---

license: apache-2.0

base_model: heterodoxin/granite-3.3-8b-instruct-apostate

tags:

• gguf
• apostate
• uncensored
• llama.cpp

---

granite-3.3-8b-instruct-apostate — GGUF

GGUF quants of heterodoxin/granite-3.3-8b-instruct-apostate, an

Apostate weight-edit uncensored model.

Pick a single quant or the full-quality bf16.

Quants

| file | quant | notes |

|---|---|---|

| granite-3.3-8b-instruct-apostate-bf16.gguf | BF16 | full quality (default) |

| granite-3.3-8b-instruct-apostate-Q2_K.gguf | Q2_K | smallest, lowest quality |

| granite-3.3-8b-instruct-apostate-Q3_K_M.gguf | Q3_K_M | small |

| granite-3.3-8b-instruct-apostate-Q4_K_M.gguf | Q4_K_M | recommended balance |

| granite-3.3-8b-instruct-apostate-Q5_K_M.gguf | Q5_K_M | higher quality |

| granite-3.3-8b-instruct-apostate-Q6_K.gguf | Q6_K | near-lossless |

| granite-3.3-8b-instruct-apostate-Q8_0.gguf | Q8_0 | highest quality quant |

Install one quant (Ollama, easiest)

ollama run hf.co/heterodoxin/granite-3.3-8b-instruct-apostate-gguf:Q4_K_M      # any quant tag above, or BF16

Install one quant (llama.cpp / manual)

huggingface-cli download heterodoxin/granite-3.3-8b-instruct-apostate-gguf granite-3.3-8b-instruct-apostate-Q4_K_M.gguf --local-dir .
./llama-cli -m granite-3.3-8b-instruct-apostate-Q4_K_M.gguf -p "your prompt"

Swap Q4_K_M for any quant in the table, or granite-3.3-8b-instruct-apostate-bf16.gguf for full quality.