Views
No views yet
1!pip install -U bitsandbytes
2!pip install -U transformers
3!pip install -U accelerate
4!pip install -U datasets
5!pip install -U peft
6from transformers import (
7 AutoModelForCausalLM,
8 AutoTokenizer,
9 BitsAndBytesConfig,
10)
11from peft import PeftModel
12import torch
13from tqdm import tqdm
14import json
15HF_TOKEN = "Hugging Face Token"
16model_id = "llm-jp/llm-jp-3-13b"
17adapter_id = "Yusuke525/llm-jp-3-13b-ft-hirano-synthetic_2000"1# QLoRA config
2bnb_config = BitsAndBytesConfig(
3 load_in_4bit=True,
4 bnb_4bit_quant_type="nf4",
5 bnb_4bit_compute_dtype=torch.bfloat16,
6)
7# Load model
8model = AutoModelForCausalLM.from_pretrained(
9 model_id,
10 quantization_config=bnb_config,
11 device_map="auto",
12 token = HF_TOKEN
13)
14# Load tokenizer
15tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True, token = HF_TOKEN)
16# 元のモデルにLoRAのアダプタを統合。
17model = PeftModel.from_pretrained(model, adapter_id, token = HF_TOKEN)1datasets = []
2with open("./elyza-tasks-100-TV_0.jsonl", "r") as f:
3 item = ""
4 for line in f:
5 line = line.strip()
6 item += line
7 if item.endswith("}"):
8 datasets.append(json.loads(item))
9 item = ""1results = []
2for data in tqdm(datasets):
3 input = data["input"]
4 prompt = f"""### 指示
5 {input}
6 ### 回答
7 """
8 tokenized_input = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
9 attention_mask = torch.ones_like(tokenized_input)
10 with torch.no_grad():
11 outputs = model.generate(
12 tokenized_input,
13 attention_mask=attention_mask,
14 max_new_tokens=100,
15 do_sample=False,
16 repetition_penalty=1.2,
17 pad_token_id=tokenizer.eos_token_id
18 )[0]
19 output = tokenizer.decode(outputs[tokenized_input.size(1):], skip_special_tokens=True)
20 results.append({"task_id": data["task_id"], "input": input, "output": output})1import re
2jsonl_id = re.sub(".*/", "", adapter_id)
3with open(f"./{jsonl_id}-outputs.jsonl", 'w', encoding='utf-8') as f:
4 for result in results:
5 json.dump(result, f, ensure_ascii=False) # ensure_ascii=False for handling non-ASCII characters
6 f.write('\n')