monamonamona/llm-jp-3-13b-finetune-2

Uploaded model

• Developed by: monamonamona
• License: apache-2.0
• Finetuned from model : llm-jp/llm-jp-3-13b

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

Sample Use

以下はelyza-tasks-100-TV_0.jsonlの回答のためのコードです。

from transformers import (
    AutoModelForCausalLM,
    AutoTokenizer,
    BitsAndBytesConfig,
)
import torch
from tqdm import tqdm
import json

HF_TOKEN = "your-token"
model_name = "monamonamona/llm-jp-3-13b-finetune-2"

# QLoRA config
bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.bfloat16,
    bnb_4bit_use_double_quant=False,
)

# Load model
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    quantization_config=bnb_config,
    device_map="auto",
    token = HF_TOKEN
)

# Load tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True, token = HF_TOKEN)

# データセットの読み込み。
datasets = []
with open("./preddata/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 = ""

# llmjp
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=100,
            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})