Update app.py

app.py

@@ -44,7 +44,7 @@ for kn in np.linspace(0.2, 2, 100):
 def plot_curve(kn, kd):
     fig = plt.figure()
     plt.plot(kns, overheads, color="black")
-    plt.scatter([kn], [compute_overhead(kn, kd)*100], marker="D", c="red", label="You are here!")
+    plt.scatter([kn], [compute_overhead(kn, kd)*100], marker="x", c="red", label="You are here!")
     plt.xlabel("Fraction of compute optimal model size")
     plt.ylabel("Compute overhead (%)")
     plt.legend(loc="best")
@@ -70,10 +70,12 @@ Compute budget (TFLOPs): {C:.2E}
 
 ## Chinchilla optimal:
 Optimal model size:\t\t {N_opt/Bn:.2f}B
+
 Optimal datset size (tokens):\t {D_opt/Bn:.2f}
 
 ## Your setting trade-off:
 Training compute overhead (%):\t {100*compute_overhead(kn, kd):.2f}
+
 Inference cost fraction (%):\t {kn*100:.2f}"""
     return text, fig