# /// script # requires-python = ">=3.10,<3.14" # dependencies = [ # "accelerate", # "bitsandbytes>=0.43.0", # "datasets==4.3.0", # "hf_transfer", # "huggingface_hub>=0.34.0", # "marimo", # "peft", # "protobuf", # "sentencepiece", # "torchao>=0.16.0", # "transformers==4.56.2", # "triton>=3.2.0", # "trl==0.22.2", # "unsloth @ git+https://github.com/unslothai/unsloth", # "unsloth_zoo @ git+https://github.com/unslothai/unsloth-zoo", # ] # # [tool.uv] # no-build-package = [ # "bitsandbytes", # "triton", # "vllm", # "xformers", # ] # /// import marimo __generated_with = "0.23.8" app = marimo.App() @app.cell def _(): import marimo as mo return (mo,) @app.cell(hide_code=True) def _(mo): mo.md(r""" To run this notebook, hit the **▶ Run all** button in the bottom-right corner - or use `Ctrl/Cmd + Shift + R`.

Join Discord if you need help + ⭐ Star us on Github ⭐

To install Unsloth on your local device, follow [our guide](https://unsloth.ai/docs/get-started/install). This notebook is licensed [LGPL-3.0](https://github.com/unslothai/notebooks?tab=LGPL-3.0-1-ov-file#readme). You will learn how to do [data prep](#Data), how to [train](#Train), how to [run the model](#Inference), & how to save it """) return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### News """) return @app.cell(hide_code=True) def _(mo): mo.md(r""" Introducing **Unsloth Studio** - a new open source, no-code web UI to train and run LLMs. [Blog](https://unsloth.ai/docs/new/studio) • [Notebook](https://github.com/unslothai/unsloth/blob/main/studio/Unsloth_Studio_Colab.ipynb)

_{Train models — no code needed}

_{Run GGUF models on Mac, Windows & Linux}

Train MoEs - DeepSeek, GLM, Qwen and gpt-oss 12x faster with 35% less VRAM. [Blog](https://unsloth.ai/docs/new/faster-moe) Ultra Long-Context Reinforcement Learning is here with 7x more context windows! [Blog](https://unsloth.ai/docs/new/grpo-long-context) New in Reinforcement Learning: [FP8 RL](https://unsloth.ai/docs/new/fp8-reinforcement-learning) • [Vision RL](https://unsloth.ai/docs/new/vision-reinforcement-learning-vlm-rl) • [Standby](https://unsloth.ai/docs/basics/memory-efficient-rl) • [gpt-oss RL](https://unsloth.ai/docs/new/gpt-oss-reinforcement-learning) Visit our docs for all our [model uploads](https://unsloth.ai/docs/get-started/unsloth-model-catalog) and [notebooks](https://unsloth.ai/docs/get-started/unsloth-notebooks). """) return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### Unsloth """) return @app.cell def _(): from unsloth import FastLanguageModel import torch fourbit_models = [ "unsloth/Qwen3-4B-Instruct-2507-unsloth-bnb-4bit", # Qwen 14B 2x faster "unsloth/Qwen3-4B-Thinking-2507-unsloth-bnb-4bit", "unsloth/Qwen3-8B-unsloth-bnb-4bit", "unsloth/Qwen3-14B-unsloth-bnb-4bit", "unsloth/Qwen3-32B-unsloth-bnb-4bit", # 4bit dynamic quants for superior accuracy and low memory use "unsloth/gemma-3-12b-it-unsloth-bnb-4bit", "unsloth/Phi-4", "unsloth/Llama-3.1-8B", "unsloth/Llama-3.2-3B", "unsloth/orpheus-3b-0.1-ft-unsloth-bnb-4bit", # [NEW] We support TTS models! ] # More models at https://huggingface.co/unsloth model, tokenizer = FastLanguageModel.from_pretrained( model_name="unsloth/Qwen3-4B-Thinking-2507", # YOUR MODEL YOU USED FOR TRAINING 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 = "YOUR_HF_TOKEN", # HF Token for gated models ) return FastLanguageModel, model, tokenizer, torch @app.cell(hide_code=True) def _(mo): mo.md(r""" We now add LoRA adapters so we only need to update a small amount of parameters! """) return @app.cell def _(FastLanguageModel, model): model_1 = FastLanguageModel.get_peft_model( model, r=32, # Choose any number > 0 ! Suggested 8, 16, 32, 64, 128 target_modules=[ "q_proj", "k_proj", "v_proj", "o_proj", "gate_proj", "up_proj", "down_proj", ], lora_alpha=32, lora_dropout=0, # Supports any, but = 0 is optimized bias="none", # Supports any, but = "none" is optimized use_gradient_checkpointing="unsloth", # True or "unsloth" for very long context random_state=3407, use_rslora=False, # We support rank stabilized LoRA loftq_config=None, # And LoftQ ) return (model_1,) @app.cell(hide_code=True) def _(mo): mo.md(r""" ### Data Prep We now use the `Qwen-3` format for conversation style finetunes. We use the [Open Math Reasoning](https://huggingface.co/datasets/unsloth/OpenMathReasoning-mini) dataset which was used to win the [AIMO](https://www.kaggle.com/competitions/ai-mathematical-olympiad-progress-prize-2/leaderboard) (AI Mathematical Olympiad - Progress Prize 2) challenge! We sample 10% of verifiable reasoning traces that used DeepSeek R1, and which got > 95% accuracy. Qwen-3 renders multi turn conversations like below: ``` <|im_start|>user Hello!<|im_end|> <|im_start|>assistant Hey there!<|im_end|> ``` We use our `get_chat_template` function to get the correct chat template. We support `zephyr, chatml, mistral, llama, alpaca, vicuna, vicuna_old, phi3, llama3, phi4, qwen2.5, gemma3` and more. """) return @app.cell def _(tokenizer): from unsloth.chat_templates import get_chat_template tokenizer_1 = get_chat_template(tokenizer, chat_template="qwen3-thinking") return (tokenizer_1,) @app.cell def _(): from datasets import load_dataset dataset = load_dataset("unsloth/OpenMathReasoning-mini", split="cot") return (dataset,) @app.cell(hide_code=True) def _(mo): mo.md(r""" We now convert the reasoning dataset into conversational format: """) return @app.cell def _(dataset): def generate_conversation(examples): problems = examples["problem"] solutions = examples["generated_solution"] conversations = [] for problem, solution in zip(problems, solutions): conversations.append( [ {"role": "user", "content": problem}, {"role": "assistant", "content": solution}, ] ) return {"conversations": conversations} dataset_1 = dataset.map(generate_conversation, batched=True) return (dataset_1,) @app.cell(hide_code=True) def _(mo): mo.md(r""" We now have to apply the chat template for `Qwen-3` onto the conversations, and save it to `text`. """) return @app.cell def _(dataset_1, tokenizer_1): def formatting_prompts_func(examples): convos = examples["conversations"] texts = [ tokenizer_1.apply_chat_template( convo, tokenize=False, add_generation_prompt=False ) for convo in convos ] return {"text": texts} dataset_2 = dataset_1.map(formatting_prompts_func, batched=True) return (dataset_2,) @app.cell(hide_code=True) def _(mo): mo.md(r""" Let's see how the chat template did! """) return @app.cell def _(dataset_2): dataset_2[100]["text"] return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### Train the model Now let's train our model. We do 60 steps to speed things up, but you can set `num_train_epochs=1` for a full run, and turn off `max_steps=None`. """) return @app.cell def _(dataset_2, model_1, tokenizer_1): from trl import SFTTrainer, SFTConfig trainer = SFTTrainer( model=model_1, tokenizer=tokenizer_1, train_dataset=dataset_2, eval_dataset=None, # Can set up evaluation! args=SFTConfig( dataset_text_field="text", per_device_train_batch_size=2, gradient_accumulation_steps=4, # Use GA to mimic batch size! warmup_steps=5, max_steps=60, learning_rate=0.0002, # Reduce to 2e-5 for long training runs logging_steps=1, optim="adamw_8bit", weight_decay=0.001, lr_scheduler_type="linear", seed=3407, report_to="none", # Use TrackIO/WandB etc ), ) return (trainer,) @app.cell(hide_code=True) def _(mo): mo.md(r""" We also use Unsloth's `train_on_completions` method to only train on the assistant outputs and ignore the loss on the user's inputs. This helps increase accuracy of finetunes! """) return @app.cell def _(trainer): from unsloth.chat_templates import train_on_responses_only trainer_1 = train_on_responses_only( trainer, instruction_part="<|im_start|>user\n", response_part="<|im_start|>assistant\n", ) return (trainer_1,) @app.cell(hide_code=True) def _(mo): mo.md(r""" Let's verify masking the instruction part is done! Let's print the 100th row again. """) return @app.cell def _(tokenizer_1, trainer_1): tokenizer_1.decode(trainer_1.train_dataset[100]["input_ids"]) return @app.cell(hide_code=True) def _(mo): mo.md(r""" Now let's print the masked out example - you should see only the answer is present: """) return @app.cell def _(tokenizer_1, trainer_1): tokenizer_1.decode( [ tokenizer_1.pad_token_id if x == -100 else x for x in trainer_1.train_dataset[100]["labels"] ] ).replace(tokenizer_1.pad_token, " ") return @app.cell def _(torch): # @title Show current memory stats gpu_stats = torch.cuda.get_device_properties(0) start_gpu_memory = round(torch.cuda.max_memory_reserved() / 1024 / 1024 / 1024, 3) max_memory = round(gpu_stats.total_memory / 1024 / 1024 / 1024, 3) print(f"GPU = {gpu_stats.name}. Max memory = {max_memory} GB.") print(f"{start_gpu_memory} GB of memory reserved.") return max_memory, start_gpu_memory @app.cell(hide_code=True) def _(mo): mo.md(r""" Let's train the model! To resume a training run, set `trainer.train(resume_from_checkpoint = True)` """) return @app.cell def _(trainer_1): trainer_stats = trainer_1.train() return (trainer_stats,) @app.cell def _(max_memory, start_gpu_memory, torch, trainer_stats): # @title Show final memory and time stats used_memory = round(torch.cuda.max_memory_reserved() / 1024 / 1024 / 1024, 3) used_memory_for_lora = round(used_memory - start_gpu_memory, 3) used_percentage = round(used_memory / max_memory * 100, 3) lora_percentage = round(used_memory_for_lora / max_memory * 100, 3) print(f"{trainer_stats.metrics['train_runtime']} seconds used for training.") print( f"{round(trainer_stats.metrics['train_runtime'] / 60, 2)} minutes used for training." ) print(f"Peak reserved memory = {used_memory} GB.") print(f"Peak reserved memory for training = {used_memory_for_lora} GB.") print(f"Peak reserved memory % of max memory = {used_percentage} %.") print(f"Peak reserved memory for training % of max memory = {lora_percentage} %.") return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### Inference Let's run the model via Unsloth native inference! According to the `Qwen-3` team, the recommended settings for instruct inference are `temperature = 0.7, top_p = 0.8, top_k = 20` For reasoning chat based inference, `temperature = 0.6, top_p = 0.95, top_k = 20` """) return @app.cell def _(model_1, tokenizer_1): messages = [{"role": "user", "content": "Solve (x + 2)^2 = 0."}] text = tokenizer_1.apply_chat_template( messages, tokenize=False, add_generation_prompt=True, enable_thinking=False ) from transformers import TextStreamer _ = model_1.generate( **tokenizer_1(text, return_tensors="pt").to("cuda"), max_new_tokens=450, # Increase for longer outputs! temperature=0.7, # For non thinking top_p=0.8, top_k=20, streamer=TextStreamer(tokenizer_1, skip_prompt=False), ) return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### Saving, loading finetuned models To save the final model as LoRA adapters, either use Hugging Face's `push_to_hub` for an online save or `save_pretrained` for a local save. **[NOTE]** This ONLY saves the LoRA adapters, and not the full model. To save to 16bit or GGUF, scroll down! """) return @app.cell def _(model_1, tokenizer_1): model_1.save_pretrained("qwen_lora") tokenizer_1.save_pretrained("qwen_lora") return @app.cell(hide_code=True) def _(mo): mo.md(r""" Now if you want to load the LoRA adapters we just saved for inference, set `False` to `True`: """) return @app.cell def _(): if False: from unsloth import FastLanguageModel as _FastLanguageModel _model, _tokenizer = _FastLanguageModel.from_pretrained( model_name="qwen_lora", max_seq_length=2048, load_in_4bit=True # YOUR MODEL YOU USED FOR TRAINING ) return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### Saving to float16 for VLLM We also support saving to `float16` directly. Select `merged_16bit` for float16 or `merged_4bit` for int4. We also allow `lora` adapters as a fallback. Use `push_to_hub_merged` to upload to your Hugging Face account! You can go to https://huggingface.co/settings/tokens for your personal tokens. See [our docs](https://unsloth.ai/docs/basics/inference-and-deployment) for more deployment options. """) return @app.cell def _(model_1, tokenizer_1): # Merge to 16bit if False: model_1.save_pretrained_merged( "qwen_finetune_16bit", tokenizer_1, save_method="merged_16bit" ) if False: # Pushing to HF Hub model_1.push_to_hub_merged( "HF_USERNAME/qwen_finetune_16bit", tokenizer_1, save_method="merged_16bit", token="YOUR_HF_TOKEN", # Get a token at https://huggingface.co/settings/tokens ) if False: # Merge to 4bit model_1.save_pretrained_merged( "qwen_finetune_4bit", tokenizer_1, save_method="merged_4bit" ) if False: model_1.push_to_hub_merged( "HF_USERNAME/qwen_finetune_4bit", tokenizer_1, save_method="merged_4bit", token="YOUR_HF_TOKEN", # Get a token at https://huggingface.co/settings/tokens ) # Pushing to HF Hub if False: model_1.save_pretrained("qwen_lora") # Just LoRA adapters tokenizer_1.save_pretrained("qwen_lora") if False: model_1.push_to_hub("HF_USERNAME/qwen_lora", token="YOUR_HF_TOKEN") tokenizer_1.push_to_hub( "HF_USERNAME/qwen_lora", token="YOUR_HF_TOKEN" ) # Pushing to HF Hub return @app.cell(hide_code=True) def _(mo): mo.md(r""" ### GGUF / llama.cpp Conversion To save to `GGUF` / `llama.cpp`, we support it natively now! We clone `llama.cpp` and we default save it to `q8_0`. We allow all methods like `q4_k_m`. Use `save_pretrained_gguf` for local saving and `push_to_hub_gguf` for uploading to HF. Some supported quant methods (full list on our [docs page](https://unsloth.ai/docs/basics/inference-and-deployment/saving-to-gguf)): * `q8_0` - Fast conversion. High resource use, but generally acceptable. * `q4_k_m` - Recommended. Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q4_K. * `q5_k_m` - Recommended. Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q5_K. [**NEW**] To finetune and auto export to Ollama, try our [Ollama notebook](https://github.com/unslothai/notebooks/blob/main/nb/Llama3_(8B)-Ollama.ipynb) """) return @app.cell(hide_code=True) def _(mo): mo.md(r""" Likewise, if you want to instead push to GGUF to your Hugging Face account, set `if False` to `if True` and add your Hugging Face token and upload location! """) return @app.cell def _(model_1, tokenizer_1): # Save to 8bit Q8_0 if False: model_1.save_pretrained_gguf("qwen_finetune", tokenizer_1) if False: model_1.push_to_hub_gguf( "HF_USERNAME/qwen_finetune", tokenizer_1, token="YOUR_HF_TOKEN" ) if False: # Save to 16bit GGUF model_1.save_pretrained_gguf( "qwen_finetune", tokenizer_1, quantization_method="f16" ) if False: model_1.push_to_hub_gguf( "HF_USERNAME/qwen_finetune", tokenizer_1, quantization_method="f16", token="YOUR_HF_TOKEN", # Get a token at https://huggingface.co/settings/tokens ) # Pushing to HF Hub if False: model_1.save_pretrained_gguf( "qwen_finetune", tokenizer_1, quantization_method="q4_k_m" ) # Save to q4_k_m GGUF if False: model_1.push_to_hub_gguf( "HF_USERNAME/qwen_finetune", tokenizer_1, quantization_method="q4_k_m", token="YOUR_HF_TOKEN", # Get a token at https://huggingface.co/settings/tokens ) if False: # Pushing to HF Hub # Save to multiple GGUF options - much faster if you want multiple! model_1.push_to_hub_gguf( "HF_USERNAME/qwen_finetune", tokenizer_1, quantization_method=["q4_k_m", "q8_0", "q5_k_m"], token="YOUR_HF_TOKEN", # Get a token at https://huggingface.co/settings/tokens ) return @app.cell(hide_code=True) def _(mo): mo.md(r""" Now, use the `qwen_finetune.Q8_0.gguf` file or `qwen_finetune.Q4_K_M.gguf` file in llama.cpp. And we're done! If you have any questions on Unsloth, we have a [Discord](https://discord.gg/unsloth) channel! If you find any bugs or want to keep updated with the latest LLM stuff, or need help, join projects etc, feel free to join our Discord! Some other resources: 1. Train your own reasoning model - Llama GRPO notebook [Open in molab](https://github.com/unslothai/notebooks/blob/main/nb/Llama3.1_(8B)-GRPO.ipynb) 2. Saving finetunes to Ollama. [Free notebook](https://github.com/unslothai/notebooks/blob/main/nb/Llama3_(8B)-Ollama.ipynb) 3. Llama 3.2 Vision finetuning - Radiography use case. [Open in molab](https://github.com/unslothai/notebooks/blob/main/nb/Llama3.2_(11B)-Vision.ipynb) 4. See notebooks for DPO, ORPO, Continued pretraining, conversational finetuning and more on our [documentation](https://unsloth.ai/docs/get-started/unsloth-notebooks)!

Join Discord if you need help + ⭐️ Star us on Github ⭐️

This notebook and all Unsloth notebooks are licensed [LGPL-3.0](https://github.com/unslothai/notebooks?tab=LGPL-3.0-1-ov-file#readme). """) return if __name__ == "__main__": app.run()