Instructions to use Mostafa8Mehrabi/qwen3-30m-fp16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Mostafa8Mehrabi/qwen3-30m-fp16 with Transformers:

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

pipe = pipeline("text-generation", model="Mostafa8Mehrabi/qwen3-30m-fp16")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Mostafa8Mehrabi/qwen3-30m-fp16")
model = AutoModelForCausalLM.from_pretrained("Mostafa8Mehrabi/qwen3-30m-fp16")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Mostafa8Mehrabi/qwen3-30m-fp16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Mostafa8Mehrabi/qwen3-30m-fp16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Mostafa8Mehrabi/qwen3-30m-fp16",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/Mostafa8Mehrabi/qwen3-30m-fp16

SGLang

How to use Mostafa8Mehrabi/qwen3-30m-fp16 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 "Mostafa8Mehrabi/qwen3-30m-fp16" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Mostafa8Mehrabi/qwen3-30m-fp16",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "Mostafa8Mehrabi/qwen3-30m-fp16" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Mostafa8Mehrabi/qwen3-30m-fp16",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use Mostafa8Mehrabi/qwen3-30m-fp16 with Docker Model Runner:
```
docker model run hf.co/Mostafa8Mehrabi/qwen3-30m-fp16
```

Qwen3-30M with GPT-2 Tokenizer (FP16)

A 30M parameter version of Qwen3-0.6B using GPT-2's tokenizer for better compatibility, optimized with FP16 precision.

Model Details

Base Model: Qwen/Qwen3-0.6B
Architecture: Qwen3 (8 layers, 224 hidden size)
Parameters: ~35M (reduced from 637M)
Tokenizer: GPT-2 (50,257 vocabulary)
Vocabulary: Reduced from 151,936 to 50,257 tokens
Precision: FP16 (half precision for memory efficiency)
Model Size: ~60MB (vs ~120MB in FP32)

Architecture Specifications

Layers: 8 transformer layers
Hidden Size: 224
Intermediate Size: 896 (4x hidden_size)
Attention Heads: 8
Key-Value Heads: 8
Max Position Embeddings: 32,768
Activation: SiLU

Usage

from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

# Load with automatic fp16 support
tokenizer = AutoTokenizer.from_pretrained("Mostafa8Mehrabi/qwen3-30m-fp16")
model = AutoModelForCausalLM.from_pretrained(
    "Mostafa8Mehrabi/qwen3-30m-fp16",
    torch_dtype=torch.float16,  # Explicitly use fp16
    device_map="auto"  # Automatically place on available device
)

# For GPU inference (recommended)
# model = model.to("cuda") # if you have a GPU

inputs = tokenizer("Hello, how are you?", return_tensors="pt")
# Move inputs to same device as model if using GPU
# inputs = {k: v.to(model.device) for k, v in inputs.items()}

outputs = model.generate(**inputs, max_length=50, do_sample=True, temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Key Features

✅ FP16 Precision: 50% smaller model size, faster inference
✅ 30M Parameters: Ultra-lightweight for edge deployment
✅ 8 Layers: Balanced depth for performance
✅ Standard GPT-2 tokenizer (no trust_remote_code)
✅ Compatible vocabulary sizes
✅ SafeTensors format for faster loading
✅ Works like any HuggingFace model
✅ 21x smaller than original Qwen3-0.6B
✅ GPU optimized for efficient inference

Architecture Comparison

Component	Original	This Model
Parameters	637M	~35M
Vocabulary	151,936	50,257
Hidden Size	1024	224
Layers	28	8
Intermediate Size	4096	896
Attention Heads	16	8
Tokenizer	Qwen3	GPT-2
Precision	FP32	FP16
Model Size	~1.2GB	~60MB