Will a single model running on a single consumer GPU (<1.5k 2020 USD) outperform GPT-3 175B on all benchmarks in the original paper by 2025?
Plus
28
Ṁ38762026
86%
chance
1D
1W
1M
ALL
There are no restrictions on the amount or kind of compute used to *train* the model. Question is about whether it will actually be done, not whether it will be possible in theory. If I judge the model to really be many specific models stuck together to look like one general model it will not count.
This question is managed and resolved by Manifold.
Get
1,000
and3.00
Sort by:
Llamas on pixel 7s https://github.com/rupeshs/alpaca.cpp/tree/linux-android-build-support (ik ik its not over 13B yet, just sharing progress)
@ValeryCherepanov By "run on a single GPU" I mean the weights + one full input vector can fit on a consumer GPU at once. Otherwise the question would be meaningless - you can always split up matrices into smaller blocks and run the computation sequentially.
Related questions
Related questions
Will there be a model that has a 75% win rate against the latest iteration of GPT-4 as of January 1st, 2025?
62% chance
Will an open source model beat GPT-4 in 2024?
39% chance
Will it cost less than 100k USD to train and run a language model that outperforms GPT-3 175B on all benchmarks by the end 2024?
85% chance
Will a language model that runs locally on a consumer cellphone beat GPT4 by EOY 2026?
70% chance
Will any open-source model achieve GPT-4 level performance on MMLU through 2024?
83% chance
Will a model be trained using at least as much compute as GPT-3 using AMD GPUs before Jan 1 2026?
84% chance
Will it be possible to disentangle most of the features learned by a model comparable to GPT-3 this decade? (1k subsidy)
56% chance
Will a 15 billion parameter LLM match or outperform GPT4 in 2024?
24% chance
Will a GPT-3 quality model be trained for under $10.000 by 2030?
83% chance
Will the performance jump from GPT4->GPT5 be less than the one from GPT3->GPT4?
75% chance