Instructions to use OpenPathAI/Onyx-1.5-2B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use OpenPathAI/Onyx-1.5-2B with Transformers:

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

pipe = pipeline("text-generation", model="OpenPathAI/Onyx-1.5-2B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("OpenPathAI/Onyx-1.5-2B")
model = AutoModelForCausalLM.from_pretrained("OpenPathAI/Onyx-1.5-2B", device_map="auto")
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

llama.cpp

How to use OpenPathAI/Onyx-1.5-2B 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 OpenPathAI/Onyx-1.5-2B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

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 OpenPathAI/Onyx-1.5-2B:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

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 OpenPathAI/Onyx-1.5-2B:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

Use Docker

docker model run hf.co/OpenPathAI/Onyx-1.5-2B:Q4_K_M

LM Studio
Jan

vLLM

How to use OpenPathAI/Onyx-1.5-2B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "OpenPathAI/Onyx-1.5-2B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "OpenPathAI/Onyx-1.5-2B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/OpenPathAI/Onyx-1.5-2B:Q4_K_M

SGLang

How to use OpenPathAI/Onyx-1.5-2B 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 "OpenPathAI/Onyx-1.5-2B" \
    --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": "OpenPathAI/Onyx-1.5-2B",
		"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 "OpenPathAI/Onyx-1.5-2B" \
        --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": "OpenPathAI/Onyx-1.5-2B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use OpenPathAI/Onyx-1.5-2B with Ollama:
```
ollama run hf.co/OpenPathAI/Onyx-1.5-2B:Q4_K_M
```
Unsloth Desktop

How to use OpenPathAI/Onyx-1.5-2B with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

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": "OpenPathAI/Onyx-1.5-2B:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use OpenPathAI/Onyx-1.5-2B with Docker Model Runner:
```
docker model run hf.co/OpenPathAI/Onyx-1.5-2B:Q4_K_M
```

Lemonade

How to use OpenPathAI/Onyx-1.5-2B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull OpenPathAI/Onyx-1.5-2B:Q4_K_M

Run and chat with the model

lemonade run user.Onyx-1.5-2B-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use OpenPathAI/Onyx-1.5-2B with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

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 OpenPathAI/Onyx-1.5-2B:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use OpenPathAI/Onyx-1.5-2B with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf OpenPathAI/Onyx-1.5-2B:Q4_K_M

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 "OpenPathAI/Onyx-1.5-2B:Q4_K_M" \
  --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"

Onyx-1.5-2B

Model Bahasa Indonesia Eksperimental

Ringkasan

Onyx-1.5-2B adalah model bahasa yang dikembangkan sebagai eksperimen untuk Bahasa Indonesia. Model ini dibangun di atas arsitektur Qwen 2 dan dilatih menggunakan metode LoRA pada dataset Bahasa Indonesia.

Peringatan: Model ini merupakan eksperimen dan belum layak untuk aplikasi produksi. Kualitas output masih terbatas dan sering menghasilkan informasi yang tidak akurat. Gunakan hanya untuk tujuan pembelajaran.

Detail Model

Komponen	Detail
Arsitektur	Qwen 2
Jenis	Causal Language Model (Decoder-only)
Total Parameter	1,5 Miliar
Parameter Trainable	~41,9 Juta (LoRA)
Metode	LoRA (Low-Rank Adaptation)
Dataset	`AksaraLLM/aksara-mega-sft` (64.797 samples)
Epoch	3
Precision	FP16
Format Tersedia	Hugging Face (safetensors) + GGUF

Konfigurasi LoRA

Parameter	Nilai
LoRA Rank (r)	16
LoRA Alpha	32
LoRA Dropout	0,05
Target Modules	`q_proj`, `k_proj`, `v_proj`, `o_proj`, `gate_proj`, `up_proj`, `down_proj`

Keterbatasan

Model ini memiliki keterbatasan yang signifikan:

Keterbatasan	Deskripsi
Akurasi Fakta	Sering menghasilkan informasi yang salah
Repetition	Kadang mengulang kata atau frasa yang sama
Bahasa Daerah	Belum mampu memahami bahasa daerah dengan baik
Identitas	Tidak mengetahui siapa pembuatnya
Konsistensi	Output tidak selalu koheren
Ukuran	1.5B parameter memiliki kapasitas terbatas

Model ini BUKAN untuk:

Aplikasi produksi
Pertanyaan faktual yang membutuhkan akurasi
Tugas yang membutuhkan penalaran kompleks
Layanan publik

Cara Penggunaan

Format Prompt


Instruction:

[Pertanyaan atau instruksi Anda]

Response:

[Jawaban model]

GGUF (untuk llama.cpp / LM Studio)

# Download GGUF
huggingface-cli download OpenPathAI/Onyx-1.5-2B \
  --include "gguf/onyx-1.5-2b-Q4_K_M.gguf" \
  --local-dir ./models

# Jalankan dengan llama.cpp
./llama-cli -m ./models/gguf/onyx-1.5-2b-Q4_K_M.gguf \
  -p "### Instruction:\nHalo, apa kabar?\n\n### Response:\n" \
  -n 256 --temp 0.7 -cnv

Hugging Face (transformers)

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

MODEL_NAME = "OpenPathAI/Onyx-1.5-2B"

tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForCausalLM.from_pretrained(
    MODEL_NAME,
    torch_dtype=torch.float16,
    device_map="auto",
)

prompt = "### Instruction:\nHalo, apa kabar?\n\n### Response:\n"
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

outputs = model.generate(
    **inputs,
    max_new_tokens=256,
    temperature=0.7,
    top_p=0.9,
    do_sample=True,
    pad_token_id=tokenizer.eos_token_id,
)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

File yang Tersedia

File Deskripsi model.safetensors Bobot model (format Hugging Face) config.json Konfigurasi model tokenizer.json Tokenizer gguf/onyx-1.5-2b-f16.gguf GGUF FP16 (~~3.5 GB) gguf/onyx-1.5-2b-Q4_K_M.gguf GGUF Q4_K_M (~~1.1 GB)

Lisensi

Model ini dilisensikan di bawah Apache License 2.0.

Sitasi

@misc{onyx-1-5-2b,
  author = {OpenPathAI},
  title = {Onyx-1.5-2B: Experimental Indonesian Language Model},
  year = {2026},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/OpenPathAI/Onyx-1.5-2B}}
}

Kontak

Saluran Detail Organisasi OpenPathAI Hugging Face OpenPathAI Dataset Aksara Mega SFT

Dikembangkan oleh OpenPathAI

Model ini merupakan eksperimen dan belum layak untuk aplikasi produksi.

Downloads last month: 3

Safetensors

Model size

2B params

Tensor type

F32

F16

Dataset used to train OpenPathAI/Onyx-1.5-2B

Collection including OpenPathAI/Onyx-1.5-2B

Onyx Models

Collection

1 item • Updated 4 days ago