Update README.md
Browse files
README.md
CHANGED
|
@@ -15,11 +15,11 @@ Inspired by human attention patterns in Visual Question Answering (VQA) tasks, T
|
|
| 15 |
The TRIM method has been extensively tested across 12 datasets, and the results demonstrate a significant reduction in computational overhead while maintaining a consistent level of performance.
|
| 16 |
This research marks a critical stride in efficient MLLM development, promoting greater accessibility and sustainability of high-performing models.
|
| 17 |
|
| 18 |
-
<img src="./main.
|
| 19 |
|
| 20 |
TRIM significantly streamlines the computational process, reducing the number of image tokens by approximately 79%, processing time by 67%, and memory usage by 30% relative to the baseline (LLaVA-1.5-7B).
|
| 21 |
|
| 22 |
-
<img src="./fig1.
|
| 23 |
|
| 24 |
## How to use?
|
| 25 |
|
|
|
|
| 15 |
The TRIM method has been extensively tested across 12 datasets, and the results demonstrate a significant reduction in computational overhead while maintaining a consistent level of performance.
|
| 16 |
This research marks a critical stride in efficient MLLM development, promoting greater accessibility and sustainability of high-performing models.
|
| 17 |
|
| 18 |
+
<img src="./main.jpg" width="600" alt="TRIM" align="center" />
|
| 19 |
|
| 20 |
TRIM significantly streamlines the computational process, reducing the number of image tokens by approximately 79%, processing time by 67%, and memory usage by 30% relative to the baseline (LLaVA-1.5-7B).
|
| 21 |
|
| 22 |
+
<img src="./fig1.jpg" width="300" alt="stat2"/>
|
| 23 |
|
| 24 |
## How to use?
|
| 25 |
|