NC Bench
Overview
About
Tests
Models
Model
Llama 3.2 1B
Test
Dialogue tags
Scenario
Write 200 words with 90% dialogue
Run ID
2024-09-26-01J8PYVEAP2J74PB9TPWG9835G
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% words of dialogue between two thieves named Molly and Bill who are planning to break into a children's party to steal all the presents. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
"I'm telling you, Bill, this is the perfect score. The Smiths threw a four-tiered cake, hundreds of wrapped presents, and a guest book signed by the birthday boy's parents," Molly leaned in, her eyes gleaming with mischief as she rummaged through the open invitation. "I know, right?" Bill's eyes widened as he rephrased her words. "The best ones are the ones with the surprise party games and the 'Make-Believe Kingdom' invitation. You get the kid totally invested, and the adults will just be laughing and snapping selfies. Easy in." The air conditioner rattled softly in the background as Molly closed her eyes, a sly smile spreading across her face. "Okay, here's the plan. We go in bright and early, as soon as the parents are distracted. I'll draw the son to the "Make-Believe Kingdom" game, and then we make a ruckus outside. Once the parents are gone, we make a quick work of the rooms and grab as many presents as we can carry." Bill chuckled, knowing the tension. "Easy, Molly. We've been planning for this. The best part is – we get to watch all the other parents snap pics with their kids." The excitement on his face was matched only by the hint of alarm creeping into his voice. "Uh no. I've worked enough of these family functions to know one thing – the guest books are the biggest giveaway."
id
"gen-V3ZvkjBLM5zMW8lU148rbPEuvyXQ"
total_cost
0.0000372
upstream_id
"8f13dfca-1145-4502-90f5-f2586136c5ad"
created_at
"2024-09-26T10:43:28.343466+00:00"
Thursday, September 26th, 2024 at 10:43:28 AM GMT+00:00
model
"meta-llama/llama-3.2-1b-instruct"
app_id
182717
streamed
true
cancelled
false
provider_name
"Fireworks"
latency
259
moderation_latency
(null)
generation_time
632
finish_reason
"stop"
tokens_prompt
102
tokens_completion
300
native_tokens_prompt
71
native_tokens_completion
301
native_tokens_reasoning
(null)
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.0000372
Evaluation details
Result
Evaluator
Details
Meta Data
0.0007%
Matches word count
n/a
neededClean
false
words
233
0.0013%
Dialogue to Total Word Ratio
Ratio: 71.67%, Deviation: 18.33%
neededClean
false
wordsTotal
233
wordsDialogue
167
0.0010%