step 4400 | val_ppl=11.29

README.md

@@ -25,9 +25,9 @@ Small language model (9.9M parameters) trained from scratch.
 | Total parameters | 9.853M |
 
 ## Training
-- Tokens seen: 4,190,208,000
-- Val loss: 2.4361
-- Val PPL: 11.43
+- Tokens seen: 4,609,228,800
+- Val loss: 2.4235
+- Val PPL: 11.29
 
 ## Usage
 ```python

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3a623c1f904bdb9f46ed0718c33f1996a9db6692de446fea9b322554bff206d3
+oid sha256:e7dfd16f8455d147e258af01d353cce7c411886bd6eda2312d775e534a81b404
 size 39421616

resume/ckpt.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:445874cf6b1f4aafe7ac6cc1436c8664c25eae2577357c64d852ca181d1add5e
+oid sha256:74383dcbf094327f996c9d95b1eeffaa4f841de83261a26fb2b1defa406c56ec
 size 119812327

resume/latest_step.txt

@@ -1 +1 @@
-4000
+4400

training_meta.json

@@ -1,8 +1,8 @@
 {
-  "step": 4000,
-  "val_loss": 2.4361114742279053,
-  "val_ppl": 11.428514155678789,
+  "step": 4400,
+  "val_loss": 2.423523959350586,
+  "val_ppl": 11.285559175233262,
   "params_M": 9.853,
-  "pushed_at": "2026-06-18T02:22:12.443456",
-  "tokens_seen": 4190208000
+  "pushed_at": "2026-06-18T05:47:07.597812",
+  "tokens_seen": 4609228800
 }