• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
AI
Reaction

Hugging Face download rates reportedly drop about 30% for many tracked top open LLMs

A user monitoring the models says download counting may have changed slightly, but the decline has continued.

Nathan LambertNL
Andrew TraskAT
5 Sources, 1h ago, first seen 1h ago

TLDR

A user tracking top open LLMs on Hugging Face reports a roughly 30% drop in download rates for many of the models monitored. They say download counting may have changed slightly, but the decline has continued. In a reply, another user guesses that fewer people are trying to host models themselves.

Combined views

11.4K

5 Sources, first seen 1h ago

96 likes42 comments36 saves5 reposts

Combined views

11.4K

5 Sources, first seen 1h ago

96 likes42 comments36 saves5 reposts

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Featured Source

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

5 Sources

Nathan Lambert@natolambertWe've been monitoring a decline in HuggingFace downloads across the top open LLMs we track for interconnects. It looks like there was a slight change in counting of downloads, as we saw a ~30% download rate drop for many top models, but this is continuing over time. What else do people think it may be?1h
Andrew Trask@iamtrask@natolambert a guess: people are spending less time trying to host models themselves.1h
Vaibhav (VB) Srivastav@reach_vb@natolambert potential thesis: developers/ tinkerers are using bigger open weights models via OpenRouter (& other gateways) (would checkout since hosting > 500B param models is non trivial) the utility of running a smaller model is almost eroded by cheaper and faster small models from labs1h
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI

    5 Sources

    Nathan Lambert@natolambertWe've been monitoring a decline in HuggingFace downloads across the top open LLMs we track for interconnects. It looks like there was a slight change in counting of downloads, as we saw a ~30% download rate drop for many top models, but this is continuing over time. What else do people think it may be?1h
    Andrew Trask@iamtrask@natolambert a guess: people are spending less time trying to host models themselves.1h
    Vaibhav (VB) Srivastav@reach_vb@natolambert potential thesis: developers/ tinkerers are using bigger open weights models via OpenRouter (& other gateways) (would checkout since hosting > 500B param models is non trivial) the utility of running a smaller model is almost eroded by cheaper and faster small models from labs1h
    Today's Rank

    #5

    Today's Rank

    #5