README.md

metadata

license: mit
language: ja
datasets:
  - kunishou/databricks-dolly-15k-ja
  - wikipedia
  - cc100
  - mc4
tags:
  - japanese
  - causal-lm
  - open-calm
inference: false

OpenCALM-LARGE

Model Description

OpenCALM is a suite of decoder-only language models pre-trained on Japanese datasets, developed by CyberAgent, Inc.

このモデルはpeftを用いてopen-calm-largeをLoRAファインチューニングしたものです。

Usage

pytorchおよびtransformers, peftをインストールして下記コードを実行してください

（pip install torch, transformers, peft）

and please execute this code.

下記コードに関しては

npakaさんの記事(https://note.com/npaka/n/na5b8e6f749ce)

を参考にさせて頂きました。 感謝致します。

import torch
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM, AutoTokenizer

model_name = "cyberagent/open-calm-large"
lora_weights = "Mizuiro-sakura/open-calm-large-finetuned-databricks-dolly"

# モデルの準備
model = AutoModelForCausalLM.from_pretrained(
    model_name
)

# トークンナイザーの準備
tokenizer = AutoTokenizer.from_pretrained(model_name)

# LoRAモデルの準備
model = PeftModel.from_pretrained(
    model, 
    lora_weights,
    adapter_name=lora_weights
)

# 評価モード
model.eval()

# プロンプトテンプレートの準備
def generate_prompt(data_point):
    if data_point["input"]:
        return f"""Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.

### Instruction:
{data_point["instruction"]}

### Input:
{data_point["input"]}

### Response:"""
    else:
        return f"""Below is an instruction that describes a task. Write a response that appropriately completes the request.

### Instruction:
{data_point["instruction"]}

### Response:"""

# テキスト生成関数の定義
def generate(instruction,input=None,maxTokens=256):
    # 推論
    prompt = generate_prompt({'instruction':instruction,'input':input})
    input_ids = tokenizer(prompt, return_tensors="pt", truncation=True).input_ids
    outputs = model.generate(
        input_ids=input_ids, 
        max_new_tokens=maxTokens, 
        do_sample=True,
        temperature=0.7, 
        top_p=0.75, 
        top_k=40,         
        no_repeat_ngram_size=2,
    )
    outputs = outputs[0].tolist()

    # EOSトークンにヒットしたらデコード完了
    if tokenizer.eos_token_id in outputs:
        eos_index = outputs.index(tokenizer.eos_token_id)
    else:
        eos_index = len(outputs)
    decoded = tokenizer.decode(outputs[:eos_index])
    
    # レスポンス内容のみ抽出
    sentinel = "### Response:"
    sentinelLoc = decoded.find(sentinel)
    if sentinelLoc >= 0:
        print(decoded[sentinelLoc+len(sentinel):])
    else:
        print('Warning: Expected prompt template to be emitted.  Ignoring output.')

generate("自然言語処理とは？")

Model Details

Model	Params	Layers	Dim	Heads	Dev ppl
cyberagent/open-calm-small	160M	12	768	12	19.7
cyberagent/open-calm-medium	400M	24	1024	16	13.8
cyberagent/open-calm-large	830M	24	1536	16	11.3
cyberagent/open-calm-1b	1.4B	24	2048	16	10.3
cyberagent/open-calm-3b	2.7B	32	2560	32	9.7
cyberagent/open-calm-7b	6.8B	32	4096	32	8.2

• Developed by: CyberAgent, Inc.
• Model type: Transformer-based Language Model
• Language: Japanese
• Library: GPT-NeoX
• License: OpenCALM is licensed under the Creative Commons Attribution-ShareAlike 4.0 International License (CC BY-SA 4.0). When using this model, please provide appropriate credit to CyberAgent, Inc.
  • Example (en): This model is a fine-tuned version of OpenCALM-XX developed by CyberAgent, Inc. The original model is released under the CC BY-SA 4.0 license, and this model is also released under the same CC BY-SA 4.0 license. For more information, please visit: https://creativecommons.org/licenses/by-sa/4.0/
  • Example (ja): 本モデルは、株式会社サイバーエージェントによるOpenCALM-XXをファインチューニングしたものです。元のモデルはCC BY-SA 4.0ライセンスのもとで公開されており、本モデルも同じくCC BY-SA 4.0ライセンスで公開します。詳しくはこちらをご覧ください: https://creativecommons.org/licenses/by-sa/4.0/

Training Dataset

• Wikipedia (ja)
• Common Crawl (ja)

Author

Ryosuke Ishigami

Citations

@software{gpt-neox-library,
  title = {{GPT-NeoX: Large Scale Autoregressive Language Modeling in PyTorch}},
  author = {Andonian, Alex and Anthony, Quentin and Biderman, Stella and Black, Sid and Gali, Preetham and Gao, Leo and Hallahan, Eric and Levy-Kramer, Josh and Leahy, Connor and Nestler, Lucas and Parker, Kip and Pieler, Michael and Purohit, Shivanshu and Songz, Tri and Phil, Wang and Weinbach, Samuel},
  url = {https://www.github.com/eleutherai/gpt-neox},
  doi = {10.5281/zenodo.5879544},
  month = {8},
  year = {2021},
  version = {0.0.1},
}