step 200 | val_ppl=45.49

README.md

@@ -25,9 +25,9 @@ Small language model (9.9M parameters) trained from scratch.
 | Total parameters | 9.902M |
 
 ## Training
-- Tokens seen: 141,419,520
-- Val loss: inf
-- Val PPL: inf
+- Tokens seen: 471,398,400
+- Val loss: 3.8175
+- Val PPL: 45.49
 
 ## Usage
 ```python

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b687326d79592e4596afff27ed591d6689f3605fbafa7b9ff0532f3d24141a28
+oid sha256:9de947ef8478ee61b29d5959a37907690d3db9ebc45e79a16a34f79de35d9c64
 size 39621248

resume/ckpt.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:11db9c36edb1cff345de6ddfa015e5b39a6a54bf19994ecf61f4857f2dc1337d
-size 122094365
+oid sha256:b790d48ea299b83280b7df95f2be2e0040b3d52dfec1a8e1c905a1a0e235ea53
+size 122094685

resume/latest_step.txt

@@ -1 +1 @@
-60
+200

training_meta.json

@@ -1,8 +1,8 @@
 {
-  "step": 60,
-  "val_loss": Infinity,
-  "val_ppl": Infinity,
+  "step": 200,
+  "val_loss": 3.8174806594848634,
+  "val_ppl": 45.48946039640565,
   "params_M": 9.902,
-  "pushed_at": "2026-06-11T06:10:41.431405",
-  "tokens_seen": 141419520
+  "pushed_at": "2026-06-11T07:42:55.482805",
+  "tokens_seen": 471398400
 }