Instructions to use microsoft/Florence-2-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use microsoft/Florence-2-base with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="microsoft/Florence-2-base", trust_remote_code=True)

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("microsoft/Florence-2-base", trust_remote_code=True)
model = AutoModelForImageTextToText.from_pretrained("microsoft/Florence-2-base", trust_remote_code=True)

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use microsoft/Florence-2-base with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "microsoft/Florence-2-base"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "microsoft/Florence-2-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/microsoft/Florence-2-base

SGLang

How to use microsoft/Florence-2-base with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "microsoft/Florence-2-base" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "microsoft/Florence-2-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "microsoft/Florence-2-base" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "microsoft/Florence-2-base",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use microsoft/Florence-2-base with Docker Model Runner:
```
docker model run hf.co/microsoft/Florence-2-base
```

Matt commited on Apr 24, 2025

Commit

1e8c51f

1 Parent(s): bef74d7

Tie weights correctly and update weights

Browse files

Files changed (2) hide show

model.safetensors +2 -2
modeling_florence2.py +6 -0

model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0931732cefbda398920d6a13bfc7eefc6c457a1463a17c7f4930b68b74b46d5
-size 542001482

 version https://git-lfs.github.com/spec/v1
+oid sha256:43c6d08c6afba4b18c373b5c0f94be59550c84ffe17993c1d3c2fe06c46f94ea
+size 463221938

modeling_florence2.py CHANGED Viewed

@@ -2066,6 +2066,12 @@ class Florence2LanguageForConditionalGeneration(Florence2LanguagePreTrainedModel
         # Initialize weights and apply final processing
         self.post_init()
     def get_encoder(self):
         return self.model.get_encoder()

         # Initialize weights and apply final processing
         self.post_init()
+    def _tie_weights(self):
+        if self.config.tie_word_embeddings:
+            self._tie_or_clone_weights(self.model.encoder.embed_tokens, self.model.shared)
+            self._tie_or_clone_weights(self.model.decoder.embed_tokens, self.model.shared)
+            self._tie_or_clone_weights(self.lm_head, self.model.shared)
     def get_encoder(self):
         return self.model.get_encoder()