Tracked topic
Qwen3 is an AI model family developed by Alibaba, released as a set of large language models for natural-language tasks.
Experimental results for pruning and distillation from Qwen3 8B using Model Optimizer show that Qwen3 Depth Pruned 6B model is 30% faster than the Qwen3 4B model, and it also performs better on the MMLU (Massive Multitask Language Understanding) benchmark. Depth pruning was applied to reduce the model from 36 to 24 layers, resulting in a 6B model, using one NVIDIA H100 80 GB HBM3. The Pruned model is distilled from Qwen3-8B using the OptimalScale/ClimbMix data processed from nvidia/ClimbMix pretraining dataset. The experiment uses 25% of the data, which is approximately 90B tokens. Distillatio
Pruning and Distilling LLMs Using NVIDIA TensorRT Model Optimizer | NVIDIA Technical Blog…500.0}) llm = LLM( model="Qwen/Qwen3-30B-A3B-Instruct-2507", sparse_attention_config=sparse_attention_config, # Other LLM arguments... ) The actual threshold value equals the threshold_scale_factor divided by the…
…llama.cpp now delivers 2x performance on Qwen 3.5 and 3.6 27B dense models, and 1.6x performance on Qwen 3.5 and 3.6 35B mixture-of-expert (MoE…
…Muon training performance on NVIDIA GB300 NVL72 Table 1 summarizes training throughput of the Kimi K2 and Qwen3 30B models with Muon and the AdamW optimizer on the NVIDIA GB300 NVL72 system…
…Engineers manually triage thousands of incidents across multiple tools, sift through a high volume of alarm and performance data, and stitch together fragmented dashboards and logs before applying a fix or dispatching…
…Qwen3 30B a3B 0.4 26,695 tokens/sec/gpu 8x B200 NVIDIA DGX B200 nemo:26.02.01 4096 1 1 1 8 FP8 512 NVIDIA B200 H100 Training Performance Framework…
Data Center / Cloud Boost Inference Performance up to 15x on NVIDIA Blackwell Using DFlash Speculative Decoding Jun 23, 2026 By Amr Elmeleegy , Benjamin Chislett , Fernando Xiong , Michael Iovine , Omri Almog and Hao…
…nvcr.io/nvidia/ai-dynamo/vllm-runtime:0.5.1 workingDir: /workspace/components/backends/vllm command: - python3 - -m - dynamo.vllm args: - --model - Qwen/Qwen3-0.6B VllmPrefillWorker: dynamoNamespace: vllm-v1-disagg-router…
…With 34 billion parameters, it leads Qwen2.5-VL (72B) by 17.0 points, Qwen3-VL (32B) by 7.0 points, Gemini 2.5 Pro by 15.1 points, and GPT-4o…
…The star_count SFT campaign trained Qwen/Qwen3-VL-2B-Instruct on the new NeMo Gym visual counting environment using LoRA SFT. We first profiled the model out of the box on…
…Qwen3-30B Similar experiments were run on mixture-of-experts (MoE) models, with results for Qwen3-30B showing matching accuracy curves. FP8 achieves similar accuracy to BF16. Speed gain is being investigated…