Instructions to use NotASI/FineTome-Llama3.2-3B-1002 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use NotASI/FineTome-Llama3.2-3B-1002 with Transformers:

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

pipe = pipeline("text-generation", model="NotASI/FineTome-Llama3.2-3B-1002")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("NotASI/FineTome-Llama3.2-3B-1002")
model = AutoModelForCausalLM.from_pretrained("NotASI/FineTome-Llama3.2-3B-1002")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use NotASI/FineTome-Llama3.2-3B-1002 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "NotASI/FineTome-Llama3.2-3B-1002"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "NotASI/FineTome-Llama3.2-3B-1002",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/NotASI/FineTome-Llama3.2-3B-1002

SGLang

How to use NotASI/FineTome-Llama3.2-3B-1002 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 "NotASI/FineTome-Llama3.2-3B-1002" \
    --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": "NotASI/FineTome-Llama3.2-3B-1002",
		"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 "NotASI/FineTome-Llama3.2-3B-1002" \
        --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": "NotASI/FineTome-Llama3.2-3B-1002",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use NotASI/FineTome-Llama3.2-3B-1002 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 NotASI/FineTome-Llama3.2-3B-1002 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 NotASI/FineTome-Llama3.2-3B-1002 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for NotASI/FineTome-Llama3.2-3B-1002 to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="NotASI/FineTome-Llama3.2-3B-1002",
    max_seq_length=2048,
)

Docker Model Runner
How to use NotASI/FineTome-Llama3.2-3B-1002 with Docker Model Runner:
```
docker model run hf.co/NotASI/FineTome-Llama3.2-3B-1002
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

IMPORTANT

In case you got the following error:

exception: data did not match any variant of untagged enum modelwrapper at line 1251003 column 3

Please upgrade your transformer package, that is, use the following code:

pip install --upgrade "transformers>=4.45"

Uploaded model

Developed by: NotASI
License: apache-2.0
Finetuned from model : unsloth/Llama-3.2-3B-Instruct-bnb-4bit

Details

This model was trained on mlabonne/FineTome-100k for 2 epochs with rslora + qlora, and achieve the final training loss: 0.596400.

This model follows the same chat template as the base model one.

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric	Value
Avg.	16.60
IFEval (0-Shot)	54.74
BBH (3-Shot)	19.52
MATH Lvl 5 (4-Shot)	5.29
GPQA (0-shot)	0.11
MuSR (0-shot)	3.96
MMLU-PRO (5-shot)	15.96

Downloads last month: 8

Safetensors

Model size

3B params

Tensor type

F16

Model tree for NotASI/FineTome-Llama3.2-3B-1002

Base model

meta-llama/Llama-3.2-3B-Instruct

Quantized

unsloth/Llama-3.2-3B-Instruct-bnb-4bit

Finetuned

(245)

this model

Merges

1 model

Quantizations

2 models

Dataset used to train NotASI/FineTome-Llama3.2-3B-1002

Collection including NotASI/FineTome-Llama3.2-3B-1002

Tiny Fine Tunes🤏

Collection

A collection of SLM (Small Language Model) fine tunes. • 4 items • Updated Oct 10, 2024

Evaluation results

strict accuracy on IFEval (0-Shot)
Open LLM Leaderboard

54.740
normalized accuracy on BBH (3-Shot)
Open LLM Leaderboard

19.520
exact match on MATH Lvl 5 (4-Shot)
Open LLM Leaderboard

5.290
acc_norm on GPQA (0-shot)
Open LLM Leaderboard

0.110
acc_norm on MuSR (0-shot)
Open LLM Leaderboard

3.960
accuracy on MMLU-PRO (5-shot)
test set Open LLM Leaderboard

15.960