Uploaded model

  • Developed by: cyumizou
  • License: CC-BY-NC-SA (Non-commercial Use Only)
  • Finetuned from model : llm-jp/llm-jp-3-13b

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

Datasets

The models have been fine-tuned on the following datasets. Language is Japanese only.

Purpose Dataset license copyright holder
Instruction tuning ichikara-instruction-003-002-1 CC-BY-NC-SA 理化学研究所
ichikara-instruction-003-003-1 CC-BY-NC-SA 理化学研究所
elyza/ELYZA-tasks-100 CC BY-SA 4.0 ELYZA
DPO weblab-GENIAC/aya-ja-nemotron-dpo-masked apache-2.0 GENIAC 松尾研LLMプロジェクト

Usage

The following script should be run for inference

# 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 = "cyumizou/llm-jp-3-13b-finetune-adapter_lora-dpo"

# Hugging Face Token を指定。
# 下記の URL から Hugging Face Token を取得できますので下記の HF_TOKEN に入れてください。
# https://huggingface.co/settings/tokens  
HF_TOKEN = "Your token" #@param {type:"string"}

# 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)

# タスクとなるデータの読み込み。
# 事前にデータをアップロードしてください。
datasets = []
with open("./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。

# 推論するためにモデルのモードを変更
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})
  

# 結果をjsonlで保存。

# ここではadapter_idを元にファイル名を決定しているが、ファイル名は任意で問題なし。
json_file_id = re.sub(".*/", "",adapter_id)
with open(f"/content/{json_file_id}_output.jsonl", 'w', encoding='utf-8') as f:
    for result in results:
        json.dump(result, f, ensure_ascii=False)
        f.write('\n')
Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model is not currently available via any of the supported Inference Providers.
The model cannot be deployed to the HF Inference API: The model has no pipeline_tag.

Model tree for cyumizou/llm-jp-3-13b-finetune-adapter_lora-dpo

Finetuned
(1121)
this model

Datasets used to train cyumizou/llm-jp-3-13b-finetune-adapter_lora-dpo