Models & Labs
NVIDIA NeMo AutoModel Boosts Transformers Fine-Tuning
Hugging Face Bloghigh confidence
Why it matters
- →NeMo AutoModel significantly reduces training time and memory usage for MoE models.
- →It maintains API compatibility with Hugging Face, easing adoption for developers.
- →The tool enhances scalability across multiple GPUs, crucial for large-scale AI projects.
NVIDIA has introduced NeMo AutoModel, an open library designed to enhance the fine-tuning of Transformers, especially for Mixture of Experts (MoE) models. This tool builds on Transformers v5, incorporating Expert Parallelism and DeepEP fused dispatch to deliver up to 3.7 times faster training and up to 32% less GPU memory usage. The library maintains compatibility with Hugging Face's from_pretrained() API, allowing users to benefit from these optimizations without changing their existing code. NeMo AutoModel is particularly effective in scaling MoE models across multiple GPUs, making it a valuable asset for developers working with large AI models.
Read originalMore in Models & Labs
Models & Labsmodels
llama.cpp b10955 release addresses heap corruption
The latest release of llama.cpp, b10955, tackles a critical issue of heap corruption by disabling the ggml-cpu precompiled header and fixing CACHE_LINE_SIZE ambiguity. This update ensures consistent CACHE_LINE_SIZE values across C++ kernels and C work-buffer sizing code, preventing heap-buffer-overflow and subsequent crashes. By restoring the natural include order and removing the std::hardware_destructive_interference_size branch, the update makes the value deterministic and include-order independent. This release is a technical fix that stabilizes the runtime environment for developers using llama.cpp.
llama.cpp Releases
Models & Labsmodels
llama.cpp b10956 release enhances SYCL backend
The latest llama.cpp release, b10956, introduces significant improvements to the SYCL backend, particularly for handling large k values in TOP_K operations. By implementing a radix select method, the update allows for efficient GPU-resident processing, avoiding previous limitations that forced operations to fall back to the CPU. This change enhances performance, especially in scenarios requiring large k values, such as qwen4exp's sparse-attention indexer. The update ensures that operations are more efficient and scalable, providing a notable boost in processing speed without regressing any measured shapes.