---

license: apache-2.0

base_model: openbmb/MiniCPM5-1B

tags:

• llama-cpp
• gguf
• text-generation
• lora
• build-small-hackathon
• thousand-token-wood

datasets:

• MarianaCodebase/lost-frequency-radio-transmissions

language:

• es
• en
• fr

pipeline_tag: text-generation

---

MiniCPM5-1B — Lost Frequency Radio (GGUF Q4_K_M)

LoRA fine-tune of openbmb/MiniCPM5-1B for Lost Frequency Radio, an interactive radio from parallel universes built for the Hugging Face Build Small Hackathon 2026 (track 🍄 An Adventure in Thousand Token Wood).

Demo: Lost Frequency Radio Space

Dataset: ~786 surreal radio transmissions (es / en / fr) with structured tokens, template-generated and hand-curated.

Task: write short in-character radio scripts (60 to 90 words): 1950s announcers, Jupiter weather reports, impossible commercials, number stations, late-night cross-universe call-in shows.

Anti prompt-leak design: the system prompts contain no instruction-shaped rules ("write only the script, 60-90 words..."). The format is learned purely from the completions, so a 1B model has nothing instruction-shaped to "recite" on air. French is taught at the weight level through its own slice of the dataset.

Structured tokens

The model emits markers that the frontend turns into audiovisual events:

| Token | Effect on the radio |

|---|---|

| [JINGLE] | light pulse + arpeggio |

| [INTERFERENCIA] | screen glitch + burst of static |

| [CORTE COMERCIAL] | click + dimming |

| [FIN DE TRANSMISION] | display fade and signal drop |

Usage with llama.cpp

from llama_cpp import Llama

llm = Llama(model_path="MiniCPM5-1B-lost-frequency-radio-Q4_K_M.gguf", n_ctx=2048)

prompt = (
    "<s><|im_start|>system\nYou are the official voice of the Jupiter Weather "
    "Service, year 2187. You write radio scripts in Spanish.<|im_end|>\n"
    "<|im_start|>user\nWrite tonight's transmission. On-air script only."
    "<|im_end|>\n<|im_start|>assistant\n<think>\n\n</think>\n\n"
)

tokens = llm.tokenize(prompt.encode(), add_bos=False, special=True)
out = llm.create_completion(prompt=tokens, max_tokens=220, temperature=0.7,
                            stop=["<|im_end|>"])
print(out["choices"][0]["text"])

Note: the <think>\n\n</think>\n\n prefill disables MiniCPM5's reasoning mode (equivalent to enable_thinking=False in the chat template).

Training

• 786 examples (es / en / fr), LoRA r=16, alpha=32, dropout 0.05, applied to every projection (q/k/v/o/gate/up/down)
• 3 epochs, lr 1e-4 cosine, bf16, max_length 768
• Hardware: a single RTX 4050 laptop (6 GB), the model is tiny by design
• Final loss ≈ 0.36-0.42, token accuracy ≈ 0.92

Files

• MiniCPM5-1B-lost-frequency-radio-Q4_K_M.gguf — Q4_K_M quantization (~651 MB), the one the Space uses
• lora-adapter/ — LoRA adapters (to reproduce or continue training)