This website requires JavaScript.
Explore
Help
Register
Sign In
lars
/
giant
Watch
1
Star
0
Fork
0
You've already forked giant
Code
Issues
4
Pull Requests
2
Actions
Packages
Projects
Releases
Wiki
Activity
Files
6744d05948bc0e3a58b03a3b366c2caa8419fbaf
giant
/
giant
T
History
lars
6744d05948
Merge branch '4-prototype-a-mixture-of-experts-routing-tree-architecture' into analysis-streaming-rewrite
...
# Conflicts: # giant/analysis.py # tests/test_analysis.py
2026-07-17 12:06:10 +02:00
..
data
Apply ruff format after merging phase2-secondary-prediction
2026-07-15 10:00:36 +02:00
model
Add ComposedRouter for multi-axis MoE gating
2026-07-15 13:53:21 +02:00
__init__.py
Initial commit: giant surrogate model with two-phase roadmap in README
2026-06-17 09:29:19 +02:00
analysis.py
Rescale secondary energies to exactly consume the e_sec budget
2026-07-13 13:45:13 +02:00
cli.py
Add EMA weights, weight decay, step-based LR schedule, and grad-norm logging to training
2026-07-17 11:02:12 +02:00
config.py
Add EMA weights, weight decay, step-based LR schedule, and grad-norm logging to training
2026-07-17 11:02:12 +02:00
constants.py
Apply ruff format across the codebase
2026-07-08 14:44:53 +02:00
geometry.py
Add fast slab lookup for the GeometryOracle, replacing knn as the default
2026-07-08 11:42:12 +02:00
pipeline.py
Add EMA weights, weight decay, step-based LR schedule, and grad-norm logging to training
2026-07-17 11:02:12 +02:00
rollout.py
Stream giant rollout output instead of buffering the whole run
2026-07-15 11:10:04 +02:00
sample.py
Apply ruff format across the codebase
2026-07-08 14:44:53 +02:00
train.py
Add EMA weights, weight decay, step-based LR schedule, and grad-norm logging to training
2026-07-17 11:02:12 +02:00
validate.py
Add ProcessRouter for physics-process-based expert gating
2026-07-08 16:15:09 +02:00