%%capture
!pip install camel-ai==0.2.16

%%capture
!pip install unsloth
# Also get the latest nightly Unsloth!
!pip uninstall unsloth -y && pip install --upgrade --no-cache-dir --no-deps git+https://github.com/unslothai/unsloth.git

import os
from datetime import datetime
import json
from camel.datagen.cotdatagen import CoTDataGenerator

from getpass import getpass

openai_api_key = getpass('Enter your OpenAI API key: ')
os.environ["OPENAI_API_KEY"] = openai_api_key

# import os
# from google.colab import userdata

# os.environ["OPENAI_API_KEY"] = userdata.get("OPENAI_API_KEY")

sys_msg = 'You are a genius at slow-thinking data and code'

from camel.models import ModelFactory
from camel.types import ModelPlatformType, ModelType
from camel.configs import ChatGPTConfig

# Define the model, here in this case we use gpt-4o-mini
model = ModelFactory.create(
    model_platform=ModelPlatformType.OPENAI,
    model_type=ModelType.GPT_4O_MINI,
    model_config_dict=ChatGPTConfig().as_dict(), # [Optional] the config for model
)

from camel.agents import ChatAgent
chat_agent = ChatAgent(
    system_message=sys_msg,
    model=model,
    message_window_size=10,
)

#get example json data

import requests
import json

# URL of the JSON file
url = 'https://raw.githubusercontent.com/zjrwtx/alldata/refs/heads/main/qa_data.json'

# Send a GET request to fetch the JSON file
response = requests.get(url)

# Check if the request was successful
if response.status_code == 200:
    # Parse the response content as JSON
    json_data = response.json()

    # Specify the file path to save the JSON data
    file_path = 'qa_data.json'

    # Write the JSON data to the file
    with open(file_path, 'w', encoding='utf-8') as json_file:
        json.dump(json_data, json_file, ensure_ascii=False, indent=4)

    print(f"JSON data successfully saved to {file_path}")
else:
    print(f"Failed to retrieve JSON file. Status code: {response.status_code}")

with open(file_path, 'r', encoding='utf-8') as f:
    qa_data = json.load(f)

# Create an instance of CoTDataGenerator
testo1 = CoTDataGenerator(chat_agent, golden_answers=qa_data)

# Record generated answers
generated_answers = {}

# Test Q&A
for question in qa_data.keys():
    print(f"Question: {question}")

    # Get AI's thought process and answer
    answer = testo1.get_answer(question)
    generated_answers[question] = answer
    print(f"AI's thought process and answer:\n{answer}")

    # Verify the answer
    is_correct = testo1.verify_answer(question, answer)
    print(f"Answer verification result: {'Correct' if is_correct else 'Incorrect'}")
    print("-" * 50)
    print()  # Add a new line at the end of each iteration

simplified_output = {
    'timestamp': datetime.now().isoformat(),
    'qa_pairs': generated_answers
}
simplified_file = f'generated_answers_{datetime.now().strftime("%Y%m%d_%H%M%S")}.json'
with open(simplified_file, 'w', encoding='utf-8') as f:
    json.dump(simplified_output, f, ensure_ascii=False, indent=2)
print(f"The generated answers have been exported to: {simplified_file}")

import json
from datetime import datetime

def transform_qa_format(input_file):
    # Read the input JSON file
    with open(input_file, 'r', encoding='utf-8') as f:
        data = json.load(f)

    # Transform the data
    transformed_data = []
    for question, answer in data['qa_pairs'].items():
        transformed_pair = {
            "instruction": question,
            "input": "",
            "output": answer
        }
        transformed_data.append(transformed_pair)

    # Generate output filename with timestamp
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    output_file = f'transformed_qa_{timestamp}.json'

    # Write the transformed data
    with open(output_file, 'w', encoding='utf-8') as f:
        json.dump(transformed_data, f, ensure_ascii=False, indent=2)

    return output_file, transformed_data

output_file, transformed_data = transform_qa_format(simplified_file)
print(f"Transformation complete. Output saved to: {output_file}")

# Import necessary modules and classes
from camel.datahubs.huggingface import HuggingFaceDatasetManager  # Manages interactions with Hugging Face datasets
from camel.datahubs.models import Record  # Represents a single record in the dataset
from datetime import datetime  # Handles date and time operations

# Main function: Upload dataset to Hugging Face
def upload_to_huggingface(transformed_data, username, dataset_name=None):
    r"""Uploads transformed data to the Hugging Face dataset platform.

    Args:
        transformed_data (list): Transformed data, typically a list of dictionaries.
        username (str): Hugging Face username.
        dataset_name (str, optional): Custom dataset name.

    Returns:
        str: URL of the uploaded dataset.
    """
    # Initialize HuggingFaceDatasetManager to interact with Hugging Face datasets
    manager = HuggingFaceDatasetManager()

    # Generate or validate the dataset name
    dataset_name = generate_or_validate_dataset_name(username, dataset_name)

    # Create the dataset on Hugging Face and get the dataset URL
    dataset_url = create_dataset(manager, dataset_name)

    # Create a dataset card to add metadata
    create_dataset_card(manager, dataset_name, username)

    # Convert the transformed data into a list of Record objects
    records = create_records(transformed_data)

    # Add the Record objects to the dataset
    add_records_to_dataset(manager, dataset_name, records)

    # Return the dataset URL
    return dataset_url

# Generate or validate the dataset name
def generate_or_validate_dataset_name(username, dataset_name):
    r"""Generates a default dataset name or validates and formats a user-provided name.

    Args:
        username (str): Hugging Face username.
        dataset_name (str, optional): User-provided custom dataset name.

    Returns:
        str: Formatted dataset name.
    """
    if dataset_name is None:
        # If no dataset name is provided, generate a default name with the username and current date
        dataset_name = f"{username}/qa-dataset-{datetime.now().strftime('%Y%m%d')}"
    else:
        # If a dataset name is provided, format it to include the username
        dataset_name = f"{username}/{dataset_name}"
    return dataset_name

# Create a dataset on Hugging Face
def create_dataset(manager, dataset_name):
    r"""Creates a new dataset on Hugging Face and returns the dataset URL.

    Args:
        manager (HuggingFaceDatasetManager): Instance of HuggingFaceDatasetManager.
        dataset_name (str): Name of the dataset.

    Returns:
        str: URL of the created dataset.
    """
    print(f"Creating dataset: {dataset_name}")
    # Use HuggingFaceDatasetManager to create the dataset
    dataset_url = manager.create_dataset(name=dataset_name)
    print(f"Dataset created: {dataset_url}")
    return dataset_url

# Create a dataset card with metadata
def create_dataset_card(manager, dataset_name, username):
    r"""Creates a dataset card to add metadata

    Args:
        manager (HuggingFaceDatasetManager): Instance of HuggingFaceDatasetManager.
        dataset_name (str): Name of the dataset.
        username (str): Hugging Face username.
    """
    print("Creating dataset card...")
    # Use HuggingFaceDatasetManager to create the dataset card
    manager.create_dataset_card(
        dataset_name=dataset_name,
        description="Question-Answer dataset generated by CAMEL CoTDataGenerator",  # Dataset description
        license="mit",  # Dataset license
        language=["en"],  # Dataset language
        size_category="<1MB",  # Dataset size category
        version="0.1.0",  # Dataset version
        tags=["camel", "question-answering"],  # Dataset tags
        task_categories=["question-answering"],  # Dataset task categories
        authors=[username]  # Dataset authors
    )
    print("Dataset card created successfully.")

# Convert transformed data into Record objects
def create_records(transformed_data):
    r"""Converts transformed data into a list of Record objects.

    Args:
        transformed_data (list): Transformed data, typically a list of dictionaries.

    Returns:
        list: List of Record objects.
    """
    records = []
    # Iterate through the transformed data and convert each dictionary into a Record object
    for item in transformed_data:
        record = Record(**item)  # Use the dictionary key-value pairs to create a Record object
        records.append(record)
    return records

# Add Record objects to the dataset
def add_records_to_dataset(manager, dataset_name, records):
    r"""Adds a list of Record objects to the dataset.

    Args:
        manager (HuggingFaceDatasetManager): Instance of HuggingFaceDatasetManager.
        dataset_name (str): Name of the dataset.
        records (list): List of Record objects.
    """
    print("Adding records to the dataset...")
    # Use HuggingFaceDatasetManager to add the records to the dataset
    manager.add_records(dataset_name=dataset_name, records=records)
    print("Records added successfully.")

HUGGING_FACE_TOKEN = getpass('Enter your HUGGING_FACE_TOKEN: ')
os.environ["HUGGING_FACE_TOKEN"] = HUGGING_FACE_TOKEN

# import os
# from google.colab import userdata

# os.environ["HUGGING_FACE_TOKEN"] = userdata.get("HUGGING_FACE_TOKEN")

# Set your personal huggingface config, then upload to HuggingFace
username = input("Enter your HuggingFace username: ")
dataset_name = input("Enter dataset name (press Enter to use default): ").strip()
if not dataset_name:
    dataset_name = None

try:
    dataset_url = upload_to_huggingface(transformed_data, username, dataset_name)
    print(f"\nData successfully uploaded to HuggingFace!")
    print(f"Dataset URL: {dataset_url}")
except Exception as e:
    print(f"Error uploading to HuggingFace: {str(e)}")

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/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(
    # Can select any from the below:
    # "unsloth/Qwen2.5-0.5B", "unsloth/Qwen2.5-1.5B", "unsloth/Qwen2.5-3B"
    # "unsloth/Qwen2.5-14B",  "unsloth/Qwen2.5-32B",  "unsloth/Qwen2.5-72B",
    # And also all Instruct versions and Math. Coding versions!
    model_name = "unsloth/Qwen2.5-1.5B",
    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("zjrwtxtechstudio/o1data06", 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,
    packing = False, # Can make training 5x faster for short sequences.
    args = TrainingArguments(
        per_device_train_batch_size = 2,
        gradient_accumulation_steps = 4,
        warmup_steps = 5,
        # num_train_epochs = 1, # Set this for 1 full training run.
        max_steps = 60,
        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
    ),
)

#@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 is copied from above
FastLanguageModel.for_inference(model)  # Enable native 2x faster inference

# Prepare the input for inference
inputs = tokenizer(
    [
        alpaca_prompt.format(
            "how many r in strawberry？",  # Instruction
            "",  # Input (empty for this example)
            "",  # Output (leave blank for generation)
        )
    ],
    return_tensors="pt"
).to("cuda")

# Generate the output
outputs = model.generate(
    **inputs,
    max_new_tokens=4096,  # Maximum number of tokens to generate
    use_cache=True        # Use cache for faster inference
)

# Decode the generated output and clean it
decoded_outputs = tokenizer.batch_decode(outputs, skip_special_tokens=True)

# Print the cleaned output
print(decoded_outputs[0])  # Print the first (and only) output

model.save_pretrained("lora_model") # Local saving
tokenizer.save_pretrained("lora_model")
model.push_to_hub("zjrwtxtechstudio/qwen2.5-1.5b-cot", token = "  ") # Online saving
tokenizer.push_to_hub("zjrwtxtechstudio/qwen2.5-1.5b-cot", token = "  ") # Online saving

if True:
    from unsloth import FastLanguageModel
    model, tokenizer = FastLanguageModel.from_pretrained(
        model_name = "zjrwtxtechstudio/qwen2.5-1.5b-cot", # YOUR MODEL YOU USED FOR TRAINING
        max_seq_length = max_seq_length,
        dtype = dtype,
        load_in_4bit = load_in_4bit,
    )
    FastLanguageModel.for_inference(model) # Enable native 2x faster inference

# alpaca_prompt = You MUST copy from above!

inputs = tokenizer(
[
    alpaca_prompt.format(
        "which one is bigger between 9.11 and 9.9？", # instruction
        "", # 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 = 4098)