CoolFace
Modelpublic

num1notsvn/wicara-56m-chat-GGUF

sourceHugging Faceupdated 22d agoView on Hugging Face
0likes187downloads
Model Card

Wicara 56M Chat — GGUF

This repository contains official GGUF quantized formats of num1notsvn/wicara-56m-chat, an Indonesian language model (56M parameters) trained from scratch on a single consumer laptop (NVIDIA RTX 4050 6 GB).

Available Files

FilenameFormatSizeDescription
wicara-56m-chat-q8_0.ggufQ8_0 (8-bit)~70.6 MBRecommended: Near-lossless 8-bit quantization, ultra-fast inference on CPU/mobile.
wicara-56m-chat-f16.ggufF16 (16-bit)~132.9 MBOriginal unquantized 16-bit floating point precision.

How to Use

1. In LM Studio

  1. 1.Open LM Studio.
  2. 2.Search for num1notsvn/wicara-56m-chat-GGUF or drag-and-drop wicara-56m-chat-q8_0.gguf into your LM Studio models directory (~/.cache/lm-studio/models).
  3. 3.Load the model and start chatting immediately!

2. In Ollama (Direct from Hugging Face)

You can run this model directly in Ollama without manual downloading:

bash
ollama run hf.co/num1notsvn/wicara-56m-chat-GGUF:Q8_0

3. In Ollama (Local Build)

Using the provided Modelfile:

dockerfile
FROM ./wicara-56m-chat-q8_0.gguf

TEMPLATE """{{- if .System }}<|system|>
{{ .System }}<|end|>
{{- end }}
{{- range .Messages }}
{{- if eq .Role "user" }}<|user|>
{{ .Content }}<|end|>
{{- else if eq .Role "assistant" }}<|assistant|>
{{ .Content }}<|end|>
{{- end }}
{{- end }}
<|assistant|>
"""

PARAMETER stop "<|end|>"
PARAMETER temperature 0.7
PARAMETER top_p 0.9
PARAMETER repeat_penalty 1.15
SYSTEM "Kamu adalah asisten AI berbahasa Indonesia yang ramah."

Create and run:

bash
ollama create wicara -f Modelfile
ollama run wicara

Original Model & License