Deep35 continual checkpoint step 52000
Browse files- last-checkpoint/continual_state.json +34 -11
- last-checkpoint/data_state.pt +2 -2
- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +1436 -4
- last-checkpoint/training_args.bin +1 -1
last-checkpoint/continual_state.json
CHANGED
|
@@ -1,12 +1,12 @@
|
|
| 1 |
{
|
| 2 |
-
"global_step":
|
| 3 |
-
"sequence_tokens_emitted":
|
| 4 |
"mix": [
|
| 5 |
-
0.
|
|
|
|
| 6 |
0.05,
|
| 7 |
-
0.
|
| 8 |
-
0.
|
| 9 |
-
0.0025
|
| 10 |
],
|
| 11 |
"fineweb": "HuggingFaceFW/fineweb-edu:default",
|
| 12 |
"ultra_multi": "openbmb/Ultra-FineWeb-L3:Ultra-FineWeb-L3-en-Multi-Style-Synthetic",
|
|
@@ -27,11 +27,34 @@
|
|
| 27 |
0.05,
|
| 28 |
0.01
|
| 29 |
],
|
| 30 |
-
"formatter_version":
|
| 31 |
"rewrite6_format": "join non-empty instruction/text/output; empty components valid",
|
| 32 |
"noprompt_format": "stored text field exactly; empty rows skipped",
|
| 33 |
-
"
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"global_step": 52000,
|
| 3 |
+
"sequence_tokens_emitted": 7045365760,
|
| 4 |
"mix": [
|
| 5 |
+
0.8,
|
| 6 |
+
0.09,
|
| 7 |
0.05,
|
| 8 |
+
0.05,
|
| 9 |
+
0.01
|
|
|
|
| 10 |
],
|
| 11 |
"fineweb": "HuggingFaceFW/fineweb-edu:default",
|
| 12 |
"ultra_multi": "openbmb/Ultra-FineWeb-L3:Ultra-FineWeb-L3-en-Multi-Style-Synthetic",
|
|
|
|
| 27 |
0.05,
|
| 28 |
0.01
|
| 29 |
],
|
| 30 |
+
"formatter_version": 8,
|
| 31 |
"rewrite6_format": "join non-empty instruction/text/output; empty components valid",
|
| 32 |
"noprompt_format": "stored text field exactly; empty rows skipped",
|
| 33 |
+
"batch_history": [
|
| 34 |
+
{
|
| 35 |
+
"start_step": 0,
|
| 36 |
+
"start_tokens": 0,
|
| 37 |
+
"batch_size": 128,
|
| 38 |
+
"gradient_accumulation": 1,
|
| 39 |
+
"tokens_per_update": 131072
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"start_step": 49000,
|
| 43 |
+
"start_tokens": 6422528000,
|
| 44 |
+
"batch_size": 128,
|
| 45 |
+
"gradient_accumulation": 1,
|
| 46 |
+
"tokens_per_update": 131072
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"start_step": 50000,
|
| 50 |
+
"start_tokens": 6553600000,
|
| 51 |
+
"batch_size": 240,
|
| 52 |
+
"gradient_accumulation": 1,
|
| 53 |
+
"tokens_per_update": 245760
|
| 54 |
+
}
|
| 55 |
+
],
|
| 56 |
+
"current_batch_size": 240,
|
| 57 |
+
"current_gradient_accumulation": 1,
|
| 58 |
+
"current_tokens_per_update": 245760,
|
| 59 |
+
"cumulative_tokens_at_save": 7045120000
|
| 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 sha256:6018e87e1d0334942faf2ca8006834d328ef0a655cc37eae5601c73df67fa2af
|
| 3 |
+
size 34815
|
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 sha256:28a0a5f0d9a2dfc1c5d4ce789ea281c3bcad887cb5f4cc649ee9eb43c221bfc4
|
| 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:ceeb7c2ff7ecc71f60f0bba40d9fb22ce159d7909ee43dfe6d023401d2e08724
|
| 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 sha256:411a2ade309b999580a46833a3bd3c8b5b3020eaf0487ae796aa8a71afb9d717
|
| 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 sha256:4f2e183f79b66d432124622640c816b53abe78dfca0ed80382d5d280004c8b22
|
| 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,
|
|
@@ -35906,6 +35906,1438 @@
|
|
| 35906 |
"eval_samples_per_second": 530.541,
|
| 35907 |
"eval_steps_per_second": 16.579,
|
| 35908 |
"step": 50000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 35909 |
}
|
| 35910 |
],
|
| 35911 |
"logging_steps": 10,
|
|
@@ -35925,8 +37357,8 @@
|
|
| 35925 |
"attributes": {}
|
| 35926 |
}
|
| 35927 |
},
|
| 35928 |
-
"total_flos":
|
| 35929 |
-
"train_batch_size":
|
| 35930 |
"trial_name": null,
|
| 35931 |
"trial_params": null
|
| 35932 |
}
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1e-06,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 52000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 35906 |
"eval_samples_per_second": 530.541,
|
| 35907 |
"eval_steps_per_second": 16.579,
|
| 35908 |
"step": 50000
|
| 35909 |
+
},
|
| 35910 |
+
{
|
| 35911 |
+
"epoch": 5e-09,
|
| 35912 |
+
"grad_norm": 0.48828125,
|
| 35913 |
+
"learning_rate": 0.00010246413961327904,
|
| 35914 |
+
"loss": 3.077,
|
| 35915 |
+
"step": 50010
|
| 35916 |
+
},
|
| 35917 |
+
{
|
| 35918 |
+
"epoch": 1e-08,
|
| 35919 |
+
"grad_norm": 0.51171875,
|
| 35920 |
+
"learning_rate": 0.00010237117601769932,
|
| 35921 |
+
"loss": 3.0892,
|
| 35922 |
+
"step": 50020
|
| 35923 |
+
},
|
| 35924 |
+
{
|
| 35925 |
+
"epoch": 1.5e-08,
|
| 35926 |
+
"grad_norm": 0.58203125,
|
| 35927 |
+
"learning_rate": 0.000102278250249607,
|
| 35928 |
+
"loss": 3.0528,
|
| 35929 |
+
"step": 50030
|
| 35930 |
+
},
|
| 35931 |
+
{
|
| 35932 |
+
"epoch": 2e-08,
|
| 35933 |
+
"grad_norm": 0.478515625,
|
| 35934 |
+
"learning_rate": 0.000102185362365129,
|
| 35935 |
+
"loss": 3.0804,
|
| 35936 |
+
"step": 50040
|
| 35937 |
+
},
|
| 35938 |
+
{
|
| 35939 |
+
"epoch": 2.5e-08,
|
| 35940 |
+
"grad_norm": 0.47265625,
|
| 35941 |
+
"learning_rate": 0.00010209251242036886,
|
| 35942 |
+
"loss": 3.0685,
|
| 35943 |
+
"step": 50050
|
| 35944 |
+
},
|
| 35945 |
+
{
|
| 35946 |
+
"epoch": 3e-08,
|
| 35947 |
+
"grad_norm": 0.462890625,
|
| 35948 |
+
"learning_rate": 0.00010199970047140765,
|
| 35949 |
+
"loss": 3.0533,
|
| 35950 |
+
"step": 50060
|
| 35951 |
+
},
|
| 35952 |
+
{
|
| 35953 |
+
"epoch": 3.5e-08,
|
| 35954 |
+
"grad_norm": 0.4453125,
|
| 35955 |
+
"learning_rate": 0.00010190692657430326,
|
| 35956 |
+
"loss": 3.0581,
|
| 35957 |
+
"step": 50070
|
| 35958 |
+
},
|
| 35959 |
+
{
|
| 35960 |
+
"epoch": 4e-08,
|
| 35961 |
+
"grad_norm": 0.482421875,
|
| 35962 |
+
"learning_rate": 0.00010181419078509062,
|
| 35963 |
+
"loss": 3.0981,
|
| 35964 |
+
"step": 50080
|
| 35965 |
+
},
|
| 35966 |
+
{
|
| 35967 |
+
"epoch": 4.5e-08,
|
| 35968 |
+
"grad_norm": 0.46875,
|
| 35969 |
+
"learning_rate": 0.0001017214931597817,
|
| 35970 |
+
"loss": 3.0777,
|
| 35971 |
+
"step": 50090
|
| 35972 |
+
},
|
| 35973 |
+
{
|
| 35974 |
+
"epoch": 5e-08,
|
| 35975 |
+
"grad_norm": 0.5078125,
|
| 35976 |
+
"learning_rate": 0.00010162883375436538,
|
| 35977 |
+
"loss": 3.0681,
|
| 35978 |
+
"step": 50100
|
| 35979 |
+
},
|
| 35980 |
+
{
|
| 35981 |
+
"epoch": 5.5e-08,
|
| 35982 |
+
"grad_norm": 0.466796875,
|
| 35983 |
+
"learning_rate": 0.0001015362126248074,
|
| 35984 |
+
"loss": 3.0786,
|
| 35985 |
+
"step": 50110
|
| 35986 |
+
},
|
| 35987 |
+
{
|
| 35988 |
+
"epoch": 6e-08,
|
| 35989 |
+
"grad_norm": 0.474609375,
|
| 35990 |
+
"learning_rate": 0.0001014436298270506,
|
| 35991 |
+
"loss": 3.0386,
|
| 35992 |
+
"step": 50120
|
| 35993 |
+
},
|
| 35994 |
+
{
|
| 35995 |
+
"epoch": 6.5e-08,
|
| 35996 |
+
"grad_norm": 0.4921875,
|
| 35997 |
+
"learning_rate": 0.00010135108541701434,
|
| 35998 |
+
"loss": 3.1031,
|
| 35999 |
+
"step": 50130
|
| 36000 |
+
},
|
| 36001 |
+
{
|
| 36002 |
+
"epoch": 7e-08,
|
| 36003 |
+
"grad_norm": 0.4765625,
|
| 36004 |
+
"learning_rate": 0.00010125857945059509,
|
| 36005 |
+
"loss": 3.0772,
|
| 36006 |
+
"step": 50140
|
| 36007 |
+
},
|
| 36008 |
+
{
|
| 36009 |
+
"epoch": 7.5e-08,
|
| 36010 |
+
"grad_norm": 0.44921875,
|
| 36011 |
+
"learning_rate": 0.00010116611198366592,
|
| 36012 |
+
"loss": 3.0655,
|
| 36013 |
+
"step": 50150
|
| 36014 |
+
},
|
| 36015 |
+
{
|
| 36016 |
+
"epoch": 8e-08,
|
| 36017 |
+
"grad_norm": 0.52734375,
|
| 36018 |
+
"learning_rate": 0.00010107368307207674,
|
| 36019 |
+
"loss": 3.0504,
|
| 36020 |
+
"step": 50160
|
| 36021 |
+
},
|
| 36022 |
+
{
|
| 36023 |
+
"epoch": 8.5e-08,
|
| 36024 |
+
"grad_norm": 0.482421875,
|
| 36025 |
+
"learning_rate": 0.00010098129277165412,
|
| 36026 |
+
"loss": 3.0748,
|
| 36027 |
+
"step": 50170
|
| 36028 |
+
},
|
| 36029 |
+
{
|
| 36030 |
+
"epoch": 9e-08,
|
| 36031 |
+
"grad_norm": 0.443359375,
|
| 36032 |
+
"learning_rate": 0.0001008889411382013,
|
| 36033 |
+
"loss": 3.0916,
|
| 36034 |
+
"step": 50180
|
| 36035 |
+
},
|
| 36036 |
+
{
|
| 36037 |
+
"epoch": 9.5e-08,
|
| 36038 |
+
"grad_norm": 0.498046875,
|
| 36039 |
+
"learning_rate": 0.00010079662822749818,
|
| 36040 |
+
"loss": 3.0565,
|
| 36041 |
+
"step": 50190
|
| 36042 |
+
},
|
| 36043 |
+
{
|
| 36044 |
+
"epoch": 1e-07,
|
| 36045 |
+
"grad_norm": 0.470703125,
|
| 36046 |
+
"learning_rate": 0.00010070435409530144,
|
| 36047 |
+
"loss": 3.0814,
|
| 36048 |
+
"step": 50200
|
| 36049 |
+
},
|
| 36050 |
+
{
|
| 36051 |
+
"epoch": 1.05e-07,
|
| 36052 |
+
"grad_norm": 0.478515625,
|
| 36053 |
+
"learning_rate": 0.00010061211879734392,
|
| 36054 |
+
"loss": 3.0656,
|
| 36055 |
+
"step": 50210
|
| 36056 |
+
},
|
| 36057 |
+
{
|
| 36058 |
+
"epoch": 1.1e-07,
|
| 36059 |
+
"grad_norm": 0.451171875,
|
| 36060 |
+
"learning_rate": 0.00010051992238933552,
|
| 36061 |
+
"loss": 3.1134,
|
| 36062 |
+
"step": 50220
|
| 36063 |
+
},
|
| 36064 |
+
{
|
| 36065 |
+
"epoch": 1.15e-07,
|
| 36066 |
+
"grad_norm": 0.4921875,
|
| 36067 |
+
"learning_rate": 0.00010042776492696218,
|
| 36068 |
+
"loss": 3.097,
|
| 36069 |
+
"step": 50230
|
| 36070 |
+
},
|
| 36071 |
+
{
|
| 36072 |
+
"epoch": 1.2e-07,
|
| 36073 |
+
"grad_norm": 0.466796875,
|
| 36074 |
+
"learning_rate": 0.00010033564646588669,
|
| 36075 |
+
"loss": 3.0581,
|
| 36076 |
+
"step": 50240
|
| 36077 |
+
},
|
| 36078 |
+
{
|
| 36079 |
+
"epoch": 1.25e-07,
|
| 36080 |
+
"grad_norm": 0.5078125,
|
| 36081 |
+
"learning_rate": 0.00010024356706174806,
|
| 36082 |
+
"loss": 3.0447,
|
| 36083 |
+
"step": 50250
|
| 36084 |
+
},
|
| 36085 |
+
{
|
| 36086 |
+
"epoch": 1.3e-07,
|
| 36087 |
+
"grad_norm": 0.47265625,
|
| 36088 |
+
"learning_rate": 0.00010015152677016179,
|
| 36089 |
+
"loss": 3.0488,
|
| 36090 |
+
"step": 50260
|
| 36091 |
+
},
|
| 36092 |
+
{
|
| 36093 |
+
"epoch": 1.35e-07,
|
| 36094 |
+
"grad_norm": 0.462890625,
|
| 36095 |
+
"learning_rate": 0.00010005952564671973,
|
| 36096 |
+
"loss": 3.076,
|
| 36097 |
+
"step": 50270
|
| 36098 |
+
},
|
| 36099 |
+
{
|
| 36100 |
+
"epoch": 1.4e-07,
|
| 36101 |
+
"grad_norm": 0.52734375,
|
| 36102 |
+
"learning_rate": 9.99675637469901e-05,
|
| 36103 |
+
"loss": 3.0726,
|
| 36104 |
+
"step": 50280
|
| 36105 |
+
},
|
| 36106 |
+
{
|
| 36107 |
+
"epoch": 1.45e-07,
|
| 36108 |
+
"grad_norm": 0.5,
|
| 36109 |
+
"learning_rate": 9.987564112651737e-05,
|
| 36110 |
+
"loss": 3.0732,
|
| 36111 |
+
"step": 50290
|
| 36112 |
+
},
|
| 36113 |
+
{
|
| 36114 |
+
"epoch": 1.5e-07,
|
| 36115 |
+
"grad_norm": 0.482421875,
|
| 36116 |
+
"learning_rate": 9.978375784082248e-05,
|
| 36117 |
+
"loss": 3.0587,
|
| 36118 |
+
"step": 50300
|
| 36119 |
+
},
|
| 36120 |
+
{
|
| 36121 |
+
"epoch": 1.55e-07,
|
| 36122 |
+
"grad_norm": 0.4921875,
|
| 36123 |
+
"learning_rate": 9.969191394540223e-05,
|
| 36124 |
+
"loss": 3.0667,
|
| 36125 |
+
"step": 50310
|
| 36126 |
+
},
|
| 36127 |
+
{
|
| 36128 |
+
"epoch": 1.6e-07,
|
| 36129 |
+
"grad_norm": 0.50390625,
|
| 36130 |
+
"learning_rate": 9.960010949573002e-05,
|
| 36131 |
+
"loss": 3.0913,
|
| 36132 |
+
"step": 50320
|
| 36133 |
+
},
|
| 36134 |
+
{
|
| 36135 |
+
"epoch": 1.65e-07,
|
| 36136 |
+
"grad_norm": 0.482421875,
|
| 36137 |
+
"learning_rate": 9.950834454725524e-05,
|
| 36138 |
+
"loss": 3.0945,
|
| 36139 |
+
"step": 50330
|
| 36140 |
+
},
|
| 36141 |
+
{
|
| 36142 |
+
"epoch": 1.7e-07,
|
| 36143 |
+
"grad_norm": 0.5234375,
|
| 36144 |
+
"learning_rate": 9.941661915540343e-05,
|
| 36145 |
+
"loss": 3.0876,
|
| 36146 |
+
"step": 50340
|
| 36147 |
+
},
|
| 36148 |
+
{
|
| 36149 |
+
"epoch": 1.75e-07,
|
| 36150 |
+
"grad_norm": 0.470703125,
|
| 36151 |
+
"learning_rate": 9.932493337557626e-05,
|
| 36152 |
+
"loss": 3.0706,
|
| 36153 |
+
"step": 50350
|
| 36154 |
+
},
|
| 36155 |
+
{
|
| 36156 |
+
"epoch": 1.8e-07,
|
| 36157 |
+
"grad_norm": 0.46875,
|
| 36158 |
+
"learning_rate": 9.923328726315147e-05,
|
| 36159 |
+
"loss": 3.0699,
|
| 36160 |
+
"step": 50360
|
| 36161 |
+
},
|
| 36162 |
+
{
|
| 36163 |
+
"epoch": 1.85e-07,
|
| 36164 |
+
"grad_norm": 0.484375,
|
| 36165 |
+
"learning_rate": 9.914168087348284e-05,
|
| 36166 |
+
"loss": 3.0659,
|
| 36167 |
+
"step": 50370
|
| 36168 |
+
},
|
| 36169 |
+
{
|
| 36170 |
+
"epoch": 1.9e-07,
|
| 36171 |
+
"grad_norm": 0.482421875,
|
| 36172 |
+
"learning_rate": 9.905011426190015e-05,
|
| 36173 |
+
"loss": 3.1036,
|
| 36174 |
+
"step": 50380
|
| 36175 |
+
},
|
| 36176 |
+
{
|
| 36177 |
+
"epoch": 1.95e-07,
|
| 36178 |
+
"grad_norm": 0.451171875,
|
| 36179 |
+
"learning_rate": 9.895858748370918e-05,
|
| 36180 |
+
"loss": 3.0681,
|
| 36181 |
+
"step": 50390
|
| 36182 |
+
},
|
| 36183 |
+
{
|
| 36184 |
+
"epoch": 2e-07,
|
| 36185 |
+
"grad_norm": 0.5390625,
|
| 36186 |
+
"learning_rate": 9.886710059419155e-05,
|
| 36187 |
+
"loss": 3.0689,
|
| 36188 |
+
"step": 50400
|
| 36189 |
+
},
|
| 36190 |
+
{
|
| 36191 |
+
"epoch": 2.05e-07,
|
| 36192 |
+
"grad_norm": 0.46484375,
|
| 36193 |
+
"learning_rate": 9.877565364860505e-05,
|
| 36194 |
+
"loss": 3.0513,
|
| 36195 |
+
"step": 50410
|
| 36196 |
+
},
|
| 36197 |
+
{
|
| 36198 |
+
"epoch": 2.1e-07,
|
| 36199 |
+
"grad_norm": 0.4921875,
|
| 36200 |
+
"learning_rate": 9.868424670218297e-05,
|
| 36201 |
+
"loss": 3.061,
|
| 36202 |
+
"step": 50420
|
| 36203 |
+
},
|
| 36204 |
+
{
|
| 36205 |
+
"epoch": 2.15e-07,
|
| 36206 |
+
"grad_norm": 0.486328125,
|
| 36207 |
+
"learning_rate": 9.859287981013483e-05,
|
| 36208 |
+
"loss": 3.0643,
|
| 36209 |
+
"step": 50430
|
| 36210 |
+
},
|
| 36211 |
+
{
|
| 36212 |
+
"epoch": 2.2e-07,
|
| 36213 |
+
"grad_norm": 0.70703125,
|
| 36214 |
+
"learning_rate": 9.850155302764554e-05,
|
| 36215 |
+
"loss": 3.0509,
|
| 36216 |
+
"step": 50440
|
| 36217 |
+
},
|
| 36218 |
+
{
|
| 36219 |
+
"epoch": 2.25e-07,
|
| 36220 |
+
"grad_norm": 0.490234375,
|
| 36221 |
+
"learning_rate": 9.841026640987615e-05,
|
| 36222 |
+
"loss": 3.076,
|
| 36223 |
+
"step": 50450
|
| 36224 |
+
},
|
| 36225 |
+
{
|
| 36226 |
+
"epoch": 2.3e-07,
|
| 36227 |
+
"grad_norm": 0.474609375,
|
| 36228 |
+
"learning_rate": 9.83190200119633e-05,
|
| 36229 |
+
"loss": 3.0528,
|
| 36230 |
+
"step": 50460
|
| 36231 |
+
},
|
| 36232 |
+
{
|
| 36233 |
+
"epoch": 2.35e-07,
|
| 36234 |
+
"grad_norm": 0.52734375,
|
| 36235 |
+
"learning_rate": 9.822781388901932e-05,
|
| 36236 |
+
"loss": 3.0495,
|
| 36237 |
+
"step": 50470
|
| 36238 |
+
},
|
| 36239 |
+
{
|
| 36240 |
+
"epoch": 2.4e-07,
|
| 36241 |
+
"grad_norm": 0.52734375,
|
| 36242 |
+
"learning_rate": 9.813664809613225e-05,
|
| 36243 |
+
"loss": 3.0791,
|
| 36244 |
+
"step": 50480
|
| 36245 |
+
},
|
| 36246 |
+
{
|
| 36247 |
+
"epoch": 2.45e-07,
|
| 36248 |
+
"grad_norm": 0.58984375,
|
| 36249 |
+
"learning_rate": 9.804552268836576e-05,
|
| 36250 |
+
"loss": 3.0763,
|
| 36251 |
+
"step": 50490
|
| 36252 |
+
},
|
| 36253 |
+
{
|
| 36254 |
+
"epoch": 2.5e-07,
|
| 36255 |
+
"grad_norm": 0.494140625,
|
| 36256 |
+
"learning_rate": 9.795443772075904e-05,
|
| 36257 |
+
"loss": 3.0677,
|
| 36258 |
+
"step": 50500
|
| 36259 |
+
},
|
| 36260 |
+
{
|
| 36261 |
+
"epoch": 2.5e-07,
|
| 36262 |
+
"eval_loss": 3.129530191421509,
|
| 36263 |
+
"eval_runtime": 0.2149,
|
| 36264 |
+
"eval_samples_per_second": 148.921,
|
| 36265 |
+
"eval_steps_per_second": 4.654,
|
| 36266 |
+
"step": 50500
|
| 36267 |
+
},
|
| 36268 |
+
{
|
| 36269 |
+
"epoch": 2.55e-07,
|
| 36270 |
+
"grad_norm": 0.453125,
|
| 36271 |
+
"learning_rate": 9.786339324832714e-05,
|
| 36272 |
+
"loss": 3.0699,
|
| 36273 |
+
"step": 50510
|
| 36274 |
+
},
|
| 36275 |
+
{
|
| 36276 |
+
"epoch": 2.6e-07,
|
| 36277 |
+
"grad_norm": 0.455078125,
|
| 36278 |
+
"learning_rate": 9.777238932606025e-05,
|
| 36279 |
+
"loss": 3.053,
|
| 36280 |
+
"step": 50520
|
| 36281 |
+
},
|
| 36282 |
+
{
|
| 36283 |
+
"epoch": 2.65e-07,
|
| 36284 |
+
"grad_norm": 0.4765625,
|
| 36285 |
+
"learning_rate": 9.768142600892437e-05,
|
| 36286 |
+
"loss": 3.0643,
|
| 36287 |
+
"step": 50530
|
| 36288 |
+
},
|
| 36289 |
+
{
|
| 36290 |
+
"epoch": 2.7e-07,
|
| 36291 |
+
"grad_norm": 0.515625,
|
| 36292 |
+
"learning_rate": 9.759050335186092e-05,
|
| 36293 |
+
"loss": 3.0705,
|
| 36294 |
+
"step": 50540
|
| 36295 |
+
},
|
| 36296 |
+
{
|
| 36297 |
+
"epoch": 2.75e-07,
|
| 36298 |
+
"grad_norm": 0.462890625,
|
| 36299 |
+
"learning_rate": 9.749962140978659e-05,
|
| 36300 |
+
"loss": 3.0879,
|
| 36301 |
+
"step": 50550
|
| 36302 |
+
},
|
| 36303 |
+
{
|
| 36304 |
+
"epoch": 2.8e-07,
|
| 36305 |
+
"grad_norm": 0.47265625,
|
| 36306 |
+
"learning_rate": 9.74087802375937e-05,
|
| 36307 |
+
"loss": 3.0484,
|
| 36308 |
+
"step": 50560
|
| 36309 |
+
},
|
| 36310 |
+
{
|
| 36311 |
+
"epoch": 2.85e-07,
|
| 36312 |
+
"grad_norm": 0.48046875,
|
| 36313 |
+
"learning_rate": 9.731797989014985e-05,
|
| 36314 |
+
"loss": 3.0952,
|
| 36315 |
+
"step": 50570
|
| 36316 |
+
},
|
| 36317 |
+
{
|
| 36318 |
+
"epoch": 2.9e-07,
|
| 36319 |
+
"grad_norm": 0.53125,
|
| 36320 |
+
"learning_rate": 9.722722042229788e-05,
|
| 36321 |
+
"loss": 3.0392,
|
| 36322 |
+
"step": 50580
|
| 36323 |
+
},
|
| 36324 |
+
{
|
| 36325 |
+
"epoch": 2.95e-07,
|
| 36326 |
+
"grad_norm": 0.52734375,
|
| 36327 |
+
"learning_rate": 9.713650188885624e-05,
|
| 36328 |
+
"loss": 3.0659,
|
| 36329 |
+
"step": 50590
|
| 36330 |
+
},
|
| 36331 |
+
{
|
| 36332 |
+
"epoch": 3e-07,
|
| 36333 |
+
"grad_norm": 0.474609375,
|
| 36334 |
+
"learning_rate": 9.704582434461823e-05,
|
| 36335 |
+
"loss": 3.05,
|
| 36336 |
+
"step": 50600
|
| 36337 |
+
},
|
| 36338 |
+
{
|
| 36339 |
+
"epoch": 3.05e-07,
|
| 36340 |
+
"grad_norm": 0.5390625,
|
| 36341 |
+
"learning_rate": 9.695518784435277e-05,
|
| 36342 |
+
"loss": 3.0352,
|
| 36343 |
+
"step": 50610
|
| 36344 |
+
},
|
| 36345 |
+
{
|
| 36346 |
+
"epoch": 3.1e-07,
|
| 36347 |
+
"grad_norm": 0.5078125,
|
| 36348 |
+
"learning_rate": 9.686459244280379e-05,
|
| 36349 |
+
"loss": 3.0766,
|
| 36350 |
+
"step": 50620
|
| 36351 |
+
},
|
| 36352 |
+
{
|
| 36353 |
+
"epoch": 3.15e-07,
|
| 36354 |
+
"grad_norm": 0.447265625,
|
| 36355 |
+
"learning_rate": 9.677403819469046e-05,
|
| 36356 |
+
"loss": 3.087,
|
| 36357 |
+
"step": 50630
|
| 36358 |
+
},
|
| 36359 |
+
{
|
| 36360 |
+
"epoch": 3.2e-07,
|
| 36361 |
+
"grad_norm": 0.48828125,
|
| 36362 |
+
"learning_rate": 9.668352515470711e-05,
|
| 36363 |
+
"loss": 3.0798,
|
| 36364 |
+
"step": 50640
|
| 36365 |
+
},
|
| 36366 |
+
{
|
| 36367 |
+
"epoch": 3.25e-07,
|
| 36368 |
+
"grad_norm": 0.466796875,
|
| 36369 |
+
"learning_rate": 9.659305337752307e-05,
|
| 36370 |
+
"loss": 3.1194,
|
| 36371 |
+
"step": 50650
|
| 36372 |
+
},
|
| 36373 |
+
{
|
| 36374 |
+
"epoch": 3.3e-07,
|
| 36375 |
+
"grad_norm": 0.48828125,
|
| 36376 |
+
"learning_rate": 9.650262291778282e-05,
|
| 36377 |
+
"loss": 3.0624,
|
| 36378 |
+
"step": 50660
|
| 36379 |
+
},
|
| 36380 |
+
{
|
| 36381 |
+
"epoch": 3.35e-07,
|
| 36382 |
+
"grad_norm": 0.486328125,
|
| 36383 |
+
"learning_rate": 9.641223383010607e-05,
|
| 36384 |
+
"loss": 3.0865,
|
| 36385 |
+
"step": 50670
|
| 36386 |
+
},
|
| 36387 |
+
{
|
| 36388 |
+
"epoch": 3.4e-07,
|
| 36389 |
+
"grad_norm": 0.482421875,
|
| 36390 |
+
"learning_rate": 9.632188616908713e-05,
|
| 36391 |
+
"loss": 3.0837,
|
| 36392 |
+
"step": 50680
|
| 36393 |
+
},
|
| 36394 |
+
{
|
| 36395 |
+
"epoch": 3.45e-07,
|
| 36396 |
+
"grad_norm": 0.52734375,
|
| 36397 |
+
"learning_rate": 9.623157998929572e-05,
|
| 36398 |
+
"loss": 3.0708,
|
| 36399 |
+
"step": 50690
|
| 36400 |
+
},
|
| 36401 |
+
{
|
| 36402 |
+
"epoch": 3.5e-07,
|
| 36403 |
+
"grad_norm": 0.4453125,
|
| 36404 |
+
"learning_rate": 9.614131534527612e-05,
|
| 36405 |
+
"loss": 3.0575,
|
| 36406 |
+
"step": 50700
|
| 36407 |
+
},
|
| 36408 |
+
{
|
| 36409 |
+
"epoch": 3.55e-07,
|
| 36410 |
+
"grad_norm": 0.474609375,
|
| 36411 |
+
"learning_rate": 9.605109229154786e-05,
|
| 36412 |
+
"loss": 3.0563,
|
| 36413 |
+
"step": 50710
|
| 36414 |
+
},
|
| 36415 |
+
{
|
| 36416 |
+
"epoch": 3.6e-07,
|
| 36417 |
+
"grad_norm": 0.5078125,
|
| 36418 |
+
"learning_rate": 9.596091088260518e-05,
|
| 36419 |
+
"loss": 3.043,
|
| 36420 |
+
"step": 50720
|
| 36421 |
+
},
|
| 36422 |
+
{
|
| 36423 |
+
"epoch": 3.65e-07,
|
| 36424 |
+
"grad_norm": 0.484375,
|
| 36425 |
+
"learning_rate": 9.587077117291711e-05,
|
| 36426 |
+
"loss": 3.0689,
|
| 36427 |
+
"step": 50730
|
| 36428 |
+
},
|
| 36429 |
+
{
|
| 36430 |
+
"epoch": 3.7e-07,
|
| 36431 |
+
"grad_norm": 0.54296875,
|
| 36432 |
+
"learning_rate": 9.578067321692766e-05,
|
| 36433 |
+
"loss": 3.0457,
|
| 36434 |
+
"step": 50740
|
| 36435 |
+
},
|
| 36436 |
+
{
|
| 36437 |
+
"epoch": 3.75e-07,
|
| 36438 |
+
"grad_norm": 0.52734375,
|
| 36439 |
+
"learning_rate": 9.569061706905547e-05,
|
| 36440 |
+
"loss": 3.0884,
|
| 36441 |
+
"step": 50750
|
| 36442 |
+
},
|
| 36443 |
+
{
|
| 36444 |
+
"epoch": 3.8e-07,
|
| 36445 |
+
"grad_norm": 0.47265625,
|
| 36446 |
+
"learning_rate": 9.560060278369402e-05,
|
| 36447 |
+
"loss": 3.0726,
|
| 36448 |
+
"step": 50760
|
| 36449 |
+
},
|
| 36450 |
+
{
|
| 36451 |
+
"epoch": 3.85e-07,
|
| 36452 |
+
"grad_norm": 0.4609375,
|
| 36453 |
+
"learning_rate": 9.551063041521143e-05,
|
| 36454 |
+
"loss": 3.0716,
|
| 36455 |
+
"step": 50770
|
| 36456 |
+
},
|
| 36457 |
+
{
|
| 36458 |
+
"epoch": 3.9e-07,
|
| 36459 |
+
"grad_norm": 0.52734375,
|
| 36460 |
+
"learning_rate": 9.542070001795062e-05,
|
| 36461 |
+
"loss": 3.0654,
|
| 36462 |
+
"step": 50780
|
| 36463 |
+
},
|
| 36464 |
+
{
|
| 36465 |
+
"epoch": 3.95e-07,
|
| 36466 |
+
"grad_norm": 0.470703125,
|
| 36467 |
+
"learning_rate": 9.533081164622898e-05,
|
| 36468 |
+
"loss": 3.0563,
|
| 36469 |
+
"step": 50790
|
| 36470 |
+
},
|
| 36471 |
+
{
|
| 36472 |
+
"epoch": 4e-07,
|
| 36473 |
+
"grad_norm": 0.462890625,
|
| 36474 |
+
"learning_rate": 9.524096535433879e-05,
|
| 36475 |
+
"loss": 3.0715,
|
| 36476 |
+
"step": 50800
|
| 36477 |
+
},
|
| 36478 |
+
{
|
| 36479 |
+
"epoch": 4.05e-07,
|
| 36480 |
+
"grad_norm": 0.466796875,
|
| 36481 |
+
"learning_rate": 9.515116119654659e-05,
|
| 36482 |
+
"loss": 3.0468,
|
| 36483 |
+
"step": 50810
|
| 36484 |
+
},
|
| 36485 |
+
{
|
| 36486 |
+
"epoch": 4.1e-07,
|
| 36487 |
+
"grad_norm": 0.5234375,
|
| 36488 |
+
"learning_rate": 9.50613992270937e-05,
|
| 36489 |
+
"loss": 3.0638,
|
| 36490 |
+
"step": 50820
|
| 36491 |
+
},
|
| 36492 |
+
{
|
| 36493 |
+
"epoch": 4.15e-07,
|
| 36494 |
+
"grad_norm": 0.484375,
|
| 36495 |
+
"learning_rate": 9.49716795001959e-05,
|
| 36496 |
+
"loss": 3.0668,
|
| 36497 |
+
"step": 50830
|
| 36498 |
+
},
|
| 36499 |
+
{
|
| 36500 |
+
"epoch": 4.2e-07,
|
| 36501 |
+
"grad_norm": 0.453125,
|
| 36502 |
+
"learning_rate": 9.488200207004341e-05,
|
| 36503 |
+
"loss": 3.0756,
|
| 36504 |
+
"step": 50840
|
| 36505 |
+
},
|
| 36506 |
+
{
|
| 36507 |
+
"epoch": 4.25e-07,
|
| 36508 |
+
"grad_norm": 0.478515625,
|
| 36509 |
+
"learning_rate": 9.479236699080096e-05,
|
| 36510 |
+
"loss": 3.0522,
|
| 36511 |
+
"step": 50850
|
| 36512 |
+
},
|
| 36513 |
+
{
|
| 36514 |
+
"epoch": 4.3e-07,
|
| 36515 |
+
"grad_norm": 0.52734375,
|
| 36516 |
+
"learning_rate": 9.470277431660766e-05,
|
| 36517 |
+
"loss": 3.0698,
|
| 36518 |
+
"step": 50860
|
| 36519 |
+
},
|
| 36520 |
+
{
|
| 36521 |
+
"epoch": 4.35e-07,
|
| 36522 |
+
"grad_norm": 0.45703125,
|
| 36523 |
+
"learning_rate": 9.4613224101577e-05,
|
| 36524 |
+
"loss": 3.0707,
|
| 36525 |
+
"step": 50870
|
| 36526 |
+
},
|
| 36527 |
+
{
|
| 36528 |
+
"epoch": 4.4e-07,
|
| 36529 |
+
"grad_norm": 0.439453125,
|
| 36530 |
+
"learning_rate": 9.452371639979695e-05,
|
| 36531 |
+
"loss": 3.0788,
|
| 36532 |
+
"step": 50880
|
| 36533 |
+
},
|
| 36534 |
+
{
|
| 36535 |
+
"epoch": 4.45e-07,
|
| 36536 |
+
"grad_norm": 0.490234375,
|
| 36537 |
+
"learning_rate": 9.443425126532956e-05,
|
| 36538 |
+
"loss": 3.0516,
|
| 36539 |
+
"step": 50890
|
| 36540 |
+
},
|
| 36541 |
+
{
|
| 36542 |
+
"epoch": 4.5e-07,
|
| 36543 |
+
"grad_norm": 0.486328125,
|
| 36544 |
+
"learning_rate": 9.434482875221145e-05,
|
| 36545 |
+
"loss": 3.075,
|
| 36546 |
+
"step": 50900
|
| 36547 |
+
},
|
| 36548 |
+
{
|
| 36549 |
+
"epoch": 4.55e-07,
|
| 36550 |
+
"grad_norm": 0.494140625,
|
| 36551 |
+
"learning_rate": 9.42554489144532e-05,
|
| 36552 |
+
"loss": 3.0823,
|
| 36553 |
+
"step": 50910
|
| 36554 |
+
},
|
| 36555 |
+
{
|
| 36556 |
+
"epoch": 4.6e-07,
|
| 36557 |
+
"grad_norm": 0.4609375,
|
| 36558 |
+
"learning_rate": 9.416611180603991e-05,
|
| 36559 |
+
"loss": 3.0312,
|
| 36560 |
+
"step": 50920
|
| 36561 |
+
},
|
| 36562 |
+
{
|
| 36563 |
+
"epoch": 4.65e-07,
|
| 36564 |
+
"grad_norm": 0.486328125,
|
| 36565 |
+
"learning_rate": 9.407681748093071e-05,
|
| 36566 |
+
"loss": 3.0439,
|
| 36567 |
+
"step": 50930
|
| 36568 |
+
},
|
| 36569 |
+
{
|
| 36570 |
+
"epoch": 4.7e-07,
|
| 36571 |
+
"grad_norm": 0.5234375,
|
| 36572 |
+
"learning_rate": 9.398756599305886e-05,
|
| 36573 |
+
"loss": 3.0656,
|
| 36574 |
+
"step": 50940
|
| 36575 |
+
},
|
| 36576 |
+
{
|
| 36577 |
+
"epoch": 4.75e-07,
|
| 36578 |
+
"grad_norm": 0.486328125,
|
| 36579 |
+
"learning_rate": 9.389835739633186e-05,
|
| 36580 |
+
"loss": 3.0561,
|
| 36581 |
+
"step": 50950
|
| 36582 |
+
},
|
| 36583 |
+
{
|
| 36584 |
+
"epoch": 4.8e-07,
|
| 36585 |
+
"grad_norm": 0.51171875,
|
| 36586 |
+
"learning_rate": 9.380919174463118e-05,
|
| 36587 |
+
"loss": 3.0647,
|
| 36588 |
+
"step": 50960
|
| 36589 |
+
},
|
| 36590 |
+
{
|
| 36591 |
+
"epoch": 4.85e-07,
|
| 36592 |
+
"grad_norm": 0.5625,
|
| 36593 |
+
"learning_rate": 9.37200690918124e-05,
|
| 36594 |
+
"loss": 3.0431,
|
| 36595 |
+
"step": 50970
|
| 36596 |
+
},
|
| 36597 |
+
{
|
| 36598 |
+
"epoch": 4.9e-07,
|
| 36599 |
+
"grad_norm": 0.50390625,
|
| 36600 |
+
"learning_rate": 9.363098949170531e-05,
|
| 36601 |
+
"loss": 3.0374,
|
| 36602 |
+
"step": 50980
|
| 36603 |
+
},
|
| 36604 |
+
{
|
| 36605 |
+
"epoch": 4.95e-07,
|
| 36606 |
+
"grad_norm": 0.498046875,
|
| 36607 |
+
"learning_rate": 9.354195299811329e-05,
|
| 36608 |
+
"loss": 3.0627,
|
| 36609 |
+
"step": 50990
|
| 36610 |
+
},
|
| 36611 |
+
{
|
| 36612 |
+
"epoch": 5e-07,
|
| 36613 |
+
"grad_norm": 0.484375,
|
| 36614 |
+
"learning_rate": 9.345295966481413e-05,
|
| 36615 |
+
"loss": 3.0979,
|
| 36616 |
+
"step": 51000
|
| 36617 |
+
},
|
| 36618 |
+
{
|
| 36619 |
+
"epoch": 5e-07,
|
| 36620 |
+
"eval_loss": 3.1213490962982178,
|
| 36621 |
+
"eval_runtime": 0.0603,
|
| 36622 |
+
"eval_samples_per_second": 530.996,
|
| 36623 |
+
"eval_steps_per_second": 16.594,
|
| 36624 |
+
"step": 51000
|
| 36625 |
+
},
|
| 36626 |
+
{
|
| 36627 |
+
"epoch": 5.05e-07,
|
| 36628 |
+
"grad_norm": 0.5,
|
| 36629 |
+
"learning_rate": 9.336400954555923e-05,
|
| 36630 |
+
"loss": 3.0963,
|
| 36631 |
+
"step": 51010
|
| 36632 |
+
},
|
| 36633 |
+
{
|
| 36634 |
+
"epoch": 5.1e-07,
|
| 36635 |
+
"grad_norm": 0.50390625,
|
| 36636 |
+
"learning_rate": 9.327510269407403e-05,
|
| 36637 |
+
"loss": 3.0656,
|
| 36638 |
+
"step": 51020
|
| 36639 |
+
},
|
| 36640 |
+
{
|
| 36641 |
+
"epoch": 5.15e-07,
|
| 36642 |
+
"grad_norm": 0.453125,
|
| 36643 |
+
"learning_rate": 9.318623916405782e-05,
|
| 36644 |
+
"loss": 3.0855,
|
| 36645 |
+
"step": 51030
|
| 36646 |
+
},
|
| 36647 |
+
{
|
| 36648 |
+
"epoch": 5.2e-07,
|
| 36649 |
+
"grad_norm": 0.4765625,
|
| 36650 |
+
"learning_rate": 9.309741900918372e-05,
|
| 36651 |
+
"loss": 3.0654,
|
| 36652 |
+
"step": 51040
|
| 36653 |
+
},
|
| 36654 |
+
{
|
| 36655 |
+
"epoch": 5.25e-07,
|
| 36656 |
+
"grad_norm": 0.4921875,
|
| 36657 |
+
"learning_rate": 9.30086422830986e-05,
|
| 36658 |
+
"loss": 3.0488,
|
| 36659 |
+
"step": 51050
|
| 36660 |
+
},
|
| 36661 |
+
{
|
| 36662 |
+
"epoch": 5.3e-07,
|
| 36663 |
+
"grad_norm": 0.50390625,
|
| 36664 |
+
"learning_rate": 9.291990903942324e-05,
|
| 36665 |
+
"loss": 3.0746,
|
| 36666 |
+
"step": 51060
|
| 36667 |
+
},
|
| 36668 |
+
{
|
| 36669 |
+
"epoch": 5.35e-07,
|
| 36670 |
+
"grad_norm": 0.50390625,
|
| 36671 |
+
"learning_rate": 9.283121933175194e-05,
|
| 36672 |
+
"loss": 3.0698,
|
| 36673 |
+
"step": 51070
|
| 36674 |
+
},
|
| 36675 |
+
{
|
| 36676 |
+
"epoch": 5.4e-07,
|
| 36677 |
+
"grad_norm": 0.498046875,
|
| 36678 |
+
"learning_rate": 9.274257321365295e-05,
|
| 36679 |
+
"loss": 3.0538,
|
| 36680 |
+
"step": 51080
|
| 36681 |
+
},
|
| 36682 |
+
{
|
| 36683 |
+
"epoch": 5.45e-07,
|
| 36684 |
+
"grad_norm": 0.466796875,
|
| 36685 |
+
"learning_rate": 9.265397073866799e-05,
|
| 36686 |
+
"loss": 3.0553,
|
| 36687 |
+
"step": 51090
|
| 36688 |
+
},
|
| 36689 |
+
{
|
| 36690 |
+
"epoch": 5.5e-07,
|
| 36691 |
+
"grad_norm": 2.03125,
|
| 36692 |
+
"learning_rate": 9.256541196031252e-05,
|
| 36693 |
+
"loss": 3.0775,
|
| 36694 |
+
"step": 51100
|
| 36695 |
+
},
|
| 36696 |
+
{
|
| 36697 |
+
"epoch": 5.55e-07,
|
| 36698 |
+
"grad_norm": 0.5703125,
|
| 36699 |
+
"learning_rate": 9.247689693207562e-05,
|
| 36700 |
+
"loss": 3.0664,
|
| 36701 |
+
"step": 51110
|
| 36702 |
+
},
|
| 36703 |
+
{
|
| 36704 |
+
"epoch": 5.6e-07,
|
| 36705 |
+
"grad_norm": 0.490234375,
|
| 36706 |
+
"learning_rate": 9.238842570741985e-05,
|
| 36707 |
+
"loss": 3.0337,
|
| 36708 |
+
"step": 51120
|
| 36709 |
+
},
|
| 36710 |
+
{
|
| 36711 |
+
"epoch": 5.65e-07,
|
| 36712 |
+
"grad_norm": 0.498046875,
|
| 36713 |
+
"learning_rate": 9.22999983397814e-05,
|
| 36714 |
+
"loss": 3.0708,
|
| 36715 |
+
"step": 51130
|
| 36716 |
+
},
|
| 36717 |
+
{
|
| 36718 |
+
"epoch": 5.7e-07,
|
| 36719 |
+
"grad_norm": 0.546875,
|
| 36720 |
+
"learning_rate": 9.221161488256997e-05,
|
| 36721 |
+
"loss": 3.0573,
|
| 36722 |
+
"step": 51140
|
| 36723 |
+
},
|
| 36724 |
+
{
|
| 36725 |
+
"epoch": 5.75e-07,
|
| 36726 |
+
"grad_norm": 0.47265625,
|
| 36727 |
+
"learning_rate": 9.212327538916862e-05,
|
| 36728 |
+
"loss": 3.0555,
|
| 36729 |
+
"step": 51150
|
| 36730 |
+
},
|
| 36731 |
+
{
|
| 36732 |
+
"epoch": 5.8e-07,
|
| 36733 |
+
"grad_norm": 0.453125,
|
| 36734 |
+
"learning_rate": 9.203497991293414e-05,
|
| 36735 |
+
"loss": 3.0541,
|
| 36736 |
+
"step": 51160
|
| 36737 |
+
},
|
| 36738 |
+
{
|
| 36739 |
+
"epoch": 5.85e-07,
|
| 36740 |
+
"grad_norm": 0.482421875,
|
| 36741 |
+
"learning_rate": 9.194672850719638e-05,
|
| 36742 |
+
"loss": 3.0425,
|
| 36743 |
+
"step": 51170
|
| 36744 |
+
},
|
| 36745 |
+
{
|
| 36746 |
+
"epoch": 5.9e-07,
|
| 36747 |
+
"grad_norm": 0.5078125,
|
| 36748 |
+
"learning_rate": 9.185852122525873e-05,
|
| 36749 |
+
"loss": 3.0713,
|
| 36750 |
+
"step": 51180
|
| 36751 |
+
},
|
| 36752 |
+
{
|
| 36753 |
+
"epoch": 5.95e-07,
|
| 36754 |
+
"grad_norm": 0.5,
|
| 36755 |
+
"learning_rate": 9.177035812039808e-05,
|
| 36756 |
+
"loss": 3.0829,
|
| 36757 |
+
"step": 51190
|
| 36758 |
+
},
|
| 36759 |
+
{
|
| 36760 |
+
"epoch": 6e-07,
|
| 36761 |
+
"grad_norm": 0.45703125,
|
| 36762 |
+
"learning_rate": 9.168223924586428e-05,
|
| 36763 |
+
"loss": 3.0628,
|
| 36764 |
+
"step": 51200
|
| 36765 |
+
},
|
| 36766 |
+
{
|
| 36767 |
+
"epoch": 6.05e-07,
|
| 36768 |
+
"grad_norm": 0.53515625,
|
| 36769 |
+
"learning_rate": 9.159416465488085e-05,
|
| 36770 |
+
"loss": 3.0838,
|
| 36771 |
+
"step": 51210
|
| 36772 |
+
},
|
| 36773 |
+
{
|
| 36774 |
+
"epoch": 6.1e-07,
|
| 36775 |
+
"grad_norm": 0.48046875,
|
| 36776 |
+
"learning_rate": 9.150613440064436e-05,
|
| 36777 |
+
"loss": 3.0512,
|
| 36778 |
+
"step": 51220
|
| 36779 |
+
},
|
| 36780 |
+
{
|
| 36781 |
+
"epoch": 6.15e-07,
|
| 36782 |
+
"grad_norm": 0.53125,
|
| 36783 |
+
"learning_rate": 9.141814853632458e-05,
|
| 36784 |
+
"loss": 3.0805,
|
| 36785 |
+
"step": 51230
|
| 36786 |
+
},
|
| 36787 |
+
{
|
| 36788 |
+
"epoch": 6.2e-07,
|
| 36789 |
+
"grad_norm": 0.4765625,
|
| 36790 |
+
"learning_rate": 9.133020711506455e-05,
|
| 36791 |
+
"loss": 3.036,
|
| 36792 |
+
"step": 51240
|
| 36793 |
+
},
|
| 36794 |
+
{
|
| 36795 |
+
"epoch": 6.25e-07,
|
| 36796 |
+
"grad_norm": 0.439453125,
|
| 36797 |
+
"learning_rate": 9.124231018998044e-05,
|
| 36798 |
+
"loss": 3.0639,
|
| 36799 |
+
"step": 51250
|
| 36800 |
+
},
|
| 36801 |
+
{
|
| 36802 |
+
"epoch": 6.3e-07,
|
| 36803 |
+
"grad_norm": 0.453125,
|
| 36804 |
+
"learning_rate": 9.115445781416147e-05,
|
| 36805 |
+
"loss": 3.0884,
|
| 36806 |
+
"step": 51260
|
| 36807 |
+
},
|
| 36808 |
+
{
|
| 36809 |
+
"epoch": 6.35e-07,
|
| 36810 |
+
"grad_norm": 0.52734375,
|
| 36811 |
+
"learning_rate": 9.106665004067018e-05,
|
| 36812 |
+
"loss": 3.0888,
|
| 36813 |
+
"step": 51270
|
| 36814 |
+
},
|
| 36815 |
+
{
|
| 36816 |
+
"epoch": 6.4e-07,
|
| 36817 |
+
"grad_norm": 0.515625,
|
| 36818 |
+
"learning_rate": 9.097888692254183e-05,
|
| 36819 |
+
"loss": 3.0709,
|
| 36820 |
+
"step": 51280
|
| 36821 |
+
},
|
| 36822 |
+
{
|
| 36823 |
+
"epoch": 6.45e-07,
|
| 36824 |
+
"grad_norm": 0.49609375,
|
| 36825 |
+
"learning_rate": 9.0891168512785e-05,
|
| 36826 |
+
"loss": 3.062,
|
| 36827 |
+
"step": 51290
|
| 36828 |
+
},
|
| 36829 |
+
{
|
| 36830 |
+
"epoch": 6.5e-07,
|
| 36831 |
+
"grad_norm": 0.462890625,
|
| 36832 |
+
"learning_rate": 9.080349486438118e-05,
|
| 36833 |
+
"loss": 3.0852,
|
| 36834 |
+
"step": 51300
|
| 36835 |
+
},
|
| 36836 |
+
{
|
| 36837 |
+
"epoch": 6.55e-07,
|
| 36838 |
+
"grad_norm": 0.482421875,
|
| 36839 |
+
"learning_rate": 9.071586603028473e-05,
|
| 36840 |
+
"loss": 3.0751,
|
| 36841 |
+
"step": 51310
|
| 36842 |
+
},
|
| 36843 |
+
{
|
| 36844 |
+
"epoch": 6.6e-07,
|
| 36845 |
+
"grad_norm": 0.466796875,
|
| 36846 |
+
"learning_rate": 9.062828206342305e-05,
|
| 36847 |
+
"loss": 3.0617,
|
| 36848 |
+
"step": 51320
|
| 36849 |
+
},
|
| 36850 |
+
{
|
| 36851 |
+
"epoch": 6.65e-07,
|
| 36852 |
+
"grad_norm": 0.498046875,
|
| 36853 |
+
"learning_rate": 9.054074301669638e-05,
|
| 36854 |
+
"loss": 3.0574,
|
| 36855 |
+
"step": 51330
|
| 36856 |
+
},
|
| 36857 |
+
{
|
| 36858 |
+
"epoch": 6.7e-07,
|
| 36859 |
+
"grad_norm": 0.48046875,
|
| 36860 |
+
"learning_rate": 9.045324894297791e-05,
|
| 36861 |
+
"loss": 3.0712,
|
| 36862 |
+
"step": 51340
|
| 36863 |
+
},
|
| 36864 |
+
{
|
| 36865 |
+
"epoch": 6.75e-07,
|
| 36866 |
+
"grad_norm": 0.4921875,
|
| 36867 |
+
"learning_rate": 9.036579989511356e-05,
|
| 36868 |
+
"loss": 3.0725,
|
| 36869 |
+
"step": 51350
|
| 36870 |
+
},
|
| 36871 |
+
{
|
| 36872 |
+
"epoch": 6.8e-07,
|
| 36873 |
+
"grad_norm": 0.51171875,
|
| 36874 |
+
"learning_rate": 9.027839592592207e-05,
|
| 36875 |
+
"loss": 3.072,
|
| 36876 |
+
"step": 51360
|
| 36877 |
+
},
|
| 36878 |
+
{
|
| 36879 |
+
"epoch": 6.85e-07,
|
| 36880 |
+
"grad_norm": 0.484375,
|
| 36881 |
+
"learning_rate": 9.019103708819514e-05,
|
| 36882 |
+
"loss": 3.0584,
|
| 36883 |
+
"step": 51370
|
| 36884 |
+
},
|
| 36885 |
+
{
|
| 36886 |
+
"epoch": 6.9e-07,
|
| 36887 |
+
"grad_norm": 0.484375,
|
| 36888 |
+
"learning_rate": 9.010372343469686e-05,
|
| 36889 |
+
"loss": 3.0186,
|
| 36890 |
+
"step": 51380
|
| 36891 |
+
},
|
| 36892 |
+
{
|
| 36893 |
+
"epoch": 6.95e-07,
|
| 36894 |
+
"grad_norm": 0.478515625,
|
| 36895 |
+
"learning_rate": 9.001645501816436e-05,
|
| 36896 |
+
"loss": 3.0659,
|
| 36897 |
+
"step": 51390
|
| 36898 |
+
},
|
| 36899 |
+
{
|
| 36900 |
+
"epoch": 7e-07,
|
| 36901 |
+
"grad_norm": 0.5390625,
|
| 36902 |
+
"learning_rate": 8.992923189130728e-05,
|
| 36903 |
+
"loss": 3.04,
|
| 36904 |
+
"step": 51400
|
| 36905 |
+
},
|
| 36906 |
+
{
|
| 36907 |
+
"epoch": 7.05e-07,
|
| 36908 |
+
"grad_norm": 0.462890625,
|
| 36909 |
+
"learning_rate": 8.984205410680794e-05,
|
| 36910 |
+
"loss": 3.0649,
|
| 36911 |
+
"step": 51410
|
| 36912 |
+
},
|
| 36913 |
+
{
|
| 36914 |
+
"epoch": 7.1e-07,
|
| 36915 |
+
"grad_norm": 0.48828125,
|
| 36916 |
+
"learning_rate": 8.975492171732129e-05,
|
| 36917 |
+
"loss": 3.034,
|
| 36918 |
+
"step": 51420
|
| 36919 |
+
},
|
| 36920 |
+
{
|
| 36921 |
+
"epoch": 7.15e-07,
|
| 36922 |
+
"grad_norm": 0.5,
|
| 36923 |
+
"learning_rate": 8.96678347754748e-05,
|
| 36924 |
+
"loss": 3.0646,
|
| 36925 |
+
"step": 51430
|
| 36926 |
+
},
|
| 36927 |
+
{
|
| 36928 |
+
"epoch": 7.2e-07,
|
| 36929 |
+
"grad_norm": 0.482421875,
|
| 36930 |
+
"learning_rate": 8.958079333386852e-05,
|
| 36931 |
+
"loss": 3.0774,
|
| 36932 |
+
"step": 51440
|
| 36933 |
+
},
|
| 36934 |
+
{
|
| 36935 |
+
"epoch": 7.25e-07,
|
| 36936 |
+
"grad_norm": 0.515625,
|
| 36937 |
+
"learning_rate": 8.949379744507516e-05,
|
| 36938 |
+
"loss": 3.0537,
|
| 36939 |
+
"step": 51450
|
| 36940 |
+
},
|
| 36941 |
+
{
|
| 36942 |
+
"epoch": 7.3e-07,
|
| 36943 |
+
"grad_norm": 0.45703125,
|
| 36944 |
+
"learning_rate": 8.940684716163963e-05,
|
| 36945 |
+
"loss": 3.0421,
|
| 36946 |
+
"step": 51460
|
| 36947 |
+
},
|
| 36948 |
+
{
|
| 36949 |
+
"epoch": 7.35e-07,
|
| 36950 |
+
"grad_norm": 0.5078125,
|
| 36951 |
+
"learning_rate": 8.931994253607957e-05,
|
| 36952 |
+
"loss": 3.054,
|
| 36953 |
+
"step": 51470
|
| 36954 |
+
},
|
| 36955 |
+
{
|
| 36956 |
+
"epoch": 7.4e-07,
|
| 36957 |
+
"grad_norm": 0.45703125,
|
| 36958 |
+
"learning_rate": 8.923308362088488e-05,
|
| 36959 |
+
"loss": 3.0506,
|
| 36960 |
+
"step": 51480
|
| 36961 |
+
},
|
| 36962 |
+
{
|
| 36963 |
+
"epoch": 7.45e-07,
|
| 36964 |
+
"grad_norm": 0.5,
|
| 36965 |
+
"learning_rate": 8.914627046851789e-05,
|
| 36966 |
+
"loss": 3.0629,
|
| 36967 |
+
"step": 51490
|
| 36968 |
+
},
|
| 36969 |
+
{
|
| 36970 |
+
"epoch": 7.5e-07,
|
| 36971 |
+
"grad_norm": 0.515625,
|
| 36972 |
+
"learning_rate": 8.905950313141332e-05,
|
| 36973 |
+
"loss": 3.0423,
|
| 36974 |
+
"step": 51500
|
| 36975 |
+
},
|
| 36976 |
+
{
|
| 36977 |
+
"epoch": 7.5e-07,
|
| 36978 |
+
"eval_loss": 3.111537218093872,
|
| 36979 |
+
"eval_runtime": 0.063,
|
| 36980 |
+
"eval_samples_per_second": 507.937,
|
| 36981 |
+
"eval_steps_per_second": 15.873,
|
| 36982 |
+
"step": 51500
|
| 36983 |
+
},
|
| 36984 |
+
{
|
| 36985 |
+
"epoch": 7.55e-07,
|
| 36986 |
+
"grad_norm": 0.515625,
|
| 36987 |
+
"learning_rate": 8.897278166197817e-05,
|
| 36988 |
+
"loss": 3.0612,
|
| 36989 |
+
"step": 51510
|
| 36990 |
+
},
|
| 36991 |
+
{
|
| 36992 |
+
"epoch": 7.6e-07,
|
| 36993 |
+
"grad_norm": 0.4765625,
|
| 36994 |
+
"learning_rate": 8.888610611259177e-05,
|
| 36995 |
+
"loss": 3.0569,
|
| 36996 |
+
"step": 51520
|
| 36997 |
+
},
|
| 36998 |
+
{
|
| 36999 |
+
"epoch": 7.65e-07,
|
| 37000 |
+
"grad_norm": 0.515625,
|
| 37001 |
+
"learning_rate": 8.879947653560578e-05,
|
| 37002 |
+
"loss": 3.0693,
|
| 37003 |
+
"step": 51530
|
| 37004 |
+
},
|
| 37005 |
+
{
|
| 37006 |
+
"epoch": 7.7e-07,
|
| 37007 |
+
"grad_norm": 0.5078125,
|
| 37008 |
+
"learning_rate": 8.871289298334383e-05,
|
| 37009 |
+
"loss": 3.0439,
|
| 37010 |
+
"step": 51540
|
| 37011 |
+
},
|
| 37012 |
+
{
|
| 37013 |
+
"epoch": 7.75e-07,
|
| 37014 |
+
"grad_norm": 0.474609375,
|
| 37015 |
+
"learning_rate": 8.862635550810219e-05,
|
| 37016 |
+
"loss": 3.0643,
|
| 37017 |
+
"step": 51550
|
| 37018 |
+
},
|
| 37019 |
+
{
|
| 37020 |
+
"epoch": 7.8e-07,
|
| 37021 |
+
"grad_norm": 0.48828125,
|
| 37022 |
+
"learning_rate": 8.853986416214884e-05,
|
| 37023 |
+
"loss": 3.0453,
|
| 37024 |
+
"step": 51560
|
| 37025 |
+
},
|
| 37026 |
+
{
|
| 37027 |
+
"epoch": 7.85e-07,
|
| 37028 |
+
"grad_norm": 0.515625,
|
| 37029 |
+
"learning_rate": 8.84534189977242e-05,
|
| 37030 |
+
"loss": 3.0771,
|
| 37031 |
+
"step": 51570
|
| 37032 |
+
},
|
| 37033 |
+
{
|
| 37034 |
+
"epoch": 7.9e-07,
|
| 37035 |
+
"grad_norm": 0.51953125,
|
| 37036 |
+
"learning_rate": 8.836702006704075e-05,
|
| 37037 |
+
"loss": 3.0559,
|
| 37038 |
+
"step": 51580
|
| 37039 |
+
},
|
| 37040 |
+
{
|
| 37041 |
+
"epoch": 7.95e-07,
|
| 37042 |
+
"grad_norm": 0.482421875,
|
| 37043 |
+
"learning_rate": 8.828066742228283e-05,
|
| 37044 |
+
"loss": 3.064,
|
| 37045 |
+
"step": 51590
|
| 37046 |
+
},
|
| 37047 |
+
{
|
| 37048 |
+
"epoch": 8e-07,
|
| 37049 |
+
"grad_norm": 0.458984375,
|
| 37050 |
+
"learning_rate": 8.819436111560719e-05,
|
| 37051 |
+
"loss": 3.0784,
|
| 37052 |
+
"step": 51600
|
| 37053 |
+
},
|
| 37054 |
+
{
|
| 37055 |
+
"epoch": 8.05e-07,
|
| 37056 |
+
"grad_norm": 0.50390625,
|
| 37057 |
+
"learning_rate": 8.81081011991423e-05,
|
| 37058 |
+
"loss": 3.067,
|
| 37059 |
+
"step": 51610
|
| 37060 |
+
},
|
| 37061 |
+
{
|
| 37062 |
+
"epoch": 8.1e-07,
|
| 37063 |
+
"grad_norm": 0.470703125,
|
| 37064 |
+
"learning_rate": 8.802188772498871e-05,
|
| 37065 |
+
"loss": 3.0762,
|
| 37066 |
+
"step": 51620
|
| 37067 |
+
},
|
| 37068 |
+
{
|
| 37069 |
+
"epoch": 8.15e-07,
|
| 37070 |
+
"grad_norm": 0.55078125,
|
| 37071 |
+
"learning_rate": 8.793572074521892e-05,
|
| 37072 |
+
"loss": 3.096,
|
| 37073 |
+
"step": 51630
|
| 37074 |
+
},
|
| 37075 |
+
{
|
| 37076 |
+
"epoch": 8.2e-07,
|
| 37077 |
+
"grad_norm": 0.482421875,
|
| 37078 |
+
"learning_rate": 8.784960031187736e-05,
|
| 37079 |
+
"loss": 3.0663,
|
| 37080 |
+
"step": 51640
|
| 37081 |
+
},
|
| 37082 |
+
{
|
| 37083 |
+
"epoch": 8.25e-07,
|
| 37084 |
+
"grad_norm": 0.46875,
|
| 37085 |
+
"learning_rate": 8.776352647698025e-05,
|
| 37086 |
+
"loss": 3.0662,
|
| 37087 |
+
"step": 51650
|
| 37088 |
+
},
|
| 37089 |
+
{
|
| 37090 |
+
"epoch": 8.3e-07,
|
| 37091 |
+
"grad_norm": 0.51171875,
|
| 37092 |
+
"learning_rate": 8.767749929251591e-05,
|
| 37093 |
+
"loss": 3.0433,
|
| 37094 |
+
"step": 51660
|
| 37095 |
+
},
|
| 37096 |
+
{
|
| 37097 |
+
"epoch": 8.35e-07,
|
| 37098 |
+
"grad_norm": 0.53515625,
|
| 37099 |
+
"learning_rate": 8.75915188104441e-05,
|
| 37100 |
+
"loss": 3.0467,
|
| 37101 |
+
"step": 51670
|
| 37102 |
+
},
|
| 37103 |
+
{
|
| 37104 |
+
"epoch": 8.4e-07,
|
| 37105 |
+
"grad_norm": 0.48046875,
|
| 37106 |
+
"learning_rate": 8.750558508269677e-05,
|
| 37107 |
+
"loss": 3.0472,
|
| 37108 |
+
"step": 51680
|
| 37109 |
+
},
|
| 37110 |
+
{
|
| 37111 |
+
"epoch": 8.45e-07,
|
| 37112 |
+
"grad_norm": 0.5,
|
| 37113 |
+
"learning_rate": 8.741969816117739e-05,
|
| 37114 |
+
"loss": 3.0681,
|
| 37115 |
+
"step": 51690
|
| 37116 |
+
},
|
| 37117 |
+
{
|
| 37118 |
+
"epoch": 8.5e-07,
|
| 37119 |
+
"grad_norm": 0.46484375,
|
| 37120 |
+
"learning_rate": 8.73338580977612e-05,
|
| 37121 |
+
"loss": 3.0703,
|
| 37122 |
+
"step": 51700
|
| 37123 |
+
},
|
| 37124 |
+
{
|
| 37125 |
+
"epoch": 8.55e-07,
|
| 37126 |
+
"grad_norm": 0.515625,
|
| 37127 |
+
"learning_rate": 8.724806494429514e-05,
|
| 37128 |
+
"loss": 3.0697,
|
| 37129 |
+
"step": 51710
|
| 37130 |
+
},
|
| 37131 |
+
{
|
| 37132 |
+
"epoch": 8.6e-07,
|
| 37133 |
+
"grad_norm": 0.498046875,
|
| 37134 |
+
"learning_rate": 8.71623187525979e-05,
|
| 37135 |
+
"loss": 3.049,
|
| 37136 |
+
"step": 51720
|
| 37137 |
+
},
|
| 37138 |
+
{
|
| 37139 |
+
"epoch": 8.65e-07,
|
| 37140 |
+
"grad_norm": 0.53125,
|
| 37141 |
+
"learning_rate": 8.707661957445964e-05,
|
| 37142 |
+
"loss": 3.0461,
|
| 37143 |
+
"step": 51730
|
| 37144 |
+
},
|
| 37145 |
+
{
|
| 37146 |
+
"epoch": 8.7e-07,
|
| 37147 |
+
"grad_norm": 0.5078125,
|
| 37148 |
+
"learning_rate": 8.699096746164237e-05,
|
| 37149 |
+
"loss": 3.0484,
|
| 37150 |
+
"step": 51740
|
| 37151 |
+
},
|
| 37152 |
+
{
|
| 37153 |
+
"epoch": 8.75e-07,
|
| 37154 |
+
"grad_norm": 0.4765625,
|
| 37155 |
+
"learning_rate": 8.690536246587934e-05,
|
| 37156 |
+
"loss": 3.0563,
|
| 37157 |
+
"step": 51750
|
| 37158 |
+
},
|
| 37159 |
+
{
|
| 37160 |
+
"epoch": 8.8e-07,
|
| 37161 |
+
"grad_norm": 0.5390625,
|
| 37162 |
+
"learning_rate": 8.681980463887575e-05,
|
| 37163 |
+
"loss": 3.0285,
|
| 37164 |
+
"step": 51760
|
| 37165 |
+
},
|
| 37166 |
+
{
|
| 37167 |
+
"epoch": 8.85e-07,
|
| 37168 |
+
"grad_norm": 0.482421875,
|
| 37169 |
+
"learning_rate": 8.673429403230783e-05,
|
| 37170 |
+
"loss": 3.0502,
|
| 37171 |
+
"step": 51770
|
| 37172 |
+
},
|
| 37173 |
+
{
|
| 37174 |
+
"epoch": 8.9e-07,
|
| 37175 |
+
"grad_norm": 0.462890625,
|
| 37176 |
+
"learning_rate": 8.664883069782372e-05,
|
| 37177 |
+
"loss": 3.0344,
|
| 37178 |
+
"step": 51780
|
| 37179 |
+
},
|
| 37180 |
+
{
|
| 37181 |
+
"epoch": 8.95e-07,
|
| 37182 |
+
"grad_norm": 0.498046875,
|
| 37183 |
+
"learning_rate": 8.656341468704282e-05,
|
| 37184 |
+
"loss": 3.0759,
|
| 37185 |
+
"step": 51790
|
| 37186 |
+
},
|
| 37187 |
+
{
|
| 37188 |
+
"epoch": 9e-07,
|
| 37189 |
+
"grad_norm": 0.494140625,
|
| 37190 |
+
"learning_rate": 8.647804605155591e-05,
|
| 37191 |
+
"loss": 3.0505,
|
| 37192 |
+
"step": 51800
|
| 37193 |
+
},
|
| 37194 |
+
{
|
| 37195 |
+
"epoch": 9.05e-07,
|
| 37196 |
+
"grad_norm": 0.515625,
|
| 37197 |
+
"learning_rate": 8.639272484292526e-05,
|
| 37198 |
+
"loss": 3.0457,
|
| 37199 |
+
"step": 51810
|
| 37200 |
+
},
|
| 37201 |
+
{
|
| 37202 |
+
"epoch": 9.1e-07,
|
| 37203 |
+
"grad_norm": 0.478515625,
|
| 37204 |
+
"learning_rate": 8.63074511126844e-05,
|
| 37205 |
+
"loss": 3.0269,
|
| 37206 |
+
"step": 51820
|
| 37207 |
+
},
|
| 37208 |
+
{
|
| 37209 |
+
"epoch": 9.15e-07,
|
| 37210 |
+
"grad_norm": 0.482421875,
|
| 37211 |
+
"learning_rate": 8.622222491233819e-05,
|
| 37212 |
+
"loss": 3.0507,
|
| 37213 |
+
"step": 51830
|
| 37214 |
+
},
|
| 37215 |
+
{
|
| 37216 |
+
"epoch": 9.2e-07,
|
| 37217 |
+
"grad_norm": 0.4765625,
|
| 37218 |
+
"learning_rate": 8.613704629336296e-05,
|
| 37219 |
+
"loss": 3.0477,
|
| 37220 |
+
"step": 51840
|
| 37221 |
+
},
|
| 37222 |
+
{
|
| 37223 |
+
"epoch": 9.25e-07,
|
| 37224 |
+
"grad_norm": 0.490234375,
|
| 37225 |
+
"learning_rate": 8.605191530720594e-05,
|
| 37226 |
+
"loss": 3.0362,
|
| 37227 |
+
"step": 51850
|
| 37228 |
+
},
|
| 37229 |
+
{
|
| 37230 |
+
"epoch": 9.3e-07,
|
| 37231 |
+
"grad_norm": 0.4765625,
|
| 37232 |
+
"learning_rate": 8.5966832005286e-05,
|
| 37233 |
+
"loss": 3.0317,
|
| 37234 |
+
"step": 51860
|
| 37235 |
+
},
|
| 37236 |
+
{
|
| 37237 |
+
"epoch": 9.35e-07,
|
| 37238 |
+
"grad_norm": 0.486328125,
|
| 37239 |
+
"learning_rate": 8.588179643899294e-05,
|
| 37240 |
+
"loss": 3.0339,
|
| 37241 |
+
"step": 51870
|
| 37242 |
+
},
|
| 37243 |
+
{
|
| 37244 |
+
"epoch": 9.4e-07,
|
| 37245 |
+
"grad_norm": 0.45703125,
|
| 37246 |
+
"learning_rate": 8.579680865968783e-05,
|
| 37247 |
+
"loss": 3.0329,
|
| 37248 |
+
"step": 51880
|
| 37249 |
+
},
|
| 37250 |
+
{
|
| 37251 |
+
"epoch": 9.45e-07,
|
| 37252 |
+
"grad_norm": 0.46875,
|
| 37253 |
+
"learning_rate": 8.571186871870279e-05,
|
| 37254 |
+
"loss": 3.0419,
|
| 37255 |
+
"step": 51890
|
| 37256 |
+
},
|
| 37257 |
+
{
|
| 37258 |
+
"epoch": 9.5e-07,
|
| 37259 |
+
"grad_norm": 0.54296875,
|
| 37260 |
+
"learning_rate": 8.562697666734117e-05,
|
| 37261 |
+
"loss": 3.0322,
|
| 37262 |
+
"step": 51900
|
| 37263 |
+
},
|
| 37264 |
+
{
|
| 37265 |
+
"epoch": 9.55e-07,
|
| 37266 |
+
"grad_norm": 0.48828125,
|
| 37267 |
+
"learning_rate": 8.554213255687725e-05,
|
| 37268 |
+
"loss": 3.0449,
|
| 37269 |
+
"step": 51910
|
| 37270 |
+
},
|
| 37271 |
+
{
|
| 37272 |
+
"epoch": 9.6e-07,
|
| 37273 |
+
"grad_norm": 0.46875,
|
| 37274 |
+
"learning_rate": 8.54573364385566e-05,
|
| 37275 |
+
"loss": 3.0549,
|
| 37276 |
+
"step": 51920
|
| 37277 |
+
},
|
| 37278 |
+
{
|
| 37279 |
+
"epoch": 9.65e-07,
|
| 37280 |
+
"grad_norm": 0.486328125,
|
| 37281 |
+
"learning_rate": 8.537258836359544e-05,
|
| 37282 |
+
"loss": 3.056,
|
| 37283 |
+
"step": 51930
|
| 37284 |
+
},
|
| 37285 |
+
{
|
| 37286 |
+
"epoch": 9.7e-07,
|
| 37287 |
+
"grad_norm": 0.4609375,
|
| 37288 |
+
"learning_rate": 8.528788838318132e-05,
|
| 37289 |
+
"loss": 3.037,
|
| 37290 |
+
"step": 51940
|
| 37291 |
+
},
|
| 37292 |
+
{
|
| 37293 |
+
"epoch": 9.75e-07,
|
| 37294 |
+
"grad_norm": 0.54296875,
|
| 37295 |
+
"learning_rate": 8.520323654847253e-05,
|
| 37296 |
+
"loss": 3.0611,
|
| 37297 |
+
"step": 51950
|
| 37298 |
+
},
|
| 37299 |
+
{
|
| 37300 |
+
"epoch": 9.8e-07,
|
| 37301 |
+
"grad_norm": 0.51171875,
|
| 37302 |
+
"learning_rate": 8.511863291059835e-05,
|
| 37303 |
+
"loss": 3.0313,
|
| 37304 |
+
"step": 51960
|
| 37305 |
+
},
|
| 37306 |
+
{
|
| 37307 |
+
"epoch": 9.85e-07,
|
| 37308 |
+
"grad_norm": 0.5390625,
|
| 37309 |
+
"learning_rate": 8.503407752065897e-05,
|
| 37310 |
+
"loss": 3.0129,
|
| 37311 |
+
"step": 51970
|
| 37312 |
+
},
|
| 37313 |
+
{
|
| 37314 |
+
"epoch": 9.9e-07,
|
| 37315 |
+
"grad_norm": 0.55859375,
|
| 37316 |
+
"learning_rate": 8.494957042972535e-05,
|
| 37317 |
+
"loss": 3.0485,
|
| 37318 |
+
"step": 51980
|
| 37319 |
+
},
|
| 37320 |
+
{
|
| 37321 |
+
"epoch": 9.95e-07,
|
| 37322 |
+
"grad_norm": 0.478515625,
|
| 37323 |
+
"learning_rate": 8.48651116888394e-05,
|
| 37324 |
+
"loss": 3.0369,
|
| 37325 |
+
"step": 51990
|
| 37326 |
+
},
|
| 37327 |
+
{
|
| 37328 |
+
"epoch": 1e-06,
|
| 37329 |
+
"grad_norm": 0.5234375,
|
| 37330 |
+
"learning_rate": 8.478070134901371e-05,
|
| 37331 |
+
"loss": 3.0345,
|
| 37332 |
+
"step": 52000
|
| 37333 |
+
},
|
| 37334 |
+
{
|
| 37335 |
+
"epoch": 1e-06,
|
| 37336 |
+
"eval_loss": 3.106865882873535,
|
| 37337 |
+
"eval_runtime": 0.064,
|
| 37338 |
+
"eval_samples_per_second": 500.325,
|
| 37339 |
+
"eval_steps_per_second": 15.635,
|
| 37340 |
+
"step": 52000
|
| 37341 |
}
|
| 37342 |
],
|
| 37343 |
"logging_steps": 10,
|
|
|
|
| 37357 |
"attributes": {}
|
| 37358 |
}
|
| 37359 |
},
|
| 37360 |
+
"total_flos": 9.6662924034048e+17,
|
| 37361 |
+
"train_batch_size": 240,
|
| 37362 |
"trial_name": null,
|
| 37363 |
"trial_params": null
|
| 37364 |
}
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5969
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:126c9c312afe7ca952c65954bf0518ea6213e6ceedab4d699c21dfeea6d89b1f
|
| 3 |
size 5969
|