guardiangod
Joined 12 years ago
Karma
3,522
Submissions
466
Member since
May 20
Recent submissions
Comment · 1 month ago
Comment · 2 months ago
Yeah I am decompiling a PS2 game. It's not easy but I think I finally have the right setup.
Comment · 2 months ago
You missed the years long work at FTX under Sam Bank Friedman all the way to FTX's bankruptcy.
Comment · 2 months ago
I am running DS v4 flash 0731 lossless at 80t/s right now. It really is not at Opus 4.5 level (for my workload). I would say it's around 3.7 Sonnet, which is still pretty good, but other models such as GLM 5.2 are still leaps better. Of course I run DSv4 flash over GLM 5.2 for a few very good reasons, but intelligence is not 1 of them.
Comment · 2 months ago
Extremely large 1 bit models are usually within 50-60% of KV divergence to lossless models. In this case I think the comparison to Opus 4.5 is a fair assessment.
Extremely large models don't suffer as much from quantization due to its weight topology also contains encoded information, so the loss of info from any one weight is somewhat mitigated.
Comment · 2 months ago
https://unsloth.ai/docs/models/qwen3.8
The 1bit quant model is at an astonishing 397GB with 95B active per MOE. This literally puts Opus 4.5 performance level into a machine a normal person could buy, and still gets usable tokens/second.
The full lossless model BF16 is clocking at 4.9TB. The model card claims the model to be between Opus 4.8 and Fable 5. Again that's astonishing as getting a machine with 7TB RAM (with context + KV cache) is still within the realm of medium size companies.
Bad things: The open source version has its vision capability removed, and the context capped at 250k . I expect someone to bolt a Kimi 2.6 vision tower to it to restore the vision capability (at less performance of course). For context, I played around with extending the context to 600k for Qwen 3.5 397b, and the context remained stable up to around 480k. It'd be interesting to see if the same can be done to Q3.8 .
Also no out of the box DSpark/DFlash support. MTP is present so we should at least get some boost in TP speed.
Comment · 2 months ago
The article is
One line after another
Hard to read it is
Comment · 3 months ago
“You are trying to kidnap what I have rightfully stolen, and I think it quite ungentlemanly."
Comment · 3 months ago
Don't ever ask GPT Sol on how to LARP Fallout games, thanks.
Comment · 3 months ago
I disagree with the idea that file is a good metaphor for LLM. Files have seek and byte streams, which is just an unneeded abstraction for LLM. The LLM doesn't need to seek or jump to the middle of a file, if you store and organize your data properly.
Why force the LLM to use files over vector database or key-value stores, just because it's a design principal for UNIX (which is designed for human users, not LLMs.)
Comment · 3 months ago
If your objective is just to hack something together quickly to test an idea, and doesn't care if the application segfault 6 hours later (eg. a prototype game engine), C++ is definitely the superior choice to Rust.
Rust is like the instructor demanding that you build your house up to the national code, down to the choice of nail for your floor board. Your house will be perfect, but building it is extremely difficult and high-friction. In many cases this is unnecessary.
Well actually, it might.