Instructions to use cloudyu/hy3-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use cloudyu/hy3-gguf with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="cloudyu/hy3-gguf",
	filename="hy3_q4k_mixed.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use cloudyu/hy3-gguf with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf cloudyu/hy3-gguf
# Run inference directly in the terminal:
llama cli -hf cloudyu/hy3-gguf

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf cloudyu/hy3-gguf
# Run inference directly in the terminal:
llama cli -hf cloudyu/hy3-gguf

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf cloudyu/hy3-gguf
# Run inference directly in the terminal:
./llama-cli -hf cloudyu/hy3-gguf

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf cloudyu/hy3-gguf
# Run inference directly in the terminal:
./build/bin/llama-cli -hf cloudyu/hy3-gguf

Use Docker

docker model run hf.co/cloudyu/hy3-gguf

LM Studio
Jan

vLLM

How to use cloudyu/hy3-gguf with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "cloudyu/hy3-gguf"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "cloudyu/hy3-gguf",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/cloudyu/hy3-gguf

Ollama
How to use cloudyu/hy3-gguf with Ollama:
```
ollama run hf.co/cloudyu/hy3-gguf
```

Unsloth Studio

How to use cloudyu/hy3-gguf with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for cloudyu/hy3-gguf to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for cloudyu/hy3-gguf to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for cloudyu/hy3-gguf to start chatting

Atomic Chat new
Docker Model Runner
How to use cloudyu/hy3-gguf with Docker Model Runner:
```
docker model run hf.co/cloudyu/hy3-gguf
```

Lemonade

How to use cloudyu/hy3-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull cloudyu/hy3-gguf

Run and chat with the model

lemonade run user.hy3-gguf-{{QUANT_TAG}}

List all available models

lemonade list

hy3-gguf / README.md

cloudyu

Update README.md

f80f66f verified 4 days ago

preview code

Raw

History Blame Contribute Delete

4.66 kB

	---
	license: apache-2.0
	base_model: tencent/Hy3
	tags:
	- hy3
	- hunyuan
	- hy_v3
	- gguf
	- mixture-of-experts
	- moe
	language:
	- en
	- zh
	library_name: hy3
	pipeline_tag: text-generation
	---

	# hy3-gguf — GGUF weights for Tencent Hy3 (tencent/Hy3)

	GGUF-format weights for tencent/Hy3 (`HYV3ForCausalLM`, `model_type: hy_v3`), a
	295B-parameter / 21B-active-parameter Mixture-of-Experts model from
	Tencent's Hunyuan ("Hy") team ([`tencent/Hy3`](https://huggingface.co/tencent/Hy3)).

	These files were produced by the [`hy3`](https://github.com/yuhai-china/hy3)
	converter (`hy3-convert`) and are meant to be run with the **`hy3` inference
	engine**, a from-scratch C/Metal/CUDA implementation.





	> ## ⚠️ This GGUF does NOT work with llama.cpp
	>
	> Despite the `.gguf` extension, these files are **only usable by the
	> [`hy3`](https://github.com/yuhai-china/hy3) engine**. `llama.cpp`,
	> `ollama`, `LM Studio`, `text-generation-webui`, `koboldcpp`, and any other
	> llama.cpp-based tool cannot load these files. Three independent reasons:
	>
	> 1. Unknown architecture. The metadata declares
	> `general.architecture = "hy_v3"`. llama.cpp only knows `hunyuan-moe`,
	> `hunyuan-dense`, `hunyuan_vl` — loading aborts with
	> `unknown model architecture: 'hy_v3'`.
	> 2. Custom metadata keys. All hyperparameters use the `hy_v3.*` prefix
	> (`hy_v3.block_count`, `hy_v3.expert_count`, …), which llama.cpp does not
	> look up.
	> 3. Non-fused expert tensors. Experts are stored one tensor per expert
	> (`blk.N.ffn_gate_exps.0.gate_proj.weight`, `…1…`, … — 46080 tensors),
	> whereas llama.cpp expects experts fused into a single stacked 3D tensor per
	> layer. This is a fundamentally different on-disk layout.
	>
	> This is a custom GGUF readable only by the `hy3` loader. Do not open
	> issues against llama.cpp for these files.

	## How to run

	Use the `hy3` engine: <https://github.com/yuhai-china/hy3>

	```bash
	git clone https://github.com/yuhai-china/hy3
	cd hy3
	make # macOS builds the Metal backend automatically

	# download a GGUF from this repo, then:
	./run_metal.sh -m /path/to/hy3_q4k_mixed.gguf -p "The capital of France is" -experts 8
	```

	> Testing scope: the `hy3` engine's performance work and benchmarks were
	> developed and verified only on macOS / Apple Silicon (Metal backend),
	> measured on an M2 Ultra (~20–27 tok/s decode depending on `-experts`). The
	> CPU and CUDA backends exist in the source but were not exercised as part of
	> that work — treat them as untested.

	## Files / quantization

	The mixed-precision GGUF follows this scheme (see `hy3_convert.c`):

	\| Tensor group \| Type \|
	\|---\|---\|
	\| Routed experts (`ffn_{gate,up,down}_exps`) — the bulk of the model \| Q4_K \|
	\| Attention q/k/v/o projections, shared-expert & dense FFN, `output.weight` \| Q8_0 \|
	\| Norms, router (`ffn_gate_inp`), biases \| F32 \|
	\| `token_embd.weight` \| F16 \|

	## Model facts

	\| \| \|
	\|---\|---\|
	\| Architecture \| `HYV3ForCausalLM` (`hy_v3`) \|
	\| Layers \| 80 (layer 0 dense, layers 1–79 MoE) \|
	\| Hidden size \| 4096 \|
	\| Attention \| 64 heads, GQA with 8 KV heads, head_dim 128 \|
	\| Experts \| 192 routed (top-8 activated) + 1 shared (always active) \|
	\| Expert intermediate size \| 1536 \|
	\| Dense (layer 0) intermediate size \| 13312 \|
	\| Vocab size \| 120832 (120818 real tokens + padding) \|
	\| RoPE \| theta 11158840, `rotate_half` pairing \|
	\| QK norm \| per-head RMSNorm on Q and K, before RoPE \|
	\| MoE routing \| `sigmoid(router_logits)`; top-8 by `sigmoid + expert_bias`, combined using unbiased sigmoid weights, renormalized to sum 1, scaled by `router_scaling_factor = 2.826` \|

	The engine supports a runtime top-k experts override (`-experts 1..8`) to
	trade quality for speed. On a small 13-question code/reasoning eval (greedy,
	no-think): experts=8 → 10/13, experts=4 → 7/13. Default is 8.

	## Chat template

	Hy3 is instruction-tuned and expects the Hunyuan V3 chat format (the `hy3`
	engine applies it automatically; use `--raw` to bypass). Single user turn,
	no-think:

	```
	<｜hy_begin_of_sentence:opensource｜><｜reasoning_mode:opensource｜>reasoning_effort:no_think<｜hy_User:opensource｜>{prompt}<｜hy_Assistant:opensource｜><think:opensource></think:opensource>
	```

	Generation stops on `<｜hy_eos:opensource｜>` (120025),
	`<｜hy_endofsentence｜>` (120001), or `<｜hy_EOT｜>` (120008).

	## License & attribution

	Weights derive from [`tencent/Hy3`](https://huggingface.co/tencent/Hy3); refer
	to the upstream repository for the governing model license. This is an
	unofficial community conversion, not affiliated with or endorsed by Tencent.