You can deploy this model and start earning money today!

awning/
llm-jp-3-13b-finetune

Name: llm-jp-3-13b-finetune
Availability: InStock

Views

No views yet

Model Card for Model ID

This repository provides a LoRA adapter fine-tuned on top of the llm-jp-13b model using Supervised Fine-Tuning (SFT).
The adapter itself is lightweight and can be integrated into the base model.

Model Details

Model Description

Developed by: awning (yt888888)
Language(s) (NLP): Japanese
License: apache-2.0
Finetuned from model [optional]: llm-jp/llm-jp-3-13b

Model Sources [optional]

Repository: https://huggingface.co/llm-jp/llm-jp-3-13b

How to Get Started with the Model

Use the code below to get started with the model.

python

# Install necessary libraries
!pip install -U bitsandbytes
!pip install -U transformers
!pip install -U accelerate
!pip install -U datasets
!pip install -U peft

# Load libraries
from transformers import (
    AutoModelForCausalLM,
    AutoTokenizer,
    BitsAndBytesConfig,
)
from peft import PeftModel
import torch
from tqdm import tqdm
import json

# Set base model and LoRa adapter
model_id = "llm-jp/llm-jp-3-13b" 
adapter_id = "awning/llm-jp-3-13b-finetune"

HF_TOKEN = "YOUR_TOKEN"

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

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

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

# Join base model and adapter
model = PeftModel.from_pretrained(model, adapter_id, token = HF_TOKEN)

# Load dataset for inference
datasets = []
with open("{★JSON Lines形式のファイル★}", "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
  """

# Inference
tokenized_input = tokenizer.encode(prompt, add_special_tokens=False, return_tensors="pt").to(model.device)
attention_mask = torch.ones_like(tokenized_input)
with torch.no_grad():
    outputs = model.generate(
        tokenized_input,
        attention_mask=attention_mask,
        max_new_tokens=100,
        do_sample=False,
        repetition_penalty=1.2,
        pad_token_id=tokenizer.eos_token_id
    )[0]
output = tokenizer.decode(outputs[tokenized_input.size(1):], skip_special_tokens=True)

results.append({"task_id": data["task_id"], "input": input, "output": output})

# Save result as JSON
import re
jsonl_id = re.sub(".*/", "", adapter_id)
with open(f"{jsonl_id}-outputs.jsonl", 'w', encoding='utf-8') as f:
    for result in results:
        json.dump(result, f, ensure_ascii=False)
        f.write('\n')

Dataset Information

SFT (Supervised Fine-Tuning) was performed using the following dataset.

Language	Dataset
Japanese	ichikara-instruction-003-003-1