"Assistant Pepe 70B" (believe it or not) https://huggingface.co/SicariusSicariiStuff/Assistant_Pepe_70B
https://huggingface.co/SicariusSicariiStuff/Assistant_Pepe_70B
I feel a bit weird suggesting this one, as "Pepe" is politicized in the US. Don't take this request as either an endorsement or a condemnation of "Pepe the Meme Frog" as a concept.
But - this model genuinely exhibits some interesting behaviors. Running it without a system prompt, it adds a surprising amount of character, and it's apparently pretty good at some unusual tasks. 70B model, I believe Llama 3.1 derived? The author created several GGUFs, but I'm hoping to encounter some intermediate sizes, so I can get closer to the performance of the top model. (48 GB of VRAM - 2x 3090)
| PP | TG | Size | Model |
|---|---|---|---|
| 564.4 | 13.3 | 37.9 | Golddiamondgold-Paperbliteration-L33-70b:IQ4_XS |
| 386.1 | 5.3 | 42.5 GB | Assistant_Pepe_70B:Q4_K_M |
we not hating anyone, we just quanting =)
It's queued!
You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#Assistant_Pepe_70B-GGUF for quants to appear.
Looks like nothing's shown up here - IDK if it's specific to this model, or if 70B ones take longer to get queued & processed; just figured I'd check in!
lol it failed with some random error, I will queue it again, it should work this time
lol it failed with some random error, I will queue it again, it should work this time
initial upload was corrupted, reuploaded, now works, sry for the inconvenience, and ty for quanting!
nice to see you here =)
feel free to cook more models and ask to quant =)