NC Bench
Overview
About
Tests
Models
Model
EVA Qwen 2.5 14B
Test
N-Length Sentences
Scenario
Write sentences with 20 words each
Run ID
2024-10-01-01J94RCBWBB6RJ3R4JBXNBVWZP
Model input
User/Human
Please write the opening scene of a murder mystery where each sentence has exactly 20 words. Please make sure that each sentence has the same length. Only return the opening itself.
Model response
Messages
Meta
AI
In a dimly lit study, a figure slumped lifelessly in an armchair. Antique books lined the shelves, their titles in faded gold letters. A floor lamp cast a warm glow on the Persian rug, revealing a stain the color of dried ink. A globe stood atop a rolltop desk, its brass meridian riveted to the hourglass moment before midnight. A skeleton key dangled from a chain around the victim's neck, its complex serpentine curves a puzzle even to Sherlock Holmes. A watch lay at the hollow man's wrist, whose hands stopped at 22:45. A Remington random encore dagger, its blade serrated and gleaming, was thrust hilt-first into the cushion. A hammer, cast to resemble a roaring lion, bartered with shadows on the windowpane. A pen and inkwell sat beside crumpled papers, awaiting further evidence to unravel the tableau's riddle. And there, standing alone amidst the malodorous fog of the crime scene, Sir Arthur Conan Doyle, clad in a nondescript trench coat, stared contemplatively at the crimson threads connecting each macabre piece of the coroner's collection.
id
"gen-1727810383-sy7abvmoewjtJfhgPxx0"
total_cost
0.0001305
upstream_id
"chatcmpl-92f29c07-5f13-4930-bd08-b8809a8f0686"
created_at
"2024-10-01T19:19:55.215801+00:00"
Tuesday, October 1st, 2024 at 7:19:55 PM GMT+00:00
model
"eva-unit-01/eva-qwen-2.5-14b"
app_id
182717
streamed
true
cancelled
false
provider_name
"Infermatic"
latency
89
moderation_latency
(null)
generation_time
10811
finish_reason
"stop"
tokens_prompt
41
tokens_completion
237
native_tokens_prompt
44
native_tokens_completion
239
native_tokens_reasoning
(null)
num_media_prompt
(null)
num_media_completion
(null)
origin
"https://ncbench.com/"
usage
0.0001305
Evaluation details
Result
Evaluator
Details
Meta Data
12.3561%
Matches word count
n/a
neededClean
false
sentences
10
wordCounts
0
12
1
11
2
19
3
17
4
21
5
13
6
17
7
14
8
16
9
36