无论文勇闯秋招算法岗——奇富科技一面

面试官状态很好，思维活跃

问了deepspeed zero，zero 1 2 3 的区别？

模型显存占用大小，全参微调 32B 要占多少显存？

sft 全参和lora的区别，什么时候用全参，什么时候用lora，lora的rank是什么意思，有什么用

grpo是on policy还是off policy的？

grpo 和 gspo的区别？

做题

超时删除，一个类，init()，get(key) value，set() key value 存下来

import timeclass TimeCacheDict:
'''
'''
def __init__(self,expire_time:int):
    self.cache = {}
    self.expire_time = expire_time
    self.key_time = {}
def set(self,key:str,value:str):
    current_time = time.time()
    self.cache[key]=value
    self.key_time[key]=current_time
def get(self,key:str):
    current_time = time.time()
    if key not in self.cache:
        return None
    if current_time - self.key_time[key] > self.expire_time:
        del self.cache[key]
        del self.key_time[key]
        return None
    return self.cache[key]
cache=TimeCacheDict(3)
cache.set("key","value")
print(cache.get('key'))
time.sleep(4)
print(cache.get('key'))
cache.set("key","value")
print(cache.get('key'))
time.sleep(2)
print(cache.get('key'))

说了一些格式上的问题

然后说异步轮训或者队列先进先出的方案会更好。

异步轮训 + OrderedDict 方案：

import time
import asyncio
from collections import OrderedDict

class AsyncTimeCacheDict:
    def __init__(self, expire_time: int):
        self.cache = OrderedDict()
        self.expire_time = expire_time

    async def set(self, key: str, value: str):
        current_time = time.time()
        if key in self.cache:
            del self.cache[key]
        self.cache[key] = (value, current_time)
        await self._clean_expired_keys()

    async def get(self, key: str):
        current_time = time.time()
        if key not in self.cache:
            return None
        value, timestamp = self.cache[key]
        if current_time - timestamp > self.expire_time:
            del self.cache[key]
            return None
        return value

    async def _clean_expired_keys(self):
        current_time = time.time()
        keys_to_delete = []
        for key, (value, timestamp) in self.cache.items():
            if current_time - timestamp > self.expire_time:
                keys_to_delete.append(key)
            else:
                break
        for key in keys_to_delete:
            del self.cache[key]

# 测试代码
async def main():
    cache = AsyncTimeCacheDict(3)
    await cache.set("key1", "value1")
    await asyncio.sleep(1)
    await cache.set("key2", "value2")
    print(await cache.get("key1"))  # 输出: value1
    await asyncio.sleep(3)
    print(await cache.get("key1"))  # 输出: None
    print(await cache.get("key2"))  # 输出: None
    await cache.set("key3", "value3")
    print(await cache.get("key3"))  # 输出: value3

asyncio.run(main())

地点是北京。

业务就是多模态文档，做一些决策，会用到VLM Qwen-vl gpt-4o等模型。