Spaces:
Running
Running
David Pomerenke
commited on
Commit
·
9aa08d7
1
Parent(s):
d178010
Fix speed issues
Browse files- index.html +2 -2
- languagebench.py +32 -25
- pyproject.toml +1 -0
- results.json +489 -81
- results_summary.json +39 -9
- uv.lock +11 -0
index.html
CHANGED
|
@@ -66,7 +66,7 @@
|
|
| 66 |
fill: "grey",
|
| 67 |
}))),
|
| 68 |
Plot.text(summaryData, Plot.stackX({
|
| 69 |
-
filter: (d) => d.speakers >
|
| 70 |
x: "speakers",
|
| 71 |
y2: "bleu",
|
| 72 |
order: "bleu",
|
|
@@ -75,7 +75,7 @@
|
|
| 75 |
frameAnchor: "bottom",
|
| 76 |
textAnchor: "end",
|
| 77 |
dy: 10,
|
| 78 |
-
rotate: 270
|
| 79 |
}))
|
| 80 |
]
|
| 81 |
});
|
|
|
|
| 66 |
fill: "grey",
|
| 67 |
}))),
|
| 68 |
Plot.text(summaryData, Plot.stackX({
|
| 69 |
+
filter: (d) => d.speakers > 5_000_000,
|
| 70 |
x: "speakers",
|
| 71 |
y2: "bleu",
|
| 72 |
order: "bleu",
|
|
|
|
| 75 |
frameAnchor: "bottom",
|
| 76 |
textAnchor: "end",
|
| 77 |
dy: 10,
|
| 78 |
+
rotate: 270,
|
| 79 |
}))
|
| 80 |
]
|
| 81 |
});
|
languagebench.py
CHANGED
|
@@ -6,49 +6,59 @@ from os import getenv
|
|
| 6 |
|
| 7 |
import evaluate
|
| 8 |
import pandas as pd
|
|
|
|
|
|
|
| 9 |
from dotenv import load_dotenv
|
| 10 |
from joblib.memory import Memory
|
| 11 |
from openai import AsyncOpenAI
|
| 12 |
from tqdm.asyncio import tqdm_asyncio
|
| 13 |
-
from tqdm.auto import tqdm
|
| 14 |
|
| 15 |
# config
|
| 16 |
models = [
|
| 17 |
"openai/gpt-4o-mini",
|
| 18 |
-
"google/gemini-flash-1.5",
|
| 19 |
"anthropic/claude-3.5-sonnet",
|
| 20 |
-
"
|
| 21 |
-
"
|
|
|
|
|
|
|
| 22 |
]
|
| 23 |
-
# models = ["gpt-4o-mini"]
|
| 24 |
original_language = "eng_Latn"
|
| 25 |
dataset = "floresp-v2.0-rc.3/dev"
|
| 26 |
random.seed(42)
|
| 27 |
target_languages = [f.split(".")[1] for f in os.listdir(dataset)]
|
| 28 |
-
target_languages = random.choices(target_languages, k=
|
| 29 |
-
# target_languages = [
|
| 30 |
-
# "eng_Latn",
|
| 31 |
-
# "deu_Latn",
|
| 32 |
-
# "fra_Latn",
|
| 33 |
-
# "spa_Latn",
|
| 34 |
-
# "cmn_Hans",
|
| 35 |
-
# "cmn_Hant",
|
| 36 |
-
# ]
|
| 37 |
|
| 38 |
# setup
|
| 39 |
load_dotenv()
|
| 40 |
client = AsyncOpenAI(
|
| 41 |
base_url="https://openrouter.ai/api/v1",
|
| 42 |
api_key=getenv("OPENROUTER_API_KEY"),
|
| 43 |
-
# api_key=getenv("OPENAI_API_KEY"),
|
| 44 |
)
|
| 45 |
cache = Memory(location=".cache", verbose=0).cache
|
| 46 |
bleu = evaluate.load("sacrebleu")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47 |
|
| 48 |
|
| 49 |
@cache
|
| 50 |
async def complete(**kwargs):
|
| 51 |
-
|
|
|
|
|
|
|
| 52 |
|
| 53 |
|
| 54 |
def reorder(language_name):
|
|
@@ -70,10 +80,11 @@ async def translate(model, target_language, target_script, sentence):
|
|
| 70 |
messages=[
|
| 71 |
{
|
| 72 |
"role": "user",
|
| 73 |
-
"content": f"Translate the following text to {target_language}
|
| 74 |
}
|
| 75 |
],
|
| 76 |
-
temperature=0,
|
|
|
|
| 77 |
)
|
| 78 |
return reply.choices[0].message.content
|
| 79 |
|
|
@@ -103,15 +114,11 @@ async def main():
|
|
| 103 |
for model in models:
|
| 104 |
stats = get_language_stats(target_language)
|
| 105 |
print(f"{model} -> {stats['name']}")
|
| 106 |
-
# predictions = [
|
| 107 |
-
# await translate(model, stats["name"], stats["script"], sentence)
|
| 108 |
-
# for sentence in tqdm(original_sentences[:n])
|
| 109 |
-
# ]
|
| 110 |
predictions = [
|
| 111 |
translate(model, stats["name"], stats["script"], sentence)
|
| 112 |
-
for sentence in
|
| 113 |
]
|
| 114 |
-
predictions = await tqdm_asyncio.gather(*predictions)
|
| 115 |
metrics = bleu.compute(
|
| 116 |
predictions=predictions,
|
| 117 |
references=target_sentences[:n],
|
|
@@ -130,11 +137,11 @@ async def main():
|
|
| 130 |
)
|
| 131 |
with open("results.json", "w") as f:
|
| 132 |
json.dump(results, f, indent=2, ensure_ascii=False)
|
| 133 |
-
# compute mean bleu for each target language
|
| 134 |
pd.DataFrame(results).groupby("target_language_name").agg(
|
| 135 |
{"bleu": "mean", "speakers": "mean"}
|
| 136 |
).reset_index().to_json("results_summary.json", indent=2, orient="records")
|
| 137 |
|
| 138 |
|
| 139 |
if __name__ == "__main__":
|
|
|
|
| 140 |
asyncio.run(main())
|
|
|
|
| 6 |
|
| 7 |
import evaluate
|
| 8 |
import pandas as pd
|
| 9 |
+
import requests
|
| 10 |
+
from aiolimiter import AsyncLimiter
|
| 11 |
from dotenv import load_dotenv
|
| 12 |
from joblib.memory import Memory
|
| 13 |
from openai import AsyncOpenAI
|
| 14 |
from tqdm.asyncio import tqdm_asyncio
|
|
|
|
| 15 |
|
| 16 |
# config
|
| 17 |
models = [
|
| 18 |
"openai/gpt-4o-mini",
|
|
|
|
| 19 |
"anthropic/claude-3.5-sonnet",
|
| 20 |
+
"meta-llama/llama-3.1-70b-instruct", # lots of slow repetitions for LRLs
|
| 21 |
+
"mistralai/mistral-nemo",
|
| 22 |
+
"google/gemini-flash-1.5", # very fast
|
| 23 |
+
"qwen/qwen-2.5-72b-instruct", # somewhat slow
|
| 24 |
]
|
|
|
|
| 25 |
original_language = "eng_Latn"
|
| 26 |
dataset = "floresp-v2.0-rc.3/dev"
|
| 27 |
random.seed(42)
|
| 28 |
target_languages = [f.split(".")[1] for f in os.listdir(dataset)]
|
| 29 |
+
target_languages = random.choices(target_languages, k=15) + ["deu_Latn"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 30 |
|
| 31 |
# setup
|
| 32 |
load_dotenv()
|
| 33 |
client = AsyncOpenAI(
|
| 34 |
base_url="https://openrouter.ai/api/v1",
|
| 35 |
api_key=getenv("OPENROUTER_API_KEY"),
|
|
|
|
| 36 |
)
|
| 37 |
cache = Memory(location=".cache", verbose=0).cache
|
| 38 |
bleu = evaluate.load("sacrebleu")
|
| 39 |
+
rate_limit = AsyncLimiter(max_rate=2, time_period=0.1)
|
| 40 |
+
|
| 41 |
+
|
| 42 |
+
def check_rate_limit():
|
| 43 |
+
print(
|
| 44 |
+
requests.get(
|
| 45 |
+
"https://openrouter.ai/api/v1/auth/key",
|
| 46 |
+
headers={"Authorization": f"Bearer {getenv('OPENROUTER_API_KEY')}"},
|
| 47 |
+
).json()
|
| 48 |
+
)
|
| 49 |
+
print(
|
| 50 |
+
requests.get(
|
| 51 |
+
"https://openrouter.ai/api/v1/models",
|
| 52 |
+
headers={"Authorization": f"Bearer {getenv('OPENROUTER_API_KEY')}"},
|
| 53 |
+
).json()
|
| 54 |
+
)
|
| 55 |
|
| 56 |
|
| 57 |
@cache
|
| 58 |
async def complete(**kwargs):
|
| 59 |
+
async with rate_limit:
|
| 60 |
+
response = await client.chat.completions.create(**kwargs)
|
| 61 |
+
return response
|
| 62 |
|
| 63 |
|
| 64 |
def reorder(language_name):
|
|
|
|
| 80 |
messages=[
|
| 81 |
{
|
| 82 |
"role": "user",
|
| 83 |
+
"content": f"Translate the following text to the {target_language} language; use the {target_script} script; reply only with the translation:\n\n{sentence}",
|
| 84 |
}
|
| 85 |
],
|
| 86 |
+
temperature=0.1,
|
| 87 |
+
max_tokens=1024,
|
| 88 |
)
|
| 89 |
return reply.choices[0].message.content
|
| 90 |
|
|
|
|
| 114 |
for model in models:
|
| 115 |
stats = get_language_stats(target_language)
|
| 116 |
print(f"{model} -> {stats['name']}")
|
|
|
|
|
|
|
|
|
|
|
|
|
| 117 |
predictions = [
|
| 118 |
translate(model, stats["name"], stats["script"], sentence)
|
| 119 |
+
for sentence in original_sentences[:n]
|
| 120 |
]
|
| 121 |
+
predictions = await tqdm_asyncio.gather(*predictions, miniters=1)
|
| 122 |
metrics = bleu.compute(
|
| 123 |
predictions=predictions,
|
| 124 |
references=target_sentences[:n],
|
|
|
|
| 137 |
)
|
| 138 |
with open("results.json", "w") as f:
|
| 139 |
json.dump(results, f, indent=2, ensure_ascii=False)
|
|
|
|
| 140 |
pd.DataFrame(results).groupby("target_language_name").agg(
|
| 141 |
{"bleu": "mean", "speakers": "mean"}
|
| 142 |
).reset_index().to_json("results_summary.json", indent=2, orient="records")
|
| 143 |
|
| 144 |
|
| 145 |
if __name__ == "__main__":
|
| 146 |
+
# check_rate_limit()
|
| 147 |
asyncio.run(main())
|
pyproject.toml
CHANGED
|
@@ -5,6 +5,7 @@ description = "Add your description here"
|
|
| 5 |
readme = "README.md"
|
| 6 |
requires-python = ">=3.10"
|
| 7 |
dependencies = [
|
|
|
|
| 8 |
"evaluate>=0.4.3",
|
| 9 |
"joblib>=1.4.2",
|
| 10 |
"openai>=1.52.2",
|
|
|
|
| 5 |
readme = "README.md"
|
| 6 |
requires-python = ">=3.10"
|
| 7 |
dependencies = [
|
| 8 |
+
"aiolimiter>=1.1.0",
|
| 9 |
"evaluate>=0.4.3",
|
| 10 |
"joblib>=1.4.2",
|
| 11 |
"openai>=1.52.2",
|
results.json
CHANGED
|
@@ -5,39 +5,47 @@
|
|
| 5 |
"target_language": "isl_Latn",
|
| 6 |
"target_language_name": "Icelandic",
|
| 7 |
"speakers": 358000,
|
| 8 |
-
"bleu": 61.
|
| 9 |
},
|
| 10 |
{
|
| 11 |
-
"model": "
|
| 12 |
"original_language": "eng_Latn",
|
| 13 |
"target_language": "isl_Latn",
|
| 14 |
"target_language_name": "Icelandic",
|
| 15 |
"speakers": 358000,
|
| 16 |
-
"bleu":
|
| 17 |
},
|
| 18 |
{
|
| 19 |
-
"model": "
|
| 20 |
"original_language": "eng_Latn",
|
| 21 |
"target_language": "isl_Latn",
|
| 22 |
"target_language_name": "Icelandic",
|
| 23 |
"speakers": 358000,
|
| 24 |
-
"bleu":
|
| 25 |
},
|
| 26 |
{
|
| 27 |
-
"model": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 28 |
"original_language": "eng_Latn",
|
| 29 |
"target_language": "isl_Latn",
|
| 30 |
"target_language_name": "Icelandic",
|
| 31 |
"speakers": 358000,
|
| 32 |
-
"bleu":
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"model": "
|
| 36 |
"original_language": "eng_Latn",
|
| 37 |
"target_language": "isl_Latn",
|
| 38 |
"target_language_name": "Icelandic",
|
| 39 |
"speakers": 358000,
|
| 40 |
-
"bleu":
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"model": "openai/gpt-4o-mini",
|
|
@@ -45,39 +53,47 @@
|
|
| 45 |
"target_language": "ind_Latn",
|
| 46 |
"target_language_name": "Indonesian",
|
| 47 |
"speakers": 198996550,
|
| 48 |
-
"bleu":
|
| 49 |
},
|
| 50 |
{
|
| 51 |
-
"model": "
|
| 52 |
"original_language": "eng_Latn",
|
| 53 |
"target_language": "ind_Latn",
|
| 54 |
"target_language_name": "Indonesian",
|
| 55 |
"speakers": 198996550,
|
| 56 |
-
"bleu":
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"model": "
|
| 60 |
"original_language": "eng_Latn",
|
| 61 |
"target_language": "ind_Latn",
|
| 62 |
"target_language_name": "Indonesian",
|
| 63 |
"speakers": 198996550,
|
| 64 |
-
"bleu":
|
| 65 |
},
|
| 66 |
{
|
| 67 |
-
"model": "
|
| 68 |
"original_language": "eng_Latn",
|
| 69 |
"target_language": "ind_Latn",
|
| 70 |
"target_language_name": "Indonesian",
|
| 71 |
"speakers": 198996550,
|
| 72 |
-
"bleu":
|
| 73 |
},
|
| 74 |
{
|
| 75 |
-
"model": "
|
| 76 |
"original_language": "eng_Latn",
|
| 77 |
"target_language": "ind_Latn",
|
| 78 |
"target_language_name": "Indonesian",
|
| 79 |
"speakers": 198996550,
|
| 80 |
-
"bleu":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
},
|
| 82 |
{
|
| 83 |
"model": "openai/gpt-4o-mini",
|
|
@@ -85,39 +101,47 @@
|
|
| 85 |
"target_language": "pol_Latn",
|
| 86 |
"target_language_name": "Polish",
|
| 87 |
"speakers": 40200000,
|
| 88 |
-
"bleu":
|
| 89 |
},
|
| 90 |
{
|
| 91 |
-
"model": "
|
| 92 |
"original_language": "eng_Latn",
|
| 93 |
"target_language": "pol_Latn",
|
| 94 |
"target_language_name": "Polish",
|
| 95 |
"speakers": 40200000,
|
| 96 |
-
"bleu":
|
| 97 |
},
|
| 98 |
{
|
| 99 |
-
"model": "
|
| 100 |
"original_language": "eng_Latn",
|
| 101 |
"target_language": "pol_Latn",
|
| 102 |
"target_language_name": "Polish",
|
| 103 |
"speakers": 40200000,
|
| 104 |
-
"bleu": 62.
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"model": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 108 |
"original_language": "eng_Latn",
|
| 109 |
"target_language": "pol_Latn",
|
| 110 |
"target_language_name": "Polish",
|
| 111 |
"speakers": 40200000,
|
| 112 |
-
"bleu":
|
| 113 |
},
|
| 114 |
{
|
| 115 |
-
"model": "
|
| 116 |
"original_language": "eng_Latn",
|
| 117 |
"target_language": "pol_Latn",
|
| 118 |
"target_language_name": "Polish",
|
| 119 |
"speakers": 40200000,
|
| 120 |
-
"bleu":
|
| 121 |
},
|
| 122 |
{
|
| 123 |
"model": "openai/gpt-4o-mini",
|
|
@@ -125,39 +149,47 @@
|
|
| 125 |
"target_language": "kas_Deva",
|
| 126 |
"target_language_name": "Kashmiri",
|
| 127 |
"speakers": 6900000,
|
| 128 |
-
"bleu": 22.
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"model": "
|
| 132 |
"original_language": "eng_Latn",
|
| 133 |
"target_language": "kas_Deva",
|
| 134 |
"target_language_name": "Kashmiri",
|
| 135 |
"speakers": 6900000,
|
| 136 |
-
"bleu":
|
| 137 |
},
|
| 138 |
{
|
| 139 |
-
"model": "
|
| 140 |
"original_language": "eng_Latn",
|
| 141 |
"target_language": "kas_Deva",
|
| 142 |
"target_language_name": "Kashmiri",
|
| 143 |
"speakers": 6900000,
|
| 144 |
-
"bleu":
|
| 145 |
},
|
| 146 |
{
|
| 147 |
-
"model": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 148 |
"original_language": "eng_Latn",
|
| 149 |
"target_language": "kas_Deva",
|
| 150 |
"target_language_name": "Kashmiri",
|
| 151 |
"speakers": 6900000,
|
| 152 |
-
"bleu": 18.
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"model": "
|
| 156 |
"original_language": "eng_Latn",
|
| 157 |
"target_language": "kas_Deva",
|
| 158 |
"target_language_name": "Kashmiri",
|
| 159 |
"speakers": 6900000,
|
| 160 |
-
"bleu":
|
| 161 |
},
|
| 162 |
{
|
| 163 |
"model": "openai/gpt-4o-mini",
|
|
@@ -165,39 +197,47 @@
|
|
| 165 |
"target_language": "lin_Latn",
|
| 166 |
"target_language_name": "Lingala",
|
| 167 |
"speakers": 20000000,
|
| 168 |
-
"bleu":
|
| 169 |
},
|
| 170 |
{
|
| 171 |
-
"model": "
|
| 172 |
"original_language": "eng_Latn",
|
| 173 |
"target_language": "lin_Latn",
|
| 174 |
"target_language_name": "Lingala",
|
| 175 |
"speakers": 20000000,
|
| 176 |
-
"bleu":
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"model": "
|
| 180 |
"original_language": "eng_Latn",
|
| 181 |
"target_language": "lin_Latn",
|
| 182 |
"target_language_name": "Lingala",
|
| 183 |
"speakers": 20000000,
|
| 184 |
-
"bleu":
|
| 185 |
},
|
| 186 |
{
|
| 187 |
-
"model": "
|
| 188 |
"original_language": "eng_Latn",
|
| 189 |
"target_language": "lin_Latn",
|
| 190 |
"target_language_name": "Lingala",
|
| 191 |
"speakers": 20000000,
|
| 192 |
-
"bleu":
|
| 193 |
},
|
| 194 |
{
|
| 195 |
-
"model": "
|
| 196 |
"original_language": "eng_Latn",
|
| 197 |
"target_language": "lin_Latn",
|
| 198 |
"target_language_name": "Lingala",
|
| 199 |
"speakers": 20000000,
|
| 200 |
-
"bleu":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 201 |
},
|
| 202 |
{
|
| 203 |
"model": "openai/gpt-4o-mini",
|
|
@@ -205,39 +245,47 @@
|
|
| 205 |
"target_language": "bjn_Latn",
|
| 206 |
"target_language_name": "Banjar",
|
| 207 |
"speakers": 3500000,
|
| 208 |
-
"bleu": 50.
|
| 209 |
},
|
| 210 |
{
|
| 211 |
-
"model": "
|
| 212 |
"original_language": "eng_Latn",
|
| 213 |
"target_language": "bjn_Latn",
|
| 214 |
"target_language_name": "Banjar",
|
| 215 |
"speakers": 3500000,
|
| 216 |
-
"bleu":
|
| 217 |
},
|
| 218 |
{
|
| 219 |
-
"model": "
|
| 220 |
"original_language": "eng_Latn",
|
| 221 |
"target_language": "bjn_Latn",
|
| 222 |
"target_language_name": "Banjar",
|
| 223 |
"speakers": 3500000,
|
| 224 |
-
"bleu":
|
| 225 |
},
|
| 226 |
{
|
| 227 |
-
"model": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 228 |
"original_language": "eng_Latn",
|
| 229 |
"target_language": "bjn_Latn",
|
| 230 |
"target_language_name": "Banjar",
|
| 231 |
"speakers": 3500000,
|
| 232 |
-
"bleu":
|
| 233 |
},
|
| 234 |
{
|
| 235 |
-
"model": "
|
| 236 |
"original_language": "eng_Latn",
|
| 237 |
"target_language": "bjn_Latn",
|
| 238 |
"target_language_name": "Banjar",
|
| 239 |
"speakers": 3500000,
|
| 240 |
-
"bleu":
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"model": "openai/gpt-4o-mini",
|
|
@@ -245,39 +293,47 @@
|
|
| 245 |
"target_language": "bho_Deva",
|
| 246 |
"target_language_name": "Bhojpuri",
|
| 247 |
"speakers": 52200000,
|
| 248 |
-
"bleu": 43.
|
| 249 |
},
|
| 250 |
{
|
| 251 |
-
"model": "
|
| 252 |
"original_language": "eng_Latn",
|
| 253 |
"target_language": "bho_Deva",
|
| 254 |
"target_language_name": "Bhojpuri",
|
| 255 |
"speakers": 52200000,
|
| 256 |
-
"bleu":
|
| 257 |
},
|
| 258 |
{
|
| 259 |
-
"model": "
|
| 260 |
"original_language": "eng_Latn",
|
| 261 |
"target_language": "bho_Deva",
|
| 262 |
"target_language_name": "Bhojpuri",
|
| 263 |
"speakers": 52200000,
|
| 264 |
-
"bleu":
|
| 265 |
},
|
| 266 |
{
|
| 267 |
-
"model": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 268 |
"original_language": "eng_Latn",
|
| 269 |
"target_language": "bho_Deva",
|
| 270 |
"target_language_name": "Bhojpuri",
|
| 271 |
"speakers": 52200000,
|
| 272 |
-
"bleu":
|
| 273 |
},
|
| 274 |
{
|
| 275 |
-
"model": "
|
| 276 |
"original_language": "eng_Latn",
|
| 277 |
"target_language": "bho_Deva",
|
| 278 |
"target_language_name": "Bhojpuri",
|
| 279 |
"speakers": 52200000,
|
| 280 |
-
"bleu":
|
| 281 |
},
|
| 282 |
{
|
| 283 |
"model": "openai/gpt-4o-mini",
|
|
@@ -285,39 +341,47 @@
|
|
| 285 |
"target_language": "ces_Latn",
|
| 286 |
"target_language_name": "Czech",
|
| 287 |
"speakers": 10700000,
|
| 288 |
-
"bleu":
|
| 289 |
},
|
| 290 |
{
|
| 291 |
-
"model": "
|
| 292 |
"original_language": "eng_Latn",
|
| 293 |
"target_language": "ces_Latn",
|
| 294 |
"target_language_name": "Czech",
|
| 295 |
"speakers": 10700000,
|
| 296 |
-
"bleu": 69.
|
| 297 |
},
|
| 298 |
{
|
| 299 |
-
"model": "
|
| 300 |
"original_language": "eng_Latn",
|
| 301 |
"target_language": "ces_Latn",
|
| 302 |
"target_language_name": "Czech",
|
| 303 |
"speakers": 10700000,
|
| 304 |
-
"bleu":
|
| 305 |
},
|
| 306 |
{
|
| 307 |
-
"model": "
|
| 308 |
"original_language": "eng_Latn",
|
| 309 |
"target_language": "ces_Latn",
|
| 310 |
"target_language_name": "Czech",
|
| 311 |
"speakers": 10700000,
|
| 312 |
-
"bleu":
|
| 313 |
},
|
| 314 |
{
|
| 315 |
-
"model": "
|
| 316 |
"original_language": "eng_Latn",
|
| 317 |
"target_language": "ces_Latn",
|
| 318 |
"target_language_name": "Czech",
|
| 319 |
"speakers": 10700000,
|
| 320 |
-
"bleu":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 321 |
},
|
| 322 |
{
|
| 323 |
"model": "openai/gpt-4o-mini",
|
|
@@ -325,38 +389,382 @@
|
|
| 325 |
"target_language": "sag_Latn",
|
| 326 |
"target_language_name": "Sango",
|
| 327 |
"speakers": 4600000,
|
| 328 |
-
"bleu":
|
| 329 |
},
|
| 330 |
{
|
| 331 |
-
"model": "
|
| 332 |
"original_language": "eng_Latn",
|
| 333 |
"target_language": "sag_Latn",
|
| 334 |
"target_language_name": "Sango",
|
| 335 |
"speakers": 4600000,
|
| 336 |
-
"bleu":
|
| 337 |
},
|
| 338 |
{
|
| 339 |
-
"model": "
|
| 340 |
"original_language": "eng_Latn",
|
| 341 |
"target_language": "sag_Latn",
|
| 342 |
"target_language_name": "Sango",
|
| 343 |
"speakers": 4600000,
|
| 344 |
-
"bleu":
|
| 345 |
},
|
| 346 |
{
|
| 347 |
-
"model": "
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 348 |
"original_language": "eng_Latn",
|
| 349 |
"target_language": "sag_Latn",
|
| 350 |
"target_language_name": "Sango",
|
| 351 |
"speakers": 4600000,
|
| 352 |
-
"bleu":
|
| 353 |
},
|
| 354 |
{
|
| 355 |
-
"model": "
|
| 356 |
"original_language": "eng_Latn",
|
| 357 |
"target_language": "sag_Latn",
|
| 358 |
"target_language_name": "Sango",
|
| 359 |
"speakers": 4600000,
|
| 360 |
-
"bleu":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 361 |
}
|
| 362 |
]
|
|
|
|
| 5 |
"target_language": "isl_Latn",
|
| 6 |
"target_language_name": "Icelandic",
|
| 7 |
"speakers": 358000,
|
| 8 |
+
"bleu": 61.11721529849627
|
| 9 |
},
|
| 10 |
{
|
| 11 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 12 |
"original_language": "eng_Latn",
|
| 13 |
"target_language": "isl_Latn",
|
| 14 |
"target_language_name": "Icelandic",
|
| 15 |
"speakers": 358000,
|
| 16 |
+
"bleu": 66.81187855948907
|
| 17 |
},
|
| 18 |
{
|
| 19 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 20 |
"original_language": "eng_Latn",
|
| 21 |
"target_language": "isl_Latn",
|
| 22 |
"target_language_name": "Icelandic",
|
| 23 |
"speakers": 358000,
|
| 24 |
+
"bleu": 58.83900250147204
|
| 25 |
},
|
| 26 |
{
|
| 27 |
+
"model": "mistralai/mistral-nemo",
|
| 28 |
+
"original_language": "eng_Latn",
|
| 29 |
+
"target_language": "isl_Latn",
|
| 30 |
+
"target_language_name": "Icelandic",
|
| 31 |
+
"speakers": 358000,
|
| 32 |
+
"bleu": 46.527810964691184
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"model": "google/gemini-flash-1.5",
|
| 36 |
"original_language": "eng_Latn",
|
| 37 |
"target_language": "isl_Latn",
|
| 38 |
"target_language_name": "Icelandic",
|
| 39 |
"speakers": 358000,
|
| 40 |
+
"bleu": 59.44313738508877
|
| 41 |
},
|
| 42 |
{
|
| 43 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 44 |
"original_language": "eng_Latn",
|
| 45 |
"target_language": "isl_Latn",
|
| 46 |
"target_language_name": "Icelandic",
|
| 47 |
"speakers": 358000,
|
| 48 |
+
"bleu": 43.301355850733806
|
| 49 |
},
|
| 50 |
{
|
| 51 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 53 |
"target_language": "ind_Latn",
|
| 54 |
"target_language_name": "Indonesian",
|
| 55 |
"speakers": 198996550,
|
| 56 |
+
"bleu": 76.72985943936231
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 60 |
"original_language": "eng_Latn",
|
| 61 |
"target_language": "ind_Latn",
|
| 62 |
"target_language_name": "Indonesian",
|
| 63 |
"speakers": 198996550,
|
| 64 |
+
"bleu": 75.27574767220828
|
| 65 |
},
|
| 66 |
{
|
| 67 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 68 |
"original_language": "eng_Latn",
|
| 69 |
"target_language": "ind_Latn",
|
| 70 |
"target_language_name": "Indonesian",
|
| 71 |
"speakers": 198996550,
|
| 72 |
+
"bleu": 74.67140101059968
|
| 73 |
},
|
| 74 |
{
|
| 75 |
+
"model": "mistralai/mistral-nemo",
|
| 76 |
"original_language": "eng_Latn",
|
| 77 |
"target_language": "ind_Latn",
|
| 78 |
"target_language_name": "Indonesian",
|
| 79 |
"speakers": 198996550,
|
| 80 |
+
"bleu": 66.83496111463917
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"model": "google/gemini-flash-1.5",
|
| 84 |
"original_language": "eng_Latn",
|
| 85 |
"target_language": "ind_Latn",
|
| 86 |
"target_language_name": "Indonesian",
|
| 87 |
"speakers": 198996550,
|
| 88 |
+
"bleu": 74.42344305993296
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 92 |
+
"original_language": "eng_Latn",
|
| 93 |
+
"target_language": "ind_Latn",
|
| 94 |
+
"target_language_name": "Indonesian",
|
| 95 |
+
"speakers": 198996550,
|
| 96 |
+
"bleu": 73.22218842351182
|
| 97 |
},
|
| 98 |
{
|
| 99 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 101 |
"target_language": "pol_Latn",
|
| 102 |
"target_language_name": "Polish",
|
| 103 |
"speakers": 40200000,
|
| 104 |
+
"bleu": 64.36371715824997
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 108 |
"original_language": "eng_Latn",
|
| 109 |
"target_language": "pol_Latn",
|
| 110 |
"target_language_name": "Polish",
|
| 111 |
"speakers": 40200000,
|
| 112 |
+
"bleu": 66.08214136732671
|
| 113 |
},
|
| 114 |
{
|
| 115 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 116 |
"original_language": "eng_Latn",
|
| 117 |
"target_language": "pol_Latn",
|
| 118 |
"target_language_name": "Polish",
|
| 119 |
"speakers": 40200000,
|
| 120 |
+
"bleu": 62.976635886624194
|
| 121 |
},
|
| 122 |
{
|
| 123 |
+
"model": "mistralai/mistral-nemo",
|
| 124 |
+
"original_language": "eng_Latn",
|
| 125 |
+
"target_language": "pol_Latn",
|
| 126 |
+
"target_language_name": "Polish",
|
| 127 |
+
"speakers": 40200000,
|
| 128 |
+
"bleu": 58.10005545101253
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"model": "google/gemini-flash-1.5",
|
| 132 |
"original_language": "eng_Latn",
|
| 133 |
"target_language": "pol_Latn",
|
| 134 |
"target_language_name": "Polish",
|
| 135 |
"speakers": 40200000,
|
| 136 |
+
"bleu": 65.21734418681433
|
| 137 |
},
|
| 138 |
{
|
| 139 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 140 |
"original_language": "eng_Latn",
|
| 141 |
"target_language": "pol_Latn",
|
| 142 |
"target_language_name": "Polish",
|
| 143 |
"speakers": 40200000,
|
| 144 |
+
"bleu": 61.96619004506918
|
| 145 |
},
|
| 146 |
{
|
| 147 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 149 |
"target_language": "kas_Deva",
|
| 150 |
"target_language_name": "Kashmiri",
|
| 151 |
"speakers": 6900000,
|
| 152 |
+
"bleu": 22.82001592906364
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 156 |
"original_language": "eng_Latn",
|
| 157 |
"target_language": "kas_Deva",
|
| 158 |
"target_language_name": "Kashmiri",
|
| 159 |
"speakers": 6900000,
|
| 160 |
+
"bleu": 27.495782316124032
|
| 161 |
},
|
| 162 |
{
|
| 163 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 164 |
"original_language": "eng_Latn",
|
| 165 |
"target_language": "kas_Deva",
|
| 166 |
"target_language_name": "Kashmiri",
|
| 167 |
"speakers": 6900000,
|
| 168 |
+
"bleu": 23.76175861036051
|
| 169 |
},
|
| 170 |
{
|
| 171 |
+
"model": "mistralai/mistral-nemo",
|
| 172 |
+
"original_language": "eng_Latn",
|
| 173 |
+
"target_language": "kas_Deva",
|
| 174 |
+
"target_language_name": "Kashmiri",
|
| 175 |
+
"speakers": 6900000,
|
| 176 |
+
"bleu": 22.35639610853823
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"model": "google/gemini-flash-1.5",
|
| 180 |
"original_language": "eng_Latn",
|
| 181 |
"target_language": "kas_Deva",
|
| 182 |
"target_language_name": "Kashmiri",
|
| 183 |
"speakers": 6900000,
|
| 184 |
+
"bleu": 18.32056378467213
|
| 185 |
},
|
| 186 |
{
|
| 187 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 188 |
"original_language": "eng_Latn",
|
| 189 |
"target_language": "kas_Deva",
|
| 190 |
"target_language_name": "Kashmiri",
|
| 191 |
"speakers": 6900000,
|
| 192 |
+
"bleu": 15.7563725169679
|
| 193 |
},
|
| 194 |
{
|
| 195 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 197 |
"target_language": "lin_Latn",
|
| 198 |
"target_language_name": "Lingala",
|
| 199 |
"speakers": 20000000,
|
| 200 |
+
"bleu": 49.63954111089247
|
| 201 |
},
|
| 202 |
{
|
| 203 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 204 |
"original_language": "eng_Latn",
|
| 205 |
"target_language": "lin_Latn",
|
| 206 |
"target_language_name": "Lingala",
|
| 207 |
"speakers": 20000000,
|
| 208 |
+
"bleu": 56.61505235063709
|
| 209 |
},
|
| 210 |
{
|
| 211 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 212 |
"original_language": "eng_Latn",
|
| 213 |
"target_language": "lin_Latn",
|
| 214 |
"target_language_name": "Lingala",
|
| 215 |
"speakers": 20000000,
|
| 216 |
+
"bleu": 10.308395614510356
|
| 217 |
},
|
| 218 |
{
|
| 219 |
+
"model": "mistralai/mistral-nemo",
|
| 220 |
"original_language": "eng_Latn",
|
| 221 |
"target_language": "lin_Latn",
|
| 222 |
"target_language_name": "Lingala",
|
| 223 |
"speakers": 20000000,
|
| 224 |
+
"bleu": 5.39489654623325
|
| 225 |
},
|
| 226 |
{
|
| 227 |
+
"model": "google/gemini-flash-1.5",
|
| 228 |
"original_language": "eng_Latn",
|
| 229 |
"target_language": "lin_Latn",
|
| 230 |
"target_language_name": "Lingala",
|
| 231 |
"speakers": 20000000,
|
| 232 |
+
"bleu": 56.907145530399205
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 236 |
+
"original_language": "eng_Latn",
|
| 237 |
+
"target_language": "lin_Latn",
|
| 238 |
+
"target_language_name": "Lingala",
|
| 239 |
+
"speakers": 20000000,
|
| 240 |
+
"bleu": 22.097904872127472
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 245 |
"target_language": "bjn_Latn",
|
| 246 |
"target_language_name": "Banjar",
|
| 247 |
"speakers": 3500000,
|
| 248 |
+
"bleu": 50.84965350648206
|
| 249 |
},
|
| 250 |
{
|
| 251 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 252 |
"original_language": "eng_Latn",
|
| 253 |
"target_language": "bjn_Latn",
|
| 254 |
"target_language_name": "Banjar",
|
| 255 |
"speakers": 3500000,
|
| 256 |
+
"bleu": 55.87800126842635
|
| 257 |
},
|
| 258 |
{
|
| 259 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 260 |
"original_language": "eng_Latn",
|
| 261 |
"target_language": "bjn_Latn",
|
| 262 |
"target_language_name": "Banjar",
|
| 263 |
"speakers": 3500000,
|
| 264 |
+
"bleu": 56.01233364686365
|
| 265 |
},
|
| 266 |
{
|
| 267 |
+
"model": "mistralai/mistral-nemo",
|
| 268 |
+
"original_language": "eng_Latn",
|
| 269 |
+
"target_language": "bjn_Latn",
|
| 270 |
+
"target_language_name": "Banjar",
|
| 271 |
+
"speakers": 3500000,
|
| 272 |
+
"bleu": 45.594880960004936
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"model": "google/gemini-flash-1.5",
|
| 276 |
"original_language": "eng_Latn",
|
| 277 |
"target_language": "bjn_Latn",
|
| 278 |
"target_language_name": "Banjar",
|
| 279 |
"speakers": 3500000,
|
| 280 |
+
"bleu": 55.05933522821672
|
| 281 |
},
|
| 282 |
{
|
| 283 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 284 |
"original_language": "eng_Latn",
|
| 285 |
"target_language": "bjn_Latn",
|
| 286 |
"target_language_name": "Banjar",
|
| 287 |
"speakers": 3500000,
|
| 288 |
+
"bleu": 45.742104208986554
|
| 289 |
},
|
| 290 |
{
|
| 291 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 293 |
"target_language": "bho_Deva",
|
| 294 |
"target_language_name": "Bhojpuri",
|
| 295 |
"speakers": 52200000,
|
| 296 |
+
"bleu": 43.19693805178097
|
| 297 |
},
|
| 298 |
{
|
| 299 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 300 |
"original_language": "eng_Latn",
|
| 301 |
"target_language": "bho_Deva",
|
| 302 |
"target_language_name": "Bhojpuri",
|
| 303 |
"speakers": 52200000,
|
| 304 |
+
"bleu": 44.54261459568271
|
| 305 |
},
|
| 306 |
{
|
| 307 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 308 |
"original_language": "eng_Latn",
|
| 309 |
"target_language": "bho_Deva",
|
| 310 |
"target_language_name": "Bhojpuri",
|
| 311 |
"speakers": 52200000,
|
| 312 |
+
"bleu": 43.97962317460702
|
| 313 |
},
|
| 314 |
{
|
| 315 |
+
"model": "mistralai/mistral-nemo",
|
| 316 |
+
"original_language": "eng_Latn",
|
| 317 |
+
"target_language": "bho_Deva",
|
| 318 |
+
"target_language_name": "Bhojpuri",
|
| 319 |
+
"speakers": 52200000,
|
| 320 |
+
"bleu": 31.161774640380557
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"model": "google/gemini-flash-1.5",
|
| 324 |
"original_language": "eng_Latn",
|
| 325 |
"target_language": "bho_Deva",
|
| 326 |
"target_language_name": "Bhojpuri",
|
| 327 |
"speakers": 52200000,
|
| 328 |
+
"bleu": 40.00304050592406
|
| 329 |
},
|
| 330 |
{
|
| 331 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 332 |
"original_language": "eng_Latn",
|
| 333 |
"target_language": "bho_Deva",
|
| 334 |
"target_language_name": "Bhojpuri",
|
| 335 |
"speakers": 52200000,
|
| 336 |
+
"bleu": 32.75450450583861
|
| 337 |
},
|
| 338 |
{
|
| 339 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 341 |
"target_language": "ces_Latn",
|
| 342 |
"target_language_name": "Czech",
|
| 343 |
"speakers": 10700000,
|
| 344 |
+
"bleu": 70.25260364509532
|
| 345 |
},
|
| 346 |
{
|
| 347 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 348 |
"original_language": "eng_Latn",
|
| 349 |
"target_language": "ces_Latn",
|
| 350 |
"target_language_name": "Czech",
|
| 351 |
"speakers": 10700000,
|
| 352 |
+
"bleu": 69.93901215857153
|
| 353 |
},
|
| 354 |
{
|
| 355 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 356 |
"original_language": "eng_Latn",
|
| 357 |
"target_language": "ces_Latn",
|
| 358 |
"target_language_name": "Czech",
|
| 359 |
"speakers": 10700000,
|
| 360 |
+
"bleu": 65.8092084143282
|
| 361 |
},
|
| 362 |
{
|
| 363 |
+
"model": "mistralai/mistral-nemo",
|
| 364 |
"original_language": "eng_Latn",
|
| 365 |
"target_language": "ces_Latn",
|
| 366 |
"target_language_name": "Czech",
|
| 367 |
"speakers": 10700000,
|
| 368 |
+
"bleu": 62.378801757235195
|
| 369 |
},
|
| 370 |
{
|
| 371 |
+
"model": "google/gemini-flash-1.5",
|
| 372 |
"original_language": "eng_Latn",
|
| 373 |
"target_language": "ces_Latn",
|
| 374 |
"target_language_name": "Czech",
|
| 375 |
"speakers": 10700000,
|
| 376 |
+
"bleu": 69.19130488388231
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 380 |
+
"original_language": "eng_Latn",
|
| 381 |
+
"target_language": "ces_Latn",
|
| 382 |
+
"target_language_name": "Czech",
|
| 383 |
+
"speakers": 10700000,
|
| 384 |
+
"bleu": 62.958317947186224
|
| 385 |
},
|
| 386 |
{
|
| 387 |
"model": "openai/gpt-4o-mini",
|
|
|
|
| 389 |
"target_language": "sag_Latn",
|
| 390 |
"target_language_name": "Sango",
|
| 391 |
"speakers": 4600000,
|
| 392 |
+
"bleu": 22.51219967404956
|
| 393 |
},
|
| 394 |
{
|
| 395 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 396 |
"original_language": "eng_Latn",
|
| 397 |
"target_language": "sag_Latn",
|
| 398 |
"target_language_name": "Sango",
|
| 399 |
"speakers": 4600000,
|
| 400 |
+
"bleu": 33.66347085808987
|
| 401 |
},
|
| 402 |
{
|
| 403 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 404 |
"original_language": "eng_Latn",
|
| 405 |
"target_language": "sag_Latn",
|
| 406 |
"target_language_name": "Sango",
|
| 407 |
"speakers": 4600000,
|
| 408 |
+
"bleu": 5.1757926109109125
|
| 409 |
},
|
| 410 |
{
|
| 411 |
+
"model": "mistralai/mistral-nemo",
|
| 412 |
+
"original_language": "eng_Latn",
|
| 413 |
+
"target_language": "sag_Latn",
|
| 414 |
+
"target_language_name": "Sango",
|
| 415 |
+
"speakers": 4600000,
|
| 416 |
+
"bleu": 3.089512428083042
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"model": "google/gemini-flash-1.5",
|
| 420 |
"original_language": "eng_Latn",
|
| 421 |
"target_language": "sag_Latn",
|
| 422 |
"target_language_name": "Sango",
|
| 423 |
"speakers": 4600000,
|
| 424 |
+
"bleu": 37.1561421416539
|
| 425 |
},
|
| 426 |
{
|
| 427 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 428 |
"original_language": "eng_Latn",
|
| 429 |
"target_language": "sag_Latn",
|
| 430 |
"target_language_name": "Sango",
|
| 431 |
"speakers": 4600000,
|
| 432 |
+
"bleu": 16.36794672654396
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"model": "openai/gpt-4o-mini",
|
| 436 |
+
"original_language": "eng_Latn",
|
| 437 |
+
"target_language": "nus_Latn",
|
| 438 |
+
"target_language_name": "Nuer",
|
| 439 |
+
"speakers": 900000,
|
| 440 |
+
"bleu": 4.395400049180613
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 444 |
+
"original_language": "eng_Latn",
|
| 445 |
+
"target_language": "nus_Latn",
|
| 446 |
+
"target_language_name": "Nuer",
|
| 447 |
+
"speakers": 900000,
|
| 448 |
+
"bleu": 17.048915940655856
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 452 |
+
"original_language": "eng_Latn",
|
| 453 |
+
"target_language": "nus_Latn",
|
| 454 |
+
"target_language_name": "Nuer",
|
| 455 |
+
"speakers": 900000,
|
| 456 |
+
"bleu": 2.1467398669412425
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"model": "mistralai/mistral-nemo",
|
| 460 |
+
"original_language": "eng_Latn",
|
| 461 |
+
"target_language": "nus_Latn",
|
| 462 |
+
"target_language_name": "Nuer",
|
| 463 |
+
"speakers": 900000,
|
| 464 |
+
"bleu": 1.3220304007482553
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"model": "google/gemini-flash-1.5",
|
| 468 |
+
"original_language": "eng_Latn",
|
| 469 |
+
"target_language": "nus_Latn",
|
| 470 |
+
"target_language_name": "Nuer",
|
| 471 |
+
"speakers": 900000,
|
| 472 |
+
"bleu": 2.8486682345713175
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 476 |
+
"original_language": "eng_Latn",
|
| 477 |
+
"target_language": "nus_Latn",
|
| 478 |
+
"target_language_name": "Nuer",
|
| 479 |
+
"speakers": 900000,
|
| 480 |
+
"bleu": 3.98395018103736
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"model": "openai/gpt-4o-mini",
|
| 484 |
+
"original_language": "eng_Latn",
|
| 485 |
+
"target_language": "srp_Cyrl",
|
| 486 |
+
"target_language_name": "Serbian",
|
| 487 |
+
"speakers": 9000000,
|
| 488 |
+
"bleu": 68.38476225789503
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 492 |
+
"original_language": "eng_Latn",
|
| 493 |
+
"target_language": "srp_Cyrl",
|
| 494 |
+
"target_language_name": "Serbian",
|
| 495 |
+
"speakers": 9000000,
|
| 496 |
+
"bleu": 69.90182117504679
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 500 |
+
"original_language": "eng_Latn",
|
| 501 |
+
"target_language": "srp_Cyrl",
|
| 502 |
+
"target_language_name": "Serbian",
|
| 503 |
+
"speakers": 9000000,
|
| 504 |
+
"bleu": 66.57363169106475
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"model": "mistralai/mistral-nemo",
|
| 508 |
+
"original_language": "eng_Latn",
|
| 509 |
+
"target_language": "srp_Cyrl",
|
| 510 |
+
"target_language_name": "Serbian",
|
| 511 |
+
"speakers": 9000000,
|
| 512 |
+
"bleu": 58.52489424015931
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"model": "google/gemini-flash-1.5",
|
| 516 |
+
"original_language": "eng_Latn",
|
| 517 |
+
"target_language": "srp_Cyrl",
|
| 518 |
+
"target_language_name": "Serbian",
|
| 519 |
+
"speakers": 9000000,
|
| 520 |
+
"bleu": 68.63835437069461
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 524 |
+
"original_language": "eng_Latn",
|
| 525 |
+
"target_language": "srp_Cyrl",
|
| 526 |
+
"target_language_name": "Serbian",
|
| 527 |
+
"speakers": 9000000,
|
| 528 |
+
"bleu": 56.895639010018655
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"model": "openai/gpt-4o-mini",
|
| 532 |
+
"original_language": "eng_Latn",
|
| 533 |
+
"target_language": "quy_Latn",
|
| 534 |
+
"target_language_name": " Ayacucho Quechua",
|
| 535 |
+
"speakers": 918200,
|
| 536 |
+
"bleu": 16.93431059735613
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 540 |
+
"original_language": "eng_Latn",
|
| 541 |
+
"target_language": "quy_Latn",
|
| 542 |
+
"target_language_name": " Ayacucho Quechua",
|
| 543 |
+
"speakers": 918200,
|
| 544 |
+
"bleu": 45.34723191551773
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 548 |
+
"original_language": "eng_Latn",
|
| 549 |
+
"target_language": "quy_Latn",
|
| 550 |
+
"target_language_name": " Ayacucho Quechua",
|
| 551 |
+
"speakers": 918200,
|
| 552 |
+
"bleu": 15.68251052610097
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"model": "mistralai/mistral-nemo",
|
| 556 |
+
"original_language": "eng_Latn",
|
| 557 |
+
"target_language": "quy_Latn",
|
| 558 |
+
"target_language_name": " Ayacucho Quechua",
|
| 559 |
+
"speakers": 918200,
|
| 560 |
+
"bleu": 6.787154528017713
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"model": "google/gemini-flash-1.5",
|
| 564 |
+
"original_language": "eng_Latn",
|
| 565 |
+
"target_language": "quy_Latn",
|
| 566 |
+
"target_language_name": " Ayacucho Quechua",
|
| 567 |
+
"speakers": 918200,
|
| 568 |
+
"bleu": 42.71572417876077
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 572 |
+
"original_language": "eng_Latn",
|
| 573 |
+
"target_language": "quy_Latn",
|
| 574 |
+
"target_language_name": " Ayacucho Quechua",
|
| 575 |
+
"speakers": 918200,
|
| 576 |
+
"bleu": 16.1290828285597
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"model": "openai/gpt-4o-mini",
|
| 580 |
+
"original_language": "eng_Latn",
|
| 581 |
+
"target_language": "ind_Latn",
|
| 582 |
+
"target_language_name": "Indonesian",
|
| 583 |
+
"speakers": 198996550,
|
| 584 |
+
"bleu": 76.72985943936231
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 588 |
+
"original_language": "eng_Latn",
|
| 589 |
+
"target_language": "ind_Latn",
|
| 590 |
+
"target_language_name": "Indonesian",
|
| 591 |
+
"speakers": 198996550,
|
| 592 |
+
"bleu": 75.27574767220828
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 596 |
+
"original_language": "eng_Latn",
|
| 597 |
+
"target_language": "ind_Latn",
|
| 598 |
+
"target_language_name": "Indonesian",
|
| 599 |
+
"speakers": 198996550,
|
| 600 |
+
"bleu": 74.67140101059968
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"model": "mistralai/mistral-nemo",
|
| 604 |
+
"original_language": "eng_Latn",
|
| 605 |
+
"target_language": "ind_Latn",
|
| 606 |
+
"target_language_name": "Indonesian",
|
| 607 |
+
"speakers": 198996550,
|
| 608 |
+
"bleu": 66.83496111463917
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"model": "google/gemini-flash-1.5",
|
| 612 |
+
"original_language": "eng_Latn",
|
| 613 |
+
"target_language": "ind_Latn",
|
| 614 |
+
"target_language_name": "Indonesian",
|
| 615 |
+
"speakers": 198996550,
|
| 616 |
+
"bleu": 74.42344305993296
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 620 |
+
"original_language": "eng_Latn",
|
| 621 |
+
"target_language": "ind_Latn",
|
| 622 |
+
"target_language_name": "Indonesian",
|
| 623 |
+
"speakers": 198996550,
|
| 624 |
+
"bleu": 73.22218842351182
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"model": "openai/gpt-4o-mini",
|
| 628 |
+
"original_language": "eng_Latn",
|
| 629 |
+
"target_language": "ban_Latn",
|
| 630 |
+
"target_language_name": "Bali (Indonesia)",
|
| 631 |
+
"speakers": 4000000,
|
| 632 |
+
"bleu": 53.59739003047012
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 636 |
+
"original_language": "eng_Latn",
|
| 637 |
+
"target_language": "ban_Latn",
|
| 638 |
+
"target_language_name": "Bali (Indonesia)",
|
| 639 |
+
"speakers": 4000000,
|
| 640 |
+
"bleu": 52.93221548087568
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 644 |
+
"original_language": "eng_Latn",
|
| 645 |
+
"target_language": "ban_Latn",
|
| 646 |
+
"target_language_name": "Bali (Indonesia)",
|
| 647 |
+
"speakers": 4000000,
|
| 648 |
+
"bleu": 50.05490673003243
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"model": "mistralai/mistral-nemo",
|
| 652 |
+
"original_language": "eng_Latn",
|
| 653 |
+
"target_language": "ban_Latn",
|
| 654 |
+
"target_language_name": "Bali (Indonesia)",
|
| 655 |
+
"speakers": 4000000,
|
| 656 |
+
"bleu": 51.163592105681474
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"model": "google/gemini-flash-1.5",
|
| 660 |
+
"original_language": "eng_Latn",
|
| 661 |
+
"target_language": "ban_Latn",
|
| 662 |
+
"target_language_name": "Bali (Indonesia)",
|
| 663 |
+
"speakers": 4000000,
|
| 664 |
+
"bleu": 50.17518305639421
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 668 |
+
"original_language": "eng_Latn",
|
| 669 |
+
"target_language": "ban_Latn",
|
| 670 |
+
"target_language_name": "Bali (Indonesia)",
|
| 671 |
+
"speakers": 4000000,
|
| 672 |
+
"bleu": 51.734846987161674
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"model": "openai/gpt-4o-mini",
|
| 676 |
+
"original_language": "eng_Latn",
|
| 677 |
+
"target_language": "als_Latn",
|
| 678 |
+
"target_language_name": " Tosk Albanian",
|
| 679 |
+
"speakers": 3000000,
|
| 680 |
+
"bleu": 66.25169488670164
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 684 |
+
"original_language": "eng_Latn",
|
| 685 |
+
"target_language": "als_Latn",
|
| 686 |
+
"target_language_name": " Tosk Albanian",
|
| 687 |
+
"speakers": 3000000,
|
| 688 |
+
"bleu": 69.48790830340513
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 692 |
+
"original_language": "eng_Latn",
|
| 693 |
+
"target_language": "als_Latn",
|
| 694 |
+
"target_language_name": " Tosk Albanian",
|
| 695 |
+
"speakers": 3000000,
|
| 696 |
+
"bleu": 64.27107595491931
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"model": "mistralai/mistral-nemo",
|
| 700 |
+
"original_language": "eng_Latn",
|
| 701 |
+
"target_language": "als_Latn",
|
| 702 |
+
"target_language_name": " Tosk Albanian",
|
| 703 |
+
"speakers": 3000000,
|
| 704 |
+
"bleu": 35.73300770047836
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"model": "google/gemini-flash-1.5",
|
| 708 |
+
"original_language": "eng_Latn",
|
| 709 |
+
"target_language": "als_Latn",
|
| 710 |
+
"target_language_name": " Tosk Albanian",
|
| 711 |
+
"speakers": 3000000,
|
| 712 |
+
"bleu": 66.62103194583597
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 716 |
+
"original_language": "eng_Latn",
|
| 717 |
+
"target_language": "als_Latn",
|
| 718 |
+
"target_language_name": " Tosk Albanian",
|
| 719 |
+
"speakers": 3000000,
|
| 720 |
+
"bleu": 51.24498484368196
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"model": "openai/gpt-4o-mini",
|
| 724 |
+
"original_language": "eng_Latn",
|
| 725 |
+
"target_language": "deu_Latn",
|
| 726 |
+
"target_language_name": " Standard German",
|
| 727 |
+
"speakers": 105000000,
|
| 728 |
+
"bleu": 74.41880377467668
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"model": "anthropic/claude-3.5-sonnet",
|
| 732 |
+
"original_language": "eng_Latn",
|
| 733 |
+
"target_language": "deu_Latn",
|
| 734 |
+
"target_language_name": " Standard German",
|
| 735 |
+
"speakers": 105000000,
|
| 736 |
+
"bleu": 77.22130383692244
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"model": "meta-llama/llama-3.1-70b-instruct",
|
| 740 |
+
"original_language": "eng_Latn",
|
| 741 |
+
"target_language": "deu_Latn",
|
| 742 |
+
"target_language_name": " Standard German",
|
| 743 |
+
"speakers": 105000000,
|
| 744 |
+
"bleu": 73.32845880107014
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"model": "mistralai/mistral-nemo",
|
| 748 |
+
"original_language": "eng_Latn",
|
| 749 |
+
"target_language": "deu_Latn",
|
| 750 |
+
"target_language_name": " Standard German",
|
| 751 |
+
"speakers": 105000000,
|
| 752 |
+
"bleu": 70.07261531493668
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"model": "google/gemini-flash-1.5",
|
| 756 |
+
"original_language": "eng_Latn",
|
| 757 |
+
"target_language": "deu_Latn",
|
| 758 |
+
"target_language_name": " Standard German",
|
| 759 |
+
"speakers": 105000000,
|
| 760 |
+
"bleu": 73.9443508325751
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"model": "qwen/qwen-2.5-72b-instruct",
|
| 764 |
+
"original_language": "eng_Latn",
|
| 765 |
+
"target_language": "deu_Latn",
|
| 766 |
+
"target_language_name": " Standard German",
|
| 767 |
+
"speakers": 105000000,
|
| 768 |
+
"bleu": 72.01646514265417
|
| 769 |
}
|
| 770 |
]
|
results_summary.json
CHANGED
|
@@ -1,47 +1,77 @@
|
|
| 1 |
[
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
{
|
| 3 |
"target_language_name":"Banjar",
|
| 4 |
-
"bleu":
|
| 5 |
"speakers":3500000.0
|
| 6 |
},
|
| 7 |
{
|
| 8 |
"target_language_name":"Bhojpuri",
|
| 9 |
-
"bleu":
|
| 10 |
"speakers":52200000.0
|
| 11 |
},
|
| 12 |
{
|
| 13 |
"target_language_name":"Czech",
|
| 14 |
-
"bleu":
|
| 15 |
"speakers":10700000.0
|
| 16 |
},
|
| 17 |
{
|
| 18 |
"target_language_name":"Icelandic",
|
| 19 |
-
"bleu":
|
| 20 |
"speakers":358000.0
|
| 21 |
},
|
| 22 |
{
|
| 23 |
"target_language_name":"Indonesian",
|
| 24 |
-
"bleu":
|
| 25 |
"speakers":198996550.0
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"target_language_name":"Kashmiri",
|
| 29 |
-
"bleu":
|
| 30 |
"speakers":6900000.0
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"target_language_name":"Lingala",
|
| 34 |
-
"bleu":
|
| 35 |
"speakers":20000000.0
|
| 36 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
{
|
| 38 |
"target_language_name":"Polish",
|
| 39 |
-
"bleu":
|
| 40 |
"speakers":40200000.0
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"target_language_name":"Sango",
|
| 44 |
-
"bleu":
|
| 45 |
"speakers":4600000.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
}
|
| 47 |
]
|
|
|
|
| 1 |
[
|
| 2 |
+
{
|
| 3 |
+
"target_language_name":" Ayacucho Quechua",
|
| 4 |
+
"bleu":23.9326690957,
|
| 5 |
+
"speakers":918200.0
|
| 6 |
+
},
|
| 7 |
+
{
|
| 8 |
+
"target_language_name":" Standard German",
|
| 9 |
+
"bleu":73.5003329505,
|
| 10 |
+
"speakers":105000000.0
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"target_language_name":" Tosk Albanian",
|
| 14 |
+
"bleu":58.9349506058,
|
| 15 |
+
"speakers":3000000.0
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"target_language_name":"Bali (Indonesia)",
|
| 19 |
+
"bleu":51.6096890651,
|
| 20 |
+
"speakers":4000000.0
|
| 21 |
+
},
|
| 22 |
{
|
| 23 |
"target_language_name":"Banjar",
|
| 24 |
+
"bleu":51.5227181365,
|
| 25 |
"speakers":3500000.0
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"target_language_name":"Bhojpuri",
|
| 29 |
+
"bleu":39.273082579,
|
| 30 |
"speakers":52200000.0
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"target_language_name":"Czech",
|
| 34 |
+
"bleu":66.754874801,
|
| 35 |
"speakers":10700000.0
|
| 36 |
},
|
| 37 |
{
|
| 38 |
"target_language_name":"Icelandic",
|
| 39 |
+
"bleu":56.0067334267,
|
| 40 |
"speakers":358000.0
|
| 41 |
},
|
| 42 |
{
|
| 43 |
"target_language_name":"Indonesian",
|
| 44 |
+
"bleu":73.5262667867,
|
| 45 |
"speakers":198996550.0
|
| 46 |
},
|
| 47 |
{
|
| 48 |
"target_language_name":"Kashmiri",
|
| 49 |
+
"bleu":21.7518148776,
|
| 50 |
"speakers":6900000.0
|
| 51 |
},
|
| 52 |
{
|
| 53 |
"target_language_name":"Lingala",
|
| 54 |
+
"bleu":33.4938226708,
|
| 55 |
"speakers":20000000.0
|
| 56 |
},
|
| 57 |
+
{
|
| 58 |
+
"target_language_name":"Nuer",
|
| 59 |
+
"bleu":5.2909507789,
|
| 60 |
+
"speakers":900000.0
|
| 61 |
+
},
|
| 62 |
{
|
| 63 |
"target_language_name":"Polish",
|
| 64 |
+
"bleu":63.1176806825,
|
| 65 |
"speakers":40200000.0
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"target_language_name":"Sango",
|
| 69 |
+
"bleu":19.6608440732,
|
| 70 |
"speakers":4600000.0
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"target_language_name":"Serbian",
|
| 74 |
+
"bleu":64.8198504575,
|
| 75 |
+
"speakers":9000000.0
|
| 76 |
}
|
| 77 |
]
|
uv.lock
CHANGED
|
@@ -93,6 +93,15 @@ wheels = [
|
|
| 93 |
{ url = "https://files.pythonhosted.org/packages/ae/63/3e1aee3e554263f3f1011cca50d78a4894ae16ce99bf78101ac3a2f0ef74/aiohttp-3.10.10-cp313-cp313-win_amd64.whl", hash = "sha256:486f7aabfa292719a2753c016cc3a8f8172965cabb3ea2e7f7436c7f5a22a151", size = 376785 },
|
| 94 |
]
|
| 95 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 96 |
[[package]]
|
| 97 |
name = "aiosignal"
|
| 98 |
version = "1.3.1"
|
|
@@ -535,6 +544,7 @@ name = "languagebench"
|
|
| 535 |
version = "0.1.0"
|
| 536 |
source = { virtual = "." }
|
| 537 |
dependencies = [
|
|
|
|
| 538 |
{ name = "evaluate" },
|
| 539 |
{ name = "joblib" },
|
| 540 |
{ name = "openai" },
|
|
@@ -546,6 +556,7 @@ dependencies = [
|
|
| 546 |
|
| 547 |
[package.metadata]
|
| 548 |
requires-dist = [
|
|
|
|
| 549 |
{ name = "evaluate", specifier = ">=0.4.3" },
|
| 550 |
{ name = "joblib", specifier = ">=1.4.2" },
|
| 551 |
{ name = "openai", specifier = ">=1.52.2" },
|
|
|
|
| 93 |
{ url = "https://files.pythonhosted.org/packages/ae/63/3e1aee3e554263f3f1011cca50d78a4894ae16ce99bf78101ac3a2f0ef74/aiohttp-3.10.10-cp313-cp313-win_amd64.whl", hash = "sha256:486f7aabfa292719a2753c016cc3a8f8172965cabb3ea2e7f7436c7f5a22a151", size = 376785 },
|
| 94 |
]
|
| 95 |
|
| 96 |
+
[[package]]
|
| 97 |
+
name = "aiolimiter"
|
| 98 |
+
version = "1.1.0"
|
| 99 |
+
source = { registry = "https://pypi.org/simple" }
|
| 100 |
+
sdist = { url = "https://files.pythonhosted.org/packages/df/62/6de944a6839a68f7d69e552e26d12234d9c556472e4c277a3a563013640a/aiolimiter-1.1.0.tar.gz", hash = "sha256:461cf02f82a29347340d031626c92853645c099cb5ff85577b831a7bd21132b5", size = 6229 }
|
| 101 |
+
wheels = [
|
| 102 |
+
{ url = "https://files.pythonhosted.org/packages/60/69/4b7dea755fafa10b248928da836a2cc8b5cff0762f363234e24218040f8e/aiolimiter-1.1.0-py3-none-any.whl", hash = "sha256:0b4997961fc58b8df40279e739f9cf0d3e255e63e9a44f64df567a8c17241e24", size = 7212 },
|
| 103 |
+
]
|
| 104 |
+
|
| 105 |
[[package]]
|
| 106 |
name = "aiosignal"
|
| 107 |
version = "1.3.1"
|
|
|
|
| 544 |
version = "0.1.0"
|
| 545 |
source = { virtual = "." }
|
| 546 |
dependencies = [
|
| 547 |
+
{ name = "aiolimiter" },
|
| 548 |
{ name = "evaluate" },
|
| 549 |
{ name = "joblib" },
|
| 550 |
{ name = "openai" },
|
|
|
|
| 556 |
|
| 557 |
[package.metadata]
|
| 558 |
requires-dist = [
|
| 559 |
+
{ name = "aiolimiter", specifier = ">=1.1.0" },
|
| 560 |
{ name = "evaluate", specifier = ">=0.4.3" },
|
| 561 |
{ name = "joblib", specifier = ">=1.4.2" },
|
| 562 |
{ name = "openai", specifier = ">=1.52.2" },
|