|
|
|
|
|
|
|
|
def generate_behavior_samples(dictionaries, count=50, source_language="en"): |
|
|
samples = [] |
|
|
|
|
|
behaviors = [ |
|
|
("warm", "steady", "medium", "natural", "high"), |
|
|
("neutral", "slow", "shallow", "formal", "medium"), |
|
|
("direct", "fast", "deep", "technical", "high"), |
|
|
("friendly", "steady", "medium", "casual", "high"), |
|
|
("precise", "slow", "deep", "structured", "high") |
|
|
] |
|
|
|
|
|
for i in range(count): |
|
|
tone, pacing, depth, style, clarity = behaviors[i % len(behaviors)] |
|
|
|
|
|
user_text = "Explain this to me." |
|
|
glyphic = f"BEHAVIOR(tone={tone}, pacing={pacing}, depth={depth})" |
|
|
realized = f"I will explain this in a {tone} and {clarity} way." |
|
|
|
|
|
samples.append({ |
|
|
"input": user_text, |
|
|
"glyphic": glyphic, |
|
|
"output": realized, |
|
|
"behavior": { |
|
|
"tone": tone, |
|
|
"pacing": pacing, |
|
|
"depth": depth, |
|
|
"style": style, |
|
|
"clarity": clarity |
|
|
}, |
|
|
"emotion": "neutral", |
|
|
"sensory": "none", |
|
|
"social": "alone" |
|
|
}) |
|
|
|
|
|
return samples |
|
|
|
|
|
|
