Update app.py
Browse files
app.py
CHANGED
|
@@ -517,10 +517,113 @@ def predict_tags(test_sentence):
|
|
| 517 |
|
| 518 |
predict_tags(test_sentence)
|
| 519 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 520 |
|
| 521 |
tagger = gr.Interface(
|
| 522 |
predict_tags,
|
| 523 |
gr.Textbox(placeholder="Enter sentence here..."),
|
|
|
|
| 524 |
["highlight"],
|
| 525 |
title="BERT Filipino Part of Speech Tagger",
|
| 526 |
description="Enter a text in Tagalog to classify the tags for each word. Each word to tag needs to be space separated.",
|
|
|
|
| 517 |
|
| 518 |
predict_tags(test_sentence)
|
| 519 |
|
| 520 |
+
def get_readme():
|
| 521 |
+
return
|
| 522 |
+
"""
|
| 523 |
+
|
| 524 |
+
This is a BERT Tagalog Base Uncased Part of Speech tagger fine-tuned model of [Jiang et. al.'s pre-trained bert-tagalog-base-uncased model](https://huggingface.co/GKLMIP/bert-tagalog-base-uncased).
|
| 525 |
+
|
| 526 |
+
## Model Details
|
| 527 |
+
|
| 528 |
+
### Model Description
|
| 529 |
+
|
| 530 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 531 |
+
|
| 532 |
+
- **Developed by:** syke9p3, mnemoria, xenoxia, riakm
|
| 533 |
+
- **Shared by:** syke9p3
|
| 534 |
+
- **Model type:** BERT Tagalog Base Uncased
|
| 535 |
+
- **Languages (NLP):** Tagalog, Filipino
|
| 536 |
+
- **Dataset:** Sagum et. al.'s annotated Tagalog Corpora based on MGNN Tagset convention. This model was trained in 800 sentences and evaluated with 200 sentences.
|
| 537 |
+
- **Finetuned from model**: [Jiang et. al.'s pre-trained bert-tagalog-base-uncased model](https://huggingface.co/GKLMIP/bert-tagalog-base-uncased)
|
| 538 |
+
|
| 539 |
+
## GitHub Link
|
| 540 |
+
[syke9p3/bert-tagalog-pos-tagger](https://github.com/syke9p3/bert-tagalog-pos-tagger)
|
| 541 |
+
|
| 542 |
+
### Tags
|
| 543 |
+
|
| 544 |
+
| Part of Speech | Tags |
|
| 545 |
+
|-----------------------------------------------|------|
|
| 546 |
+
| **Noun** | NNC |
|
| 547 |
+
| Common Noun | NNC |
|
| 548 |
+
| Proper Noun | NNP |
|
| 549 |
+
| Proper Noun Abbreviation | NNPA |
|
| 550 |
+
| Common Noun Abbreviation | NNCA |
|
| 551 |
+
| **Pronoun** | PR |
|
| 552 |
+
| as Subject (Palagyo)/Personal Pronouns Singular | PRS |
|
| 553 |
+
| Personal Pronouns | PRP |
|
| 554 |
+
| Possessive Subject (Paari) | PRSP |
|
| 555 |
+
| Pointing to an Object Demonstrative/(Paturol/Pamatlig) | PRO |
|
| 556 |
+
| Question/Interrogative (Pananong)/Singular | PRQ |
|
| 557 |
+
| Question/Interrogative Plural | PRQP |
|
| 558 |
+
| Location (Panlunan) | PRL |
|
| 559 |
+
| Comparison (Panulad) | PRC |
|
| 560 |
+
| Found (Pahimaton) | PRF |
|
| 561 |
+
| Indefinite | PRI |
|
| 562 |
+
| **Determiner** | DT |
|
| 563 |
+
| Determiner (Pantukoy) for Common Noun Plural | DTC |
|
| 564 |
+
| Determiner (Pantukoy) for Proper Noun | DTP |
|
| 565 |
+
| Determiner (Pantukoy) for Proper Noun Plural | DTPP |
|
| 566 |
+
| Lexical Marker | LM |
|
| 567 |
+
| Conjunctions (Pang-ugnay) | CC, CCT, CCR, CCB, CCA |
|
| 568 |
+
| Ligatures (Pang-angkop) | CCP |
|
| 569 |
+
| Preposition (Pang-ukol) | CCU |
|
| 570 |
+
| **Verb (Pandiwa)** | VB |
|
| 571 |
+
| Neutral/Infinitive | VBW |
|
| 572 |
+
| Auxiliary, Modal/Pseudo-verbs | VBS |
|
| 573 |
+
| Existential | VBH |
|
| 574 |
+
| Non-existential | VBN |
|
| 575 |
+
| Time Past (Perfective) | VBTS |
|
| 576 |
+
| Time Present (Imperfective) | VBTR |
|
| 577 |
+
| Time Future (Contemplative) | VBTF |
|
| 578 |
+
| Recent past | VBTP |
|
| 579 |
+
| Actor Focus | VBAF |
|
| 580 |
+
| Object/Goal Focus | VBOF |
|
| 581 |
+
| Benefactive Focus | VBOB |
|
| 582 |
+
| Locative Focus | VBOL |
|
| 583 |
+
| Instrumental Focus | VBOI |
|
| 584 |
+
| Referential/Measurement Focus | VBRF |
|
| 585 |
+
| **Adjective (Pang-uri)** | JJ |
|
| 586 |
+
| Describing (Panlarawan) | JJD |
|
| 587 |
+
| Used for Comparison (same level) (Pahambing Magkatulad) | JJC |
|
| 588 |
+
| Comparison Comparative (more) (Palamang) | JJCC |
|
| 589 |
+
| Comparison Superlative (most) (Pasukdol) | JJCS |
|
| 590 |
+
| Comparison Negation (not quite) (Di-Magkatulad) | JJCN |
|
| 591 |
+
| Describing Number (Pamilang) | JJN |
|
| 592 |
+
| **Adverb (Pang-Abay)** | RB |
|
| 593 |
+
| Describing “How” (Pamaraan) | RBD |
|
| 594 |
+
| Number (Panggaano/Panukat) | RBN |
|
| 595 |
+
| Conditional (Kondisyunal) | RBK |
|
| 596 |
+
| Causative (Pananhi) | RBP |
|
| 597 |
+
| Benefactive (Benepaktibo) | RBB |
|
| 598 |
+
| Referential (Pangkaukulan) | RBR |
|
| 599 |
+
| Question (Pananong) | RBQ |
|
| 600 |
+
| Agree (Panang-ayon) | RBT |
|
| 601 |
+
| Disagree (Pananggi) | RBF |
|
| 602 |
+
| Frequency (Pamanahon) | RBW |
|
| 603 |
+
| Possibility (Pang-agam) | RBM |
|
| 604 |
+
| Place (Panlunan) | RBL |
|
| 605 |
+
| Enclitics (Paningit) | RBI |
|
| 606 |
+
| Interjections (Sambitla) | RBJ |
|
| 607 |
+
| Social Formula (Pormularyong Panlipunan) | RBS |
|
| 608 |
+
|**Cardinal Number (Bilang)** | CD |
|
| 609 |
+
| Digit, Rank, Count | CDB |
|
| 610 |
+
| **Topicless (Walang Paksa)** | TS |
|
| 611 |
+
| Foreign Words | FW |
|
| 612 |
+
| **Punctuation (Pananda)** | PM |
|
| 613 |
+
| Period | PMP |
|
| 614 |
+
| Exclamation Point | PME |
|
| 615 |
+
| Question Mark | PMQ |
|
| 616 |
+
| Comma | PMC |
|
| 617 |
+
| Semi-colon | PMSC |
|
| 618 |
+
| **Symbols** | PMS |
|
| 619 |
+
|
| 620 |
+
"""
|
| 621 |
+
|
| 622 |
|
| 623 |
tagger = gr.Interface(
|
| 624 |
predict_tags,
|
| 625 |
gr.Textbox(placeholder="Enter sentence here..."),
|
| 626 |
+
gr.Markdown(get_readme()),
|
| 627 |
["highlight"],
|
| 628 |
title="BERT Filipino Part of Speech Tagger",
|
| 629 |
description="Enter a text in Tagalog to classify the tags for each word. Each word to tag needs to be space separated.",
|