Upload llama_3_1_8b_+_unsloth_2x_faster_finetuning.py

llama_3_1_8b_+_unsloth_2x_faster_finetuning.py

@@ -0,0 +1,209 @@
+# -*- coding: utf-8 -*-
+"""Mine Llama-3.1 8b + Unsloth 2x faster finetuning.ipynb
+
+Automatically generated by Colab.
+
+Original file is located at
+    https://colab.research.google.com/drive/1eI70VMZWms-GCr4j3vf-tcprenY6JCnP
+"""
+
+# Commented out IPython magic to ensure Python compatibility.
+# %%capture
+# # Installs Unsloth, Xformers (Flash Attention) and all other packages!
+# !pip install "unsloth[colab-new] @ git+https://github.com/unslothai/unsloth.git"
+# 
+# # We have to check which Torch version for Xformers (2.3 -> 0.0.27)
+# from torch import __version__; from packaging.version import Version as V
+# xformers = "xformers==0.0.27" if V(__version__) < V("2.4.0") else "xformers"
+# !pip install --no-deps {xformers} trl peft accelerate bitsandbytes triton
+
+from unsloth import FastLanguageModel
+import torch
+max_seq_length = 512 # 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/Meta-Llama-3.1-8B-bnb-4bit",      # Llama-3.1 15 trillion tokens model 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",    # We also uploaded 4bit for 405b!
+    "unsloth/Mistral-Nemo-Base-2407-bnb-4bit", # New Mistral 12b 2x faster!
+    "unsloth/Mistral-Nemo-Instruct-2407-bnb-4bit",
+    "unsloth/mistral-7b-v0.3-bnb-4bit",        # Mistral v3 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!
+] # More models at https://huggingface.co/unsloth
+
+model, tokenizer = FastLanguageModel.from_pretrained(
+    model_name = "meta-llama/Meta-Llama-3.1-8B",
+    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 = 128, # 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 = 10,
+    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 a conversation between a healthcare provider and a patient. The healthcare provider should respond appropriately to the patient's query.
+
+### Question:
+{}
+
+### Response:
+{}"""
+
+EOS_TOKEN = tokenizer.eos_token  # Must add EOS_TOKEN
+
+def formatting_prompts_func(examples):
+    # contexts = examples["Context"]
+    questions = examples["Question"]
+    answers = examples["Answer"]
+    texts = []
+    for question, answer in zip(questions, answers):
+        # Must add EOS_TOKEN, otherwise your generation will go on forever!
+        text = alpaca_prompt.format(question, answer) + EOS_TOKEN
+        texts.append(text)
+    return {"text": texts}
+
+import pandas as pd
+from datasets import Dataset, DatasetDict
+from sklearn.model_selection import train_test_split
+
+# Load your data
+df = pd.read_csv('/content/data/train.csv')
+train_df, test_df = train_test_split(df, test_size=0.001, random_state=42)
+train_df, val_df = train_test_split(train_df, test_size=0.001, random_state=42)
+
+# Create datasets
+train_dataset = Dataset.from_pandas(train_df)
+val_dataset = Dataset.from_pandas(val_df)
+test_dataset = Dataset.from_pandas(test_df)
+
+# Combine datasets into a DatasetDict
+dataset_dict = DatasetDict({
+    'train': train_dataset,
+    'validation': val_dataset,
+    'test': test_dataset
+})
+
+# Apply the formatting function to the datasets
+dataset_dict = dataset_dict.map(formatting_prompts_func, batched=True)
+
+from trl import SFTTrainer
+from transformers import TrainingArguments
+from unsloth import is_bfloat16_supported
+
+dataset=dataset_dict
+trainer = SFTTrainer(
+    model = model,
+    tokenizer = tokenizer,
+    train_dataset = dataset['train'],
+    dataset_text_field = "text",
+    max_seq_length = max_seq_length,
+    dataset_num_proc = 2,
+    packing = False, # Can make training 5x faster for short sequences.
+    args = TrainingArguments(
+        per_device_train_batch_size = 4,
+        gradient_accumulation_steps = 32,
+        warmup_steps = 5,
+        # num_train_epochs = 1, # Set this for 1 full training run.
+        max_steps = 160,
+        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",
+    ),
+)
+
+#@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.")
+
+trainer_stats = trainer.train()
+
+#@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} %.")
+
+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.
+
+### Input:
+{}
+
+### Response:
+{}"""
+
+FastLanguageModel.for_inference(model) # Enable native 2x faster inference
+inputs = tokenizer(
+[
+    alpaca_prompt.format(
+        "What is alternate for Rifagut 400 mg?", # 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, skip_special_tokens=True)
+
+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.
+
+### Input:
+{}
+
+### Response:
+{}"""
+
+FastLanguageModel.for_inference(model) # Enable native 2x faster inference
+inputs = tokenizer(
+[
+    alpaca_prompt.format(
+        "what is alternate of Ocurax 400mg tablets?", # input
+        "", # output - leave this blank for generation!
+    )
+], return_tensors = "pt").to("cuda")
+
+from transformers import TextStreamer
+text_streamer = TextStreamer(tokenizer)
+_ = model.generate(**inputs, streamer = text_streamer, max_new_tokens = 200)
+
+# | f16 | q4_k_m | q5_k_m | q8_0 |
+
+# Save to 8-bit quantization (f16) GGUF
+model.save_pretrained_gguf("model-16bit", tokenizer, quantization_method="f16")
+
+# Push to Hugging Face hub with 8-bit quantization (f16)
+model.push_to_hub_gguf("student-abdullah/Llama3.1_medicine_fine-tuned_24-09_16bit_gguf", tokenizer, quantization_method="f16", token="hf_...")