Hugging Face's logo

xtuner
/
llava-phi-3-mini

Image-Text-to-Text
Safetensors
xtuner
llava_llama
conversational
Model card Files Community
2
LZHgrla commited on
Commit
617fcc5
•
1 Parent(s): fb5d35c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +19 -6
README.md CHANGED
@@ -21,17 +21,30 @@ llava-phi-3-mini is a LLaVA model fine-tuned from [microsoft/Phi-3-mini-4k-instr
21
  **Note: This model is in official LLaVA format. The models in xtuner LLaVA format and HuggingFace LLaVA format can be found on [xtuner/llava-phi-3-mini-xtuner](https://huggingface.co/xtuner/llava-phi-3-mini-xtuner) and [xtuner/llava-phi-3-mini-hf](https://huggingface.co/xtuner/llava-phi-3-mini-hf).**
22
 
23
 
 
24
  ## Details
25
 
26
- | Model | Visual Encoder | Projector | Resolution | Pretraining Strategy | Fine-tuning Strategy | Pretrain Dataset | Fine-tune Dataset |
27
- | :-------------------- | ------------------: | --------: | ---------: | ---------------------: | ------------------------: | ------------------------: | -----------------------: |
28
- | LLaVA-v1.5-7B | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, Frozen ViT | LLaVA-PT (558K) | LLaVA-Mix (665K) |
29
- | LLaVA-Llama-3-8B | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, LoRA ViT | LLaVA-PT (558K) | LLaVA-Mix (665K) |
30
- | LLaVA-Llama-3-8B-v1.1 | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, LoRA ViT | ShareGPT4V-PT (1246K) | InternVL-SFT (1268K) |
31
- | LLaVA-Phi-3-mini | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, Full ViT | ShareGPT4V-PT (1246K) | InternVL-SFT (1268K) |
32
 
33
  ## Results
34
 
 
 
 
 
 
 
 
 
 
 
 
 
35
 
36
  ## Quickstart
37
 
 
21
  **Note: This model is in official LLaVA format. The models in xtuner LLaVA format and HuggingFace LLaVA format can be found on [xtuner/llava-phi-3-mini-xtuner](https://huggingface.co/xtuner/llava-phi-3-mini-xtuner) and [xtuner/llava-phi-3-mini-hf](https://huggingface.co/xtuner/llava-phi-3-mini-hf).**
22
 
23
 
24
+
25
  ## Details
26
 
27
+ | Model | Visual Encoder | Projector | Resolution | Pretraining Strategy | Fine-tuning Strategy | Pretrain Dataset | Fine-tune Dataset | Pretrain Epoch | Fine-tune Epoch |
28
+ | :-------------------- | ------------------: | --------: | ---------: | ---------------------: | ------------------------: | ------------------------: | -----------------------: | -------------- | --------------- |
29
+ | LLaVA-v1.5-7B | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, Frozen ViT | LLaVA-PT (558K) | LLaVA-Mix (665K) | 1 | 1 |
30
+ | LLaVA-Llama-3-8B | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, LoRA ViT | LLaVA-PT (558K) | LLaVA-Mix (665K) | 1 | 1 |
31
+ | LLaVA-Llama-3-8B-v1.1 | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, LoRA ViT | ShareGPT4V-PT (1246K) | InternVL-SFT (1268K) | 1 | 1 |
32
+ | **LLaVA-Phi-3-mini** | CLIP-L | MLP | 336 | Frozen LLM, Frozen ViT | Full LLM, Full ViT | ShareGPT4V-PT (1246K) | InternVL-SFT (1268K) | 1 | 2 |
33
 
34
  ## Results
35
 
36
+ <div align="center">
37
+ <img src="https://github.com/InternLM/xtuner/assets/36994684/78524f65-260d-4ae3-a687-03fc5a19dcbb" alt="Image" width=500" />
38
+ </div>
39
+
40
+ | Model | MMBench Test (EN) | MMMU Val | SEED-IMG | AI2D Test | ScienceQA Test | HallusionBench aAcc | POPE | GQA | TextVQA | MME | MMStar |
41
+ | :-------------------- | :---------------: | :-------: | :------: | :-------: | :------------: | :-----------------: | :--: | :--: | :-----: | :------: | :----: |
42
+ | LLaVA-v1.5-7B | 66.5 | 35.3 | 60.5 | 54.8 | 70.4 | 44.9 | 85.9 | 62.0 | 58.2 | 1511/348 | 30.3 |
43
+ | LLaVA-Llama-3-8B | 68.9 | 36.8 | 69.8 | 60.9 | 73.3 | 47.3 | 87.2 | 63.5 | 58.0 | 1506/295 | 38.2 |
44
+ | LLaVA-Llama-3-8B-v1.1 | 72.3 | 37.1 | 70.1 | 70.0 | 72.9 | 47.7 | 86.4 | 62.6 | 59.0 | 1469/349 | 45.1 |
45
+ | **LLaVA-Phi-3-mini** | 69.2 | 41.4 | 70.0 | 69.3 | 73.7 | 49.8 | 87.3 | 61.5 | 57.8 | 1477/313 | 43.7 |
46
+
47
+
48
 
49
  ## Quickstart
50