Hugging Face's logo

GODELEV
/
Exp-1

Safetensors
English
llama
causal-lm
gqa
rope
swiglu
Model card Files
xet
 Community
1
GODELEV commited on
Commit
ffd5de0
·
verified ·
1 Parent(s): afe5753

step 4800 | val_ppl=11.18

Browse files
Files changed (5) hide show
  1. README.md +3 -3
  2. model.safetensors +1 -1
  3. resume/ckpt.pt +1 -1
  4. resume/latest_step.txt +1 -1
  5. training_meta.json +5 -5
README.md CHANGED
@@ -25,9 +25,9 @@ Small language model (9.9M parameters) trained from scratch.
25
  | Total parameters | 9.853M |
26
 
27
  ## Training
28
- - Tokens seen: 4,609,228,800
29
- - Val loss: 2.4235
30
- - Val PPL: 11.29
31
 
32
  ## Usage
33
  ```python
 
25
  | Total parameters | 9.853M |
26
 
27
  ## Training
28
+ - Tokens seen: 5,028,249,600
29
+ - Val loss: 2.4145
30
+ - Val PPL: 11.18
31
 
32
  ## Usage
33
  ```python
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e7dfd16f8455d147e258af01d353cce7c411886bd6eda2312d775e534a81b404
3
  size 39421616
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e34e57b55af297cecb8a522368b79288dd25ac39a7d39b1336f31dc51769fd9
3
  size 39421616
resume/ckpt.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:74383dcbf094327f996c9d95b1eeffaa4f841de83261a26fb2b1defa406c56ec
3
  size 119812327
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e542a611020536fbbda2187c18314a69cae02cbb51fce840799a4e4e35c923e0
3
  size 119812327
resume/latest_step.txt CHANGED
@@ -1 +1 @@
1
- 4400
 
1
+ 4800
training_meta.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
- "step": 4400,
3
- "val_loss": 2.423523959350586,
4
- "val_ppl": 11.285559175233262,
5
  "params_M": 9.853,
6
- "pushed_at": "2026-06-18T05:47:07.597812",
7
- "tokens_seen": 4609228800
8
  }
 
1
  {
2
+ "step": 4800,
3
+ "val_loss": 2.4144561973571776,
4
+ "val_ppl": 11.18368698508662,
5
  "params_M": 9.853,
6
+ "pushed_at": "2026-06-18T07:47:49.668654",
7
+ "tokens_seen": 5028249600
8
  }