Instructions to use prism-ml/Ternary-Bonsai-2-27B-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 prism-ml/Ternary-Bonsai-2-27B-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 prism-ml/Ternary-Bonsai-2-27B-gguf:F16
# Run inference directly in the terminal:
llama cli -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16
# Run inference directly in the terminal:
llama cli -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16

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 prism-ml/Ternary-Bonsai-2-27B-gguf:F16
# Run inference directly in the terminal:
./llama-cli -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16

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 prism-ml/Ternary-Bonsai-2-27B-gguf:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16

Use Docker

docker model run hf.co/prism-ml/Ternary-Bonsai-2-27B-gguf:F16

LM Studio
Jan

vLLM

How to use prism-ml/Ternary-Bonsai-2-27B-gguf with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "prism-ml/Ternary-Bonsai-2-27B-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": "prism-ml/Ternary-Bonsai-2-27B-gguf",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/prism-ml/Ternary-Bonsai-2-27B-gguf:F16

Ollama
How to use prism-ml/Ternary-Bonsai-2-27B-gguf with Ollama:
```
ollama run hf.co/prism-ml/Ternary-Bonsai-2-27B-gguf:F16
```
Unsloth Desktop

How to use prism-ml/Ternary-Bonsai-2-27B-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "prism-ml/Ternary-Bonsai-2-27B-gguf:F16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use prism-ml/Ternary-Bonsai-2-27B-gguf with Docker Model Runner:
```
docker model run hf.co/prism-ml/Ternary-Bonsai-2-27B-gguf:F16
```

Lemonade

How to use prism-ml/Ternary-Bonsai-2-27B-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull prism-ml/Ternary-Bonsai-2-27B-gguf:F16

Run and chat with the model

lemonade run user.Ternary-Bonsai-2-27B-gguf-F16

List all available models

lemonade list

Hermes Agent

How to use prism-ml/Ternary-Bonsai-2-27B-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 prism-ml/Ternary-Bonsai-2-27B-gguf:F16

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 prism-ml/Ternary-Bonsai-2-27B-gguf:F16

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use prism-ml/Ternary-Bonsai-2-27B-gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf prism-ml/Ternary-Bonsai-2-27B-gguf:F16

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 "prism-ml/Ternary-Bonsai-2-27B-gguf:F16" \
  --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"

New discussion

Resources

View closed (6)

I took Bonsai 2 27B apart, part 2: the format is public, the last 8% is ordinary ternary QAT.

#62 opened 25 minutes ago by

SkyIsNotGreen

Set the model name and min_p in the GGUF metadata (weights unchanged)

#59 opened about 11 hours ago by

bri-prism

for 16gb vram context size <=114688 is fast. Pelican is gorgeous :)

👍🤯 3

#56 opened about 14 hours ago by

KottCh

intel alchemist kernel

#55 opened about 14 hours ago by

Demilenos

Real KLD testing against Unsloth's gguf BF16 of Qwen 3.8-27B

👀👍 2

#54 opened about 18 hours ago by

mrumel

When will they release the official version with ROCM or Vulkan support for the 6900XT?

#53 opened 1 day ago by

aidoluiz

AMD works just wonderfully, here is how:

#52 opened 2 days ago by

RegisteredWednesday

issue : 6800 (rdna 2) gpu card on PrismML-Eng llama.cpp . cant run model by unsloth Studio and ..

#51 opened 2 days ago by

myhugginfacegacc

SYCL backend: any speculative type collapses performance (even target prefill drops ~200x) - draft model itself is healthy

#50 opened 2 days ago by

Yoo00ooOO

How does the 2080 Ti 22G perform with this model?

#49 opened 2 days ago by

coresen

PLEASE stop lying about the "intelligence" of the model.

🔥👍 21

#47 opened 3 days ago by

Splarkszter

在4060笔记本，8G下，优化到45Token/S之后，再提10%到50Token/S，但这还不是上限...理论上限可能高达65Token/S

👍 2

#46 opened 3 days ago by

SuperLogic

Solved "xhigh" loop and include MTP. Test on 4080 12G VRAM laptop, reach ~60 t/s

👍🚀 2

#45 opened 3 days ago by

zhijin123

Built a multi agent orchestrator with it

🚀 2

#43 opened 4 days ago by

anubhav200

dspark will improve intelligence/bit

#42 opened 4 days ago by

john1248

3060 rtx 8 gb vram here ... share your preset !

#41 opened 4 days ago by

oytaub

First time used... Failed basic tool calls immediately. (Q2 variant)

👍 5

#40 opened 4 days ago by

laser50

在4060笔记本，8G下，优化到45Token/S，已经达到大厂的Token接口速度，彻底实现Token自由，大家再也不需要去大厂订阅。

#39 opened 4 days ago by

SuperLogic

Ternary-Bonsai-2-27B on an RTX 2060 SUPER 8GB — Japanese supplied-context reasoning is the biggest surprise

#38 opened 4 days ago by

mktnhr

LM Studio says no

#37 opened 4 days ago by

Vort

No sure why all the hate?

❤️ 1

#36 opened 4 days ago by

Mogsie

It is fighting with the harness, chat template and blaming the human operator

#35 opened 4 days ago by

gbuzhf

OMG

#34 opened 4 days ago by

darkmatter2222

Density table divides Bonsai by the ideal 5.80 GB; competitors by shipped files

#33 opened 5 days ago by

makerportal

NB啊

#32 opened 5 days ago by

lpiao3825

Really useful quantified small models PQ2_0. Have good intelligence and pretty fast. (With evalscope testing result)

🧠🚀 4

#31 opened 5 days ago by

zhijin123

confused on sampling parameters

👀 1

#30 opened 5 days ago by

Jcamacho05

Bonsai 2 27B Uncensored

#29 opened 5 days ago by

e-RHM-e

Joining your group

#28 opened 5 days ago by

Mohammedkarimi

Surprisingly good PTQ1_0 quantization – and ~39.5 tok/s on an RTX 4070

👍🔥 7

#27 opened 5 days ago by

TheWegemann

出一个Q6的试试？看看能力能不能保留FP16的99.99%？

#26 opened 5 days ago by

kelei999999

Update README.md

#25 opened 5 days ago by

rikunarita-3

Something is wrong

#24 opened 5 days ago by

kashish4u

In-file MTP speculation refused on Hadamard-folded weights (prism-b10683); ~1.34x decode after the 7-line embedding fix

#23 opened 5 days ago by

zhaokeqi

This model is HORRIBLE for coding. don't waste your time.

🧠👍 14

#22 opened 5 days ago by

avicohen

Hi, can we have the safetensor pls.

#21 opened 6 days ago by

jezzza1401

swift bonsai 🚀

🚀 3

#20 opened 6 days ago by

iqueak

Well, your released fork on GH ends on a "not found" page.

👍 1

#19 opened 6 days ago by

TheWegemann

Potential quality issue: Stuck on Reasoning Loop on a hard task

#18 opened 6 days ago by

DemetriusMichael

I wonder how the model would have performed without quantizing embedding and other sensitive parts down to 1.72bpw.

#17 opened 6 days ago by

tigerjjw53

This breaks the model.

👀👍 3

#16 opened 6 days ago by

phil111

How well does this model perform on agentic tasks?

#15 opened 6 days ago by

akiko19191

Other formats ? - AWB , GPTQ , etc

#14 opened 6 days ago by

TahirC

will this work on AMD card

#13 opened 6 days ago by

tapanpatro

Request: a Q2_0 group-64 (mainline-compatible) export for Bonsai 2 27B?

➕👀 2

#12 opened 6 days ago by

AlexGabbia

生成一个svg动画：鹈鹕骑自行车，用H5给我展示下

😔 7

#11 opened 6 days ago by

sunnyboxs

The PQ2_0 tensor does not support GPU buffers, which results in the inference speed of mi50/mi60 being only 0.1 t/s.

➕ 2

#10 opened 6 days ago by

shenghedadi

GGUF versions available via Torrent

#9 opened 6 days ago by

baiomys

Request: Refusal-free (abliterated) weights for industrial, SIEM & agentic workflows

➕ 3

#8 opened 6 days ago by

amaxcz

when vulkan?!

👍➕ 18

#7 opened 6 days ago by

Amin420