unsloth

Finetune Gemma 3n, Qwen3, Llama 4, Phi-4 & Mistral 2x faster with 80% less VRAM!

Finetune for Free

Notebooks are beginner friendly. Read our guide. Add your dataset, click \”Run All\”, and export your finetuned model to GGUF, Ollama, vLLM or Hugging Face.

• See all our notebooks for: Kaggle, GRPO, TTS & Vision
• See all our models and all our notebooks
• See detailed documentation for Unsloth here

⚡ Quickstart

• Install with pip (recommended) for Linux devices:

pip install unsloth

For Windows install instructions, see here.

? Unsloth.ai News

• Gemma 3n by Google: Read Blog. We uploaded GGUFs, 4-bit models.
• Text-to-Speech (TTS) is now supported, including sesame/csm-1b and STT openai/whisper-large-v3.
• Qwen3 is now supported. Qwen3-30B-A3B fits on 17.5GB VRAM.
• Introducing Dynamic 2.0 quants that set new benchmarks on 5-shot MMLU & KL Divergence.
• Llama 4 by Meta, including Scout & Maverick are now supported.
• EVERYTHING is now supported – all models (BERT, diffusion, Cohere, Mamba), FFT, etc. MultiGPU coming soon. Enable FFT with full_finetuning = True, 8-bit with load_in_8bit = True.
• Introducing Long-context Reasoning (GRPO) in Unsloth. Train your own reasoning model with just 5GB VRAM. Transform Llama, Phi, Mistral etc. into reasoning LLMs!
• DeepSeek-R1 – run or fine-tune them with our guide. All model uploads: here.

? Links and Resources

Key Features

• Supports full-finetuning, pretraining, 4b-bit, 16-bit and 8-bit training
• Supports all transformer-style models including TTS, STT, multimodal, diffusion, BERT and more!
• All kernels written in OpenAI\’s Triton language. Manual backprop engine.
• 0% loss in accuracy – no approximation methods – all exact.
• No change of hardware. Supports NVIDIA GPUs since 2018+. Minimum CUDA Capability 7.0 (V100, T4, Titan V, RTX 20, 30, 40x, A100, H100, L40 etc) Check your GPU! GTX 1070, 1080 works, but is slow.
• Works on Linux and Windows
• If you trained a model with ?Unsloth, you can use this cool sticker!  

? Install Unsloth

You can also see our documentation for more detailed installation and updating instructions here.

Pip Installation

Install with pip (recommended) for Linux devices:

pip install unsloth

To update Unsloth:

pip install --upgrade --force-reinstall --no-cache-dir unsloth unsloth_zoo

See here for advanced pip install instructions.

Windows Installation

Warning

Python 3.13 does not support Unsloth. Use 3.12, 3.11 or 3.10

1. Install NVIDIA Video Driver:
   You should install the latest version of your GPUs driver. Download drivers here: NVIDIA GPU Drive.

2. Install Visual Studio C++:
   You will need Visual Studio, with C++ installed. By default, C++ is not installed with Visual Studio, so make sure you select all of the C++ options. Also select options for Windows 10/11 SDK. For detailed instructions with options, see here.

3. Install CUDA Toolkit:
   Follow the instructions to install CUDA Toolkit.

4. Install PyTorch:
   You will need the correct version of PyTorch that is compatible with your CUDA drivers, so make sure to select them carefully.
   Install PyTorch.

5. Install Unsloth:

pip install unsloth

Notes

To run Unsloth directly on Windows:

• Install Triton from this Windows fork and follow the instructions here (be aware that the Windows fork requires PyTorch >= 2.4 and CUDA 12)
• In the SFTConfig, set dataset_num_proc=1 to avoid a crashing issue:

SFTConfig(
    dataset_num_proc=1,
    ...
)

Advanced/Troubleshooting

For advanced installation instructions or if you see weird errors during installations:

1. Install torch and triton. Go to https://pyt*orc*h.or*g to install it. For example pip install torch torchvision torchaudio triton
2. Confirm if CUDA is installed correctly. Try nvcc. If that fails, you need to install cudatoolkit or CUDA drivers.
3. Install xformers manually. You can try installing vllm and seeing if vllm succeeds. Check if xformers succeeded with python -m xformers.info Go to https://github.com***/facebookresearch/xformers. Another option is to install flash-attn for Ampere GPUs.
4. Double check that your versions of Python, CUDA, CUDNN, torch, triton, and xformers are compatible with one another. The PyTorch Compatibility Matrix may be useful.
5. Finally, install bitsandbytes and check it with python -m bitsandbytes

Conda Installation (Optional)

️Only use Conda if you have it. If not, use Pip. Select either pytorch-cuda=11.8,12.1 for CUDA 11.8 or CUDA 12.1. We support python=3.10,3.11,3.12.

conda create --name unsloth_env \\
    python=3.11 \\
    pytorch-cuda=12.1 \\
    pytorch cudatoolkit xformers -c pytorch -c nvidia -c xformers \\
    -y
conda activate unsloth_env

pip install unsloth

mkdir -p ~/miniconda3
wget https://repo.*ana*c*onda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -O ~/miniconda3/miniconda.sh
bash ~/miniconda3/miniconda.sh -b -u -p ~/miniconda3
rm -rf ~/miniconda3/miniconda.sh
~/miniconda3/bin/conda init bash
~/miniconda3/bin/conda init zsh

Advanced Pip Installation

️Do **NOT** use this if you have Conda. Pip is a bit more complex since there are dependency issues. The pip command is different for torch 2.2,2.3,2.4,2.5 and CUDA versions.

For other torch versions, we support torch211, torch212, torch220, torch230, torch240 and for CUDA versions, we support cu118 and cu121 and cu124. For Ampere devices (A100, H100, RTX3090) and above, use cu118-ampere or cu121-ampere or cu124-ampere.

For example, if you have torch 2.4 and CUDA 12.1, use:

pip install --upgrade pip
pip install \"unsloth[cu121-torch240] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"

Another example, if you have torch 2.5 and CUDA 12.4, use:

pip install --upgrade pip
pip install \"unsloth[cu124-torch250] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"

And other examples:

pip install \"unsloth[cu121-ampere-torch240] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"
pip install \"unsloth[cu118-ampere-torch240] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"
pip install \"unsloth[cu121-torch240] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"
pip install \"unsloth[cu118-torch240] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"

pip install \"unsloth[cu121-torch230] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"
pip install \"unsloth[cu121-ampere-torch230] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"

pip install \"unsloth[cu121-torch250] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"
pip install \"unsloth[cu124-ampere-torch250] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"

Or, run the below in a terminal to get the optimal pip installation command:

wget -qO- https://raw.githubuse*rconte**nt.com/unslothai/unsloth/main/unsloth/_auto_install.py | python -

Or, run the below manually in a Python REPL:

try: import torch
except: raise ImportError(\'Install torch via `pip install torch`\')
from packaging.version import Version as V
v = V(torch.__version__)
cuda = str(torch.version.cuda)
is_ampere = torch.cuda.get_device_capability()[0] >= 8
if cuda != \"12.1\" and cuda != \"11.8\" and cuda != \"12.4\": raise RuntimeError(f\"CUDA = {cuda} not supported!\")
if   v <= V(\'2.1.0\'): raise RuntimeError(f\"Torch = {v} too old!\")
elif v <= V(\'2.1.1\'): x = \'cu{}{}-torch211\'
elif v <= V(\'2.1.2\'): x = \'cu{}{}-torch212\'
elif v  < V(\'2.3.0\'): x = \'cu{}{}-torch220\'
elif v  < V(\'2.4.0\'): x = \'cu{}{}-torch230\'
elif v  < V(\'2.5.0\'): x = \'cu{}{}-torch240\'
elif v  < V(\'2.6.0\'): x = \'cu{}{}-torch250\'
else: raise RuntimeError(f\"Torch = {v} too new!\")
x = x.format(cuda.replace(\".\", \"\"), \"-ampere\" if is_ampere else \"\")
print(f\'pip install --upgrade pip && pip install \"unsloth[{x}] @ git+https://g*i*thub.com*/unslothai/unsloth.git\"\')

Documentation

• Go to our official Documentation for saving to GGUF, checkpointing, evaluation and more!
• We support Huggingface\’s TRL, Trainer, Seq2SeqTrainer or even Pytorch code!
• We\’re in ?Hugging Face\’s official docs! Check out the SFT docs and DPO docs!
• If you want to download models from the ModelScope community, please use an environment variable: UNSLOTH_USE_MODELSCOPE=1, and install the modelscope library by: pip install modelscope -U.

unsloth_cli.py also supports UNSLOTH_USE_MODELSCOPE=1 to download models and datasets. please remember to use the model and dataset id in the ModelScope community.

from unsloth import FastLanguageModel, FastModel
import torch
from trl import SFTTrainer, SFTConfig
from datasets import load_dataset
max_seq_length = 2048 # Supports RoPE Scaling internally, so choose any!
# Get LAION dataset
url = \"https://hu**g*gingface.co/datasets/laion/OIG/resolve/main/unified_chip2.jsonl\"
dataset = load_dataset(\"json\", data_files = {\"train\" : url}, split = \"train\")

# 4bit pre quantized models we support for 4x faster downloading + no OOMs.
fourbit_models = [
    \"unsloth/Meta-Llama-3.1-8B-bnb-4bit\",      # Llama-3.1 2x faster
    \"unsloth/Meta-Llama-3.1-8B-Instruct-bnb-4bit\",
    \"unsloth/Meta-Llama-3.1-70B-bnb-4bit\",
    \"unsloth/Meta-Llama-3.1-405B-bnb-4bit\",    # 4bit for 405b!
    \"unsloth/Mistral-Small-Instruct-2409\",     # Mistral 22b 2x faster!
    \"unsloth/mistral-7b-instruct-v0.3-bnb-4bit\",
    \"unsloth/Phi-3.5-mini-instruct\",           # Phi-3.5 2x faster!
    \"unsloth/Phi-3-medium-4k-instruct\",
    \"unsloth/gemma-2-9b-bnb-4bit\",
    \"unsloth/gemma-2-27b-bnb-4bit\",            # Gemma 2x faster!

    \"unsloth/Llama-3.2-1B-bnb-4bit\",           # NEW! Llama 3.2 models
    \"unsloth/Llama-3.2-1B-Instruct-bnb-4bit\",
    \"unsloth/Llama-3.2-3B-bnb-4bit\",
    \"unsloth/Llama-3.2-3B-Instruct-bnb-4bit\",

    \"unsloth/Llama-3.3-70B-Instruct-bnb-4bit\" # NEW! Llama 3.3 70B!
] # More models at https://hugg*in**gface.co/unsloth

model, tokenizer = FastModel.from_pretrained(
    model_name = \"unsloth/gemma-3-4B-it\",
    max_seq_length = 2048, # Choose any for long context!
    load_in_4bit = True,  # 4 bit quantization to reduce memory
    load_in_8bit = False, # [NEW!] A bit more accurate, uses 2x memory
    full_finetuning = False, # [NEW!] We have full finetuning now!
    # token = \"hf_...\", # use one if using gated models
)

# Do model patching and add fast LoRA weights
model = FastLanguageModel.get_peft_model(
    model,
    r = 16,
    target_modules = [\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\",
                      \"gate_proj\", \"up_proj\", \"down_proj\",],
    lora_alpha = 16,
    lora_dropout = 0, # Supports any, but = 0 is optimized
    bias = \"none\",    # Supports any, but = \"none\" is optimized
    # [NEW] \"unsloth\" uses 30% less VRAM, fits 2x larger batch sizes!
    use_gradient_checkpointing = \"unsloth\", # True or \"unsloth\" for very long context
    random_state = 3407,
    max_seq_length = max_seq_length,
    use_rslora = False,  # We support rank stabilized LoRA
    loftq_config = None, # And LoftQ
)

trainer = SFTTrainer(
    model = model,
    train_dataset = dataset,
    tokenizer = tokenizer,
    args = SFTConfig(
        max_seq_length = max_seq_length,
        per_device_train_batch_size = 2,
        gradient_accumulation_steps = 4,
        warmup_steps = 10,
        max_steps = 60,
        logging_steps = 1,
        output_dir = \"outputs\",
        optim = \"adamw_8bit\",
        seed = 3407,
    ),
)
trainer.train()

# Go to https://git**hu*b.com/unslothai/unsloth/wiki for advanced tips like
# (1) Saving to GGUF / merging to 16bit for vLLM
# (2) Continued training from a saved LoRA adapter
# (3) Adding an evaluation loop / OOMs
# (4) Customized chat templates

Reinforcement Learning

RL including DPO, GRPO, PPO, Reward Modelling, Online DPO all work with Unsloth. We\’re in ?Hugging Face\’s official docs! We\’re on the GRPO docs and the DPO docs! List of RL notebooks:

• Advanced Qwen3 GRPO notebook: Link
• ORPO notebook: Link
• DPO Zephyr notebook: Link
• KTO notebook: Link
• SimPO notebook: Link

import os
os.environ[\"CUDA_VISIBLE_DEVICES\"] = \"0\" # Optional set GPU device ID

from unsloth import FastLanguageModel
import torch
from trl import DPOTrainer, DPOConfig
max_seq_length = 2048

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name = \"unsloth/zephyr-sft-bnb-4bit\",
    max_seq_length = max_seq_length,
    load_in_4bit = True,
)

# Do model patching and add fast LoRA weights
model = FastLanguageModel.get_peft_model(
    model,
    r = 64,
    target_modules = [\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\",
                      \"gate_proj\", \"up_proj\", \"down_proj\",],
    lora_alpha = 64,
    lora_dropout = 0, # Supports any, but = 0 is optimized
    bias = \"none\",    # Supports any, but = \"none\" is optimized
    # [NEW] \"unsloth\" uses 30% less VRAM, fits 2x larger batch sizes!
    use_gradient_checkpointing = \"unsloth\", # True or \"unsloth\" for very long context
    random_state = 3407,
    max_seq_length = max_seq_length,
)

dpo_trainer = DPOTrainer(
    model = model,
    ref_model = None,
    train_dataset = YOUR_DATASET_HERE,
    # eval_dataset = YOUR_DATASET_HERE,
    tokenizer = tokenizer,
    args = DPOConfig(
        per_device_train_batch_size = 4,
        gradient_accumulation_steps = 8,
        warmup_ratio = 0.1,
        num_train_epochs = 3,
        logging_steps = 1,
        optim = \"adamw_8bit\",
        seed = 42,
        output_dir = \"outputs\",
        max_length = 1024,
        max_prompt_length = 512,
        beta = 0.1,
    ),
)
dpo_trainer.train()

? Performance Benchmarking

• For our most detailed benchmarks, read our Llama 3.3 Blog.
• Benchmarking of Unsloth was also conducted by ?Hugging Face.

We tested using the Alpaca Dataset, a batch size of 2, gradient accumulation steps of 4, rank = 32, and applied QLoRA on all linear layers (q, k, v, o, gate, up, down):

Context length benchmarks

Llama 3.1 (8B) max. context length

We tested Llama 3.1 (8B) Instruct and did 4bit QLoRA on all linear layers (Q, K, V, O, gate, up and down) with rank = 32 with a batch size of 1. We padded all sequences to a certain maximum sequence length to mimic long context finetuning workloads.

Llama 3.3 (70B) max. context length

We tested Llama 3.3 (70B) Instruct on a 80GB A100 and did 4bit QLoRA on all linear layers (Q, K, V, O, gate, up and down) with rank = 32 with a batch size of 1. We padded all sequences to a certain maximum sequence length to mimic long context finetuning workloads.

Citation

You can cite the Unsloth repo as follows:

@software{unsloth,
  author = {Daniel Han, Michael Han and Unsloth team},
  title = {Unsloth},
  url = {http://gi*th*ub.*com/unslothai/unsloth},
  year = {2023}
}

Thank You to

• The llama.cpp library that lets users save models with Unsloth
• The Hugging Face team and their TRL library
• Erik for his help adding Apple\’s ML Cross Entropy in Unsloth
• Etherl for adding support for TTS, diffusion and BERT models
• And of course for every single person who has contributed or has used Unsloth!