README.md

---
language:
- en
license: gemma
license_name: gemma-terms
license_link: https://ai.google.dev/gemma/terms
base_model: google/gemma-2b-it
tags:
- LLM
- MMFM
- Intel
model-index:
- name: llava-gemma-2b
  results:
  - task:
      type: Large Language Model
      name: Large Language Model
    metrics:
    - type: GQA
      name: GQA
      value: 0.531
    - type: MME Cog.
      name: MME Cog.
      value: 236
    - type: MME Per.
      name: MME Per.
      value: 1130
    - type: MM-Vet
      name: MM-Vet
      value: 17.7  
    - type: POPE Acc.
      name: POPE Acc.
      value: 0.850
    - type: POPE F1
      name: POPE F1
      value: 0.839
    - type: VQAv2
      name: VQAv2
      value: 70.7
    - type: MMVP
      name: MMVP
      value: 0.287  
    - type: ScienceQA Image
      name: ScienceQA Image
      value: 0.564
library_name: transformers
pipeline_tag: image-text-to-text
---

## Model Details:  LLaVA-Gemma-2b

`llava-gemma-2b` is a large multimodal model (LMM) trained using the [LLaVA-v1.5 framework](https://arxiv.org/abs/2310.03744) with the 2-billion parameter [google/gemma-2b-it](https://huggingface.co/google/gemma-2b-it) model as language backbone and the CLIP-based vision encoder.

| Model Details | Description |
| ----------- | ----------- | 
| Authors | Intel: [Musashi Hinck](https://huggingface.co/musashihinck), [Matthew Olson](https://huggingface.co/matthewlyleolson), [David Cobbley](https://huggingface.co/djcobble), [Shao-Yen Tseng](https://huggingface.co/shaoyent), [Vasudev Lal](https://huggingface.co/vasudevlal) | 
| Date | March 2024 | 
| Version | 1 | 
| Type | Large multimodal model (LMM) | 
| Paper or Other Resources | [LLaVA-Gemma: Accelerating Multimodal Foundation Models with a Compact Language Model](https://arxiv.org/abs/2404.01331) | 
| License | [Gemma](https://ai.google.dev/gemma/terms) |
| Questions or Comments | [Community Tab](https://huggingface.co/Intel/llava-gemma-2b/discussions) and [Intel DevHub Discord](https://discord.gg/rv2Gp55UJQ)|

This model card was created by [Benjamin Consolvo](https://huggingface.co/bconsolvo) and the authors listed above.

## Intended Use

| Intended Use | Description |
| ----------- | ----------- | 
| Primary intended uses | The model has been finetuned for multimodal benchmark evaluations, but can also be used as a multimodal chatbot. | 
| Primary intended users | Anyone using or evaluating multimodal models. | 
| Out-of-scope uses | This model is not intended for uses that require high levels of factuality, high stakes situations, mental health or medical applications, generating misinformation or disinformation, impersonating others, facilitating or inciting harassment or violence, any use that could lead to the violation of a human right under the UN Declaration of Human Rights. |

### How to use

Currently, using `llava-gemma` requires a [modified preprocessor](./processing_llavagemma.py). _We are currently working on modifying the `LlavaProcessor` class to streamline usage (see [PR #30030](https://github.com/huggingface/transformers/pull/30030)). Expect updates soon._

For current usage, see [`usage.py`](./usage.py) or the following code block:

```python
import requests
from PIL import Image
from transformers import (
  LlavaForConditionalGeneration,
  AutoTokenizer,
  CLIPImageProcessor
)
from processing_llavagemma import LlavaGemmaProcessor # This is in this repo

checkpoint = "Intel/llava-gemma-2b"

# Load model
model = LlavaForConditionalGeneration.from_pretrained(checkpoint)
processor = LlavaGemmaProcessor(
    tokenizer=AutoTokenizer.from_pretrained(checkpoint),
    image_processor=CLIPImageProcessor.from_pretrained(checkpoint)
)

# Prepare inputs
# Use gemma chat template
prompt = processor.tokenizer.apply_chat_template(
    [{'role': 'user', 'content': "<image>\nWhat's the content of the image?"}],
    tokenize=False,
    add_generation_prompt=True
)
url = "https://www.ilankelman.org/stopsigns/australia.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = processor(text=prompt, images=image, return_tensors="pt")

# Generate
generate_ids = model.generate(**inputs, max_length=30)
output = processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
print(output)
```

For straightforward use as a chatbot (without images), you can modify the last portion of code to the following:

```python
# Prepare inputs
# Use gemma chat template
prompt = processor.tokenizer.apply_chat_template(
    [{'role': 'user', 'content': "Summarize the following paragraph? In this paper, we introduced LLaVA-Gemma, a compact vision-language model leveraging the Gemma Large Language Model in two variants, Gemma-2B and Gemma-7B. Our work provides a unique opportunity for researchers to explore the trade-offs between computational efficiency and multimodal understanding in small-scale models. The availability of both variants allows for a comparative analysis that sheds light on how model size impacts performance in various tasks. Our evaluations demonstrate the versatility and effectiveness of LLaVA-Gemma across a range of datasets, highlighting its potential as a benchmark for future research in small-scale vision-language models. With these models, future practitioners can optimize the performance of small-scale multimodal models more directly."}],
    tokenize=False,
    add_generation_prompt=True
)
# url = "https://www.ilankelman.org/stopsigns/australia.jpg"
# image = Image.open(requests.get(url, stream=True).raw)
inputs = processor(text=prompt, images=None, return_tensors="pt")

# Generate
generate_ids = model.generate(**inputs, max_length=300)
output = processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
print(output)
```

## Factors

| Factors | Description | 
| ----------- | ----------- | 
| Groups | - | 
| Instrumentation | - |
| Environment | Trained for 4 hours on 8 Intel Gaudi 2 AI accelerators. |
| Card Prompts | Model training and deployment on alternate hardware and software will change model performance |

## Metrics

| Metrics | Description | 
| ----------- | ----------- | 
| Model performance measures | We evaluate the LlaVA-Gemma models on a similar collection of benchmarks to other LMM works: GQA; MME; MM-Vet; POPE (accuracy and F1); VQAv2; MMVP; the image subset of ScienceQA. Our experiments provide insights into the efficacy of various design choices within the LLaVA framework. |
| Decision thresholds | - | 
| Approaches to uncertainty and variability | - | 

## Training Data

The model was trained using the LLaVA-v1.5 data mixture. This is listed as follows:

- 558K filtered image-text pairs from LAION/CC/SBU, captioned by BLIP.
- 158K GPT-generated multimodal instruction-following data.
- 450K academic-task-oriented VQA data mixture.
- 40K ShareGPT data.

## Quantitative Analyses

Performance of LLaVA-Gemma models across seven benchmarks. Highlighted box indicates strongest performance amongst LLaVA-Gemma models. Bottom two rows show self-reported performance of Llava Phi-2 and LLaVA-v1.5 respectively. The bolded **gemma-2b-it** is the current model used here in this model card.

| LM Backbone | Vision Model | Pretrained Connector | GQA   | MME cognition | MME perception | MM-Vet | POPE accuracy | POPE F1 | VQAv2 | ScienceQA Image | MMVP  |
| ----------- | ------------ | -------------------- | ----- | ------------- | -------------- | ------ | ------------- | ------- | ----- | --------------- | ----- |
| **gemma-2b-it** | CLIP         | Yes                  | 0.531 | 236           | 1130           | 17.7   | 0.850         |<mark>0.839</mark>| 70.65 | 0.564  | 0.287 |
| gemma-2b-it | CLIP         | No                   | 0.481 | 248           | 935            | 13.1   | 0.784         | 0.762   | 61.74 | 0.549           | 0.180 |
| gemma-2b-it | DinoV2       | Yes                  |<mark>0.587</mark>| 307| <mark>1133</mark>   |<mark>19.1</mark>| <mark>0.853</mark>   | 0.838   |<mark>71.37</mark>| 0.555         | 0.227 |
| gemma-2b-it | DinoV2       | No                   | 0.501 | <mark>309</mark>| 959          | 14.5   | 0.793         | 0.772   | 61.65 | 0.568           | 0.180 |
|             |              |                      |       |               |                |        |               |         |       |                 |       |
| gemma-7b-it | CLIP         | Yes                  | 0.472 | 253           | 895            | 18.2   | 0.848         | 0.829   | 68.7  | 0.625           | <mark>0.327</mark> |
| gemma-7b-it | CLIP         | No                   | 0.472 | 278           | 857            | 19.1   | 0.782         | 0.734   | 65.1 | <mark>0.636</mark>           | 0.240 |
| gemma-7b-it | DinoV2       | Yes                  | 0.519 | 257           | 1021           | 14.3   | 0.794         | 0.762   | 65.2 | 0.628           | <mark>0.327</mark> |
| gemma-7b-it | DinoV2       | No                   | 0.459 | 226           | 771            | 12.2   | 0.693         | 0.567   | 57.4 | 0.598           | 0.267 |
|             |              |                      |       |               |                |        |               |         |       |                 |       |
| Phi-2b      | CLIP         | Yes                  | -     | -             | 1335           | 28.9   | -             | 0.850   | 71.4  | 0.684           | - |
| Llama-2-7b  | CLIP         | Yes                  | 0.620 | 348           | 1511           | 30.6   | 0.850         | 0.859   | 78.5  | 0.704           | 46.1 |

## Ethical Considerations

Intel is committed to respecting human rights and avoiding causing or contributing to adverse impacts on human rights. See [Intel’s Global Human Rights Principles](https://www.intel.com/content/dam/www/central-libraries/us/en/documents/policy-human-rights.pdf). Intel’s products and software are intended only to be used in applications that do not cause or contribute to adverse impacts on human rights.

| Ethical Considerations | Description | 
| ----------- | ----------- | 
| Data | The model was trained using the LLaVA-v1.5 data mixture as described above. |
| Human life | The model is not intended to inform decisions central to human life or flourishing. | 
| Mitigations | No additional risk mitigation strategies were considered during model development. |
| Risks and harms | This model has not been assessed for harm or biases, and should not be used for sensitive applications where it may cause harm. |
| Use cases | - | 

## Caveats and Recommendations

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model.

## Citation details
```bibtex
@misc{hinck2024llavagemma,
      title={LLaVA-Gemma: Accelerating Multimodal Foundation Models with a Compact Language Model}, 
      author={Musashi Hinck and Matthew L. Olson and David Cobbley and Shao-Yen Tseng and Vasudev Lal},
      year={2024},
      eprint={2404.01331},
      url={https://arxiv.org/abs/2404.01331},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```