Update README.md

README.md

@@ -225,7 +225,11 @@ This model card documents **Ornith-1.5-35B-A3B**, the mid-size mixture-of-expert
 <li><b>vLLM</b> ≥ 0.19.1</li>
 <li><b>SGLang</b> ≥ 0.5.9</li>
 </ul>
-<p style="margin:0">Recommended sampling parameters: <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">temperature=0.6</code>, <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">top_p=0.95</code>, <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">top_k=20</code> (use <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">temperature=1.0</code> to reproduce the reported benchmark setup).</p>
+<p style="margin:0 0 6px">Recommended sampling parameters:</p>
+<ul style="margin:0;padding-left:20px">
+<li><b>For general tasks:</b> <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">temperature=0.6</code>, <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">top_p=0.95</code>, <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">top_k=20</code></li>
+<li><b>To reproduce the reported benchmarks:</b> <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">temperature=1.0</code></li>
+</ul>
 </div>
 
 
@@ -262,58 +266,42 @@ python -m sglang.launch_server \
     --reasoning-parser qwen3
 ```
 
-#### Hugging Face Transformers
+#### For Long-Context
 
-For a quick local test (or to script offline generation), load the model directly with Transformers. Make sure you have a recent release installed — see the [Transformers installation guide](https://huggingface.co/docs/transformers/installation); Ornith-1.5-35B-A3B requires `transformers >= 5.8.1`.
+Ornith-1.5-35B-A3B handles context windows of up to 262,144 tokens. When a task's combined input and output must go beyond this limit, we suggest extending the effective window with RoPE scaling — YaRN is the technique we validate against, and it is already built into both vLLM and SGLang. With a scaling factor of 4.0, the usable window grows to roughly 1M tokens.
 
-```python
-from transformers import AutoModelForCausalLM, AutoTokenizer
+You can turn YaRN on in either of two ways:
 
-model_name = "ornith-ai/Ornith-1.5-35B-A3B"
+- **Edit the checkpoint's `config.json`.** Add a `rope_scaling` block to the model configuration:
 
-tokenizer = AutoTokenizer.from_pretrained(model_name)
-model = AutoModelForCausalLM.from_pretrained(
-    model_name,
-    dtype="auto",
-    device_map="auto",
-)
+  ```json
+  {
+      "rope_scaling": {
+          "rope_type": "yarn",
+          "factor": 4.0,
+          "original_max_position_embeddings": 262144
+      }
+  }
+  ```
 
-messages = [
-    {"role": "user", "content": "Write a Python function is_prime(n). Keep it short."}
-]
-text = tokenizer.apply_chat_template(
-    messages,
-    tokenize=False,
-    add_generation_prompt=True,
-)
+- **Override at launch time.** Leave the checkpoint untouched and extend the serve commands above with the equivalent flags.
 
-inputs = tokenizer(text, return_tensors="pt").to(model.device)
-generated = model.generate(
-    **inputs,
-    max_new_tokens=512,
-    do_sample=True,
-    temperature=0.6,
-    top_p=0.95,
-    top_k=20,
-)
-output_ids = generated[0][inputs.input_ids.shape[1]:]
+  vLLM:
 
-# The reply contains a <think> ... </think> reasoning block followed by the answer.
-content = tokenizer.decode(output_ids, skip_special_tokens=True)
-print(content)
-```
+  ```bash
+  VLLM_ALLOW_LONG_MAX_MODEL_LEN=1 vllm serve ornith-ai/Ornith-1.5-35B-A3B ... --hf-overrides '{"rope_scaling": {"rope_type": "yarn", "factor": 4.0, "original_max_position_embeddings": 262144}}' --max-model-len 1000000
+  ```
 
-To split the reasoning trace from the final answer, parse on the `</think>` marker:
+  SGLang:
 
-```python
-text = tokenizer.decode(output_ids, skip_special_tokens=True)
-if "</think>" in text:
-    reasoning, answer = text.split("</think>", 1)
-    reasoning = reasoning.replace("<think>", "").strip()
-    answer = answer.strip()
-else:
-    reasoning, answer = "", text.strip()
-```
+  ```bash
+  SGLANG_ALLOW_OVERWRITE_LONGER_CONTEXT_LEN=1 python -m sglang.launch_server ... --json-model-override-args '{"rope_scaling": {"rope_type": "yarn", "factor": 4.0, "original_max_position_embeddings": 262144}}' --context-length 1000000
+  ```
+
+<div style="border-left:4px solid #FD8E5B;background:rgba(253,142,91,0.1);border-radius:6px;padding:12px 16px;font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',Roboto,sans-serif;font-size:14px;line-height:1.6">
+<div style="font-weight:700;color:#FD8E5B;margin-bottom:6px">📝 NOTE</div>
+<p style="margin:0">Open-source runtimes implement YaRN <i>statically</i>: the same scaling factor is applied to every request regardless of its length, which can slightly hurt quality on ordinary-length inputs. Only enable <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">rope_scaling</code> when your workload genuinely needs the longer window, and size <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">factor</code> to match it — the target window is roughly <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">factor</code> × 262,144, so if your requests top out around 524,288 tokens, <code style="background:rgba(253,142,91,0.15);padding:1px 5px;border-radius:4px">factor: 2.0</code> is the better setting.</p>
+</div>
 
 ### Using Ornith-1.5-35B-A3B via the Chat Completions API
 
@@ -381,51 +369,29 @@ You can point any OpenAI-compatible SDK (Python, Node.js, etc.) or `curl` at the
 
 ## Agentic Usage
 
-Ornith-1.5-35B-A3B excels in tool-calling and agentic coding capabilities.
+Ornith-1.5-35B-A3B excels in tool-calling and agentic coding. It exposes an OpenAI-compatible endpoint with tool calling and works out of the box with standard agent frameworks.
 
-### Agent Frameworks
+**Examples of using Ornith with agents:**
 
-Because Ornith-1.5-35B-A3B exposes an OpenAI-compatible endpoint with tool calling, it works out of the box with standard agent frameworks. Below is a minimal example that connects Ornith-1.5-35B-A3B to tools through an MCP server.
 
-```python
-import os
-from openai import OpenAI
-
-client = OpenAI(
-    base_url=os.getenv("OPENAI_BASE_URL", "http://localhost:8000/v1"),
-    api_key=os.getenv("OPENAI_API_KEY", "EMPTY"),
-)
-
-tools = [
-    {
-        "type": "function",
-        "function": {
-            "name": "run_shell",
-            "description": "Run a shell command and return its output.",
-            "parameters": {
-                "type": "object",
-                "properties": {
-                    "command": {"type": "string", "description": "The command to run"}
-                },
-                "required": ["command"],
-            },
-        },
-    }
-]
+#### Ollama
+```bash
+ollama run hf.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF
+```
 
-messages = [{"role": "user", "content": "List the Python files in the current directory."}]
 
-response = client.chat.completions.create(
-    model="ornith-ai/Ornith-1.5-35B-A3B",
-    messages=messages,
-    tools=tools,
-    temperature=0.6,
-    top_p=0.95,
-)
-print(response.choices[0].message)
+#### Atomic.chat
+```bash
+# Atomic.chat loads a GGUF build of Ornith (ornith-ai/Ornith-1.5-35B-A3B-GGUF)
+# through llama.cpp's OpenAI-compatible API on port 8000.
+llama-server -hf ornith-ai/Ornith-1.5-35B-A3B-GGUF --port 8000 -c 262144
 ```
 
-**Examples of using Ornith with agent harness:**
+#### llama.cpp
+```bash
+# llama.cpp — serve an OpenAI-compatible API on port 8000.
+llama-server -hf ornith-ai/Ornith-1.5-35B-A3B-GGUF --port 8000 -c 262144
+```
 
 #### Hermes Agent
 ```bash
@@ -435,18 +401,6 @@ export OPENAI_API_KEY="EMPTY"
 export MODEL="ornith-ai/Ornith-1.5-35B-A3B"
 ```
 
-
-#### Atomic.chat / Ollama / llama.cpp
-```bash
-# Both runtimes load a GGUF build of Ornith (publish one at ornith-ai/Ornith-1.5-35B-A3B-GGUF).
-
-# llama.cpp — serve an OpenAI-compatible API on port 8000.
-llama-server -hf ornith-ai/Ornith-1.5-35B-A3B-GGUF --port 8000 -c 262144
-
-# Ollama — pull and chat with the same GGUF straight from Hugging Face.
-ollama run hf.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF
-```
-
 #### OpenClaw
 
 ```bash
@@ -470,18 +424,6 @@ pip install unsloth
 #   )
 ```
 
-#### OpenHands
-```bash
-pip install openhands-ai
-
-# OpenHands routes through LiteLLM; the "openai/" prefix selects the OpenAI-compatible path.
-export LLM_MODEL="openai/ornith-ai/Ornith-1.5-35B-A3B"
-export LLM_BASE_URL="http://localhost:8000/v1"
-export LLM_API_KEY="EMPTY"
-
-# Launch the CLI (or run the official OpenHands Docker image with the same env vars).
-openhands
-```
 
 ### Coding CLIs