Nov 13 7 days ago 18:55 Split the architecture figure from the pruning figure The model is now its own figure (assets/architecture.png), with one transformer layer opened up; assets/pruning.png shows only the before-and-after pruning comparison. 18:43 Redraw the architecture figure The figure is redrawn in the Netra Lab house style with the layers opened up; the alt text describes it. Nov 6 14 days ago 10:10 Add v1 evaluation results (student, fine-tuned teacher, stock 1B, inference time) Oct 25 26 days ago 11:23 Card: inference time at 0% vs 80% sparsity 10:58 Card: Figure 2 as teacher vs SOTA bars 10:56 Card: redesigned speed/error figure 10:50 Card: pruning diagram and speed/error trade-off figure 10:42 Card: sota package usage, GitHub link, DDD naming Oct 23 28 days ago 15:04 inference.py: put the bundled omniprune package on sys.path 15:03 Release pruned (80% sparsity) Khmer + English omniASR CTC model 14:59 initial commit