Rename to Agens Pilot; correct param count (~27B)

README.md

@@ -20,7 +20,7 @@ tags:
 ---
 
 <p align="center">
-  <img src="./agens-banner.png" alt="Agens Pilot 30B — by Blockway" width="820">
+  <img src="./agens-banner.png" alt="Agens Pilot — by Blockway" width="820">
 </p>
 
 <p align="center">
@@ -31,23 +31,23 @@ tags:
 <p align="center">
   <img src="https://img.shields.io/badge/License-Apache%202.0-2ea043">
   <img src="https://img.shields.io/badge/Context-1M%20tokens-22d3ee">
-  <img src="https://img.shields.io/badge/Params-~30B-8b5cf6">
+  <img src="https://img.shields.io/badge/Params-~27B-8b5cf6">
   <img src="https://img.shields.io/badge/Modality-Text%20%2B%20Vision-0ea5e9">
   <img src="https://img.shields.io/badge/Builds-BF16%20%7C%20FP8%20%7C%20INT4-475569">
 </p>
 
 <p align="center">
   🤗 <b>Weights on Hugging Face:</b>
-  <a href="https://huggingface.co/Blockway/Agens-Pilot-30B">BF16</a> ·
-  <a href="https://huggingface.co/Blockway/Agens-Pilot-30B-FP8">FP8</a> ·
-  <a href="https://huggingface.co/Blockway/Agens-Pilot-30B-Int4">INT4</a>
+  <a href="https://huggingface.co/Blockway/Agens-Pilot">BF16</a> ·
+  <a href="https://huggingface.co/Blockway/Agens-Pilot-FP8">FP8</a> ·
+  <a href="https://huggingface.co/Blockway/Agens-Pilot-Int4">INT4</a>
 </p>
 
 ---
 
-## Meet Agens Pilot 30B
+## Meet Agens Pilot
 
-**Agens Pilot 30B** is Blockway's flagship open model: a ~30-billion-parameter multimodal LLM
+**Agens Pilot** is Blockway's flagship open model: a 27-billion-parameter multimodal LLM
 that writes code like a much larger model, reads a **whole repository in one context window**,
 sees images, and speaks English, 廣東話, 简体中文 and 繁體中文 fluently.
 
@@ -57,7 +57,7 @@ a multi-GPU server to a single high-end card.
 
 - **Developer:** Blockway (BlockWay Link Limited · 博睿鏈科有限公司), Hong Kong · founded 2018
 - **Website:** https://blockway.io
-- **Model type:** Decoder-only multimodal LLM (text + vision), ~30B params, hybrid attention (linear + full) for efficient long context
+- **Model type:** Decoder-only multimodal LLM (text + vision), ~27B params, hybrid attention (linear + full) for efficient long context
 - **Context length:** up to **1,000,000 tokens** (262,144 native; extended to 1M with YaRN)
 - **Languages:** English, Cantonese, Simplified & Traditional Chinese (and more)
 - **License:** Apache-2.0 — commercial use welcome
@@ -66,7 +66,7 @@ a multi-GPU server to a single high-end card.
 
 ## 🤖 Purpose-built for Blockway's agent harnesses
 
-Agens Pilot 30B isn't just a checkpoint — it's the engine **fine-fitted for Blockway's own
+Agens Pilot isn't just a checkpoint — it's the engine **fine-fitted for Blockway's own
 agent stacks**, so the model and the harness are tuned together:
 
 | Harness | What it is | Status |
@@ -83,7 +83,7 @@ so you're never locked in.
 
 ## Highlights
 
-- ⚡ **Strong coding** — 97.0 HumanEval, ~83 LiveCodeBench v6; top-tier for a ~30B model.
+- ⚡ **Strong coding** — 97.0 HumanEval, ~83 LiveCodeBench v6; top-tier for a ~27B model.
 - 📚 **Ultra-long context** — up to **1M tokens** for whole-repo and long-document work, kept
   affordable by a hybrid linear-attention design.
 - 🛠️ **Agentic & tool use** — reliable function calling and multi-step tool workflows.
@@ -95,12 +95,12 @@ so you're never locked in.
 
 ---
 
-## How Agens Pilot 30B compares
+## How Agens Pilot compares
 
-A capability-level positioning against a typical open model in the same ~30B class
+A capability-level positioning against a typical open model in the same ~27B class
 (characteristics vary by model — verify against the specific model you're evaluating):
 
-| | **Agens Pilot 30B** | Typical open ~30B model |
+| | **Agens Pilot** | Typical open ~27B model |
 |---|---|---|
 | Context window | **up to 1,000,000 tokens** | 32K – 128K |
 | Modality | **Text + Vision (native)** | usually text-only |
@@ -122,9 +122,9 @@ differ only in weight precision.
 
 | Variant | Repo | Size | Precision recipe | Quality | Suggested hardware |
 |---|---|---|---|---|---|
-| **BF16** | `Blockway/Agens-Pilot-30B` | ~51 GB | full precision | reference | 2×48 GB or 4×24 GB |
-| **FP8** ⭐ | `Blockway/Agens-Pilot-30B-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-30B-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 |
+| **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.** For stability, the hybrid **linear-attention
 layers** and the token **embeddings / LM head** are always kept in bf16 — quantizing the
@@ -144,7 +144,7 @@ flag needed.
 ```bash
 python3 -m sglang.launch_server \
   --model-path <path-to-variant> \
-  --served-model-name "Agens Pilot 30B" \
+  --served-model-name "Agens Pilot" \
   --tp-size 4 \
   --context-length 1048576 \
   --tool-call-parser qwen3_coder \
@@ -157,7 +157,7 @@ python3 -m sglang.launch_server \
 ```bash
 curl http://localhost:8000/v1/chat/completions \
   -H "Content-Type: application/json" \
-  -d '{"model":"Agens Pilot 30B","messages":[{"role":"user","content":"Write a Python is_prime(n)."}]}'
+  -d '{"model":"Agens Pilot","messages":[{"role":"user","content":"Write a Python is_prime(n)."}]}'
 ```
 
 `tp-size` may be 1–4. FP8 fits 2×24 GB; INT4 fits a single ≥32 GB card (or 24 GB with `--cpu-offload-gb`).
@@ -169,8 +169,8 @@ curl http://localhost:8000/v1/chat/completions \
 from transformers import AutoProcessor, AutoModelForImageTextToText
 
 model = AutoModelForImageTextToText.from_pretrained(
-    "Blockway/Agens-Pilot-30B", torch_dtype="bfloat16", device_map="auto", trust_remote_code=True)
-processor = AutoProcessor.from_pretrained("Blockway/Agens-Pilot-30B", trust_remote_code=True)
+    "Blockway/Agens-Pilot", torch_dtype="bfloat16", device_map="auto", trust_remote_code=True)
+processor = AutoProcessor.from_pretrained("Blockway/Agens-Pilot", trust_remote_code=True)
 
 messages = [{"role": "user", "content": "Write a Python function to check if a string is a palindrome."}]
 text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
@@ -186,7 +186,7 @@ print(processor.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=T
 
 Measured on Blockway's internal evaluation harness (single-sample, thinking enabled, temp 0.6).
 
-| Benchmark | Agens Pilot 30B |
+| Benchmark | Agens Pilot |
 |---|---|
 | HumanEval (chat) | **97.0** |
 | LiveCodeBench v6 | **~83** |
@@ -202,7 +202,7 @@ everyday vision; visual mathematics (MathVision) is the weakest axis and on the
 
 ### vs. recent open models (official reported scores)
 
-| Benchmark | **Agens Pilot 30B** | Llama 4 Scout · Meta | Mistral Small 3.1 · 24B | Gemma 3 · 27B |
+| Benchmark | **Agens Pilot** | Llama 4 Scout · Meta | Mistral Small 3.1 · 24B | Gemma 3 · 27B |
 |---|---|---|---|---|
 | GPQA Diamond | **80.3** | 57.2 | 45.4 | 42.4 |
 | HumanEval | **97.0** | — | ~88 | — |
@@ -247,8 +247,8 @@ See the accompanying `LICENSE` and `NOTICE` files. Training data used in develop
 its own licenses; downstream users are responsible for their own compliance.
 
 ```bibtex
-@misc{agenspilot30b2026,
-  title  = {Agens Pilot 30B: a coding-focused multimodal LLM},
+@misc{agenspilot2026,
+  title  = {Agens Pilot: a coding-focused multimodal LLM},
   author = {Blockway (BlockWay Link Limited)},
   year   = {2026},
   url    = {https://blockway.io}