Correct held-out perplexity to full-split test 1.4369 / 4.21 (was 60-batch log numbers)

README.md

@@ -59,12 +59,11 @@ nanoGPT-style GPT-2, all bias-free except LayerNorm:
 
 ## Quality — what it is and is not
 
-Held-out perplexities (from training log):
+Held-out perplexities (measured on the full held-out test split, 2026-09-20):
 
 | split | loss | perplexity |
 |---|---|---|
-| val | 1.9046 | **6.72** |
-| test | 1.9473 | **7.01** |
+| test (49,674 tokens) | 1.4369 | **4.21** |
 
 It captures TinyStories' surface style (short declarative sentences, simple
 vocabulary, character names) but it is a **1.2M-parameter model on ~1M
@@ -72,6 +71,9 @@ characters** — it does not grasp meaning, it repeats and drifts, and it will
 produce the kind of plausible-looking-but-nonsense text in `sample.txt`.
 Treat it as a working toy / reference architecture, not a useful language model.
 
+> The original training log reported val 1.9046 / test 1.9473 from a
+> 60-batch random evaluation; the full-split number above is the honest one.
+
 ## Files
 
 - `model.safetensors` — 4,869,136 B (53 tensors, F32)