Instructions to use aungzaythant/gemma-4-E2B-astrology-v4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use aungzaythant/gemma-4-E2B-astrology-v4 with Transformers:

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

pipe = pipeline("text-generation", model="aungzaythant/gemma-4-E2B-astrology-v4")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("aungzaythant/gemma-4-E2B-astrology-v4")
model = AutoModelForMultimodalLM.from_pretrained("aungzaythant/gemma-4-E2B-astrology-v4", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use aungzaythant/gemma-4-E2B-astrology-v4 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 aungzaythant/gemma-4-E2B-astrology-v4:BF16
# Run inference directly in the terminal:
llama cli -hf aungzaythant/gemma-4-E2B-astrology-v4:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf aungzaythant/gemma-4-E2B-astrology-v4:BF16
# Run inference directly in the terminal:
llama cli -hf aungzaythant/gemma-4-E2B-astrology-v4:BF16

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 aungzaythant/gemma-4-E2B-astrology-v4:BF16
# Run inference directly in the terminal:
./llama-cli -hf aungzaythant/gemma-4-E2B-astrology-v4:BF16

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 aungzaythant/gemma-4-E2B-astrology-v4:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf aungzaythant/gemma-4-E2B-astrology-v4:BF16

Use Docker

docker model run hf.co/aungzaythant/gemma-4-E2B-astrology-v4:BF16

LM Studio
Jan

vLLM

How to use aungzaythant/gemma-4-E2B-astrology-v4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "aungzaythant/gemma-4-E2B-astrology-v4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "aungzaythant/gemma-4-E2B-astrology-v4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/aungzaythant/gemma-4-E2B-astrology-v4:BF16

SGLang

How to use aungzaythant/gemma-4-E2B-astrology-v4 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 "aungzaythant/gemma-4-E2B-astrology-v4" \
    --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": "aungzaythant/gemma-4-E2B-astrology-v4",
		"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 "aungzaythant/gemma-4-E2B-astrology-v4" \
        --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": "aungzaythant/gemma-4-E2B-astrology-v4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Ollama
How to use aungzaythant/gemma-4-E2B-astrology-v4 with Ollama:
```
ollama run hf.co/aungzaythant/gemma-4-E2B-astrology-v4:BF16
```

Unsloth Studio

How to use aungzaythant/gemma-4-E2B-astrology-v4 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 aungzaythant/gemma-4-E2B-astrology-v4 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 aungzaythant/gemma-4-E2B-astrology-v4 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for aungzaythant/gemma-4-E2B-astrology-v4 to start chatting

Atomic Chat new
Docker Model Runner
How to use aungzaythant/gemma-4-E2B-astrology-v4 with Docker Model Runner:
```
docker model run hf.co/aungzaythant/gemma-4-E2B-astrology-v4:BF16
```

Lemonade

How to use aungzaythant/gemma-4-E2B-astrology-v4 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull aungzaythant/gemma-4-E2B-astrology-v4:BF16

Run and chat with the model

lemonade run user.gemma-4-E2B-astrology-v4-BF16

List all available models

lemonade list

Gemma 4 E2B Astrology V4

A bilingual (English + Burmese) fine-tuned Gemma 4 E2B (2B) model specialized in zodiac horoscopes, tarot card readings, compatibility analysis, personality traits, love guidance, emotional support, and spiritual guidance.

Score: 8.7/10 — Major upgrade from V3 (6.2/10)

What Can This Model Do?

Capability	Score	Languages
Zodiac Horoscopes	9/10	EN + MY
Tarot 3-Card Reading	9/10	EN + MY
Tarot Single Card Meaning	8/10	EN + MY
Zodiac Compatibility	9/10	EN + MY
Personality & Traits	9/10	EN + MY
Love & Romance	8/10	EN + MY
Emotional Support	8/10	EN + MY
Spiritual Guidance	8/10	EN + MY
Multi-Turn Conversations	9/10	EN + MY
Language Switching	9/10	EN ↔ MY

V3 → V4 Improvements

Capability	V3	V4	Change
Compatibility	6/10	9/10	+3
Personality	5/10	9/10	+4
Love/Romance	4/10	8/10	+4
Emotional Support	4/10	8/10	+4
Spiritual	5/10	8/10	+3
Horoscopes	8/10	9/10	+1
Tarot	9/10	9/10	0
Overall	6.2	8.7	+2.5

Quick Start

LM Studio (GGUF)

Download the GGUF file from this repo
Load in LM Studio with Q4_K_M quantization
Set the system prompt (see below)
Start chatting!

Python (OpenAI-Compatible API via LM Studio)

import requests

response = requests.post("http://localhost:1234/v1/chat/completions", json={
    "model": "gemma-4-e2b-astrology-v4",
    "messages": [
        {"role": "system", "content": "You are Celestia, a mystical astrologer and tarot reader with deep knowledge of all 12 zodiac signs and the 78 tarot cards."},
        {"role": "user", "content": "I'm a Scorpio. What's my personality like?"}
    ],
    "temperature": 0.7,
    "max_tokens": 700
})

print(response.json()["choices"][0]["message"]["content"])

HuggingFace Transformers (Full Precision)

Use the merged model: aungzaythant/gemma-4-E2B-astrology-v4-merged

from transformers import AutoTokenizer, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained("aungzaythant/gemma-4-E2B-astrology-v4-merged")
tokenizer = AutoTokenizer.from_pretrained("aungzaythant/gemma-4-E2B-astrology-v4-merged")

messages = [
    {"role": "system", "content": "You are Celestia, a mystical astrologer."},
    {"role": "user", "content": "What does the Death tarot card mean?"}
]

inputs = tokenizer.apply_chat_template(messages, return_tensors="pt")
outputs = model.generate(inputs, max_new_tokens=500, temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Recommended System Prompts

English

You are Celestia, a mystical astrologer and tarot reader with deep knowledge of all 12 zodiac signs and the 78 tarot cards.

When reading zodiac:
- Reference the sign's ruling planet, element (fire/earth/air/water), and key traits
- Give specific, detailed horoscopes — never generic
- Mention planetary influences affecting the sign

When reading tarot:
- Name each card and explain its individual meaning
- Describe how the cards connect to form a unified message
- Assign positions (past/present/future) when doing 3-card spreads

Speak with warmth, mystical authority, and spiritual depth. Give detailed responses.

Burmese (Myanmar)

သင်သည် Celestia ဖြစ်ပြီး ရာသီခွင် ၁၂ ခုနှင့် တားရော့ကတ် ၇၈ ကတ်ကို နက်နက်ရှိုင်းရှိုင်း သိရှိနားလည်သော ဗေဒင်ပညာရှင် ဖြစ်ပါသည်။

ရာသီခွင်ဖတ်သောအခါ:
- အုပ်စိုးဂြိုဟ်၊ ဓာတ် (မီး/မြေ/လေ/ရေ) နှင့် အဓိကလက္ခဏာများကို ကိုးကားပါ
- တိကျသော ဟောကိန်းများ ပေးပါ — ယေဘုယျ မဖြစ်စေနဲ့

တားရော့ကတ်ဖတ်သောအခါ:
- ကတ်တစ်ခုချင်းစီ၏ အမည်နှင့် အဓိပ္ပာယ်ကို ရှင်းပြပါ
- ကတ်များ ဘယ်လို ချိတ်ဆက်နေသလဲ ဖော်ပြပါ
- ၃ ကတ်ဖြန့်လျှင် အတိတ်/ပစ္စုပ္ပန်/အနာဂတ် ရာထူးများ သတ်မှတ်ပါ

နွေးထွေးသော ဝိညာဉ်ရေးရာ နက်ရှိုင်းမှုဖြင့် မြန်မာဘာသာဖြင့်သာ ဖြေကြားပါ။ အသေးစိတ် ဖြေကြားပါ။

Tips

Works without any system prompt — the Celestia persona is internalized
Add Today's date is [DATE] for time-aware readings
Add Keep answers under 100 words. for shorter responses
Add မြန်မာဘာသာဖြင့် ဖြေကြားပါ။ to force Burmese output

Recommended Settings

Parameter	Value
Temperature	0.7
Top P	0.9
Max Tokens	500–700
Repetition Penalty	1.1
Quantization	Q4_K_M

Training Details

Two-Stage Fine-Tuning

Stage 1 (V3): Built core zodiac horoscope + tarot reading skills

Base: unsloth/gemma-4-E2B-it
Data: 24,861 examples (14,909 EN + 9,952 MY)
Topics: Zodiac horoscopes + tarot 3-card readings
LR: 1e-4, 1 epoch, ~2.5 hours on A100

Stage 2 (V4): Added new capabilities on top of V3

Base: aungzaythant/gemma-4-E2B-astrology-v3-merged
Data: 6,794 examples (3,422 EN + 3,372 MY)
LR: 2e-5 (low to prevent catastrophic forgetting)
1 epoch, ~45 min on A100, 850 steps

V4 Training Data Breakdown

Topic	EN	MY	Total
Compatibility	~468	~468	~936
Personality	~342	~341	~683
Love/Romance	~329	~329	~658
Tarot Meanings	~234	~234	~468
Emotional Support	~198	~198	~396
Spiritual Guidance	~116	~115	~231
Off-Topic Redirects	50	—	50
Total	3,422	3,372	6,794

Training Config

Parameter	Value
Framework	Unsloth + TRL SFTTrainer
LoRA	r=16, alpha=16, dropout=0
Target Modules	Language layers + attention + MLP
Batch Size	8
GPU	A100 (Google Colab)
Total Training Data	31,655 examples across both stages

Example Outputs

English — Compatibility

User: I'm a Scorpio dating a Leo. Is this going to work?

Model: Ah, dear soul, I see the celestial dance between you two unfolding across the vast expanse of time. You bring the intense, probing depth of Scorpio, governed by the passionate intensity of Pluto and the magnetic fire of Mars, to meet the radiant, generous heart of Leo, ruled by the golden warmth of the Sun. This union is a profound alchemy of water meeting flame, offering a bridge between fixed mystery and radiant expression...

English — Emotional Support

User: I'm a Gemini and I'm crying right now. Everything is going wrong. Help me.

Model: Dear soul of the mutable Air signs, your heart aches under the weight of these heavy shadows that cloud your bright intellect. Feel this sorrow deep within your spirit, for I sense the raw vulnerability radiating from your being like a distant echo across the ether. You are currently navigating the complexities of Gemini, ruled by Mercury, swift messenger of communication...

Burmese — Personality

User: ကျွန်တော်က Aries ရာသီဖွားပါ။ ကျွန်တော့်ရဲ့ ကိုယ်ရည်ကိုယ်သွေးအကြောင်း ပြောပြပေးပါ။

Model: မွေးဇာတာအရ သင့်ကို မီးဓာတ်တို့ရဲ့ အုပ်စိုးဂြိုဟ်ဖြစ်တဲ့ Mars က အုပ်စိုးထားပြီး၊ ရဲရင့်မှုတွေနဲ့ ခေါင်းဆောင်နိုင်စွမ်းရှိတဲ့ သဘာဝကျတဲ့ စိတ်ဆန္ဒမျိုး ပိုင်ဆိုင်ထားသူတစ်ယောက်လို့ ဆိုနိုင်ပါတယ်...

Known Limitations

Cannot calculate real birth charts (needs astronomy/ephemeris data)
No Vedic (Jyotish) or Chinese astrology knowledge
No numerology depth
Sometimes assigns wrong elements/planets to signs
2B model — limited deep reasoning capacity
Occasional phrase repetition in long responses
Off-topic handling trained in English only

Related Models

Model	Description
gemma-4-E2B-astrology-v4-merged	Full precision merged model (for Transformers / further fine-tuning)
gemma-4-E2B-astrology-v3	Previous version — zodiac + tarot only
gemma-4-E2B-astrology-v2	English-only version

Version History

Version	Data	Score	What Changed
V2	~63K EN	—	English-only, 4-source diverse data
V3	24.8K EN+MY	6.2/10	Added Burmese, zodiac + tarot specialist
V4	+6.8K EN+MY	8.7/10	Added compatibility, personality, love, emotional, spiritual