fine-tune Microsoft/Phi-3-mini-128k-instruct

使用transformers在聊天指令数据集上使用LoRA对Phi-3语言模型进行微调，以改进会话理解和响应生成。

logging

日志记录:设置日志记录以跟踪训练过程。

import sys
import logging

import datasets
from datasets import load_dataset
from peft import LoraConfig
import torch
import transformers
from trl import SFTTrainer
from transformers import AutoModelForCausalLM, AutoTokenizer, TrainingArguments, BitsAndBytesConfig

# Logging setup (you can customize this as needed)
logging.basicConfig(
    format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
    datefmt="%Y-%m-%d %H:%M:%S",
    handlers=[logging.StreamHandler(sys.stdout)],
)
logger = logging.getLogger(__name__)

超参数

超参数:定义两个字典，training_config和peft_config，分别存储用于训练和PEFT(Parameter-Efficient Fine-Tuning)的超参数。
训练参数:从training_config字典中创建一个TrainingArguments对象。
PEFT配置:从peft_config字典创建一个LoraConfig对象，指定LoRA (Low-Rank Adaptation)设置以进行有效的微调。

# Training hyperparameters
training_config = {
   
    "bf16": True,  # Use mixed precision
    "do_eval": False,
    "learning_rate": 5.0e-06,
    "log_level": "info",
    "logging_steps": 20,
    "logging_strategy": "steps",
    "lr_scheduler_type": "cosine",
    "num_train_epochs": 1,
    "max_steps": -1,
    "output_dir": "./checkpoint_dir",