GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
JC1DA
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Model Details
Model Provider
JC1DA
Model Tree
DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU
Input Modalities
Output Modalities
Supported Functionality
| Parameter | Value |
|---|---|
| Algorithm | AutoRound + GPTQ |
| Weight bits | 4 |
| Activation bits | 16 |
| Group size | 128 |
| Version | AutoRound 0.12.3 |
Model size: ~19GB (vs ~54GB FP16) — 65% reduction
Both versions benchmarked on identical hardware (NVIDIA A100 80GB) using vLLM 0.27.1.
| Benchmark | FP16 | W4A16 | Delta |
|---|---|---|---|
| ARC-Easy (acc) | 86.91% | 87.25% | +0.34% |
| ARC-Easy (acc_norm) | 85.69% | 86.49% | +0.80% |
| HumanEval (pass@1) | 79.27% | 79.88% | +0.61% |
| BBH (exact_match) | 89.94% | 90.34% | +0.40% |
| MMLU (acc) | 86.89% | 86.69% | -0.20% |
Key findings:
Works with any framework supporting GPTQ/AutoRound checkpoints:
python
from transformers import AutoModelForCausalLM, AutoTokenizermodel_id = "DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-W4A16/DavidAU_Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-w4g128"tokenizer = AutoTokenizer.from_pretrained(model_id)model = AutoModelForCausalLM.from_pretrained(model_id,device_map="auto",torch_dtype="auto")
Or via vLLM for high-throughput serving:
bash
vllm serve DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-W4A16/DavidAU_Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NM-DAU-w4g128 \--dtype auto --tensor-parallel-size 1
See the FP16 original model card for full details on training methodology, stages, and capabilities.
This quantized version preserves all characteristics of the original:
| WikiText2 PPL |
| 8.04 |
| 8.17 |
| +0.13 |