Instructions to use promzeus/gh0stx-bf16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use promzeus/gh0stx-bf16 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="promzeus/gh0stx-bf16")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("promzeus/gh0stx-bf16")
model = AutoModelForCausalLM.from_pretrained("promzeus/gh0stx-bf16")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use promzeus/gh0stx-bf16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "promzeus/gh0stx-bf16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "promzeus/gh0stx-bf16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/promzeus/gh0stx-bf16

SGLang

How to use promzeus/gh0stx-bf16 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "promzeus/gh0stx-bf16" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "promzeus/gh0stx-bf16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "promzeus/gh0stx-bf16" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "promzeus/gh0stx-bf16",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use promzeus/gh0stx-bf16 with Docker Model Runner:
```
docker model run hf.co/promzeus/gh0stx-bf16
```

gh0stx-bf16

Expert-pruned BF16 derivative of Qwen3.5-397B-A17B. Full-precision weights — portable to any CUDA stack; quantize to whatever your hardware needs.

Specs

Base: Qwen3.5-397B-A17B (hybrid GatedDeltaNet + sparse MoE + periodic full-attention)
Prune: expert-pruned 512 → 174 experts with a domain-calibrated keep-set (routing-frequency over a balanced corpus, not raw router-norm) → ~141B total / ~17B active
Format: BF16, ~280 GB (16 shards)
Context: up to 131K (native up to 256K)

Hardware

BF16 needs ~280 GB → multi-GPU, or quantize first:

8× H100/A100 80GB (TP=8) — runs BF16 directly.
Fewer / smaller cards — quantize: AWQ / GPTQ (INT4) ≈ 75–90 GB, or FP8 (H100) ≈ ~140 GB (llm-compressor / AutoAWQ).
Single NVIDIA Blackwell card (GB10 / DGX Spark) → use the NVFP4 sibling promzeus/gh0stx-nvfp4 (fits in 128 GB, DFlash ~17 tok/s).

Serve (vLLM, BF16, tensor-parallel)

vllm serve promzeus/gh0stx-bf16 \
  --served-model-name gh0stx --tensor-parallel-size 8 \
  --trust-remote-code --max-model-len 131072 \
  --gpu-memory-utilization 0.90 \
  --enable-chunked-prefill --enable-prefix-caching \
  --reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_coder

Set --tensor-parallel-size to your GPU count. Requires a vLLM build with Qwen3.5-MoE support.

Speculative decoding (recommended: DFlash)

DFlash is the fastest path and is format-agnostic — the drafter is an external ~0.8B model, no file surgery on this repo. Add to the serve command:

  --mamba-cache-dtype float32 \
  --speculative-config '{"method":"dflash","model":"z-lab/Qwen3.5-397B-A17B-DFlash","num_speculative_tokens":10}'

Drafter: z-lab/Qwen3.5-397B-A17B-DFlash (auto-downloads, matches this model's base → healthy acceptance).
--mamba-cache-dtype float32 is required with DFlash on this hybrid (GatedDeltaNet) model: in fp16 the Mamba KV page size doesn't reconcile with the drafter's sliding-window page size and engine init fails in unify_kv_cache_spec_page_size. Use n=10 (n≥12 is crash-correlated on the validated stack).
On a single Blackwell card this yields ~17 tok/s (see the NVFP4 sibling); on multi-GPU BF16 the acceptance win carries over, absolute t/s depends on your interconnect.

Alternative: built-in MTP head. A grafted qwen3_5_mtp head is shipped in the NVFP4 sibling in NVFP4 form (Blackwell-only). For BF16/other hardware, DFlash above is the recommended accelerator.

Notes

KV cache: keep BF16. Do not set --kv-cache-dtype fp8 — unstable with the linear-attention + spec-decode path (and mandatory BF16 for the DFlash drafter).
Thinking mode is ON by default (Qwen3.5). Disable per request with chat_template_kwargs: {"enable_thinking": false} — recommended for agentic/tool use.

Credits

Qwen for Qwen3.5-397B-A17B.
z-lab for DFlash and the Qwen3.5-397B-A17B-DFlash drafter.
AEON-7 for the DGX-Spark / sm_121a DFlash recipe (the --mamba-cache-dtype float32 unlock).

Sibling

promzeus/gh0stx-nvfp4 — NVFP4 (W4A4) for a single Blackwell card (GB10), with DFlash (~~17 tok/s @ 65K) and a built-in NVFP4 MTP head (~~11–13 tok/s @ 131K).

Downloads last month: 1,082

Safetensors

Model size

141B params

Tensor type

BF16

Model tree for promzeus/gh0stx-bf16

Base model

Qwen/Qwen3.5-397B-A17B

Finetuned

(38)

this model