Chasottco's picture
Update README.md
ede632e verified
|
raw
history blame
2.73 kB
metadata
base_model: llm-jp/llm-jp-3-13b
tags:
  - text-generation-inference
  - transformers
  - unsloth
  - llama
  - trl
license: apache-2.0
language:
  - en

Uploaded model

  • Developed by: Chasottco
  • License: apache-2.0
  • Finetuned from model : llm-jp/llm-jp-3-13b

This llama model was trained 2x faster with Unsloth and Huggingface's TRL library.


#Google Colabでの動作を想定

必要なライブラリをインストール

%%capture !pip install unsloth !pip uninstall unsloth -y && pip install --upgrade --no-cache-dir "unsloth[colab-new] @ git+https://github.com/unslothai/unsloth.git" !pip install -U torch !pip install -U peft

必要なライブラリを読み込み

from unsloth import FastLanguageModel from peft import PeftModel import torch import json from tqdm import tqdm import re

ベースとなるモデルと学習したLoRAのアダプタ(Hugging FaceのIDを指定)

model_id = "llm-jp/llm-jp-3-13b" adapter_id = "Chasottco/llm-jp-3-13b-it-Chasottco"

Hugging Face Token を指定

HF_TOKEN = ""

unslothのFastLanguageModelで元のモデルをロード。

dtype = None # Noneにしておけば自動で設定 load_in_4bit = True # 今回は13Bモデルを扱うためTrue

model, tokenizer = FastLanguageModel.from_pretrained( model_name=model_id, dtype=dtype, load_in_4bit=load_in_4bit, trust_remote_code=True, )

元のモデルにLoRAのアダプタを統合。

model = PeftModel.from_pretrained(model, adapter_id, token = HF_TOKEN)

google drive mount(事前にデータをアップロード)

from google.colab import drive drive.mount('/content/drive')

タスクとなるデータの読み込み。

datasets = [] with open("/content/drive/MyDrive/2024松尾研LLM/elyza-tasks-100-TV_0.jsonl", "r") as f: item = "" for line in f: line = line.strip() item += line if item.endswith("}"): datasets.append(json.loads(item)) item = ""

モデルを用いてタスクの推論

FastLanguageModel.for_inference(model)

results = [] for dt in tqdm(datasets): input = dt["input"]

prompt = f"""### 指示\n{input}\n### 回答\n"""

inputs = tokenizer([prompt], return_tensors = "pt").to(model.device)

outputs = model.generate(**inputs, max_new_tokens = 512, use_cache = True, do_sample=False, repetition_penalty=1.2) prediction = tokenizer.decode(outputs[0], skip_special_tokens=True).split('\n### 回答')[-1]

results.append({"task_id": dt["task_id"], "input": input, "output": prediction})