Deep35 continual checkpoint step 68000
Browse files
last-checkpoint/continual_state.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"global_step":
|
| 3 |
-
"sequence_tokens_emitted":
|
| 4 |
"mix": [
|
| 5 |
0.8,
|
| 6 |
0.09,
|
|
@@ -56,5 +56,5 @@
|
|
| 56 |
"current_batch_size": 240,
|
| 57 |
"current_gradient_accumulation": 1,
|
| 58 |
"current_tokens_per_update": 245760,
|
| 59 |
-
"cumulative_tokens_at_save":
|
| 60 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"global_step": 68000,
|
| 3 |
+
"sequence_tokens_emitted": 10977525760,
|
| 4 |
"mix": [
|
| 5 |
0.8,
|
| 6 |
0.09,
|
|
|
|
| 56 |
"current_batch_size": 240,
|
| 57 |
"current_gradient_accumulation": 1,
|
| 58 |
"current_tokens_per_update": 245760,
|
| 59 |
+
"cumulative_tokens_at_save": 10977280000
|
| 60 |
}
|
last-checkpoint/data_state.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83839bb4a1a1dd1bdb13ae46419226765e1f4932d23409118ae48682e4b81711
|
| 3 |
+
size 25087
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 70326256
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9dbe9733d86e444acf76351bdc0d01c90c9ebfadaa415e128566c836b0e34d58
|
| 3 |
size 70326256
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 140740171
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:506921a66c8e8d24363c844357c33b51d81726d4f2505352e43c93b354f7b1ad
|
| 3 |
size 140740171
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d24c706aabd9e13c1509b36523f7b9ba4d813be3ac7bd6766eb3b55a97ca216
|
| 3 |
size 14645
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83683208050f46041252eea4909d760c7161c0633154e96362abc1e12795e57e
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -47362,6 +47362,1438 @@
|
|
| 47362 |
"eval_samples_per_second": 550.129,
|
| 47363 |
"eval_steps_per_second": 17.192,
|
| 47364 |
"step": 66000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 47365 |
}
|
| 47366 |
],
|
| 47367 |
"logging_steps": 10,
|
|
@@ -47381,7 +48813,7 @@
|
|
| 47381 |
"attributes": {}
|
| 47382 |
}
|
| 47383 |
},
|
| 47384 |
-
"total_flos": 1.
|
| 47385 |
"train_batch_size": 240,
|
| 47386 |
"trial_name": null,
|
| 47387 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 9e-06,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 68000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 47362 |
"eval_samples_per_second": 550.129,
|
| 47363 |
"eval_steps_per_second": 17.192,
|
| 47364 |
"step": 66000
|
| 47365 |
+
},
|
| 47366 |
+
{
|
| 47367 |
+
"epoch": 8.005e-06,
|
| 47368 |
+
"grad_norm": 0.47265625,
|
| 47369 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47370 |
+
"loss": 3.0445,
|
| 47371 |
+
"step": 66010
|
| 47372 |
+
},
|
| 47373 |
+
{
|
| 47374 |
+
"epoch": 8.01e-06,
|
| 47375 |
+
"grad_norm": 0.458984375,
|
| 47376 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47377 |
+
"loss": 3.0545,
|
| 47378 |
+
"step": 66020
|
| 47379 |
+
},
|
| 47380 |
+
{
|
| 47381 |
+
"epoch": 8.015e-06,
|
| 47382 |
+
"grad_norm": 0.45703125,
|
| 47383 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47384 |
+
"loss": 3.0385,
|
| 47385 |
+
"step": 66030
|
| 47386 |
+
},
|
| 47387 |
+
{
|
| 47388 |
+
"epoch": 8.02e-06,
|
| 47389 |
+
"grad_norm": 0.47265625,
|
| 47390 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47391 |
+
"loss": 3.0215,
|
| 47392 |
+
"step": 66040
|
| 47393 |
+
},
|
| 47394 |
+
{
|
| 47395 |
+
"epoch": 8.025e-06,
|
| 47396 |
+
"grad_norm": 0.462890625,
|
| 47397 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47398 |
+
"loss": 2.9968,
|
| 47399 |
+
"step": 66050
|
| 47400 |
+
},
|
| 47401 |
+
{
|
| 47402 |
+
"epoch": 8.03e-06,
|
| 47403 |
+
"grad_norm": 0.4609375,
|
| 47404 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47405 |
+
"loss": 3.0261,
|
| 47406 |
+
"step": 66060
|
| 47407 |
+
},
|
| 47408 |
+
{
|
| 47409 |
+
"epoch": 8.035e-06,
|
| 47410 |
+
"grad_norm": 0.47265625,
|
| 47411 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47412 |
+
"loss": 3.037,
|
| 47413 |
+
"step": 66070
|
| 47414 |
+
},
|
| 47415 |
+
{
|
| 47416 |
+
"epoch": 8.04e-06,
|
| 47417 |
+
"grad_norm": 0.470703125,
|
| 47418 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47419 |
+
"loss": 3.0226,
|
| 47420 |
+
"step": 66080
|
| 47421 |
+
},
|
| 47422 |
+
{
|
| 47423 |
+
"epoch": 8.045e-06,
|
| 47424 |
+
"grad_norm": 0.4765625,
|
| 47425 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47426 |
+
"loss": 3.0362,
|
| 47427 |
+
"step": 66090
|
| 47428 |
+
},
|
| 47429 |
+
{
|
| 47430 |
+
"epoch": 8.05e-06,
|
| 47431 |
+
"grad_norm": 0.486328125,
|
| 47432 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47433 |
+
"loss": 3.0285,
|
| 47434 |
+
"step": 66100
|
| 47435 |
+
},
|
| 47436 |
+
{
|
| 47437 |
+
"epoch": 8.055e-06,
|
| 47438 |
+
"grad_norm": 0.4453125,
|
| 47439 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47440 |
+
"loss": 3.0267,
|
| 47441 |
+
"step": 66110
|
| 47442 |
+
},
|
| 47443 |
+
{
|
| 47444 |
+
"epoch": 8.06e-06,
|
| 47445 |
+
"grad_norm": 0.453125,
|
| 47446 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47447 |
+
"loss": 3.0302,
|
| 47448 |
+
"step": 66120
|
| 47449 |
+
},
|
| 47450 |
+
{
|
| 47451 |
+
"epoch": 8.065e-06,
|
| 47452 |
+
"grad_norm": 0.50390625,
|
| 47453 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47454 |
+
"loss": 3.0164,
|
| 47455 |
+
"step": 66130
|
| 47456 |
+
},
|
| 47457 |
+
{
|
| 47458 |
+
"epoch": 8.07e-06,
|
| 47459 |
+
"grad_norm": 0.4453125,
|
| 47460 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47461 |
+
"loss": 3.0595,
|
| 47462 |
+
"step": 66140
|
| 47463 |
+
},
|
| 47464 |
+
{
|
| 47465 |
+
"epoch": 8.075e-06,
|
| 47466 |
+
"grad_norm": 0.48046875,
|
| 47467 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47468 |
+
"loss": 3.0509,
|
| 47469 |
+
"step": 66150
|
| 47470 |
+
},
|
| 47471 |
+
{
|
| 47472 |
+
"epoch": 8.08e-06,
|
| 47473 |
+
"grad_norm": 0.462890625,
|
| 47474 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47475 |
+
"loss": 3.0647,
|
| 47476 |
+
"step": 66160
|
| 47477 |
+
},
|
| 47478 |
+
{
|
| 47479 |
+
"epoch": 8.085e-06,
|
| 47480 |
+
"grad_norm": 0.46875,
|
| 47481 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47482 |
+
"loss": 3.0425,
|
| 47483 |
+
"step": 66170
|
| 47484 |
+
},
|
| 47485 |
+
{
|
| 47486 |
+
"epoch": 8.09e-06,
|
| 47487 |
+
"grad_norm": 0.455078125,
|
| 47488 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47489 |
+
"loss": 3.0515,
|
| 47490 |
+
"step": 66180
|
| 47491 |
+
},
|
| 47492 |
+
{
|
| 47493 |
+
"epoch": 8.095e-06,
|
| 47494 |
+
"grad_norm": 0.5,
|
| 47495 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47496 |
+
"loss": 3.0459,
|
| 47497 |
+
"step": 66190
|
| 47498 |
+
},
|
| 47499 |
+
{
|
| 47500 |
+
"epoch": 8.1e-06,
|
| 47501 |
+
"grad_norm": 0.458984375,
|
| 47502 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47503 |
+
"loss": 3.0318,
|
| 47504 |
+
"step": 66200
|
| 47505 |
+
},
|
| 47506 |
+
{
|
| 47507 |
+
"epoch": 8.105e-06,
|
| 47508 |
+
"grad_norm": 0.51171875,
|
| 47509 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47510 |
+
"loss": 3.04,
|
| 47511 |
+
"step": 66210
|
| 47512 |
+
},
|
| 47513 |
+
{
|
| 47514 |
+
"epoch": 8.11e-06,
|
| 47515 |
+
"grad_norm": 0.48828125,
|
| 47516 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47517 |
+
"loss": 3.0442,
|
| 47518 |
+
"step": 66220
|
| 47519 |
+
},
|
| 47520 |
+
{
|
| 47521 |
+
"epoch": 8.115e-06,
|
| 47522 |
+
"grad_norm": 0.48046875,
|
| 47523 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47524 |
+
"loss": 3.0354,
|
| 47525 |
+
"step": 66230
|
| 47526 |
+
},
|
| 47527 |
+
{
|
| 47528 |
+
"epoch": 8.12e-06,
|
| 47529 |
+
"grad_norm": 0.45703125,
|
| 47530 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47531 |
+
"loss": 3.0485,
|
| 47532 |
+
"step": 66240
|
| 47533 |
+
},
|
| 47534 |
+
{
|
| 47535 |
+
"epoch": 8.125e-06,
|
| 47536 |
+
"grad_norm": 0.50390625,
|
| 47537 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47538 |
+
"loss": 3.0376,
|
| 47539 |
+
"step": 66250
|
| 47540 |
+
},
|
| 47541 |
+
{
|
| 47542 |
+
"epoch": 8.13e-06,
|
| 47543 |
+
"grad_norm": 0.458984375,
|
| 47544 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47545 |
+
"loss": 3.0317,
|
| 47546 |
+
"step": 66260
|
| 47547 |
+
},
|
| 47548 |
+
{
|
| 47549 |
+
"epoch": 8.135e-06,
|
| 47550 |
+
"grad_norm": 0.458984375,
|
| 47551 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47552 |
+
"loss": 3.0437,
|
| 47553 |
+
"step": 66270
|
| 47554 |
+
},
|
| 47555 |
+
{
|
| 47556 |
+
"epoch": 8.14e-06,
|
| 47557 |
+
"grad_norm": 0.48046875,
|
| 47558 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47559 |
+
"loss": 3.0361,
|
| 47560 |
+
"step": 66280
|
| 47561 |
+
},
|
| 47562 |
+
{
|
| 47563 |
+
"epoch": 8.145e-06,
|
| 47564 |
+
"grad_norm": 0.458984375,
|
| 47565 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47566 |
+
"loss": 3.0363,
|
| 47567 |
+
"step": 66290
|
| 47568 |
+
},
|
| 47569 |
+
{
|
| 47570 |
+
"epoch": 8.15e-06,
|
| 47571 |
+
"grad_norm": 0.53125,
|
| 47572 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47573 |
+
"loss": 3.0467,
|
| 47574 |
+
"step": 66300
|
| 47575 |
+
},
|
| 47576 |
+
{
|
| 47577 |
+
"epoch": 8.155e-06,
|
| 47578 |
+
"grad_norm": 0.470703125,
|
| 47579 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47580 |
+
"loss": 3.0571,
|
| 47581 |
+
"step": 66310
|
| 47582 |
+
},
|
| 47583 |
+
{
|
| 47584 |
+
"epoch": 8.16e-06,
|
| 47585 |
+
"grad_norm": 0.49609375,
|
| 47586 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47587 |
+
"loss": 3.0546,
|
| 47588 |
+
"step": 66320
|
| 47589 |
+
},
|
| 47590 |
+
{
|
| 47591 |
+
"epoch": 8.165e-06,
|
| 47592 |
+
"grad_norm": 0.4453125,
|
| 47593 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47594 |
+
"loss": 3.0254,
|
| 47595 |
+
"step": 66330
|
| 47596 |
+
},
|
| 47597 |
+
{
|
| 47598 |
+
"epoch": 8.17e-06,
|
| 47599 |
+
"grad_norm": 0.4765625,
|
| 47600 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47601 |
+
"loss": 3.0371,
|
| 47602 |
+
"step": 66340
|
| 47603 |
+
},
|
| 47604 |
+
{
|
| 47605 |
+
"epoch": 8.175e-06,
|
| 47606 |
+
"grad_norm": 0.439453125,
|
| 47607 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47608 |
+
"loss": 3.0444,
|
| 47609 |
+
"step": 66350
|
| 47610 |
+
},
|
| 47611 |
+
{
|
| 47612 |
+
"epoch": 8.18e-06,
|
| 47613 |
+
"grad_norm": 0.474609375,
|
| 47614 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47615 |
+
"loss": 3.0448,
|
| 47616 |
+
"step": 66360
|
| 47617 |
+
},
|
| 47618 |
+
{
|
| 47619 |
+
"epoch": 8.185e-06,
|
| 47620 |
+
"grad_norm": 0.447265625,
|
| 47621 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47622 |
+
"loss": 3.0486,
|
| 47623 |
+
"step": 66370
|
| 47624 |
+
},
|
| 47625 |
+
{
|
| 47626 |
+
"epoch": 8.19e-06,
|
| 47627 |
+
"grad_norm": 0.50390625,
|
| 47628 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47629 |
+
"loss": 3.0656,
|
| 47630 |
+
"step": 66380
|
| 47631 |
+
},
|
| 47632 |
+
{
|
| 47633 |
+
"epoch": 8.195e-06,
|
| 47634 |
+
"grad_norm": 0.5078125,
|
| 47635 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47636 |
+
"loss": 3.0372,
|
| 47637 |
+
"step": 66390
|
| 47638 |
+
},
|
| 47639 |
+
{
|
| 47640 |
+
"epoch": 8.2e-06,
|
| 47641 |
+
"grad_norm": 0.55078125,
|
| 47642 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47643 |
+
"loss": 3.0541,
|
| 47644 |
+
"step": 66400
|
| 47645 |
+
},
|
| 47646 |
+
{
|
| 47647 |
+
"epoch": 8.205e-06,
|
| 47648 |
+
"grad_norm": 0.45703125,
|
| 47649 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47650 |
+
"loss": 3.0388,
|
| 47651 |
+
"step": 66410
|
| 47652 |
+
},
|
| 47653 |
+
{
|
| 47654 |
+
"epoch": 8.21e-06,
|
| 47655 |
+
"grad_norm": 0.53125,
|
| 47656 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47657 |
+
"loss": 3.0438,
|
| 47658 |
+
"step": 66420
|
| 47659 |
+
},
|
| 47660 |
+
{
|
| 47661 |
+
"epoch": 8.215e-06,
|
| 47662 |
+
"grad_norm": 0.451171875,
|
| 47663 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47664 |
+
"loss": 3.0318,
|
| 47665 |
+
"step": 66430
|
| 47666 |
+
},
|
| 47667 |
+
{
|
| 47668 |
+
"epoch": 8.22e-06,
|
| 47669 |
+
"grad_norm": 0.490234375,
|
| 47670 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47671 |
+
"loss": 3.0375,
|
| 47672 |
+
"step": 66440
|
| 47673 |
+
},
|
| 47674 |
+
{
|
| 47675 |
+
"epoch": 8.225e-06,
|
| 47676 |
+
"grad_norm": 0.470703125,
|
| 47677 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47678 |
+
"loss": 3.0149,
|
| 47679 |
+
"step": 66450
|
| 47680 |
+
},
|
| 47681 |
+
{
|
| 47682 |
+
"epoch": 8.23e-06,
|
| 47683 |
+
"grad_norm": 0.447265625,
|
| 47684 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47685 |
+
"loss": 3.0372,
|
| 47686 |
+
"step": 66460
|
| 47687 |
+
},
|
| 47688 |
+
{
|
| 47689 |
+
"epoch": 8.235e-06,
|
| 47690 |
+
"grad_norm": 0.474609375,
|
| 47691 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47692 |
+
"loss": 3.0313,
|
| 47693 |
+
"step": 66470
|
| 47694 |
+
},
|
| 47695 |
+
{
|
| 47696 |
+
"epoch": 8.24e-06,
|
| 47697 |
+
"grad_norm": 0.494140625,
|
| 47698 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47699 |
+
"loss": 3.0464,
|
| 47700 |
+
"step": 66480
|
| 47701 |
+
},
|
| 47702 |
+
{
|
| 47703 |
+
"epoch": 8.245e-06,
|
| 47704 |
+
"grad_norm": 0.5078125,
|
| 47705 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47706 |
+
"loss": 3.0351,
|
| 47707 |
+
"step": 66490
|
| 47708 |
+
},
|
| 47709 |
+
{
|
| 47710 |
+
"epoch": 8.25e-06,
|
| 47711 |
+
"grad_norm": 0.482421875,
|
| 47712 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47713 |
+
"loss": 3.091,
|
| 47714 |
+
"step": 66500
|
| 47715 |
+
},
|
| 47716 |
+
{
|
| 47717 |
+
"epoch": 8.25e-06,
|
| 47718 |
+
"eval_loss": 3.0829520225524902,
|
| 47719 |
+
"eval_runtime": 0.0582,
|
| 47720 |
+
"eval_samples_per_second": 549.386,
|
| 47721 |
+
"eval_steps_per_second": 17.168,
|
| 47722 |
+
"step": 66500
|
| 47723 |
+
},
|
| 47724 |
+
{
|
| 47725 |
+
"epoch": 8.255e-06,
|
| 47726 |
+
"grad_norm": 0.486328125,
|
| 47727 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47728 |
+
"loss": 3.0257,
|
| 47729 |
+
"step": 66510
|
| 47730 |
+
},
|
| 47731 |
+
{
|
| 47732 |
+
"epoch": 8.26e-06,
|
| 47733 |
+
"grad_norm": 0.478515625,
|
| 47734 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47735 |
+
"loss": 3.0434,
|
| 47736 |
+
"step": 66520
|
| 47737 |
+
},
|
| 47738 |
+
{
|
| 47739 |
+
"epoch": 8.265e-06,
|
| 47740 |
+
"grad_norm": 0.447265625,
|
| 47741 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47742 |
+
"loss": 3.0337,
|
| 47743 |
+
"step": 66530
|
| 47744 |
+
},
|
| 47745 |
+
{
|
| 47746 |
+
"epoch": 8.27e-06,
|
| 47747 |
+
"grad_norm": 0.578125,
|
| 47748 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47749 |
+
"loss": 3.0175,
|
| 47750 |
+
"step": 66540
|
| 47751 |
+
},
|
| 47752 |
+
{
|
| 47753 |
+
"epoch": 8.275e-06,
|
| 47754 |
+
"grad_norm": 0.490234375,
|
| 47755 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47756 |
+
"loss": 3.0537,
|
| 47757 |
+
"step": 66550
|
| 47758 |
+
},
|
| 47759 |
+
{
|
| 47760 |
+
"epoch": 8.28e-06,
|
| 47761 |
+
"grad_norm": 0.46484375,
|
| 47762 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47763 |
+
"loss": 3.0327,
|
| 47764 |
+
"step": 66560
|
| 47765 |
+
},
|
| 47766 |
+
{
|
| 47767 |
+
"epoch": 8.285e-06,
|
| 47768 |
+
"grad_norm": 0.455078125,
|
| 47769 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47770 |
+
"loss": 3.0479,
|
| 47771 |
+
"step": 66570
|
| 47772 |
+
},
|
| 47773 |
+
{
|
| 47774 |
+
"epoch": 8.29e-06,
|
| 47775 |
+
"grad_norm": 0.458984375,
|
| 47776 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47777 |
+
"loss": 3.0206,
|
| 47778 |
+
"step": 66580
|
| 47779 |
+
},
|
| 47780 |
+
{
|
| 47781 |
+
"epoch": 8.295e-06,
|
| 47782 |
+
"grad_norm": 0.4609375,
|
| 47783 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47784 |
+
"loss": 3.0485,
|
| 47785 |
+
"step": 66590
|
| 47786 |
+
},
|
| 47787 |
+
{
|
| 47788 |
+
"epoch": 8.3e-06,
|
| 47789 |
+
"grad_norm": 0.4609375,
|
| 47790 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47791 |
+
"loss": 3.033,
|
| 47792 |
+
"step": 66600
|
| 47793 |
+
},
|
| 47794 |
+
{
|
| 47795 |
+
"epoch": 8.305e-06,
|
| 47796 |
+
"grad_norm": 0.4609375,
|
| 47797 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47798 |
+
"loss": 3.0513,
|
| 47799 |
+
"step": 66610
|
| 47800 |
+
},
|
| 47801 |
+
{
|
| 47802 |
+
"epoch": 8.31e-06,
|
| 47803 |
+
"grad_norm": 0.45703125,
|
| 47804 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47805 |
+
"loss": 3.0018,
|
| 47806 |
+
"step": 66620
|
| 47807 |
+
},
|
| 47808 |
+
{
|
| 47809 |
+
"epoch": 8.315e-06,
|
| 47810 |
+
"grad_norm": 0.4765625,
|
| 47811 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47812 |
+
"loss": 3.0578,
|
| 47813 |
+
"step": 66630
|
| 47814 |
+
},
|
| 47815 |
+
{
|
| 47816 |
+
"epoch": 8.32e-06,
|
| 47817 |
+
"grad_norm": 0.46484375,
|
| 47818 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47819 |
+
"loss": 3.0488,
|
| 47820 |
+
"step": 66640
|
| 47821 |
+
},
|
| 47822 |
+
{
|
| 47823 |
+
"epoch": 8.325e-06,
|
| 47824 |
+
"grad_norm": 0.470703125,
|
| 47825 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47826 |
+
"loss": 3.0414,
|
| 47827 |
+
"step": 66650
|
| 47828 |
+
},
|
| 47829 |
+
{
|
| 47830 |
+
"epoch": 8.33e-06,
|
| 47831 |
+
"grad_norm": 0.474609375,
|
| 47832 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47833 |
+
"loss": 3.0539,
|
| 47834 |
+
"step": 66660
|
| 47835 |
+
},
|
| 47836 |
+
{
|
| 47837 |
+
"epoch": 8.335e-06,
|
| 47838 |
+
"grad_norm": 0.6015625,
|
| 47839 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47840 |
+
"loss": 3.018,
|
| 47841 |
+
"step": 66670
|
| 47842 |
+
},
|
| 47843 |
+
{
|
| 47844 |
+
"epoch": 8.34e-06,
|
| 47845 |
+
"grad_norm": 0.43359375,
|
| 47846 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47847 |
+
"loss": 3.0815,
|
| 47848 |
+
"step": 66680
|
| 47849 |
+
},
|
| 47850 |
+
{
|
| 47851 |
+
"epoch": 8.345e-06,
|
| 47852 |
+
"grad_norm": 0.466796875,
|
| 47853 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47854 |
+
"loss": 3.0623,
|
| 47855 |
+
"step": 66690
|
| 47856 |
+
},
|
| 47857 |
+
{
|
| 47858 |
+
"epoch": 8.35e-06,
|
| 47859 |
+
"grad_norm": 0.47265625,
|
| 47860 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47861 |
+
"loss": 3.0369,
|
| 47862 |
+
"step": 66700
|
| 47863 |
+
},
|
| 47864 |
+
{
|
| 47865 |
+
"epoch": 8.355e-06,
|
| 47866 |
+
"grad_norm": 0.45703125,
|
| 47867 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47868 |
+
"loss": 3.0238,
|
| 47869 |
+
"step": 66710
|
| 47870 |
+
},
|
| 47871 |
+
{
|
| 47872 |
+
"epoch": 8.36e-06,
|
| 47873 |
+
"grad_norm": 0.455078125,
|
| 47874 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47875 |
+
"loss": 3.0206,
|
| 47876 |
+
"step": 66720
|
| 47877 |
+
},
|
| 47878 |
+
{
|
| 47879 |
+
"epoch": 8.365e-06,
|
| 47880 |
+
"grad_norm": 0.484375,
|
| 47881 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47882 |
+
"loss": 3.0457,
|
| 47883 |
+
"step": 66730
|
| 47884 |
+
},
|
| 47885 |
+
{
|
| 47886 |
+
"epoch": 8.37e-06,
|
| 47887 |
+
"grad_norm": 0.455078125,
|
| 47888 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47889 |
+
"loss": 3.0436,
|
| 47890 |
+
"step": 66740
|
| 47891 |
+
},
|
| 47892 |
+
{
|
| 47893 |
+
"epoch": 8.375e-06,
|
| 47894 |
+
"grad_norm": 0.4375,
|
| 47895 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47896 |
+
"loss": 3.0285,
|
| 47897 |
+
"step": 66750
|
| 47898 |
+
},
|
| 47899 |
+
{
|
| 47900 |
+
"epoch": 8.38e-06,
|
| 47901 |
+
"grad_norm": 0.4453125,
|
| 47902 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47903 |
+
"loss": 3.041,
|
| 47904 |
+
"step": 66760
|
| 47905 |
+
},
|
| 47906 |
+
{
|
| 47907 |
+
"epoch": 8.385e-06,
|
| 47908 |
+
"grad_norm": 0.515625,
|
| 47909 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47910 |
+
"loss": 3.0537,
|
| 47911 |
+
"step": 66770
|
| 47912 |
+
},
|
| 47913 |
+
{
|
| 47914 |
+
"epoch": 8.39e-06,
|
| 47915 |
+
"grad_norm": 0.48828125,
|
| 47916 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47917 |
+
"loss": 3.0163,
|
| 47918 |
+
"step": 66780
|
| 47919 |
+
},
|
| 47920 |
+
{
|
| 47921 |
+
"epoch": 8.395e-06,
|
| 47922 |
+
"grad_norm": 0.4765625,
|
| 47923 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47924 |
+
"loss": 3.0504,
|
| 47925 |
+
"step": 66790
|
| 47926 |
+
},
|
| 47927 |
+
{
|
| 47928 |
+
"epoch": 8.4e-06,
|
| 47929 |
+
"grad_norm": 0.453125,
|
| 47930 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47931 |
+
"loss": 3.0378,
|
| 47932 |
+
"step": 66800
|
| 47933 |
+
},
|
| 47934 |
+
{
|
| 47935 |
+
"epoch": 8.405e-06,
|
| 47936 |
+
"grad_norm": 0.4609375,
|
| 47937 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47938 |
+
"loss": 3.0341,
|
| 47939 |
+
"step": 66810
|
| 47940 |
+
},
|
| 47941 |
+
{
|
| 47942 |
+
"epoch": 8.41e-06,
|
| 47943 |
+
"grad_norm": 0.54296875,
|
| 47944 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47945 |
+
"loss": 3.028,
|
| 47946 |
+
"step": 66820
|
| 47947 |
+
},
|
| 47948 |
+
{
|
| 47949 |
+
"epoch": 8.415e-06,
|
| 47950 |
+
"grad_norm": 0.546875,
|
| 47951 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47952 |
+
"loss": 3.0443,
|
| 47953 |
+
"step": 66830
|
| 47954 |
+
},
|
| 47955 |
+
{
|
| 47956 |
+
"epoch": 8.42e-06,
|
| 47957 |
+
"grad_norm": 0.45703125,
|
| 47958 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47959 |
+
"loss": 3.0426,
|
| 47960 |
+
"step": 66840
|
| 47961 |
+
},
|
| 47962 |
+
{
|
| 47963 |
+
"epoch": 8.425e-06,
|
| 47964 |
+
"grad_norm": 0.48046875,
|
| 47965 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47966 |
+
"loss": 3.0519,
|
| 47967 |
+
"step": 66850
|
| 47968 |
+
},
|
| 47969 |
+
{
|
| 47970 |
+
"epoch": 8.43e-06,
|
| 47971 |
+
"grad_norm": 0.484375,
|
| 47972 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47973 |
+
"loss": 3.0303,
|
| 47974 |
+
"step": 66860
|
| 47975 |
+
},
|
| 47976 |
+
{
|
| 47977 |
+
"epoch": 8.435e-06,
|
| 47978 |
+
"grad_norm": 0.4921875,
|
| 47979 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47980 |
+
"loss": 3.0315,
|
| 47981 |
+
"step": 66870
|
| 47982 |
+
},
|
| 47983 |
+
{
|
| 47984 |
+
"epoch": 8.44e-06,
|
| 47985 |
+
"grad_norm": 0.478515625,
|
| 47986 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47987 |
+
"loss": 3.0364,
|
| 47988 |
+
"step": 66880
|
| 47989 |
+
},
|
| 47990 |
+
{
|
| 47991 |
+
"epoch": 8.445e-06,
|
| 47992 |
+
"grad_norm": 0.458984375,
|
| 47993 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 47994 |
+
"loss": 3.0443,
|
| 47995 |
+
"step": 66890
|
| 47996 |
+
},
|
| 47997 |
+
{
|
| 47998 |
+
"epoch": 8.45e-06,
|
| 47999 |
+
"grad_norm": 0.486328125,
|
| 48000 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48001 |
+
"loss": 3.0219,
|
| 48002 |
+
"step": 66900
|
| 48003 |
+
},
|
| 48004 |
+
{
|
| 48005 |
+
"epoch": 8.455e-06,
|
| 48006 |
+
"grad_norm": 0.453125,
|
| 48007 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48008 |
+
"loss": 3.0278,
|
| 48009 |
+
"step": 66910
|
| 48010 |
+
},
|
| 48011 |
+
{
|
| 48012 |
+
"epoch": 8.46e-06,
|
| 48013 |
+
"grad_norm": 0.48828125,
|
| 48014 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48015 |
+
"loss": 3.0317,
|
| 48016 |
+
"step": 66920
|
| 48017 |
+
},
|
| 48018 |
+
{
|
| 48019 |
+
"epoch": 8.465e-06,
|
| 48020 |
+
"grad_norm": 0.486328125,
|
| 48021 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48022 |
+
"loss": 3.0449,
|
| 48023 |
+
"step": 66930
|
| 48024 |
+
},
|
| 48025 |
+
{
|
| 48026 |
+
"epoch": 8.47e-06,
|
| 48027 |
+
"grad_norm": 0.478515625,
|
| 48028 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48029 |
+
"loss": 3.0392,
|
| 48030 |
+
"step": 66940
|
| 48031 |
+
},
|
| 48032 |
+
{
|
| 48033 |
+
"epoch": 8.475e-06,
|
| 48034 |
+
"grad_norm": 0.52734375,
|
| 48035 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48036 |
+
"loss": 3.0548,
|
| 48037 |
+
"step": 66950
|
| 48038 |
+
},
|
| 48039 |
+
{
|
| 48040 |
+
"epoch": 8.48e-06,
|
| 48041 |
+
"grad_norm": 0.466796875,
|
| 48042 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48043 |
+
"loss": 3.0396,
|
| 48044 |
+
"step": 66960
|
| 48045 |
+
},
|
| 48046 |
+
{
|
| 48047 |
+
"epoch": 8.485e-06,
|
| 48048 |
+
"grad_norm": 0.44921875,
|
| 48049 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48050 |
+
"loss": 3.0645,
|
| 48051 |
+
"step": 66970
|
| 48052 |
+
},
|
| 48053 |
+
{
|
| 48054 |
+
"epoch": 8.49e-06,
|
| 48055 |
+
"grad_norm": 0.4609375,
|
| 48056 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48057 |
+
"loss": 3.0365,
|
| 48058 |
+
"step": 66980
|
| 48059 |
+
},
|
| 48060 |
+
{
|
| 48061 |
+
"epoch": 8.495e-06,
|
| 48062 |
+
"grad_norm": 0.45703125,
|
| 48063 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48064 |
+
"loss": 3.0402,
|
| 48065 |
+
"step": 66990
|
| 48066 |
+
},
|
| 48067 |
+
{
|
| 48068 |
+
"epoch": 8.5e-06,
|
| 48069 |
+
"grad_norm": 0.47265625,
|
| 48070 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48071 |
+
"loss": 3.0291,
|
| 48072 |
+
"step": 67000
|
| 48073 |
+
},
|
| 48074 |
+
{
|
| 48075 |
+
"epoch": 8.5e-06,
|
| 48076 |
+
"eval_loss": 3.0823678970336914,
|
| 48077 |
+
"eval_runtime": 0.0605,
|
| 48078 |
+
"eval_samples_per_second": 528.633,
|
| 48079 |
+
"eval_steps_per_second": 16.52,
|
| 48080 |
+
"step": 67000
|
| 48081 |
+
},
|
| 48082 |
+
{
|
| 48083 |
+
"epoch": 8.505e-06,
|
| 48084 |
+
"grad_norm": 0.466796875,
|
| 48085 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48086 |
+
"loss": 3.0439,
|
| 48087 |
+
"step": 67010
|
| 48088 |
+
},
|
| 48089 |
+
{
|
| 48090 |
+
"epoch": 8.51e-06,
|
| 48091 |
+
"grad_norm": 0.48828125,
|
| 48092 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48093 |
+
"loss": 3.0319,
|
| 48094 |
+
"step": 67020
|
| 48095 |
+
},
|
| 48096 |
+
{
|
| 48097 |
+
"epoch": 8.515e-06,
|
| 48098 |
+
"grad_norm": 0.447265625,
|
| 48099 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48100 |
+
"loss": 3.0308,
|
| 48101 |
+
"step": 67030
|
| 48102 |
+
},
|
| 48103 |
+
{
|
| 48104 |
+
"epoch": 8.52e-06,
|
| 48105 |
+
"grad_norm": 0.462890625,
|
| 48106 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48107 |
+
"loss": 3.0466,
|
| 48108 |
+
"step": 67040
|
| 48109 |
+
},
|
| 48110 |
+
{
|
| 48111 |
+
"epoch": 8.525e-06,
|
| 48112 |
+
"grad_norm": 0.48046875,
|
| 48113 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48114 |
+
"loss": 3.0416,
|
| 48115 |
+
"step": 67050
|
| 48116 |
+
},
|
| 48117 |
+
{
|
| 48118 |
+
"epoch": 8.53e-06,
|
| 48119 |
+
"grad_norm": 0.44921875,
|
| 48120 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48121 |
+
"loss": 3.0737,
|
| 48122 |
+
"step": 67060
|
| 48123 |
+
},
|
| 48124 |
+
{
|
| 48125 |
+
"epoch": 8.535e-06,
|
| 48126 |
+
"grad_norm": 0.5625,
|
| 48127 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48128 |
+
"loss": 3.0387,
|
| 48129 |
+
"step": 67070
|
| 48130 |
+
},
|
| 48131 |
+
{
|
| 48132 |
+
"epoch": 8.54e-06,
|
| 48133 |
+
"grad_norm": 0.5234375,
|
| 48134 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48135 |
+
"loss": 3.0537,
|
| 48136 |
+
"step": 67080
|
| 48137 |
+
},
|
| 48138 |
+
{
|
| 48139 |
+
"epoch": 8.545e-06,
|
| 48140 |
+
"grad_norm": 0.453125,
|
| 48141 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48142 |
+
"loss": 3.0475,
|
| 48143 |
+
"step": 67090
|
| 48144 |
+
},
|
| 48145 |
+
{
|
| 48146 |
+
"epoch": 8.55e-06,
|
| 48147 |
+
"grad_norm": 0.4765625,
|
| 48148 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48149 |
+
"loss": 3.0571,
|
| 48150 |
+
"step": 67100
|
| 48151 |
+
},
|
| 48152 |
+
{
|
| 48153 |
+
"epoch": 8.555e-06,
|
| 48154 |
+
"grad_norm": 0.470703125,
|
| 48155 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48156 |
+
"loss": 3.0331,
|
| 48157 |
+
"step": 67110
|
| 48158 |
+
},
|
| 48159 |
+
{
|
| 48160 |
+
"epoch": 8.56e-06,
|
| 48161 |
+
"grad_norm": 0.455078125,
|
| 48162 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48163 |
+
"loss": 3.0373,
|
| 48164 |
+
"step": 67120
|
| 48165 |
+
},
|
| 48166 |
+
{
|
| 48167 |
+
"epoch": 8.565e-06,
|
| 48168 |
+
"grad_norm": 0.4609375,
|
| 48169 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48170 |
+
"loss": 3.0535,
|
| 48171 |
+
"step": 67130
|
| 48172 |
+
},
|
| 48173 |
+
{
|
| 48174 |
+
"epoch": 8.57e-06,
|
| 48175 |
+
"grad_norm": 0.44921875,
|
| 48176 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48177 |
+
"loss": 3.0475,
|
| 48178 |
+
"step": 67140
|
| 48179 |
+
},
|
| 48180 |
+
{
|
| 48181 |
+
"epoch": 8.575e-06,
|
| 48182 |
+
"grad_norm": 0.46875,
|
| 48183 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48184 |
+
"loss": 3.0467,
|
| 48185 |
+
"step": 67150
|
| 48186 |
+
},
|
| 48187 |
+
{
|
| 48188 |
+
"epoch": 8.58e-06,
|
| 48189 |
+
"grad_norm": 0.458984375,
|
| 48190 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48191 |
+
"loss": 3.0431,
|
| 48192 |
+
"step": 67160
|
| 48193 |
+
},
|
| 48194 |
+
{
|
| 48195 |
+
"epoch": 8.585e-06,
|
| 48196 |
+
"grad_norm": 0.458984375,
|
| 48197 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48198 |
+
"loss": 3.0411,
|
| 48199 |
+
"step": 67170
|
| 48200 |
+
},
|
| 48201 |
+
{
|
| 48202 |
+
"epoch": 8.59e-06,
|
| 48203 |
+
"grad_norm": 0.486328125,
|
| 48204 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48205 |
+
"loss": 3.0203,
|
| 48206 |
+
"step": 67180
|
| 48207 |
+
},
|
| 48208 |
+
{
|
| 48209 |
+
"epoch": 8.595e-06,
|
| 48210 |
+
"grad_norm": 0.43359375,
|
| 48211 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48212 |
+
"loss": 3.025,
|
| 48213 |
+
"step": 67190
|
| 48214 |
+
},
|
| 48215 |
+
{
|
| 48216 |
+
"epoch": 8.6e-06,
|
| 48217 |
+
"grad_norm": 0.49609375,
|
| 48218 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48219 |
+
"loss": 3.0311,
|
| 48220 |
+
"step": 67200
|
| 48221 |
+
},
|
| 48222 |
+
{
|
| 48223 |
+
"epoch": 8.605e-06,
|
| 48224 |
+
"grad_norm": 0.447265625,
|
| 48225 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48226 |
+
"loss": 3.0402,
|
| 48227 |
+
"step": 67210
|
| 48228 |
+
},
|
| 48229 |
+
{
|
| 48230 |
+
"epoch": 8.61e-06,
|
| 48231 |
+
"grad_norm": 0.470703125,
|
| 48232 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48233 |
+
"loss": 3.0429,
|
| 48234 |
+
"step": 67220
|
| 48235 |
+
},
|
| 48236 |
+
{
|
| 48237 |
+
"epoch": 8.615e-06,
|
| 48238 |
+
"grad_norm": 0.443359375,
|
| 48239 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48240 |
+
"loss": 3.0174,
|
| 48241 |
+
"step": 67230
|
| 48242 |
+
},
|
| 48243 |
+
{
|
| 48244 |
+
"epoch": 8.62e-06,
|
| 48245 |
+
"grad_norm": 0.490234375,
|
| 48246 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48247 |
+
"loss": 3.0341,
|
| 48248 |
+
"step": 67240
|
| 48249 |
+
},
|
| 48250 |
+
{
|
| 48251 |
+
"epoch": 8.625e-06,
|
| 48252 |
+
"grad_norm": 0.482421875,
|
| 48253 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48254 |
+
"loss": 3.0394,
|
| 48255 |
+
"step": 67250
|
| 48256 |
+
},
|
| 48257 |
+
{
|
| 48258 |
+
"epoch": 8.63e-06,
|
| 48259 |
+
"grad_norm": 0.462890625,
|
| 48260 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48261 |
+
"loss": 3.0389,
|
| 48262 |
+
"step": 67260
|
| 48263 |
+
},
|
| 48264 |
+
{
|
| 48265 |
+
"epoch": 8.635e-06,
|
| 48266 |
+
"grad_norm": 0.4609375,
|
| 48267 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48268 |
+
"loss": 3.0345,
|
| 48269 |
+
"step": 67270
|
| 48270 |
+
},
|
| 48271 |
+
{
|
| 48272 |
+
"epoch": 8.64e-06,
|
| 48273 |
+
"grad_norm": 0.443359375,
|
| 48274 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48275 |
+
"loss": 3.0341,
|
| 48276 |
+
"step": 67280
|
| 48277 |
+
},
|
| 48278 |
+
{
|
| 48279 |
+
"epoch": 8.645e-06,
|
| 48280 |
+
"grad_norm": 0.4609375,
|
| 48281 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48282 |
+
"loss": 3.0441,
|
| 48283 |
+
"step": 67290
|
| 48284 |
+
},
|
| 48285 |
+
{
|
| 48286 |
+
"epoch": 8.65e-06,
|
| 48287 |
+
"grad_norm": 0.470703125,
|
| 48288 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48289 |
+
"loss": 3.0313,
|
| 48290 |
+
"step": 67300
|
| 48291 |
+
},
|
| 48292 |
+
{
|
| 48293 |
+
"epoch": 8.655e-06,
|
| 48294 |
+
"grad_norm": 0.47265625,
|
| 48295 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48296 |
+
"loss": 3.037,
|
| 48297 |
+
"step": 67310
|
| 48298 |
+
},
|
| 48299 |
+
{
|
| 48300 |
+
"epoch": 8.66e-06,
|
| 48301 |
+
"grad_norm": 0.47265625,
|
| 48302 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48303 |
+
"loss": 3.0681,
|
| 48304 |
+
"step": 67320
|
| 48305 |
+
},
|
| 48306 |
+
{
|
| 48307 |
+
"epoch": 8.665e-06,
|
| 48308 |
+
"grad_norm": 0.5,
|
| 48309 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48310 |
+
"loss": 3.0329,
|
| 48311 |
+
"step": 67330
|
| 48312 |
+
},
|
| 48313 |
+
{
|
| 48314 |
+
"epoch": 8.67e-06,
|
| 48315 |
+
"grad_norm": 0.474609375,
|
| 48316 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48317 |
+
"loss": 3.0445,
|
| 48318 |
+
"step": 67340
|
| 48319 |
+
},
|
| 48320 |
+
{
|
| 48321 |
+
"epoch": 8.675e-06,
|
| 48322 |
+
"grad_norm": 0.482421875,
|
| 48323 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48324 |
+
"loss": 3.0357,
|
| 48325 |
+
"step": 67350
|
| 48326 |
+
},
|
| 48327 |
+
{
|
| 48328 |
+
"epoch": 8.68e-06,
|
| 48329 |
+
"grad_norm": 0.48046875,
|
| 48330 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48331 |
+
"loss": 3.048,
|
| 48332 |
+
"step": 67360
|
| 48333 |
+
},
|
| 48334 |
+
{
|
| 48335 |
+
"epoch": 8.685e-06,
|
| 48336 |
+
"grad_norm": 0.45703125,
|
| 48337 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48338 |
+
"loss": 3.0432,
|
| 48339 |
+
"step": 67370
|
| 48340 |
+
},
|
| 48341 |
+
{
|
| 48342 |
+
"epoch": 8.69e-06,
|
| 48343 |
+
"grad_norm": 0.50390625,
|
| 48344 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48345 |
+
"loss": 3.0405,
|
| 48346 |
+
"step": 67380
|
| 48347 |
+
},
|
| 48348 |
+
{
|
| 48349 |
+
"epoch": 8.695e-06,
|
| 48350 |
+
"grad_norm": 0.44921875,
|
| 48351 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48352 |
+
"loss": 3.053,
|
| 48353 |
+
"step": 67390
|
| 48354 |
+
},
|
| 48355 |
+
{
|
| 48356 |
+
"epoch": 8.7e-06,
|
| 48357 |
+
"grad_norm": 0.45703125,
|
| 48358 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48359 |
+
"loss": 3.0219,
|
| 48360 |
+
"step": 67400
|
| 48361 |
+
},
|
| 48362 |
+
{
|
| 48363 |
+
"epoch": 8.705e-06,
|
| 48364 |
+
"grad_norm": 0.53125,
|
| 48365 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48366 |
+
"loss": 3.0494,
|
| 48367 |
+
"step": 67410
|
| 48368 |
+
},
|
| 48369 |
+
{
|
| 48370 |
+
"epoch": 8.71e-06,
|
| 48371 |
+
"grad_norm": 0.515625,
|
| 48372 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48373 |
+
"loss": 3.0391,
|
| 48374 |
+
"step": 67420
|
| 48375 |
+
},
|
| 48376 |
+
{
|
| 48377 |
+
"epoch": 8.715e-06,
|
| 48378 |
+
"grad_norm": 0.5390625,
|
| 48379 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48380 |
+
"loss": 3.0182,
|
| 48381 |
+
"step": 67430
|
| 48382 |
+
},
|
| 48383 |
+
{
|
| 48384 |
+
"epoch": 8.72e-06,
|
| 48385 |
+
"grad_norm": 0.48828125,
|
| 48386 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48387 |
+
"loss": 3.0384,
|
| 48388 |
+
"step": 67440
|
| 48389 |
+
},
|
| 48390 |
+
{
|
| 48391 |
+
"epoch": 8.725e-06,
|
| 48392 |
+
"grad_norm": 0.46875,
|
| 48393 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48394 |
+
"loss": 3.051,
|
| 48395 |
+
"step": 67450
|
| 48396 |
+
},
|
| 48397 |
+
{
|
| 48398 |
+
"epoch": 8.73e-06,
|
| 48399 |
+
"grad_norm": 0.47265625,
|
| 48400 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48401 |
+
"loss": 3.0575,
|
| 48402 |
+
"step": 67460
|
| 48403 |
+
},
|
| 48404 |
+
{
|
| 48405 |
+
"epoch": 8.735e-06,
|
| 48406 |
+
"grad_norm": 0.453125,
|
| 48407 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48408 |
+
"loss": 3.0291,
|
| 48409 |
+
"step": 67470
|
| 48410 |
+
},
|
| 48411 |
+
{
|
| 48412 |
+
"epoch": 8.74e-06,
|
| 48413 |
+
"grad_norm": 0.458984375,
|
| 48414 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48415 |
+
"loss": 3.0491,
|
| 48416 |
+
"step": 67480
|
| 48417 |
+
},
|
| 48418 |
+
{
|
| 48419 |
+
"epoch": 8.745e-06,
|
| 48420 |
+
"grad_norm": 0.4453125,
|
| 48421 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48422 |
+
"loss": 3.0519,
|
| 48423 |
+
"step": 67490
|
| 48424 |
+
},
|
| 48425 |
+
{
|
| 48426 |
+
"epoch": 8.75e-06,
|
| 48427 |
+
"grad_norm": 0.451171875,
|
| 48428 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48429 |
+
"loss": 3.0268,
|
| 48430 |
+
"step": 67500
|
| 48431 |
+
},
|
| 48432 |
+
{
|
| 48433 |
+
"epoch": 8.75e-06,
|
| 48434 |
+
"eval_loss": 3.081859588623047,
|
| 48435 |
+
"eval_runtime": 0.0608,
|
| 48436 |
+
"eval_samples_per_second": 526.317,
|
| 48437 |
+
"eval_steps_per_second": 16.447,
|
| 48438 |
+
"step": 67500
|
| 48439 |
+
},
|
| 48440 |
+
{
|
| 48441 |
+
"epoch": 8.755e-06,
|
| 48442 |
+
"grad_norm": 0.458984375,
|
| 48443 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48444 |
+
"loss": 3.0417,
|
| 48445 |
+
"step": 67510
|
| 48446 |
+
},
|
| 48447 |
+
{
|
| 48448 |
+
"epoch": 8.76e-06,
|
| 48449 |
+
"grad_norm": 0.455078125,
|
| 48450 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48451 |
+
"loss": 3.0442,
|
| 48452 |
+
"step": 67520
|
| 48453 |
+
},
|
| 48454 |
+
{
|
| 48455 |
+
"epoch": 8.765e-06,
|
| 48456 |
+
"grad_norm": 0.4765625,
|
| 48457 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48458 |
+
"loss": 3.053,
|
| 48459 |
+
"step": 67530
|
| 48460 |
+
},
|
| 48461 |
+
{
|
| 48462 |
+
"epoch": 8.77e-06,
|
| 48463 |
+
"grad_norm": 0.443359375,
|
| 48464 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48465 |
+
"loss": 3.0442,
|
| 48466 |
+
"step": 67540
|
| 48467 |
+
},
|
| 48468 |
+
{
|
| 48469 |
+
"epoch": 8.775e-06,
|
| 48470 |
+
"grad_norm": 0.458984375,
|
| 48471 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48472 |
+
"loss": 3.0293,
|
| 48473 |
+
"step": 67550
|
| 48474 |
+
},
|
| 48475 |
+
{
|
| 48476 |
+
"epoch": 8.78e-06,
|
| 48477 |
+
"grad_norm": 0.45703125,
|
| 48478 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48479 |
+
"loss": 3.0321,
|
| 48480 |
+
"step": 67560
|
| 48481 |
+
},
|
| 48482 |
+
{
|
| 48483 |
+
"epoch": 8.785e-06,
|
| 48484 |
+
"grad_norm": 0.484375,
|
| 48485 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48486 |
+
"loss": 3.0391,
|
| 48487 |
+
"step": 67570
|
| 48488 |
+
},
|
| 48489 |
+
{
|
| 48490 |
+
"epoch": 8.79e-06,
|
| 48491 |
+
"grad_norm": 0.4765625,
|
| 48492 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48493 |
+
"loss": 3.0509,
|
| 48494 |
+
"step": 67580
|
| 48495 |
+
},
|
| 48496 |
+
{
|
| 48497 |
+
"epoch": 8.795e-06,
|
| 48498 |
+
"grad_norm": 0.46875,
|
| 48499 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48500 |
+
"loss": 3.038,
|
| 48501 |
+
"step": 67590
|
| 48502 |
+
},
|
| 48503 |
+
{
|
| 48504 |
+
"epoch": 8.8e-06,
|
| 48505 |
+
"grad_norm": 0.44921875,
|
| 48506 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48507 |
+
"loss": 3.0559,
|
| 48508 |
+
"step": 67600
|
| 48509 |
+
},
|
| 48510 |
+
{
|
| 48511 |
+
"epoch": 8.805e-06,
|
| 48512 |
+
"grad_norm": 0.47265625,
|
| 48513 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48514 |
+
"loss": 3.0248,
|
| 48515 |
+
"step": 67610
|
| 48516 |
+
},
|
| 48517 |
+
{
|
| 48518 |
+
"epoch": 8.81e-06,
|
| 48519 |
+
"grad_norm": 0.443359375,
|
| 48520 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48521 |
+
"loss": 3.0079,
|
| 48522 |
+
"step": 67620
|
| 48523 |
+
},
|
| 48524 |
+
{
|
| 48525 |
+
"epoch": 8.815e-06,
|
| 48526 |
+
"grad_norm": 0.439453125,
|
| 48527 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48528 |
+
"loss": 3.0462,
|
| 48529 |
+
"step": 67630
|
| 48530 |
+
},
|
| 48531 |
+
{
|
| 48532 |
+
"epoch": 8.82e-06,
|
| 48533 |
+
"grad_norm": 0.43359375,
|
| 48534 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48535 |
+
"loss": 3.0348,
|
| 48536 |
+
"step": 67640
|
| 48537 |
+
},
|
| 48538 |
+
{
|
| 48539 |
+
"epoch": 8.825e-06,
|
| 48540 |
+
"grad_norm": 0.484375,
|
| 48541 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48542 |
+
"loss": 3.013,
|
| 48543 |
+
"step": 67650
|
| 48544 |
+
},
|
| 48545 |
+
{
|
| 48546 |
+
"epoch": 8.83e-06,
|
| 48547 |
+
"grad_norm": 0.443359375,
|
| 48548 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48549 |
+
"loss": 3.0469,
|
| 48550 |
+
"step": 67660
|
| 48551 |
+
},
|
| 48552 |
+
{
|
| 48553 |
+
"epoch": 8.835e-06,
|
| 48554 |
+
"grad_norm": 0.462890625,
|
| 48555 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48556 |
+
"loss": 3.0416,
|
| 48557 |
+
"step": 67670
|
| 48558 |
+
},
|
| 48559 |
+
{
|
| 48560 |
+
"epoch": 8.84e-06,
|
| 48561 |
+
"grad_norm": 0.458984375,
|
| 48562 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48563 |
+
"loss": 3.0634,
|
| 48564 |
+
"step": 67680
|
| 48565 |
+
},
|
| 48566 |
+
{
|
| 48567 |
+
"epoch": 8.845e-06,
|
| 48568 |
+
"grad_norm": 0.490234375,
|
| 48569 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48570 |
+
"loss": 3.0573,
|
| 48571 |
+
"step": 67690
|
| 48572 |
+
},
|
| 48573 |
+
{
|
| 48574 |
+
"epoch": 8.85e-06,
|
| 48575 |
+
"grad_norm": 0.44140625,
|
| 48576 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48577 |
+
"loss": 3.026,
|
| 48578 |
+
"step": 67700
|
| 48579 |
+
},
|
| 48580 |
+
{
|
| 48581 |
+
"epoch": 8.855e-06,
|
| 48582 |
+
"grad_norm": 0.439453125,
|
| 48583 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48584 |
+
"loss": 3.03,
|
| 48585 |
+
"step": 67710
|
| 48586 |
+
},
|
| 48587 |
+
{
|
| 48588 |
+
"epoch": 8.86e-06,
|
| 48589 |
+
"grad_norm": 0.484375,
|
| 48590 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48591 |
+
"loss": 3.0547,
|
| 48592 |
+
"step": 67720
|
| 48593 |
+
},
|
| 48594 |
+
{
|
| 48595 |
+
"epoch": 8.865e-06,
|
| 48596 |
+
"grad_norm": 0.48828125,
|
| 48597 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48598 |
+
"loss": 3.0614,
|
| 48599 |
+
"step": 67730
|
| 48600 |
+
},
|
| 48601 |
+
{
|
| 48602 |
+
"epoch": 8.87e-06,
|
| 48603 |
+
"grad_norm": 0.4609375,
|
| 48604 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48605 |
+
"loss": 3.0263,
|
| 48606 |
+
"step": 67740
|
| 48607 |
+
},
|
| 48608 |
+
{
|
| 48609 |
+
"epoch": 8.875e-06,
|
| 48610 |
+
"grad_norm": 0.470703125,
|
| 48611 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48612 |
+
"loss": 3.0236,
|
| 48613 |
+
"step": 67750
|
| 48614 |
+
},
|
| 48615 |
+
{
|
| 48616 |
+
"epoch": 8.88e-06,
|
| 48617 |
+
"grad_norm": 0.462890625,
|
| 48618 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48619 |
+
"loss": 3.0457,
|
| 48620 |
+
"step": 67760
|
| 48621 |
+
},
|
| 48622 |
+
{
|
| 48623 |
+
"epoch": 8.885e-06,
|
| 48624 |
+
"grad_norm": 0.52734375,
|
| 48625 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48626 |
+
"loss": 3.0458,
|
| 48627 |
+
"step": 67770
|
| 48628 |
+
},
|
| 48629 |
+
{
|
| 48630 |
+
"epoch": 8.89e-06,
|
| 48631 |
+
"grad_norm": 0.4609375,
|
| 48632 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48633 |
+
"loss": 3.0455,
|
| 48634 |
+
"step": 67780
|
| 48635 |
+
},
|
| 48636 |
+
{
|
| 48637 |
+
"epoch": 8.895e-06,
|
| 48638 |
+
"grad_norm": 0.466796875,
|
| 48639 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48640 |
+
"loss": 3.0376,
|
| 48641 |
+
"step": 67790
|
| 48642 |
+
},
|
| 48643 |
+
{
|
| 48644 |
+
"epoch": 8.9e-06,
|
| 48645 |
+
"grad_norm": 0.5,
|
| 48646 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48647 |
+
"loss": 3.0155,
|
| 48648 |
+
"step": 67800
|
| 48649 |
+
},
|
| 48650 |
+
{
|
| 48651 |
+
"epoch": 8.905e-06,
|
| 48652 |
+
"grad_norm": 0.484375,
|
| 48653 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48654 |
+
"loss": 3.0315,
|
| 48655 |
+
"step": 67810
|
| 48656 |
+
},
|
| 48657 |
+
{
|
| 48658 |
+
"epoch": 8.91e-06,
|
| 48659 |
+
"grad_norm": 0.44921875,
|
| 48660 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48661 |
+
"loss": 3.0545,
|
| 48662 |
+
"step": 67820
|
| 48663 |
+
},
|
| 48664 |
+
{
|
| 48665 |
+
"epoch": 8.915e-06,
|
| 48666 |
+
"grad_norm": 0.4609375,
|
| 48667 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48668 |
+
"loss": 3.0477,
|
| 48669 |
+
"step": 67830
|
| 48670 |
+
},
|
| 48671 |
+
{
|
| 48672 |
+
"epoch": 8.92e-06,
|
| 48673 |
+
"grad_norm": 0.45703125,
|
| 48674 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48675 |
+
"loss": 3.0478,
|
| 48676 |
+
"step": 67840
|
| 48677 |
+
},
|
| 48678 |
+
{
|
| 48679 |
+
"epoch": 8.925e-06,
|
| 48680 |
+
"grad_norm": 0.44140625,
|
| 48681 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48682 |
+
"loss": 3.0601,
|
| 48683 |
+
"step": 67850
|
| 48684 |
+
},
|
| 48685 |
+
{
|
| 48686 |
+
"epoch": 8.93e-06,
|
| 48687 |
+
"grad_norm": 0.447265625,
|
| 48688 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48689 |
+
"loss": 3.0103,
|
| 48690 |
+
"step": 67860
|
| 48691 |
+
},
|
| 48692 |
+
{
|
| 48693 |
+
"epoch": 8.935e-06,
|
| 48694 |
+
"grad_norm": 0.458984375,
|
| 48695 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48696 |
+
"loss": 3.0529,
|
| 48697 |
+
"step": 67870
|
| 48698 |
+
},
|
| 48699 |
+
{
|
| 48700 |
+
"epoch": 8.94e-06,
|
| 48701 |
+
"grad_norm": 0.466796875,
|
| 48702 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48703 |
+
"loss": 3.0444,
|
| 48704 |
+
"step": 67880
|
| 48705 |
+
},
|
| 48706 |
+
{
|
| 48707 |
+
"epoch": 8.945e-06,
|
| 48708 |
+
"grad_norm": 0.494140625,
|
| 48709 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48710 |
+
"loss": 3.027,
|
| 48711 |
+
"step": 67890
|
| 48712 |
+
},
|
| 48713 |
+
{
|
| 48714 |
+
"epoch": 8.95e-06,
|
| 48715 |
+
"grad_norm": 0.478515625,
|
| 48716 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48717 |
+
"loss": 3.0434,
|
| 48718 |
+
"step": 67900
|
| 48719 |
+
},
|
| 48720 |
+
{
|
| 48721 |
+
"epoch": 8.955e-06,
|
| 48722 |
+
"grad_norm": 0.458984375,
|
| 48723 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48724 |
+
"loss": 3.0388,
|
| 48725 |
+
"step": 67910
|
| 48726 |
+
},
|
| 48727 |
+
{
|
| 48728 |
+
"epoch": 8.96e-06,
|
| 48729 |
+
"grad_norm": 0.46875,
|
| 48730 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48731 |
+
"loss": 3.0308,
|
| 48732 |
+
"step": 67920
|
| 48733 |
+
},
|
| 48734 |
+
{
|
| 48735 |
+
"epoch": 8.965e-06,
|
| 48736 |
+
"grad_norm": 0.51171875,
|
| 48737 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48738 |
+
"loss": 3.0558,
|
| 48739 |
+
"step": 67930
|
| 48740 |
+
},
|
| 48741 |
+
{
|
| 48742 |
+
"epoch": 8.97e-06,
|
| 48743 |
+
"grad_norm": 0.447265625,
|
| 48744 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48745 |
+
"loss": 3.0221,
|
| 48746 |
+
"step": 67940
|
| 48747 |
+
},
|
| 48748 |
+
{
|
| 48749 |
+
"epoch": 8.975e-06,
|
| 48750 |
+
"grad_norm": 0.494140625,
|
| 48751 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48752 |
+
"loss": 3.0485,
|
| 48753 |
+
"step": 67950
|
| 48754 |
+
},
|
| 48755 |
+
{
|
| 48756 |
+
"epoch": 8.98e-06,
|
| 48757 |
+
"grad_norm": 0.58203125,
|
| 48758 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48759 |
+
"loss": 3.0545,
|
| 48760 |
+
"step": 67960
|
| 48761 |
+
},
|
| 48762 |
+
{
|
| 48763 |
+
"epoch": 8.985e-06,
|
| 48764 |
+
"grad_norm": 0.48828125,
|
| 48765 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48766 |
+
"loss": 3.0355,
|
| 48767 |
+
"step": 67970
|
| 48768 |
+
},
|
| 48769 |
+
{
|
| 48770 |
+
"epoch": 8.99e-06,
|
| 48771 |
+
"grad_norm": 0.494140625,
|
| 48772 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48773 |
+
"loss": 3.0394,
|
| 48774 |
+
"step": 67980
|
| 48775 |
+
},
|
| 48776 |
+
{
|
| 48777 |
+
"epoch": 8.995e-06,
|
| 48778 |
+
"grad_norm": 0.494140625,
|
| 48779 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48780 |
+
"loss": 3.0447,
|
| 48781 |
+
"step": 67990
|
| 48782 |
+
},
|
| 48783 |
+
{
|
| 48784 |
+
"epoch": 9e-06,
|
| 48785 |
+
"grad_norm": 0.439453125,
|
| 48786 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 48787 |
+
"loss": 3.0356,
|
| 48788 |
+
"step": 68000
|
| 48789 |
+
},
|
| 48790 |
+
{
|
| 48791 |
+
"epoch": 9e-06,
|
| 48792 |
+
"eval_loss": 3.0828726291656494,
|
| 48793 |
+
"eval_runtime": 0.0615,
|
| 48794 |
+
"eval_samples_per_second": 520.008,
|
| 48795 |
+
"eval_steps_per_second": 16.25,
|
| 48796 |
+
"step": 68000
|
| 48797 |
}
|
| 48798 |
],
|
| 48799 |
"logging_steps": 10,
|
|
|
|
| 48813 |
"attributes": {}
|
| 48814 |
}
|
| 48815 |
},
|
| 48816 |
+
"total_flos": 1.50614323494912e+18,
|
| 48817 |
"train_batch_size": 240,
|
| 48818 |
"trial_name": null,
|
| 48819 |
"trial_params": null
|