mtakashi/llm-jp-3-13b_qlora

Uploaded model

• Developed by: mtakashi
• License: apache-2.0, CC-BY-NC-SA
• Finetuned from model : llm-jp/llm-jp-3-13b

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

Usage

以下は、ELYZA-tasks-100-TVに対する回答の出力方法です

#from huggingface_hub import notebook_login
from google.colab import userdata
HF_TOKEN=userdata.get('HF_TOKEN')

# unsloth
# 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

#モデル名
model_id = "llm-jp/llm-jp-3-13b"
adapter_id = "mtakashi/llm-jp-3-13b_qlora" # 学習モデルを指定（Lolaアダプタ）

## unslothのFastLanguageModelで元のモデルをロード
dtype = None
load_in_4bit = True

model, tokenizer = FastLanguageModel.from_pretrained(
    model_name=model_id,
    dtype=dtype,
    load_in_4bit=load_in_4bit,
    trust_remote_code=True,
)
from peft import PeftModel

# オリジナルモデルにLoRAのアダプタを統合
model = PeftModel.from_pretrained(model, adapter_id, token = HF_TOKEN)

# 推論するためにモデルのモードを変更
FastLanguageModel.for_inference(model)

import json
datasets = []
with open("/content/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 = ""

#llm-jpモデルを用いてタスクを実行
from tqdm import tqdm
results = []
for dt in tqdm(datasets):
  input = dt["input"]

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

  tokenized_input = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
  outputs = model.generate(tokenized_input, max_new_tokens = 512, do_sample=True, top_p=0.95, temperature=0.7, repetition_penalty=1.05,)
  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})

# jsonlで保存
with open(f"{model_id}_output.jsonl", 'w', encoding='utf-8') as f:
    for result in results:
        json.dump(result, f, ensure_ascii=False)
        f.write('\n')
    print(f"JSONLファイルを出力しました: {model_id}_output.jsonl")