Instructions to use rogersam/tinyllama-instruct-lite-v1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use rogersam/tinyllama-instruct-lite-v1 with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("unsloth/tinyllama-bnb-4bit")
model = PeftModel.from_pretrained(base_model, "rogersam/tinyllama-instruct-lite-v1")

Transformers

How to use rogersam/tinyllama-instruct-lite-v1 with Transformers:

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

pipe = pipeline("text-generation", model="rogersam/tinyllama-instruct-lite-v1")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("rogersam/tinyllama-instruct-lite-v1", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use rogersam/tinyllama-instruct-lite-v1 with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/rogersam/tinyllama-instruct-lite-v1

SGLang

How to use rogersam/tinyllama-instruct-lite-v1 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 "rogersam/tinyllama-instruct-lite-v1" \
    --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": "rogersam/tinyllama-instruct-lite-v1",
		"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 "rogersam/tinyllama-instruct-lite-v1" \
        --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": "rogersam/tinyllama-instruct-lite-v1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Unsloth Studio new

How to use rogersam/tinyllama-instruct-lite-v1 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 rogersam/tinyllama-instruct-lite-v1 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 rogersam/tinyllama-instruct-lite-v1 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for rogersam/tinyllama-instruct-lite-v1 to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="rogersam/tinyllama-instruct-lite-v1",
    max_seq_length=2048,
)

Docker Model Runner
How to use rogersam/tinyllama-instruct-lite-v1 with Docker Model Runner:
```
docker model run hf.co/rogersam/tinyllama-instruct-lite-v1
```

TinyLlama Instruct Lite v1

📌 Model Summary

rogersam/tinyllama-instruct-lite-v1 is a LoRA fine-tuned TinyLlama model using Unsloth.
It is designed for instruction-following tasks in English + Tamil, such as:

General Q&A
Summarization
Basic math & reasoning
English ↔ Tamil translation

This project demonstrates how a lightweight 1B model can be adapted for multiple domains with limited resources.

🔎 Model Details

Developed by: Roger Samuel J (Hugging Face Profile)
Model type: Causal LM (decoder-only)
Languages: English, Tamil
License: Same as base model (TinyLlama)
Fine-tuned from: unsloth/tinyllama-bnb-4bit
Method: LoRA via PEFT + Unsloth

📂 Model Sources

Model Repo: rogersam/tinyllama-instruct-lite-v1
Base Model: unsloth/tinyllama-bnb-4bit

💡 Uses

Direct Use

Running lightweight instruction tasks on CPU/GPU
Translating English ↔ Tamil sentences
Answering short questions and reasoning queries
Summarizing small texts

Out-of-Scope

Sensitive decision-making (finance, healthcare, law)
Long context generation (>512 tokens)
Production-grade chatbots

⚠️ Bias, Risks & Limitations

Small dataset → may hallucinate facts
Not aligned for safety or toxicity filtering
Limited Tamil coverage (basic sentences only)

Recommendation: Use for demo & educational purposes only.

🚀 How to Get Started

from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

model_id = "rogersam/tinyllama-instruct-lite-v1"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto")

pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)

prompt = "Translate English to Tamil: How are you?"
print(pipe(prompt, max_new_tokens=50)[0]["generated_text"])

Downloads last month: 4

Model tree for rogersam/tinyllama-instruct-lite-v1

Base model

unsloth/tinyllama-bnb-4bit

Adapter

(38)

this model