Gokulnath2003 commited on
Commit
98c29f4
·
verified ·
1 Parent(s): ae63231

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +20 -12
app.py CHANGED
@@ -19,17 +19,7 @@ model_links = {
19
  # Pull info about the model to display
20
  model_info = {
21
  "Meta-Llama-3-8B": {
22
- 'description': """The **Meta-Llama 3 (8B)** is a cutting-edge **Large Language Model (LLM)** developed by Meta's AI team, comprising over 8 billion parameters. This model has been specifically fine-tuned for educational purposes to excel in interactive question-and-answer sessions.\n
23
- \n### Training Process:
24
- This Llama model was meticulously fine-tuned using science textbooks from the NCERT curriculum, which covers a wide range of subjects including Physics, Chemistry, Biology, and Environmental Science. The fine-tuning process utilized **Docker AutoTrain**, enabling scalable and automated training pipelines. The model was trained on datasets focusing on providing detailed, accurate responses in line with the NCERT syllabus.
25
- \n### Purpose:
26
- Llama-3 8B is designed to assist both students and educators by delivering clear, concise explanations to science-related questions. With a deep understanding of the NCERT curriculum, it helps break down complex scientific concepts, making learning easier and more engaging for students, while acting as an intuitive guide for teachers.
27
- \n### Specialized Features:
28
- - **Contextual Understanding**: Optimized to handle detailed science-related queries, ensuring high relevance in responses.
29
- - **Fine-Grained Knowledge**: Equipped to offer explanations on subjects ranging from basic scientific principles to advanced concepts, ideal for various educational levels.
30
- - **Accuracy and Reliability**: Trained with a focus on minimizing misinformation, this model prioritizes delivering trustworthy responses, tailored specifically for the education sector.\n
31
- This model is a testament to the potential of AI in revolutionizing education by offering students a personal, reliable assistant to clarify doubts and enrich their understanding of science.
32
- """
33
  }
34
  }
35
 
@@ -49,7 +39,25 @@ st.sidebar.write("Base-Model used: Meta Llama, trained using: Docker AutoTrain."
49
 
50
  # Add technical details in the sidebar
51
  st.sidebar.markdown(model_info["Meta-Llama-3-8B"]['description'])
52
- st.sidebar.markdown("*By Gokulnath ♔ *")
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
53
 
54
  # If model selection was needed (now removed)
55
  selected_model = "Meta-Llama-3-8B" # Only one model remains
 
19
  # Pull info about the model to display
20
  model_info = {
21
  "Meta-Llama-3-8B": {
22
+ 'description': """The **Meta-Llama 3 (8B)** is a cutting-edge **Large Language Model (LLM)** developed by Meta's AI team, comprising over 8 billion parameters. This model has been specifically fine-tuned for educational purposes to excel in interactive question-and-answer sessions.\n"""
 
 
 
 
 
 
 
 
 
 
23
  }
24
  }
25
 
 
39
 
40
  # Add technical details in the sidebar
41
  st.sidebar.markdown(model_info["Meta-Llama-3-8B"]['description'])
42
+ st.sidebar.markdown("""
43
+ ### Meta-Llama 3 (8B)
44
+ Yo, this **Meta-Llama 3 (8B)** is a next-level AI model built by Meta's genius squad. It packs a whopping **8 billion parameters** and is totally fine-tuned for school stuff, especially science Q&As.
45
+
46
+ ### How it’s Trained:
47
+ We fed it all the juicy info from **NCERT science textbooks** (yep, the same ones you use), covering Physics, Chem, Bio, and more. With **Docker AutoTrain**, we made sure it learns fast and scales like a boss.
48
+
49
+ ### What it Does:
50
+ This Llama’s mission? Make science super easy. Whether you're stuck on a tricky topic or just need a quick answer, it’s here to help you (and your teacher) break things down like a pro.
51
+
52
+ ### Cool Stuff:
53
+ - **Knows Its Stuff**: It's laser-focused on science questions, so no random answers.
54
+ - **Super Smart**: Whether it's easy or advanced, it’s got you covered.
55
+ - **Trustworthy**: It's trained to get things right, so you can count on its answers.
56
+
57
+ Basically, it's like having a personal science tutor right in your pocket.
58
+ """)
59
+
60
+ st.sidebar.markdown("By Gokulnath ♔")
61
 
62
  # If model selection was needed (now removed)
63
  selected_model = "Meta-Llama-3-8B" # Only one model remains