Upload checkpoint-4156
Browse files
checkpoint-4156/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:008dd09a3d6e4dcf91cb80e9cfcc39b67b01bdce6650a50d3d1c27328fc70126
|
| 3 |
+
size 16389
|
checkpoint-4156/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f5ef6becba428d8cddfca6fb712d0c32c6dc017a8b8e17b615a89a81176ea927
|
| 3 |
+
size 1465
|
checkpoint-4156/trainer_state.json
ADDED
|
@@ -0,0 +1,1235 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 4156,
|
| 3 |
+
"best_metric": 2.6353492736816406,
|
| 4 |
+
"best_model_checkpoint": "outputs/chess-explainer/checkpoint-4156",
|
| 5 |
+
"epoch": 4.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 4156,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0009624639076034649,
|
| 14 |
+
"grad_norm": 883.1925048828125,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 17.491071701049805,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02406159769008662,
|
| 21 |
+
"grad_norm": 55.103694915771484,
|
| 22 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 23 |
+
"loss": 11.750035603841146,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.04812319538017324,
|
| 28 |
+
"grad_norm": 9.345708847045898,
|
| 29 |
+
"learning_rate": 3.141025641025641e-05,
|
| 30 |
+
"loss": 4.889132080078125,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07218479307025986,
|
| 35 |
+
"grad_norm": 10.196633338928223,
|
| 36 |
+
"learning_rate": 4.7435897435897435e-05,
|
| 37 |
+
"loss": 3.7856689453125,
|
| 38 |
+
"step": 75
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09624639076034648,
|
| 42 |
+
"grad_norm": 8.611700057983398,
|
| 43 |
+
"learning_rate": 6.346153846153847e-05,
|
| 44 |
+
"loss": 3.5517022705078123,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.12030798845043311,
|
| 49 |
+
"grad_norm": 7.812922477722168,
|
| 50 |
+
"learning_rate": 7.948717948717948e-05,
|
| 51 |
+
"loss": 3.4541973876953125,
|
| 52 |
+
"step": 125
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.14436958614051973,
|
| 56 |
+
"grad_norm": 7.820157527923584,
|
| 57 |
+
"learning_rate": 9.551282051282052e-05,
|
| 58 |
+
"loss": 3.3625674438476563,
|
| 59 |
+
"step": 150
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.16843118383060635,
|
| 63 |
+
"grad_norm": 7.358847141265869,
|
| 64 |
+
"learning_rate": 0.00011153846153846154,
|
| 65 |
+
"loss": 3.33571533203125,
|
| 66 |
+
"step": 175
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.19249278152069296,
|
| 70 |
+
"grad_norm": 7.274508953094482,
|
| 71 |
+
"learning_rate": 0.00012756410256410257,
|
| 72 |
+
"loss": 3.3122216796875,
|
| 73 |
+
"step": 200
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2165543792107796,
|
| 77 |
+
"grad_norm": 7.056718349456787,
|
| 78 |
+
"learning_rate": 0.0001435897435897436,
|
| 79 |
+
"loss": 3.28180419921875,
|
| 80 |
+
"step": 225
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.24061597690086622,
|
| 84 |
+
"grad_norm": 6.19630241394043,
|
| 85 |
+
"learning_rate": 0.00015961538461538462,
|
| 86 |
+
"loss": 3.2475381469726563,
|
| 87 |
+
"step": 250
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2646775745909528,
|
| 91 |
+
"grad_norm": 6.5128397941589355,
|
| 92 |
+
"learning_rate": 0.00017564102564102566,
|
| 93 |
+
"loss": 3.234718017578125,
|
| 94 |
+
"step": 275
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.28873917228103946,
|
| 98 |
+
"grad_norm": 6.359344482421875,
|
| 99 |
+
"learning_rate": 0.00019166666666666667,
|
| 100 |
+
"loss": 3.2271856689453124,
|
| 101 |
+
"step": 300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.3128007699711261,
|
| 105 |
+
"grad_norm": 6.695319175720215,
|
| 106 |
+
"learning_rate": 0.0001999993003464737,
|
| 107 |
+
"loss": 3.1996661376953126,
|
| 108 |
+
"step": 325
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.3368623676612127,
|
| 112 |
+
"grad_norm": 5.9124321937561035,
|
| 113 |
+
"learning_rate": 0.00019999334849877745,
|
| 114 |
+
"loss": 3.1934442138671875,
|
| 115 |
+
"step": 350
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.36092396535129934,
|
| 119 |
+
"grad_norm": 5.623684406280518,
|
| 120 |
+
"learning_rate": 0.00019998132369740206,
|
| 121 |
+
"loss": 3.1889007568359373,
|
| 122 |
+
"step": 375
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.3849855630413859,
|
| 126 |
+
"grad_norm": 5.664840221405029,
|
| 127 |
+
"learning_rate": 0.00019996322667265672,
|
| 128 |
+
"loss": 3.173824462890625,
|
| 129 |
+
"step": 400
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.40904716073147257,
|
| 133 |
+
"grad_norm": 6.227374076843262,
|
| 134 |
+
"learning_rate": 0.0001999390585236385,
|
| 135 |
+
"loss": 3.15881591796875,
|
| 136 |
+
"step": 425
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.4331087584215592,
|
| 140 |
+
"grad_norm": 5.214109420776367,
|
| 141 |
+
"learning_rate": 0.0001999088207181655,
|
| 142 |
+
"loss": 3.09910888671875,
|
| 143 |
+
"step": 450
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.4571703561116458,
|
| 147 |
+
"grad_norm": 5.23874568939209,
|
| 148 |
+
"learning_rate": 0.0001998725150926878,
|
| 149 |
+
"loss": 3.1576531982421874,
|
| 150 |
+
"step": 475
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.48123195380173245,
|
| 154 |
+
"grad_norm": 5.108809471130371,
|
| 155 |
+
"learning_rate": 0.0001998301438521759,
|
| 156 |
+
"loss": 3.1337042236328125,
|
| 157 |
+
"step": 500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.5052935514918191,
|
| 161 |
+
"grad_norm": 5.708388805389404,
|
| 162 |
+
"learning_rate": 0.00019978170956998676,
|
| 163 |
+
"loss": 3.128594970703125,
|
| 164 |
+
"step": 525
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.5293551491819056,
|
| 168 |
+
"grad_norm": 5.174871921539307,
|
| 169 |
+
"learning_rate": 0.00019972721518770757,
|
| 170 |
+
"loss": 3.1022515869140626,
|
| 171 |
+
"step": 550
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.5534167468719923,
|
| 175 |
+
"grad_norm": 5.113072872161865,
|
| 176 |
+
"learning_rate": 0.00019966666401497705,
|
| 177 |
+
"loss": 3.063781433105469,
|
| 178 |
+
"step": 575
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.5774783445620789,
|
| 182 |
+
"grad_norm": 5.224897384643555,
|
| 183 |
+
"learning_rate": 0.00019960005972928445,
|
| 184 |
+
"loss": 3.0889013671875,
|
| 185 |
+
"step": 600
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.6015399422521656,
|
| 189 |
+
"grad_norm": 5.108465671539307,
|
| 190 |
+
"learning_rate": 0.00019952740637574633,
|
| 191 |
+
"loss": 3.0858636474609376,
|
| 192 |
+
"step": 625
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.6256015399422522,
|
| 196 |
+
"grad_norm": 5.350228786468506,
|
| 197 |
+
"learning_rate": 0.00019944870836686063,
|
| 198 |
+
"loss": 3.0400531005859377,
|
| 199 |
+
"step": 650
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.6496631376323387,
|
| 203 |
+
"grad_norm": 4.423816204071045,
|
| 204 |
+
"learning_rate": 0.0001993639704822389,
|
| 205 |
+
"loss": 3.06386474609375,
|
| 206 |
+
"step": 675
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.6737247353224254,
|
| 210 |
+
"grad_norm": 5.499719619750977,
|
| 211 |
+
"learning_rate": 0.00019927319786831594,
|
| 212 |
+
"loss": 3.0412890625,
|
| 213 |
+
"step": 700
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.697786333012512,
|
| 217 |
+
"grad_norm": 4.638278007507324,
|
| 218 |
+
"learning_rate": 0.0001991763960380373,
|
| 219 |
+
"loss": 3.0481976318359374,
|
| 220 |
+
"step": 725
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.7218479307025987,
|
| 224 |
+
"grad_norm": 5.008415222167969,
|
| 225 |
+
"learning_rate": 0.00019907357087052424,
|
| 226 |
+
"loss": 3.0042901611328126,
|
| 227 |
+
"step": 750
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.7459095283926853,
|
| 231 |
+
"grad_norm": 4.793009281158447,
|
| 232 |
+
"learning_rate": 0.00019896472861071702,
|
| 233 |
+
"loss": 2.9917413330078126,
|
| 234 |
+
"step": 775
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.7699711260827719,
|
| 238 |
+
"grad_norm": 5.019183158874512,
|
| 239 |
+
"learning_rate": 0.0001988498758689953,
|
| 240 |
+
"loss": 3.0368035888671874,
|
| 241 |
+
"step": 800
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.7940327237728585,
|
| 245 |
+
"grad_norm": 4.315158843994141,
|
| 246 |
+
"learning_rate": 0.00019872901962077687,
|
| 247 |
+
"loss": 2.9863958740234375,
|
| 248 |
+
"step": 825
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.8180943214629451,
|
| 252 |
+
"grad_norm": 5.23874044418335,
|
| 253 |
+
"learning_rate": 0.00019860216720609395,
|
| 254 |
+
"loss": 2.9992340087890623,
|
| 255 |
+
"step": 850
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.8421559191530318,
|
| 259 |
+
"grad_norm": 4.575445652008057,
|
| 260 |
+
"learning_rate": 0.00019846932632914733,
|
| 261 |
+
"loss": 3.0112255859375,
|
| 262 |
+
"step": 875
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.8662175168431184,
|
| 266 |
+
"grad_norm": 4.955294609069824,
|
| 267 |
+
"learning_rate": 0.0001983305050578386,
|
| 268 |
+
"loss": 3.0229232788085936,
|
| 269 |
+
"step": 900
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.890279114533205,
|
| 273 |
+
"grad_norm": 4.7536773681640625,
|
| 274 |
+
"learning_rate": 0.00019818571182328003,
|
| 275 |
+
"loss": 3.0158026123046877,
|
| 276 |
+
"step": 925
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.9143407122232916,
|
| 280 |
+
"grad_norm": 4.718550205230713,
|
| 281 |
+
"learning_rate": 0.00019803495541928262,
|
| 282 |
+
"loss": 2.983572998046875,
|
| 283 |
+
"step": 950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.9384023099133783,
|
| 287 |
+
"grad_norm": 4.429309844970703,
|
| 288 |
+
"learning_rate": 0.000197878245001822,
|
| 289 |
+
"loss": 2.9714495849609377,
|
| 290 |
+
"step": 975
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.9624639076034649,
|
| 294 |
+
"grad_norm": 4.627259254455566,
|
| 295 |
+
"learning_rate": 0.00019771559008848223,
|
| 296 |
+
"loss": 2.961048889160156,
|
| 297 |
+
"step": 1000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.9865255052935515,
|
| 301 |
+
"grad_norm": 5.209016799926758,
|
| 302 |
+
"learning_rate": 0.00019754700055787797,
|
| 303 |
+
"loss": 2.9716583251953126,
|
| 304 |
+
"step": 1025
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.0,
|
| 308 |
+
"eval_loss": 2.934025287628174,
|
| 309 |
+
"eval_runtime": 12.8238,
|
| 310 |
+
"eval_samples_per_second": 116.97,
|
| 311 |
+
"eval_steps_per_second": 1.872,
|
| 312 |
+
"step": 1039
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.0105871029836382,
|
| 316 |
+
"grad_norm": 4.551164150238037,
|
| 317 |
+
"learning_rate": 0.00019737248664905435,
|
| 318 |
+
"loss": 2.9250759887695312,
|
| 319 |
+
"step": 1050
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.0346487006737248,
|
| 323 |
+
"grad_norm": 5.032167911529541,
|
| 324 |
+
"learning_rate": 0.00019719205896086522,
|
| 325 |
+
"loss": 2.8127606201171873,
|
| 326 |
+
"step": 1075
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.0587102983638113,
|
| 330 |
+
"grad_norm": 5.156538009643555,
|
| 331 |
+
"learning_rate": 0.00019700572845132936,
|
| 332 |
+
"loss": 2.836297607421875,
|
| 333 |
+
"step": 1100
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.082771896053898,
|
| 337 |
+
"grad_norm": 5.519596099853516,
|
| 338 |
+
"learning_rate": 0.00019681350643696506,
|
| 339 |
+
"loss": 2.84802490234375,
|
| 340 |
+
"step": 1125
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.1068334937439845,
|
| 344 |
+
"grad_norm": 4.649038314819336,
|
| 345 |
+
"learning_rate": 0.00019661540459210271,
|
| 346 |
+
"loss": 2.8203363037109375,
|
| 347 |
+
"step": 1150
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.1308950914340712,
|
| 351 |
+
"grad_norm": 4.99331521987915,
|
| 352 |
+
"learning_rate": 0.00019641143494817582,
|
| 353 |
+
"loss": 2.862440185546875,
|
| 354 |
+
"step": 1175
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.1549566891241578,
|
| 358 |
+
"grad_norm": 4.924983501434326,
|
| 359 |
+
"learning_rate": 0.0001962016098929904,
|
| 360 |
+
"loss": 2.8315557861328124,
|
| 361 |
+
"step": 1200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.1790182868142445,
|
| 365 |
+
"grad_norm": 5.165991306304932,
|
| 366 |
+
"learning_rate": 0.0001959859421699725,
|
| 367 |
+
"loss": 2.8693508911132812,
|
| 368 |
+
"step": 1225
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.2030798845043311,
|
| 372 |
+
"grad_norm": 4.788819313049316,
|
| 373 |
+
"learning_rate": 0.00019576444487739425,
|
| 374 |
+
"loss": 2.836847229003906,
|
| 375 |
+
"step": 1250
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.2271414821944178,
|
| 379 |
+
"grad_norm": 5.095317840576172,
|
| 380 |
+
"learning_rate": 0.0001955371314675784,
|
| 381 |
+
"loss": 2.8328265380859374,
|
| 382 |
+
"step": 1275
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.2512030798845044,
|
| 386 |
+
"grad_norm": 5.162815570831299,
|
| 387 |
+
"learning_rate": 0.00019530401574608134,
|
| 388 |
+
"loss": 2.8390570068359375,
|
| 389 |
+
"step": 1300
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.275264677574591,
|
| 393 |
+
"grad_norm": 4.838395595550537,
|
| 394 |
+
"learning_rate": 0.00019506511187085456,
|
| 395 |
+
"loss": 2.855579833984375,
|
| 396 |
+
"step": 1325
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.2993262752646775,
|
| 400 |
+
"grad_norm": 4.820464611053467,
|
| 401 |
+
"learning_rate": 0.00019482043435138478,
|
| 402 |
+
"loss": 2.835367431640625,
|
| 403 |
+
"step": 1350
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.3233878729547641,
|
| 407 |
+
"grad_norm": 4.540008544921875,
|
| 408 |
+
"learning_rate": 0.00019456999804781297,
|
| 409 |
+
"loss": 2.822310485839844,
|
| 410 |
+
"step": 1375
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.3474494706448508,
|
| 414 |
+
"grad_norm": 4.731804847717285,
|
| 415 |
+
"learning_rate": 0.00019431381817003142,
|
| 416 |
+
"loss": 2.8431842041015627,
|
| 417 |
+
"step": 1400
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.3715110683349374,
|
| 421 |
+
"grad_norm": 4.5690741539001465,
|
| 422 |
+
"learning_rate": 0.00019405191027676034,
|
| 423 |
+
"loss": 2.8410440063476563,
|
| 424 |
+
"step": 1425
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.395572666025024,
|
| 428 |
+
"grad_norm": 4.800422668457031,
|
| 429 |
+
"learning_rate": 0.0001937842902746029,
|
| 430 |
+
"loss": 2.8300006103515627,
|
| 431 |
+
"step": 1450
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.4196342637151107,
|
| 435 |
+
"grad_norm": 5.346854209899902,
|
| 436 |
+
"learning_rate": 0.00019351097441707883,
|
| 437 |
+
"loss": 2.8334808349609375,
|
| 438 |
+
"step": 1475
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.4436958614051973,
|
| 442 |
+
"grad_norm": 5.628939628601074,
|
| 443 |
+
"learning_rate": 0.00019323197930363776,
|
| 444 |
+
"loss": 2.8730230712890625,
|
| 445 |
+
"step": 1500
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.467757459095284,
|
| 449 |
+
"grad_norm": 4.565985202789307,
|
| 450 |
+
"learning_rate": 0.00019294732187865072,
|
| 451 |
+
"loss": 2.856542053222656,
|
| 452 |
+
"step": 1525
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.4918190567853706,
|
| 456 |
+
"grad_norm": 5.309317111968994,
|
| 457 |
+
"learning_rate": 0.0001926570194303812,
|
| 458 |
+
"loss": 2.797247314453125,
|
| 459 |
+
"step": 1550
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.515880654475457,
|
| 463 |
+
"grad_norm": 5.3165693283081055,
|
| 464 |
+
"learning_rate": 0.00019236108958993519,
|
| 465 |
+
"loss": 2.81298828125,
|
| 466 |
+
"step": 1575
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.539942252165544,
|
| 470 |
+
"grad_norm": 4.712189674377441,
|
| 471 |
+
"learning_rate": 0.0001920595503301902,
|
| 472 |
+
"loss": 2.82627685546875,
|
| 473 |
+
"step": 1600
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.5640038498556303,
|
| 477 |
+
"grad_norm": 4.948224067687988,
|
| 478 |
+
"learning_rate": 0.00019175241996470402,
|
| 479 |
+
"loss": 2.820457763671875,
|
| 480 |
+
"step": 1625
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.588065447545717,
|
| 484 |
+
"grad_norm": 5.298257827758789,
|
| 485 |
+
"learning_rate": 0.00019143971714660212,
|
| 486 |
+
"loss": 2.8194976806640626,
|
| 487 |
+
"step": 1650
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.6121270452358036,
|
| 491 |
+
"grad_norm": 5.093235015869141,
|
| 492 |
+
"learning_rate": 0.00019112146086744505,
|
| 493 |
+
"loss": 2.81148193359375,
|
| 494 |
+
"step": 1675
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.6361886429258903,
|
| 498 |
+
"grad_norm": 4.273094654083252,
|
| 499 |
+
"learning_rate": 0.00019079767045607492,
|
| 500 |
+
"loss": 2.7786264038085937,
|
| 501 |
+
"step": 1700
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.660250240615977,
|
| 505 |
+
"grad_norm": 4.984681606292725,
|
| 506 |
+
"learning_rate": 0.00019046836557744135,
|
| 507 |
+
"loss": 2.815713195800781,
|
| 508 |
+
"step": 1725
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.6843118383060636,
|
| 512 |
+
"grad_norm": 5.150766849517822,
|
| 513 |
+
"learning_rate": 0.00019013356623140748,
|
| 514 |
+
"loss": 2.837911376953125,
|
| 515 |
+
"step": 1750
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.7083734359961502,
|
| 519 |
+
"grad_norm": 5.059906959533691,
|
| 520 |
+
"learning_rate": 0.00018979329275153493,
|
| 521 |
+
"loss": 2.7880908203125,
|
| 522 |
+
"step": 1775
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.7324350336862366,
|
| 526 |
+
"grad_norm": 4.918034076690674,
|
| 527 |
+
"learning_rate": 0.00018944756580384914,
|
| 528 |
+
"loss": 2.800741882324219,
|
| 529 |
+
"step": 1800
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.7564966313763235,
|
| 533 |
+
"grad_norm": 5.184783458709717,
|
| 534 |
+
"learning_rate": 0.00018909640638558415,
|
| 535 |
+
"loss": 2.806388244628906,
|
| 536 |
+
"step": 1825
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.78055822906641,
|
| 540 |
+
"grad_norm": 4.884471416473389,
|
| 541 |
+
"learning_rate": 0.0001887398358239073,
|
| 542 |
+
"loss": 2.8272454833984373,
|
| 543 |
+
"step": 1850
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.8046198267564968,
|
| 547 |
+
"grad_norm": 5.7572922706604,
|
| 548 |
+
"learning_rate": 0.0001883778757746241,
|
| 549 |
+
"loss": 2.7949139404296877,
|
| 550 |
+
"step": 1875
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.8286814244465832,
|
| 554 |
+
"grad_norm": 5.056210517883301,
|
| 555 |
+
"learning_rate": 0.0001880105482208629,
|
| 556 |
+
"loss": 2.783699951171875,
|
| 557 |
+
"step": 1900
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.8527430221366699,
|
| 561 |
+
"grad_norm": 4.134166240692139,
|
| 562 |
+
"learning_rate": 0.00018763787547173975,
|
| 563 |
+
"loss": 2.7708892822265625,
|
| 564 |
+
"step": 1925
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.8768046198267565,
|
| 568 |
+
"grad_norm": 4.962014198303223,
|
| 569 |
+
"learning_rate": 0.00018725988016100357,
|
| 570 |
+
"loss": 2.8055865478515627,
|
| 571 |
+
"step": 1950
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.9008662175168431,
|
| 575 |
+
"grad_norm": 4.6821088790893555,
|
| 576 |
+
"learning_rate": 0.0001868765852456615,
|
| 577 |
+
"loss": 2.809751281738281,
|
| 578 |
+
"step": 1975
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.9249278152069298,
|
| 582 |
+
"grad_norm": 4.961895942687988,
|
| 583 |
+
"learning_rate": 0.0001864880140045846,
|
| 584 |
+
"loss": 2.837526550292969,
|
| 585 |
+
"step": 2000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.9489894128970162,
|
| 589 |
+
"grad_norm": 4.305720806121826,
|
| 590 |
+
"learning_rate": 0.000186094190037094,
|
| 591 |
+
"loss": 2.8118356323242186,
|
| 592 |
+
"step": 2025
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.973051010587103,
|
| 596 |
+
"grad_norm": 4.6312174797058105,
|
| 597 |
+
"learning_rate": 0.00018569513726152782,
|
| 598 |
+
"loss": 2.8202804565429687,
|
| 599 |
+
"step": 2050
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.9971126082771895,
|
| 603 |
+
"grad_norm": 5.322890281677246,
|
| 604 |
+
"learning_rate": 0.0001852908799137883,
|
| 605 |
+
"loss": 2.7954449462890625,
|
| 606 |
+
"step": 2075
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.0,
|
| 610 |
+
"eval_loss": 2.796738386154175,
|
| 611 |
+
"eval_runtime": 10.4549,
|
| 612 |
+
"eval_samples_per_second": 143.473,
|
| 613 |
+
"eval_steps_per_second": 2.296,
|
| 614 |
+
"step": 2078
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.0211742059672764,
|
| 618 |
+
"grad_norm": 4.973223686218262,
|
| 619 |
+
"learning_rate": 0.00018488144254586998,
|
| 620 |
+
"loss": 2.634290771484375,
|
| 621 |
+
"step": 2100
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.045235803657363,
|
| 625 |
+
"grad_norm": 4.723742961883545,
|
| 626 |
+
"learning_rate": 0.0001844668500243686,
|
| 627 |
+
"loss": 2.6185684204101562,
|
| 628 |
+
"step": 2125
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.0692974013474497,
|
| 632 |
+
"grad_norm": 4.866776466369629,
|
| 633 |
+
"learning_rate": 0.0001840471275289707,
|
| 634 |
+
"loss": 2.63686767578125,
|
| 635 |
+
"step": 2150
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.093358999037536,
|
| 639 |
+
"grad_norm": 5.0636887550354,
|
| 640 |
+
"learning_rate": 0.00018362230055092465,
|
| 641 |
+
"loss": 2.6171771240234376,
|
| 642 |
+
"step": 2175
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.1174205967276225,
|
| 646 |
+
"grad_norm": 5.5221943855285645,
|
| 647 |
+
"learning_rate": 0.0001831923948914921,
|
| 648 |
+
"loss": 2.6215435791015627,
|
| 649 |
+
"step": 2200
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 2.1414821944177094,
|
| 653 |
+
"grad_norm": 4.635878562927246,
|
| 654 |
+
"learning_rate": 0.00018275743666038144,
|
| 655 |
+
"loss": 2.6706222534179687,
|
| 656 |
+
"step": 2225
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.165543792107796,
|
| 660 |
+
"grad_norm": 5.1727705001831055,
|
| 661 |
+
"learning_rate": 0.00018231745227416155,
|
| 662 |
+
"loss": 2.63311279296875,
|
| 663 |
+
"step": 2250
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 2.1896053897978827,
|
| 667 |
+
"grad_norm": 4.907919406890869,
|
| 668 |
+
"learning_rate": 0.00018187246845465787,
|
| 669 |
+
"loss": 2.64796630859375,
|
| 670 |
+
"step": 2275
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 2.213666987487969,
|
| 674 |
+
"grad_norm": 4.771077632904053,
|
| 675 |
+
"learning_rate": 0.0001814225122273292,
|
| 676 |
+
"loss": 2.6579470825195313,
|
| 677 |
+
"step": 2300
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 2.237728585178056,
|
| 681 |
+
"grad_norm": 4.675436496734619,
|
| 682 |
+
"learning_rate": 0.0001809676109196265,
|
| 683 |
+
"loss": 2.655972900390625,
|
| 684 |
+
"step": 2325
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.2617901828681424,
|
| 688 |
+
"grad_norm": 5.554975986480713,
|
| 689 |
+
"learning_rate": 0.00018050779215933314,
|
| 690 |
+
"loss": 2.6388140869140626,
|
| 691 |
+
"step": 2350
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 2.2858517805582292,
|
| 695 |
+
"grad_norm": 5.329654216766357,
|
| 696 |
+
"learning_rate": 0.00018004308387288695,
|
| 697 |
+
"loss": 2.681984558105469,
|
| 698 |
+
"step": 2375
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.3099133782483157,
|
| 702 |
+
"grad_norm": 5.244048595428467,
|
| 703 |
+
"learning_rate": 0.0001795735142836842,
|
| 704 |
+
"loss": 2.65951416015625,
|
| 705 |
+
"step": 2400
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.3339749759384025,
|
| 709 |
+
"grad_norm": 4.824357509613037,
|
| 710 |
+
"learning_rate": 0.00017909911191036547,
|
| 711 |
+
"loss": 2.6685150146484373,
|
| 712 |
+
"step": 2425
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 2.358036573628489,
|
| 716 |
+
"grad_norm": 5.436433792114258,
|
| 717 |
+
"learning_rate": 0.00017861990556508356,
|
| 718 |
+
"loss": 2.639136962890625,
|
| 719 |
+
"step": 2450
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 2.3820981713185754,
|
| 723 |
+
"grad_norm": 5.015676975250244,
|
| 724 |
+
"learning_rate": 0.0001781359243517537,
|
| 725 |
+
"loss": 2.652701416015625,
|
| 726 |
+
"step": 2475
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.4061597690086622,
|
| 730 |
+
"grad_norm": 5.517067909240723,
|
| 731 |
+
"learning_rate": 0.00017764719766428594,
|
| 732 |
+
"loss": 2.6934796142578126,
|
| 733 |
+
"step": 2500
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 2.4302213666987487,
|
| 737 |
+
"grad_norm": 4.892432689666748,
|
| 738 |
+
"learning_rate": 0.00017715375518479994,
|
| 739 |
+
"loss": 2.647161865234375,
|
| 740 |
+
"step": 2525
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.4542829643888355,
|
| 744 |
+
"grad_norm": 5.116722106933594,
|
| 745 |
+
"learning_rate": 0.0001766556268818222,
|
| 746 |
+
"loss": 2.6621990966796876,
|
| 747 |
+
"step": 2550
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.478344562078922,
|
| 751 |
+
"grad_norm": 4.984952926635742,
|
| 752 |
+
"learning_rate": 0.0001761528430084662,
|
| 753 |
+
"loss": 2.6337930297851564,
|
| 754 |
+
"step": 2575
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 2.502406159769009,
|
| 758 |
+
"grad_norm": 6.104625701904297,
|
| 759 |
+
"learning_rate": 0.00017564543410059473,
|
| 760 |
+
"loss": 2.6945245361328123,
|
| 761 |
+
"step": 2600
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 2.5264677574590952,
|
| 765 |
+
"grad_norm": 5.737064361572266,
|
| 766 |
+
"learning_rate": 0.0001751334309749655,
|
| 767 |
+
"loss": 2.6489166259765624,
|
| 768 |
+
"step": 2625
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.550529355149182,
|
| 772 |
+
"grad_norm": 5.924130916595459,
|
| 773 |
+
"learning_rate": 0.0001746168647273595,
|
| 774 |
+
"loss": 2.649996337890625,
|
| 775 |
+
"step": 2650
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 2.5745909528392685,
|
| 779 |
+
"grad_norm": 5.325296878814697,
|
| 780 |
+
"learning_rate": 0.00017409576673069245,
|
| 781 |
+
"loss": 2.6771316528320312,
|
| 782 |
+
"step": 2675
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.598652550529355,
|
| 786 |
+
"grad_norm": 5.358754634857178,
|
| 787 |
+
"learning_rate": 0.0001735701686331094,
|
| 788 |
+
"loss": 2.6654647827148437,
|
| 789 |
+
"step": 2700
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 2.622714148219442,
|
| 793 |
+
"grad_norm": 4.84153938293457,
|
| 794 |
+
"learning_rate": 0.00017304010235606257,
|
| 795 |
+
"loss": 2.63984619140625,
|
| 796 |
+
"step": 2725
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 2.6467757459095282,
|
| 800 |
+
"grad_norm": 5.001486778259277,
|
| 801 |
+
"learning_rate": 0.00017250560009237276,
|
| 802 |
+
"loss": 2.650990295410156,
|
| 803 |
+
"step": 2750
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 2.670837343599615,
|
| 807 |
+
"grad_norm": 5.516018867492676,
|
| 808 |
+
"learning_rate": 0.0001719666943042741,
|
| 809 |
+
"loss": 2.6812786865234375,
|
| 810 |
+
"step": 2775
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 2.6948989412897015,
|
| 814 |
+
"grad_norm": 4.497090816497803,
|
| 815 |
+
"learning_rate": 0.00017142341772144248,
|
| 816 |
+
"loss": 2.6683084106445314,
|
| 817 |
+
"step": 2800
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 2.7189605389797884,
|
| 821 |
+
"grad_norm": 5.212906360626221,
|
| 822 |
+
"learning_rate": 0.00017087580333900778,
|
| 823 |
+
"loss": 2.646448974609375,
|
| 824 |
+
"step": 2825
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.743022136669875,
|
| 828 |
+
"grad_norm": 5.7047600746154785,
|
| 829 |
+
"learning_rate": 0.00017032388441555,
|
| 830 |
+
"loss": 2.654071044921875,
|
| 831 |
+
"step": 2850
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 2.7670837343599617,
|
| 835 |
+
"grad_norm": 5.212134838104248,
|
| 836 |
+
"learning_rate": 0.00016976769447107928,
|
| 837 |
+
"loss": 2.635329284667969,
|
| 838 |
+
"step": 2875
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.791145332050048,
|
| 842 |
+
"grad_norm": 5.364657878875732,
|
| 843 |
+
"learning_rate": 0.00016920726728500017,
|
| 844 |
+
"loss": 2.6405084228515623,
|
| 845 |
+
"step": 2900
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 2.8152069297401345,
|
| 849 |
+
"grad_norm": 5.30003023147583,
|
| 850 |
+
"learning_rate": 0.00016864263689406002,
|
| 851 |
+
"loss": 2.656103210449219,
|
| 852 |
+
"step": 2925
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.8392685274302214,
|
| 856 |
+
"grad_norm": 5.683626174926758,
|
| 857 |
+
"learning_rate": 0.00016807383759028186,
|
| 858 |
+
"loss": 2.6557977294921873,
|
| 859 |
+
"step": 2950
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.863330125120308,
|
| 863 |
+
"grad_norm": 6.03591251373291,
|
| 864 |
+
"learning_rate": 0.00016750090391888173,
|
| 865 |
+
"loss": 2.651222839355469,
|
| 866 |
+
"step": 2975
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.8873917228103947,
|
| 870 |
+
"grad_norm": 5.038674354553223,
|
| 871 |
+
"learning_rate": 0.00016692387067617062,
|
| 872 |
+
"loss": 2.6435317993164062,
|
| 873 |
+
"step": 3000
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 2.911453320500481,
|
| 877 |
+
"grad_norm": 4.938141822814941,
|
| 878 |
+
"learning_rate": 0.00016634277290744103,
|
| 879 |
+
"loss": 2.64944580078125,
|
| 880 |
+
"step": 3025
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.935514918190568,
|
| 884 |
+
"grad_norm": 5.381541728973389,
|
| 885 |
+
"learning_rate": 0.0001657576459048388,
|
| 886 |
+
"loss": 2.665923767089844,
|
| 887 |
+
"step": 3050
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 2.9595765158806544,
|
| 891 |
+
"grad_norm": 5.302907943725586,
|
| 892 |
+
"learning_rate": 0.00016516852520521947,
|
| 893 |
+
"loss": 2.6620660400390626,
|
| 894 |
+
"step": 3075
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.9836381135707413,
|
| 898 |
+
"grad_norm": 5.368643283843994,
|
| 899 |
+
"learning_rate": 0.00016457544658799011,
|
| 900 |
+
"loss": 2.6165301513671877,
|
| 901 |
+
"step": 3100
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 3.0,
|
| 905 |
+
"eval_loss": 2.7015655040740967,
|
| 906 |
+
"eval_runtime": 9.9742,
|
| 907 |
+
"eval_samples_per_second": 150.388,
|
| 908 |
+
"eval_steps_per_second": 2.406,
|
| 909 |
+
"step": 3117
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 3.0076997112608277,
|
| 913 |
+
"grad_norm": 5.0673828125,
|
| 914 |
+
"learning_rate": 0.00016397844607293627,
|
| 915 |
+
"loss": 2.61294677734375,
|
| 916 |
+
"step": 3125
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 3.0317613089509146,
|
| 920 |
+
"grad_norm": 5.165353775024414,
|
| 921 |
+
"learning_rate": 0.00016337755991803444,
|
| 922 |
+
"loss": 2.4642942810058592,
|
| 923 |
+
"step": 3150
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 3.055822906641001,
|
| 927 |
+
"grad_norm": 5.3854079246521,
|
| 928 |
+
"learning_rate": 0.00016277282461724988,
|
| 929 |
+
"loss": 2.4820993041992185,
|
| 930 |
+
"step": 3175
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 3.0798845043310874,
|
| 934 |
+
"grad_norm": 5.312911510467529,
|
| 935 |
+
"learning_rate": 0.00016216427689832015,
|
| 936 |
+
"loss": 2.489494171142578,
|
| 937 |
+
"step": 3200
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 3.1039461020211743,
|
| 941 |
+
"grad_norm": 5.348562240600586,
|
| 942 |
+
"learning_rate": 0.0001615519537205247,
|
| 943 |
+
"loss": 2.4705026245117185,
|
| 944 |
+
"step": 3225
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 3.1280076997112607,
|
| 948 |
+
"grad_norm": 5.841189384460449,
|
| 949 |
+
"learning_rate": 0.00016093589227244017,
|
| 950 |
+
"loss": 2.46598388671875,
|
| 951 |
+
"step": 3250
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 3.1520692974013476,
|
| 955 |
+
"grad_norm": 5.61829948425293,
|
| 956 |
+
"learning_rate": 0.00016031612996968156,
|
| 957 |
+
"loss": 2.4969444274902344,
|
| 958 |
+
"step": 3275
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 3.176130895091434,
|
| 962 |
+
"grad_norm": 5.292537212371826,
|
| 963 |
+
"learning_rate": 0.0001596927044526301,
|
| 964 |
+
"loss": 2.473064422607422,
|
| 965 |
+
"step": 3300
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 3.200192492781521,
|
| 969 |
+
"grad_norm": 5.158451557159424,
|
| 970 |
+
"learning_rate": 0.0001590656535841471,
|
| 971 |
+
"loss": 2.4472772216796876,
|
| 972 |
+
"step": 3325
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 3.2242540904716073,
|
| 976 |
+
"grad_norm": 5.177428722381592,
|
| 977 |
+
"learning_rate": 0.00015843501544727446,
|
| 978 |
+
"loss": 2.5115626525878905,
|
| 979 |
+
"step": 3350
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 3.248315688161694,
|
| 983 |
+
"grad_norm": 4.971099853515625,
|
| 984 |
+
"learning_rate": 0.00015780082834292172,
|
| 985 |
+
"loss": 2.5359817504882813,
|
| 986 |
+
"step": 3375
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 3.2723772858517806,
|
| 990 |
+
"grad_norm": 5.984713554382324,
|
| 991 |
+
"learning_rate": 0.00015716313078753986,
|
| 992 |
+
"loss": 2.518363342285156,
|
| 993 |
+
"step": 3400
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 3.2964388835418674,
|
| 997 |
+
"grad_norm": 5.768255233764648,
|
| 998 |
+
"learning_rate": 0.00015652196151078217,
|
| 999 |
+
"loss": 2.4845159912109374,
|
| 1000 |
+
"step": 3425
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 3.320500481231954,
|
| 1004 |
+
"grad_norm": 5.554891586303711,
|
| 1005 |
+
"learning_rate": 0.0001558773594531519,
|
| 1006 |
+
"loss": 2.521295166015625,
|
| 1007 |
+
"step": 3450
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 3.3445620789220403,
|
| 1011 |
+
"grad_norm": 5.35845422744751,
|
| 1012 |
+
"learning_rate": 0.00015522936376363746,
|
| 1013 |
+
"loss": 2.5144276428222656,
|
| 1014 |
+
"step": 3475
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 3.368623676612127,
|
| 1018 |
+
"grad_norm": 5.9035210609436035,
|
| 1019 |
+
"learning_rate": 0.00015457801379733464,
|
| 1020 |
+
"loss": 2.4923356628417968,
|
| 1021 |
+
"step": 3500
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 3.3926852743022136,
|
| 1025 |
+
"grad_norm": 5.4024786949157715,
|
| 1026 |
+
"learning_rate": 0.0001539233491130564,
|
| 1027 |
+
"loss": 2.5232693481445314,
|
| 1028 |
+
"step": 3525
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 3.4167468719923004,
|
| 1032 |
+
"grad_norm": 5.661716461181641,
|
| 1033 |
+
"learning_rate": 0.00015326540947093042,
|
| 1034 |
+
"loss": 2.510073699951172,
|
| 1035 |
+
"step": 3550
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 3.440808469682387,
|
| 1039 |
+
"grad_norm": 5.541335582733154,
|
| 1040 |
+
"learning_rate": 0.00015260423482998427,
|
| 1041 |
+
"loss": 2.512746124267578,
|
| 1042 |
+
"step": 3575
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 3.4648700673724737,
|
| 1046 |
+
"grad_norm": 5.722592353820801,
|
| 1047 |
+
"learning_rate": 0.00015193986534571863,
|
| 1048 |
+
"loss": 2.5191300964355468,
|
| 1049 |
+
"step": 3600
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 3.48893166506256,
|
| 1053 |
+
"grad_norm": 6.089288711547852,
|
| 1054 |
+
"learning_rate": 0.0001512723413676683,
|
| 1055 |
+
"loss": 2.4799041748046875,
|
| 1056 |
+
"step": 3625
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 3.512993262752647,
|
| 1060 |
+
"grad_norm": 5.490766525268555,
|
| 1061 |
+
"learning_rate": 0.00015060170343695196,
|
| 1062 |
+
"loss": 2.501981658935547,
|
| 1063 |
+
"step": 3650
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 3.5370548604427334,
|
| 1067 |
+
"grad_norm": 5.429261684417725,
|
| 1068 |
+
"learning_rate": 0.0001499279922838096,
|
| 1069 |
+
"loss": 2.528828430175781,
|
| 1070 |
+
"step": 3675
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 3.56111645813282,
|
| 1074 |
+
"grad_norm": 5.104456424713135,
|
| 1075 |
+
"learning_rate": 0.00014925124882512908,
|
| 1076 |
+
"loss": 2.506761016845703,
|
| 1077 |
+
"step": 3700
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.5851780558229067,
|
| 1081 |
+
"grad_norm": 5.07927942276001,
|
| 1082 |
+
"learning_rate": 0.000148571514161961,
|
| 1083 |
+
"loss": 2.476543121337891,
|
| 1084 |
+
"step": 3725
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 3.609239653512993,
|
| 1088 |
+
"grad_norm": 5.794335842132568,
|
| 1089 |
+
"learning_rate": 0.00014788882957702257,
|
| 1090 |
+
"loss": 2.5332624816894533,
|
| 1091 |
+
"step": 3750
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 3.63330125120308,
|
| 1095 |
+
"grad_norm": 5.124516010284424,
|
| 1096 |
+
"learning_rate": 0.00014720323653219018,
|
| 1097 |
+
"loss": 2.5622378540039064,
|
| 1098 |
+
"step": 3775
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 3.6573628488931664,
|
| 1102 |
+
"grad_norm": 5.550374984741211,
|
| 1103 |
+
"learning_rate": 0.00014651477666598151,
|
| 1104 |
+
"loss": 2.531949462890625,
|
| 1105 |
+
"step": 3800
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 3.681424446583253,
|
| 1109 |
+
"grad_norm": 5.177489280700684,
|
| 1110 |
+
"learning_rate": 0.00014582349179102644,
|
| 1111 |
+
"loss": 2.5247061157226565,
|
| 1112 |
+
"step": 3825
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 3.7054860442733397,
|
| 1116 |
+
"grad_norm": 5.406796932220459,
|
| 1117 |
+
"learning_rate": 0.00014512942389152783,
|
| 1118 |
+
"loss": 2.5437385559082033,
|
| 1119 |
+
"step": 3850
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 3.7295476419634266,
|
| 1123 |
+
"grad_norm": 5.12544584274292,
|
| 1124 |
+
"learning_rate": 0.00014443261512071156,
|
| 1125 |
+
"loss": 2.4803817749023436,
|
| 1126 |
+
"step": 3875
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 3.753609239653513,
|
| 1130 |
+
"grad_norm": 4.869628429412842,
|
| 1131 |
+
"learning_rate": 0.00014373310779826638,
|
| 1132 |
+
"loss": 2.5459124755859377,
|
| 1133 |
+
"step": 3900
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 3.7776708373435994,
|
| 1137 |
+
"grad_norm": 5.2594099044799805,
|
| 1138 |
+
"learning_rate": 0.0001430309444077738,
|
| 1139 |
+
"loss": 2.4942247009277345,
|
| 1140 |
+
"step": 3925
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 3.8017324350336863,
|
| 1144 |
+
"grad_norm": 5.267584800720215,
|
| 1145 |
+
"learning_rate": 0.00014232616759412772,
|
| 1146 |
+
"loss": 2.5201406860351563,
|
| 1147 |
+
"step": 3950
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 3.8257940327237727,
|
| 1151 |
+
"grad_norm": 5.06338357925415,
|
| 1152 |
+
"learning_rate": 0.00014161882016094473,
|
| 1153 |
+
"loss": 2.5072589111328125,
|
| 1154 |
+
"step": 3975
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.8498556304138596,
|
| 1158 |
+
"grad_norm": 5.17350435256958,
|
| 1159 |
+
"learning_rate": 0.00014090894506796426,
|
| 1160 |
+
"loss": 2.54510986328125,
|
| 1161 |
+
"step": 4000
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 3.873917228103946,
|
| 1165 |
+
"grad_norm": 4.9889678955078125,
|
| 1166 |
+
"learning_rate": 0.00014019658542843951,
|
| 1167 |
+
"loss": 2.4879167175292967,
|
| 1168 |
+
"step": 4025
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 3.897978825794033,
|
| 1172 |
+
"grad_norm": 6.220742225646973,
|
| 1173 |
+
"learning_rate": 0.0001394817845065192,
|
| 1174 |
+
"loss": 2.517736358642578,
|
| 1175 |
+
"step": 4050
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 3.9220404234841193,
|
| 1179 |
+
"grad_norm": 5.632843017578125,
|
| 1180 |
+
"learning_rate": 0.00013876458571461978,
|
| 1181 |
+
"loss": 2.4728121948242188,
|
| 1182 |
+
"step": 4075
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 3.946102021174206,
|
| 1186 |
+
"grad_norm": 5.72349214553833,
|
| 1187 |
+
"learning_rate": 0.0001380450326107889,
|
| 1188 |
+
"loss": 2.5195367431640623,
|
| 1189 |
+
"step": 4100
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 3.9701636188642926,
|
| 1193 |
+
"grad_norm": 6.055045127868652,
|
| 1194 |
+
"learning_rate": 0.00013732316889606005,
|
| 1195 |
+
"loss": 2.530181579589844,
|
| 1196 |
+
"step": 4125
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 3.994225216554379,
|
| 1200 |
+
"grad_norm": 4.998103141784668,
|
| 1201 |
+
"learning_rate": 0.00013659903841179832,
|
| 1202 |
+
"loss": 2.5351657104492187,
|
| 1203 |
+
"step": 4150
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.0,
|
| 1207 |
+
"eval_loss": 2.6353492736816406,
|
| 1208 |
+
"eval_runtime": 10.2956,
|
| 1209 |
+
"eval_samples_per_second": 145.693,
|
| 1210 |
+
"eval_steps_per_second": 2.331,
|
| 1211 |
+
"step": 4156
|
| 1212 |
+
}
|
| 1213 |
+
],
|
| 1214 |
+
"logging_steps": 25,
|
| 1215 |
+
"max_steps": 10390,
|
| 1216 |
+
"num_input_tokens_seen": 0,
|
| 1217 |
+
"num_train_epochs": 10,
|
| 1218 |
+
"save_steps": 500,
|
| 1219 |
+
"stateful_callbacks": {
|
| 1220 |
+
"TrainerControl": {
|
| 1221 |
+
"args": {
|
| 1222 |
+
"should_epoch_stop": false,
|
| 1223 |
+
"should_evaluate": false,
|
| 1224 |
+
"should_log": false,
|
| 1225 |
+
"should_save": true,
|
| 1226 |
+
"should_training_stop": false
|
| 1227 |
+
},
|
| 1228 |
+
"attributes": {}
|
| 1229 |
+
}
|
| 1230 |
+
},
|
| 1231 |
+
"total_flos": 0.0,
|
| 1232 |
+
"train_batch_size": 8,
|
| 1233 |
+
"trial_name": null,
|
| 1234 |
+
"trial_params": null
|
| 1235 |
+
}
|