Uploaded model

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

1.ファインチューニングした本モデルを使用して推論するモデルとトークナイザを読み出すコードの例を以下に示します。

from unsloth import FastLanguageModel

model_name = "SusumuDou/llm-jp-3-13b-finetune-2"

max_seq_length = 2048

dtype = None

load_in_4bit = True

model, tokenizer = FastLanguageModel.from_pretrained( model_name = model_name, max_seq_length = max_seq_length, dtype = dtype, load_in_4bit = load_in_4bit, token = HF TOKEN, )

FastLanguageModel.for_inference(model)

2.上記1の推論モデルとトークナイザを使って推論したoutput.jsonlの出力方法を以下に示します。

 モデルに推論させる入力ファイル:LLM_2024/最終課題/elyza-tasks-100-TV_0.jsonl

(1) 入力ファイルの読み込みコード

datasets = []

with open("/content/drive/MyDrive/LLM_2024/最終課題/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 = ""
    

(2) 推論コード

from tqdm import tqdm

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})

(3) 推論結果output.jsonlの出力コード

with open(f"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 API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for SusumuDou/llm-jp-3-13b-finetune-2

Finetuned
(1140)
this model