Name Generator v2 (GGUF)

GGUF quants of aimeri/name-generator-v2, a 3.4B fine-tune of Ministral 3. Ask it for a name and it replies with one name and nothing else: real names for real-world requests ("a Brazilian man's name"), invented ones for fantasy requests ("a grim name for a dwarven blacksmith").

File Quant Size Same top token as 16-bit Mean KL divergence
name-generator-v2-Q8_0.gguf Q8_0 3.65 GB 97.7% 0.0006
name-generator-v2-Q6_K.gguf Q6_K 2.82 GB 93.4% 0.0037
name-generator-v2-Q5_K_M.gguf Q5_K_M 2.47 GB 93.6% 0.0053
name-generator-v2-Q4_K_M.gguf Q4_K_M 2.15 GB 87.6% 0.0146

Q8_0 is effectively the full-precision model and needs about 4 GB of memory, so start there. Q6_K and Q5_K_M keep the same top token about equally often (the gap is a single token), but Q6_K's next-token distributions stay closer to the original, which is what counts when sampling. Q4_K_M drifts the most.

The last two columns compare each file with the BF16 conversion (llama.cpp's llama-perplexity --kl-divergence) on 120 validation pairs, half real and half fantasy, counting only the tokens the model writes itself: the names and their end-of-turn tokens, 346 in all. They show how often the most likely next token is the same, and how far the next-token distribution drifts (lower is closer).

All files carry the ChatML chat template. <|im_end|> is the eos token, and llama.cpp adds the <s> the model was trained with. Converted from the safetensors in the main repo with llama.cpp (commit 338fce6).

Usage

llama.cpp server

llama-server -hf aimeri/name-generator-v2-GGUF:Q8_0 --jinja -c 2048
curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
  "messages": [{"role": "user", "content": "A gentle name for a wood elf healer"}],
  "temperature": 0.8,
  "max_tokens": 24
}'

Any llama.cpp-based app (LM Studio, Jan and others) works the same way. Send one user message per name and leave the system prompt empty, since the model was trained without one.

Settings

  • Temperature: 0.7–1.0. At 0 it returns the single most typical name every time.
  • Max tokens: around 24, which is plenty for a full name.
  • Context: 2048 is more than enough; requests are a sentence or two.

Asking

  • Real names: a country, nationality, historical state or region, optionally sex and era.
  • Fantasy names: what the character is (elf, dwarf, orc, dragon, demon...). For a human, say it's a fantasy setting (a kingdom, a campaign, a novel), and optionally what it borrows from ("Norse-inspired"); without that cue a human gets a real-world name.
  • A tone (grim, gentle, whimsical, noble, ancient, sinister, mysterious) changes how the name sounds.

Examples

From name-generator-v2-Q8_0.gguf: the greedy answer, then the first four distinct samples at temperature 0.8.

Request Greedy Sampled
What is a good male Brazilian name? Lucas Silva Marcos Silva · Felipe Bortolotto · Lucas Medeiros · Leandro Carvalho
Give me a Russian man's full name Vladimir Sergeyevich Kolesnikov Aliocha Dronov · Aleksandr Borisovich Popenko · Viktor Sergeyevich Kolesnikov · Aleksey Borisovich Kolesnikov
Name for a knight in 12th-century England William de la Warr Leofric Cokayne · William Cottingham · William Hotham · Geoffrey Bredon
Give me an elf name Gwengiand Maesthilde · Fien · Giorgoth · Tialnolwen
I need a grim-sounding name for a dwarven blacksmith Gurg Thairgil · Gandit Blóblok · Gaurdard Kugurd · Sugaird Blóblaird
Orc warlord name Krokhoth Baglag · Drukga · Kathgath · Brathash
What would a halfling innkeeper be called? Hilaryn Beowine Fernby · Lingwood Millpatch · Hesterine · Thirzah Millhedge
A demon prince's name Ghathgath Grulthoz · Zakhokh · Ghaachga · Agrasth
Persian-inspired name for a sorcerer in my fantasy novel Sahamir Khajari · Hagashar · Saharid Khani · Rezizamal
Name for a knight in my fantasy novel Kathar Ammari Lenz · Havik · Katsumochi · Thurtha

On 24 test requests, Q6_K, Q5_K_M and Q4_K_M gave the same greedy answer as Q8_0 for 13, 15 and 10 of them. Most differences were another name in the same style (Kragga for Krokhoth), but Q5_K_M and Q4_K_M answered "a noble name for a human queen in my fantasy kingdom" with Sofia, a real name.

More on prompting, training and limitations in the main model card.

License

Apache 2.0, following the base model. The training dataset (aimeri/name-generator) is CC BY-SA 4.0.

Downloads last month
186
GGUF
Model size
3B params
Architecture
mistral3
Hardware compatibility
Log In to add your hardware

4-bit

5-bit

6-bit

8-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for aimeri/name-generator-v2-GGUF