Instructions to use hari02/llava_report_4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use hari02/llava_report_4bit with Transformers:

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

pipe = pipeline("image-text-to-text", model="hari02/llava_report_4bit")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("hari02/llava_report_4bit")
model = AutoModelForMultimodalLM.from_pretrained("hari02/llava_report_4bit", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use hari02/llava_report_4bit with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "hari02/llava_report_4bit"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "hari02/llava_report_4bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/hari02/llava_report_4bit

SGLang

How to use hari02/llava_report_4bit 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 "hari02/llava_report_4bit" \
    --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": "hari02/llava_report_4bit",
		"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 "hari02/llava_report_4bit" \
        --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": "hari02/llava_report_4bit",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use hari02/llava_report_4bit with Docker Model Runner:
```
docker model run hf.co/hari02/llava_report_4bit
```

llava_report_4bit / config.json

hari02

Upload LlavaForConditionalGeneration

9fb18f5 verified over 2 years ago

Raw History Blame Contribute Delete

1.45 kB

	{
	"_name_or_path": "llava-hf/llava-1.5-7b-hf",
	"architectures": [
	"LlavaForConditionalGeneration"
	],
	"ignore_index": -100,
	"image_token_index": 32000,
	"model_type": "llava",
	"pad_token_id": 32001,
	"projector_hidden_act": "gelu",
	"quantization_config": {
	"_load_in_4bit": true,
	"_load_in_8bit": false,
	"bnb_4bit_compute_dtype": "float32",
	"bnb_4bit_quant_storage": "uint8",
	"bnb_4bit_quant_type": "fp4",
	"bnb_4bit_use_double_quant": false,
	"llm_int8_enable_fp32_cpu_offload": false,
	"llm_int8_has_fp16_weight": false,
	"llm_int8_skip_modules": null,
	"llm_int8_threshold": 6.0,
	"load_in_4bit": true,
	"load_in_8bit": false,
	"quant_method": "bitsandbytes"
	},
	"text_config": {
	"_name_or_path": "lmsys/vicuna-7b-v1.5",
	"architectures": [
	"LlamaForCausalLM"
	],
	"max_position_embeddings": 4096,
	"model_type": "llama",
	"rms_norm_eps": 1e-05,
	"torch_dtype": "float16",
	"vocab_size": 32064
	},
	"tie_word_embeddings": false,
	"torch_dtype": "float32",
	"transformers_version": "4.41.2",
	"vision_config": {
	"hidden_size": 1024,
	"image_size": 336,
	"intermediate_size": 4096,
	"model_type": "clip_vision_model",
	"num_attention_heads": 16,
	"num_hidden_layers": 24,
	"patch_size": 14,
	"projection_dim": 768,
	"vocab_size": 32000
	},
	"vision_feature_layer": -2,
	"vision_feature_select_strategy": "default"
	}