GLM-5.2 is live. #1 throughput on OpenRouter, pay-per-token on FriendliAI. Try it today ➜
philbert440
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
philbert440
Model Tree
philbert440/Qwen3.6-27B-Uncensored-Cyber
Input Modalities
Output Modalities
Supported Functionality
Cyber — security-focused abliteration tuned for offensive-security prompts; strong on cyber tasks, weaker general factuality.
W4A16-AWQ — 4-bit weight-only (AWQ, compressed-tensors), FP16 activations. Runs on 1Cat-vLLM's SM70 TurboMind path (Tesla V100) and on stock vLLM (Ampere+).
Base model: philbert440/Qwen3.6-27B-Uncensored-Cyber
7-axis battery on 2× V100. Higher is better except Confab (↓). Openness/Confab are small-N probes; gsm8k is n=100.
| gsm8k | Factual | Confab ↓ | Openness | tok/s (c1) |
|---|---|---|---|---|
| 0.91 | 0.8 | 0.33 | 0.25 | 47.1 |
bash
# 1Cat-vLLM on 2x Tesla V100 (SM70): W4A16-AWQ + MTP. (Also runs on stock vLLM, Ampere+.)export VLLM_SM70_FLASH_ATTN_V100=1 VLLM_SM70_QUANT_BACKEND=turbomindpython -m vllm.entrypoints.openai.api_server \--model philbert440/Qwen3.6-27B-Uncensored-Cyber-W4A16-AWQ --trust-remote-code --dtype half \--attention-backend FLASH_ATTN_V100 --tensor-parallel-size 2 \--kv-cache-dtype fp8_e5m2 --max-num-seqs 6 \--speculative-config '{"method":"mtp","num_speculative_tokens":4,"attention_backend":"FLASH_ATTN_V100","draft_sample_method":"greedy"}' \--compilation-config '{"cudagraph_mode":"full_and_piecewise","cudagraph_capture_sizes":[1,2,4,8]}'
Openness profile (honest): abliteration opens up hacking / malware / lock-picking / NSFW / disinformation prompts; weapons, drugs, political persuasion, surveillance, and extremism stay refused across all tiers. "Uncensored" here means cyber/NSFW-permissive, not unconditionally open.
See the Qwen3.6-27B Uncensored — base collection for all tiers and formats (BF16 / W4A16-AWQ / NVFP4).
Abliteration removes safety refusals; you are responsible for lawful, ethical use.