last 2026-10-25 16:22:51
Browse files- last/ckpt.pt +1 -1
- last/state.json +1 -1
- logs/train_log.jsonl +9 -0
last/ckpt.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 329499691
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ef67ac536c22668245114c445bc737bc055888eefc9dd72858c1e1a647a7db0
|
| 3 |
size 329499691
|
last/state.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"step":
|
|
|
|
| 1 |
+
{"step": 10000, "best_val": 0.923981524555829, "tokens": 132070451, "time": 1792945371.4815211}
|
logs/train_log.jsonl
CHANGED
|
@@ -209,3 +209,12 @@
|
|
| 209 |
{"step": 9500, "val": 0.9355872968198322}
|
| 210 |
{"step": 9550, "loss": 0.9577423483133316, "lr": 0.00030032065015668627, "tok_s": 9534.848452471158}
|
| 211 |
{"step": 9600, "loss": 0.9909506291151047, "lr": 0.0002985189042352221, "tok_s": 12062.44303286884}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 209 |
{"step": 9500, "val": 0.9355872968198322}
|
| 210 |
{"step": 9550, "loss": 0.9577423483133316, "lr": 0.00030032065015668627, "tok_s": 9534.848452471158}
|
| 211 |
{"step": 9600, "loss": 0.9909506291151047, "lr": 0.0002985189042352221, "tok_s": 12062.44303286884}
|
| 212 |
+
{"step": 9650, "loss": 0.8807881623506546, "lr": 5.150541507043142e-05, "tok_s": 12206.208918833974}
|
| 213 |
+
{"step": 9700, "loss": 0.8910769075155258, "lr": 5.110634692737243e-05, "tok_s": 10753.538505919938}
|
| 214 |
+
{"step": 9750, "loss": 0.8895992785692215, "lr": 5.076848907349929e-05, "tok_s": 11421.998198297353}
|
| 215 |
+
{"step": 9800, "loss": 0.8433485478162766, "lr": 5.049193387577573e-05, "tok_s": 10341.461850629155}
|
| 216 |
+
{"step": 9850, "loss": 0.8942935466766357, "lr": 5.0276756941631436e-05, "tok_s": 11306.013554883324}
|
| 217 |
+
{"step": 9900, "loss": 0.9275110810995102, "lr": 5.012301709829164e-05, "tok_s": 11747.883706671695}
|
| 218 |
+
{"step": 9950, "loss": 0.8948111087083817, "lr": 5.003075637669448e-05, "tok_s": 11318.57667123224}
|
| 219 |
+
{"step": 10000, "loss": 0.8525521159172058, "lr": 5e-05, "tok_s": 11364.966562504143}
|
| 220 |
+
{"step": 10000, "val": 0.923981524555829}
|