Training in progress, step 282, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 101752088
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41c04db13401440bb120e3569a23dbda67cd78267d7c0b1c77f3d3b3cee4cdee
|
| 3 |
size 101752088
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 203713238
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1a032c7471714a5d4a253e904e854da99f9722e45c96bc0da82257681a15490
|
| 3 |
size 203713238
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:45144a3e80d33a7835b701c1b7b63faebde586b75158a47eb826cd0228136ec0
|
| 3 |
size 14512
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:feb6925b0db33b6f02f0ccbd50be336d8d47178a933641d2c637051d854a6c60
|
| 3 |
size 14512
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c28833a5c9fe2e108390575900c0ade8d470ff95484328f12052b199c28b6360
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 141,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1010,6 +1010,1001 @@
|
|
| 1010 |
"eval_samples_per_second": 5.938,
|
| 1011 |
"eval_steps_per_second": 1.489,
|
| 1012 |
"step": 141
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1013 |
}
|
| 1014 |
],
|
| 1015 |
"logging_steps": 1,
|
|
@@ -1029,7 +2024,7 @@
|
|
| 1029 |
"attributes": {}
|
| 1030 |
}
|
| 1031 |
},
|
| 1032 |
-
"total_flos":
|
| 1033 |
"train_batch_size": 2,
|
| 1034 |
"trial_name": null,
|
| 1035 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5007769145394007,
|
| 5 |
"eval_steps": 141,
|
| 6 |
+
"global_step": 282,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1010 |
"eval_samples_per_second": 5.938,
|
| 1011 |
"eval_steps_per_second": 1.489,
|
| 1012 |
"step": 141
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.25216426193118757,
|
| 1016 |
+
"grad_norm": 0.3713844120502472,
|
| 1017 |
+
"learning_rate": 0.0001845441764722514,
|
| 1018 |
+
"loss": 0.7688,
|
| 1019 |
+
"step": 142
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.2539400665926748,
|
| 1023 |
+
"grad_norm": 0.352450430393219,
|
| 1024 |
+
"learning_rate": 0.00018421553219875658,
|
| 1025 |
+
"loss": 0.7769,
|
| 1026 |
+
"step": 143
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.25571587125416206,
|
| 1030 |
+
"grad_norm": 0.3609173893928528,
|
| 1031 |
+
"learning_rate": 0.00018388372960726228,
|
| 1032 |
+
"loss": 0.7718,
|
| 1033 |
+
"step": 144
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.25749167591564925,
|
| 1037 |
+
"grad_norm": 0.36195874214172363,
|
| 1038 |
+
"learning_rate": 0.00018354878114129367,
|
| 1039 |
+
"loss": 0.7375,
|
| 1040 |
+
"step": 145
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.2592674805771365,
|
| 1044 |
+
"grad_norm": 0.3802485466003418,
|
| 1045 |
+
"learning_rate": 0.00018321069936235503,
|
| 1046 |
+
"loss": 0.7778,
|
| 1047 |
+
"step": 146
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.26104328523862375,
|
| 1051 |
+
"grad_norm": 0.38449469208717346,
|
| 1052 |
+
"learning_rate": 0.00018286949694945866,
|
| 1053 |
+
"loss": 0.7458,
|
| 1054 |
+
"step": 147
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.262819089900111,
|
| 1058 |
+
"grad_norm": 0.3975572884082794,
|
| 1059 |
+
"learning_rate": 0.00018252518669864936,
|
| 1060 |
+
"loss": 0.7367,
|
| 1061 |
+
"step": 148
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.26459489456159824,
|
| 1065 |
+
"grad_norm": 0.49581316113471985,
|
| 1066 |
+
"learning_rate": 0.0001821777815225245,
|
| 1067 |
+
"loss": 0.7948,
|
| 1068 |
+
"step": 149
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.2663706992230855,
|
| 1072 |
+
"grad_norm": 0.5556712746620178,
|
| 1073 |
+
"learning_rate": 0.00018182729444974992,
|
| 1074 |
+
"loss": 0.8143,
|
| 1075 |
+
"step": 150
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.2681465038845727,
|
| 1079 |
+
"grad_norm": 0.3207700848579407,
|
| 1080 |
+
"learning_rate": 0.00018147373862457107,
|
| 1081 |
+
"loss": 0.8578,
|
| 1082 |
+
"step": 151
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.2699223085460599,
|
| 1086 |
+
"grad_norm": 0.3484250605106354,
|
| 1087 |
+
"learning_rate": 0.00018111712730632022,
|
| 1088 |
+
"loss": 0.8757,
|
| 1089 |
+
"step": 152
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.27169811320754716,
|
| 1093 |
+
"grad_norm": 0.33792024850845337,
|
| 1094 |
+
"learning_rate": 0.0001807574738689193,
|
| 1095 |
+
"loss": 0.8464,
|
| 1096 |
+
"step": 153
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.2734739178690344,
|
| 1100 |
+
"grad_norm": 0.3430371582508087,
|
| 1101 |
+
"learning_rate": 0.000180394791800378,
|
| 1102 |
+
"loss": 0.8607,
|
| 1103 |
+
"step": 154
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.27524972253052166,
|
| 1107 |
+
"grad_norm": 0.3120534420013428,
|
| 1108 |
+
"learning_rate": 0.00018002909470228842,
|
| 1109 |
+
"loss": 0.8392,
|
| 1110 |
+
"step": 155
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.2770255271920089,
|
| 1114 |
+
"grad_norm": 0.3126620054244995,
|
| 1115 |
+
"learning_rate": 0.00017966039628931446,
|
| 1116 |
+
"loss": 0.8191,
|
| 1117 |
+
"step": 156
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.2788013318534961,
|
| 1121 |
+
"grad_norm": 0.32269468903541565,
|
| 1122 |
+
"learning_rate": 0.00017928871038867784,
|
| 1123 |
+
"loss": 0.8164,
|
| 1124 |
+
"step": 157
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.28057713651498334,
|
| 1128 |
+
"grad_norm": 0.3052617907524109,
|
| 1129 |
+
"learning_rate": 0.00017891405093963938,
|
| 1130 |
+
"loss": 0.8268,
|
| 1131 |
+
"step": 158
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.2823529411764706,
|
| 1135 |
+
"grad_norm": 0.29926028847694397,
|
| 1136 |
+
"learning_rate": 0.00017853643199297633,
|
| 1137 |
+
"loss": 0.7847,
|
| 1138 |
+
"step": 159
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.28412874583795783,
|
| 1142 |
+
"grad_norm": 0.2997240722179413,
|
| 1143 |
+
"learning_rate": 0.00017815586771045535,
|
| 1144 |
+
"loss": 0.8143,
|
| 1145 |
+
"step": 160
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.2859045504994451,
|
| 1149 |
+
"grad_norm": 0.29772111773490906,
|
| 1150 |
+
"learning_rate": 0.0001777723723643014,
|
| 1151 |
+
"loss": 0.7412,
|
| 1152 |
+
"step": 161
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.2876803551609323,
|
| 1156 |
+
"grad_norm": 0.3138352632522583,
|
| 1157 |
+
"learning_rate": 0.0001773859603366626,
|
| 1158 |
+
"loss": 0.7747,
|
| 1159 |
+
"step": 162
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.2894561598224195,
|
| 1163 |
+
"grad_norm": 0.32726818323135376,
|
| 1164 |
+
"learning_rate": 0.00017699664611907072,
|
| 1165 |
+
"loss": 0.8123,
|
| 1166 |
+
"step": 163
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.29123196448390676,
|
| 1170 |
+
"grad_norm": 0.3244825005531311,
|
| 1171 |
+
"learning_rate": 0.0001766044443118978,
|
| 1172 |
+
"loss": 0.7705,
|
| 1173 |
+
"step": 164
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.293007769145394,
|
| 1177 |
+
"grad_norm": 0.35875847935676575,
|
| 1178 |
+
"learning_rate": 0.00017620936962380856,
|
| 1179 |
+
"loss": 0.7881,
|
| 1180 |
+
"step": 165
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.29478357380688125,
|
| 1184 |
+
"grad_norm": 0.36488401889801025,
|
| 1185 |
+
"learning_rate": 0.00017581143687120875,
|
| 1186 |
+
"loss": 0.7956,
|
| 1187 |
+
"step": 166
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.2965593784683685,
|
| 1191 |
+
"grad_norm": 0.33817097544670105,
|
| 1192 |
+
"learning_rate": 0.00017541066097768963,
|
| 1193 |
+
"loss": 0.7719,
|
| 1194 |
+
"step": 167
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.29833518312985574,
|
| 1198 |
+
"grad_norm": 0.36390411853790283,
|
| 1199 |
+
"learning_rate": 0.0001750070569734681,
|
| 1200 |
+
"loss": 0.8172,
|
| 1201 |
+
"step": 168
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.30011098779134293,
|
| 1205 |
+
"grad_norm": 0.34076422452926636,
|
| 1206 |
+
"learning_rate": 0.00017460063999482316,
|
| 1207 |
+
"loss": 0.7419,
|
| 1208 |
+
"step": 169
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.3018867924528302,
|
| 1212 |
+
"grad_norm": 0.39437592029571533,
|
| 1213 |
+
"learning_rate": 0.00017419142528352817,
|
| 1214 |
+
"loss": 0.7519,
|
| 1215 |
+
"step": 170
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.3036625971143174,
|
| 1219 |
+
"grad_norm": 0.4019312560558319,
|
| 1220 |
+
"learning_rate": 0.00017377942818627942,
|
| 1221 |
+
"loss": 0.7944,
|
| 1222 |
+
"step": 171
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.30543840177580467,
|
| 1226 |
+
"grad_norm": 0.40751898288726807,
|
| 1227 |
+
"learning_rate": 0.00017336466415412028,
|
| 1228 |
+
"loss": 0.7827,
|
| 1229 |
+
"step": 172
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.3072142064372919,
|
| 1233 |
+
"grad_norm": 0.4780448079109192,
|
| 1234 |
+
"learning_rate": 0.0001729471487418621,
|
| 1235 |
+
"loss": 0.7872,
|
| 1236 |
+
"step": 173
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.30899001109877916,
|
| 1240 |
+
"grad_norm": 0.40511685609817505,
|
| 1241 |
+
"learning_rate": 0.0001725268976075005,
|
| 1242 |
+
"loss": 0.7642,
|
| 1243 |
+
"step": 174
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.31076581576026635,
|
| 1247 |
+
"grad_norm": 0.5618127584457397,
|
| 1248 |
+
"learning_rate": 0.0001721039265116285,
|
| 1249 |
+
"loss": 0.872,
|
| 1250 |
+
"step": 175
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.3125416204217536,
|
| 1254 |
+
"grad_norm": 0.294917494058609,
|
| 1255 |
+
"learning_rate": 0.00017167825131684513,
|
| 1256 |
+
"loss": 0.8545,
|
| 1257 |
+
"step": 176
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.31431742508324084,
|
| 1261 |
+
"grad_norm": 0.3281805217266083,
|
| 1262 |
+
"learning_rate": 0.00017124988798716083,
|
| 1263 |
+
"loss": 0.8404,
|
| 1264 |
+
"step": 177
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.3160932297447281,
|
| 1268 |
+
"grad_norm": 0.33336278796195984,
|
| 1269 |
+
"learning_rate": 0.00017081885258739846,
|
| 1270 |
+
"loss": 0.8495,
|
| 1271 |
+
"step": 178
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.31786903440621533,
|
| 1275 |
+
"grad_norm": 0.3366440236568451,
|
| 1276 |
+
"learning_rate": 0.00017038516128259115,
|
| 1277 |
+
"loss": 0.8659,
|
| 1278 |
+
"step": 179
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.3196448390677026,
|
| 1282 |
+
"grad_norm": 0.32397955656051636,
|
| 1283 |
+
"learning_rate": 0.00016994883033737582,
|
| 1284 |
+
"loss": 0.8292,
|
| 1285 |
+
"step": 180
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.32142064372918977,
|
| 1289 |
+
"grad_norm": 0.2874945402145386,
|
| 1290 |
+
"learning_rate": 0.00016950987611538324,
|
| 1291 |
+
"loss": 0.7949,
|
| 1292 |
+
"step": 181
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.323196448390677,
|
| 1296 |
+
"grad_norm": 0.3074096143245697,
|
| 1297 |
+
"learning_rate": 0.00016906831507862443,
|
| 1298 |
+
"loss": 0.8076,
|
| 1299 |
+
"step": 182
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.32497225305216426,
|
| 1303 |
+
"grad_norm": 0.30116966366767883,
|
| 1304 |
+
"learning_rate": 0.0001686241637868734,
|
| 1305 |
+
"loss": 0.8058,
|
| 1306 |
+
"step": 183
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.3267480577136515,
|
| 1310 |
+
"grad_norm": 0.3052218556404114,
|
| 1311 |
+
"learning_rate": 0.00016817743889704565,
|
| 1312 |
+
"loss": 0.8067,
|
| 1313 |
+
"step": 184
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.32852386237513875,
|
| 1317 |
+
"grad_norm": 0.3073555827140808,
|
| 1318 |
+
"learning_rate": 0.00016772815716257412,
|
| 1319 |
+
"loss": 0.8496,
|
| 1320 |
+
"step": 185
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.33029966703662594,
|
| 1324 |
+
"grad_norm": 0.289145290851593,
|
| 1325 |
+
"learning_rate": 0.0001672763354327804,
|
| 1326 |
+
"loss": 0.7362,
|
| 1327 |
+
"step": 186
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.3320754716981132,
|
| 1331 |
+
"grad_norm": 0.31561294198036194,
|
| 1332 |
+
"learning_rate": 0.00016682199065224307,
|
| 1333 |
+
"loss": 0.802,
|
| 1334 |
+
"step": 187
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.33385127635960044,
|
| 1338 |
+
"grad_norm": 0.2900339365005493,
|
| 1339 |
+
"learning_rate": 0.00016636513986016213,
|
| 1340 |
+
"loss": 0.7432,
|
| 1341 |
+
"step": 188
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.3356270810210877,
|
| 1345 |
+
"grad_norm": 0.3267146646976471,
|
| 1346 |
+
"learning_rate": 0.0001659058001897201,
|
| 1347 |
+
"loss": 0.7771,
|
| 1348 |
+
"step": 189
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 0.3374028856825749,
|
| 1352 |
+
"grad_norm": 0.3258307874202728,
|
| 1353 |
+
"learning_rate": 0.00016544398886743933,
|
| 1354 |
+
"loss": 0.7345,
|
| 1355 |
+
"step": 190
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 0.3391786903440622,
|
| 1359 |
+
"grad_norm": 0.32989659905433655,
|
| 1360 |
+
"learning_rate": 0.000164979723212536,
|
| 1361 |
+
"loss": 0.7383,
|
| 1362 |
+
"step": 191
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 0.34095449500554936,
|
| 1366 |
+
"grad_norm": 0.3265599310398102,
|
| 1367 |
+
"learning_rate": 0.00016451302063627066,
|
| 1368 |
+
"loss": 0.6977,
|
| 1369 |
+
"step": 192
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 0.3427302996670366,
|
| 1373 |
+
"grad_norm": 0.39376598596572876,
|
| 1374 |
+
"learning_rate": 0.00016404389864129533,
|
| 1375 |
+
"loss": 0.7851,
|
| 1376 |
+
"step": 193
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.34450610432852385,
|
| 1380 |
+
"grad_norm": 0.40358301997184753,
|
| 1381 |
+
"learning_rate": 0.00016357237482099684,
|
| 1382 |
+
"loss": 0.7928,
|
| 1383 |
+
"step": 194
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 0.3462819089900111,
|
| 1387 |
+
"grad_norm": 0.3747034966945648,
|
| 1388 |
+
"learning_rate": 0.00016309846685883726,
|
| 1389 |
+
"loss": 0.7751,
|
| 1390 |
+
"step": 195
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 0.34805771365149835,
|
| 1394 |
+
"grad_norm": 0.4160248041152954,
|
| 1395 |
+
"learning_rate": 0.00016262219252769064,
|
| 1396 |
+
"loss": 0.8035,
|
| 1397 |
+
"step": 196
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 0.3498335183129856,
|
| 1401 |
+
"grad_norm": 0.39067476987838745,
|
| 1402 |
+
"learning_rate": 0.00016214356968917648,
|
| 1403 |
+
"loss": 0.6726,
|
| 1404 |
+
"step": 197
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.3516093229744728,
|
| 1408 |
+
"grad_norm": 0.4980023205280304,
|
| 1409 |
+
"learning_rate": 0.00016166261629298995,
|
| 1410 |
+
"loss": 0.7917,
|
| 1411 |
+
"step": 198
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 0.35338512763596003,
|
| 1415 |
+
"grad_norm": 0.4774058163166046,
|
| 1416 |
+
"learning_rate": 0.0001611793503762285,
|
| 1417 |
+
"loss": 0.7599,
|
| 1418 |
+
"step": 199
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.3551609322974473,
|
| 1422 |
+
"grad_norm": 0.5196167230606079,
|
| 1423 |
+
"learning_rate": 0.00016069379006271566,
|
| 1424 |
+
"loss": 0.7608,
|
| 1425 |
+
"step": 200
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 0.3569367369589345,
|
| 1429 |
+
"grad_norm": 0.2735799551010132,
|
| 1430 |
+
"learning_rate": 0.00016020595356232135,
|
| 1431 |
+
"loss": 0.8588,
|
| 1432 |
+
"step": 201
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 0.35871254162042177,
|
| 1436 |
+
"grad_norm": 0.30770814418792725,
|
| 1437 |
+
"learning_rate": 0.00015971585917027862,
|
| 1438 |
+
"loss": 0.8222,
|
| 1439 |
+
"step": 202
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 0.360488346281909,
|
| 1443 |
+
"grad_norm": 0.317123144865036,
|
| 1444 |
+
"learning_rate": 0.00015922352526649803,
|
| 1445 |
+
"loss": 0.7941,
|
| 1446 |
+
"step": 203
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.3622641509433962,
|
| 1450 |
+
"grad_norm": 0.32672154903411865,
|
| 1451 |
+
"learning_rate": 0.00015872897031487791,
|
| 1452 |
+
"loss": 0.867,
|
| 1453 |
+
"step": 204
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 0.36403995560488345,
|
| 1457 |
+
"grad_norm": 0.3169744610786438,
|
| 1458 |
+
"learning_rate": 0.00015823221286261215,
|
| 1459 |
+
"loss": 0.8781,
|
| 1460 |
+
"step": 205
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.3658157602663707,
|
| 1464 |
+
"grad_norm": 0.30588722229003906,
|
| 1465 |
+
"learning_rate": 0.00015773327153949465,
|
| 1466 |
+
"loss": 0.7827,
|
| 1467 |
+
"step": 206
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 0.36759156492785794,
|
| 1471 |
+
"grad_norm": 0.3179618716239929,
|
| 1472 |
+
"learning_rate": 0.0001572321650572205,
|
| 1473 |
+
"loss": 0.8178,
|
| 1474 |
+
"step": 207
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 0.3693673695893452,
|
| 1478 |
+
"grad_norm": 0.3094286322593689,
|
| 1479 |
+
"learning_rate": 0.00015672891220868432,
|
| 1480 |
+
"loss": 0.7966,
|
| 1481 |
+
"step": 208
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.37114317425083243,
|
| 1485 |
+
"grad_norm": 0.31584280729293823,
|
| 1486 |
+
"learning_rate": 0.00015622353186727544,
|
| 1487 |
+
"loss": 0.7982,
|
| 1488 |
+
"step": 209
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.3729189789123196,
|
| 1492 |
+
"grad_norm": 0.29120850563049316,
|
| 1493 |
+
"learning_rate": 0.0001557160429861702,
|
| 1494 |
+
"loss": 0.7789,
|
| 1495 |
+
"step": 210
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.37469478357380687,
|
| 1499 |
+
"grad_norm": 0.29743698239326477,
|
| 1500 |
+
"learning_rate": 0.000155206464597621,
|
| 1501 |
+
"loss": 0.7799,
|
| 1502 |
+
"step": 211
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.3764705882352941,
|
| 1506 |
+
"grad_norm": 0.31440189480781555,
|
| 1507 |
+
"learning_rate": 0.00015469481581224272,
|
| 1508 |
+
"loss": 0.7661,
|
| 1509 |
+
"step": 212
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.37824639289678136,
|
| 1513 |
+
"grad_norm": 0.3395606279373169,
|
| 1514 |
+
"learning_rate": 0.00015418111581829574,
|
| 1515 |
+
"loss": 0.7657,
|
| 1516 |
+
"step": 213
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 0.3800221975582686,
|
| 1520 |
+
"grad_norm": 0.31749066710472107,
|
| 1521 |
+
"learning_rate": 0.0001536653838809667,
|
| 1522 |
+
"loss": 0.7913,
|
| 1523 |
+
"step": 214
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.38179800221975585,
|
| 1527 |
+
"grad_norm": 0.3586166501045227,
|
| 1528 |
+
"learning_rate": 0.0001531476393416456,
|
| 1529 |
+
"loss": 0.7774,
|
| 1530 |
+
"step": 215
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.38357380688124304,
|
| 1534 |
+
"grad_norm": 0.32895100116729736,
|
| 1535 |
+
"learning_rate": 0.0001526279016172008,
|
| 1536 |
+
"loss": 0.7882,
|
| 1537 |
+
"step": 216
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 0.3853496115427303,
|
| 1541 |
+
"grad_norm": 0.3541489839553833,
|
| 1542 |
+
"learning_rate": 0.00015210619019925066,
|
| 1543 |
+
"loss": 0.7708,
|
| 1544 |
+
"step": 217
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.38712541620421753,
|
| 1548 |
+
"grad_norm": 0.3232908546924591,
|
| 1549 |
+
"learning_rate": 0.00015158252465343242,
|
| 1550 |
+
"loss": 0.7238,
|
| 1551 |
+
"step": 218
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.3889012208657048,
|
| 1555 |
+
"grad_norm": 0.36565467715263367,
|
| 1556 |
+
"learning_rate": 0.00015105692461866874,
|
| 1557 |
+
"loss": 0.7685,
|
| 1558 |
+
"step": 219
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 0.390677025527192,
|
| 1562 |
+
"grad_norm": 0.3799486756324768,
|
| 1563 |
+
"learning_rate": 0.000150529409806431,
|
| 1564 |
+
"loss": 0.7296,
|
| 1565 |
+
"step": 220
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.39245283018867927,
|
| 1569 |
+
"grad_norm": 0.4193985164165497,
|
| 1570 |
+
"learning_rate": 0.00015000000000000001,
|
| 1571 |
+
"loss": 0.7731,
|
| 1572 |
+
"step": 221
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.39422863485016646,
|
| 1576 |
+
"grad_norm": 0.4226386845111847,
|
| 1577 |
+
"learning_rate": 0.00014946871505372425,
|
| 1578 |
+
"loss": 0.8048,
|
| 1579 |
+
"step": 222
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 0.3960044395116537,
|
| 1583 |
+
"grad_norm": 0.40805166959762573,
|
| 1584 |
+
"learning_rate": 0.00014893557489227517,
|
| 1585 |
+
"loss": 0.7389,
|
| 1586 |
+
"step": 223
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.39778024417314095,
|
| 1590 |
+
"grad_norm": 0.5135468244552612,
|
| 1591 |
+
"learning_rate": 0.0001484005995098999,
|
| 1592 |
+
"loss": 0.779,
|
| 1593 |
+
"step": 224
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.3995560488346282,
|
| 1597 |
+
"grad_norm": 0.6674650311470032,
|
| 1598 |
+
"learning_rate": 0.0001478638089696716,
|
| 1599 |
+
"loss": 0.82,
|
| 1600 |
+
"step": 225
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.40133185349611544,
|
| 1604 |
+
"grad_norm": 0.3206911087036133,
|
| 1605 |
+
"learning_rate": 0.00014732522340273684,
|
| 1606 |
+
"loss": 0.8985,
|
| 1607 |
+
"step": 226
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 0.4031076581576027,
|
| 1611 |
+
"grad_norm": 0.33583980798721313,
|
| 1612 |
+
"learning_rate": 0.0001467848630075608,
|
| 1613 |
+
"loss": 0.8171,
|
| 1614 |
+
"step": 227
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.4048834628190899,
|
| 1618 |
+
"grad_norm": 0.3324304223060608,
|
| 1619 |
+
"learning_rate": 0.00014624274804916958,
|
| 1620 |
+
"loss": 0.8531,
|
| 1621 |
+
"step": 228
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 0.4066592674805771,
|
| 1625 |
+
"grad_norm": 0.32210710644721985,
|
| 1626 |
+
"learning_rate": 0.00014569889885839037,
|
| 1627 |
+
"loss": 0.8349,
|
| 1628 |
+
"step": 229
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.40843507214206437,
|
| 1632 |
+
"grad_norm": 0.30829885601997375,
|
| 1633 |
+
"learning_rate": 0.00014515333583108896,
|
| 1634 |
+
"loss": 0.8176,
|
| 1635 |
+
"step": 230
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.4102108768035516,
|
| 1639 |
+
"grad_norm": 0.31730225682258606,
|
| 1640 |
+
"learning_rate": 0.00014460607942740468,
|
| 1641 |
+
"loss": 0.8109,
|
| 1642 |
+
"step": 231
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 0.41198668146503886,
|
| 1646 |
+
"grad_norm": 0.32128164172172546,
|
| 1647 |
+
"learning_rate": 0.00014405715017098335,
|
| 1648 |
+
"loss": 0.8049,
|
| 1649 |
+
"step": 232
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.4137624861265261,
|
| 1653 |
+
"grad_norm": 0.32257241010665894,
|
| 1654 |
+
"learning_rate": 0.00014350656864820733,
|
| 1655 |
+
"loss": 0.79,
|
| 1656 |
+
"step": 233
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.4155382907880133,
|
| 1660 |
+
"grad_norm": 0.29663363099098206,
|
| 1661 |
+
"learning_rate": 0.0001429543555074237,
|
| 1662 |
+
"loss": 0.7606,
|
| 1663 |
+
"step": 234
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 0.41731409544950054,
|
| 1667 |
+
"grad_norm": 0.3175968527793884,
|
| 1668 |
+
"learning_rate": 0.00014240053145816967,
|
| 1669 |
+
"loss": 0.8093,
|
| 1670 |
+
"step": 235
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.4190899001109878,
|
| 1674 |
+
"grad_norm": 0.30839797854423523,
|
| 1675 |
+
"learning_rate": 0.00014184511727039612,
|
| 1676 |
+
"loss": 0.8033,
|
| 1677 |
+
"step": 236
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.42086570477247504,
|
| 1681 |
+
"grad_norm": 0.32169485092163086,
|
| 1682 |
+
"learning_rate": 0.0001412881337736885,
|
| 1683 |
+
"loss": 0.7583,
|
| 1684 |
+
"step": 237
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.4226415094339623,
|
| 1688 |
+
"grad_norm": 0.3165202736854553,
|
| 1689 |
+
"learning_rate": 0.00014072960185648577,
|
| 1690 |
+
"loss": 0.7864,
|
| 1691 |
+
"step": 238
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.4244173140954495,
|
| 1695 |
+
"grad_norm": 0.3507262170314789,
|
| 1696 |
+
"learning_rate": 0.00014016954246529696,
|
| 1697 |
+
"loss": 0.8196,
|
| 1698 |
+
"step": 239
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.4261931187569367,
|
| 1702 |
+
"grad_norm": 0.3330634534358978,
|
| 1703 |
+
"learning_rate": 0.0001396079766039157,
|
| 1704 |
+
"loss": 0.7356,
|
| 1705 |
+
"step": 240
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.42796892341842396,
|
| 1709 |
+
"grad_norm": 0.3456502854824066,
|
| 1710 |
+
"learning_rate": 0.00013904492533263244,
|
| 1711 |
+
"loss": 0.7636,
|
| 1712 |
+
"step": 241
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.4297447280799112,
|
| 1716 |
+
"grad_norm": 0.3290559649467468,
|
| 1717 |
+
"learning_rate": 0.00013848040976744457,
|
| 1718 |
+
"loss": 0.6921,
|
| 1719 |
+
"step": 242
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.43152053274139845,
|
| 1723 |
+
"grad_norm": 0.34343284368515015,
|
| 1724 |
+
"learning_rate": 0.00013791445107926478,
|
| 1725 |
+
"loss": 0.7661,
|
| 1726 |
+
"step": 243
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.4332963374028857,
|
| 1730 |
+
"grad_norm": 0.34806933999061584,
|
| 1731 |
+
"learning_rate": 0.00013734707049312673,
|
| 1732 |
+
"loss": 0.7266,
|
| 1733 |
+
"step": 244
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.43507214206437295,
|
| 1737 |
+
"grad_norm": 0.3577682375907898,
|
| 1738 |
+
"learning_rate": 0.00013677828928738934,
|
| 1739 |
+
"loss": 0.7337,
|
| 1740 |
+
"step": 245
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.43684794672586014,
|
| 1744 |
+
"grad_norm": 0.37708649039268494,
|
| 1745 |
+
"learning_rate": 0.00013620812879293863,
|
| 1746 |
+
"loss": 0.6949,
|
| 1747 |
+
"step": 246
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.4386237513873474,
|
| 1751 |
+
"grad_norm": 0.3661216199398041,
|
| 1752 |
+
"learning_rate": 0.00013563661039238785,
|
| 1753 |
+
"loss": 0.7049,
|
| 1754 |
+
"step": 247
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.44039955604883463,
|
| 1758 |
+
"grad_norm": 0.4453539550304413,
|
| 1759 |
+
"learning_rate": 0.00013506375551927547,
|
| 1760 |
+
"loss": 0.7957,
|
| 1761 |
+
"step": 248
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.4421753607103219,
|
| 1765 |
+
"grad_norm": 0.46171826124191284,
|
| 1766 |
+
"learning_rate": 0.00013448958565726144,
|
| 1767 |
+
"loss": 0.7175,
|
| 1768 |
+
"step": 249
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.4439511653718091,
|
| 1772 |
+
"grad_norm": 0.6314205527305603,
|
| 1773 |
+
"learning_rate": 0.00013391412233932149,
|
| 1774 |
+
"loss": 0.8853,
|
| 1775 |
+
"step": 250
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.4457269700332963,
|
| 1779 |
+
"grad_norm": 0.29680782556533813,
|
| 1780 |
+
"learning_rate": 0.00013333738714693956,
|
| 1781 |
+
"loss": 0.8789,
|
| 1782 |
+
"step": 251
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.44750277469478356,
|
| 1786 |
+
"grad_norm": 0.30771735310554504,
|
| 1787 |
+
"learning_rate": 0.00013275940170929843,
|
| 1788 |
+
"loss": 0.8126,
|
| 1789 |
+
"step": 252
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.4492785793562708,
|
| 1793 |
+
"grad_norm": 0.3242880403995514,
|
| 1794 |
+
"learning_rate": 0.00013218018770246858,
|
| 1795 |
+
"loss": 0.7787,
|
| 1796 |
+
"step": 253
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.45105438401775805,
|
| 1800 |
+
"grad_norm": 0.33549076318740845,
|
| 1801 |
+
"learning_rate": 0.00013159976684859527,
|
| 1802 |
+
"loss": 0.8113,
|
| 1803 |
+
"step": 254
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.4528301886792453,
|
| 1807 |
+
"grad_norm": 0.34281155467033386,
|
| 1808 |
+
"learning_rate": 0.00013101816091508388,
|
| 1809 |
+
"loss": 0.8371,
|
| 1810 |
+
"step": 255
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.45460599334073254,
|
| 1814 |
+
"grad_norm": 0.3422442078590393,
|
| 1815 |
+
"learning_rate": 0.0001304353917137836,
|
| 1816 |
+
"loss": 0.8362,
|
| 1817 |
+
"step": 256
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.45638179800221973,
|
| 1821 |
+
"grad_norm": 0.3019155263900757,
|
| 1822 |
+
"learning_rate": 0.00012985148110016947,
|
| 1823 |
+
"loss": 0.7317,
|
| 1824 |
+
"step": 257
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.458157602663707,
|
| 1828 |
+
"grad_norm": 0.32793429493904114,
|
| 1829 |
+
"learning_rate": 0.0001292664509725226,
|
| 1830 |
+
"loss": 0.7861,
|
| 1831 |
+
"step": 258
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 0.4599334073251942,
|
| 1835 |
+
"grad_norm": 0.32433855533599854,
|
| 1836 |
+
"learning_rate": 0.00012868032327110904,
|
| 1837 |
+
"loss": 0.7708,
|
| 1838 |
+
"step": 259
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 0.46170921198668147,
|
| 1842 |
+
"grad_norm": 0.31858816742897034,
|
| 1843 |
+
"learning_rate": 0.00012809311997735696,
|
| 1844 |
+
"loss": 0.7754,
|
| 1845 |
+
"step": 260
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 0.4634850166481687,
|
| 1849 |
+
"grad_norm": 0.3172609210014343,
|
| 1850 |
+
"learning_rate": 0.00012750486311303218,
|
| 1851 |
+
"loss": 0.7839,
|
| 1852 |
+
"step": 261
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.46526082130965596,
|
| 1856 |
+
"grad_norm": 0.2951931953430176,
|
| 1857 |
+
"learning_rate": 0.00012691557473941243,
|
| 1858 |
+
"loss": 0.7261,
|
| 1859 |
+
"step": 262
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 0.46703662597114315,
|
| 1863 |
+
"grad_norm": 0.31385374069213867,
|
| 1864 |
+
"learning_rate": 0.00012632527695645993,
|
| 1865 |
+
"loss": 0.8221,
|
| 1866 |
+
"step": 263
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 0.4688124306326304,
|
| 1870 |
+
"grad_norm": 0.31157392263412476,
|
| 1871 |
+
"learning_rate": 0.0001257339919019925,
|
| 1872 |
+
"loss": 0.7711,
|
| 1873 |
+
"step": 264
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 0.47058823529411764,
|
| 1877 |
+
"grad_norm": 0.32580870389938354,
|
| 1878 |
+
"learning_rate": 0.00012514174175085345,
|
| 1879 |
+
"loss": 0.7592,
|
| 1880 |
+
"step": 265
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 0.4723640399556049,
|
| 1884 |
+
"grad_norm": 0.33285781741142273,
|
| 1885 |
+
"learning_rate": 0.00012454854871407994,
|
| 1886 |
+
"loss": 0.7349,
|
| 1887 |
+
"step": 266
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 0.47413984461709213,
|
| 1891 |
+
"grad_norm": 0.3179035186767578,
|
| 1892 |
+
"learning_rate": 0.0001239544350380699,
|
| 1893 |
+
"loss": 0.7338,
|
| 1894 |
+
"step": 267
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.4759156492785794,
|
| 1898 |
+
"grad_norm": 0.31393003463745117,
|
| 1899 |
+
"learning_rate": 0.00012335942300374788,
|
| 1900 |
+
"loss": 0.7088,
|
| 1901 |
+
"step": 268
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 0.47769145394006657,
|
| 1905 |
+
"grad_norm": 0.33285436034202576,
|
| 1906 |
+
"learning_rate": 0.00012276353492572935,
|
| 1907 |
+
"loss": 0.7069,
|
| 1908 |
+
"step": 269
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 0.4794672586015538,
|
| 1912 |
+
"grad_norm": 0.38329485058784485,
|
| 1913 |
+
"learning_rate": 0.00012216679315148386,
|
| 1914 |
+
"loss": 0.7093,
|
| 1915 |
+
"step": 270
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 0.48124306326304106,
|
| 1919 |
+
"grad_norm": 0.3584016263484955,
|
| 1920 |
+
"learning_rate": 0.00012156922006049702,
|
| 1921 |
+
"loss": 0.7513,
|
| 1922 |
+
"step": 271
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 0.4830188679245283,
|
| 1926 |
+
"grad_norm": 0.3995126187801361,
|
| 1927 |
+
"learning_rate": 0.00012097083806343103,
|
| 1928 |
+
"loss": 0.7384,
|
| 1929 |
+
"step": 272
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.48479467258601555,
|
| 1933 |
+
"grad_norm": 0.4097007215023041,
|
| 1934 |
+
"learning_rate": 0.00012037166960128443,
|
| 1935 |
+
"loss": 0.7794,
|
| 1936 |
+
"step": 273
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.4865704772475028,
|
| 1940 |
+
"grad_norm": 0.4780315160751343,
|
| 1941 |
+
"learning_rate": 0.00011977173714455034,
|
| 1942 |
+
"loss": 0.7437,
|
| 1943 |
+
"step": 274
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 0.48834628190899,
|
| 1947 |
+
"grad_norm": 0.5396427512168884,
|
| 1948 |
+
"learning_rate": 0.00011917106319237386,
|
| 1949 |
+
"loss": 0.7542,
|
| 1950 |
+
"step": 275
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 0.49012208657047723,
|
| 1954 |
+
"grad_norm": 0.29439178109169006,
|
| 1955 |
+
"learning_rate": 0.00011856967027170818,
|
| 1956 |
+
"loss": 0.8389,
|
| 1957 |
+
"step": 276
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 0.4918978912319645,
|
| 1961 |
+
"grad_norm": 0.3243663012981415,
|
| 1962 |
+
"learning_rate": 0.00011796758093646989,
|
| 1963 |
+
"loss": 0.8767,
|
| 1964 |
+
"step": 277
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 0.4936736958934517,
|
| 1968 |
+
"grad_norm": 0.342454195022583,
|
| 1969 |
+
"learning_rate": 0.00011736481776669306,
|
| 1970 |
+
"loss": 0.8538,
|
| 1971 |
+
"step": 278
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.49544950055493897,
|
| 1975 |
+
"grad_norm": 0.30882903933525085,
|
| 1976 |
+
"learning_rate": 0.00011676140336768236,
|
| 1977 |
+
"loss": 0.7766,
|
| 1978 |
+
"step": 279
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.4972253052164262,
|
| 1982 |
+
"grad_norm": 0.3247200548648834,
|
| 1983 |
+
"learning_rate": 0.00011615736036916549,
|
| 1984 |
+
"loss": 0.8268,
|
| 1985 |
+
"step": 280
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 0.4990011098779134,
|
| 1989 |
+
"grad_norm": 0.3077162504196167,
|
| 1990 |
+
"learning_rate": 0.00011555271142444433,
|
| 1991 |
+
"loss": 0.7786,
|
| 1992 |
+
"step": 281
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.5007769145394007,
|
| 1996 |
+
"grad_norm": 0.3300260603427887,
|
| 1997 |
+
"learning_rate": 0.00011494747920954545,
|
| 1998 |
+
"loss": 0.7853,
|
| 1999 |
+
"step": 282
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 0.5007769145394007,
|
| 2003 |
+
"eval_loss": 0.7658749222755432,
|
| 2004 |
+
"eval_runtime": 158.4653,
|
| 2005 |
+
"eval_samples_per_second": 5.989,
|
| 2006 |
+
"eval_steps_per_second": 1.502,
|
| 2007 |
+
"step": 282
|
| 2008 |
}
|
| 2009 |
],
|
| 2010 |
"logging_steps": 1,
|
|
|
|
| 2024 |
"attributes": {}
|
| 2025 |
}
|
| 2026 |
},
|
| 2027 |
+
"total_flos": 3.7141566324945715e+17,
|
| 2028 |
"train_batch_size": 2,
|
| 2029 |
"trial_name": null,
|
| 2030 |
"trial_params": null
|