Instructions to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("/storage/ukp/shared/shared_model_weights/models--unsloth--Phi-4-mini-reasoning-unsloth-bnb-4bit")
model = PeftModel.from_pretrained(base_model, "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT")

Transformers

How to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT with Transformers:

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

pipe = pipeline("text-generation", model="haritzpuerto/unsloth-Phi-4-3.8B-IF-RT")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("haritzpuerto/unsloth-Phi-4-3.8B-IF-RT", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/haritzpuerto/unsloth-Phi-4-3.8B-IF-RT

SGLang

How to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT 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 "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT" \
    --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": "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT",
		"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 "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT" \
        --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": "haritzpuerto/unsloth-Phi-4-3.8B-IF-RT",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Studio

How to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT 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 haritzpuerto/unsloth-Phi-4-3.8B-IF-RT 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 haritzpuerto/unsloth-Phi-4-3.8B-IF-RT to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for haritzpuerto/unsloth-Phi-4-3.8B-IF-RT to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="haritzpuerto/unsloth-Phi-4-3.8B-IF-RT",
    max_seq_length=2048,
)

Docker Model Runner
How to use haritzpuerto/unsloth-Phi-4-3.8B-IF-RT with Docker Model Runner:
```
docker model run hf.co/haritzpuerto/unsloth-Phi-4-3.8B-IF-RT
```

Model Card

This model is a fine-tuned version of unsloth/Phi-4-mini-reasoning-unsloth-bnb-4bit.

Quick start

from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="None", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

Training procedure

This model was trained with SFT.

Framework versions

PEFT 0.17.1
TRL: 0.23.0
Transformers: 4.57.1
Pytorch: 2.8.0
Datasets: 4.3.0
Tokenizers: 0.22.1

Citations

@misc{puerto2026controllablereasoningmodelsprivate,
      title={Controllable Reasoning Models Are Private Thinkers}, 
      author={Haritz Puerto and Haonan Li and Xudong Han and Timothy Baldwin and Iryna Gurevych},
      year={2026},
      eprint={2602.24210},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2602.24210}, 
}