Update to new GPT 3.5-Turbo type

[tyu1996]

From our existing ChatDev typing.py:

class ModelType(Enum):
    GPT_3_5_TURBO = "gpt-3.5-turbo-16k-0613"
    GPT_3_5_TURBO_NEW = "gpt-3.5-turbo-16k"
    GPT_4 = "gpt-4"
    GPT_4_32k = "gpt-4-32k"
    GPT_4_TURBO = "gpt-4-turbo"
    GPT_4_TURBO_V = "gpt-4-turbo"

    STUB = "stub"

    @property
    def value_for_tiktoken(self):
        return self.value if self.name != "STUB" else "gpt-3.5-turbo-16k-0613"

Where in official OpenAI's models page:

The current GPT 3.5-Turbo types (gpt-3.5-turbo-16k-0613 & gpt-3.5-turbo-16k) used in ChatDev is considered legacy and will be deprecated very soon (next week).

Advise to update the types to remain updated to use OpenAI's latest GPT 3.5-Turbo, by just using the value "gpt-3.5-turbo" should be sufficient (which is already supports 16k context window by default).

[tyu1996]

However, I suggest do some testing before pushing this changes. In my local repo I modified it to use "gpt-3.5-turbo", ChatDev seems underperforming. More observation from different users is required to prove the latest "gpt-3.5-turbo" type works well with ChatDev.