Add final step 22327 to the evaluation table

README.md

@@ -20,9 +20,9 @@ Decoder-only language model, 1 073 912 851 parameters, Polish-dominant. **Slayer
 Slayer1B 1.06** from its final checkpoint (step 13848) on 8 891 005 707 training tokens (about 7.0 B new and 1.9 B
 replayed from the 1.06 data), at the same constant learning rate and global batch.
 
-**Status: training complete at step 22327 (6 November 2026).** The training-time evaluation table below currently
-ends at step 19000; the checkpoints after it (20000, 21000, 21450, 22000, 22327) are being evaluated with the same
-protocol, and each result is added here when it is ready. A 1.07 model release (converted weights) has not been
+**Status: training complete at step 22327 (6 November 2026).** The training-time evaluation table below includes the
+final checkpoint (22327); the checkpoints between 19000 and 22327 (20000, 21000, 21450, 22000) are being evaluated
+with the same protocol, and each result is added here when it is ready. A 1.07 model release (converted weights) has not been
 published in this repository yet. Native training checkpoints (raw PyTorch training state `ckpt_N.pt`, not a
 ready-to-use model) are published file by file in
 [SlayerLab/Slayer1B-1.07-checkpoints](https://huggingface.co/SlayerLab/Slayer1B-1.07-checkpoints), with sha256 sums.
@@ -81,18 +81,20 @@ ARC-Easy-PL eu20). 1.07 checkpoints are evaluated on CPU; the 1.06 checkpoint (s
 setup matches the 1.06 card values at card rounding. These are **not** the lm-evaluation-harness benchmark used in the
 Results table of 1.00–1.03.
 
-| metric | 1.06 (13848) | 13998 | 14000 | 15000 | 15730 | 16000 | 17000 | 18000 | 18590 | **19000** |
-|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
-| BLiMP (EN), accuracy % | 76.12 | 76.29 | 76.33 | 77.94 | 77.32 | 77.13 | 77.61 | 77.88 | 77.46 | **77.74** |
-| ARC-Easy (EN), accuracy % | 52.99 | 54.34 | 54.21 | 54.59 | 55.64 | 56.73 | 56.48 | 56.52 | 56.52 | **56.65** |
-| WikiText-2, byte perplexity ↓ | 2.2849 | 2.2618 | 2.2601 | 2.2321 | 2.2237 | 2.2205 | 2.2138 | 2.2042 | 2.2145 | **2.2029** |
-| WikiText-2, token perplexity ↓ | 30.03 | 28.80 | 28.71 | 27.28 | 26.85 | 26.70 | 26.37 | 25.90 | 26.40 | **25.84** |
-| MultiBLiMP (PL), accuracy % | 98.41 | 98.66 | 98.59 | 98.62 | 98.62 | 98.81 | 98.87 | 98.84 | 98.62 | **98.53** |
-| ARC-Easy (PL, eu20), accuracy % | 45.03 | 45.37 | 44.91 | 45.92 | 45.75 | 46.25 | 45.41 | 46.21 | 47.14 | **46.80** |
+| metric | 1.06 (13848) | 13998 | 14000 | 15000 | 15730 | 16000 | 17000 | 18000 | 18590 | 19000 | **22327 (final)** |
+|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
+| BLiMP (EN), accuracy % | 76.12 | 76.29 | 76.33 | 77.94 | 77.32 | 77.13 | 77.61 | 77.88 | 77.46 | 77.74 | **77.90** |
+| ARC-Easy (EN), accuracy % | 52.99 | 54.34 | 54.21 | 54.59 | 55.64 | 56.73 | 56.48 | 56.52 | 56.52 | 56.65 | **58.67** |
+| WikiText-2, byte perplexity ↓ | 2.2849 | 2.2618 | 2.2601 | 2.2321 | 2.2237 | 2.2205 | 2.2138 | 2.2042 | 2.2145 | 2.2029 | **2.1837** |
+| WikiText-2, token perplexity ↓ | 30.03 | 28.80 | 28.71 | 27.28 | 26.85 | 26.70 | 26.37 | 25.90 | 26.40 | 25.84 | **24.92** |
+| MultiBLiMP (PL), accuracy % | 98.41 | 98.66 | 98.59 | 98.62 | 98.62 | 98.81 | 98.87 | 98.84 | 98.62 | 98.53 | **98.84** |
+| ARC-Easy (PL, eu20), accuracy % | 45.03 | 45.37 | 44.91 | 45.92 | 45.75 | 46.25 | 45.41 | 46.21 | 47.14 | 46.80 | **46.93** |
 
 ARC-Easy has 2 376 questions per language, so differences below about one percentage point between neighbouring
-checkpoints are within noise. WikiText-2 byte perplexity at 18590 was 0.010 higher than at 18000; at 19000 it is the
-lowest so far (2.2029). Each point is a single measurement of a constant-learning-rate run. The training loss in the figure is measured
+checkpoints are within noise. WikiText-2 byte perplexity at 18590 was 0.010 higher than at 18000. The final checkpoint
+(22327) has the lowest WikiText-2 perplexity and the highest English ARC-Easy accuracy of the run (+2.0 points over
+19000, together with small gains on all other metrics); the checkpoints in between are not measured yet, so the chart
+joins 19000 and 22327 directly. Each point is a single measurement of a constant-learning-rate run. The training loss in the figure is measured
 on the 1.07 data mixture (44 % English), so it is not directly comparable with the training loss of 1.00–1.06.
 
 The benchmark caveat of the 1.03 card applies to the 1.00–1.06 part of the training. For 1.07, all nine data files