TheBloke

/

WizardLM-30B-Uncensored-GGML

like 119

Really impressive.
WizardLM-30B-Uncensored .
I was making reasoning tests and I am really impressed.
Easily beat all 7B , 13B and 30b models.
Is even better than alpaca-lora-65B.ggmlv3.q5_1.

It has level of gpt4-alpaca-lora_mlp-65B.ggmlv3.q5_1... a slightly worse in writing... reasoning same level.

Solved this one - only 65b solving it properly (only gpt4-alpaca-lora_mlp-65B actually)

solve this equation and explain each step 2Y-12=-16
To solve the equation 2Y - 12 = -16, we need to isolate the variable Y on one side of the equation. Here are the steps to do that:

Step 1: Subtract 12 from both sides of the equation to isolate the term 2Y:
2Y - 12 - 12 = -16 - 12
Simplifying this step, we get:
2Y = -4

Step 2: Divide both sides of the equation by 2 to isolate Y:
2Y/2 = (-4)/2
Simplifying this step, we get:
Y = -2

Therefore, the solution to the equation 2Y - 12 = -16 is Y = -2. This means that if we substitute -2 for Y in the original equation, we will get a true statement:
2(-2) - 12 = -16
-4 - 12 = -16
-16 = -16
This shows that the solution is valid because it satisfies the original equation.

Even that question answered ... so close ..only GPT4 or Bing AI can do that

Mark is faster than Mary , Mary is faster than Joe. Is Joe faster than Mark?

Based on the given information, we can make a comparison chart:

From the chart, we can see that Mary is faster than Joe. However, we cannot determine whether Joe is faster than Mark or not since there is no information about their relative speeds.

Interesting thing is ... I asked to write a story on 2000 words so the model make it .. BUT I asked to continue the story and did it perfectly!
Continued the story after 2000 words still keeps story line intact and coherent.
I never met any model that could do that after 2000 tokens... that is strange ;)