297 lines
11 KiB
Python
297 lines
11 KiB
Python
"""
|
|
Tool conversation prompt templates. Basically copy from FastChat.
|
|
"""
|
|
|
|
import dataclasses
|
|
from enum import auto, Enum
|
|
from typing import List, Any, Dict
|
|
|
|
|
|
class SeparatorStyle(Enum):
|
|
"""Separator styles."""
|
|
|
|
ADD_COLON_SINGLE = auto()
|
|
ADD_COLON_TWO = auto()
|
|
ADD_COLON_SPACE_SINGLE = auto()
|
|
NO_COLON_SINGLE = auto()
|
|
ADD_NEW_LINE_SINGLE = auto()
|
|
DOLLY = auto()
|
|
RWKV = auto()
|
|
PHOENIX = auto()
|
|
ONLY_LAST_ASSISTANT = auto()
|
|
|
|
|
|
@dataclasses.dataclass
|
|
class Conversation:
|
|
"""A class that keeps all conversation history."""
|
|
|
|
# The name of this template
|
|
name: str
|
|
# The System prompt
|
|
system: str
|
|
# Two roles
|
|
roles: List[str]
|
|
# All messages
|
|
messages: List[List[str]]
|
|
# Offset of few shot examples
|
|
offset: int
|
|
# Separators
|
|
sep_style: SeparatorStyle
|
|
sep: str
|
|
sep2: str = None
|
|
# Stop criteria (the default one is EOS token)
|
|
stop_str: str = None
|
|
# Stops generation if meeting any token in this list
|
|
stop_token_ids: List[int] = None
|
|
|
|
def get_prompt(self) -> str:
|
|
"""Get the prompt for generation."""
|
|
if self.sep_style == SeparatorStyle.ONLY_LAST_ASSISTANT:
|
|
seps = [self.sep, self.sep2]
|
|
ret = ""
|
|
for i, (role, message) in enumerate(self.messages):
|
|
if i + 1 == len(self.messages) and message:
|
|
ret += role + ": " + str(message) + seps[1]
|
|
elif message:
|
|
ret += role + ": " + str(message) + seps[0]
|
|
else:
|
|
ret += role + ":"
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.ADD_COLON_SINGLE:
|
|
ret = self.system + self.sep
|
|
for role, message in self.messages:
|
|
if message:
|
|
ret += role + ": " + message + self.sep
|
|
else:
|
|
ret += role + ":"
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.ADD_COLON_TWO:
|
|
seps = [self.sep, self.sep2]
|
|
ret = self.system + seps[0]
|
|
for i, (role, message) in enumerate(self.messages):
|
|
try:
|
|
if message:
|
|
ret += role + ": " + message + seps[i % 2]
|
|
else:
|
|
ret += role + ":"
|
|
except:
|
|
continue
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.ADD_COLON_SPACE_SINGLE:
|
|
ret = self.system + self.sep
|
|
for role, message in self.messages:
|
|
if message:
|
|
ret += role + ": " + message + self.sep
|
|
else:
|
|
ret += role + ": " # must be end with a space
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.NO_COLON_SINGLE:
|
|
ret = self.system
|
|
for role, message in self.messages:
|
|
if message:
|
|
ret += role + message + self.sep
|
|
else:
|
|
ret += role
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.ADD_NEW_LINE_SINGLE:
|
|
ret = self.system + self.sep
|
|
for role, message in self.messages:
|
|
if message:
|
|
ret += role + "\n" + message + self.sep
|
|
else:
|
|
ret += role + "\n"
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.DOLLY:
|
|
seps = [self.sep, self.sep2]
|
|
ret = self.system
|
|
for i, (role, message) in enumerate(self.messages):
|
|
if message:
|
|
ret += role + ":\n" + message + seps[i % 2]
|
|
if i % 2 == 1:
|
|
ret += "\n\n"
|
|
else:
|
|
ret += role + ":\n"
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.RWKV:
|
|
ret = self.system
|
|
for i, (role, message) in enumerate(self.messages):
|
|
if message:
|
|
ret += (
|
|
role
|
|
+ ": "
|
|
+ message.replace("\r\n", "\n").replace("\n\n", "\n")
|
|
)
|
|
ret += "\n\n"
|
|
else:
|
|
ret += role + ":"
|
|
return ret
|
|
elif self.sep_style == SeparatorStyle.PHOENIX:
|
|
ret = self.system
|
|
for role, message in self.messages:
|
|
if message:
|
|
ret += role + ": " + "<s>" + message + "</s>"
|
|
else:
|
|
ret += role + ": " + "<s>"
|
|
return ret
|
|
else:
|
|
raise ValueError(f"Invalid style: {self.sep_style}")
|
|
|
|
def append_message(self, role: str, message: str):
|
|
"""Append a new message."""
|
|
self.messages.append([role, message])
|
|
|
|
def to_gradio_chatbot(self):
|
|
"""Convert the history to gradio chatbot format"""
|
|
ret = []
|
|
for i, (role, msg) in enumerate(self.messages[self.offset :]):
|
|
if i % 2 == 0:
|
|
ret.append([msg, None])
|
|
else:
|
|
ret[-1][-1] = msg
|
|
return ret
|
|
|
|
def to_openai_api_messages(self):
|
|
"""Convert the conversation to OpenAI chat completion format."""
|
|
ret = [{"role": "system", "content": self.system}]
|
|
|
|
for i, (_, msg) in enumerate(self.messages[self.offset :]):
|
|
if i % 2 == 0:
|
|
ret.append({"role": "user", "content": msg})
|
|
else:
|
|
if msg is not None:
|
|
ret.append({"role": "assistant", "content": msg})
|
|
return ret
|
|
|
|
def copy(self):
|
|
return Conversation(
|
|
name=self.name,
|
|
system=self.system,
|
|
roles=self.roles,
|
|
messages=[[x, y] for x, y in self.messages],
|
|
offset=self.offset,
|
|
sep_style=self.sep_style,
|
|
sep=self.sep,
|
|
sep2=self.sep2,
|
|
stop_str=self.stop_str,
|
|
stop_token_ids=self.stop_token_ids,
|
|
)
|
|
|
|
def dict(self):
|
|
return {
|
|
"name": self.name,
|
|
"system": self.system,
|
|
"roles": self.roles,
|
|
"messages": self.messages,
|
|
"offset": self.offset,
|
|
}
|
|
|
|
|
|
# A global registry for all conversation templates
|
|
conv_templates: Dict[str, Conversation] = {}
|
|
|
|
|
|
def register_conv_template(template: Conversation, override: bool = False):
|
|
"""Register a new conversation template."""
|
|
if not override:
|
|
assert template.name not in conv_templates, f"{name} has been registered."
|
|
conv_templates[template.name] = template
|
|
|
|
|
|
def get_conv_template(name: str) -> Conversation:
|
|
"""Get a conversation template."""
|
|
return conv_templates[name].copy()
|
|
|
|
|
|
# A template with one conversation example
|
|
register_conv_template(
|
|
Conversation(
|
|
name="one_shot",
|
|
system="A chat between a curious human and an artificial intelligence assistant. "
|
|
"The assistant gives helpful, detailed, and polite answers to the human's questions.",
|
|
roles=("Human", "Assistant"),
|
|
messages=(
|
|
(
|
|
"Human",
|
|
"What are the key differences between renewable and non-renewable energy sources?",
|
|
),
|
|
(
|
|
"Assistant",
|
|
"Renewable energy sources are those that can be replenished naturally in a relatively "
|
|
"short amount of time, such as solar, wind, hydro, geothermal, and biomass. "
|
|
"Non-renewable energy sources, on the other hand, are finite and will eventually be "
|
|
"depleted, such as coal, oil, and natural gas. Here are some key differences between "
|
|
"renewable and non-renewable energy sources:\n"
|
|
"1. Availability: Renewable energy sources are virtually inexhaustible, while non-renewable "
|
|
"energy sources are finite and will eventually run out.\n"
|
|
"2. Environmental impact: Renewable energy sources have a much lower environmental impact "
|
|
"than non-renewable sources, which can lead to air and water pollution, greenhouse gas emissions, "
|
|
"and other negative effects.\n"
|
|
"3. Cost: Renewable energy sources can be more expensive to initially set up, but they typically "
|
|
"have lower operational costs than non-renewable sources.\n"
|
|
"4. Reliability: Renewable energy sources are often more reliable and can be used in more remote "
|
|
"locations than non-renewable sources.\n"
|
|
"5. Flexibility: Renewable energy sources are often more flexible and can be adapted to different "
|
|
"situations and needs, while non-renewable sources are more rigid and inflexible.\n"
|
|
"6. Sustainability: Renewable energy sources are more sustainable over the long term, while "
|
|
"non-renewable sources are not, and their depletion can lead to economic and social instability.",
|
|
),
|
|
),
|
|
offset=2,
|
|
sep_style=SeparatorStyle.ADD_COLON_SINGLE,
|
|
sep="\n### ",
|
|
stop_str="###",
|
|
)
|
|
)
|
|
|
|
# Vicuna v1.1 template
|
|
register_conv_template(
|
|
Conversation(
|
|
name="vicuna-v1.1",
|
|
system="A chat between a curious user and an artificial intelligence assistant. "
|
|
"The assistant gives helpful, detailed, and polite answers to the user's questions.",
|
|
roles=("USER", "ASSISTANT"),
|
|
messages=(),
|
|
offset=0,
|
|
sep_style=SeparatorStyle.ADD_COLON_TWO,
|
|
sep=" ",
|
|
sep2="</s>",
|
|
)
|
|
)
|
|
|
|
# tool-llama template
|
|
register_conv_template(
|
|
Conversation(
|
|
name="tool-llama",
|
|
system="A chat between a curious user and an artificial intelligence assistant who can use external tools and APIs to solve the user's question. "
|
|
"The assistant gives tools and APIs calling processes or final answer to the human's question.",
|
|
roles=("Human", "Assistant"),
|
|
messages=(),
|
|
offset=0,
|
|
sep_style=SeparatorStyle.ADD_COLON_TWO,
|
|
sep=" ",
|
|
sep2="</s>",
|
|
)
|
|
)
|
|
|
|
# tool_llama_v2 with openai function template
|
|
register_conv_template(
|
|
Conversation(
|
|
name="tool-llama-single-round",
|
|
system="", # We put the system message in the specific SFT data. Remember to use the same system message in inference.
|
|
roles=("System", "User", "Function", "Assistant"),
|
|
messages=(),
|
|
offset=0,
|
|
sep_style=SeparatorStyle.ONLY_LAST_ASSISTANT,
|
|
sep="\n",
|
|
sep2="</s>",
|
|
)
|
|
)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
conv = get_conv_template("vicuna_v1.1")
|
|
conv.append_message(conv.roles[0], "Hello!")
|
|
conv.append_message(conv.roles[1], "Hi!")
|
|
conv.append_message(conv.roles[0], "How are you?")
|
|
conv.append_message(conv.roles[1], None)
|
|
print(conv.get_prompt())
|