Model Architecture: Qwen3NextForCausalLM
- Input: Text
- Output: Text
Model Optimizations:
- Weight quantization: FP8
- Activation quantization: FP8
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 FP8 data type. This optimization reduces the number of bits per parameter from 16 to 8, reducing the disk size and GPU memory requirements by approximately 50%. Only the weights and activations of the linear operators within transformers blocks of the language model are quantized.

Deployment

Use with vLLM

Initialize vLLM server:

vllm serve inference-optimization/Qwen3-Coder-Next-FP8-dynamic --port 8000 --tensor-parallel-size 2 --enable-auto-tool-choice --tool-call-parser qwen3_coder

Send requests to the server:

python

# 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-FP8-dynamic",
    max_tokens=65536,
    tools=tools,
)

print(completion.choices[0])

Creation

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

Creation details

python

from datasets import load_dataset
from transformers import AutoModelForCausalLM, AutoTokenizer

from llmcompressor import oneshot
from llmcompressor.modifiers.quantization import QuantizationModifier
from llmcompressor.utils import dispatch_for_generation

# NOTE: Requires a minimum of transformers 4.57.0

MODEL_ID = "Qwen/Qwen3-Coder-Next"

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


# Configure the quantization algorithm and scheme.
# In this case, we:
#   * quantize the weights to fp8 with per channel via ptq
#   * quantize the activations to fp8 with dynamic per token
recipe = QuantizationModifier(
    targets="Linear", scheme="FP8_DYNAMIC", ignore=[
        "re:.*lm_head",
        "re:.*mlp.gate$",
        "re:.*mlp.shared_expert_gate$",
        "re:.*linear_attn.*",
    ],
    weight_observer="mse"
)

# Apply quantization.
oneshot(model=model, recipe=recipe)

# Confirm generations of the quantized model look sane.
print("========== SAMPLE GENERATION ==============")
dispatch_for_generation(model)
input_ids = tokenizer("Hello my name is", return_tensors="pt").input_ids.to(
    model.device
)
output = model.generate(input_ids, max_new_tokens=20)
print(tokenizer.decode(output[0]))
print("==========================================")

# Save to disk in compressed-tensors format.
SAVE_DIR = MODEL_ID.rstrip("/").split("/")[-1] + "-FP8-dynamic"
model.save_pretrained(SAVE_DIR)
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.

Evaluation details

**Coding Benchmarks **

SWE-Bench

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

Accuracy

Category	Metric	Qwen3-Coder-Next	Qwen3-Coder-Next-FP8-dynamic	Recovery (%)
SWE-Bench	Lite	49.33	53	107.4