Document optimizer shard / resuming pretraining

README.md

@@ -17,3 +17,17 @@ This is a **base** (pretrained, not instruction-tuned) checkpoint at step 2162
 (final step of pretraining). `meta_002162.json` has the full model/training config.
 `tokenizer/` contains the matching BPE tokenizer (`tokenizer.pkl`) and token-byte
 table (`token_bytes.pt`) needed to reproduce BPB evaluation.
+
+## Resuming pretraining
+
+`optim_002162_rank0.pt` is the matching Muon/AdamW optimizer state (momentum
+buffers etc.) for this step. To continue pretraining, place both files under
+`$NANOCHAT_BASE_DIR/base_checkpoints/lfm2-quantum-128m/` (named `model_002162.pt`
+and `optim_002162_rank0.pt`) alongside `meta_002162.json`, then run:
+
+```bash
+RESUME_FROM_STEP=2162 MODEL_TAG=lfm2-quantum-128m bash runs/run_lfm2_quantum_mini.sh
+```
+
+For SFT/RL instead of continued pretraining, only `model_002162.pt` +
+`meta_002162.json` + `tokenizer/` are needed (the optimizer shard is pretraining-only).