Dataset Viewer
Auto-converted to Parquet Duplicate
id
stringlengths
17
28
subset
stringclasses
1 value
task
stringclasses
2 values
task_type
stringclasses
2 values
input_type
stringclasses
1 value
lang
stringclasses
37 values
switch
stringclasses
26 values
pair_in_training_table
bool
2 classes
prompt
stringlengths
66
939
input
stringlengths
52
920
english
stringlengths
52
920
source_dataset
stringclasses
5 values
source_register
stringclasses
3 values
reference
stringclasses
0 values
reference_judge
dict
candidates
listlengths
8
8
output
listlengths
8
8
persona:7cad82bb1996
generated
hm
eng_hm
english
kon
fra
true
<hm> <kon>+<fra>: Hey there! Hi there! I love to read books. What are your favorite books? I love to read books too. My favorite books are the ones that make me think about my life and the world around me. I like those kinds of books too. I love to be challenged by the books I read.
Hey there! Hi there! I love to read books. What are your favorite books? I love to read books too. My favorite books are the ones that make me think about my life and the world around me. I like those kinds of books too. I love to be challenged by the books I read.
Hey there! Hi there! I love to read books. What are your favorite books? I love to read books too. My favorite books are the ones that make me think about my life and the world around me. I like those kinds of books too. I love to be challenged by the books I read.
persona
chat
null
null
[ { "k": 0, "greedy": true, "text": "<kon><fra><dev>les livres | Tes | les livres | mes | vie | les | moi | les livres que | suis</dev> Eh, mono kele kuna! Mono ke zola lire les livres. Tes livres préférés kele nki? Mono mosi ke zola lire les livres. Mes livres préférés kele yina ke pusa mono na mes pensé...
[ "Eh, mono kele kuna! Mono ke zola lire les livres. Tes livres préférés kele nki? Mono mosi ke zola lire les livres. Mes livres préférés kele yina ke pusa mono na mes pensées na nge mpi na le monde. Mono mosi ke zola les types de livres yayi. Mono suis intéressée na les livres que je lis.", "Eh, je m'appelle Hi th...
soda:ed166d67c02a
generated
hm
eng_hm
english
ibo
pcm
true
<hm> <ibo>+<pcm>: Oh, of course I do. He was such a big part of my life for so many years. I miss him every day. It's hard to lose somebody you love. Yes, it is. But even though he's gone, I know that our love is still there. It's like a little flame inside my heart that keeps me going.
Oh, of course I do. He was such a big part of my life for so many years. I miss him every day. It's hard to lose somebody you love. Yes, it is. But even though he's gone, I know that our love is still there. It's like a little flame inside my heart that keeps me going.
Oh, of course I do. He was such a big part of my life for so many years. I miss him every day. It's hard to lose somebody you love. Yes, it is. But even though he's gone, I know that our love is still there. It's like a little flame inside my heart that keeps me going.
soda
chat
null
null
[ { "k": 0, "greedy": true, "text": "<ibo><pcm><dev>true | for many years | dey miss am | lose person wey you love | true true | even though | still dey | like small fire | keep | going</dev> O, true, mụ onwe m mụ onwe m. He was such a big part of my life for many years. I dey miss am every day, truly. Ọ ...
[ "O, true, mụ onwe m mụ onwe m. He was such a big part of my life for many years. I dey miss am every day, truly. Ọ siri ike to lose person wey you love, true true. Eziokwu, ọ bụ eziokwu. Mana even though ọ lara, m ma na our love still dey. Ọ dị like small fire n'ime obi m wey dey keep me going, i nọ na-aga.", "Ah...
soda:d9ffa2a0238c
generated
hm
eng_hm
english
nde
eng
true
<hm> <nde>+<eng>: Yeah, I was just trying it out. But I know it's bad for me, so I want to quit now. Aura, you know how I feel about smoking. It's really harmful to your health and I don't want you doing it anymore. Yeah, I know. I'm sorry for doing it in the first place. It won't happen again. Okay, well as long as y...
Yeah, I was just trying it out. But I know it's bad for me, so I want to quit now. Aura, you know how I feel about smoking. It's really harmful to your health and I don't want you doing it anymore. Yeah, I know. I'm sorry for doing it in the first place. It won't happen again. Okay, well as long as you're committed to ...
Yeah, I was just trying it out. But I know it's bad for me, so I want to quit now. Aura, you know how I feel about smoking. It's really harmful to your health and I don't want you doing it anymore. Yeah, I know. I'm sorry for doing it in the first place. It won't happen again. Okay, well as long as you're committed to ...
soda
chat
null
null
[ { "k": 0, "greedy": true, "text": "<nde><eng><dev>was just trying out | know | bad for | want | quit | you know how | really harmful | and | don want you | doing this anymore</dev> Yebo, I was just trying out it. Kodwa I know it's bad for me, so I want to quit manje. Aura, you know how ngizizwa ngomhlab...
[ "Yebo, I was just trying out it. Kodwa I know it's bad for me, so I want to quit manje. Aura, you know how ngizizwa ngomhlaba wokutshaya. It's really harmful empilweni yakho and I don't want you to do this anymore. Yebo, I know. I'm sorry ngokukwenza lokhu at first. It won't happen again. Okay, provided ukuthi you'...
persona:5bdb224d56f2
generated
hm
eng_hm
english
ful
eng
true
<hm> <ful>+<eng>: That's a long time! I'm sure you've learned a lot. I have! I've learned a lot about the human body, and I've also learned a lot about how to care for people. That's great! I'm sure your patients are lucky to have you. I hope so! I try my best to help them. I'm sure you do. So, what made you want to b...
That's a long time! I'm sure you've learned a lot. I have! I've learned a lot about the human body, and I've also learned a lot about how to care for people. That's great! I'm sure your patients are lucky to have you. I hope so! I try my best to help them. I'm sure you do. So, what made you want to be a nurse? I always...
That's a long time! I'm sure you've learned a lot. I have! I've learned a lot about the human body, and I've also learned a lot about how to care for people. That's great! I'm sure your patients are lucky to have you. I hope so! I try my best to help them. I'm sure you do. So, what made you want to be a nurse? I always...
persona
chat
null
null
[ { "k": 0, "greedy": true, "text": "<ful><eng><dev>Long time | sure | have learned | lot about the human body | how | care for people | That | great | your patients | hope so</dev> Long time woni! I'm sure a anndii fof. I have learned a lot about the human body, e how to care for people. That is great! M...
[ "Long time woni! I'm sure a anndii fof. I have learned a lot about the human body, e how to care for people. That is great! Mi mawnuni wonde your patients ɓe njaaraama ngam a woni e ɓe. I hope so! Mi ɗon try my best ngam mballude ɓe. I'm sure a ɗon yidde ɗum. So, ko honno a yidde to be a nurse? I've always wanted t...
writingprompts:7587b337d54c
generated
hm
eng_hm
english
lin
kon
false
<hm> <lin>+<kon>: They realize that such a place has no value. Joy without effort is hollow, it's exactly the drama and the suffering of human existence that makes it have value. They'd rather die and face oblivion but have a chance to change the course of history than to exist unchanging forever.
They realize that such a place has no value. Joy without effort is hollow, it's exactly the drama and the suffering of human existence that makes it have value. They'd rather die and face oblivion but have a chance to change the course of history than to exist unchanging forever.
They realize that such a place has no value. Joy without effort is hollow, it's exactly the drama and the suffering of human existence that makes it have value. They'd rather die and face oblivion but have a chance to change the course of history than to exist unchanging forever.
writingprompts
story
null
null
[ { "k": 0, "greedy": true, "text": "<lin><ibo><dev>nothing | effort | life | change | forever</dev> Baya bamonaki ete esika oyo eza na valeur te, nothing. Esengo soki eza na effort te eza pamba, eza drama mpe souffrance ya life oyo esalaka ete eza na valeur. Balingi kokufa mpe kokanisa ete bamonaki eloko...
[ "Baya bamonaki ete esika oyo eza na valeur te, nothing. Esengo soki eza na effort te eza pamba, eza drama mpe souffrance ya life oyo esalaka ete eza na valeur. Balingi kokufa mpe kokanisa ete bamonaki eloko te kasi bazo zala na chance ya change l'histoire na esika ya kozala kaka forever.", "Baita ete ete ete esik...
soda:95a05c379c66
generated
hm
eng_hm
english
amh
ara
false
<hm> <amh>+<ara>: That sounds like a good day. Did anything else happen? Nope, not really. Just a normal day.
That sounds like a good day. Did anything else happen? Nope, not really. Just a normal day.
That sounds like a good day. Did anything else happen? Nope, not really. Just a normal day.
soda
chat
null
null
[ { "k": 0, "greedy": true, "text": "<amh><ara><dev>wallahi | something else happened | not | just | normal day</dev> ያ መልካም ቀን ይመስላል፣ wallahi! ሌላ something else happened? አላደረገም፣ not just it. It is just a normal day.", "finished": true, "devices": [ "wallahi", "something else happened...
[ "ያ መልካም ቀን ይመስላል፣ wallahi! ሌላ something else happened? አላደረገም፣ not just it. It is just a normal day.", "ያ መልካም ቀን ይመስላል፣ yaani! ሌላ የሆነ ነገር ደግሞ happened? በቃ አላደረገም፣ wallahi፣ maybe የሆነ normal day ነበር፣ please!", "ያ በጣም ጥሩ ቀን ይመስላል። ሌላ activity ደገመዎት ወይስ? አላደረግኩም፣ wallahi አይደለም። just a normal day ነበር።", "ያ ጥሩ ቀን ...
soda:4599d8ec5ac8
generated
hm
eng_hm
english
swh
sheng
true
<hm> <swh>+<sheng>: Hey, how was your day? It was good. I had a lot of fun. That's great. I'm glad you had a good time. Yeah, me too. I'm really tired though. Yeah, it sounds like it was a long day for you. Do you want to take a shower and then we can go to bed?
Hey, how was your day? It was good. I had a lot of fun. That's great. I'm glad you had a good time. Yeah, me too. I'm really tired though. Yeah, it sounds like it was a long day for you. Do you want to take a shower and then we can go to bed?
Hey, how was your day? It was good. I had a lot of fun. That's great. I'm glad you had a good time. Yeah, me too. I'm really tired though. Yeah, it sounds like it was a long day for you. Do you want to take a shower and then we can go to bed?
soda
chat
null
null
[ { "k": 0, "greedy": true, "text": "<swh><sheng><dev>ume stay | wallahi | yaani | time | seriously | but | out | long day | you know | sleep</dev> Hujambo, ume-stay vipi siku yako? Ilikuwa poa, wallahi. Nilikuwa na furaha sana, yaani. Ni poa. Nimefurahi kwamba uli-enjoy time yako. Eeh, mimi pia. Nimechok...
[ "Hujambo, ume-stay vipi siku yako? Ilikuwa poa, wallahi. Nilikuwa na furaha sana, yaani. Ni poa. Nimefurahi kwamba uli-enjoy time yako. Eeh, mimi pia. Nimechoka seriously, but it's out. Eeh, inaonekana ulikuwa na long day kwako. Unataka ku-shower kwanza kisha tunaweza ku-go to sleep?", "Shikamoo mpenzi, mbona ime...
writingprompts:7662bad72b28
generated
hm
eng_hm
english
kik
swh
true
<hm> <kik>+<swh>: Before I was thirty five the house was awash with the laughter and play of children. First the first time in my life I felt a sort of peace. & nbsp; I learned what true patience was at the hands of Prince Torsten. White streaked my red hair and my hands sometimes trembled. My torn heart and lung had ...
Before I was thirty five the house was awash with the laughter and play of children. First the first time in my life I felt a sort of peace. & nbsp; I learned what true patience was at the hands of Prince Torsten. White streaked my red hair and my hands sometimes trembled. My torn heart and lung had not agreed with the...
Before I was thirty five the house was awash with the laughter and play of children. First the first time in my life I felt a sort of peace. & nbsp; I learned what true patience was at the hands of Prince Torsten. White streaked my red hair and my hands sometimes trembled. My torn heart and lung had not agreed with the...
writingprompts
story
null
null
[ { "k": 0, "greedy": true, "text": "<kik><swh><dev>yaani | amani | wakati | nguvu | sana</dev> Mbere ya kũnyia mĩaka mĩrongo ĩtatũ na ithano, nyũmba ĩyo nĩ yaigĩte na kũtheka na kucheza kwa ciana, yaani chaos kabisa. Kĩambĩrĩria-inĩ, thaa cia kĩambĩrĩria muoyo wakwa nĩ ndaigĩire amani ya ngoro. Nĩ ndaigĩ...
[ "Mbere ya kũnyia mĩaka mĩrongo ĩtatũ na ithano, nyũmba ĩyo nĩ yaigĩte na kũtheka na kucheza kwa ciana, yaani chaos kabisa. Kĩambĩrĩria-inĩ, thaa cia kĩambĩrĩria muoyo wakwa nĩ ndaigĩire amani ya ngoro. Nĩ ndaigĩire ũrĩa subira ya ma ĩrĩ thĩinĩ wa moko ma Prince Torsten. Rangi ya rangi ya rangi ya rangi ya rangi ya ...
fineweb_edu:48170ff698bd
generated
hm
eng_hm
english
ewe
fra
true
<hm> <ewe>+<fra>: It turns out that there is no solid consensus on the genuine origins of the holiday we now call Thanksgiving, unless you count that Abraham Lincoln officially designated it a national holiday in 1863 "to set apart and observe the last Thursday of November next, as a day of Thanksgiving and Praise to ...
It turns out that there is no solid consensus on the genuine origins of the holiday we now call Thanksgiving, unless you count that Abraham Lincoln officially designated it a national holiday in 1863 "to set apart and observe the last Thursday of November next, as a day of Thanksgiving and Praise to our beneficent Fath...
It turns out that there is no solid consensus on the genuine origins of the holiday we now call Thanksgiving, unless you count that Abraham Lincoln officially designated it a national holiday in 1863 "to set apart and observe the last Thursday of November next, as a day of Thanksgiving and Praise to our beneficent Fath...
fineweb_edu
complex
null
null
[ { "k": 0, "greedy": true, "text": "<ewe><fra><dev>est pas | nom | pour | dernier | comme | jour | pour | bien | est passé | pour les</dev> Eyae nye nu si fia be c'est pas de consensus ŋu be thanksgiving la tso nu ka me, unless you count be Abraham Lincoln wòna le nom sia le 1863 be wònye national holida...
[ "Eyae nye nu si fia be c'est pas de consensus ŋu be thanksgiving la tso nu ka me, unless you count be Abraham Lincoln wòna le nom sia le 1863 be wònye national holiday pour célébrer le dernier Thursday de novembre, comme un jour de Thanksgiving pour la miséricorde de notre père qui vit dans le ciel. C'est passé en ...
writingprompts:76a4f90a4e25
generated
hm
eng_hm
english
bam
wol
false
<hm> <bam>+<wol>: Reaching out under the pot I withdrew the key. Once I'd opened the door, the smell became worse. The hallway was filled with soggy newspapers and large stacks of papers. Envelopes, more newspapers, and almost a swimming pool of coupons and magazines. I had to wade through it and try and keep myself f...
Reaching out under the pot I withdrew the key. Once I'd opened the door, the smell became worse. The hallway was filled with soggy newspapers and large stacks of papers. Envelopes, more newspapers, and almost a swimming pool of coupons and magazines. I had to wade through it and try and keep myself from gagging from th...
Reaching out under the pot I withdrew the key. Once I'd opened the door, the smell became worse. The hallway was filled with soggy newspapers and large stacks of papers. Envelopes, more newspapers, and almost a swimming pool of coupons and magazines. I had to wade through it and try and keep myself from gagging from th...
writingprompts
story
null
null
[ { "k": 0, "greedy": true, "text": "<bam><wol><dev>est incroyable | comme | était une blague | problème eau | solve</dev> N bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n b...
[ "N bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ sɔrɔ n bɛ ...
persona:ab7e8ad548f1
generated
hm
eng_hm
english
zul
afr
false
<hm> <zul>+<afr>: I'm taking a marketing class, and I'm really enjoying it! Marketing is fun! I graduated with a business degree, and marketing is all that I really enjoyed. That's awesome! What are you doing with your business degree?
I'm taking a marketing class, and I'm really enjoying it! Marketing is fun! I graduated with a business degree, and marketing is all that I really enjoyed. That's awesome! What are you doing with your business degree?
I'm taking a marketing class, and I'm really enjoying it! Marketing is fun! I graduated with a business degree, and marketing is all that I really enjoyed. That's awesome! What are you doing with your business degree?
persona
chat
null
null
[ { "k": 0, "greedy": true, "text": "<zul><afr><dev>die | regtig | business | wat | regtig | wat jou business degree</dev> Ngithatha ʼn marketing klas, futhi ngiyakujabulela kakhulu, die marketing is regtig lekker! Marketing uyajabulisa! Ngiphothule ngesitifiketi se-business, futhi marketing yikho okungith...
[ "Ngithatha ʼn marketing klas, futhi ngiyakujabulela kakhulu, die marketing is regtig lekker! Marketing uyajabulisa! Ngiphothule ngesitifiketi se-business, futhi marketing yikho okungithandile okukhulu, wat's wat regtig. Lokhu kuhle kakhulu! Kwenzani ngale business degree yakho, mfwethu?", "Ngithatha i-marketing cl...
writingprompts:9efe3839b252
generated
hm
eng_hm
english
tir
amh
true
<hm> <tir>+<amh>: Are you ready to go?'' Derek was not happy. He was disheveled, his hair much less in the latest style than a general suggestion of what combed hair may look like. His clothing was less carefully crafted than just quickly grabbed out of his closet, resulting in a bland combination of a white shirt and...
Are you ready to go?'' Derek was not happy. He was disheveled, his hair much less in the latest style than a general suggestion of what combed hair may look like. His clothing was less carefully crafted than just quickly grabbed out of his closet, resulting in a bland combination of a white shirt and the motto `` Other...
Are you ready to go?'' Derek was not happy. He was disheveled, his hair much less in the latest style than a general suggestion of what combed hair may look like. His clothing was less carefully crafted than just quickly grabbed out of his closet, resulting in a bland combination of a white shirt and the motto `` Other...
writingprompts
story
null
null
[ { "k": 0, "greedy": true, "text": "<tir><amh><dev>በጣም | style | የለም | correct | fight</dev> ክትከይድ ተዳልየካዶ፧ ዴሪክ በጣም ኣይሕጉሰን። ጸጉሩ ኣብቲ አዲሱ style ካብቲ ዝተዛዘመ style ዝተሓተተ እዩ፡ በጣም ዝተዘበረቐ ይመስል ነበር። ልብሱ ብቕልጡፍ ካብቲ ከም ዝተወሰደ ዝተገብረ ኣይኮነን፡ ስለዚ ብቀሊል መንገዲ ሓደ ነጭ ሸሚዝን ምስቲ የለም፡ 'ሌላው ስህተት ነው፣ አንተ ትክክል ነህ' ዝብል ቋሚ ቋንቋን ብቀሊል ቡኒ ...
[ "ክትከይድ ተዳልየካዶ፧ ዴሪክ በጣም ኣይሕጉሰን። ጸጉሩ ኣብቲ አዲሱ style ካብቲ ዝተዛዘመ style ዝተሓተተ እዩ፡ በጣም ዝተዘበረቐ ይመስል ነበር። ልብሱ ብቕልጡፍ ካብቲ ከም ዝተወሰደ ዝተገብረ ኣይኮነን፡ ስለዚ ብቀሊል መንገዲ ሓደ ነጭ ሸሚዝን ምስቲ የለም፡ 'ሌላው ስህተት ነው፣ አንተ ትክክል ነህ' ዝብል ቋሚ ቋንቋን ብቀሊል ቡኒ ዝተለበሰ ነጭ ሸሚዝን ተሓዊሱ ነበረ። እዚ ድማ ተረድኣይ እዩ፤ ኩሉ ሰብ ብሓደ ለይቲ ብዝተሓተተ መንገዲ ኣብ ከምዚ ዝተሓተተ ጉዳይ ብሓባር ክዛረብ ኣይክእልን እዩ፡...
writingprompts:e80780015cb0
generated
hm
eng_hm
english
nso
eng
true
<hm> <nso>+<eng>: An elderly lady had fallen carrying in groceries. This Tower acts as a community, and as such we are required to maintain its cleanliness as a family.
An elderly lady had fallen carrying in groceries. This Tower acts as a community, and as such we are required to maintain its cleanliness as a family.
An elderly lady had fallen carrying in groceries. This Tower acts as a community, and as such we are required to maintain its cleanliness as a family.
writingprompts
story
null
null
[ { "k": 0, "greedy": true, "text": "<nso><eng><dev> carrying groceries | This Tower | community | and therefore | maintain its cleanliness | family</dev> Go na le mma yo mogolo yo a welego a carrying groceries. This Tower e šoma bjalo ka community, and therefore re swanetše go maintain its cleanliness bj...
[ "Go na le mma yo mogolo yo a welego a carrying groceries. This Tower e šoma bjalo ka community, and therefore re swanetše go maintain its cleanliness bjalo ka family.", "Go na le mohlokomedi yo a welego carrying groceries. This Tower e šoma bjalo ka community, and therefore re swanetše go maintain its cleanliness...
writingprompts:22957d02c8b7
generated
hm
eng_hm
english
tso
zul
true
<hm> <tso>+<zul>: The mountain men came next, and when ruling, showed very little tolerance to any criminal, but especially robbers or burglars. The fools had a boy ruler.
The mountain men came next, and when ruling, showed very little tolerance to any criminal, but especially robbers or burglars. The fools had a boy ruler.
The mountain men came next, and when ruling, showed very little tolerance to any criminal, but especially robbers or burglars. The fools had a boy ruler.
writingprompts
story
null
null
[ { "k": 0, "greedy": true, "text": "<tso><zul><dev>ama | ama | uyezwa</dev> Vamanorhi va le xitsongweni va fike endzhaku, kutani loko va fika va hlawula, va kombise ukungavumeli okuncane eka ama-criminal, kambe ngopfu ama-robbers kumbe ama-burglars. Lava va nga ri na vutlhari va va na umbusi wumfana, uye...
[ "Vamanorhi va le xitsongweni va fike endzhaku, kutani loko va fika va hlawula, va kombise ukungavumeli okuncane eka ama-criminal, kambe ngopfu ama-robbers kumbe ama-burglars. Lava va nga ri na vutlhari va va na umbusi wumfana, uyezwa?", "Vaman’wana va tindhawu va fika endzhaku, futhi ngenkathi ba-decida, va kombi...
End of preview. Expand in Data Studio

African Code-Switch Benchmark

A dataset for code-switching and idiom-rich writing in 37 African languages. Every row is a prompt (English text, or text in the base language) with candidate rewrites and a cleaned output column you can use directly. The test split is meant to become the benchmark for measuring progress on African-language code-switching, and the train split is for building better models. Nobody has checked the candidates with native speakers yet.

Two rewrite tasks:

  • hm (human-style mixing) writes the text in a base language with a second language mixed in, the way people really talk. Example: Yoruba with English words in the middle of a sentence.
  • ood rewrites the text in lively, everyday language full of idiomatic expressions (including sayings and comparisons) just like a native speaker would.

Train and test

Split Subset English texts or items Rows hm rows ood rows Candidates
train generated 8,919 17,794 8,875 8,919 142,352
test generated 11,100 22,147 11,048 11,099 177,176
test curated 6,230 6,230 2,380 3,850 18,675

The subset column tells the two kinds of rows apart.

test is the benchmark. It has two parts:

  • generated: 11,100 English texts, exactly 300 per language (37 languages), so no language counts more than another when you average scores. Every text has an hm and an ood row with eight candidates written by our 1B model (AletheiaAING/gemma-3-1b-csw-best). The language pairs and text types follow the same mix as train (under 1% difference on source mix, under 0.5% on language pairs), and all 81 base-switch pairs appear, with 66 to 166 hm rows each.
  • curated: 6,230 held-out items in 34 languages and five tasks (eng_hm, m_hm, eng_ood, m_ood, hm_ood, explained under Columns). Each item has a curated reference and the outputs of three csw systems, each with four scores: ev, reward, chrf and judge (how they are derived is explained under How the scores were derived). The reference has judge grades too.

Report scores per language as well as the average.

train holds the remaining 8,919 generated texts.

How the split was made. The split is by English text, so the hm and ood rows of one text always stay together. Near-duplicate texts (for example conversations that share most of their sentences) are kept on the same side, so the model never sees a near copy of a test text during training. Texts were sampled per language to match the overall mix of switch-language pairs and text types.

What you can use it for

  • Turn single-language chat or instruction data into code-switched user queries or assistant replies.
  • Build text for ASR language models, TTS scripts and word lists with switched phrases (have a native speaker check it first).
  • Test chatbots, translation and language-ID systems on code-switched inputs.
  • Train and evaluate code-switch detectors with the labelled switched phrases (the devices field).
  • Measure progress on African-language code-switching with the test split.
  • Study how to pick good code-switched text from several candidates.

Columns

Each row is one input text with one task. Columns that don't apply to a row are null.

Row columns

Column Meaning
id id of the item. For generated rows the hm and ood rows of one text share it.
subset generated (English texts with candidates written by our 1B model) or curated (the held-out test items with a reference, see Train and test).
task what is asked: hm (code-switching) or ood (idiom-rich rewrite).
task_type the full task name: eng_hm, m_hm, eng_ood, m_ood or hm_ood. eng_ means the input is English, m_ means it is already in the base language, and hm_ood means it is already code-switched and you want the idiom-rich version. generated rows are all eng_hm or eng_ood.
input_type the language of the input: english, base_language or code_switched.
lang base language code, for example yor.
switch switch language code for hm tasks and hm_ood, otherwise null.
pair_in_training_table false for the 3,348 hm rows whose base-switch pair is outside the 67 pairs our 1B model was trained on (see the pair table), true otherwise. Use it to keep only the supported pairs.
prompt the exact prompt given to the model, for example <hm> <kon>+<fra>: Hey there! ....
input the text the model had to rewrite (also inside prompt).
english the English source text, null when the input is not English.
source_dataset, source_register where the English text comes from and its register (chat, story, complex); curated rows have source_dataset = curated.
reference the curated reference rewrite, with tags and device block removed. curated rows only.
reference_judge judge grades of the reference (see below). curated rows only.
candidates the candidate outputs, one entry each (columns below).
output the cleaned text of each candidate, in the same order as candidates: language tags and the <dev>...</dev> block removed, so you can use it directly.

Columns inside each entry of candidates

Column Meaning
k candidate number. For generated rows 0 is the greedy output and 1 to 7 are samples; for curated rows it numbers the three systems.
greedy true if this is a greedy (temperature 0) output.
text the raw output: language tags, <dev> block and text.
finished true if the model stopped by itself and was not cut off at the length limit (null for curated rows).
devices the switched phrases (hm) or idioms (ood) the model listed, as they appear in the text.
final the text without tags and device block.
format_ok true if the output has the language tags, one device block and a non-empty text.
ev learned quality score, 1 to 10, null when missing.
system curated rows only: sft (supervised), rl_best or rl_last (reinforcement-learned, best and last checkpoint).
reward curated rows only: the rule-based reward (can be negative; higher is better).
chrf curated rows only: chrF++ of final against reference, 0 to 100.
judge curated rows only: judge grades of this output, a record with the criteria below.

Use output[i] for the text of candidates[i].

How the scores were derived

None of these scores comes from a person.

  • ev: a learned quality model (a multilingual encoder we fine-tuned on rated code-switched outputs). It reads the prompt and an output and predicts a 1 to 10 quality score. It is only computed for outputs with the right format. It likes short, simple code-switching, so don't use it as the only filter.
  • reward: the reward used to train the two reinforcement-learned systems. It is the learned quality score combined with a format check and rule penalties for things like unused devices, the wrong language, repetition, wrong length, English leaking in and copying the input.
  • chrf: sentence-level chrF++ (word order 2, sacreBLEU) between the cleaned output and the cleaned reference. It measures closeness to the reference, not quality on its own.
  • judge, reference_judge: grades from 1 to 5 (mostly whole numbers) given by a large-language-model judge with a fixed rubric. The judge saw the input and the output but not which system wrote it. Anchors: 5 means a native speaker would change nothing, 1 means unusable. hm rows are graded on meaning, base_language, switch_language, language_compliance, switching_naturalness and overall; ood rows on meaning, language_correctness, idiom_use, naturalness, language_compliance and overall. A criterion the judge could not assess is null, and criteria that don't apply to the task are null. A few outputs could not be parsed and have no grades.

Example row, shortened:

{"id": "persona:7cad82bb1996", "task": "hm", "lang": "kon", "switch": "fra", "pair_in_training_table": true,
 "prompt": "<hm> <kon>+<fra>:  Hey there! Hi there! I love to read books. ...", "english": "Hey there! Hi there! I love to read books. ...",
 "source_dataset": "persona", "source_register": "chat",
 "candidates": [{"k": 0, "greedy": true, "text": "<kon><fra><dev>les livres | Tes | ...</dev> Eh, mono kele kuna! ...", "finished": true,
                 "devices": ["les livres", "Tes"], "final": "Eh, mono kele kuna! ...", "format_ok": true, "ev": 3.1}, "... 7 more ..."],
 "output": ["Eh, mono kele kuna! ...", "... 7 more ..."]}

How to use it

from datasets import load_dataset

ds = load_dataset("AletheiaAING/african-code-switch-benchmark")          # splits: train, test
row = ds["test"][0]
print(row["prompt"]); print(row["output"][0])    # the greedy candidate as clean text

To pick the highest-scoring valid candidate of every prompt:

best = []
for r in ds["test"]:
    ok = [(c["ev"], o) for c, o in zip(r["candidates"], r["output"]) if c["format_ok"] and "ev" in c]
    if ok: best.append((max(ok)[0], r["lang"], r["task"], max(ok)[1]))

Statistics

Rows are (English text, task) pairs; every generated text has an hm and an ood row (a few have only one). The tables below describe the generated subset, except the curated tables at the end.

Overall. 20,019 generated English texts, 39,941 rows, 319,528 candidates, 37 base languages, 81 base-switch pairs, plus 6,230 curated test items. The mean quality score ev of a valid candidate is 4.12 (train 4.12, test 4.12); ev comes from a learned model, not from people.

English source texts. The texts keep the licences of their sources (see License). Rows by source:

source_dataset train test
soda 5744 7238
writingprompts 5296 6572
persona 3682 4497
fineweb_edu 1876 2266
gutenberg 1196 1574

By base language. test has the same number of texts for every language.

Language Code train texts test texts train rows test rows
Afrikaans afr 240 300 480 600
Amharic amh 240 300 480 600
Bambara bam 241 300 482 600
Cameroonian Pidgin wes 241 300 446 557
Chichewa nya 239 300 478 600
Ewe ewe 241 300 482 599
Fon fon 241 300 482 600
Fulah ful 241 300 482 600
Ghanaian Pidgin gpe 242 300 476 591
Hausa hau 241 300 482 600
Igbo ibo 241 300 482 600
Kikuyu kik 241 300 482 600
Kinyarwanda kin 242 300 484 600
Kirundi run 241 300 482 600
Kongo kon 241 300 482 600
Lingala lin 242 300 484 600
Luganda lug 242 300 484 600
Luo luo 241 300 482 600
Malagasy mlg 242 300 484 600
Nigerian Pidgin pcm 242 300 484 600
North Ndebele nde 241 300 482 600
Oromo orm 240 300 480 600
Sepedi nso 242 300 484 600
Sesotho sot 241 300 482 600
Setswana tsn 241 300 482 600
Shona sna 241 300 482 600
Somali som 241 300 482 600
Swahili swh 242 300 484 600
Swati ssw 239 300 478 600
Tigrinya tir 241 300 482 600
Tsonga tso 242 300 484 600
Tumbuka tum 241 300 482 600
Twi twi 241 300 482 600
Wolof wol 241 300 482 600
Xhosa xho 242 300 484 600
Yoruba yor 240 300 480 600
Zulu zul 241 300 482 600

By base-switch language pair (hm rows). The last column tells whether our 1B model was trained on this pair. The 3,348 rows of the pairs marked "no" come from an earlier, wider pair list; the model's candidates for them are less reliable, so filter on pair_in_training_table if you need only the supported pairs.

Base Code Switch Code train test total in the 1B training table
Afrikaans afr English eng 117 146 263 yes
Afrikaans afr Zulu zul 123 154 277 no
Amharic amh Arabic ara 131 164 295 no
Amharic amh English eng 109 136 245 yes
Bambara bam French fra 115 143 258 yes
Bambara bam Wolof wol 126 157 283 no
Cameroonian Pidgin wes Duala dua 53 68 121 no
Cameroonian Pidgin wes Ewondo ewo 79 97 176 no
Cameroonian Pidgin wes French fra 73 92 165 yes
Chichewa nya English eng 111 141 252 yes
Chichewa nya Tumbuka tum 128 159 287 yes
Ewe ewe English eng 117 147 264 yes
Ewe ewe French fra 124 153 277 yes
Fon fon French fra 128 160 288 yes
Fon fon Yoruba yor 113 140 253 yes
Fulah ful English eng 121 151 272 yes
Fulah ful French fra 120 149 269 yes
Ghanaian Pidgin gpe Ewe ewe 78 96 174 yes
Ghanaian Pidgin gpe Ga gaa 73 92 165 yes
Ghanaian Pidgin gpe Twi twi 83 103 186 yes
Hausa hau Arabic ara 119 150 269 yes
Hausa hau English eng 122 150 272 yes
Igbo ibo English eng 112 139 251 yes
Igbo ibo Nigerian Pidgin pcm 129 161 290 yes
Kikuyu kik English eng 116 143 259 yes
Kikuyu kik Swahili swh 125 157 282 yes
Kinyarwanda kin English eng 83 103 186 yes
Kinyarwanda kin French fra 78 96 174 yes
Kinyarwanda kin Swahili swh 81 101 182 yes
Kirundi run French fra 128 159 287 yes
Kirundi run Swahili swh 113 141 254 yes
Kongo kon French fra 117 145 262 yes
Kongo kon Lingala lin 124 155 279 yes
Lingala lin French fra 108 134 242 yes
Lingala lin Kongo kon 134 166 300 no
Luganda lug English eng 114 142 256 yes
Luganda lug Swahili swh 128 158 286 yes
Luo luo English eng 117 146 263 yes
Luo luo Swahili swh 124 154 278 yes
Malagasy mlg English eng 120 149 269 no
Malagasy mlg French fra 122 151 273 yes
Nigerian Pidgin pcm Hausa hau 83 103 186 yes
Nigerian Pidgin pcm Igbo ibo 81 102 183 yes
Nigerian Pidgin pcm Yoruba yor 78 95 173 yes
North Ndebele nde English eng 125 155 280 yes
North Ndebele nde Shona sna 116 145 261 yes
Oromo orm Amharic amh 117 145 262 yes
Oromo orm English eng 123 155 278 yes
Sepedi nso Afrikaans afr 124 154 278 no
Sepedi nso English eng 118 146 264 yes
Sesotho sot Afrikaans afr 123 152 275 yes
Sesotho sot English eng 118 148 266 yes
Setswana tsn Afrikaans afr 115 143 258 yes
Setswana tsn English eng 126 157 283 yes
Shona sna English eng 131 162 293 yes
Shona sna North Ndebele nde 110 138 248 no
Somali som Amharic amh 54 66 120 no
Somali som Arabic ara 65 81 146 yes
Somali som English eng 58 74 132 yes
Somali som Italian ita 64 79 143 no
Swahili swh Arabic ara 73 91 164 yes
Swahili swh English eng 86 108 194 yes
Swahili swh Sheng sheng 83 101 184 yes
Swati ssw English eng 116 145 261 yes
Swati ssw Zulu zul 123 155 278 yes
Tigrinya tir Amharic amh 118 147 265 yes
Tigrinya tir English eng 123 153 276 yes
Tsonga tso English eng 122 151 273 yes
Tsonga tso Zulu zul 120 149 269 yes
Tumbuka tum Chichewa nya 112 139 251 yes
Tumbuka tum English eng 129 161 290 yes
Twi twi English eng 126 155 281 yes
Twi twi Ghanaian Pidgin gpe 115 145 260 yes
Wolof wol Arabic ara 127 158 285 no
Wolof wol French fra 114 142 256 yes
Xhosa xho Afrikaans afr 122 151 273 no
Xhosa xho English eng 120 149 269 yes
Yoruba yor English eng 115 143 258 yes
Yoruba yor Nigerian Pidgin pcm 125 157 282 yes
Zulu zul Afrikaans afr 125 155 280 no
Zulu zul English eng 116 145 261 yes

Curated test items by language and task (test, subset = curated).

Language Code eng_hm m_hm eng_ood m_ood hm_ood total
Afrikaans afr 100 100 99 99 99 497
Amharic amh 99 99 97 97 97 489
Bambara bam 2 2 2 2 2 10
Cameroonian Pidgin wes 1 1 72 72 1 147
Chichewa nya 52 52 53 53 51 261
Fon fon 8 8 10 10 8 44
Ghanaian Pidgin gpe 2 2 71 71 1 147
Hausa hau 74 74 99 99 74 420
Igbo ibo 21 21 25 25 21 113
Kikuyu kik 30 30 25 25 25 135
Kinyarwanda kin 9 9 9 9 9 45
Kirundi run 5 5 5 5 5 25
Kongo kon 4 4 6 6 3 23
Lingala lin 5 5 6 6 5 27
Luo luo 11 11 10 10 10 52
Malagasy mlg 84 84 91 91 79 429
Nigerian Pidgin pcm 74 74 90 90 74 402
North Ndebele nde 99 99 74 74 73 419
Oromo orm 27 27 27 27 27 135
Sepedi nso 37 37 37 37 37 185
Sesotho sot 23 23 22 22 22 112
Setswana tsn 25 25 25 25 25 125
Shona sna 23 23 23 23 23 115
Somali som 91 91 92 92 90 456
Swahili swh 95 95 99 99 93 481
Swati ssw 45 45 43 43 41 217
Tigrinya tir 15 15 14 14 14 72
Tsonga tso 21 21 20 20 20 102
Tumbuka tum 35 35 33 33 32 168
Twi twi 5 5 11 11 5 37
Wolof wol 13 13 14 14 13 67
Xhosa xho 15 15 15 15 14 74
Yoruba yor 31 31 31 31 30 154
Zulu zul 9 9 9 9 9 45
Total 1190 1190 1359 1359 1132 6230

Scores of the three systems on the curated items (mean over the curated items; none of these scores comes from a person; see How the scores were derived).

system what it is items ev (1-10) chrf (0-100) judge overall (1-5)
sft supervised 6,225 3.53 43.9 2.27
rl_best reinforcement-learned, best checkpoint 6,225 3.75 43.5 2.15
rl_last reinforcement-learned, last checkpoint 6,225 3.73 43.5 2.15
reference the curated reference text 6,225 n/a n/a 3.99

Supported prompt formats

Prompts use the formats <hm> <base>+<switch>: TEXT and <ood> <base>: TEXT, with the three-letter language codes in angle brackets and two spaces after the colon.

Limitations

  • Our goal is for this to become the standard benchmark for African-language code-switching, so that progress can be measured and the languages are better represented. It is not there yet. The candidates were written by a model and have not been checked by native speakers, so for now treat them as references of unknown quality. Human-verified references will replace them, and we welcome contributions, especially from speakers of the 37 languages.
  • Expect unnatural phrasing, invented idioms, wrong tone marks and changes of meaning in the candidates. Pidgin outputs keep a lot of English.
  • The sampled candidates are plain temperature samples, and many are worse than the greedy one. Scores are not comparable across languages, and ev likes short, simple code-switching, so don't rank by it blindly.
  • Rows with pair_in_training_table set to false are less reliable, as explained above.

License

The English source texts keep the licences of their sources: SODA and Synthetic-Persona-Chat are CC BY 4.0, WritingPrompts is MIT, the FineWeb-Edu sample is ODC-By, and Project Gutenberg texts are public domain. CC BY 4.0 and ODC-By require attribution to the original datasets. The generated outputs come from a Gemma-based model, so the Gemma Terms of Use apply to them. Credit "AletheiaAING" if you use the dataset.

Downloads last month
43