the quality of GLM 5.2 in just 155 or 162 GB?
unsloth/DeepSeek-V4-Flash-0731-GGUF · Hugging Face
https://huggingface.co/unsloth/DeepSeek-V4-Flash-0731-GGUF
https://huggingface.co/unsloth/DeepSeek-V4-Flash-0731-GGUF
the quality of GLM 5.2 in just 155 or 162 GB?
8 Comments
hummingbird@lemmy.world · 6 pts · 19d
Huh the unsloth Q4_K_XL variant running with 5-6 t/s on a 16GB 9070 XT, 128 GB RAM and an NVME ssd. This is shockingly usable 🤯
DacoTaco@lemmy.world · 3 pts · 19d
Did some more testing on it? This isnt a small model and i havent found the space to save it on my machines yet haha
BeefAndPoultry@lemmus.org · 3 pts · 19d
make sure that holds up with large context, you might need to step down to Q3 (which I've heard is still good for this model, many people are even using IQ2)
Multiplexer@discuss.tchncs.de · 5 pts · 20d
Wow, already unsloth quantizations available? That was fast.
Didn't even know they released the weights until just about now.
This will probably also mean that we will have alternative providers on Openrouter soon. Can't wait.
DS V4 will have a huge impact on many usecases. Dirt cheap and insanely capable, if only a part of the benchmarks translate to real life.
Increasingly interesting times for the big US AI companies, I guess...
Multiplexer@discuss.tchncs.de · 3 pts · 20d
Anyone knows, why 4bit quant is only marginally smaller than 8bit quant, though?
And shouldn't 8bit be roughly equal to the number of parameters anyway, so ~300GB?
BeefAndPoultry@lemmus.org · 4 pts · 20d
the original model has a lot of parts that were natively trained in 4 bit, so those layers can't go higher
BeefAndPoultry@lemmus.org · 2 pts · 20d
this appears to be the untouched model in GGUF format, 180 GB, MXFP4
https://huggingface.co/bartowski/DeepSeek-V4-Flash-0731-GGUF
notfromhere@lemmy.ml · 1 pts · 18d
Does antirez have new quants for ds4 ready?