Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks
Paper โข 1908.10084 โข Published โข 20
How to use check2207/CS221_AITeamVN_reranker with sentence-transformers:
from sentence_transformers import CrossEncoder
model = CrossEncoder("check2207/CS221_AITeamVN_reranker")
query = "Which planet is known as the Red Planet?"
passages = [
"Venus is often called Earth's twin because of its similar size and proximity.",
"Mars, known for its reddish appearance, is often referred to as the Red Planet.",
"Jupiter, the largest planet in our solar system, has a prominent red spot.",
"Saturn, famous for its rings, is sometimes mistaken for the Red Planet."
]
scores = model.predict([(query, passage) for passage in passages])
print(scores)This is a Cross Encoder model finetuned from AITeamVN/Vietnamese_Reranker using the sentence-transformers library. It computes scores for pairs of texts, which can be used for text reranking and semantic search.
CrossEncoder(
(0): Transformer({'transformer_task': 'sequence-classification', 'modality_config': {'text': {'method': 'forward', 'method_output_name': 'logits'}}, 'module_output_name': 'scores', 'architecture': 'XLMRobertaForSequenceClassification'})
)
First install the Sentence Transformers library:
pip install -U sentence-transformers
Then you can load this model and run inference.
from sentence_transformers import CrossEncoder
# Download from the ๐ค Hub
model = CrossEncoder("cross_encoder_model_id")
# Get scores for pairs of inputs
pairs = [
['Nhรขn viรชn phแปฅc vแปฅ xe khรกch khรดng hแป trแปฃ, giรบp ฤแปก hร nh khรกch lร ngฦฐแปi cao tuแปi, trแบป em bแป phแบกt bao nhiรชu?', 'ฤiแปu 59 - Luแบญt ฤฦฐแปng bแป : Quyแปn vร nghฤฉa vแปฅ cแปงa ngฦฐแปi lรกi xe, nhรขn viรชn phแปฅc vแปฅ trรชn xe รด tรด kinh doanh vแบญn tแบฃi hร nh khรกch\n1. Tแปซ chแปi vแบญn tแบฃi hร nh khรกch cรณ mแปt trong cรกc hร nh vi sau ฤรขy:\na) Gรขy rแปi trแบญt tแปฑ cรดng cแปng trรชn xe รด tรด;\nb) Cแบฃn trแป cรดng viแปc cแปงa ngฦฐแปi lรกi xe, nhรขn viรชn phแปฅc vแปฅ trรชn xe รด tรด;\nc) Gรขy แบฃnh hฦฐแปng ฤแบฟn tรญnh mแบกng, sแปฉc khoแบป, tร i sแบฃn cแปงa ngฦฐแปi khรกc trรชn xe รด tรด;\nd) Gian lแบญn vรฉ;\nฤ) Vi phแบกm quy ฤแปnh tแบกi ฤiแปm c khoแบฃn 2 ฤiแปu 60 cแปงa Luแบญt nร y.\n2. Tแปซ chแปi ฤiแปu khiแปn phฦฐฦกng tiแปn khi phรกt hiแปn phฦฐฦกng tiแปn khรดng bแบฃo ฤแบฃm cรกc ฤiแปu kiแปn vแป an toร n theo quy ฤแปnh cแปงa phรกp luแบญt vแป trแบญt tแปฑ, an toร n giao thรดng ฤฦฐแปng bแป.\n3. Khรดng ฤฦฐแปฃc tแปซ chแปi vแบญn tแบฃi hร nh khรกch, trแปซ lรฝ do quy ฤแปnh tแบกi khoแบฃn 1 ฤiแปu nร y; khรดng ฤฦฐแปฃc gรขy khรณ khฤn ฤแปi vแปi hร nh khรกch lร ngฦฐแปi cao tuแปi, ngฦฐแปi khuyแบฟt tแบญt, trแบป em, phแปฅ nแปฏ mang thai; khรดng ฤฦฐแปฃc chuyแปn hร nh khรกch sang phฦฐฦกng tiแปn khรกc khi chฦฐa ฤฦฐแปฃc sแปฑ ฤแปng รฝ cแปงa hร nh khรกch, trแปซ trฦฐแปng hแปฃp bแบฅt khแบฃ khรกng.\n4. Thu tiแปn ฤรบng giรก vรฉ niรชm yแบฟt hoแบทc theo hแปฃp ฤแปng vแบญn tแบฃi.\n5. Cรณ thรกi ฤแป, lแปi nรณi, hร nh vi vฤn minh, lแปch sแปฑ; hฦฐแปng dแบซn hร nh khรกch ngแปi ฤรบng nฦกi quy ฤแปnh, bแป trรญ chแป ngแปi ฦฐu tiรชn cho ngฦฐแปi cao tuแปi, ngฦฐแปi khuyแบฟt tแบญt, trแบป em vร phแปฅ nแปฏ mang thai.\n6. Thแปฑc hiแปn trรกch nhiแปm theo quy ฤแปnh cแปงa phรกp luแบญt vแป trแบญt tแปฑ, an toร n giao thรดng ฤฦฐแปng bแป vร quy ฤแปnh khรกc cแปงa phรกp luแบญt cรณ liรชn quan.'],
['Trรฌnh tแปฑ vร nแปi dung lร m viแปc cแปงa hแปi ฤแปng tฦฐ vแบฅn tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc chแปง trรฌ dแปฑ รกn Chฦฐฦกng trรฌnh phรกt triแปn tร i sแบฃn trรญ tuแป ฤฦฐแปฃc quy ฤแปnh nhฦฐ thแบฟ nร o?', 'ฤiแปu 18. Lแบฅy รฝ kiแบฟn tฦฐ vแบฅn cแปงa chuyรชn gia ฤแปc lแบญp vแป viแปc tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc, cรก nhรขn chแปง trรฌ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh : 1. ฤฦกn vแป quแบฃn lรฝ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh trรฌnh lรฃnh ฤแบกo Bแป quyแบฟt ฤแปnh lแปฑa chแปn 03 chuyรชn gia ฤแปc lแบญp ฤแป lแบฅy รฝ kiแบฟn tฦฐ vแบฅn vแป viแปc tuyแปn chแปn tแป chแปฉc, tแป chแปฉc chแปง trรฌ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh.\n2. Chuyรชn gia ฤแปc lแบญp cรณ รฝ kiแบฟn tฦฐ vแบฅn bแบฑng vฤn bแบฃn sau khi ฤรกnh giรก cรกc hแป sฦก tham gia tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc, tแป chแปฉc chแปง trรฌ nhiแปm vแปฅ. Chuyรชn gia ฤแปc lแบญp lร m viแปc ฤแปc lแบญp vร chแปu trรกch nhiแปm trฦฐแปc phรกp luแบญt vแป nhแบญn xรฉt, ฤรกnh giรก vร kแบฟt luแบญn cแปงa mรฌnh.\n3. ฤฦกn vแป quแบฃn lรฝ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh tแป chแปฉc giao hแป sฦก cho cรกc chuyรชn gia ฤแปc lแบญp ฤแป thแปฑc hiแปn viแปc ฤรกnh giรก lแบกi hแป sฦก cรกc nhiแปm vแปฅ tham gia tuyแปn chแปn, giao trแปฑc tiแบฟp. Thแปi gian ฤรกnh giรก hแป sฦก tแปi ฤa 15 ngร y kแป tแปซ khi nhแบญn hแป sฦก.\n4. Hแป sฦก tuyแปn chแปn, giao trแปฑc tiแบฟp ฤฦฐแปฃc xem xรฉt phรช duyแปt nแบฟu cรณ รญt nhแบฅt 02 chuyรชn gia ฤแปc lแบญp ฤรกnh giรก 70 ฤiแปm trแป lรชn.'],
['Cรกc bฦฐแปc thแปฑc hiแปn viแปc mแป vร ฤรกnh giรก hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi cแปงa nhร ฤแบงu tฦฐ dแปฑ รกn ฤแบงu tฦฐ cรณ sแปญ dแปฅng ฤแบฅt ฤฦฐแปฃc quy ฤแปnh nhฦฐ thแบฟ nร o?', 'ฤiแปu 27. Quy trรฌnh chi tiแบฟt : 1. Chuแบฉn bแป lแปฑa chแปn nhร ฤแบงu tฦฐ, bao gแปm:\na) Lแบญp hแป sฦก mแปi thแบงu;\nb) Thแบฉm ฤแปnh vร phรช duyแปt hแป sฦก mแปi thแบงu.\n2. Tแป chแปฉc lแปฑa chแปn nhร ฤแบงu tฦฐ, bao gแปm:\na) Mแปi thแบงu;\nb) Phรกt hร nh, sแปญa ฤแปi, lร m rรต hแป sฦก mแปi thแบงu;\nc) Chuแบฉn bแป, nแปp, tiแบฟp nhแบญn, quแบฃn lรฝ, sแปญa ฤแปi, rรบt hแป sฦก dแปฑ thแบงu;\nd) Mแป hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt.\n3. ฤรกnh giรก hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt, bao gแปm:\na) Kiแปm tra, ฤรกnh giรก tรญnh hแปฃp lแป cแปงa hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt;\nb) ฤรกnh giรก chi tiแบฟt hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt;\nc) Thแบฉm ฤแปnh, phรช duyแปt danh sรกch nhร ฤแบงu tฦฐ ฤรกp แปฉng yรชu cแบงu vแป kแปน thuแบญt.\n4. Mแป vร ฤรกnh giรก hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi, bao gแปm:\na) Mแป hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi;\nb) Kiแปm tra, ฤรกnh giรก tรญnh hแปฃp lแป cแปงa hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi;\nc) ฤรกnh giรก chi tiแบฟt hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi vร xแบฟp hแบกng nhร ฤแบงu tฦฐ.\n5. Trรฌnh, thแบฉm ฤแปnh, phรช duyแปt vร cรดng khai kแบฟt quแบฃ lแปฑa chแปn nhร ฤแบงu tฦฐ.\n6. ฤร m phรกn, hoร n thiแปn, kรฝ kแบฟt hแปฃp ฤแปng, bao gแปm:\na) ฤร m phรกn, hoร n thiแปn hแปฃp ฤแปng;\nb) Kรฝ kแบฟt hแปฃp ฤแปng vร cรดng khai thรดng tin hแปฃp ฤแปng dแปฑ รกn.\n7. Trฦฐแปng hแปฃp dแปฑ รกn แปฉng dแปฅng cรดng nghแป cao, quy trรฌnh lแปฑa chแปn nhร ฤแบงu tฦฐ thแปฑc hiแปn theo quy ฤแปnh tแบกi cรกc khoแบฃn 1, 2, 3, 4 vร 5 ฤiแปu nร y. Nhร ฤแบงu tฦฐ ฤฦฐแปฃc lแปฑa chแปn sแบฝ tiแบฟn hร nh lแบญp bรกo cรกo nghiรชn cแปฉu khแบฃ thi. Trรชn cฦก sแป bรกo cรกo nghiรชn cแปฉu khแบฃ thi ฤฦฐแปฃc duyแปt, cฦก quan nhร nฦฐแปc cรณ thแบฉm quyแปn vร nhร ฤแบงu tฦฐ tiแบฟn hร nh ฤร m phรกn, hoร n thiแปn vร kรฝ kแบฟt hแปฃp ฤแปng. Nแปi dung bรกo cรกo nghiรชn cแปฉu khแบฃ thi vร hแปฃp ฤแปng phแบฃi bแบฃo ฤแบฃm phรน hแปฃp vแปi mแปฅc tiรชu thแปฑc hiแปn dแปฑ รกn, quy mรด ฤแบงu tฦฐ, cรกc tiรชu chรญ cฦก bแบฃn ฤแป xรฉt duyแปt nhร ฤแบงu tฦฐ trรบng thแบงu.'],
['Trฦฐแปng chuyรชn ฤฦฐแปฃc thร nh lแบญp แป cแบฅp trung hแปc phแป thรดng dร nh cho hแปc sinh ฤแบกt kแบฟt quแบฃ xuแบฅt sแบฏc trong hแปc tแบญp nhแบฑm phรกt triแปn nฤng khiแบฟu vแป mแปt sแป mรดn hแปc trรชn cฦก sแป nร o sau ฤรขy?', 'ฤiแปu 28 - Luแบญt Giรกo dแปฅc : Cแบฅp hแปc vร ฤแป tuแปi cแปงa giรกo dแปฅc phแป thรดng\n\n1. Cรกc cแบฅp hแปc vร ฤแป tuแปi cแปงa giรกo dแปฅc phแป thรดng ฤฦฐแปฃc quy ฤแปnh nhฦฐ sau:\n\na) Giรกo dแปฅc tiแปu hแปc ฤฦฐแปฃc thแปฑc hiแปn trong 05 nฤm hแปc, tแปซ lแปp mแปt ฤแบฟn hแบฟt lแปp nฤm. Tuแปi cแปงa hแปc sinh vร o hแปc lแปp mแปt lร 06 tuแปi vร ฤฦฐแปฃc tรญnh theo nฤm;\n\nb) Giรกo dแปฅc trung hแปc cฦก sแป ฤฦฐแปฃc thแปฑc hiแปn trong 04 nฤm hแปc, tแปซ lแปp sรกu ฤแบฟn hแบฟt lแปp chรญn. Hแปc sinh vร o hแปc lแปp sรกu phแบฃi hoร n thร nh chฦฐฦกng trรฌnh tiแปu hแปc. Tuแปi cแปงa hแปc sinh vร o hแปc lแปp sรกu lร 11 tuแปi vร ฤฦฐแปฃc tรญnh theo nฤm;\n\nc) Giรกo dแปฅc trung hแปc phแป thรดng ฤฦฐแปฃc thแปฑc hiแปn trong 03 nฤm hแปc, tแปซ lแปp mฦฐแปi ฤแบฟn hแบฟt lแปp mฦฐแปi hai. Hแปc sinh vร o hแปc lแปp mฦฐแปi phแบฃi cรณ bแบฑng tแปt nghiแปp trung hแปc cฦก sแป. Tuแปi cแปงa hแปc sinh vร o hแปc lแปp mฦฐแปi lร 15 tuแปi vร ฤฦฐแปฃc tรญnh theo nฤm.\n\n2. Trฦฐแปng hแปฃp hแปc sinh ฤฦฐแปฃc hแปc vฦฐแปฃt lแปp, hแปc แป ฤแป tuแปi cao hฦกn tuแปi quy ฤแปnh tแบกi khoแบฃn 1 ฤiแปu nร y bao gแปm:\n\na) Hแปc sinh hแปc vฦฐแปฃt lแปp trong trฦฐแปng hแปฃp phรกt triแปn sแปm vแป trรญ tuแป;\n\nb) Hแปc sinh hแปc แป ฤแป tuแปi cao hฦกn tuแปi quy ฤแปnh trong trฦฐแปng hแปฃp hแปc sinh hแปc lฦฐu ban, hแปc sinh แป vรนng cรณ ฤiแปu kiแปn kinh tแบฟ - xรฃ hแปi ฤแบทc biแปt khรณ khฤn, hแปc sinh lร ngฦฐแปi dรขn tแปc thiแปu sแป, hแปc sinh lร ngฦฐแปi khuyแบฟt tแบญt, hแปc sinh kรฉm phรกt triแปn vแป thแป lแปฑc hoแบทc trรญ tuแป, hแปc sinh mแป cรดi khรดng nฦกi nฦฐฦกng tแปฑa, hแปc sinh thuแปc hแป nghรจo, hแปc sinh แป nฦฐแปc ngoร i vแป nฦฐแปc vร trฦฐแปng hแปฃp khรกc theo quy ฤแปnh cแปงa phรกp luแบญt.\n\n3. Giรกo dแปฅc phแป thรดng ฤฦฐแปฃc chia thร nh giai ฤoแบกn giรกo dแปฅc cฦก bแบฃn vร giai ฤoแบกn giรกo dแปฅc ฤแปnh hฦฐแปng nghแป nghiแปp. Giai ฤoแบกn giรกo dแปฅc cฦก bแบฃn gแปm cแบฅp tiแปu hแปc vร cแบฅp trung hแปc cฦก sแป; giai ฤoแบกn giรกo dแปฅc ฤแปnh hฦฐแปng nghแป nghiแปp lร cแบฅp trung hแปc phแป thรดng. Hoฬฃc sinh trong cฦก sแป giรกo dแปฅc nghรชฬ nghiรชฬฃp ฤฦฐฦกฬฃc hoฬฃc khแปi lฦฐแปฃng kiรชฬn thฦฐฬc vฤn hรณa trung hแปc phรดฬ thรดng.\n\n4. Bแป trฦฐแปng Bแป Giรกo dแปฅc vร ฤร o tแบกo quy ฤแปnh viแปc dแบกy vร hแปc tiแบฟng Viแปt cho trแบป em lร ngฦฐแปi dรขn tแปc thiแปu sแป trฦฐแปc khi vร o hแปc lแปp mแปt; viแปc giแบฃng dแบกy khแปi lฦฐแปฃng kiแบฟn thฦฐฬc vฤn hoฬa trung hoฬฃc phรดฬ thรดng trong cฦก sแป giรกo dแปฅc nghแป nghiแปp; cรกc trฦฐแปng hแปฃp quy ฤแปnh tแบกi khoแบฃn 2 ฤiแปu nร y.'],
['Dแปch vแปฅ viแปc lร m gแปm cรณ nhแปฏng hoแบกt ฤแปng nร o?', 'ฤiแปu 12. Hร nh vi vแบญn chuyแปn trรกi phรฉp hร ng hรณa, tiแปn tแป qua biรชn giแปi ngoร i khu vแปฑc cแปญa khแบฉu nhฦฐng chฦฐa ฤแบฟn mแปฉc truy cแปฉu trรกch nhiแปm hรฌnh sแปฑ : 1. Phแบกt tiแปn tแปซ 1.000.000 ฤแปng ฤแบฟn 5.000.000 ฤแปng ฤแปi vแปi hร nh vi vi phแบกm mร tang vแบญt vi phแบกm cรณ giรก trแป dฦฐแปi 10.000.000 ฤแปng.\n2. Phแบกt tiแปn tแปซ 5.000.000 ฤแปng ฤแบฟn 10.000.000 ฤแปng ฤแปi vแปi mแปt trong nhแปฏng hร nh vi sau:\na) Hร nh vi vi phแบกm mร tang vแบญt vi phแบกm cรณ giรก trแป tแปซ 10.000.000 ฤแปng ฤแบฟn dฦฐแปi 50.000.000 ฤแปng;\nb) Vแบญn chuyแปn trรกi phรฉp qua biรชn giแปi hร ng hรณa lร vฤn hรณa phแบฉm.\n3. Phแบกt tiแปn tแปซ 20.000.000 ฤแปng ฤแบฟn 40.000.000 ฤแปng ฤแปi vแปi hร nh vi vi phแบกm mร tang vแบญt vi phแบกm cรณ giรก trแป tแปซ 50.000.000 ฤแปng ฤแบฟn dฦฐแปi 100.000.000 ฤแปng?\n4. Phแบกt tiแปn tแปซ 40.000.000 ฤแปng ฤแบฟn 50.000.000 ฤแปng ฤแปi vแปi tang vแบญt vi phแบกm cรณ giรก trแป tแปซ 100.000.000 ฤแปng trแป lรชn.\n5. Hรฌnh thแปฉc xแปญ phแบกt bแป sung: Tแปch thu tang vแบญt ฤแปi vแปi hร nh vi vi phแบกm quy ฤแปnh tแบกi ฤiแปu nร y, trแปซ trฦฐแปng hแปฃp ฤรฃ รกp dแปฅng biแปn phรกp khแบฏc phแปฅc hแบญu quแบฃ ฤฦฐแปฃc quy ฤแปnh tแบกi khoแบฃn 6 ฤiแปu nร y.\n6. Biแปn phรกp khแบฏc phแปฅc hแบญu quแบฃ: Buแปc tiรชu hแปงy hร ng hรณa, vแบญt phแบฉm gรขy hแบกi cho sแปฉc khแปe con ngฦฐแปi, vแบญt nuรดi, cรขy trแปng, mรดi trฦฐแปng vร vฤn hรณa phแบฉm cรณ nแปi dung ฤแปc hแบกi ฤแปi vแปi hร nh vi vi phแบกm quy ฤแปnh tแบกi ฤiแปu nร y.'],
]
scores = model.predict(pairs)
print(scores)
# [1.3022e-05 1.2965e-05 1.7884e-03 2.5695e-04 1.2972e-05]
# Or rank different texts based on similarity to a single text
ranks = model.rank(
'Nhรขn viรชn phแปฅc vแปฅ xe khรกch khรดng hแป trแปฃ, giรบp ฤแปก hร nh khรกch lร ngฦฐแปi cao tuแปi, trแบป em bแป phแบกt bao nhiรชu?',
[
'ฤiแปu 59 - Luแบญt ฤฦฐแปng bแป : Quyแปn vร nghฤฉa vแปฅ cแปงa ngฦฐแปi lรกi xe, nhรขn viรชn phแปฅc vแปฅ trรชn xe รด tรด kinh doanh vแบญn tแบฃi hร nh khรกch\n1. Tแปซ chแปi vแบญn tแบฃi hร nh khรกch cรณ mแปt trong cรกc hร nh vi sau ฤรขy:\na) Gรขy rแปi trแบญt tแปฑ cรดng cแปng trรชn xe รด tรด;\nb) Cแบฃn trแป cรดng viแปc cแปงa ngฦฐแปi lรกi xe, nhรขn viรชn phแปฅc vแปฅ trรชn xe รด tรด;\nc) Gรขy แบฃnh hฦฐแปng ฤแบฟn tรญnh mแบกng, sแปฉc khoแบป, tร i sแบฃn cแปงa ngฦฐแปi khรกc trรชn xe รด tรด;\nd) Gian lแบญn vรฉ;\nฤ) Vi phแบกm quy ฤแปnh tแบกi ฤiแปm c khoแบฃn 2 ฤiแปu 60 cแปงa Luแบญt nร y.\n2. Tแปซ chแปi ฤiแปu khiแปn phฦฐฦกng tiแปn khi phรกt hiแปn phฦฐฦกng tiแปn khรดng bแบฃo ฤแบฃm cรกc ฤiแปu kiแปn vแป an toร n theo quy ฤแปnh cแปงa phรกp luแบญt vแป trแบญt tแปฑ, an toร n giao thรดng ฤฦฐแปng bแป.\n3. Khรดng ฤฦฐแปฃc tแปซ chแปi vแบญn tแบฃi hร nh khรกch, trแปซ lรฝ do quy ฤแปnh tแบกi khoแบฃn 1 ฤiแปu nร y; khรดng ฤฦฐแปฃc gรขy khรณ khฤn ฤแปi vแปi hร nh khรกch lร ngฦฐแปi cao tuแปi, ngฦฐแปi khuyแบฟt tแบญt, trแบป em, phแปฅ nแปฏ mang thai; khรดng ฤฦฐแปฃc chuyแปn hร nh khรกch sang phฦฐฦกng tiแปn khรกc khi chฦฐa ฤฦฐแปฃc sแปฑ ฤแปng รฝ cแปงa hร nh khรกch, trแปซ trฦฐแปng hแปฃp bแบฅt khแบฃ khรกng.\n4. Thu tiแปn ฤรบng giรก vรฉ niรชm yแบฟt hoแบทc theo hแปฃp ฤแปng vแบญn tแบฃi.\n5. Cรณ thรกi ฤแป, lแปi nรณi, hร nh vi vฤn minh, lแปch sแปฑ; hฦฐแปng dแบซn hร nh khรกch ngแปi ฤรบng nฦกi quy ฤแปnh, bแป trรญ chแป ngแปi ฦฐu tiรชn cho ngฦฐแปi cao tuแปi, ngฦฐแปi khuyแบฟt tแบญt, trแบป em vร phแปฅ nแปฏ mang thai.\n6. Thแปฑc hiแปn trรกch nhiแปm theo quy ฤแปnh cแปงa phรกp luแบญt vแป trแบญt tแปฑ, an toร n giao thรดng ฤฦฐแปng bแป vร quy ฤแปnh khรกc cแปงa phรกp luแบญt cรณ liรชn quan.',
'ฤiแปu 18. Lแบฅy รฝ kiแบฟn tฦฐ vแบฅn cแปงa chuyรชn gia ฤแปc lแบญp vแป viแปc tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc, cรก nhรขn chแปง trรฌ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh : 1. ฤฦกn vแป quแบฃn lรฝ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh trรฌnh lรฃnh ฤแบกo Bแป quyแบฟt ฤแปnh lแปฑa chแปn 03 chuyรชn gia ฤแปc lแบญp ฤแป lแบฅy รฝ kiแบฟn tฦฐ vแบฅn vแป viแปc tuyแปn chแปn tแป chแปฉc, tแป chแปฉc chแปง trรฌ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh.\n2. Chuyรชn gia ฤแปc lแบญp cรณ รฝ kiแบฟn tฦฐ vแบฅn bแบฑng vฤn bแบฃn sau khi ฤรกnh giรก cรกc hแป sฦก tham gia tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc, tแป chแปฉc chแปง trรฌ nhiแปm vแปฅ. Chuyรชn gia ฤแปc lแบญp lร m viแปc ฤแปc lแบญp vร chแปu trรกch nhiแปm trฦฐแปc phรกp luแบญt vแป nhแบญn xรฉt, ฤรกnh giรก vร kแบฟt luแบญn cแปงa mรฌnh.\n3. ฤฦกn vแป quแบฃn lรฝ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh tแป chแปฉc giao hแป sฦก cho cรกc chuyรชn gia ฤแปc lแบญp ฤแป thแปฑc hiแปn viแปc ฤรกnh giรก lแบกi hแป sฦก cรกc nhiแปm vแปฅ tham gia tuyแปn chแปn, giao trแปฑc tiแบฟp. Thแปi gian ฤรกnh giรก hแป sฦก tแปi ฤa 15 ngร y kแป tแปซ khi nhแบญn hแป sฦก.\n4. Hแป sฦก tuyแปn chแปn, giao trแปฑc tiแบฟp ฤฦฐแปฃc xem xรฉt phรช duyแปt nแบฟu cรณ รญt nhแบฅt 02 chuyรชn gia ฤแปc lแบญp ฤรกnh giรก 70 ฤiแปm trแป lรชn.',
'ฤiแปu 27. Quy trรฌnh chi tiแบฟt : 1. Chuแบฉn bแป lแปฑa chแปn nhร ฤแบงu tฦฐ, bao gแปm:\na) Lแบญp hแป sฦก mแปi thแบงu;\nb) Thแบฉm ฤแปnh vร phรช duyแปt hแป sฦก mแปi thแบงu.\n2. Tแป chแปฉc lแปฑa chแปn nhร ฤแบงu tฦฐ, bao gแปm:\na) Mแปi thแบงu;\nb) Phรกt hร nh, sแปญa ฤแปi, lร m rรต hแป sฦก mแปi thแบงu;\nc) Chuแบฉn bแป, nแปp, tiแบฟp nhแบญn, quแบฃn lรฝ, sแปญa ฤแปi, rรบt hแป sฦก dแปฑ thแบงu;\nd) Mแป hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt.\n3. ฤรกnh giรก hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt, bao gแปm:\na) Kiแปm tra, ฤรกnh giรก tรญnh hแปฃp lแป cแปงa hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt;\nb) ฤรกnh giรก chi tiแบฟt hแป sฦก ฤแป xuแบฅt vแป kแปน thuแบญt;\nc) Thแบฉm ฤแปnh, phรช duyแปt danh sรกch nhร ฤแบงu tฦฐ ฤรกp แปฉng yรชu cแบงu vแป kแปน thuแบญt.\n4. Mแป vร ฤรกnh giรก hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi, bao gแปm:\na) Mแป hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi;\nb) Kiแปm tra, ฤรกnh giรก tรญnh hแปฃp lแป cแปงa hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi;\nc) ฤรกnh giรก chi tiแบฟt hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi vร xแบฟp hแบกng nhร ฤแบงu tฦฐ.\n5. Trรฌnh, thแบฉm ฤแปnh, phรช duyแปt vร cรดng khai kแบฟt quแบฃ lแปฑa chแปn nhร ฤแบงu tฦฐ.\n6. ฤร m phรกn, hoร n thiแปn, kรฝ kแบฟt hแปฃp ฤแปng, bao gแปm:\na) ฤร m phรกn, hoร n thiแปn hแปฃp ฤแปng;\nb) Kรฝ kแบฟt hแปฃp ฤแปng vร cรดng khai thรดng tin hแปฃp ฤแปng dแปฑ รกn.\n7. Trฦฐแปng hแปฃp dแปฑ รกn แปฉng dแปฅng cรดng nghแป cao, quy trรฌnh lแปฑa chแปn nhร ฤแบงu tฦฐ thแปฑc hiแปn theo quy ฤแปnh tแบกi cรกc khoแบฃn 1, 2, 3, 4 vร 5 ฤiแปu nร y. Nhร ฤแบงu tฦฐ ฤฦฐแปฃc lแปฑa chแปn sแบฝ tiแบฟn hร nh lแบญp bรกo cรกo nghiรชn cแปฉu khแบฃ thi. Trรชn cฦก sแป bรกo cรกo nghiรชn cแปฉu khแบฃ thi ฤฦฐแปฃc duyแปt, cฦก quan nhร nฦฐแปc cรณ thแบฉm quyแปn vร nhร ฤแบงu tฦฐ tiแบฟn hร nh ฤร m phรกn, hoร n thiแปn vร kรฝ kแบฟt hแปฃp ฤแปng. Nแปi dung bรกo cรกo nghiรชn cแปฉu khแบฃ thi vร hแปฃp ฤแปng phแบฃi bแบฃo ฤแบฃm phรน hแปฃp vแปi mแปฅc tiรชu thแปฑc hiแปn dแปฑ รกn, quy mรด ฤแบงu tฦฐ, cรกc tiรชu chรญ cฦก bแบฃn ฤแป xรฉt duyแปt nhร ฤแบงu tฦฐ trรบng thแบงu.',
'ฤiแปu 28 - Luแบญt Giรกo dแปฅc : Cแบฅp hแปc vร ฤแป tuแปi cแปงa giรกo dแปฅc phแป thรดng\n\n1. Cรกc cแบฅp hแปc vร ฤแป tuแปi cแปงa giรกo dแปฅc phแป thรดng ฤฦฐแปฃc quy ฤแปnh nhฦฐ sau:\n\na) Giรกo dแปฅc tiแปu hแปc ฤฦฐแปฃc thแปฑc hiแปn trong 05 nฤm hแปc, tแปซ lแปp mแปt ฤแบฟn hแบฟt lแปp nฤm. Tuแปi cแปงa hแปc sinh vร o hแปc lแปp mแปt lร 06 tuแปi vร ฤฦฐแปฃc tรญnh theo nฤm;\n\nb) Giรกo dแปฅc trung hแปc cฦก sแป ฤฦฐแปฃc thแปฑc hiแปn trong 04 nฤm hแปc, tแปซ lแปp sรกu ฤแบฟn hแบฟt lแปp chรญn. Hแปc sinh vร o hแปc lแปp sรกu phแบฃi hoร n thร nh chฦฐฦกng trรฌnh tiแปu hแปc. Tuแปi cแปงa hแปc sinh vร o hแปc lแปp sรกu lร 11 tuแปi vร ฤฦฐแปฃc tรญnh theo nฤm;\n\nc) Giรกo dแปฅc trung hแปc phแป thรดng ฤฦฐแปฃc thแปฑc hiแปn trong 03 nฤm hแปc, tแปซ lแปp mฦฐแปi ฤแบฟn hแบฟt lแปp mฦฐแปi hai. Hแปc sinh vร o hแปc lแปp mฦฐแปi phแบฃi cรณ bแบฑng tแปt nghiแปp trung hแปc cฦก sแป. Tuแปi cแปงa hแปc sinh vร o hแปc lแปp mฦฐแปi lร 15 tuแปi vร ฤฦฐแปฃc tรญnh theo nฤm.\n\n2. Trฦฐแปng hแปฃp hแปc sinh ฤฦฐแปฃc hแปc vฦฐแปฃt lแปp, hแปc แป ฤแป tuแปi cao hฦกn tuแปi quy ฤแปnh tแบกi khoแบฃn 1 ฤiแปu nร y bao gแปm:\n\na) Hแปc sinh hแปc vฦฐแปฃt lแปp trong trฦฐแปng hแปฃp phรกt triแปn sแปm vแป trรญ tuแป;\n\nb) Hแปc sinh hแปc แป ฤแป tuแปi cao hฦกn tuแปi quy ฤแปnh trong trฦฐแปng hแปฃp hแปc sinh hแปc lฦฐu ban, hแปc sinh แป vรนng cรณ ฤiแปu kiแปn kinh tแบฟ - xรฃ hแปi ฤแบทc biแปt khรณ khฤn, hแปc sinh lร ngฦฐแปi dรขn tแปc thiแปu sแป, hแปc sinh lร ngฦฐแปi khuyแบฟt tแบญt, hแปc sinh kรฉm phรกt triแปn vแป thแป lแปฑc hoแบทc trรญ tuแป, hแปc sinh mแป cรดi khรดng nฦกi nฦฐฦกng tแปฑa, hแปc sinh thuแปc hแป nghรจo, hแปc sinh แป nฦฐแปc ngoร i vแป nฦฐแปc vร trฦฐแปng hแปฃp khรกc theo quy ฤแปnh cแปงa phรกp luแบญt.\n\n3. Giรกo dแปฅc phแป thรดng ฤฦฐแปฃc chia thร nh giai ฤoแบกn giรกo dแปฅc cฦก bแบฃn vร giai ฤoแบกn giรกo dแปฅc ฤแปnh hฦฐแปng nghแป nghiแปp. Giai ฤoแบกn giรกo dแปฅc cฦก bแบฃn gแปm cแบฅp tiแปu hแปc vร cแบฅp trung hแปc cฦก sแป; giai ฤoแบกn giรกo dแปฅc ฤแปnh hฦฐแปng nghแป nghiแปp lร cแบฅp trung hแปc phแป thรดng. Hoฬฃc sinh trong cฦก sแป giรกo dแปฅc nghรชฬ nghiรชฬฃp ฤฦฐฦกฬฃc hoฬฃc khแปi lฦฐแปฃng kiรชฬn thฦฐฬc vฤn hรณa trung hแปc phรดฬ thรดng.\n\n4. Bแป trฦฐแปng Bแป Giรกo dแปฅc vร ฤร o tแบกo quy ฤแปnh viแปc dแบกy vร hแปc tiแบฟng Viแปt cho trแบป em lร ngฦฐแปi dรขn tแปc thiแปu sแป trฦฐแปc khi vร o hแปc lแปp mแปt; viแปc giแบฃng dแบกy khแปi lฦฐแปฃng kiแบฟn thฦฐฬc vฤn hoฬa trung hoฬฃc phรดฬ thรดng trong cฦก sแป giรกo dแปฅc nghแป nghiแปp; cรกc trฦฐแปng hแปฃp quy ฤแปnh tแบกi khoแบฃn 2 ฤiแปu nร y.',
'ฤiแปu 12. Hร nh vi vแบญn chuyแปn trรกi phรฉp hร ng hรณa, tiแปn tแป qua biรชn giแปi ngoร i khu vแปฑc cแปญa khแบฉu nhฦฐng chฦฐa ฤแบฟn mแปฉc truy cแปฉu trรกch nhiแปm hรฌnh sแปฑ : 1. Phแบกt tiแปn tแปซ 1.000.000 ฤแปng ฤแบฟn 5.000.000 ฤแปng ฤแปi vแปi hร nh vi vi phแบกm mร tang vแบญt vi phแบกm cรณ giรก trแป dฦฐแปi 10.000.000 ฤแปng.\n2. Phแบกt tiแปn tแปซ 5.000.000 ฤแปng ฤแบฟn 10.000.000 ฤแปng ฤแปi vแปi mแปt trong nhแปฏng hร nh vi sau:\na) Hร nh vi vi phแบกm mร tang vแบญt vi phแบกm cรณ giรก trแป tแปซ 10.000.000 ฤแปng ฤแบฟn dฦฐแปi 50.000.000 ฤแปng;\nb) Vแบญn chuyแปn trรกi phรฉp qua biรชn giแปi hร ng hรณa lร vฤn hรณa phแบฉm.\n3. Phแบกt tiแปn tแปซ 20.000.000 ฤแปng ฤแบฟn 40.000.000 ฤแปng ฤแปi vแปi hร nh vi vi phแบกm mร tang vแบญt vi phแบกm cรณ giรก trแป tแปซ 50.000.000 ฤแปng ฤแบฟn dฦฐแปi 100.000.000 ฤแปng?\n4. Phแบกt tiแปn tแปซ 40.000.000 ฤแปng ฤแบฟn 50.000.000 ฤแปng ฤแปi vแปi tang vแบญt vi phแบกm cรณ giรก trแป tแปซ 100.000.000 ฤแปng trแป lรชn.\n5. Hรฌnh thแปฉc xแปญ phแบกt bแป sung: Tแปch thu tang vแบญt ฤแปi vแปi hร nh vi vi phแบกm quy ฤแปnh tแบกi ฤiแปu nร y, trแปซ trฦฐแปng hแปฃp ฤรฃ รกp dแปฅng biแปn phรกp khแบฏc phแปฅc hแบญu quแบฃ ฤฦฐแปฃc quy ฤแปnh tแบกi khoแบฃn 6 ฤiแปu nร y.\n6. Biแปn phรกp khแบฏc phแปฅc hแบญu quแบฃ: Buแปc tiรชu hแปงy hร ng hรณa, vแบญt phแบฉm gรขy hแบกi cho sแปฉc khแปe con ngฦฐแปi, vแบญt nuรดi, cรขy trแปng, mรดi trฦฐแปng vร vฤn hรณa phแบฉm cรณ nแปi dung ฤแปc hแบกi ฤแปi vแปi hร nh vi vi phแบกm quy ฤแปnh tแบกi ฤiแปu nร y.',
]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]
ALQAC-ValCEBinaryClassificationEvaluator| Metric | Value |
|---|---|
| accuracy | 0.9393 |
| accuracy_threshold | 0.8522 |
| f1 | 0.7453 |
| f1_threshold | 0.2632 |
| precision | 0.75 |
| recall | 0.7406 |
| average_precision | 0.8104 |
sentence_0, sentence_1, and label| sentence_0 | sentence_1 | label | |
|---|---|---|---|
| type | string | string | float |
| modality | text | text | |
| details |
|
|
|
| sentence_0 | sentence_1 | label |
|---|---|---|
Nhรขn viรชn phแปฅc vแปฅ xe khรกch khรดng hแป trแปฃ, giรบp ฤแปก hร nh khรกch lร ngฦฐแปi cao tuแปi, trแบป em bแป phแบกt bao nhiรชu? |
ฤiแปu 59 - Luแบญt ฤฦฐแปng bแป : Quyแปn vร nghฤฉa vแปฅ cแปงa ngฦฐแปi lรกi xe, nhรขn viรชn phแปฅc vแปฅ trรชn xe รด tรด kinh doanh vแบญn tแบฃi hร nh khรกch |
0.0 |
Trรฌnh tแปฑ vร nแปi dung lร m viแปc cแปงa hแปi ฤแปng tฦฐ vแบฅn tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc chแปง trรฌ dแปฑ รกn Chฦฐฦกng trรฌnh phรกt triแปn tร i sแบฃn trรญ tuแป ฤฦฐแปฃc quy ฤแปnh nhฦฐ thแบฟ nร o? |
ฤiแปu 18. Lแบฅy รฝ kiแบฟn tฦฐ vแบฅn cแปงa chuyรชn gia ฤแปc lแบญp vแป viแปc tuyแปn chแปn, giao trแปฑc tiแบฟp tแป chแปฉc, cรก nhรขn chแปง trรฌ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh : 1. ฤฦกn vแป quแบฃn lรฝ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh trรฌnh lรฃnh ฤแบกo Bแป quyแบฟt ฤแปnh lแปฑa chแปn 03 chuyรชn gia ฤแปc lแบญp ฤแป lแบฅy รฝ kiแบฟn tฦฐ vแบฅn vแป viแปc tuyแปn chแปn tแป chแปฉc, tแป chแปฉc chแปง trรฌ nhiแปm vแปฅ cแปงa Chฦฐฦกng trรฌnh. |
0.0 |
Cรกc bฦฐแปc thแปฑc hiแปn viแปc mแป vร ฤรกnh giรก hแป sฦก ฤแป xuแบฅt vแป tร i chรญnh - thฦฐฦกng mแบกi cแปงa nhร ฤแบงu tฦฐ dแปฑ รกn ฤแบงu tฦฐ cรณ sแปญ dแปฅng ฤแบฅt ฤฦฐแปฃc quy ฤแปnh nhฦฐ thแบฟ nร o? |
ฤiแปu 27. Quy trรฌnh chi tiแบฟt : 1. Chuแบฉn bแป lแปฑa chแปn nhร ฤแบงu tฦฐ, bao gแปm: |
0.0 |
BinaryCrossEntropyLoss with these parameters:{
"activation_fn": "torch.nn.modules.linear.Identity",
"pos_weight": null
}
per_device_train_batch_size: 16per_device_eval_batch_size: 16per_device_train_batch_size: 16num_train_epochs: 3max_steps: -1learning_rate: 5e-05lr_scheduler_type: linearlr_scheduler_kwargs: Nonewarmup_steps: 0optim: adamw_torch_fusedoptim_args: Noneweight_decay: 0.0adam_beta1: 0.9adam_beta2: 0.999adam_epsilon: 1e-08optim_target_modules: Nonegradient_accumulation_steps: 1average_tokens_across_devices: Truemax_grad_norm: 1label_smoothing_factor: 0.0bf16: Falsefp16: Falsebf16_full_eval: Falsefp16_full_eval: Falsetf32: Nonegradient_checkpointing: Falsegradient_checkpointing_kwargs: Nonetorch_compile: Falsetorch_compile_backend: Nonetorch_compile_mode: Noneuse_liger_kernel: Falseliger_kernel_config: Noneuse_cache: Falseneftune_noise_alpha: Nonetorch_empty_cache_steps: Noneauto_find_batch_size: Falselog_on_each_node: Truelogging_nan_inf_filter: Trueinclude_num_input_tokens_seen: nolog_level: passivelog_level_replica: warningdisable_tqdm: Falseproject: huggingfacetrackio_space_id: Nonetrackio_bucket_id: Nonetrackio_static_space_id: Noneper_device_eval_batch_size: 16prediction_loss_only: Trueeval_on_start: Falseeval_do_concat_batches: Trueeval_use_gather_object: Falseeval_accumulation_steps: Noneinclude_for_metrics: []batch_eval_metrics: Falsesave_only_model: Falsesave_on_each_node: Falseenable_jit_checkpoint: Falsepush_to_hub: Falsehub_private_repo: Nonehub_model_id: Nonehub_strategy: every_savehub_always_push: Falsehub_revision: Noneload_best_model_at_end: Falseignore_data_skip: Falserestore_callback_states_from_checkpoint: Falsefull_determinism: Falseseed: 42data_seed: Noneuse_cpu: Falseaccelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}parallelism_config: Nonedataloader_drop_last: Falsedataloader_num_workers: 0dataloader_pin_memory: Truedataloader_persistent_workers: Falsedataloader_prefetch_factor: Noneremove_unused_columns: Truelabel_names: Nonetrain_sampling_strategy: randomlength_column_name: lengthddp_find_unused_parameters: Noneddp_bucket_cap_mb: Noneddp_broadcast_buffers: Falseddp_static_graph: Noneddp_backend: Noneddp_timeout: 1800fsdp: Nonefsdp_config: Nonedeepspeed: Nonedebug: []skip_memory_metrics: Truedo_predict: Falseresume_from_checkpoint: Nonewarmup_ratio: Nonelocal_rank: -1prompts: Nonebatch_sampler: batch_samplermulti_dataset_batch_sampler: proportionalrouter_mapping: {}learning_rate_mapping: {}| Epoch | Step | Training Loss | ALQAC-Val_average_precision |
|---|---|---|---|
| 0.3504 | 500 | 0.3142 | 0.6915 |
| 0.7008 | 1000 | 0.2368 | 0.7599 |
| 1.0 | 1427 | - | 0.7710 |
| 1.0512 | 1500 | 0.2117 | 0.7587 |
| 1.4015 | 2000 | 0.1728 | 0.7901 |
| 1.7519 | 2500 | 0.1587 | 0.7909 |
| 2.0 | 2854 | - | 0.8107 |
| 2.1023 | 3000 | 0.1394 | 0.8014 |
| 2.4527 | 3500 | 0.0981 | 0.8104 |
@inproceedings{reimers-2019-sentence-bert,
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
author = "Reimers, Nils and Gurevych, Iryna",
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
month = "11",
year = "2019",
publisher = "Association for Computational Linguistics",
url = "https://arxiv.org/abs/1908.10084",
}