You can deploy this model and start earning money today!

textclf/
Qwen3.8-27B-TQ-4bit

Name: Qwen3.8-27B-TQ-4bit
Availability: InStock

Views

No views yet

Qwen3.8-27B-TQ-4bit

27B-parameter base model · TextCLF Quant (TQ) 4-bit

Note on Hugging Face's model-size display: Hugging Face may report roughly 3B stored parameters for this repository because TQ stores the weights in a packed 4-bit representation. The underlying model is Qwen3.8-27B (27B parameters). The smaller number reflects the packed storage tensors counted by the Hub, not the parameter count of the original model architecture.

Qwen3.8 27B quantized to 4-bit with TextCLF Quant (TQ).

TextCLF Quant: https://textclf.com

TQ is a calibration-free 4-bit quantization method designed to preserve the behavior of the original model without requiring a calibration dataset. Because TQ does not optimize the quantization around a fixed calibration corpus, it is designed to generalize beyond calibration-specific data and workloads.

This repository contains the 27B-parameter Qwen3.8 model quantized with TextCLF Quant (TQ) 4-bit.

Quantization fidelity

TQ 4-bit · Qwen 3.8 27B · WikiText Raw

Metric	Result
Mean KL divergence	0.0282
Top-1 agreement	92.4%

Lower mean KL divergence indicates closer agreement between the quantized model's output distribution and the original model. Top-1 agreement measures how often the quantized model and original model select the same highest-probability next token.

Why TQ?

Calibration-free. TQ does not require a representative calibration dataset before quantization.

This avoids tying the quantization procedure to a particular calibration corpus and is intended to provide better generalization beyond the data that would otherwise have been used for calibration.

4-bit inference. Model weights are quantized to 4-bit for substantially lower weight memory requirements than BF16/FP16 deployment.

Native vLLM integration. TQ models run through the TQ vLLM quantization plugin and custom CUDA kernels included in the TextCLF TQ Docker image.

Model


Base model	`Qwen/Qwen3.8-27B`
Original model parameters	27B
HF displayed stored tensor count	May appear as ~3B due to TQ 4-bit packing
Quantization	TextCLF Quant (TQ) 4-bit
Calibration	None — calibration-free
Runtime	vLLM + TQ custom kernels
Quantization name	`tq_quant`

Run with Docker

The recommended way to run this model is with the TQ Docker image, which contains the compatible vLLM installation, TQ quantization plugin, and TQ CUDA kernels.

1. Make sure NVIDIA Docker support works

bash

docker run --rm --gpus all \
  nvidia/cuda:13.0.2-base-ubuntu22.04 \
  nvidia-smi

Your GPU should appear in the output.

2. Start Qwen3.8-27B-TQ-4bit

bash

docker run --rm --gpus all \
  -p 8000:8000 \
  docker.io/textclf/tq-quant:4bit \
  vllm serve textclf/Qwen3.8-27B-TQ-4bit \
  --quantization tq_quant

The model is then available through vLLM's OpenAI-compatible API on port 8000.

WSL2

If you are running Docker through WSL2 and vLLM reports that UVA is unavailable, enable vLLM's WSL2 pinned-memory support:

bash

docker run --rm --gpus all \
  -e VLLM_WSL2_ENABLE_PIN_MEMORY=1 \
  -p 8000:8000 \
  docker.io/textclf/tq-quant:4bit \
  vllm serve textclf/Qwen3.8-27B-TQ-4bit \
  --quantization tq_quant

Send a request

Once the server is ready:

bash

curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "textclf/Qwen3.8-27B-TQ-4bit",
    "messages": [
      {
        "role": "user",
        "content": "Explain quantization in one paragraph."
      }
    ]
  }'

Hugging Face authentication

For higher Hugging Face Hub rate limits, pass your Hugging Face token into the container:

bash

docker run --rm --gpus all \
  -e HF_TOKEN="$HF_TOKEN" \
  -p 8000:8000 \
  docker.io/textclf/tq-quant:4bit \
  vllm serve textclf/Qwen3.8-27B-TQ-4bit \
  --quantization tq_quant

On WSL2, combine it with:

-e VLLM_WSL2_ENABLE_PIN_MEMORY=1

About TextCLF Quant (TQ)

Learn more at https://textclf.com.

TextCLF Quant (TQ) is a calibration-free quantization approach for efficient LLM inference.

Traditional post-training quantization methods can depend on a calibration dataset to estimate quantization parameters. The resulting quantization can therefore be influenced by the distribution and composition of that calibration data.

TQ removes that calibration-data requirement. The goal is to retain strong fidelity to the original model while allowing the quantized representation to generalize beyond any particular calibration corpus.

For this Qwen3.8 27B checkpoint, the TQ 4-bit model achieves:

0.0282 mean KL divergence on WikiText Raw
92.4% top-1 agreement with the original model on WikiText Raw
No calibration dataset required

Base model

This checkpoint is derived from Qwen/Qwen3.8-27B, a 27B-parameter Qwen3.8 model.

Please refer to the original Qwen model card for architecture details, capabilities, usage guidance, limitations, and upstream licensing information.

License

The base model is released under the Apache 2.0 License. See the repository license and the original Qwen model card for applicable terms.