Update experiment measurements and status
Browse files
reports/configs/dclm_edu-1337.toml
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[model]
|
| 2 |
+
vocab = 12288
|
| 3 |
+
n_layer = 19
|
| 4 |
+
n_embd = 768
|
| 5 |
+
n_head = 12
|
| 6 |
+
block = 1024
|
| 7 |
+
ffn_mult = 2.8125
|
| 8 |
+
[train]
|
| 9 |
+
seed = 1337
|
| 10 |
+
micro_batch = 64
|
| 11 |
+
global_batch = 512
|
| 12 |
+
steps = 1000
|
| 13 |
+
warmup = 50
|
| 14 |
+
lr = 0.0001
|
| 15 |
+
min_lr = 0.00001
|
| 16 |
+
muon_lr = 0.0033333333333333335
|
| 17 |
+
checkpoint_every = 500
|
| 18 |
+
validate_every = 500
|
| 19 |
+
log_every = 20
|
| 20 |
+
compile = true
|
| 21 |
+
[data]
|
| 22 |
+
train = "/data/ablation/mixtures/dclm_edu/train.bin"
|
| 23 |
+
validation = "/data/final/val.bin"
|
| 24 |
+
sha256 = "77afd3772e46ab001b452017857fe0cf2e2c72a7566865e7fece386091c8686d"
|
reports/eval/cosmopedia-1337-1000.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|