Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,19 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Mirage 12B
|
| 2 |
+
|
| 3 |
+
Mirage represents a novel series of multi-modal models, positioned at a higher tier than the Bumblebee series.
|
| 4 |
+
|
| 5 |
+
Our preceding 7B model (within the Bumblebee series) had attained a high score on numerous benchmarks. To further enhance its performance, we trained a 12B model.
|
| 6 |
+
It comprises a ViT vision encoder and a Mistral nemo 12B model.
|
| 7 |
+
|
| 8 |
+
The results indicate a substantial improvement when compared to many other open-source models.
|
| 9 |
+
|
| 10 |
+
Our Mirage 12B has surpassed Cambrian-34B and LLava-Next-34B with a single 12B parameter.
|
| 11 |
+
|
| 12 |
+
The benchmark scores can be viewed on the MMBench official leaderboard.
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
## Roadmap
|
| 16 |
+
|
| 17 |
+
The currently released version is 2.0 of Mirage. Our 2.1 version is in progress and is likely to be even more potent with multi-lingual OCR support.
|
| 18 |
+
|
| 19 |
+
If you have any inquiries, feel free to contact us through the issue!
|