Instructions to use qgallouedec/idefics3-8B-tiny with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use qgallouedec/idefics3-8B-tiny with Transformers:

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

pipe = pipeline("image-text-to-text", model="qgallouedec/idefics3-8B-tiny")

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("qgallouedec/idefics3-8B-tiny")
model = AutoModelForMultimodalLM.from_pretrained("qgallouedec/idefics3-8B-tiny")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use qgallouedec/idefics3-8B-tiny with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "qgallouedec/idefics3-8B-tiny"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "qgallouedec/idefics3-8B-tiny",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/qgallouedec/idefics3-8B-tiny

SGLang

How to use qgallouedec/idefics3-8B-tiny 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 "qgallouedec/idefics3-8B-tiny" \
    --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": "qgallouedec/idefics3-8B-tiny",
		"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 "qgallouedec/idefics3-8B-tiny" \
        --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": "qgallouedec/idefics3-8B-tiny",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use qgallouedec/idefics3-8B-tiny with Docker Model Runner:
```
docker model run hf.co/qgallouedec/idefics3-8B-tiny
```

qgallouedec HF Staff commited on Aug 13, 2025

Commit

7148e37

verified ·

1 Parent(s): 4af9483

Add tiny Idefics3 model

Browse files

Files changed (2) hide show

config.json +4 -3
model.safetensors +2 -2

config.json CHANGED Viewed

@@ -33,9 +33,9 @@
     "mlp_bias": false,
     "model_type": "llama",
     "neftune_noise_alpha": 0.0,
-    "num_attention_heads": 32,
     "num_hidden_layers": 2,
-    "num_key_value_heads": 8,
     "pad_token_id": 128002,
     "perceiver_config": {
       "_name_or_path": "",
@@ -138,9 +138,10 @@
     "intermediate_size": 4304,
     "layer_norm_eps": 1e-06,
     "model_type": "idefics3_vision",
-    "num_attention_heads": 16,
     "num_channels": 3,
     "num_hidden_layers": 2,
     "old_vision_model_name": "/fsx/hugo/siglip-so400m-14-364-flash-attn2-navit",
     "pad_token_id": 128002,
     "patch_size": 14

     "mlp_bias": false,
     "model_type": "llama",
     "neftune_noise_alpha": 0.0,
+    "num_attention_heads": 4,
     "num_hidden_layers": 2,
+    "num_key_value_heads": 2,
     "pad_token_id": 128002,
     "perceiver_config": {
       "_name_or_path": "",
     "intermediate_size": 4304,
     "layer_norm_eps": 1e-06,
     "model_type": "idefics3_vision",
+    "num_attention_heads": 4,
     "num_channels": 3,
     "num_hidden_layers": 2,
+    "num_key_value_heads": 2,
     "old_vision_model_name": "/fsx/hugo/siglip-so400m-14-364-flash-attn2-navit",
     "pad_token_id": 128002,
     "patch_size": 14

model.safetensors CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ba81671ed3aa79e3cb33a783bbef1fec1de64f88885816f2c271a2e54ff1ed77
-size 12239160

 version https://git-lfs.github.com/spec/v1
+oid sha256:f5894593ad1d7885be43ab7c1166017e43452eb90368846f4feeceb7e574ddb8
+size 11682096