Update README.md
Browse files
README.md
CHANGED
@@ -19,6 +19,60 @@ tags:
|
|
19 |
This model was converted to GGUF format from [`prithivMLmods/LatexMind-2B-Codec`](https://huggingface.co/prithivMLmods/LatexMind-2B-Codec) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
|
20 |
Refer to the [original model card](https://huggingface.co/prithivMLmods/LatexMind-2B-Codec) for more details on the model.
|
21 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
## Use with llama.cpp
|
23 |
Install llama.cpp through brew (works on Mac and Linux)
|
24 |
|
|
|
19 |
This model was converted to GGUF format from [`prithivMLmods/LatexMind-2B-Codec`](https://huggingface.co/prithivMLmods/LatexMind-2B-Codec) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
|
20 |
Refer to the [original model card](https://huggingface.co/prithivMLmods/LatexMind-2B-Codec) for more details on the model.
|
21 |
|
22 |
+
---
|
23 |
+
The LatexMind-2B-Codec model is a fine-tuned version of Qwen2-VL-2B-Instruct, optimized for Optical Character Recognition (OCR), image-to-text conversion, and mathematical expression extraction with LaTeX formatting.
|
24 |
+
This model integrates a conversational approach with visual and textual
|
25 |
+
understanding to handle multi-modal tasks effectively.
|
26 |
+
|
27 |
+
|
28 |
+
|
29 |
+
|
30 |
+
|
31 |
+
|
32 |
+
|
33 |
+
Key Enhancements:
|
34 |
+
|
35 |
+
|
36 |
+
|
37 |
+
|
38 |
+
SoTA understanding of images with various resolutions & aspect ratios:
|
39 |
+
LatexMind-2B-Codec achieves state-of-the-art performance on visual
|
40 |
+
understanding benchmarks, including MathVista, DocVQA, RealWorldQA,
|
41 |
+
MTVQA, etc.
|
42 |
+
|
43 |
+
|
44 |
+
Advanced LaTeX extraction: The model specializes
|
45 |
+
in extracting structured mathematical expressions from images and
|
46 |
+
documents, converting them into LaTeX format for precise rendering and
|
47 |
+
further computation.
|
48 |
+
|
49 |
+
|
50 |
+
Understanding long-duration videos (20min+):
|
51 |
+
LatexMind-2B-Codec can process videos over 20 minutes long, enabling
|
52 |
+
high-quality video-based question answering, mathematical solution
|
53 |
+
explanation, and educational content creation.
|
54 |
+
|
55 |
+
|
56 |
+
Agent capabilities for automated operations:
|
57 |
+
With complex reasoning and decision-making abilities, the model can be
|
58 |
+
integrated with mobile devices, robots, and assistive technologies to
|
59 |
+
automate tasks based on visual and textual inputs.
|
60 |
+
|
61 |
+
|
62 |
+
Multilingual Support: To serve global users, in
|
63 |
+
addition to English and Chinese, the model supports text recognition
|
64 |
+
inside images across multiple languages, including European languages,
|
65 |
+
Japanese, Korean, Arabic, Vietnamese, etc.
|
66 |
+
|
67 |
+
|
68 |
+
|
69 |
+
|
70 |
+
This model is particularly effective in retrieving mathematical notations and equations
|
71 |
+
from scanned documents, whiteboard images, and handwritten notes,
|
72 |
+
ensuring accurate conversion to LaTeX code for further academic and
|
73 |
+
computational applications.
|
74 |
+
|
75 |
+
---
|
76 |
## Use with llama.cpp
|
77 |
Install llama.cpp through brew (works on Mac and Linux)
|
78 |
|