Rulga commited on
Commit
801484b
·
1 Parent(s): 773406d

No code changes detected.
```

Files changed (1) hide show
  1. app.py +4 -4
app.py CHANGED
@@ -821,16 +821,16 @@ with gr.Blocks() as demo:
821
  gr.Markdown("""
822
  <small>
823
 
824
- **Epochs:**
825
  Lower = Faster training -> Higher = Model learns more thoroughly
826
  Best for small datasets: 3-5 -> Best for large datasets: 1-2
827
 
828
  **Batch Size:**
829
- Lower = Slower but more stable -> Higher = Faster but needs more RAM
830
  4 = Good for 16GB RAM -> 8 = Good for 32GB RAM
831
 
832
- **Learning Rate:**
833
- Lower = Learns slower but more reliable -> Higher = Learns faster but may be unstable
834
  2e-4 (0.0002) = Usually works best -> 1e-4 = Safer choice for fine-tuning
835
  </small>
836
  """)
 
821
  gr.Markdown("""
822
  <small>
823
 
824
+ **Epochs:**
825
  Lower = Faster training -> Higher = Model learns more thoroughly
826
  Best for small datasets: 3-5 -> Best for large datasets: 1-2
827
 
828
  **Batch Size:**
829
+ Lower = Slower but more stable -> Higher = Faster but needs more RAM
830
  4 = Good for 16GB RAM -> 8 = Good for 32GB RAM
831
 
832
+ **Learning Rate:**
833
+ Lower = Learns slower but more reliable -> Higher = Learns faster but may be unstable
834
  2e-4 (0.0002) = Usually works best -> 1e-4 = Safer choice for fine-tuning
835
  </small>
836
  """)