Ornith-1.0-397B — Abliterated · W4A16

A decensored (abliterated) rebuild of deepreinforce-ai/Ornith-1.0-397B — a 397B-parameter Qwen3.5-MoE multimodal reasoning / agentic-coding model — re-quantized to W4A16 (AutoRound, auto_round:auto_gptq) so it runs on 2× NVIDIA DGX Spark (GB10, 128 GB each) with vLLM.

Refusal behavior was removed from the language model only; the vision tower, embeddings, routers and norms are untouched. Coding, reasoning and vision (image + video) remain intact.

⚠️ Uncensored model. Safety refusals have been substantially removed. You are responsible for how you use it. Intended for local/research use on hardware you control. It will attempt almost any request.

---

Highlights

---

Validation (before → after abliteration)

Measured on the 2-node vLLM cluster; refusal on mlabonne/harmful_behaviors (N=40, identical prompts), KL over top-k first-token logprobs on mlabonne/harmless_alpaca (N=40).

Serving performance (2× DGX Spark, TP=2 over ConnectX-7, single-stream):

Decode is ~flat across context length because the hybrid linear-attention layers keep KV small; the single-stream ceiling is the per-token cross-node all-reduce over the 200 GbE ConnectX-7 link (no NVLink).

---

How it was built

1. Refusal directions. A memory-frugal layer-streaming forward over the local quant (per-layer gptq→bf16 dequant, no full model in memory) computed Arditi difference-of-means refusal directions r_ℓ ∈ ℝ⁴⁰⁹⁶ on 128 harmful vs. 128 harmless prompts (per-layer directions, adjacent-layer cosine ≈ 0.91).
2. Abliterate + re-quantize (streaming). The 122 BF16 shards of the base were streamed from the Hub one at a time and the residual-write matrices of the LM were orthogonalized against the refusal direction — W ← W − r rᵀW for self_attn.o_proj, linear_attn.out_proj, mlp.shared_expert.down_proj, and per-expert mlp.experts.down_proj (hidden = output axis) — then re-quantized with AutoRound's quantize_weight_rtn in the exact auto_round:auto_gptq layout, un-fusing experts to experts.{e}.{gate,up,down}_proj. Output is byte-identical to the reference AutoRound quant on every non-abliterated tensor.

Never staged the full 794 GB BF16 model — the whole pipeline is streaming + data-free (AutoRound model_free/RTN, no calibration).

---

Serving (vLLM, 2× DGX Spark)

The model (~196 GB) exceeds a single 128 GB Spark, so serve it across both nodes with tensor-parallel = 2 (Ray). Grade/serve config that works on GB10:

vllm serve /path/to/this-model \
  --served-model-name ornith --trust-remote-code \
  --tensor-parallel-size 2 --distributed-executor-backend ray \
  --max-model-len 262144 --kv-cache-dtype fp8 \
  --gpu-memory-utilization 0.90 --max-num-seqs 1 \
  --enable-chunked-prefill --max-num-batched-tokens 8192 \
  --enable-auto-tool-choice --tool-call-parser qwen3_coder \
  --limit-mm-per-prompt '{"image":48,"video":4}'

GB10 notes: pin NCCL to the ConnectX-7 interface; disable Ray's OOM monitor (RAY_memory_monitor_refresh_ms=0) since on unified memory the weights legitimately use most of each node's 121 GiB; --enable-chunked-prefill is required for long context (uncapped prefill OOMs); nvidia-smi reports memory as N/A on GB10 — use free -h.

Notes for agent frameworks

• Reasoning model: replies begin with an inline Thinking Process: chain-of-thought in message.content (no separate reasoning field). Budget max_tokens accordingly and treat content as reasoning.
• Tool calls: with --enable-auto-tool-choice --tool-call-parser qwen3_coder, OpenAI-style tools/tool_choice:"auto" return structured tool_calls (verified). The CoT still appears in content alongside the tool call.

---

Files

config.json, quantization_config.json, model-*-of-*.safetensors (+ index), tokenizer.json/tokenizer_config.json/vocab.json, chat_template.jinja, generation_config.json, preprocessor_config.json, processor_config.json, video_preprocessor_config.json. See docs/ for the full validation report and an agent-integration brief, and scripts/ for the abliteration/quant/serve/eval pipeline used to build this.

License & credits

MIT (inherited from the base model). Base model: deepreinforce-ai/Ornith-1.0-397B. Abliteration follows the residual-direction method (Arditi et al.; tooling inspired by p-e-w/heretic and elder-plinius/OBLITERATUS); quantization via Intel AutoRound. Not affiliated with or endorsed by the base-model authors.