lars/giant
1
0
Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity

Commit Graph

  • 74012fe049 Calibrate auto batch size separately for inference vs training lars 2026-06-22 08:40:38 +02:00
  • 92d38cbed4 Add --batch-size auto to predict, matching train lars 2026-06-22 08:23:58 +02:00
  • 46068f356f Add hyperparameter scan lars 2026-06-22 07:27:02 +02:00
  • aef0a588ce Add --batch-size auto to estimate batch size from free GPU memory lars 2026-06-19 13:24:42 +02:00
  • 74d0883868 Remove scripts/train.py in favor of the giant train CLI lars 2026-06-19 13:17:15 +02:00
  • cf653f7664 Add linear warmup before cosine LR decay lars 2026-06-19 10:43:31 +02:00
  • 7f62141445 Make sampler step count configurable for validation lars 2026-06-18 17:52:14 +02:00
  • e505e2d141 Fix miniCaloSim link in README lars 2026-06-18 17:43:24 +02:00
  • 0cdb5db947 Update README to match current architecture and tooling lars 2026-06-18 17:43:04 +02:00
  • 8cebc4809d Apply ruff format and document lint/type tooling in CLAUDE.md lars 2026-06-18 17:40:27 +02:00
  • 53fd2e4405 Add ruff and ty as dev dependencies, fix lint/type findings lars 2026-06-18 17:38:40 +02:00
  • c1a8a900d3 Split torch into cpu/cuda extras and pin dependency version bounds lars 2026-06-18 17:30:35 +02:00
  • 80b198c1a5 Add lazy polars I/O and duplicate KL/constraint checks for giant.analysis lars 2026-06-18 17:06:58 +02:00
  • a867fc4aae Add giant.analysis module for notebook-based model quality diagnostics lars 2026-06-18 16:36:19 +02:00
  • da84801625 Add configurable dropout to ResBlocks lars 2026-06-18 15:59:20 +02:00
  • 5d161a52b2 Add graceful shutdown on SIGINT/SIGTERM lars 2026-06-18 14:00:01 +02:00
  • 893d91e749 Add KL divergence to marginal validation and hook it into the training loop lars 2026-06-18 13:45:12 +02:00
  • 43634ef77a Dedup training pipeline, add seeding/resume and per-epoch metrics logging lars 2026-06-18 13:32:37 +02:00
  • f1a82b5853 Add --coord local mode to predict for raw-space prediction debugging lars 2026-06-18 10:55:58 +02:00
  • 72bd65ff9f Add post_pos as a model target via travel_dir decomposition lars 2026-06-18 10:36:55 +02:00
  • c3b7b2744c Batch StreamingStepsDataset internally instead of per-row collate lars 2026-06-18 10:06:15 +02:00
  • 6e2fe12a6d Fix installed torch version to be compatible with cuda drivers lars 2026-06-18 10:06:53 +02:00
  • 9e97fb8159 Rename direction columns from pre_dir_x/y/z to pre_dx/dy/dz lars 2026-06-17 14:45:27 +02:00
  • d402cdace3 Rename pre_energy/post_energy columns to pre_E/post_E lars 2026-06-17 14:31:41 +02:00
  • 877f83a5cd Handle material column as string type lars 2026-06-17 14:29:11 +02:00
  • 78c2a61ecd Fix column names to match actual parquet schema lars 2026-06-17 14:26:57 +02:00
  • 2e4b4d91d1 Add ROOT-to-parquet conversion script with convert dependency group lars 2026-06-17 11:27:49 +02:00
  • 646a9d7a72 Add giant predict command lars 2026-06-17 11:10:27 +02:00
  • 93c4d6b74d Add streaming data pipeline and giant CLI entry point lars 2026-06-17 11:03:48 +02:00
  • 9277d79dff Implement Phase 1: full data pipeline, model, training, and config support lars 2026-06-17 10:48:03 +02:00
  • c3bf3abebf Add CLAUDE.md with architecture overview and dev commands lars 2026-06-17 09:31:45 +02:00
  • 943a044608 Initial commit: giant surrogate model with two-phase roadmap in README lars 2026-06-17 09:29:19 +02:00