-
Notifications
You must be signed in to change notification settings - Fork 3.3k
Pull requests: kyegomez/OpenMythos
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
Add test coverage for open_mythos/moda.py (MoDA + DeepSeek MoE)
#89
opened Aug 23, 2026 by
helpdeskdcp
Loading…
4 tasks done
Fix LTIInjection.get_A() float32 underflow at clamp boundary
#88
opened Aug 23, 2026 by
helpdeskdcp
Loading…
3 tasks done
Fix RoPE broadcast mismatch in apply_rope
#87
opened Aug 23, 2026 by
helpdeskdcp
Loading…
3 tasks done
Fix float32 precision leak in LTI stability guarantee and stale RoPE test contract
#86
opened Aug 7, 2026 by
kelseyjmccorkle-stack
Loading…
2 tasks done
fix: stabilize deep recurrence (T>=32) and correct scaling strategy
#80
opened May 23, 2026 by
skyhighbg22-jpg
Loading…
feat: Finance domain adapters + GGUF export for local inference
#77
opened May 20, 2026 by
oyi77
Loading…
feat: Ring Attention + KV Cache compression for 1M context on consumer hardware
#76
opened May 20, 2026 by
oyi77
Loading…
feat: LoRA training pipeline + Colab notebook for free GPU fine-tuning
#75
opened May 20, 2026 by
oyi77
Loading…
feat: INT4/INT8 quantization + expert offloading for consumer hardware
#74
opened May 20, 2026 by
oyi77
Loading…
Improvements pack: torch pin fix, eigvals docs fix, CI + tests + TINY_CONFIG + getting started
#73
opened May 16, 2026 by
0riginal-claw
Loading…
7 tasks done
docs: add CLAUDE.md with architecture guide and development commands
#69
opened May 5, 2026 by
Feibo112358
Loading…
feat: MLX-native training stack and upstream MythosConfig compatibility
#66
opened May 1, 2026 by
santome5954-lang
Loading…
4 tasks
100x Enhancement: Vectorized MoE, Training Framework, Benchmarking Suite
#64
opened Apr 28, 2026 by
protonpunkz
Loading…
feat: Add latent reasoning trajectory extraction (return_latents)
#62
opened Apr 25, 2026 by
triunex
Loading…
docs(readme): fix spectral-radius example for 1-D diagonal A
#60
opened Apr 25, 2026 by
NFTAC
Loading…
3 tasks done
Add ablation flags (disable_act, break_recurrence, etc.) to enable depth-extrapolation training
#58
opened Apr 23, 2026 by
tonyzdev
Loading…
3 tasks done
Model improvements: FSDP dtype + MoE dispatch + ACT deadlock fix + stochastic-depth (Option B)
#56
opened Apr 23, 2026 by
pzerfos
Loading…
3 tasks done
docs: fix incorrect spectral radius computation in README example
#50
opened Apr 22, 2026 by
sunnyc0206
Loading…
feat: upgrade MiniMax default model to M3 (512K context, 128K output)
#49
opened Apr 22, 2026 by
octo-patch
Loading…
Previous Next
ProTip!
Updated in the last three days: updated:>2026-09-01.