Initial Query
For each of the following ten movies give at most 5 emoji that would best describe the movie.
Give your answer as a JSON array. So If I asked for
you might might answer
| {"Finding Nemo": ["ðŸ ", "ðŸŸ", "ðŸ¡", "ðŸ¬", "ðŸ³"],
"Toy Story": ["🚀", "âš”ï¸,", "🤖", "👽", "🌌"]}
|
.
Each emoji must be a single utf8 character. ABSOLUTELY NO ZERO WIDTH JOINING. (So, max(len(emoji) for movie in result.values() for emoji in movie) == 1)
Now give me answers for these movies:
Give ONLY a JSON output. Nothing else.
LLM Generation
Query
For each of the following ten movies give at most 5 emoji that would best describe the movie.
Give your answer as a JSON array. So If I asked for
you might might answer
| {"Finding Nemo": ["ðŸ ", "ðŸŸ", "ðŸ¡", "ðŸ¬", "ðŸ³"],
"Toy Story": ["🚀", "âš”ï¸,", "🤖", "👽", "🌌"]}
|
.
Each emoji must be a single utf8 character. ABSOLUTELY NO ZERO WIDTH JOINING. (So, max(len(emoji) for movie in result.values() for emoji in movie) == 1)
Now give me answers for these movies:
Give ONLY a JSON output. Nothing else.
Output
1
2
3
4
5
6
7
8
9
10
11
12 | {
"The Lion King": ["ðŸ¦", "👑", "🌅", "ðŸ—", "🌿"],
"The Nightmare Before Christmas": ["🎃", "🎄", "👻", "🕷ï¸", "🎅"],
"The Godfather": ["ðŸ·", "🔫", "ðŸ’", "🕴ï¸", "🖤"],
"The Matrix": ["💊", "🕶ï¸", "🔮", "🤖", "ðŸŒ"],
"Casablanca": ["✈ï¸", "🎶", "💔", "🕴ï¸", "🌹"],
"Raiders of the Lost Ark": ["ðŸ§", "💎", "ðŸ", "🤠", "âš”ï¸"],
"V for Vendetta": ["ðŸŽ", "💣", "🔥", "🕊ï¸", "âš–ï¸"],
"The Princess Bride": ["âš”ï¸", "👑", "💋", "ðŸ€", "â¤ï¸"],
"Back to the Future": ["🚗", "â°", "âš¡", "👨â€ðŸ”¬", "🕰ï¸"],
"Dune": ["ðŸœï¸", "ðŸ›", "👑", "âš”ï¸", "🌌"]
}
|
Extract Json
I extracted the following JSON from that output:
1
2
3
4
5
6
7
8
9
10
11
12 | {
"The Lion King": ["ðŸ¦", "👑", "🌅", "ðŸ—", "🌿"],
"The Nightmare Before Christmas": ["🎃", "🎄", "👻", "🕷ï¸", "🎅"],
"The Godfather": ["ðŸ·", "🔫", "ðŸ’", "🕴ï¸", "🖤"],
"The Matrix": ["💊", "🕶ï¸", "🔮", "🤖", "ðŸŒ"],
"Casablanca": ["✈ï¸", "🎶", "💔", "🕴ï¸", "🌹"],
"Raiders of the Lost Ark": ["ðŸ§", "💎", "ðŸ", "🤠", "âš”ï¸"],
"V for Vendetta": ["ðŸŽ", "💣", "🔥", "🕊ï¸", "âš–ï¸"],
"The Princess Bride": ["âš”ï¸", "👑", "💋", "ðŸ€", "â¤ï¸"],
"Back to the Future": ["🚗", "â°", "âš¡", "👨â€ðŸ”¬", "🕰ï¸"],
"Dune": ["ðŸœï¸", "ðŸ›", "👑", "âš”ï¸", "🌌"]
}
|
PyFunc
Resulting in output:
[['ðŸ¦', '👑', '🌅', 'ðŸ—', '🌿'],
['🎃', '🎄', '👻', '🕷ï¸', '🎅'],
['ðŸ·', '🔫', 'ðŸ’', '🕴ï¸', '🖤'],
['💊', '🕶ï¸', '🔮', '🤖', 'ðŸŒ'],
['✈ï¸', '🎶', '💔', '🕴ï¸', '🌹'],
['ðŸ§', '💎', 'ðŸ', '🤠', 'âš”ï¸'],
['ðŸŽ', '💣', '🔥', '🕊ï¸', 'âš–ï¸'],
['âš”ï¸', '👑', '💋', 'ðŸ€', 'â¤ï¸'],
['🚗', 'â°', 'âš¡', '🕰ï¸'],
['ðŸœï¸', 'ðŸ›', '👑', 'âš”ï¸', '🌌']]
LLM Generation
Query
For each of the following ten movies described by 5 emoji, give the movie title that best matches.
Give your answer as a JSON list. So If I asked for
You might answer
| ["Finding Nemo", "Toy Story"]]
|
.
Now give me answers for these movies:
| ['🎃', '🎄', '👻', '🕷ï¸', '🎅'],
['ðŸ·', '🔫', 'ðŸ’', '🕴ï¸', '🖤'],
['💊', '🕶ï¸', '🔮', '🤖', 'ðŸŒ'],
['✈ï¸', '🎶', '💔', '🕴ï¸', '🌹'],
['ðŸ§', '💎', 'ðŸ', '🤠', 'âš”ï¸'],
['ðŸŽ', '💣', '🔥', '🕊ï¸', 'âš–ï¸'],
['âš”ï¸', '👑', '💋', 'ðŸ€', 'â¤ï¸'],
['🚗', 'â°', 'âš¡', '🕰ï¸'],
['ðŸœï¸', 'ðŸ›', '👑', 'âš”ï¸', '🌌']]
|
What are the names of the movie titles?
Output
| ["The Lion King", "The Nightmare Before Christmas", "The Godfather", "The Matrix", "Casablanca", "Indiana Jones and the Raiders of the Lost Ark", "Les Misérables", "The Princess Bride", "Back to the Future", "Dune"]
|
Extract Json
I extracted the following JSON from that output:
| ["The Lion King", "The Nightmare Before Christmas", "The Godfather", "The Matrix", "Casablanca", "Indiana Jones and the Raiders of the Lost Ark", "Les Misérables", "The Princess Bride", "Back to the Future", "Dune"]
|
PyFunc
OK
Resulting in output:
True