t_wの輪郭

あれ

2025/4/13 10:01:00

from unsloth import FastLanguageModel
import torch

max_seq_length = 2048 # Choose any! We auto support RoPE Scaling internally!
dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
load_in_4bit = True # Use 4bit quantization to reduce memory usage. Can be False.

# 4bit pre quantized models we support for 4x faster downloading + no OOMs.
fourbit_models = [
    "unsloth/Qwen2.5-Coder-1.5B-bnb-4bit",
    "unsloth/mistral-7b-v0.3-bnb-4bit",      # New Mistral v3 2x faster!
    "unsloth/llama-3-8b-bnb-4bit",           # Llama-3 15 trillion tokens model 2x faster!
    "unsloth/llama-3-8b-Instruct-bnb-4bit",
    "unsloth/llama-3-70b-bnb-4bit",
    "unsloth/Phi-3-mini-4k-instruct",        # Phi-3 2x faster!
    "unsloth/Phi-3-medium-4k-instruct",
    "unsloth/Qwen2-0.5b-bnb-4bit",           # Qwen2 2x faster!
    "unsloth/Qwen2-1.5b-bnb-4bit",
    "unsloth/Qwen2-7b-bnb-4bit",
    "unsloth/Qwen2-72b-bnb-4bit",
    "unsloth/gemma-7b-bnb-4bit",             # Gemma 2.2x faster!
] # Try more models at https://huggingface.co/unsloth!


if __name__ == '__main__':
    model, tokenizer = FastLanguageModel.from_pretrained(
        model_name = "unsloth/Qwen2.5-Coder-0.5B-bnb-4bit", # Reminder we support ANY Hugging Face model!
        max_seq_length = max_seq_length,
        dtype = dtype,
        load_in_4bit = load_in_4bit,
        # token = "hf_...", # use one if using gated models like meta-llama/Llama-2-7b-hf
    )


    model = FastLanguageModel.get_peft_model(
        model,
        r = 16, # 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 = 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,
        use_rslora = False,  # We support rank stabilized LoRA
        loftq_config = None, # And LoftQ
    )


    alpaca_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:
    {}"""

    EOS_TOKEN = tokenizer.eos_token # Must add EOS_TOKEN
    def formatting_prompts_func(examples):
        instructions = examples["instruction"]
        inputs       = examples["input"]
        outputs      = examples["output"]
        texts = []
        for instruction, input, output in zip(instructions, inputs, outputs):
            # Must add EOS_TOKEN, otherwise your generation will go on forever!
            text = alpaca_prompt.format(instruction, input, output) + EOS_TOKEN
            texts.append(text)
        return { "text" : texts, }
    pass

    from datasets import load_dataset
    dataset = load_dataset("yahma/alpaca-cleaned", split = "train")
    dataset = dataset.map(formatting_prompts_func, batched = True,)


    from trl import SFTTrainer
    from transformers import TrainingArguments
    from unsloth import is_bfloat16_supported

    trainer = SFTTrainer(
        model = model,
        tokenizer = tokenizer,
        train_dataset = dataset,
        dataset_text_field = "text",
        max_seq_length = max_seq_length,
        dataset_num_proc = 2,
        args = TrainingArguments(
            per_device_train_batch_size = 2,
            gradient_accumulation_steps = 4,

            # Use num_train_epochs = 1, warmup_ratio for full training runs!
            # warmup_steps = 5,
            # max_steps = 2500,
            max_steps = 100,
            # num_train_epochs = 1,

            learning_rate = 2e-4,
            fp16 = not is_bfloat16_supported(),
            bf16 = is_bfloat16_supported(),
            logging_steps = 1,
            optim = "adamw_8bit",
            weight_decay = 0.01,
            lr_scheduler_type = "linear",
            seed = 3407,
            output_dir = "outputs",
            report_to = "none", # Use this for WandB etc
        ),
    )


    trainer_stats = trainer.train()


    # alpaca_prompt = Copied from above
    FastLanguageModel.for_inference(model) # Unsloth has 2x faster inference!
    inputs = tokenizer(
    [
        alpaca_prompt.format(
            "Continue the fibonnaci sequence.", # instruction
            "1, 1, 2, 3, 5, 8", # input
            "", # output - leave this blank for generation!
        )
    ], return_tensors = "pt").to("cuda")

    outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True)
    tokenizer.batch_decode(outputs)


    model.save_pretrained("lora_model") # Local saving
    tokenizer.save_pretrained("lora_model")
    # model.push_to_hub("your_name/lora_model", token = "...") # Online saving
    # tokenizer.push_to_hub("your_name/lora_model", token = "...") # Online saving

from unsloth import FastLanguageModel import torch max_seq_length = 2048 # Choose any! We auto support RoPE Scaling internally! dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+ load_in_4bit = True # Use 4bit quantization to reduce memory usage. Can be False. # 4bit pre quantized models we support for 4x faster downloading + no OOMs. fourbit_models = [ "unsloth/Qwen2.5-Coder-1.5B-bnb-4bit", "unsloth/mistral-7b-v0.3-bnb-4bit", # New Mistral v3 2x faster! "unsloth/llama-3-8b-bnb-4bit", # Llama-3 15 trillion tokens model 2x faster! "unsloth/llama-3-8b-Instruct-bnb-4bit", "unsloth/llama-3-70b-bnb-4bit", "unsloth/Phi-3-mini-4k-instruct", # Phi-3 2x faster! "unsloth/Phi-3-medium-4k-instruct", "unsloth/Qwen2-0.5b-bnb-4bit", # Qwen2 2x faster! "unsloth/Qwen2-1.5b-bnb-4bit", "unsloth/Qwen2-7b-bnb-4bit", "unsloth/Qwen2-72b-bnb-4bit", "unsloth/gemma-7b-bnb-4bit", # Gemma 2.2x faster! ] # Try more models at https://huggingface.co/unsloth! if __name__ == '__main__': model, tokenizer = FastLanguageModel.from_pretrained( model_name = "unsloth/Qwen2.5-Coder-0.5B-bnb-4bit", # Reminder we support ANY Hugging Face model! max_seq_length = max_seq_length, dtype = dtype, load_in_4bit = load_in_4bit, # token = "hf_...", # use one if using gated models like meta-llama/Llama-2-7b-hf ) model = FastLanguageModel.get_peft_model( model, r = 16, # 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 = 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, use_rslora = False, # We support rank stabilized LoRA loftq_config = None, # And LoftQ ) alpaca_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: {}""" EOS_TOKEN = tokenizer.eos_token # Must add EOS_TOKEN def formatting_prompts_func(examples): instructions = examples["instruction"] inputs = examples["input"] outputs = examples["output"] texts = [] for instruction, input, output in zip(instructions, inputs, outputs): # Must add EOS_TOKEN, otherwise your generation will go on forever! text = alpaca_prompt.format(instruction, input, output) + EOS_TOKEN texts.append(text) return { "text" : texts, } pass from datasets import load_dataset dataset = load_dataset("yahma/alpaca-cleaned", split = "train") dataset = dataset.map(formatting_prompts_func, batched = True,) from trl import SFTTrainer from transformers import TrainingArguments from unsloth import is_bfloat16_supported trainer = SFTTrainer( model = model, tokenizer = tokenizer, train_dataset = dataset, dataset_text_field = "text", max_seq_length = max_seq_length, dataset_num_proc = 2, args = TrainingArguments( per_device_train_batch_size = 2, gradient_accumulation_steps = 4, # Use num_train_epochs = 1, warmup_ratio for full training runs! # warmup_steps = 5, # max_steps = 2500, max_steps = 100, # num_train_epochs = 1, learning_rate = 2e-4, fp16 = not is_bfloat16_supported(), bf16 = is_bfloat16_supported(), logging_steps = 1, optim = "adamw_8bit", weight_decay = 0.01, lr_scheduler_type = "linear", seed = 3407, output_dir = "outputs", report_to = "none", # Use this for WandB etc ), ) trainer_stats = trainer.train() # alpaca_prompt = Copied from above FastLanguageModel.for_inference(model) # Unsloth has 2x faster inference! inputs = tokenizer( [ alpaca_prompt.format( "Continue the fibonnaci sequence.", # instruction "1, 1, 2, 3, 5, 8", # input "", # output - leave this blank for generation! ) ], return_tensors = "pt").to("cuda") outputs = model.generate(**inputs, max_new_tokens = 64, use_cache = True) tokenizer.batch_decode(outputs) model.save_pretrained("lora_model") # Local saving tokenizer.save_pretrained("lora_model") # model.push_to_hub("your_name/lora_model", token = "...") # Online saving # tokenizer.push_to_hub("your_name/lora_model", token = "...") # Online saving

t_wの輪郭

やり2025年4月13日

やった

明日

あれ

あれ