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=NOUN`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=NOUN`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=DET`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=NOUN`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=PRON\|Person=1`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=DET`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=DET`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `POS=SCONJ`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=DET`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=DET`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=PRON`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=PRON`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=DET`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=PRON`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=DET`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=NOUN`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=PRON`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=PRON`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=NOUN`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Neut\|Number=Sing\|POS=NOUN`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=NOUN`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=NOUN`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=1`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=NOUN`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=PRON`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=DET`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=PRON`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=NOUN`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=PRON`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=DET`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=DET`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=NOUN`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Masc\|Number=Sing\|POS=NOUN`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=DET`, `POS=VERB\|Tense=Pres\|VerbForm=Inf\|Voice=Act`, `POS=VERB\|Tense=Pres\|VerbForm=Inf\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=PRON`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=1`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=NOUN`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=DET`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=PRON`, `POS=VERB\|Tense=Past\|VerbForm=Inf\|Voice=Pass`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=NOUN`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Neut\|Number=Sing\|POS=DET`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=DET`, `POS=VERB\|Tense=Past\|VerbForm=Inf\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=DET`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=PRON`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=DET`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=DET`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=PRON`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=DET`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=DET`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=NOUN`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=NOUN`, `Case=Dat\|Gender=Neut\|Number=Sing\|POS=ADJ`, `POS=ADJ`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|POS=VERB\|Tense=Past\|VerbForm=Inf\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=2`, `POS=INTJ`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=DET`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=DET`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=1`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=2`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Degree=Cmp\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=3`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Degree=Cmp\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Voc\|Gender=Fem\|Number=Sing\|POS=NOUN`, `Case=Voc\|Gender=Masc\|Number=Plur\|POS=NOUN`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=PRON`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=PRON`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=PRON`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|POS=VERB\|Tense=Past\|VerbForm=Inf\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=PRON`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=PRON`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Degree=Cmp\|POS=ADV`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Gender=Neut\|Number=Sing\|POS=PRON`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=1`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=PRON`, `Case=Acc\|Degree=Cmp\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=DET`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Act`, `POS=NUM`, `Gender=Masc\|Number=Plur\|POS=NUM`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=PRON`, `Aspect=Perf\|Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Degree=Sup\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=DET`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Degree=Sup\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Mid`, `POS=X`, `Case=Acc\|Degree=Sup\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Case=Acc\|Degree=Cmp\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=2`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Number=Plur\|POS=PRON\|Person=2`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=2`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Acc\|Degree=Cmp\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `POS=PART`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=PRON`, `Degree=Sup\|POS=ADV`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=PRON`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Voc\|Gender=Neut\|Number=Sing\|POS=NOUN`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pqp\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Degree=Sup\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Degree=Sup\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `POS=VERB\|Tense=Fut\|VerbForm=Inf\|Voice=Act`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=PRON`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Degree=Cmp\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=PRON`, `POS=DET`, `Case=Voc\|Gender=Fem\|Number=Plur\|POS=NOUN`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Number=Sing\|POS=PRON`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part`, `Case=Nom\|Number=Sing\|POS=NOUN`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Degree=Cmp\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Number=Sing\|POS=PRON`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Gender=Masc\|Number=Sing\|POS=NOUN`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pqp\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Pass`, `Aspect=Perf\|Mood=Sub\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Degree=Cmp\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Pass`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Gender=Masc\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `POS=PRON`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Degree=Sup\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Voc\|Degree=Sup\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin`, `Case=Gen\|Degree=Sup\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Aspect=Perf\|Mood=Imp\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `POS=VERB\|Tense=Fut\|VerbForm=Inf\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Voc\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Dual\|POS=NOUN`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Gen\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Case=Dat\|Gender=Fem\|Number=Dual\|POS=PRON`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=NOUN`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=DET`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=ADJ`, `Case=Dat\|Gender=Fem\|Number=Dual\|POS=ADJ`, `Case=Dat\|Gender=Fem\|Number=Dual\|POS=NOUN`, `Case=Nom\|Gender=Fem\|Number=Dual\|POS=ADJ`, `Case=Nom\|Gender=Fem\|Number=Dual\|POS=PRON`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Fem\|Number=Dual\|POS=NOUN`, `Case=Nom\|Gender=Fem\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Dual\|POS=DET`, `Case=Gen\|Gender=Masc\|Number=Dual\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=PRON`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Masc\|Number=Dual\|POS=ADJ`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Degree=Sup\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Dat\|Degree=Cmp\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Aspect=Perf\|Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|Voice=Act`, `Aspect=Perf\|Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=X`, `Case=Nom\|Degree=Cmp\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pqp\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Case=Nom\|Degree=Cmp\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Degree=Cmp\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Acc\|Gender=Fem\|Number=Dual\|POS=DET`, `Case=Acc\|Gender=Fem\|Number=Dual\|POS=NOUN`, `Case=Acc\|Gender=Fem\|Number=Dual\|POS=PRON`, `Case=Acc\|Gender=Fem\|Number=Dual\|POS=ADJ`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Degree=Sup\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Voc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Number=Sing\|POS=PRON`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Voc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Number=Sing\|POS=NOUN`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Pass`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Dat\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Mood=Imp\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Degree=Cmp\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `POS=VERB\|Tense=Past\|VerbForm=Inf`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Degree=Sup\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Nom\|Degree=Sup\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Gen\|Gender=Fem\|Number=Dual\|POS=NOUN`, `Case=Nom\|Number=Sing\|POS=PRON`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Degree=Cmp\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Number=Dual\|POS=PRON`, `Aspect=Perf\|Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Neut\|Number=Plur\|POS=NOUN`, `Aspect=Perf\|Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Aspect=Perf\|Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Degree=Cmp\|Number=Sing\|POS=ADJ`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=X`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Case=Gen\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Neut\|Number=Dual\|POS=ADJ`, `Aspect=Imp\|Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Degree=Sup\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Dat\|Number=Plur\|POS=PRON`, `Aspect=Perf\|Mood=Opt\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Fem\|Number=Dual\|POS=PRON`, `Case=Gen\|Gender=Fem\|Number=Dual\|POS=DET`, `Case=Gen\|Gender=Fem\|Number=Dual\|POS=ADJ`, `Case=Gen\|Gender=Fem\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin`, `Case=Gen\|Degree=Cmp\|Number=Sing\|POS=ADJ`, `Mood=Imp\|POS=VERB\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Imp\|Number=Dual\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Voc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Degree=Cmp\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Dat\|Degree=Sup\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Dat\|Gender=Masc\|Number=Dual\|POS=NOUN`, `Aspect=Perf\|Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Degree=Cmp\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Degree=Cmp\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Aspect=Perf\|Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Voc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Voc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=PRON`, `Case=Nom\|Gender=Neut\|Number=Dual\|POS=NOUN`, `Case=Dat\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Degree=Sup\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Aspect=Perf\|Case=Voc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Number=Plur\|POS=NOUN`, `Case=Nom\|Gender=Neut\|Number=Dual\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Degree=Cmp\|Gender=Masc\|Number=Dual\|POS=ADJ`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Degree=Cmp\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Imp\|Number=Dual\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Dual\|POS=PRON`, `Aspect=Perf\|Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Number=Sing\|POS=ADJ`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Number=Dual\|POS=PRON`, `Aspect=Perf\|Case=Dat\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Fem\|POS=NOUN`, `Case=Acc\|Gender=Neut\|POS=NOUN`, `Aspect=Perf\|POS=VERB\|Tense=Past\|VerbForm=Inf\|Voice=Pass`, `Case=Dat\|Degree=Cmp\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Pass`, `Case=Dat\|Number=Plur\|POS=NOUN`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Voc\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Voc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Gender=Masc\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `POS=VERB\|Tense=Fut\|VerbForm=Inf\|Voice=Pass`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=ADJ`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=NOUN`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Aspect=Imp\|Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Neut\|Number=Dual\|POS=NOUN`, `Case=Gen\|Gender=Masc\|Number=Dual\|POS=PRON`, `Case=Voc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Number=Sing\|POS=ADJ`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin`, `Aspect=Perf\|Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Number=Sing\|POS=NOUN`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Imp\|Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Pqp\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Pass`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Mood=Sub\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Neut\|Number=Dual\|POS=ADJ`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin`, `POS=VERB\|VerbForm=Inf\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part`, `Aspect=Perf\|Case=Acc\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Voc\|Gender=Masc\|Number=Dual\|POS=NOUN`, `Case=Nom\|Number=Plur\|POS=PRON`, `Aspect=Perf\|POS=VERB\|Tense=Past\|VerbForm=Inf`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Dual\|POS=DET`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Degree=Cmp\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Degree=Sup\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Acc\|Degree=Sup\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pqp\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Degree=Sup\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Gen\|Degree=Cmp\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=1\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=X\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Degree=Sup\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pqp\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Neut\|Number=Dual\|POS=DET`, `Case=Nom\|Gender=Neut\|POS=NOUN`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|POS=VERB\|Tense=Fut\|VerbForm=Inf\|Voice=Act`, `Case=Voc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Pass`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Nom\|Degree=Sup\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Voc\|Gender=Masc\|Number=Sing\|POS=PRON`, `Aspect=Perf\|Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Imp\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=2\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Voc\|Gender=Masc\|Number=Dual\|POS=ADJ`, `Mood=Sub\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Dat\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Mood=Sub\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Pass`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Imp\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Dat\|Degree=Sup\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Dual\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `POS=VERB\|Tense=Pres\|VerbForm=Inf\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=X\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Voc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Mood=Opt\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Mood=Imp\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Voc\|Gender=Masc\|Number=Dual\|POS=PRON`, `Mood=Sub\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Imp\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Mood=Imp\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Mid`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Sub\|Number=Plur\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Aspect=Imp\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Aspect=Perf\|Case=Acc\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Degree=Sup\|Gender=Fem\|Number=Sing\|POS=ADJ`, `Mood=Opt\|Number=Dual\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Mood=Imp\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=X\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Voc\|Number=Sing\|POS=NOUN`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part`, `Aspect=Perf\|Mood=Opt\|Number=Sing\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=1\|Tense=Past\|VerbForm=Fin`, `Aspect=Imp\|Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Number=Plur\|POS=PRON`, `Case=Voc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Mood=Opt\|Number=Plur\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Degree=Cmp\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin`, `Mood=Sub\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin`, `Aspect=Imp\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Mood=Imp\|Number=Dual\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part`, `Aspect=Perf\|Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Mood=Ind\|Number=Dual\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin\|Voice=Mid`, `Case=Dat\|Degree=Sup\|Gender=Neut\|Number=Sing\|POS=ADJ`, `Mood=Imp\|Number=Plur\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Mid`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Case=Dat\|Number=Sing\|POS=PRON\|Person=2`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=PRON\|Person=1`, `Case=Nom\|Number=Sing\|POS=PRON\|Person=2`, `Case=Gen\|Number=Sing\|POS=PRON\|Person=2`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=PRON\|Person=1`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=PRON\|Person=1`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Case=Acc\|Number=Sing\|POS=PRON\|Person=2`, `Case=Nom\|Gender=Fem\|Number=Sing\|POS=PRON\|Person=1`, `Case=Dat\|Number=Plur\|POS=PRON\|Person=2`, `Case=Voc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=1`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=PRON\|Person=3`, `Aspect=Perf\|Mood=Ind\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin\|Voice=Pass`, `Case=Acc\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=1`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=1`, `Degree=Sup\|POS=ADJ`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=2`, `Case=Nom\|Gender=Masc\|Mood=Ind\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Number=Plur\|POS=X`, `Case=Dat\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Gen\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Pass`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Pass`, `Aspect=Perf\|Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Pres\|VerbForm=Fin\|Voice=Pass`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=NUM`, `Aspect=Perf\|Case=Dat\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Mood=Ind\|Number=Plur\|POS=VERB\|Person=1\|Tense=Pqp\|VerbForm=Fin\|Voice=Mid`, `Case=Nom\|Gender=Masc\|Mood=Ind\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Fin\|Voice=Act`, `Case=Acc\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Masc\|Number=Sing\|POS=PRON\|Person=1`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=PRON\|Person=2`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Fut\|VerbForm=Fin\|Voice=Act`, `Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|Voice=Act`, `Case=Dat\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Gen\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Masc\|Number=Plur\|POS=NUM`, `Case=Voc\|Degree=Cmp\|Gender=Masc\|Number=Sing\|POS=ADJ`, `Case=Gen\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Mid`, `POS=VERB\|Tense=Pres\|VerbForm=Inf`, `Aspect=Perf\|Case=Voc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Case=Nom\|Gender=Fem\|Number=Dual\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Acc\|Number=Dual\|POS=NOUN`, `Case=Acc\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Gen\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Nom\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Pres\|VerbForm=Part`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Degree=Sup\|Gender=Masc\|Number=Plur\|POS=ADJ`, `Case=Nom\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Pass`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Dat\|Degree=Sup\|Gender=Neut\|Number=Plur\|POS=ADJ`, `Case=Nom\|Degree=Cmp\|Gender=Fem\|Number=Plur\|POS=ADJ`, `Aspect=Perf\|Case=Dat\|Gender=Neut\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Act`, `Aspect=Perf\|Case=Dat\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Case=Dat\|Gender=Fem\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Inf\|Voice=Act`, `Case=Nom\|Gender=Neut\|Number=Plur\|POS=X`, `Case=Acc\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Gender=Masc\|Number=Plur\|POS=VERB\|Tense=Fut\|VerbForm=Part\|Voice=Act`, `Case=Acc\|Gender=Masc\|Number=Plur\|POS=VERB\|VerbForm=Part\|Voice=Mid`, `Case=Gen\|Gender=Fem\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Pass`, `Case=Gen\|Gender=Neut\|Number=Plur\|POS=VERB\|Tense=Past\|VerbForm=Part\|Voice=Mid`, `Aspect=Perf\|Case=Nom\|Gender=Masc\|Number=Sing\|POS=VERB\|Tense=Past\|VerbForm=Part`, `Case=Acc\|Gender=Fem\|POS=VERB\|Tense=Pres\|VerbForm=Part\|Voice=Mid`, `Mood=Opt\|Number=Sing\|POS=VERB\|Person=3\|Tense=Past\|VerbForm=Fin`, `Case=Acc\|Number=Plur\|POS=PRON`, `Mood=Ind\|Number=Sing\|POS=VERB\|Person=2\|Tense=Fut\|VerbForm=Fin` |
| **`tagger`** | `---------`, `--p---fa-`, `--s---ma-`, `-3paia---`, `-3paim---`, `-3siia---`, `a`, `c`, `d`, `g`, `i`, `l`, `m`, `n`, `p`, `r`, `u`, `v`, `x--------`, `x-p----d-`, `x-p---nn-` |
| **`parser`** | `ROOT`, `acl`, `advcl`, `advmod`, `amod`, `appos`, `case`, `cc`, `ccomp`, `conj`, `cop`, `csubj`, `dep`, `det`, `iobj`, `mark`, `nmod`, `nsubj`, `nummod`, `obj`, `obl`, `parataxis`, `punct`, `vocative`, `xcomp` |
| **`senter`** | `I`, `S` |
</details>
### Accuracy
| Type | Score |
| --- | --- |
| `POS_ACC` | 96.72 |
| `MORPH_ACC` | 94.45 |
| `LEMMA_ACC` | 92.23 |
| `TAG_ACC` | 96.72 |
| `DEP_UAS` | 78.97 |
| `DEP_LAS` | 74.05 |
| `SENTS_P` | 98.16 |
| `SENTS_R` | 98.42 |
| `SENTS_F` | 98.29 |
| `TRANSFORMER_LOSS` | 1689634.70 |
| `MORPHOLOGIZER_LOSS` | 105328.58 |
| `TAGGER_LOSS` | 27313.43 |
| `PARSER_LOSS` | 2331148.85 |
| `SENTER_LOSS` | 139276.31 |
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.