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

Libraries

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

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

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

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("microsoft/Florence-2-large", trust_remote_code=True)
model = AutoModelForMultimodalLM.from_pretrained("microsoft/Florence-2-large", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

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

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "microsoft/Florence-2-large"
# 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-large",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

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

SGLang

How to use microsoft/Florence-2-large 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-large" \
    --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-large",
		"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-large" \
        --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-large",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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

masterofaudio2077 commited on 4 days ago

Commit

9cb81b6

verified ·

1 Parent(s): c13af62

Fix TypeError: EncoderDecoderCache is not subscriptable in prepare_inputs_for_generation

Browse files

prepare_inputs_for_generation (both Florence2LanguageForConditionalGeneration and Florence2ForConditionalGeneration -- identical code in both) read past_length via past_key_values[0][0].shape[2], assuming the legacy tuple-of-tuples KV cache format. Current transformers passes a Cache object (EncoderDecoderCache for encoder-decoder models like this one), which isn't subscriptable. Every Cache subclass implements get_seq_length() for exactly this purpose -- use it when available, falling back to the legacy tuple indexing for older transformers/custom cache objects that predate the Cache class.

Files changed (1) hide show

modeling_florence2.py +14 -2

modeling_florence2.py CHANGED Viewed

@@ -2194,7 +2194,13 @@ class Florence2LanguageForConditionalGeneration(Florence2LanguagePreTrainedModel
     ):
         # cut decoder_input_ids if past_key_values is used
         if past_key_values is not None:
-            past_length = past_key_values[0][0].shape[2]
             # Some generation methods already pass only the last input ID
             if decoder_input_ids.shape[1] > past_length:
@@ -2823,7 +2829,13 @@ class Florence2ForConditionalGeneration(Florence2PreTrainedModel):
     ):
         # cut decoder_input_ids if past_key_values is used
         if past_key_values is not None:
-            past_length = past_key_values[0][0].shape[2]
             # Some generation methods already pass only the last input ID
             if decoder_input_ids.shape[1] > past_length:

     ):
         # cut decoder_input_ids if past_key_values is used
         if past_key_values is not None:
+            # past_key_values is a Cache object (e.g. EncoderDecoderCache) in current
+            # transformers, not the legacy tuple-of-tuples this code was written for --
+            # it isn't subscriptable, but every Cache subclass implements get_seq_length()
+            if hasattr(past_key_values, "get_seq_length"):
+                past_length = past_key_values.get_seq_length()
+            else:
+                past_length = past_key_values[0][0].shape[2]
             # Some generation methods already pass only the last input ID
             if decoder_input_ids.shape[1] > past_length:
     ):
         # cut decoder_input_ids if past_key_values is used
         if past_key_values is not None:
+            # past_key_values is a Cache object (e.g. EncoderDecoderCache) in current
+            # transformers, not the legacy tuple-of-tuples this code was written for --
+            # it isn't subscriptable, but every Cache subclass implements get_seq_length()
+            if hasattr(past_key_values, "get_seq_length"):
+                past_length = past_key_values.get_seq_length()
+            else:
+                past_length = past_key_values[0][0].shape[2]
             # Some generation methods already pass only the last input ID
             if decoder_input_ids.shape[1] > past_length: