Instructions to use alsersugasawa/udling with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use alsersugasawa/udling 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 alsersugasawa/udling:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf alsersugasawa/udling:Q4_K_M

Install from WinGet (Windows)

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

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf alsersugasawa/udling:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf alsersugasawa/udling:Q4_K_M

Use Docker

docker model run hf.co/alsersugasawa/udling:Q4_K_M

Llama new
LM Studio
Jan
Ollama
How to use alsersugasawa/udling with Ollama:
```
ollama run hf.co/alsersugasawa/udling:Q4_K_M
```
Unsloth Desktop

How to use alsersugasawa/udling with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf alsersugasawa/udling: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": "alsersugasawa/udling:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use alsersugasawa/udling with Docker Model Runner:
```
docker model run hf.co/alsersugasawa/udling:Q4_K_M
```

Lemonade

How to use alsersugasawa/udling with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull alsersugasawa/udling:Q4_K_M

Run and chat with the model

lemonade run user.udling-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use alsersugasawa/udling with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf alsersugasawa/udling: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 alsersugasawa/udling:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use alsersugasawa/udling with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf alsersugasawa/udling: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 "alsersugasawa/udling: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"

udling

A Hiligaynon (Ilonggo) chat model: a LoRA fine-tune of mistralai/Mistral-7B-Instruct-v0.3, merged and quantised to GGUF (Q4_K_M) for llama.cpp on a Raspberry Pi 5.

udling 1.0.0, trained 2026-10-10 on Hugging Face Jobs (a10g-large, 2 h 40 min, 940 steps, held-out loss 0.884).

It speaks Hiligaynon. It does not know the dictionary: asked what an unfamiliar word means, it is usually wrong. The Udling service answers word questions from a word list instead of from this model for that reason. Do not use the weights alone for translation.


Base model	mistralai/Mistral-7B-Instruct-v0.3 (7.25B parameters)
Method	LoRA, r=16, all linear layers, 2 epochs (scripts/train_udling.py)
Data	`alsersugasawa/udling-hiligaynon`, 8,351 rows; sources and references in SOURCES.md
File	`udling-1.0.0-Q4_K_M.gguf` in `alsersugasawa/udling` (private)
SHA-256	`c9b90534e7d025399b7d88fdbea1c3b45f0dba9756dae5a4ae03072fc06b945f`
Adapter	`alsersugasawa/udling-lora-1.0.0`; merged weights `alsersugasawa/udling-merged-1.0.0`
Code	alser-udling-ai, commit `2fd66a3`
Licence	CC-BY-NC-4.0. Mistral-7B is Apache-2.0; the non-commercial term comes from the training data (PLTAT/alpaca_hiligaynon)
Prompt	`src/udling/prompts.py` - the model expects exactly this system prompt

Assessment

scripts/eval_udling.py, 59 questions, asked of the model directly (reports in docs/eval/).

	Stock Mistral-7B-Instruct-v0.3 Q4_K_M	udling 1.0.0
Replies in Hiligaynon	2/10	8/10
Hiligaynon whatever is asked (English, Tagalog)	1/6	5/6
Word → English (held out)	2/20	4/20
English → word (held out)	0/20	0/20
Identity	0/3	3/3

Limits

It is a model taught a low-resource language from about eight thousand examples, a large share of them machine-translated. It will make mistakes in grammar and vocabulary, and it can state things that are not true. Its replies have not been reviewed by a Hiligaynon speaker. Not for commercial use.

Downloads last month: 23

GGUF

Model size

7B params

Architecture

llama

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for alsersugasawa/udling

Base model

mistralai/Mistral-7B-v0.3

Finetuned

mistralai/Mistral-7B-Instruct-v0.3

Finetuned

(544)

this model