inference-optimization / Qwen3-Coder-Next.w4a16

Qwen3-Coder-Next.w4a16

Model Overview

• Model Architecture: Qwen3NextForCausalLM
  • Input: Text
  • Output: Text
• Model Optimizations:
  • Weight quantization: INT4
  • Activation quantization: FP16
• Release Date:
• Version: 1.0
• Model Developers: : Red Hat

Quantized version of Qwen/Qwen3-Coder-Next .

Model Optimizations

This model was obtained by quantizing the weights and activations of Qwen/Qwen3-Coder-Next to INT4 data type. This optimization reduces the number of bits per parameter from 16 to 4, reducing the disk size and GPU memory requirements by approximately 75%. Only the weights and activations of the linear operators within transformers blocks of the language model are quantized.

Deployment

Use with vLLM

1. Initialize vLLM server:

vllm serve inference-optimization/Qwen3-Coder-Next.w4a16 --port 8000 --tensor-parallel-size 2 --enable-auto-tool-choice --tool-call-parser qwen3_coder
 

1. Send requests to the server:

# Your tool implementation
def square_the_number(num: float) -> dict:
    return num ** 2

# Define Tools
tools=[
    {
        "type":"function",
        "function":{
            "name": "square_the_number",
            "description": "output the square of the number.",
            "parameters": {
                "type": "object",
                "required": ["input_num"],
                "properties": {
                    'input_num': {
                        'type': 'number', 
                        'description': 'input_num is a number that will be squared'
                        }
                },
            }
        }
    }
]

from openai import OpenAI
# Define LLM
client = OpenAI(
    # Use a custom endpoint compatible with OpenAI API
    base_url='http://localhost:8000/v1',  # api_base
    api_key="EMPTY"
)
 
messages = [{'role': 'user', 'content': 'square the number 1024'}]

completion = client.chat.completions.create(
    messages=messages,
    model="RedHatAI/Qwen3-Coder-Next.w4a16",
    max_tokens=65536,
    tools=tools,
)

print(completion.choices[0])

Creation

This model was quantized using the llm-compressor library as shown below.

from datasets import load_dataset
from transformers import AutoModelForCausalLM, AutoTokenizer
from llmcompressor import oneshot
from llmcompressor.modifiers.quantization import GPTQModifier

MODEL_ID = "Qwen/Qwen3-Coder-Next"

# Load model.
model = AutoModelForCausalLM.from_pretrained(MODEL_ID, torch_dtype="auto")
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)


NUM_CALIBRATION_SAMPLES=512
MAX_SEQUENCE_LENGTH=2048

# Load dataset.
ds = load_dataset("HuggingFaceH4/ultrachat_200k", split=f"train_sft[:{NUM_CALIBRATION_SAMPLES}]")
ds = ds.shuffle(seed=42)

# Preprocess the data into the format the model is trained with.
def preprocess(example):
    return {"text": tokenizer.apply_chat_template(example["messages"], tokenize=False, )}

ds = ds.map(preprocess)

# Tokenize the data (be careful with bos tokens - we need add_special_tokens=False since the chat_template already added it).
def tokenize(sample):
    return tokenizer(sample["text"], padding=False, max_length=MAX_SEQUENCE_LENGTH, truncation=True, add_special_tokens=False)
ds = ds.map(tokenize, remove_columns=ds.column_names)

# Configure the quantization algorithm to run.
recipe = GPTQModifier(targets="Linear", scheme="W4A16", weight_observer="mse", ignore= ['re:.*lm_head', 're:.*mlp.gate$', 're:.*mlp.shared_expert_gate$', 're:.*linear_attn.*'])

# Apply quantization.
oneshot(
    model=model, dataset=ds,
    recipe=recipe,
    max_seq_length=MAX_SEQUENCE_LENGTH,
    num_calibration_samples=NUM_CALIBRATION_SAMPLES,
)

# Save to disk compressed.
SAVE_DIR = MODEL_ID.rstrip("/").split("/")[-1] + "-W4A16-G128"
model.save_pretrained(SAVE_DIR, save_compressed=True)
tokenizer.save_pretrained(SAVE_DIR)

Evaluation

The model was evaluated on the OpenLLM leaderboard task, using lm-evaluation-harness . vLLM was used for all evaluations.

python -m swebench.harness.run_evaluation \
  --dataset_name princeton-nlp/SWE-bench_Lite \
  --predictions_path preds.json \
  --run_id validate-preds

Accuracy