Instructions to use htlou/mm-interp-AA_text_image_to_text-llava-mistral with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use htlou/mm-interp-AA_text_image_to_text-llava-mistral with Transformers:

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

pipe = pipeline("image-text-to-text", model="htlou/mm-interp-AA_text_image_to_text-llava-mistral")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("htlou/mm-interp-AA_text_image_to_text-llava-mistral")
model = AutoModelForMultimodalLM.from_pretrained("htlou/mm-interp-AA_text_image_to_text-llava-mistral", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use htlou/mm-interp-AA_text_image_to_text-llava-mistral with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "htlou/mm-interp-AA_text_image_to_text-llava-mistral"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "htlou/mm-interp-AA_text_image_to_text-llava-mistral",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/htlou/mm-interp-AA_text_image_to_text-llava-mistral

SGLang

How to use htlou/mm-interp-AA_text_image_to_text-llava-mistral 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 "htlou/mm-interp-AA_text_image_to_text-llava-mistral" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "htlou/mm-interp-AA_text_image_to_text-llava-mistral",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "htlou/mm-interp-AA_text_image_to_text-llava-mistral" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "htlou/mm-interp-AA_text_image_to_text-llava-mistral",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use htlou/mm-interp-AA_text_image_to_text-llava-mistral with Docker Model Runner:
```
docker model run hf.co/htlou/mm-interp-AA_text_image_to_text-llava-mistral
```

AA_text_image_to_text

This model is a fine-tuned version of llava-hf/llava-v1.6-mistral-7b-hf on the AA_text_image_to_text dataset. It achieves the following results on the evaluation set:

Loss: 0.4527
Rewards/chosen: -0.6857
Rewards/rejected: -4.3940
Rewards/accuracies: 0.8165
Rewards/margins: 3.7083
Logps/rejected: -242.1480
Logps/chosen: -207.1762
Logits/rejected: -2.3240
Logits/chosen: -2.3485

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 1e-06
train_batch_size: 8
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
num_devices: 8
gradient_accumulation_steps: 4
total_train_batch_size: 256
total_eval_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 10
num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Validation Loss	Rewards/chosen	Rewards/rejected	Rewards/accuracies	Rewards/margins	Logps/rejected	Logps/chosen	Logits/rejected	Logits/chosen
0.4889	0.2899	40	0.4642	1.1544	-0.1887	0.7944	1.3431	-200.0950	-188.7752	-1.9876	-2.0351
0.3941	0.5797	80	0.4218	-0.2275	-2.2919	0.8044	2.0644	-221.1273	-202.5944	-1.9449	-1.9901
0.3717	0.8696	120	0.4387	-0.2101	-2.4885	0.8286	2.2784	-223.0936	-202.4208	-2.0902	-2.1229
0.1459	1.1594	160	0.4288	-0.4029	-3.3928	0.8286	2.9899	-232.1363	-204.3488	-2.2733	-2.3007
0.1455	1.4493	200	0.4255	-0.5338	-3.6331	0.8165	3.0992	-234.5387	-205.6577	-2.2466	-2.2697
0.1358	1.7391	240	0.4247	-0.2714	-3.6715	0.8327	3.4001	-234.9227	-203.0333	-2.3605	-2.3806
0.0938	2.0290	280	0.4128	-0.3136	-3.7007	0.8266	3.3870	-235.2147	-203.4556	-2.3725	-2.3933
0.0592	2.3188	320	0.4438	-0.5767	-4.1235	0.8165	3.5467	-239.4429	-206.0869	-2.3109	-2.3358
0.0673	2.6087	360	0.4553	-0.6264	-4.3005	0.8206	3.6740	-241.2126	-206.5837	-2.3254	-2.3497
0.0728	2.8986	400	0.4520	-0.6855	-4.3942	0.8185	3.7087	-242.1503	-207.1744	-2.3247	-2.3492

Framework versions

Transformers 4.45.2
Pytorch 2.4.0+cu121
Datasets 2.21.0
Tokenizers 0.20.3

Downloads last month: 5

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for htlou/mm-interp-AA_text_image_to_text-llava-mistral

Base model

llava-hf/llava-v1.6-mistral-7b-hf

Finetuned

(189)

this model