Instructions to use DSTI/DS-RLHF-1.7B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DSTI/DS-RLHF-1.7B with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("unsloth/SmolLM2-1.7B-Instruct")
model = PeftModel.from_pretrained(base_model, "DSTI/DS-RLHF-1.7B")

Transformers

How to use DSTI/DS-RLHF-1.7B with Transformers:

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

pipe = pipeline("text-generation", model="DSTI/DS-RLHF-1.7B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("DSTI/DS-RLHF-1.7B", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use DSTI/DS-RLHF-1.7B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "DSTI/DS-RLHF-1.7B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "DSTI/DS-RLHF-1.7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/DSTI/DS-RLHF-1.7B

SGLang

How to use DSTI/DS-RLHF-1.7B 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 "DSTI/DS-RLHF-1.7B" \
    --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": "DSTI/DS-RLHF-1.7B",
		"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 "DSTI/DS-RLHF-1.7B" \
        --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": "DSTI/DS-RLHF-1.7B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio

How to use DSTI/DS-RLHF-1.7B 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 DSTI/DS-RLHF-1.7B 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 DSTI/DS-RLHF-1.7B to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for DSTI/DS-RLHF-1.7B to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="DSTI/DS-RLHF-1.7B",
    max_seq_length=2048,
)

Docker Model Runner
How to use DSTI/DS-RLHF-1.7B with Docker Model Runner:
```
docker model run hf.co/DSTI/DS-RLHF-1.7B
```

Model Card for Model ID

Model Details

This model is a fine-tuned version of SmolLM2-1.7B-Instruct using ORPO (Odds Ratio Preference Optimization), a reinforcement learning from human feedback (RLHF) method.

Model Description

Base Model: unsloth/SmolLM2-1.7B-Instruct
Fine-tuning Method: ORPO (RLHF alignment)
Dataset: ~1,000 data science–related preference samples (chosen vs. rejected responses).
Objective: Improve model’s ability to generate higher-quality, relevant, and well-structured responses in data science
Language(s) (NLP): English
License: apache-2.0
Dataset: Anas989898/DPO-datascience

Uses

Direct Use

Assisting in data science education (explanations of ML concepts, statistical methods, etc.).
Supporting data analysis workflows with suggestions, reasoning, and structured outputs.
Acting as a teaching assistant for coding/data-related queries.
Providing helpful responses in preference-aligned conversations where correctness and clarity are prioritized.

Bias, Risks, and Limitations

Hallucinations: May still produce incorrect or fabricated facts, code, or references.
Dataset Size: Fine-tuned on only 1K preference pairs, which limits generalization.
Domain Focus: Optimized for data science, but may underperform on other domains.
Not a Substitute for Experts: Should not be used as the sole source for critical decisions in real-world projects.
Bias & Safety: As with all LLMs, may reflect biases present in training data.

How to Get Started with the Model

Use the code below to get started with the model.

from huggingface_hub import login
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel


tokenizer = AutoTokenizer.from_pretrained("unsloth/SmolLM2-1.7B-Instruct",)
base_model = AutoModelForCausalLM.from_pretrained(
    "unsloth/SmolLM2-1.7B-Instruct",
    device_map={"": 0}
)

model = PeftModel.from_pretrained(base_model,"DSTI/DS-RLHF-1.7B")


prompt = """Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.

### Instruction:
{}

### Input:
{}

### Response:
{}"""


inputs = tokenizer(
    [
        prompt.format(
            "You are an AI assistant that helps people find information",
            "What is the k-Means Clustering algorithm and what is it's purpose?", 
            "",  
        )
    ],
    return_tensors="pt",
).to("cuda")


from transformers import TextStreamer

text_streamer = TextStreamer(tokenizer)
_ = model.generate(**inputs, streamer=text_streamer, max_new_tokens=1800)

Citation

If you use this model, please cite:

The source dataset: Anas989898/DPO-datascience

@misc{DS-RLHF-1.7B,
  title  = {ORPO (Odds Ratio Preference Optimization) on data science–related samples},
  author = {Rustam Shiriyev},
  year   = {2025}
}

Framework versions

PEFT 0.15.2

Downloads last month: 2

Model tree for DSTI/DS-RLHF-1.7B

Base model

HuggingFaceTB/SmolLM2-1.7B

Quantized

HuggingFaceTB/SmolLM2-1.7B-Instruct

Finetuned

unsloth/SmolLM2-1.7B-Instruct

Adapter

(178)

this model

DSTI
/

DS-RLHF-1.7B