20231088/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[VLM] Florence-2 supports online serving (#16164)

Browse Source 
Signed-off-by: Isotr0py <2037008807@qq.com>
This commit is contained in:
Isotr0py 2025-04-07 19:04:02 +08:00 committed by GitHub
parent 95d63f38c0
commit 7c80368710
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
3 changed files with 22 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
examples/template_florence2.jinja Normal file
View File

@ -0,0 +1,7 @@
{%- for message in messages -%}
{%- if message['role'] == 'user' -%}
{{- message['content'] -}}
{%- elif message['role'] == 'assistant' -%}
{{- message['content'] -}}
{%- endif -%}
{%- endfor -%}

4
vllm/entrypoints/chat_utils.py
View File

@ -487,8 +487,8 @@ class BaseMultiModalItemTracker(ABC, Generic[_T]):
return "<|endoftext10|>" # 200010 (see vocab.json in hf model)
if model_type in ("minicpmo", "minicpmv"):
return "(<image>./</image>)"
if model_type in ("blip-2", "fuyu", "paligemma", "pixtral",
"mistral3"):
if model_type in ("blip-2", "florence2", "fuyu", "paligemma",
"pixtral", "mistral3"):
# These models do not use image tokens in the prompt
return None
if model_type == "qwen":

14
vllm/model_executor/models/florence2.py
View File

@ -10,7 +10,7 @@ import torch
import torch.nn as nn
import torch.nn.functional as F
from einops import rearrange
from transformers import BatchFeature, PretrainedConfig
from transformers import BartTokenizer, BatchFeature, PretrainedConfig
from vllm.config import VllmConfig
from vllm.model_executor.layers.logits_processor import LogitsProcessor
@ -826,6 +826,18 @@ class Florence2MultiModalProcessor(
) -> Union[str, list[int]]:
return [self.info.get_hf_config().eos_token_id]
def _apply_hf_processor_tokens_only(
self,
prompt_tokens: list[int],
) -> list[int]:
hf_processor = self.info.get_hf_processor()
tokenizer: BartTokenizer = hf_processor.tokenizer
prompt_text = tokenizer.decode(prompt_tokens)
# convert task tokens to prompt
prompt_text = hf_processor._construct_prompts([prompt_text])[0]
prompt_tokens = tokenizer.encode(prompt_text, add_special_tokens=False)
return prompt_tokens
def _call_hf_processor(
self,
prompt: str,