[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]
Showing only X posts for non-authenticated requests. Use your API key in requests for full results.
"@code_star I think the odds of SFT working on a small budget are mid whereas RL will be low. But on high budget SFT stays mid yet RL is high. RL is simply less compute efficient per bit"
X Link @JoshPurtell 2025-10-16T22:19Z 2560 followers, XX engagements
"@JoshPurtell What does that mean Do you require different compute budgets to run an experiments on Qwen X 0.6B with SFT vs GRPO"
X Link @code_star 2025-10-16T22:16Z 4919 followers, XX engagements
"for learning a new programming language youll want more than just RL mix of continued pretraining SFT and then agentic RL would almost certainly beat frontier models in-domain if done properly but its a major project report i worked on along similar lines"
X Link @willccbb 2025-10-16T19:58Z 35.2K followers, XXX engagements
"#keinemusik #egypt #techno #party #sft #afrohouse #technomusic #housemusiclovers #tulum #shorts"
YouTube Link @deephousesft 2025-10-16T19:06Z XXX followers, XXX engagements