在询问ChatGPT互联网上并不存在内容的时候,能给出较好答案(如用ChatGPT学建模);ChatGPT能通过信
息猜你心中的想法;你可以制定一个全新的游戏规则让ChatGPT和你玩,ChatGPT可以理解。
5. 大模型生成时的参数怎么设置?
生成模型预测调参建议:
建议去调整下 top_p, num_beams, repetition_renalty, temperature, do_sample=True;
数据生成有重复,调高repetition_renalty;
生成任务表达单一的,样本也不多的,可适当调低 temperature,生成的样子跟训练集的比较像;如果要复现训
练集的效果,temperature=0.01即可。
以上是经验参数,具体调参根据任务而定,不是固定的。
6. 有哪些省内存的大语言模型训练/微调/推理方法?
举例来说,即使 RTX 3090 有着 24GB 的 RAM,是除了 A100 之外显存最大的显卡。但使用一块 RTX 3090 依
然无法 fp32 精度训练最小号的 LLaMA-6B。
• 参数解释:
top_p=0.9,
#Moderately increase the probability threshold of nucleus sampling to increase the
quantity of candidate tokens and increase generation diversity.
temperature=1.0,
#The previous low temperature parameter could lead to a severe polarization in the
probability distribution of generated words, which degenerates the generation
strategy into greedy decoding.
do_sample=True,
#do_sample parameter is set to False by default. After setting to True, the
generation methods turn into beam-search multinomial sampling decoding strategy.
no_repeat_ngram_size=6,
#Configure the probability of the next repeating n-gram to 0, to ensure that there
are no n-grams appearing twice. This setting is an empirical preliminary
exploration.
repetition_penalty=1.8,
#For words that have appeared before, in the subsequent prediction process, we
reduce the probability of their reoccurrence by introducing the repetition_penalty
parameter. This setting is an empirical preliminary exploration.
• 动机:大模型(LLMs)现在是 NLP 领域的最主流方法之一,但是大模型的训练/微调/推理需要的内存也越来
越多。
• Memory-Efficient 的 LLMs 的训练/微调/推理方法
评论