an LLM as capable as GPT-4 runs on one 3090 by March 2025
26
1kṀ1720
Mar 2
33%
chance

e.g. Winograde >= 87.5%

Get
Ṁ1,000
to start trading!

People are also trading

an LLM as capable as GPT-4 runs on one 4090 by March 2025
35% chance
Which next-gen frontier LLMs will be released before GPT-5? (2025)
China will make a LLM approximately as good or better than GPT4 before 2025
89% chance
Will it be possible to run an LLM of GPT-4 (or higher) capability on a portable device by 2027?
47% chance
Size of smallest open-source LLM marching GPT 3.5's performance in 2025? (GB)
2.13
Will there be an LLM (as good as GPT-4) that was trained with 1/100th the energy consumed to train GPT-4, by 2026?
82% chance
Will xAI develop a more capable LLM than GPT-5 by 2026
51% chance
Will an open-source LLM beat or match GPT-4 by the end of 2024?
83% chance
Will any open source LLM with <20 billion parameters outperform GPT-4 on most language benchmarks by the end of 2024?
13% chance
Will an open-source fully functional Auto-GPT like LLM exist by the end of 2025?
90% chance
Sort by:

I sorta see the argument, llama-32b-r1 scores 88% on MMLU vs gpt-4's 86.4, you could run it on a 3090 with offloading or fp4

I imagine the weights need to fit in VRAM; does that sound right? Or can the solution involve moving weights on/off the GPU

@NoaNabeshima weights have to fit in vram

@sylv ram and not necessarily vram?

*vram sorry

predictedYES

new market for 4090: https://manifold.markets/sylv/an-llm-as-capable-as-gpt4-runs-on-o-8da33ccbe4b1

bought Ṁ80 NO

@sylv The difference between the 3090 market and the 4090 market is because.. the 4090 supports fp8??

@NoaNabeshima I mean it has better specs in general but fp8 is pretty cool

@NoaNabeshima Oh it's different dates

People are also trading

an LLM as capable as GPT-4 runs on one 4090 by March 2025
35% chance
Which next-gen frontier LLMs will be released before GPT-5? (2025)
China will make a LLM approximately as good or better than GPT4 before 2025
89% chance
Will it be possible to run an LLM of GPT-4 (or higher) capability on a portable device by 2027?
47% chance
Size of smallest open-source LLM marching GPT 3.5's performance in 2025? (GB)
2.13
Will there be an LLM (as good as GPT-4) that was trained with 1/100th the energy consumed to train GPT-4, by 2026?
82% chance
Will xAI develop a more capable LLM than GPT-5 by 2026
51% chance
Will an open-source LLM beat or match GPT-4 by the end of 2024?
83% chance
Will any open source LLM with <20 billion parameters outperform GPT-4 on most language benchmarks by the end of 2024?
13% chance
Will an open-source fully functional Auto-GPT like LLM exist by the end of 2025?
90% chance
© Manifold Markets, Inc.•Terms•Privacy