7.7 Extracting and saving responses

Chapter 7-Fine-tuning to follow instructions

7.7 Extracting and saving responses

• 在本节中，我们保存测试集响应以便在下一节中评分，除此之外保存模型的副本以供将来使用。

  ​

• 首先，让我们简单看看finetuned模型生成的响应

  torch.manual_seed(123)for entry in test_data[:3]:input_text = format_input(entry)token_ids = generate(model=model,idx=text_to_token_ids(input_text, tokenizer).to(device),max_new_tokens=256,context_size=BASE_CONFIG["context_length"],eos_id=50256)generated_text = token_ids_to_text(token_ids, tokenizer)response_text = (generated_text[len(input_text):].replace("### Response:", "").strip()
  )print(input_text)print(f"\nCorrect response:\n>> {entry['output']}")print(f"\nModel response:\n>> {response_text.strip()}")print("-------------------------------------")"""输出"""
  Below is an instruction that describes a task. Write a response that appropriately completes the request.### Instruction:
  Rewrite the sentence using a simile.### Input:
  The car is very fast.Correct response:
  >> The car is as fast as lightning.Model response:
  >> The car is as fast as a cheetah.
  -------------------------------------
  Below is an instruction that describes a task. Write a response that appropriately completes the request.### Instruction:
  What type of cloud is typically associated with thunderstorms?Correct response:
  >> The type of cloud typically associated with thunderstorms is cumulonimbus.Model response:
  >> The type of cloud associated with thunderstorms is a cumulus cloud.
  -------------------------------------
  Below is an instruction that describes a task. Write a response that appropriately completes the request.### Instruction:
  Name the author of 'Pride and Prejudice'.Correct response:
  >> Jane Austen.Model response:
  >> The author of 'Pride and Prejudice' is Jane Austen.
  -------------------------------------
  

  正如我们根据测试集指令、给定响应和模型响应所看到的，模型的性能相对较好，第一个和最后一个说明的答案显然是正确的，第二个答案很接近；模型用“cumulus cloud”而不是“cumulonimbus”来回答（但是，请注意前者可以发展成后者，后者能够产生thunderstorms）

• 最重要的是，我们可以看到模型评估不像上一章我们只需要计算正确垃圾邮件/非垃圾邮件类别标签的百分比即可获得分类准确性的那样简单。

• 在实践中，聊天机器人等instruction-finetunedLLM通过多种方法进行评估

  1. 短答案和多项选择基准，例如MMLU（“测量大规模多任务语言理解”，[https://arxiv.org/pdf/2009.03300]（https://arxiv.org/pdf/2009.03300）），用于测试模型的知识
  2. 与其他LLM的人类偏好比较，例如LMSYS聊天机器人竞技场（[https://arena.lmsys.org]（https://arena.lmsys.org））
  3. 自动会话基准测试，其中使用另一个LLM（如GPT-4）来评估响应，例如AlpackaEval（[https://tatsu-lab.github.io/alpaca_eval/]（https://tatsu-lab.github.io/alpaca_eval/））

• 在下一节中，我们将使用类似于AlpackaEval的方法，并使用另一个LLM来评估我们模型的响应；但是，我们将使用我们自己的测试集，而不是使用公开可用的基准数据集。为此，我们将模型响应添加到“test_data”字典中，并将其保存为“instruction-data-with-response. json”文件以进行记录保存，以便我们可以在需要时在单独的Python会话中加载和分析它

  from tqdm import tqdmfor i, entry in tqdm(enumerate(test_data), total=len(test_data)):input_text = format_input(entry)token_ids = generate(model=model,idx=text_to_token_ids(input_text, tokenizer).to(device),max_new_tokens=256,context_size=BASE_CONFIG["context_length"],eos_id=50256)generated_text = token_ids_to_text(token_ids, tokenizer)response_text = generated_text[len(input_text):].replace("### Response:", "").strip()test_data[i]["model_response"] = response_textwith open("instruction-data-with-response.json", "w") as file:json.dump(test_data, file, indent=4)  # "indent" for pretty-printing"""输出"""
  100%|██████████| 110/110 [00:59<00:00,  1.86it/s]
  

  检查其中一个条目，看看响应是否已正确添加到“test_data”字典中

  print(test_data[0])"""输出"""
  {'instruction': 'Rewrite the sentence using a simile.', 'input': 'The car is very fast.', 'output': 'The car is as fast as lightning.', 'model_response': 'The car is as fast as a cheetah.'}
  

  可以看到原始的output和模型的model_response都在

• 最后我们保存微调后的模型以供将来用

  import refile_name = f"E:\\LLM\\gpt2\\{re.sub(r'[ ()]', '', CHOOSE_MODEL) }-sft.pth"
  torch.save(model.state_dict(), file_name)
  print(f"Model saved as {file_name}")# Load model via
  # model.load_state_dict(torch.load("E:\\LLM\\gpt2\\gpt2-medium355M-sft.pth"))"""输出"""
  Model saved as E:\LLM\gpt2\gpt2-medium355M-sft.pth
  

  ​

  ---

7.8 Evaluating the finetuned LLM