metadata
datasets:
- Lin-Chen/ShareGPT4V
pipeline_tag: image-to-text
Model
llava-llama-3-8b-v1_1 is a LLaVA model fine-tuned from meta-llama/Meta-Llama-3-8B-Instruct and CLIP-ViT-Large-patch14-336 with ShareGPT4V-PT and InternVL-SFT by XTuner.
Note: This model is in GGUF format.
Resources:
- GitHub: xtuner
- HuggingFace LLaVA format model: xtuner/llava-llama-3-8b-v1_1-transformers
- Official LLaVA format model: xtuner/llava-llama-3-8b-v1_1-hf
- XTuner LLaVA format model: xtuner/llava-llama-3-8b-v1_1
Details
| Model | Visual Encoder | Projector | Resolution | Pretraining Strategy | Fine-tuning Strategy | Pretrain Dataset | Fine-tune Dataset |
|---|---|---|---|---|---|---|---|
| LLaVA-v1.5-7B | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, Frozen ViT | LLaVA-PT (558K) | LLaVA-Mix (665K) |
| LLaVA-Llama-3-8B | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, LoRA ViT | LLaVA-PT (558K) | LLaVA-Mix (665K) |
| LLaVA-Llama-3-8B-v1.1 | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, LoRA ViT | ShareGPT4V-PT (1246K) | InternVL-SFT (1268K) |
Results
| Model | MMBench Test (EN) | MMBench Test (CN) | CCBench Dev | MMMU Val | SEED-IMG | AI2D Test | ScienceQA Test | HallusionBench aAcc | POPE | GQA | TextVQA | MME | MMStar |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| LLaVA-v1.5-7B | 66.5 | 59.0 | 27.5 | 35.3 | 60.5 | 54.8 | 70.4 | 44.9 | 85.9 | 62.0 | 58.2 | 1511/348 | 30.3 |
| LLaVA-Llama-3-8B | 68.9 | 61.6 | 30.4 | 36.8 | 69.8 | 60.9 | 73.3 | 47.3 | 87.2 | 63.5 | 58.0 | 1506/295 | 38.2 |
| LLaVA-Llama-3-8B-v1.1 | 72.3 | 66.4 | 31.6 | 36.8 | 70.1 | 70.0 | 72.9 | 47.7 | 86.4 | 62.6 | 59.0 | 1469/349 | 45.1 |
Quickstart
Download models
# mmproj
wget https://huggingface.co./xtuner/llava-llama-3-8b-v1_1-gguf/resolve/main/llava-llama-3-8b-v1_1-mmproj-f16.gguf
# fp16 llm
wget https://huggingface.co./xtuner/llava-llama-3-8b-v1_1-gguf/resolve/main/llava-llama-3-8b-v1_1-f16.gguf
# int4 llm
wget https://huggingface.co./xtuner/llava-llama-3-8b-v1_1-gguf/resolve/main/llava-llama-3-8b-v1_1-int4.gguf
# (optional) ollama fp16 modelfile
wget https://huggingface.co./xtuner/llava-llama-3-8b-v1_1-gguf/resolve/main/OLLAMA_MODELFILE_F16
# (optional) ollama int4 modelfile
wget https://huggingface.co./xtuner/llava-llama-3-8b-v1_1-gguf/resolve/main/OLLAMA_MODELFILE_INT4
Chat by ollama
# fp16
ollama create llava-llama3-f16 -f ./OLLAMA_MODELFILE_F16
ollama run llava-llama3-f16 "xx.png Describe this image"
# int4
ollama create llava-llama3-int4 -f ./OLLAMA_MODELFILE_INT4
ollama run llava-llama3-int4 "xx.png Describe this image"
Chat by llama.cpp
Note: llava-llama-3-8b-v1_1 uses the Llama-3-instruct chat template.
# fp16
./llava-cli -m ./llava-llama-3-8b-v1_1-f16.gguf --mmproj ./llava-llama-3-8b-v1_1-mmproj-f16.gguf --image YOUR_IMAGE.jpg -c 4096 -e -p "<|start_header_id|>user<|end_header_id|>\n\n<image>\nDescribe this image<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
# int4
./llava-cli -m ./llava-llama-3-8b-v1_1-int4.gguf --mmproj ./llava-llama-3-8b-v1_1-mmproj-f16.gguf --image YOUR_IMAGE.jpg -c 4096 -e -p "<|start_header_id|>user<|end_header_id|>\n\n<image>\nDescribe this image<|eot_id|><|start_header_id|>assistant<|end_header_id|>\n\n"
Reproduce
Please refer to docs.
Citation
@misc{2023xtuner,
title={XTuner: A Toolkit for Efficiently Fine-tuning LLM},
author={XTuner Contributors},
howpublished = {\url{https://github.com/InternLM/xtuner}},
year={2023}
}