Super-Slop-Machina-XXL-v0.1-4B: Small Scale Roleplay Model
"Now larger, with higher quality, better understanding."
Improvements:
- Better adapted for longer context
- Higher output quality
- Better creative writing
- High quality dataset with exceptional long term chatlogs was used
Quants(Recommend either Q8_0 in this repo or i1 Q6_K and lower from mradermacher when comes out):
- BF16: Overkill
- Q8_0: Extreme quality
- Q6_K: Very good quality
- Q5_K_M: Good quality, faster speed
- Q4_K_M: Decent quality, fastest
Note:
I used my new experimental dataset for training this model, same dataset that was used for Super-Slop-Machina-Roleplay-1.2b-V4-ERP-Tolerant-it, still pretty much very experimental.
Under 1B models soon! I tried my best with them, results were rather unusual, when I'm done with them, I'll probably return my attention back to ministral 3 3b.
Big announcement:
I experimented with Qwen 3.5 9b and I think I am somewhat succeeding, I will soon release the finetuned weights, result that I was able to achieve with it was VERY interesting...