Update README.md

README.md

@@ -17,6 +17,135 @@ language:
 - **License:** apache-2.0
 - **Finetuned from model :** llm-jp/llm-jp-3-13b
 
-This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+このモデルは日本語タスクに特化し、指示に基づく応答を行うためにファインチューニングされたLLM用アダプタです。  
+`llm-jp/llm-jp-3-13b`をベースモデルとして、ichikara-instructionデータやELYZA-tasks-100を用いて微調整を行っています。
 
+This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
 [<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
+
+# Instruction Tuning
+
+以下の日本語データセットで微調整を行いました:
+
+| Language | Dataset | Description |
+|:---|:---|:---|
+| Japanese | [ichikara-instruction-003-001-1.json](https://liat-aip.sakura.ne.jp/wp/llm%E3%81%AE%E3%81%9F%E3%82%81%E3%81%AE%E6%97%A5%E6%9C%AC%E8%AA%9E%E3%82%A4%E3%83%B3%E3%82%B9%E3%83%88%E3%83%A9%E3%82%AF%E3%82%B7%E3%83%A7%E3%83%B3%E3%83%87%E3%83%BC%E3%82%BF-%E5%85%AC%E9%96%8B/) | 手動構築の日本語指示データセット |
+| Japanese | ichikara-instruction-003-003-1.fixed.json | 手動修正・拡張データセット |
+| Japanese | [elyza/ELYZA-tasks-100](https://huggingface.co/datasets/elyza/ELYZA-tasks-100) | 日本語多目的タスク用データセット |
+
+ichikara-instruction:
+関根聡, 安藤まや, 後藤美知子, 鈴木久美, 河原大輔, 井之上直也, 乾健太郎. ichikara-instruction: LLMのための日本語インストラクションデータの構築. 言語処理学会第30回年次大会(2024)
+
+## Usage
+
+下記コマンドで環境整備:
+
+```bash
+!pip install -U bitsandbytes
+!pip install -U transformers
+!pip install -U accelerate
+!pip install -U datasets
+```
+
+```python
+from transformers import (
+    AutoModelForCausalLM,
+    AutoTokenizer,
+    BitsAndBytesConfig,
+)
+import torch
+from tqdm import tqdm
+import json
+import re
+```
+
+```python
+HF_TOKEN = "YOUR-HF-TOKEN"
+```
+
+```python
+model_name = "KKFurudate/llm-jp-3-13b-v6_lora"
+```
+
+```python
+bnb_config = BitsAndBytesConfig(
+    load_in_4bit=True,
+    bnb_4bit_quant_type="nf4",
+    bnb_4bit_compute_dtype=torch.bfloat16,
+    bnb_4bit_use_double_quant=False,
+)
+```
+
+```python
+model = AutoModelForCausalLM.from_pretrained(
+    model_name,
+    quantization_config=bnb_config,
+    device_map="auto",
+    token=HF_TOKEN
+)
+
+tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True, token=HF_TOKEN)
+```
+
+```python
+datasets = []
+with open("./YOUR-DATA.jsonl", "r") as f:
+    item = ""
+    for line in f:
+        line = line.strip()
+        item += line
+        if item.endswith("}"):
+            datasets.append(json.loads(item))
+            item = ""
+
+results = []
+for data in tqdm(datasets):
+    input = data["input"]
+    prompt = f"""### 指示
+{input}
+### 回答：
+"""
+    tokenized_input = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
+    with torch.no_grad():
+        outputs = model.generate(
+            tokenized_input,
+            max_new_tokens=512,
+            do_sample=False,
+            repetition_penalty=1.2
+        )[0]
+    output = tokenizer.decode(outputs[tokenized_input.size(1):], skip_special_tokens=True)
+    results.append({"task_id": data["task_id"], "input": input, "output": output})
+```
+
+```python
+model_name_clean = re.sub(".*/", "", model_name)
+with open(f"./{model_name_clean}-outputs.jsonl", 'w', encoding='utf-8') as f:
+    for result in results:
+        json.dump(result, f, ensure_ascii=False)
+        f.write('\n')
+```
+
+## License
+
+[Apache License, Version 2.0](https://www.apache.org/licenses/LICENSE-2.0)
+
+
+## Base Model
+
+base_model:- llm-jp/llm-jp-3-13b
+
+## Citations
+
+@software{unsloth,
+  author = {Daniel Han, Michael Han and Unsloth team},
+  title = {Unsloth},
+  url = {http://github.com/unslothai/unsloth},
+  year = {2023}
+}
+
+@misc{elyzatasks100,
+      title={ELYZA-tasks-100: 日本語instructionモデル評価データセット},
+      url={https://huggingface.co/elyza/ELYZA-tasks-100},
+      author={Akira Sasaki and Masato Hirakawa and Shintaro Horie and Tomoaki Nakamura},
+      year={2023},
+}