Downloads · 30 days
0
E-tanaka/llm-jp-3-13b-it-etanaka16
llm-jp-3-13b-it-etanaka16 is a machine learning model from E-tanaka. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.
- Developed by: E-tanaka - License: apache-2.0 - Finetuned from model : llm-jp/llm-jp-3-13b
Downloads · 30 days
0
Access
Public
Updated Dec 10, 2024
Repo size
501 MB
Likes
0
Public
Click a slice to open those files.
.safetensors501 MB · 99%
From the Hugging Face model README
This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.
trial history: 1.dataset:ichikara-instruction-003-001-5.2.json model:llm-jp-3-13b-it-etanaka16
score=2.94 final 2.dataset:my original data(7items) model:llm-jp-3-13b-it-etanaka16_rev1 score=2.21 3.dataset:my original data(7items)+ichikara-instruction-003-001-5.2.json model:llm-jp-3-13b-it-etanaka16_rev2 score=2.84 4.dataset:ichikara-instruction-003-001-1.json model:llm-jp-3-13b-it-etanaka16_rev3 score=2.9 5.dataset:my original data(25items) model:llm-jp-3-13b-it-etanaka16_rev4 score=2.25 6.dataset:ichikara-instruction-003-001-1.json", ichikara-instruction-003-001-2.1.json", ichikara-instruction-003-001-2.2.json", ichikara-instruction-003-001-5.1.json", ichikara-instruction-003-001-5.2.json model:llm-jp-3-13b-it-etanaka16_rev5 score=2.73
Code:LoRA_template_unsloth2.ipynb >>>>
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig from unsloth import FastLanguageModel import torch max_seq_length = 1024#512 # unslothではRoPEをサポートしているのでコンテキスト長は自由に設定可能 dtype = None # Noneにしておけば自動で設定 load_in_4bit = True # 今回は8Bクラスのモデルを扱うためTrue
model_id = "llm-jp/llm-jp-3-13b" new_model_id = "llm-jp-3-13b-it-etanaka16_rev8" #Fine-Tuningしたモデルにつけたい名前、it: Instruction Tuning
model, tokenizer = FastLanguageModel.from_pretrained( model_name=model_id, dtype=dtype, load_in_4bit=load_in_4bit, trust_remote_code=True, )
model = FastLanguageModel.get_peft_model( model, r = 32, target_modules = ["q_proj", "k_proj", "v_proj", "o_proj", "gate_proj", "up_proj", "down_proj",], lora_alpha = 32, lora_dropout = 0.05, bias = "none", use_gradient_checkpointing = "unsloth", random_state = 3407, use_rslora = False, loftq_config = None, max_seq_length = max_seq_length, )
HF_TOKEN = "" #@param {type:"string"}
from datasets import load_dataset
data_dir = "/content/drive/MyDrive/LLM/Distribution20241221_all/" data_files = [
#f"{data_dir}ichikara-instruction-003-001-2.1.json",
#f"{data_dir}ichikara-instruction-003-001-2.2.json",
f"{data_dir}ichikara-instruction-003-001-5.1.json",
f"{data_dir}ichikara-instruction-003-001-5.2.json"
]
dataset = load_dataset("json", data_files=data_files)
print(dataset)
prompt = """### 指示 {}
{}"""
""" formatting_prompts_func: 各データをプロンプトに合わせた形式に合わせる """ EOS_TOKEN = tokenizer.eos_token # トークナイザーのEOSトークン(文末トークン) def formatting_prompts_func(examples): input = examples["text"] # 入力データ output = examples["output"] # 出力データ text = prompt.format(input, output) + EOS_TOKEN # プロンプトの作成 return { "formatted_text" : text, } # 新しいフィールド "formatted_text" を返す pass
dataset = dataset.map( formatting_prompts_func, num_proc= 4, # 並列処理数を指定 )
dataset
from trl import SFTTrainer from transformers import TrainingArguments from unsloth import is_bfloat16_supported
trainer = SFTTrainer( model = model, tokenizer = tokenizer, train_dataset=dataset["train"], max_seq_length = max_seq_length, dataset_text_field="formatted_text", packing = False, args = TrainingArguments( per_device_train_batch_size = 2, gradient_accumulation_steps = 4, num_train_epochs = 1, logging_steps = 10, warmup_steps = 10, save_steps=100, save_total_limit=2, max_steps=-1, learning_rate = 2e-4, fp16 = not is_bfloat16_supported(), bf16 = is_bfloat16_supported(), group_by_length=True, seed = 3407, output_dir = "outputs", report_to = "none", ), )
#@title 学習実行 trainer_stats = trainer.train()