Dataset Viewer
Auto-converted to Parquet Duplicate
groups
list
schema
string
sequence_function_without_structure_trainable
int64
totals
dict
trimodal_structure_candidate
int64
[ { "file_column_presence": { "annotation_text": 374, "atom_modalities": 242, "atom_symbols": 123, "b_factors": 119, "backbone_coordinates": 248, "ca_coordinates": 248, "chain_ids": 490, "compound_id": 123, "content_hash": 374, "convextok_dag_json": 374,...
toricblm.training_corpus_audit.v1
87,960,907
{ "row_class_counts": { "graph_fot_trainable": 13227, "sequence_function_without_structure_trainable": 87960907, "sequence_only_trainable": 11677608, "structure_graph": 957804, "trimodal_structure_candidate": 808578 }, "rows_total": 101418124 }
808,578

ToricBLM dataset state: toricblm-structure-priority-balanced-3day-20260709T185034Z epoch 001

This dataset repo records the exact local training-data state visible to the dynamic epoch launcher. It intentionally stores manifests and audit records rather than duplicating large Parquet shards.

  • Special checkpoint: toricblm-structure-priority-balanced-3day-20260709T185034Z_epoch_001_special_structure_current_step_002000.pt
  • Checkpoint repo: AmelieSchreiber/ToricGT_160M_FoT
  • Curriculum mode: structure_current
  • Selected rows: 1382903
  • Selected files: 612
  • Structure-flow rows: 883191
  • Coverage steps: 2000

Files:

  • available_shards.json: refreshed dataset-state snapshot for this epoch
  • training_corpus_audit.json: sampled corpus audit produced before training
  • *_epoch_*.env / epoch_data_overrides.env: resolved epoch launch configuration
  • dynamic_epoch_summary.tsv: epoch handoff/checkpoint summary
  • train_log_tail.txt: tail of the epoch training log
Downloads last month
80