mirror of
https://github.com/deepseek-ai/DeepSeek-VL
synced 2025-06-26 18:27:43 +00:00
update
This commit is contained in:
StevenLiuWen
0
deepseek_vl/utils/__init__.py
Normal file
0
deepseek_vl/utils/__init__.py
Normal file
326
deepseek_vl/utils/conversation.py
Normal file
326
deepseek_vl/utils/conversation.py
Normal file
@@ -0,0 +1,326 @@
|
||||
"""
|
||||
From https://github.com/lm-sys/FastChat/blob/main/fastchat/conversation.py
|
||||
"""
|
||||
|
||||
import dataclasses
|
||||
from enum import IntEnum, auto
|
||||
from typing import Dict, List
|
||||
|
||||
|
||||
class SeparatorStyle(IntEnum):
|
||||
"""Separator styles."""
|
||||
|
||||
ADD_COLON_SINGLE = auto()
|
||||
ADD_COLON_TWO = auto()
|
||||
ADD_COLON_SPACE_SINGLE = auto()
|
||||
NO_COLON_SINGLE = auto()
|
||||
NO_COLON_TWO = auto()
|
||||
ADD_NEW_LINE_SINGLE = auto()
|
||||
LLAMA2 = auto()
|
||||
CHATGLM = auto()
|
||||
CHATML = auto()
|
||||
CHATINTERN = auto()
|
||||
DOLLY = auto()
|
||||
RWKV = auto()
|
||||
PHOENIX = auto()
|
||||
ROBIN = auto()
|
||||
DeepSeek = auto()
|
||||
PLAIN = auto()
|
||||
ALIGNMENT = auto()
|
||||
|
||||
|
||||
@dataclasses.dataclass
|
||||
class Conversation:
|
||||
"""A class that manages prompt templates and keeps all conversation history."""
|
||||
|
||||
# The name of this template
|
||||
name: str
|
||||
# The template of the system prompt
|
||||
system_template: str = "{system_message}"
|
||||
# The system message
|
||||
system_message: str = ""
|
||||
# The names of two roles
|
||||
roles: List[str] = (("USER", "ASSISTANT"),)
|
||||
# All messages. Each item is (role, message).
|
||||
messages: List[List[str]] = ()
|
||||
# The number of few shot examples
|
||||
offset: int = 0
|
||||
# The separator style and configurations
|
||||
sep_style: SeparatorStyle = SeparatorStyle.ADD_COLON_SINGLE
|
||||
sep: str = "\n"
|
||||
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."""
|
||||
system_prompt = self.system_template.format(system_message=self.system_message)
|
||||
|
||||
if self.sep_style == SeparatorStyle.DeepSeek:
|
||||
seps = [self.sep, self.sep2]
|
||||
if system_prompt == "" or system_prompt is None:
|
||||
ret = ""
|
||||
else:
|
||||
ret = system_prompt + seps[0]
|
||||
for i, (role, message) in enumerate(self.messages):
|
||||
if message:
|
||||
ret += role + ": " + message + seps[i % 2]
|
||||
else:
|
||||
ret += role + ":"
|
||||
return ret
|
||||
elif self.sep_style == SeparatorStyle.LLAMA2:
|
||||
seps = [self.sep, self.sep2]
|
||||
if self.system_message:
|
||||
ret = system_prompt
|
||||
else:
|
||||
ret = "[INST] "
|
||||
for i, (role, message) in enumerate(self.messages):
|
||||
tag = self.roles[i % 2]
|
||||
if message:
|
||||
if type(message) is tuple: # multimodal message
|
||||
message, _ = message
|
||||
if i == 0:
|
||||
ret += message + " "
|
||||
else:
|
||||
ret += tag + " " + message + seps[i % 2]
|
||||
else:
|
||||
ret += tag
|
||||
return ret
|
||||
elif self.sep_style == SeparatorStyle.PLAIN:
|
||||
seps = [self.sep, self.sep2]
|
||||
ret = ""
|
||||
for i, (role, message) in enumerate(self.messages):
|
||||
if message:
|
||||
if type(message) is tuple:
|
||||
message, _, _ = message
|
||||
if i % 2 == 0:
|
||||
ret += message + seps[i % 2]
|
||||
else:
|
||||
ret += message + seps[i % 2]
|
||||
else:
|
||||
ret += ""
|
||||
return ret
|
||||
elif self.sep_style == SeparatorStyle.ALIGNMENT:
|
||||
seps = [self.sep, self.sep2]
|
||||
ret = ""
|
||||
for i, (role, message) in enumerate(self.messages):
|
||||
if message:
|
||||
if type(message) is tuple:
|
||||
message, _, _ = message
|
||||
if i % 2 == 0:
|
||||
ret += '<image>\n' + seps[i % 2]
|
||||
else:
|
||||
ret += message + seps[i % 2]
|
||||
else:
|
||||
ret += ""
|
||||
return ret
|
||||
else:
|
||||
raise ValueError(f"Invalid style: {self.sep_style}")
|
||||
|
||||
def get_prompt_for_current_round(self, content=None):
|
||||
"""Get current round formatted question prompt during sft training"""
|
||||
if self.sep_style == SeparatorStyle.PLAIN:
|
||||
formatted_question = "<image>\n"
|
||||
elif self.sep_style == SeparatorStyle.DeepSeek:
|
||||
formatted_question = f"{self.roles[0]}: " +content.strip() + self.sep + f"{self.roles[1]}:"
|
||||
else:
|
||||
raise ValueError(f"Unsupported sep_style: {self.sep_style}")
|
||||
return formatted_question
|
||||
|
||||
def set_system_message(self, system_message: str):
|
||||
"""Set the system message."""
|
||||
self.system_message = system_message
|
||||
|
||||
def append_message(self, role: str, message: str):
|
||||
"""Append a new message."""
|
||||
self.messages.append([role, message])
|
||||
|
||||
def reset_message(self):
|
||||
"""Reset a new message."""
|
||||
self.messages = []
|
||||
|
||||
def update_last_message(self, message: str):
|
||||
"""Update the last output.
|
||||
|
||||
The last message is typically set to be None when constructing the prompt,
|
||||
so we need to update it in-place after getting the response from a model.
|
||||
"""
|
||||
self.messages[-1][1] = message
|
||||
|
||||
def to_gradio_chatbot(self):
|
||||
"""Convert the conversation 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."""
|
||||
system_prompt = self.system_template.format(system_message=self.system_message)
|
||||
ret = [{"role": "system", "content": system_prompt}]
|
||||
|
||||
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_template=self.system_template,
|
||||
system_message=self.system_message,
|
||||
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 {
|
||||
"template_name": self.name,
|
||||
"system_message": self.system_message,
|
||||
"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"{template.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()
|
||||
|
||||
|
||||
# llava_llama2 template
|
||||
register_conv_template(
|
||||
Conversation(
|
||||
name="llava_llama2",
|
||||
system_message="You are a helpful language and vision assistant. "
|
||||
"You are able to understand the visual content that the user provides, "
|
||||
"and assist the user with a variety of tasks using natural language.",
|
||||
system_template="[INST] <<SYS>>\n{system_message}\n<</SYS>>\n\n",
|
||||
roles=("[INST]", "[/INST]"),
|
||||
messages=(),
|
||||
offset=0,
|
||||
sep_style=SeparatorStyle.LLAMA2,
|
||||
sep=" ",
|
||||
sep2=" </s><s>",
|
||||
stop_token_ids=[2],
|
||||
)
|
||||
)
|
||||
|
||||
# llama2 template
|
||||
# reference: https://github.com/facebookresearch/llama/blob/cfc3fc8c1968d390eb830e65c63865e980873a06/llama/generation.py#L212
|
||||
register_conv_template(
|
||||
Conversation(
|
||||
name="llama-2",
|
||||
system_template="[INST] <<SYS>>\n{system_message}\n<</SYS>>\n\n",
|
||||
roles=("[INST]", "[/INST]"),
|
||||
messages=(),
|
||||
offset=0,
|
||||
sep_style=SeparatorStyle.LLAMA2,
|
||||
sep=" ",
|
||||
sep2=" </s><s>",
|
||||
stop_token_ids=[2],
|
||||
)
|
||||
)
|
||||
|
||||
|
||||
# deepseek template
|
||||
register_conv_template(
|
||||
Conversation(
|
||||
name="deepseek",
|
||||
system_template="{system_message}",
|
||||
# system_message="You are a helpful assistant. Please answer truthfully and write out your "
|
||||
# "thinking step by step to be sure you get the right answer.",
|
||||
system_message="",
|
||||
roles=("User", "Assistant"),
|
||||
messages=(),
|
||||
offset=0,
|
||||
sep_style=SeparatorStyle.DeepSeek,
|
||||
sep="\n\n",
|
||||
sep2="<｜end▁of▁sentence｜>",
|
||||
stop_token_ids=[100001],
|
||||
stop_str=["User:", "<｜end▁of▁sentence｜>"]
|
||||
)
|
||||
)
|
||||
|
||||
register_conv_template(
|
||||
Conversation(
|
||||
name="plain",
|
||||
system_template="",
|
||||
system_message="",
|
||||
roles=("", ""),
|
||||
messages=(),
|
||||
offset=0,
|
||||
sep_style=SeparatorStyle.PLAIN,
|
||||
sep="",
|
||||
sep2="",
|
||||
stop_token_ids=[2],
|
||||
stop_str=['</s>'],
|
||||
)
|
||||
)
|
||||
|
||||
|
||||
register_conv_template(
|
||||
Conversation(
|
||||
name="alignment",
|
||||
system_template="",
|
||||
system_message="",
|
||||
roles=("", ""),
|
||||
messages=(),
|
||||
offset=0,
|
||||
sep_style=SeparatorStyle.ALIGNMENT,
|
||||
sep="",
|
||||
sep2="",
|
||||
stop_token_ids=[2],
|
||||
stop_str=['</s>'],
|
||||
)
|
||||
)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
|
||||
# print("Llama-2 template:")
|
||||
# conv = get_conv_template("llama-2")
|
||||
# conv.set_system_message("You are a helpful, respectful and honest assistant.")
|
||||
# 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())
|
||||
|
||||
# print("\n")
|
||||
|
||||
print("deepseek template:")
|
||||
conv = get_conv_template("deepseek")
|
||||
conv.append_message(conv.roles[0], "Hello!")
|
||||
conv.append_message(conv.roles[1], "Hi! This is Tony.")
|
||||
conv.append_message(conv.roles[0], "Who are you?")
|
||||
conv.append_message(conv.roles[1], "I am a helpful assistant.")
|
||||
conv.append_message(conv.roles[0], "How are you?")
|
||||
conv.append_message(conv.roles[1], None)
|
||||
print(conv.get_prompt())
|
||||
56
deepseek_vl/utils/io.py
Normal file
56
deepseek_vl/utils/io.py
Normal file
@@ -0,0 +1,56 @@
|
||||
import json
|
||||
import PIL.Image
|
||||
from typing import Dict, List
|
||||
import torch
|
||||
from transformers import AutoModelForCausalLM
|
||||
|
||||
from deepseek_vl.models import VLChatProcessor, MultiModalityCausalLM
|
||||
|
||||
|
||||
def load_pretrained_model(model_path: str):
|
||||
vl_chat_processor: VLChatProcessor = VLChatProcessor.from_pretrained(model_path)
|
||||
tokenizer = vl_chat_processor.tokenizer
|
||||
|
||||
vl_gpt: MultiModalityCausalLM = AutoModelForCausalLM.from_pretrained(model_path, trust_remote_code=True)
|
||||
vl_gpt = vl_gpt.to(torch.bfloat16).cuda().eval()
|
||||
|
||||
return tokenizer, vl_chat_processor, vl_gpt
|
||||
|
||||
|
||||
def load_pil_images(conversations: List[Dict[str, str]]) -> List[PIL.Image.Image]:
|
||||
"""
|
||||
|
||||
Args:
|
||||
conversations (List[Dict[str, str]]): the conversations with a list of messages. An example is :
|
||||
[
|
||||
{
|
||||
"role": "User",
|
||||
"content": "<image_placeholder>\nExtract all information from this image and convert them into markdown format.",
|
||||
"images": ["./examples/table_datasets.png"]
|
||||
},
|
||||
{"role": "Assistant", "content": ""},
|
||||
]
|
||||
|
||||
Returns:
|
||||
pil_images (List[PIL.Image.Image]): the list of PIL images.
|
||||
|
||||
"""
|
||||
|
||||
pil_images = []
|
||||
|
||||
for message in conversations:
|
||||
if "images" not in message:
|
||||
continue
|
||||
|
||||
for image_path in message["images"]:
|
||||
pil_img = PIL.Image.open(image_path)
|
||||
pil_img = pil_img.convert("RGB")
|
||||
pil_images.append(pil_img)
|
||||
|
||||
return pil_images
|
||||
|
||||
|
||||
def load_json(filepath):
|
||||
with open(filepath, "r") as f:
|
||||
data = json.load(f)
|
||||
return data
|
||||
Reference in New Issue
Block a user