step 11600 | val_ppl=10.24

README.md

@@ -25,7 +25,7 @@ Small language model (3.0M parameters) trained from scratch.
 | Total parameters | 3.000M |
 
 ## Training
-- Tokens seen: 11,732,582,400
+- Tokens seen: 12,151,603,200
 - Val loss: 2.3259
 - Val PPL: 10.24
 

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c3f66e7342e04d75c4bc6c3fb2c95c0960e0fe8de414d8670025d951f6392208
+oid sha256:80d5f34fba78cc144e92cf8808b7d487692b95c0aee777ae30ef609b843cd967
 size 12006544

resume/ckpt.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aaec4ebc415b94989e94beae82e4baea1fdc3968c2795d6f91c8c39d9e255349
+oid sha256:e3df380bf17855206fe3e61cd551b26257f3adc8412623f9446479d00b1cc3a8
 size 36728175

resume/latest_step.txt

@@ -1 +1 @@
-11200
+11600

training_meta.json

@@ -1,8 +1,8 @@
 {
-  "step": 11200,
+  "step": 11600,
   "val_loss": 2.3259349578094484,
   "val_ppl": 10.236246071401576,
   "params_M": 3.0,
-  "pushed_at": "2026-07-03T16:12:02.381536",
-  "tokens_seen": 11732582400
+  "pushed_at": "2026-07-03T16:59:16.008421",
+  "tokens_seen": 12151603200
 }