Topics
Results for “moe”
25 results
- FreeToken: Efficient Edge-Native MoE Serving with Bandwidth-Adaptive Execution — Hugging Face Papers · Seen · 111 points
- The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction — Hugging Face Papers · Seen · 33 points
- fat-pig/moe-blogpost-zh — Hugging Face Spaces · Seen
- The MoE Training Handbook — 2 sources · Seen
- How MoEngage built agentic customer engagement on Amazon Bedrock AgentCore — AWS Machine Learning Blog · Seen
- Edge0/Edge0-145B-q3k — Hugging Face Models · Seen
- Show HN: DeepSeek V3 in ~200 lines of pure Jax — Hacker News · Seen · 272 points
- PhysMoE: Fast Spatial Grounding with Single-Token Coordinate Representation in Modality-Decoupled VLMs — arXiv AI Papers · Seen
- I ran 120B MoE models on my 32GB M1 Max using under 16 GB — Hacker News · Seen · 34 points
- M4 Pro vs. M5 Pro (24GB): Qwen 3.8 Flash Next at 374 Tok/S Prefill with MoEShift — Hacker News · Seen · 12 points
- Show HN: Maple – Sub-2-BPW Moe Pruning to Run Gemma 4 26B at 45 TPS on a 6GB GPU — Hacker News · Seen · 20 points
- I Retested My MoE Offloading Benchmarks. Some of My Conclusions Changed. — DEV Community · Seen
- MoE Offloading Isn’t Always Faster: I Tested 5 Models on a 16GB GPU — DEV Community · Seen
- The architecture of open-weight LLMs: MoE, hybrid layers, ternary weights — Hacker News · Seen · 16 points
- How MoEngage turned over 400 APIs into agent tools with AgentCore Gateway — AWS Machine Learning Blog · Seen
- LRMoE: Demand-Side Load Re-Calibration for Efficient MoE Training on Memory-Constrained Nodes — arXiv AI Papers · Seen
- 3 Ways to Boost MoE Inference on NVIDIA DGX Spark — NVIDIA Developer Blog · Seen
- ornith-ai/Ornith-1.5-122B — Hugging Face Models · Seen
- GMoE-FTI: A Universal Facial Template Inversion Method Adopting Genuine Mixture-of-Experts Mechanism — arXiv AI Papers · Seen
- MoEZip: Routing-Aware KV Cache Compression for Sparse Mixture-of-Experts LLMs — arXiv AI Papers · Seen
- Can a 26B MoE Run on an 8 GB GPU? Gemma 4 26B A4B on an RTX 3050 — DEV Community · Seen
- OlmoEarth v1.3: State-of-the-art embeddings for the whole planet — Hugging Face Blog · Seen
- Where Prefill Expert Reduction Hurts MoE Models: A Few Prompt Tokens Carry Most of the Short-Answer F1 Loss — arXiv AI Papers · Seen
- One Model, Multiple Compute Budgets: Multi-Ratio ConceptMoE for Flexible Compute Across Turns — arXiv AI Papers · Seen
- PRIME: Frequency-Aware Unified Pruning and Merging for MoE Compression — arXiv AI Papers · Seen