[Feature] add tool role in BaseChatTemplate as tool response in messages

[Huarong]

lmdeploy/lmdeploy/model.py

Lines 224 to 237 in 9c16443

	box_map = dict(user=self.user,
	assistant=self.assistant,
	system=self.system)
	eox_map = dict(user=self.eoh,
	assistant=self.eoa + self.separator,
	system=self.eosys)
	ret = ''
	if self.meta_instruction is not None and sequence_start:
	if len(messages) and messages[0]['role'] != 'system':
	ret += f'{self.system}{self.meta_instruction}{self.eosys}'
	for message in messages:
	role = message['role']
	content = get_text(message['content'])
	ret += f'{box_map[role]}{content}{eox_map[role]}'

request payload to /chat/completions endpoint

{
    "model": "xxxx",
    "messages": [
        {
            "role": "system",
            "content": "....”"
        },
        {
            "role": "user",
            "content": "吃了阿司匹林有什么副作用？"
        },
        {
            "role": "assistant",
            "content": "....."
        },
        {
            "role": "tool",
            "content": "..."
        }
    ],
    "max_tokens": 512,
    "stream": false,
    "temperature": 0.2,
    "top_p": 0.9,
    "user": "user-1234"
}

current error:

  File "/opt/lmdeploy/lmdeploy/serve/openai/api_server.py", line 483, in chat_completions_v1
    async for res in result_generator:
  File "/opt/lmdeploy/lmdeploy/serve/async_engine.py", line 524, in generate
    prompt_input = await self._get_prompt_input(prompt,
  File "/opt/lmdeploy/lmdeploy/serve/async_engine.py", line 455, in _get_prompt_input
    prompt = chat_template.messages2prompt(prompt,
  File "/opt/lmdeploy/lmdeploy/model.py", line 237, in messages2prompt
    ret += f'{box_map[role]}{content}{eox_map[role]}'
KeyError: 'tool'

The toolrole is in the OpenAI API document.