nntsuzu/Phi3.5-new-tokenizer

Model Description

Phi-3.5-mini-instruct tokenizer strong in Chinese, Japanese, and Korean.

Uses

from transformers import LlamaTokenizer

tokenizer = LlamaTokenizer.from_pretrained("nntsuzu/Phi3.5-new-tokenizer")
text = "我愛你。愛しています。사랑해. I love you."
print(tokenizer.tokenize(text))

Fine-tune Code

Use Phi-3.5-mini-instruct when you want to fine-tune it with Chinese, Japanese and Korean data.

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

logger = logging.getLogger(__name__)

# Hyper-parameters
training_config = {
    "bf16": True,
    "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",
    "overwrite_output_dir": True,
    "per_device_eval_batch_size": 4,
    "per_device_train_batch_size": 4,
    "remove_unused_columns": True,
    "save_steps": 100,
    "save_total_limit": 1,
    "seed": 0,
    "gradient_checkpointing": True,
    "gradient_checkpointing_kwargs":{"use_reentrant": False},
    "gradient_accumulation_steps": 1,
    "warmup_ratio": 0.2,
    }

peft_config = {
    "r": 8,
    "lora_alpha": 16,
    "lora_dropout": 0.05,
    "bias": "none",
    "task_type": "CAUSAL_LM",
    "target_modules": "all-linear",
    "modules_to_save": None,
}
train_conf = TrainingArguments(**training_config)
peft_conf = LoraConfig(**peft_config)

# Setup logging
logging.basicConfig(
    format="%(asctime)s - %(levelname)s - %(name)s - %(message)s",
    datefmt="%Y-%m-%d %H:%M:%S",
    handlers=[logging.StreamHandler(sys.stdout)],
)
log_level = train_conf.get_process_log_level()
logger.setLevel(log_level)
datasets.utils.logging.set_verbosity(log_level)
transformers.utils.logging.set_verbosity(log_level)
transformers.utils.logging.enable_default_handler()
transformers.utils.logging.enable_explicit_format()

# Log on each process a small summary
logger.warning(
    f"Process rank: {train_conf.local_rank}, device: {train_conf.device}, n_gpu: {train_conf.n_gpu}"
    + f" distributed training: {bool(train_conf.local_rank != -1)}, 16-bits training: {train_conf.fp16}"
)
logger.info(f"Training/evaluation parameters {train_conf}")
logger.info(f"PEFT parameters {peft_conf}")

# Model Loading
checkpoint_path = "microsoft/Phi-3.5-mini-instruct"
tokenizer_path = "nntsuzu/Phi3.5-new-tokenizer"

model_kwargs = dict(
    use_cache=False,
    trust_remote_code=True,
    attn_implementation="flash_attention_2",  # loading the model with flash-attenstion support
    torch_dtype=torch.bfloat16,
    device_map=None
)
model = AutoModelForCausalLM.from_pretrained(checkpoint_path, **model_kwargs)
tokenizer = AutoTokenizer.from_pretrained(tokenizer_path)
model.resize_token_embeddings(len(tokenizer))

tokenizer.model_max_length = 2048
tokenizer.pad_token = tokenizer.unk_token  # use unk rather than eos token to prevent endless generation
tokenizer.pad_token_id = tokenizer.convert_tokens_to_ids(tokenizer.pad_token)
tokenizer.padding_side = 'right'

# Data Processing
def apply_chat_template(
    example,
    tokenizer,
):
    messages = example["messages"]
    example["text"] = tokenizer.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=False)
    return example

raw_dataset = load_dataset("HuggingFaceH4/ultrachat_200k")
train_dataset = raw_dataset["train_sft"]
test_dataset = raw_dataset["test_sft"]
column_names = list(train_dataset.features)

processed_train_dataset = train_dataset.map(
    apply_chat_template,
    fn_kwargs={"tokenizer": tokenizer},
    num_proc=10,
    remove_columns=column_names,
    desc="Applying chat template to train_sft",
)

processed_test_dataset = test_dataset.map(
    apply_chat_template,
    fn_kwargs={"tokenizer": tokenizer},
    num_proc=10,
    remove_columns=column_names,
    desc="Applying chat template to test_sft",
)

# Training
trainer = SFTTrainer(
    model=model,
    args=train_conf,
    peft_config=peft_conf,
    train_dataset=processed_train_dataset,
    eval_dataset=processed_test_dataset,
    max_seq_length=2048,
    dataset_text_field="text",
    tokenizer=tokenizer,
    packing=True
)
train_result = trainer.train()
metrics = train_result.metrics
trainer.log_metrics("train", metrics)
trainer.save_metrics("train", metrics)
trainer.save_state()

# Evaluation
tokenizer.padding_side = 'left'
metrics = trainer.evaluate()
metrics["eval_samples"] = len(processed_test_dataset)
trainer.log_metrics("eval", metrics)
trainer.save_metrics("eval", metrics)

# Save model
trainer.save_model(train_conf.output_dir)