Upload README.md

README.md

@@ -381,8 +381,45 @@ Purely Experimental - Works well enough I think?
 
 Prompt Template: Vicuna 1.1 / Alpaca
 
+```
+Below is an instruction that describes a task. Write a response that appropriately completes the request.
+
+### Instruction:
+{prompt}
+
+### Response:
+
+```
+
+or
+
+```
+User: <Prompt>
+
+Assistant:
+```
+
+***
+
+Mini Rant:
+
+Honestly I don't know if 13b Llama2 models will get any better. Mythomax immediately hit a peak, and from then on its all minor upgrades/sidegrades (unbiased, but from Stheno / Mlewd onwards I honestly did not care much for other models).
+With so many new blind merges and new finetunes it all feels like everyone is trying to hit that high,I know I tried. Maybe it'll never be hit again? I don't know though, just my opinion.
+
+There is heavy model incest now, people blindly merging heavily merged models together, and then someone else merging that with another merge. Some merges now contain like over 100+ models, and that number will only go up.
+
+At that point Sampler settings & prompts have way more meaningful effects compared to merges. People also do not care to compare older models with new Sampler settings, like Mythomax still holds up pretty damn well even now with the cool as fuck min_P sampler.
+
+Newer finetunes and loras are cool, but after a while they get stale fast, maybe its the nature of smaller models, I do not know. Or I'm just a hopeless addict who spent way too much time talking to my characters.
+
+Maybe I am being a bitch or whatever, or I'm too spoiled by 70Bs. Fuck. I don't know.
+
+Rant end.
+
 ***
 
+*Schizo Brainrot below.*
+
 Please help. 
 
 Stheno has held me hostage.