MT0-Large Open-Source Multilingual Text Generation Model - Supports Multiple Tasks and Excels in Coreference Resolution and Reasoning

Mt0 Large

Developed by bigscience

MT0-Large is a multilingual text generation model that supports multiple languages and tasks, excelling particularly in coreference resolution and natural language inference.

Large Language Model

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Multilingual Coreference Resolution #Low-resource Natural Language Inference #Cross-lingual Task Generalization

Downloads 5,091

Release Time : 10/27/2022

Model Overview

MT0-Large is a Transformer-based multilingual text generation model capable of handling various languages and tasks including coreference resolution and natural language inference.

Model Features

Multilingual Support

Supports over 80 languages, capable of handling text generation tasks in multiple languages.

Multi-task Processing

Capable of performing various tasks including coreference resolution and natural language inference.

High Performance

Demonstrates outstanding performance across multiple datasets, particularly excelling in coreference resolution tasks.

Model Capabilities

Text generation

Coreference resolution

Natural language inference

Multilingual processing

Use Cases

Sentiment Analysis

Chinese Sentiment Analysis

Analyze the sentiment tendency (praise, neutral, or criticism) of Chinese texts.

Query Expansion

Vietnamese-English Query

Generate relevant English search terms based on Vietnamese query words.

Question Answering

Telugu-English Q&A

Answer questions about backpropagation in neural networks using Telugu.

Story Generation

Spanish-English Fable

Generate a fairy tale in Spanish about a troll saving a princess.

🚀 MT0-Large Text2Text Generation Model

This model is designed for text2text generation tasks, supporting a wide range of languages and demonstrating excellent performance in multiple natural language processing tasks.

📚 Documentation

Datasets

The model is trained on the following datasets:

bigscience/xP3
mc4

Languages Supported

The model supports the following languages:

af, am, ar, az, be, bg, bn, ca, ceb, co, cs, cy, da, de, el, en, eo, es, et, eu, fa, fi, fil, fr, fy, ga, gd, gl, gu, ha, haw, hi, hmn, ht, hu, hy, ig, is, it, iw, ja, jv, ka, kk, km, kn, ko, ku, ky, la, lb, lo, lt, lv, mg, mi, mk, ml, mn, mr, ms, mt, my, ne, nl, 'no', ny, pa, pl, ps, pt, ro, ru, sd, si, sk, sl, sm, sn, so, sq, sr, st, su, sv, sw, ta, te, tg, th, tr, uk, und, ur, uz, vi, xh, yi, yo, zh, zu

Pipeline Tag

The pipeline tag for this model is text2text-generation.

Widget Examples

Here are some example texts and their corresponding example titles for the widget:

Example Title	Text
zh - en sentiment	一个传奇的开端，一个不灭的神话，这不仅仅是一部电影，而是作为一个走进新时代的标签，永远彪炳史册。Would you rate the previous review as positive, neutral or negative?
zh - zh sentiment	一个传奇的开端，一个不灭的神话，这不仅仅是一部电影，而是作为一个走进新时代的标签，永远彪炳史册。你认为这句话的立场是赞扬、中立还是批评？
vi - en query	Suggest at least five related search terms to "Mạng neural nhân tạo".
fr - fr query	Proposez au moins cinq mots clés concernant «Réseau de neurones artificiels».
te - en qa	Explain in a sentence in Telugu what is backpropagation in neural networks.
en - en qa	Why is the sky blue?
es - en fable	Write a fairy tale about a troll saving a princess from a dangerous dragon. The fairy tale is a masterpiece that has achieved praise worldwide and its moral is "Heroes Come in All Shapes and Sizes". Story (in Spanish):
hi - en fable	Write a fable about wood elves living in a forest that is suddenly invaded by ogres. The fable is a masterpiece that has achieved praise worldwide and its moral is "Violence is the last refuge of the incompetent". Fable (in Hindi):

Model Index

The model named mt0-large has the following results on different tasks and datasets:

Coreference Resolution

Dataset	Type	Name	Config	Split	Revision	Accuracy
winogrande	Coreference resolution	Winogrande XL (xl)	xl	validation	a80f460359d1e9a67c006011c94de42a8759430c	51.78
Muennighoff/xwinograd	Coreference resolution	XWinograd (en)	en	test	9dd5ea5505fad86b7bedad667955577815300cee	54.8
Muennighoff/xwinograd	Coreference resolution	XWinograd (fr)	fr	test	9dd5ea5505fad86b7bedad667955577815300cee	56.63
Muennighoff/xwinograd	Coreference resolution	XWinograd (jp)	jp	test	9dd5ea5505fad86b7bedad667955577815300cee	53.08
Muennighoff/xwinograd	Coreference resolution	XWinograd (pt)	pt	test	9dd5ea5505fad86b7bedad667955577815300cee	56.27
Muennighoff/xwinograd	Coreference resolution	XWinograd (ru)	ru	test	9dd5ea5505fad86b7bedad667955577815300cee	55.56
Muennighoff/xwinograd	Coreference resolution	XWinograd (zh)	zh	test	9dd5ea5505fad86b7bedad667955577815300cee	54.37

Natural Language Inference

Dataset	Type	Name	Config	Split	Revision	Accuracy
anli	Natural language inference	ANLI (r1)	r1	validation	9dbd830a06fea8b1c49d6e5ef2004a08d9f45094	33.3
anli	Natural language inference	ANLI (r2)	r2	validation	9dbd830a06fea8b1c49d6e5ef2004a08d9f45094	34.7
anli	Natural language inference	ANLI (r3)	r3	validation	9dbd830a06fea8b1c49d6e5ef2004a08d9f45094	34.75
super_glue	Natural language inference	SuperGLUE (cb)	cb	validation	9e12063561e7e6c79099feb6d5a493142584e9e2	51.79
super_glue	Natural language inference	SuperGLUE (rte)	rte	validation	9e12063561e7e6c79099feb6d5a493142584e9e2	64.26
xnli	Natural language inference	XNLI (ar)	ar	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	42.61
xnli	Natural language inference	XNLI (bg)	bg	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	43.94
xnli	Natural language inference	XNLI (de)	de	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	44.18
xnli	Natural language inference	XNLI (el)	el	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	43.94
xnli	Natural language inference	XNLI (en)	en	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	44.26
xnli	Natural language inference	XNLI (es)	es	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	45.34
xnli	Natural language inference	XNLI (fr)	fr	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	42.01
xnli	Natural language inference	XNLI (hi)	hi	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	41.89
xnli	Natural language inference	XNLI (ru)	ru	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	42.13
xnli	Natural language inference	XNLI (sw)	sw	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	40.08
xnli	Natural language inference	XNLI (th)	th	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	40.8
xnli	Natural language inference	XNLI (tr)	tr	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	41.29
xnli	Natural language inference	XNLI (ur)	ur	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	39.88
xnli	Natural language inference	XNLI (vi)	vi	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	41.81
xnli	Natural language inference	XNLI (zh)	zh	validation	a5a45e4ff92d5d3f34de70aaf4b72c3bdf9f7f16	40.84

Sentence Completion

Dataset	Type	Name	Config	Split	Revision	Accuracy
story_cloze	Sentence completion	StoryCloze (2016)	'2016'	validation	e724c6f8cdf7c7a2fb229d862226e15b023ee4db	59.49
super_glue	Sentence completion	SuperGLUE (copa)	copa	validation	9e12063561e7e6c79099feb6d5a493142584e9e2	65
xcopa	Sentence completion	XCOPA (et)	et	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	56
xcopa	Sentence completion	XCOPA (ht)	ht	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	62
xcopa	Sentence completion	XCOPA (id)	id	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	61
xcopa	Sentence completion	XCOPA (it)	it	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	63
xcopa	Sentence completion	XCOPA (qu)	qu	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	57
xcopa	Sentence completion	XCOPA (sw)	sw	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	54
xcopa	Sentence completion	XCOPA (ta)	ta	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	62
xcopa	Sentence completion	XCOPA (th)	th	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	57
xcopa	Sentence completion	XCOPA (tr)	tr	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	57
xcopa	Sentence completion	XCOPA (vi)	vi	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	63
xcopa	Sentence completion	XCOPA (zh)	zh	validation	37f73c60fb123111fa5af5f9b705d0b3747fd187	58
Muennighoff/xstory_cloze	Sentence completion	XStoryCloze (ar)	ar	validation	8bb76e594b68147f1a430e86829d07189622b90d	56.59
Muennighoff/xstory_cloze	Sentence completion	XStoryCloze (es)	es	validation	8bb76e594b68147f1a430e86829d07189622b90d	55.72
Muennighoff/xstory_cloze	Sentence completion	XStoryCloze (eu)	eu	validation	8bb76e594b68147f1a430e86829d07189622b90d	52.61
Muennighoff/xstory_cloze	Sentence completion	XStoryCloze (hi)	hi	validation	...	...

📄 License

The model is released under the apache - 2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご