Instructions to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-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 kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP
# Run inference directly in the terminal:
llama cli -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP
# Run inference directly in the terminal:
llama cli -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

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 kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP
# Run inference directly in the terminal:
./llama-cli -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

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 kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP
# Run inference directly in the terminal:
./build/bin/llama-cli -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Use Docker

docker model run hf.co/kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

LM Studio
Jan

vLLM

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Ollama
How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with Ollama:
```
ollama run hf.co/kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP
```

Unsloth Studio

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-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 kingjones777/Ling-3.0-tiny-ROCmFP4-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 kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF to start chatting

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with Docker Model Runner:
```
docker model run hf.co/kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP
```

Lemonade

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Run and chat with the model

lemonade run user.Ling-3.0-tiny-ROCmFP4-GGUF-Q4_0_ROCMFP

List all available models

lemonade list

Hermes Agent

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF:Q4_0_ROCMFP" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Ling-3.0-tiny-ROCmFP4-GGUF / README.md

kingjones777

card: add all-quant-variants table (4-bit / FP8 / FP8-AGENT)

529afca verified 1 day ago

preview code

Raw

History Blame Contribute Delete

9.77 kB

	---
	license: mit
	base_model: inclusionAI/Ling-3.0-tiny
	base_model_relation: quantized
	tags:
	- rocmfpx
	- ai-max-395
	- ryzen-ai-max-395
	- amd
	- gguf
	- llama.cpp
	- rocm
	- gfx1151
	- strix-halo
	- bailingmoe3
	- moe
	- hybrid-linear-attention
	language:
	- en
	pipeline_tag: text-generation
	---
	> ### ⚠️ STOCK `llama.cpp` WILL NOT LOAD THIS MODEL
	> `bailingmoe3` is not merged upstream, and Ling-3.0-tiny additionally needs the
	> Q-LoRA attention path (`q_lora_rank: 256`) that flash-era builds lack. Ignore the
	> auto-generated "Use this model" commands above — use the patch in `patches/`.
	>
	> 🚀 101.08 tok/s on AMD Ryzen AI MAX+ 395 (gfx1151 / Strix Halo) —
	> 4.30 GiB, smaller than Q4_K_M (4.49 GiB). Verified on two independent machines.


	## ✅ The patch you need is in this repo

	`patches/rocmfpx-2809dc5-add-bailingmoe3qlora-mellum-zaya.patch` — applies to
	[`charlie12345/ROCmFPX`](https://github.com/charlie12345/ROCmFPX) at commit `2809dc5`,
	verified with `git apply --check`.

	```bash
	git clone https://github.com/charlie12345/ROCmFPX.git && cd ROCmFPX
	git checkout 2809dc5
	git apply patches/rocmfpx-2809dc5-add-bailingmoe3qlora-mellum-zaya.patch
	cmake -B build -S . -DGGML_HIP=ON -DAMDGPU_TARGETS=gfx1151 -DCMAKE_BUILD_TYPE=Release
	cmake --build build -j$(nproc)
	```

	Full build notes, per-architecture details and licence: `patches/README.md` in this repo.

	⚠️ If you add files under `src/models/`, re-run `cmake -B build -S .` — the `models/*.cpp` GLOB
	is configure-time, so `cmake --build` alone will not link them.

	---

	# Ling-3.0-tiny — ROCmFP4 (tier 102 `COHERENT`) GGUF

	A 4-bit ROCmFP4 quantization of [`inclusionAI/Ling-3.0-tiny`](https://huggingface.co/inclusionAI/Ling-3.0-tiny),
	built for AMD gfx1151 (Ryzen AI MAX+ 395 / Strix Halo) with per-tensor protection of the
	LM head, token embeddings, shared experts, router, and the hybrid model's recurrent state.

	\| \| \|
	\|---\|---\|
	\| File \| `Ling-3.0-tiny-Q4_0_ROCMFP4_COHERENT.gguf` \|
	\| Size \| 4.2987 GiB (4,615,656,288 bytes) \|
	\| BPW \| 4.676 \|
	\| ftype \| `Q4_0_ROCMFP4_COHERENT` (102) \|
	\| Source \| BF16 GGUF (14.72 GiB) — lossless source, not a requantization \|
	\| sha256 \| `fd9af463569509aca718f6ccacab388c74d8abe108511d5b6e9c00e3243b42b4` \|

	---

	## ⛔ REQUIRES A PATCHED llama.cpp — STOCK WILL NOT LOAD THIS

	Two independent reasons, both unavoidable:

	1. `bailingmoe3` is not in upstream llama.cpp. Support is still open in
	[PR #26608](https://github.com/ggml-org/llama.cpp/pull/26608) (unmerged at time of writing).
	2. Ling-3.0-tiny needs the Q-LoRA attention path. Its config sets `q_lora_rank: 256`
	(`q_a_proj → q_a_layernorm → q_b_proj`). Several existing `bailingmoe3` implementations were
	written against Ling-3.0-flash, which has `q_lora_rank: null` and therefore no query
	compression. On such a build, every GGUF of tiny fails — including the BF16 and Q4_K_M
	ones — typically at `missing tensor 'blk.0.ssm_f.weight'`, before the Q-LoRA gap is even
	reached.

	You need a build with both `bailingmoe3` and its Q-LoRA path. The reference
	implementation is the branch behind PR #26608
	(`aetherbird/llama.cpp`, branch `bailingmoe3-support`). The ROCmFP4 quant types additionally
	require a fork that implements them; upstream llama.cpp does not have `Q4_0_ROCMFP4_*`.

	If your build loads Ling-3.0-flash but not tiny, you are missing the Q-LoRA path specifically.

	### ⚠️ `strings` is not a capability check

	We tested a second gfx1151 machine whose `libllama.so` contained **`bailingmoe3` (60 matches),
	`ssm_f_a`, and `attn_q_a`** — it looked fully capable. It still failed with the exact same
	`missing tensor 'blk.0.ssm_f.weight'`.

	Those symbols live in the tensor-name table. The fallback logic that maps `ssm_f` →
	`ssm_f_a`, and the Q-LoRA branch itself, are separate code. **Grepping the binary tells you
	nothing — attempt the load.**

	---

	## All quant variants

	Three builds of this model, all measured in one session on one box with one binary
	(Ryzen AI MAX+ 395, gfx1151, ROCm 7.2.4, `ROCmFPX-2809dc5`) — so these rows are directly
	comparable. Median of 3, warm-up discarded, otherwise-idle box.

	\| variant \| ftype \| size \| bpw \| decode (median) \| range \| repo \|
	\|---\|---\|---\|---\|---\|---\|---\|
	\| 4-bit COHERENT \| 102 \| 4.30 GiB \| 4.67 \| 104.04 \| 104.00 – 104.24 \| [Ling-3.0-tiny-ROCmFP4-GGUF](https://huggingface.co/kingjones777/Ling-3.0-tiny-ROCmFP4-GGUF) \|
	\| 8-bit AGENT \| 115 \| 7.72 GiB \| 8.40 \| 88.82 \| 88.80 – 88.83 \| [Ling-3.0-tiny-ROCmFPX-Q8_0-AGENT-GGUF](https://huggingface.co/kingjones777/Ling-3.0-tiny-ROCmFPX-Q8_0-AGENT-GGUF) \|
	\| 8-bit plain \| 111 \| 7.62 GiB \| 8.28 \| 89.51 \| 89.48 – 89.51 \| [Ling-3.0-tiny-ROCmFPX-Q8_0-GGUF](https://huggingface.co/kingjones777/Ling-3.0-tiny-ROCmFPX-Q8_0-GGUF) \|

	⚠️ The 4-bit build is faster (104.04 vs ~89 tok/s) and 44% smaller. These 8-bit builds exist for accuracy headroom, not speed — pick them only if you need the extra precision.

	What `AGENT` actually changes: it keeps far more tensors at true `Q8_0` instead of the
	packed 8-bit type — measured in these files, 135 tensors vs 2 tensors. On models with an
	MTP draft head that raises draft acceptance and wins ~6%; these two models have no MTP head,
	and here the two 8-bit builds are within noise of each other.

	## Measured results

	Verified on two independent gfx1151 machines, using the model's official sampling
	(`temperature 0.6`, `top_p 0.95`, `top_k 20`).

	\| \| machine A \| machine B \|
	\|---\|---\|---\|
	\| SoC \| Ryzen AI MAX+ 395 (gfx1151) \| Ryzen AI MAX+ 395 (gfx1151) \|
	\| memory \| 128 GB unified \| 125 GB unified \|
	\| ROCm \| 7.2.4 \| 7.13.0 \|
	\| flags \| `-ngl 99 -c 4096 -fa on` \| `-ngl 999 -c 32768 -fa on -fit off --no-mmap` \|
	\| loads \| ✅ `arch=bailingmoe3`, 526 tensors \| ✅ \|
	\| 17 × 23 \| ✅ `391` \| ✅ `391` \|
	\| capital of Japan \| ✅ `Tokyo` \| ✅ `Tokyo` \|
	\| days in 2024 \| ✅ `366` \| ✅ `366` \|
	\| reasoning separation \| ✅ clean, in `reasoning_content` \| ✅ \|
	\| decode speed \| 97.64 tok/s \| 101.08 tok/s \|

	### ⭐ The build is portable across ROCm minor versions

	The binaries were compiled against ROCm 7.2.4 and run unmodified on a ROCm 7.13.0 host —
	all 9 `Q4_0_ROCMFP4_*` quant types still enumerated, model loads, 101 tok/s. Both hosts are
	gfx1151. Useful if you have one build box and several Strix Halo machines: you can copy
	`llama-server` + `lib.so` rather than rebuilding per host (set `LD_LIBRARY_PATH` to the
	directory you copied them into).

	### Serving configuration that works

	Long-running deployment on machine B (systemd, always-hot):

	```
	-dev ROCm0 -ngl 999 -fa on -c 32768 -fit off -np 1 --no-mmap --jinja \
	--temp 0.6 --top-p 0.95 --top-k 20
	```
	plus `LimitMEMLOCK=infinity`, `HSA_OVERRIDE_GFX_VERSION=11.5.1`,
	`GGML_HIP_ENABLE_UNIFIED_MEMORY=1`.

	Verified on the running process, not just at launch: memlock unlimited (the 8 MB default
	will hobble the model), `n_ctx = 32768` actually granted — `--fit` is on by default and can
	silently shrink context or push tensors to CPU, so `-fit off` and then confirm the number — and
	zero "tensor override to CPU" lines in the log.

	### Per-tensor protection (audited in the finished file)

	\| tensor class \| type \|
	\|---\|---\|
	\| `output.weight` (LM head) \| Q6_K \|
	\| `token_embd.weight` \| Q6_K \|
	\| `_shexp` shared experts (69) \| Q8_0* \|
	\| `ffn_gate_inp` router (23) \| F32 \|
	\| `ssm_a`, `ssm_dt.bias` (36) \| F32 \|
	\| `ssm_conv1d_{q,k,v}` (54) \| F32 \|
	\| norms (79) \| F32 \|
	\| routed experts, attention projections \| 4-bit \|

	Why this matters. Tier `_STRIX` (105) protects attention K/V but not the LM head — on this
	model's 157,184-token vocabulary that leaves every logit passing through a 4-bit tensor.
	Tier `102 COHERENT` carries Q6_K token embeddings, and the head/shared-expert protections above
	were applied explicitly. Shared experts matter because they are dense — they process every
	token, so their error is systematic rather than averaged across the 128 routed experts.

	The recurrent/linear-attention state (`ssm_a`, `ssm_dt`, `conv1d`) is kept at F32: these are
	float32 in the source model, and quantizing hybrid state is a known way to produce a model that
	loads, runs, and emits fluent nonsense.

	### Size comparison (same source, same machine)

	\| build \| size \|
	\|---\|---\|
	\| BF16 \| 14.72 GiB \|
	\| Q4_K_M \| 4.4926 GiB \|
	\| this build \| 4.2987 GiB \|

	---

	## What was NOT measured

	Stated plainly so you can judge fitness for your use case:

	- No perplexity run, and no quality A/B against Q4_K_M or BF16. The correctness checks
	above are memorized-fact prompts — they are necessary but not sufficient, and a damaged model
	can pass them.
	- No long-context testing. All generations were short. The 32,768-token context was granted
	and confirmed at load on machine B, but nothing exercised rope/KV behaviour at depth, and
	nothing was run near the model's 131,072 ceiling.
	- No tool-calling evaluation.
	- MTP / speculative decoding untested — Ling-3.0-tiny has `num_nextn_predict_layers: 0`,
	so it has no MTP layer to exercise.

	---

	## Model

	`BailingMoeV3ForCausalLM` / `bailing_hybrid`, GGUF arch `bailingmoe3`.
	24 layers in a 3:1 stack of KDA (Kimi Delta Attention, 18 layers) and MLA
	(Multi-head Latent Attention, 6 layers) · hidden 1536 · 128 routed experts, 8 active ·
	shared experts · vocab 157,184 · `q_lora_rank 256` · `kv_lora_rank 512` · context 131,072.

	Base model licence: MIT (inherited). All credit for the model itself goes to
	[inclusionAI](https://huggingface.co/inclusionAI).