←Search results

Search “Hugging Face”

By relevance · Total 1193 results Clear search

Saturday

  6. PyTorch Blog

    PyTorch introduces Zephon elastic deterministic data loading for TorchTitan

    The PyTorch blog introduces Zephon, bringing elastic deterministic data loading to TorchTitan: after training a 1B-parameter model to step 150 on 16 GPUs, training was resumed on 8, 32, and 64 GPUs, and Zephon's loss curves overlapped exactly, with the SHA-256 of every global batch matching the baseline. With the standard data loader, training the same 1B model on 20B tokens across 8 to 64 GPUs produced fluctuations of up to 0.82 points on DCLM Core v1 and 0.64 points on FineWeb, whereas Zephon showed only 0.05 and 0.011 points. Zephon performs tokenization, packing, shuffling, and mixing on the fly within the training job, with no need for preprocessed corpus copies.

    The PyTorch blog introduces Zephon, bringing elastic deterministic data loading to TorchTitan: after training a 1B-parameter model to step 150 on 16 GPUs, training was resumed on 8, 32, and 64 GPUs, and Zephon's loss curves overlapped exactly, with the SHA-256 of every global batch matching the baseline. With the standard data loader, training the same 1B model on 20B tokens across 8 to 64 GPUs produced fluctuations of up to 0.82 points on DCLM Core v1 and 0.64 points on FineWeb, whereas Zephon

    Matched fieldsBody

    #engineering#English text is being prepared.#English text is being prepared.#English text is being prepared.

Friday

  2. Wired AI

    Should chatbots be banned from using "I" and "we"? WIRED interviews Sherry Turkle

    MIT professor Sherry Turkle warns in her new book "Artificial Intimacy: Who We Become When We Talk to Machines" that humans' excessive reliance on chatbots has become a crisis, and argues that chatbots should be banned from using the first person. She criticizes the workplace practice of "human in the loop" as merely having humans do the cleanup work of fixing model hallucinations, when the actual process has already been handed over to AI.

    Matched fieldsBody

    #tech-media#English text is being prepared.#Safety / alignment
  4. Analytics India Magazine

    NVIDIA Expands Nemotron Model Family with New Super VL and Two Bolt Models

    NVIDIA has added three models to the Nemotron open-source model family, covering enterprise workflows, local AI agents, and conversational applications across data centers, desktops, and edge devices. These include the 120-billion-parameter Nemotron 3.5 Super VL, which NVIDIA claims delivers up to 5x the throughput of the next-best comparable model, with accuracy approaching that of its largest reasoning model at roughly one-fifth the parameter count; there are also two small models in the Nemotron 3.5 Bolt series: 74B-A9B 24GB for everyday agent tasks on RTX and DGX Spark, and E4B 4GB for conversational and physical AI agents. Super VL and Bolt 74B-A9B are already available, and Bolt E4B is scheduled for release on October 23.

    NVIDIA has added three models to the Nemotron open-source model family, covering enterprise workflows, local AI agents, and conversational applications across data centers, desktops, and edge devices. These include the 120-billion-parameter Nemotron 3.5 Super VL, which NVIDIA claims delivers up to 5x the throughput of the next-best comparable model, with accuracy approaching that of its largest reasoning model at roughly one-fifth the parameter count; there are also two small models in the Nemot

    Matched fieldsBody

    #hot-signal#Model releases#Open-source ecosystem#English text is being prepared.
  8. heise online

    How the Kill-Switch architecture for AI applications works

    A heise article points out that terminating processes by PID to shut down AI agents is unreliable, because agents are not processes, and the number of running agents also changes dynamically. The article proposes a Kill-Switch architecture that combines five layers: cgroups, network and access permission revocation, permission enforcement points, an authoritative control layer, and short-lived permission tokens. However, this architecture can only guarantee that agents become ineffective after a limited time; it does not provide an instant shutdown switch.

    A heise article points out that terminating processes by PID to shut down AI agents is unreliable, because agents are not processes, and the number of running agents also changes dynamically. The article proposes a Kill-Switch architecture that combines five layers: cgroups, network and access permission revocation, permission enforcement points, an authoritative control layer, and short-lived permission tokens. However, this architecture can only guarantee that agents become ineffective after a

    Matched fieldsBody

    #hot-signal#Tutorials / practice#Agents#English text is being prepared.

Thursday