Instructions to use HuggingFaceM4/tiny-random-idefics with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use HuggingFaceM4/tiny-random-idefics with Transformers:

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

pipe = pipeline("image-text-to-text", model="HuggingFaceM4/tiny-random-idefics")

# Load model directly
from transformers import AutoProcessor, AutoModelForImageTextToText

processor = AutoProcessor.from_pretrained("HuggingFaceM4/tiny-random-idefics")
model = AutoModelForImageTextToText.from_pretrained("HuggingFaceM4/tiny-random-idefics")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use HuggingFaceM4/tiny-random-idefics with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "HuggingFaceM4/tiny-random-idefics"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "HuggingFaceM4/tiny-random-idefics",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/HuggingFaceM4/tiny-random-idefics

SGLang

How to use HuggingFaceM4/tiny-random-idefics 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 "HuggingFaceM4/tiny-random-idefics" \
    --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": "HuggingFaceM4/tiny-random-idefics",
		"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 "HuggingFaceM4/tiny-random-idefics" \
        --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": "HuggingFaceM4/tiny-random-idefics",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use HuggingFaceM4/tiny-random-idefics with Docker Model Runner:
```
docker model run hf.co/HuggingFaceM4/tiny-random-idefics
```

tiny-random-idefics / config.json

Leyo

Update config.json

2513f14 over 2 years ago

raw

history blame

1.47 kB

	{
	"additional_vocab_size": 2,
	"alpha_initializer": "ones",
	"alpha_type": "vector",
	"alphas_initializer_range": 0.0,
	"architectures": [
	"IdeficsForVisionText2Text"
	],
	"bos_token_id": 1,
	"cross_layer_activation_function": "swiglu",
	"cross_layer_interval": 1,
	"dropout": 0.0,
	"eos_token_id": 2,
	"ffn_dim": 64,
	"freeze_lm_head": false,
	"freeze_text_layers": false,
	"freeze_text_module_exceptions": [],
	"freeze_vision_layers": false,
	"freeze_vision_module_exceptions": [],
	"hidden_act": "silu",
	"hidden_size": 16,
	"initializer_range": 0.02,
	"intermediate_size": 11008,
	"max_new_tokens": 128,
	"max_position_embeddings": 128,
	"model_type": "idefics",
	"num_attention_heads": 4,
	"num_hidden_layers": 2,
	"pad_token_id": 0,
	"qk_layer_norms": false,
	"rms_norm_eps": 1e-06,
	"tie_word_embeddings": false,
	"torch_dtype": "float16",
	"transformers_version": "4.27.0.dev0",
	"use_cache": true,
	"use_resampler": true,
	"vocab_size": 32000,
	"word_embed_proj_dim": 16,
	"vision_config": {
	"hidden_act": "gelu",
	"embed_dim": 32,
	"image_size": 30,
	"intermediate_size": 37,
	"patch_size": 2,
	"num_attention_heads": 4,
	"num_hidden_layers": 5,
	"vision_model_name": "hf-internal-testing/tiny-random-clip"
	},
	"perceiver_config": {
	"qk_layer_norms_perceiver": false,
	"resampler_depth": 2,
	"resampler_head_dim": 8,
	"resampler_n_heads": 2,
	"resampler_n_latents": 16
	}
	}