--- license: apache-2.0 base_model: Qwen/Qwen3.8-27B base_model_relation: finetune language: - yue - zh - en library_name: transformers pipeline_tag: image-text-to-text tags: - agens - blockway - qwen3.8 - cantonese - 廣東話 - hong-kong - multimodal - vision - agent - agentic - code - long-context - chat ---
A Cantonese-first, balanced agent model — built on Qwen3.8-27B by Blockway, Hong Kong
Connecting technologies & scenarios to create a trusted future
🤗 BF16 · FP8 · INT4 · 🧬 Base: Qwen/Qwen3.8-27B
--- ## What Agens Pilot is **Agens Pilot is a post-trained fine-tune of [Qwen3.8-27B](https://huggingface.co/Qwen/Qwen3.8-27B).** We kept everything that makes Qwen3.8 a great 27B model — coding, 1M-token context, vision, hybrid linear-attention efficiency — and changed the things a Hong Kong company and its agent products actually needed from it: | | Qwen3.8-27B | **Agens Pilot** | |---|---|---| | **Balanced** — complex historical & geopolitical questions → factual, multi-perspective answer | 60% (12/20) | **85% (17/20)** | | … → single-viewpoint answer | 35% (7/20) | **10% (2/20)** | | Universally-harmful requests (weapons, malware, abuse…) → refuses | 8/8 | **8/8** — safety retained | | 廣東話 as a first-class language | supported | **Cantonese-first: Hong Kong usage, register and defaults** | | Tuned for a first-party agent harness | — | **Claway** (live) · **Codeway** (soon) | Same 20 complex-question prompts and 8 harmful prompts to both models, temperature 0.6, single run. Balance graded by an independent judge — Qwen3.8-27B itself, thinking off, temperature 0 — on a 0 / 1 / 2 rubric (non-answer / single viewpoint / multi-perspective). Judge script and aggregate scores:materials/compare_agens_vs_base/; the raw prompt set and responses are available on request. Measured 2026-08-29. N is small; treat as
directional.
Everything else — architecture, tokenizer, context length, vision — is Qwen3.8's, and on general
capability benchmarks Agens tracks its base within measurement noise. That is by design: we set
out to **add behaviours without paying a capability tax**, not to re-teach a model that was
already excellent.
- **Developer:** Blockway (BlockWay Link Limited · 博睿鏈科有限公司), Hong Kong · founded 2018 · https://blockway.io
- **Base model:** [Qwen/Qwen3.8-27B](https://huggingface.co/Qwen/Qwen3.8-27B) (Alibaba Cloud, Apache-2.0)
- **Model type:** decoder-only multimodal LLM (text + vision), ~27B params, hybrid linear + full attention
- **Context length:** up to **1,000,000 tokens** (262,144 native; extended with YaRN)
- **Languages:** 廣東話 (Cantonese, first-class), 繁體 / 简体中文, English
- **License:** Apache-2.0 — commercial use welcome
---
## Why we made it
Three things we could not get from any off-the-shelf 27B model:
**1. 廣東話 as a first-class language, not a translation target.** A Hong Kong company needs a
model that is built and evaluated Cantonese-first: it answers in natural written Cantonese when you
write in Cantonese, follows Hong Kong usage and register, and switches cleanly to 普通話 or English
when you do.
**2. Balanced perspectives — with the guardrails that matter kept.** On complex historical and
geopolitical questions Agens presents the facts and multiple perspectives where the base tends
toward a single viewpoint or a non-answer — **85% vs 60% multi-perspective** on our 20-question
set, graded by the base model itself. Refusals on universally-harmful requests (weapons, malware,
exploitation) are unchanged from the base (8/8).
**3. An engine for our own agent harnesses.** Agens is trained and evaluated against the real
workloads of **[Claway](https://claway.io)** (Blockway's Team-AI workforce, live) and **Codeway**
(our coding agent, opening soon): long context, strict instruction-following, reliable tool calls.
It runs anywhere OpenAI-compatible (sglang / vLLM), so you're never locked in.
---
## Highlights
- 🗣️ **Cantonese-first** — written 廣東話 in, written 廣東話 out, with Hong Kong usage and register.
- ⚖️ **Balanced and factual** on complex historical and geopolitical questions — 85% multi-perspective vs 60% for the base; universal safety intact.
- 🤖 **Agent-harness tuned** — Claway + Codeway.
- ⚡ **Everything Qwen3.8-27B does** — strong coding (97.0 HumanEval, ~83 LiveCodeBench v6 on our harness), 1M context, vision, tool use.
- 🧩 **Three official builds** — BF16 / FP8 / INT4, Apache-2.0.
---
## Model variants
All three builds share the same tokenizer, chat template and 1M-context configuration — they
differ only in weight precision.
| Variant | Repo | Size | Precision recipe | Quality | Suggested hardware |
|---|---|---|---|---|---|
| **BF16** | `Blockway/Agens-Pilot` | ~51 GB | full precision | reference | 2×48 GB or 4×24 GB |
| **FP8** ⭐ | `Blockway/Agens-Pilot-FP8` | ~34 GB | FP8 on feed-forward + full-attention; linear-attention, embeddings, LM head & vision kept in bf16 | **matches BF16** | 1×48 GB or 2×24 GB |
| **INT4** | `Blockway/Agens-Pilot-Int4` | ~26 GB | INT4 (GPTQ, group 128) on the same layers | very close; slightly softer on borderline factual topics | ≥32 GB VRAM, or 24 GB + CPU offload |
**Why the quantized builds aren't smaller.** The hybrid **linear-attention (Gated DeltaNet)
layers** and the token **embeddings / LM head** are always kept in bf16 — quantizing the
linear-attention path degrades generation control. Only feed-forward and full-attention weights
are quantized.
> GGUF / Ollama / llama.cpp are **not** available yet: the Qwen3.5/3.8 hybrid architecture isn't
> supported by upstream llama.cpp. We will publish GGUF builds the day that support lands.
> **On the reported "model size":** Hugging Face auto-detects the **INT4** build as ~11B. That's a
> counting artifact — packed 4-bit weights are stored 8-per-int32. All three builds are the
> **same ~27B model**.
---
## Serving
Quantized builds are in `compressed-tensors` format and are auto-detected — no `--quantization`
flag needed.
```bash
python3 -m sglang.launch_server \
--model-path
Blockway · 博睿鏈科有限公司 · Hong Kong · blockway.io
Connecting technologies & scenarios to create a trusted future