Training in progress, step 574, 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 104902272
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb92f5b93676b029d303892df387ebc6a8cd2beff92bf20ac5dcafd576ebaac7
|
| 3 |
size 104902272
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 53623316
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cf3ad1c6ae60fbd9236c2f7f7c62f4758ebfcc4fa40f384c1130b5ac9cae8456
|
| 3 |
size 53623316
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87bf700d5748b901a6308976958c2eff6608d62035f65196eb6e761fba71ccf9
|
| 3 |
size 14244
|
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:dd38dfbaa942f56b37cf6d5f7d1bc4c135ef8e7cb70adeb75a379ad7a9c2294f
|
| 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":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3031,6 +3031,1008 @@
|
|
| 3031 |
"learning_rate": 1.4595893198547888e-05,
|
| 3032 |
"loss": 6.0023,
|
| 3033 |
"step": 432
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3034 |
}
|
| 3035 |
],
|
| 3036 |
"logging_steps": 1,
|
|
@@ -3045,12 +4047,12 @@
|
|
| 3045 |
"should_evaluate": false,
|
| 3046 |
"should_log": false,
|
| 3047 |
"should_save": true,
|
| 3048 |
-
"should_training_stop":
|
| 3049 |
},
|
| 3050 |
"attributes": {}
|
| 3051 |
}
|
| 3052 |
},
|
| 3053 |
-
"total_flos":
|
| 3054 |
"train_batch_size": 4,
|
| 3055 |
"trial_name": null,
|
| 3056 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0004357298474946,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 574,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3031 |
"learning_rate": 1.4595893198547888e-05,
|
| 3032 |
"loss": 6.0023,
|
| 3033 |
"step": 432
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.7546840958605664,
|
| 3037 |
+
"grad_norm": 3.0819337368011475,
|
| 3038 |
+
"learning_rate": 1.4401497698974876e-05,
|
| 3039 |
+
"loss": 6.2708,
|
| 3040 |
+
"step": 433
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 0.7564270152505447,
|
| 3044 |
+
"grad_norm": 3.0940604209899902,
|
| 3045 |
+
"learning_rate": 1.4208187389534256e-05,
|
| 3046 |
+
"loss": 5.8773,
|
| 3047 |
+
"step": 434
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 0.7581699346405228,
|
| 3051 |
+
"grad_norm": 2.979501247406006,
|
| 3052 |
+
"learning_rate": 1.401596816312673e-05,
|
| 3053 |
+
"loss": 5.4585,
|
| 3054 |
+
"step": 435
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 0.7599128540305011,
|
| 3058 |
+
"grad_norm": 3.0984838008880615,
|
| 3059 |
+
"learning_rate": 1.3824845879392301e-05,
|
| 3060 |
+
"loss": 5.6387,
|
| 3061 |
+
"step": 436
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 0.7616557734204793,
|
| 3065 |
+
"grad_norm": 3.110401153564453,
|
| 3066 |
+
"learning_rate": 1.3634826364531616e-05,
|
| 3067 |
+
"loss": 5.7843,
|
| 3068 |
+
"step": 437
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 0.7633986928104575,
|
| 3072 |
+
"grad_norm": 3.3461947441101074,
|
| 3073 |
+
"learning_rate": 1.3445915411128295e-05,
|
| 3074 |
+
"loss": 6.2311,
|
| 3075 |
+
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 0.7651416122004358,
|
| 3079 |
+
"grad_norm": 3.493839740753174,
|
| 3080 |
+
"learning_rate": 1.325811877797245e-05,
|
| 3081 |
+
"loss": 5.9396,
|
| 3082 |
+
"step": 439
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.7668845315904139,
|
| 3086 |
+
"grad_norm": 3.3828165531158447,
|
| 3087 |
+
"learning_rate": 1.307144218988507e-05,
|
| 3088 |
+
"loss": 5.4118,
|
| 3089 |
+
"step": 440
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.7686274509803922,
|
| 3093 |
+
"grad_norm": 3.6239407062530518,
|
| 3094 |
+
"learning_rate": 1.2885891337543538e-05,
|
| 3095 |
+
"loss": 6.2809,
|
| 3096 |
+
"step": 441
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.7703703703703704,
|
| 3100 |
+
"grad_norm": 3.926117420196533,
|
| 3101 |
+
"learning_rate": 1.2701471877308091e-05,
|
| 3102 |
+
"loss": 6.1078,
|
| 3103 |
+
"step": 442
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 0.7721132897603485,
|
| 3107 |
+
"grad_norm": 3.8363847732543945,
|
| 3108 |
+
"learning_rate": 1.2518189431049498e-05,
|
| 3109 |
+
"loss": 5.9138,
|
| 3110 |
+
"step": 443
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 0.7738562091503268,
|
| 3114 |
+
"grad_norm": 3.887464761734009,
|
| 3115 |
+
"learning_rate": 1.2336049585977567e-05,
|
| 3116 |
+
"loss": 5.7176,
|
| 3117 |
+
"step": 444
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 0.775599128540305,
|
| 3121 |
+
"grad_norm": 4.025938034057617,
|
| 3122 |
+
"learning_rate": 1.2155057894470928e-05,
|
| 3123 |
+
"loss": 6.1095,
|
| 3124 |
+
"step": 445
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.7773420479302833,
|
| 3128 |
+
"grad_norm": 4.677386283874512,
|
| 3129 |
+
"learning_rate": 1.1975219873907679e-05,
|
| 3130 |
+
"loss": 6.3295,
|
| 3131 |
+
"step": 446
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 0.7790849673202614,
|
| 3135 |
+
"grad_norm": 4.534229278564453,
|
| 3136 |
+
"learning_rate": 1.179654100649727e-05,
|
| 3137 |
+
"loss": 5.6338,
|
| 3138 |
+
"step": 447
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 0.7808278867102396,
|
| 3142 |
+
"grad_norm": 4.835741996765137,
|
| 3143 |
+
"learning_rate": 1.1619026739113331e-05,
|
| 3144 |
+
"loss": 5.5942,
|
| 3145 |
+
"step": 448
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.7825708061002179,
|
| 3149 |
+
"grad_norm": 5.928588390350342,
|
| 3150 |
+
"learning_rate": 1.1442682483127687e-05,
|
| 3151 |
+
"loss": 6.0807,
|
| 3152 |
+
"step": 449
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 0.7843137254901961,
|
| 3156 |
+
"grad_norm": 7.3419108390808105,
|
| 3157 |
+
"learning_rate": 1.126751361424529e-05,
|
| 3158 |
+
"loss": 6.658,
|
| 3159 |
+
"step": 450
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.7860566448801742,
|
| 3163 |
+
"grad_norm": 3.0041606426239014,
|
| 3164 |
+
"learning_rate": 1.109352547234047e-05,
|
| 3165 |
+
"loss": 6.5957,
|
| 3166 |
+
"step": 451
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.7877995642701525,
|
| 3170 |
+
"grad_norm": 2.815629720687866,
|
| 3171 |
+
"learning_rate": 1.0920723361294099e-05,
|
| 3172 |
+
"loss": 6.3961,
|
| 3173 |
+
"step": 452
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 0.7895424836601307,
|
| 3177 |
+
"grad_norm": 2.712938070297241,
|
| 3178 |
+
"learning_rate": 1.0749112548831847e-05,
|
| 3179 |
+
"loss": 6.1991,
|
| 3180 |
+
"step": 453
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.791285403050109,
|
| 3184 |
+
"grad_norm": 2.4517555236816406,
|
| 3185 |
+
"learning_rate": 1.0578698266363734e-05,
|
| 3186 |
+
"loss": 6.0921,
|
| 3187 |
+
"step": 454
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 0.7930283224400871,
|
| 3191 |
+
"grad_norm": 2.6747450828552246,
|
| 3192 |
+
"learning_rate": 1.0409485708824507e-05,
|
| 3193 |
+
"loss": 6.284,
|
| 3194 |
+
"step": 455
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 0.7947712418300653,
|
| 3198 |
+
"grad_norm": 2.5269649028778076,
|
| 3199 |
+
"learning_rate": 1.0241480034515405e-05,
|
| 3200 |
+
"loss": 5.9745,
|
| 3201 |
+
"step": 456
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.7965141612200436,
|
| 3205 |
+
"grad_norm": 2.4382894039154053,
|
| 3206 |
+
"learning_rate": 1.0074686364946822e-05,
|
| 3207 |
+
"loss": 6.5809,
|
| 3208 |
+
"step": 457
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.7982570806100218,
|
| 3212 |
+
"grad_norm": 2.4671292304992676,
|
| 3213 |
+
"learning_rate": 9.909109784682241e-06,
|
| 3214 |
+
"loss": 6.4672,
|
| 3215 |
+
"step": 458
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.8,
|
| 3219 |
+
"grad_norm": 2.292046070098877,
|
| 3220 |
+
"learning_rate": 9.744755341183181e-06,
|
| 3221 |
+
"loss": 5.8945,
|
| 3222 |
+
"step": 459
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.8017429193899782,
|
| 3226 |
+
"grad_norm": 2.535379648208618,
|
| 3227 |
+
"learning_rate": 9.581628044655394e-06,
|
| 3228 |
+
"loss": 5.7448,
|
| 3229 |
+
"step": 460
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.8034858387799564,
|
| 3233 |
+
"grad_norm": 2.3226754665374756,
|
| 3234 |
+
"learning_rate": 9.419732867896048e-06,
|
| 3235 |
+
"loss": 6.3243,
|
| 3236 |
+
"step": 461
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.8052287581699347,
|
| 3240 |
+
"grad_norm": 2.4594292640686035,
|
| 3241 |
+
"learning_rate": 9.259074746142238e-06,
|
| 3242 |
+
"loss": 6.1204,
|
| 3243 |
+
"step": 462
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.8069716775599128,
|
| 3247 |
+
"grad_norm": 2.2522027492523193,
|
| 3248 |
+
"learning_rate": 9.099658576920466e-06,
|
| 3249 |
+
"loss": 5.5731,
|
| 3250 |
+
"step": 463
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.8087145969498911,
|
| 3254 |
+
"grad_norm": 2.384915590286255,
|
| 3255 |
+
"learning_rate": 8.941489219897353e-06,
|
| 3256 |
+
"loss": 6.2491,
|
| 3257 |
+
"step": 464
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.8104575163398693,
|
| 3261 |
+
"grad_norm": 2.3492512702941895,
|
| 3262 |
+
"learning_rate": 8.78457149673152e-06,
|
| 3263 |
+
"loss": 5.8509,
|
| 3264 |
+
"step": 465
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.8122004357298475,
|
| 3268 |
+
"grad_norm": 2.289766788482666,
|
| 3269 |
+
"learning_rate": 8.628910190926599e-06,
|
| 3270 |
+
"loss": 5.5891,
|
| 3271 |
+
"step": 466
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.8139433551198257,
|
| 3275 |
+
"grad_norm": 2.4919962882995605,
|
| 3276 |
+
"learning_rate": 8.474510047685408e-06,
|
| 3277 |
+
"loss": 6.4247,
|
| 3278 |
+
"step": 467
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.8156862745098039,
|
| 3282 |
+
"grad_norm": 2.338817596435547,
|
| 3283 |
+
"learning_rate": 8.321375773765256e-06,
|
| 3284 |
+
"loss": 6.1883,
|
| 3285 |
+
"step": 468
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.8174291938997821,
|
| 3289 |
+
"grad_norm": 2.4002230167388916,
|
| 3290 |
+
"learning_rate": 8.169512037334554e-06,
|
| 3291 |
+
"loss": 5.7954,
|
| 3292 |
+
"step": 469
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 0.8191721132897604,
|
| 3296 |
+
"grad_norm": 2.4609017372131348,
|
| 3297 |
+
"learning_rate": 8.018923467830403e-06,
|
| 3298 |
+
"loss": 5.9777,
|
| 3299 |
+
"step": 470
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 0.8209150326797385,
|
| 3303 |
+
"grad_norm": 2.534255027770996,
|
| 3304 |
+
"learning_rate": 7.869614655817575e-06,
|
| 3305 |
+
"loss": 6.098,
|
| 3306 |
+
"step": 471
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 0.8226579520697168,
|
| 3310 |
+
"grad_norm": 2.5172669887542725,
|
| 3311 |
+
"learning_rate": 7.721590152848474e-06,
|
| 3312 |
+
"loss": 5.4329,
|
| 3313 |
+
"step": 472
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.824400871459695,
|
| 3317 |
+
"grad_norm": 2.6065902709960938,
|
| 3318 |
+
"learning_rate": 7.574854471324461e-06,
|
| 3319 |
+
"loss": 5.8089,
|
| 3320 |
+
"step": 473
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 0.8261437908496732,
|
| 3324 |
+
"grad_norm": 2.8059046268463135,
|
| 3325 |
+
"learning_rate": 7.429412084358262e-06,
|
| 3326 |
+
"loss": 6.0739,
|
| 3327 |
+
"step": 474
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.8278867102396514,
|
| 3331 |
+
"grad_norm": 2.6289923191070557,
|
| 3332 |
+
"learning_rate": 7.28526742563762e-06,
|
| 3333 |
+
"loss": 6.2158,
|
| 3334 |
+
"step": 475
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 0.8296296296296296,
|
| 3338 |
+
"grad_norm": 2.877394676208496,
|
| 3339 |
+
"learning_rate": 7.142424889290139e-06,
|
| 3340 |
+
"loss": 5.5858,
|
| 3341 |
+
"step": 476
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 0.8313725490196079,
|
| 3345 |
+
"grad_norm": 2.82645583152771,
|
| 3346 |
+
"learning_rate": 7.0008888297492915e-06,
|
| 3347 |
+
"loss": 6.5205,
|
| 3348 |
+
"step": 477
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 0.8331154684095861,
|
| 3352 |
+
"grad_norm": 2.6874711513519287,
|
| 3353 |
+
"learning_rate": 6.860663561621766e-06,
|
| 3354 |
+
"loss": 6.1677,
|
| 3355 |
+
"step": 478
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.8348583877995642,
|
| 3359 |
+
"grad_norm": 2.713819980621338,
|
| 3360 |
+
"learning_rate": 6.721753359555833e-06,
|
| 3361 |
+
"loss": 5.9635,
|
| 3362 |
+
"step": 479
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.8366013071895425,
|
| 3366 |
+
"grad_norm": 2.8522517681121826,
|
| 3367 |
+
"learning_rate": 6.584162458111148e-06,
|
| 3368 |
+
"loss": 5.9607,
|
| 3369 |
+
"step": 480
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.8383442265795207,
|
| 3373 |
+
"grad_norm": 3.018490791320801,
|
| 3374 |
+
"learning_rate": 6.447895051629549e-06,
|
| 3375 |
+
"loss": 5.7084,
|
| 3376 |
+
"step": 481
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.840087145969499,
|
| 3380 |
+
"grad_norm": 2.8867599964141846,
|
| 3381 |
+
"learning_rate": 6.312955294107303e-06,
|
| 3382 |
+
"loss": 5.7403,
|
| 3383 |
+
"step": 482
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.8418300653594771,
|
| 3387 |
+
"grad_norm": 3.0370161533355713,
|
| 3388 |
+
"learning_rate": 6.179347299068378e-06,
|
| 3389 |
+
"loss": 5.7763,
|
| 3390 |
+
"step": 483
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.8435729847494553,
|
| 3394 |
+
"grad_norm": 2.995213270187378,
|
| 3395 |
+
"learning_rate": 6.047075139439151e-06,
|
| 3396 |
+
"loss": 5.5223,
|
| 3397 |
+
"step": 484
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.8453159041394336,
|
| 3401 |
+
"grad_norm": 3.0914831161499023,
|
| 3402 |
+
"learning_rate": 5.916142847424127e-06,
|
| 3403 |
+
"loss": 5.3967,
|
| 3404 |
+
"step": 485
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.8470588235294118,
|
| 3408 |
+
"grad_norm": 3.2199859619140625,
|
| 3409 |
+
"learning_rate": 5.786554414383127e-06,
|
| 3410 |
+
"loss": 6.3307,
|
| 3411 |
+
"step": 486
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.8488017429193899,
|
| 3415 |
+
"grad_norm": 3.375204086303711,
|
| 3416 |
+
"learning_rate": 5.658313790709569e-06,
|
| 3417 |
+
"loss": 5.8462,
|
| 3418 |
+
"step": 487
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.8505446623093682,
|
| 3422 |
+
"grad_norm": 3.433002471923828,
|
| 3423 |
+
"learning_rate": 5.531424885710012e-06,
|
| 3424 |
+
"loss": 6.1894,
|
| 3425 |
+
"step": 488
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.8522875816993464,
|
| 3429 |
+
"grad_norm": 3.8351926803588867,
|
| 3430 |
+
"learning_rate": 5.405891567485066e-06,
|
| 3431 |
+
"loss": 6.4691,
|
| 3432 |
+
"step": 489
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.8540305010893247,
|
| 3436 |
+
"grad_norm": 3.6944942474365234,
|
| 3437 |
+
"learning_rate": 5.281717662811381e-06,
|
| 3438 |
+
"loss": 5.8172,
|
| 3439 |
+
"step": 490
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.8557734204793028,
|
| 3443 |
+
"grad_norm": 4.206511974334717,
|
| 3444 |
+
"learning_rate": 5.158906957025078e-06,
|
| 3445 |
+
"loss": 6.2588,
|
| 3446 |
+
"step": 491
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.857516339869281,
|
| 3450 |
+
"grad_norm": 3.974405527114868,
|
| 3451 |
+
"learning_rate": 5.037463193906294e-06,
|
| 3452 |
+
"loss": 6.1301,
|
| 3453 |
+
"step": 492
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.8592592592592593,
|
| 3457 |
+
"grad_norm": 4.069376468658447,
|
| 3458 |
+
"learning_rate": 4.917390075565098e-06,
|
| 3459 |
+
"loss": 5.5271,
|
| 3460 |
+
"step": 493
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.8610021786492374,
|
| 3464 |
+
"grad_norm": 4.004759311676025,
|
| 3465 |
+
"learning_rate": 4.798691262328586e-06,
|
| 3466 |
+
"loss": 5.9545,
|
| 3467 |
+
"step": 494
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.8627450980392157,
|
| 3471 |
+
"grad_norm": 4.0203938484191895,
|
| 3472 |
+
"learning_rate": 4.681370372629368e-06,
|
| 3473 |
+
"loss": 5.696,
|
| 3474 |
+
"step": 495
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.8644880174291939,
|
| 3478 |
+
"grad_norm": 4.475993633270264,
|
| 3479 |
+
"learning_rate": 4.565430982895174e-06,
|
| 3480 |
+
"loss": 5.5264,
|
| 3481 |
+
"step": 496
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.8662309368191721,
|
| 3485 |
+
"grad_norm": 5.309076309204102,
|
| 3486 |
+
"learning_rate": 4.450876627439943e-06,
|
| 3487 |
+
"loss": 5.6954,
|
| 3488 |
+
"step": 497
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.8679738562091504,
|
| 3492 |
+
"grad_norm": 5.302606105804443,
|
| 3493 |
+
"learning_rate": 4.337710798355959e-06,
|
| 3494 |
+
"loss": 6.1407,
|
| 3495 |
+
"step": 498
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.8697167755991285,
|
| 3499 |
+
"grad_norm": 5.7808732986450195,
|
| 3500 |
+
"learning_rate": 4.225936945407494e-06,
|
| 3501 |
+
"loss": 6.0457,
|
| 3502 |
+
"step": 499
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.8714596949891068,
|
| 3506 |
+
"grad_norm": 8.515405654907227,
|
| 3507 |
+
"learning_rate": 4.1155584759256015e-06,
|
| 3508 |
+
"loss": 7.0914,
|
| 3509 |
+
"step": 500
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 0.873202614379085,
|
| 3513 |
+
"grad_norm": 2.5194711685180664,
|
| 3514 |
+
"learning_rate": 4.006578754704254e-06,
|
| 3515 |
+
"loss": 6.4179,
|
| 3516 |
+
"step": 501
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 0.8749455337690631,
|
| 3520 |
+
"grad_norm": 2.5622870922088623,
|
| 3521 |
+
"learning_rate": 3.89900110389777e-06,
|
| 3522 |
+
"loss": 6.1856,
|
| 3523 |
+
"step": 502
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 0.8766884531590414,
|
| 3527 |
+
"grad_norm": 2.5136539936065674,
|
| 3528 |
+
"learning_rate": 3.7928288029195292e-06,
|
| 3529 |
+
"loss": 6.4684,
|
| 3530 |
+
"step": 503
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.8784313725490196,
|
| 3534 |
+
"grad_norm": 2.6357421875,
|
| 3535 |
+
"learning_rate": 3.688065088342041e-06,
|
| 3536 |
+
"loss": 6.4146,
|
| 3537 |
+
"step": 504
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.8801742919389978,
|
| 3541 |
+
"grad_norm": 2.5983378887176514,
|
| 3542 |
+
"learning_rate": 3.5847131537982137e-06,
|
| 3543 |
+
"loss": 6.4045,
|
| 3544 |
+
"step": 505
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 0.881917211328976,
|
| 3548 |
+
"grad_norm": 2.259760856628418,
|
| 3549 |
+
"learning_rate": 3.482776149884093e-06,
|
| 3550 |
+
"loss": 6.3314,
|
| 3551 |
+
"step": 506
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 0.8836601307189542,
|
| 3555 |
+
"grad_norm": 2.2606775760650635,
|
| 3556 |
+
"learning_rate": 3.382257184062709e-06,
|
| 3557 |
+
"loss": 5.8027,
|
| 3558 |
+
"step": 507
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 0.8854030501089325,
|
| 3562 |
+
"grad_norm": 2.304555892944336,
|
| 3563 |
+
"learning_rate": 3.283159320569451e-06,
|
| 3564 |
+
"loss": 6.1733,
|
| 3565 |
+
"step": 508
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 0.8871459694989107,
|
| 3569 |
+
"grad_norm": 2.253995895385742,
|
| 3570 |
+
"learning_rate": 3.1854855803185645e-06,
|
| 3571 |
+
"loss": 6.1169,
|
| 3572 |
+
"step": 509
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 0.8888888888888888,
|
| 3576 |
+
"grad_norm": 2.271144390106201,
|
| 3577 |
+
"learning_rate": 3.089238940811162e-06,
|
| 3578 |
+
"loss": 5.9557,
|
| 3579 |
+
"step": 510
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 0.8906318082788671,
|
| 3583 |
+
"grad_norm": 2.289597511291504,
|
| 3584 |
+
"learning_rate": 2.9944223360443445e-06,
|
| 3585 |
+
"loss": 6.1503,
|
| 3586 |
+
"step": 511
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 0.8923747276688453,
|
| 3590 |
+
"grad_norm": 2.4109086990356445,
|
| 3591 |
+
"learning_rate": 2.9010386564218617e-06,
|
| 3592 |
+
"loss": 6.4016,
|
| 3593 |
+
"step": 512
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.8941176470588236,
|
| 3597 |
+
"grad_norm": 2.2449753284454346,
|
| 3598 |
+
"learning_rate": 2.8090907486659325e-06,
|
| 3599 |
+
"loss": 6.0639,
|
| 3600 |
+
"step": 513
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 0.8958605664488017,
|
| 3604 |
+
"grad_norm": 2.3619470596313477,
|
| 3605 |
+
"learning_rate": 2.7185814157304814e-06,
|
| 3606 |
+
"loss": 5.8261,
|
| 3607 |
+
"step": 514
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 0.8976034858387799,
|
| 3611 |
+
"grad_norm": 2.487009286880493,
|
| 3612 |
+
"learning_rate": 2.6295134167157343e-06,
|
| 3613 |
+
"loss": 5.9042,
|
| 3614 |
+
"step": 515
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 0.8993464052287582,
|
| 3618 |
+
"grad_norm": 2.6147444248199463,
|
| 3619 |
+
"learning_rate": 2.5418894667840364e-06,
|
| 3620 |
+
"loss": 6.1961,
|
| 3621 |
+
"step": 516
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 0.9010893246187364,
|
| 3625 |
+
"grad_norm": 2.4794037342071533,
|
| 3626 |
+
"learning_rate": 2.455712237077157e-06,
|
| 3627 |
+
"loss": 5.7284,
|
| 3628 |
+
"step": 517
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 0.9028322440087146,
|
| 3632 |
+
"grad_norm": 2.5165326595306396,
|
| 3633 |
+
"learning_rate": 2.3709843546347864e-06,
|
| 3634 |
+
"loss": 6.0962,
|
| 3635 |
+
"step": 518
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 0.9045751633986928,
|
| 3639 |
+
"grad_norm": 2.541537046432495,
|
| 3640 |
+
"learning_rate": 2.287708402314542e-06,
|
| 3641 |
+
"loss": 6.0823,
|
| 3642 |
+
"step": 519
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 0.906318082788671,
|
| 3646 |
+
"grad_norm": 2.418776273727417,
|
| 3647 |
+
"learning_rate": 2.2058869187131515e-06,
|
| 3648 |
+
"loss": 5.8923,
|
| 3649 |
+
"step": 520
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.9080610021786493,
|
| 3653 |
+
"grad_norm": 2.547147750854492,
|
| 3654 |
+
"learning_rate": 2.1255223980891027e-06,
|
| 3655 |
+
"loss": 5.4706,
|
| 3656 |
+
"step": 521
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 0.9098039215686274,
|
| 3660 |
+
"grad_norm": 2.498025417327881,
|
| 3661 |
+
"learning_rate": 2.046617290286601e-06,
|
| 3662 |
+
"loss": 6.0115,
|
| 3663 |
+
"step": 522
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 0.9115468409586057,
|
| 3667 |
+
"grad_norm": 2.5290932655334473,
|
| 3668 |
+
"learning_rate": 1.969174000660906e-06,
|
| 3669 |
+
"loss": 5.7543,
|
| 3670 |
+
"step": 523
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 0.9132897603485839,
|
| 3674 |
+
"grad_norm": 2.6089823246002197,
|
| 3675 |
+
"learning_rate": 1.8931948900049789e-06,
|
| 3676 |
+
"loss": 5.7353,
|
| 3677 |
+
"step": 524
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 0.9150326797385621,
|
| 3681 |
+
"grad_norm": 2.6269378662109375,
|
| 3682 |
+
"learning_rate": 1.8186822744775234e-06,
|
| 3683 |
+
"loss": 5.8222,
|
| 3684 |
+
"step": 525
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 0.9167755991285403,
|
| 3688 |
+
"grad_norm": 2.752235174179077,
|
| 3689 |
+
"learning_rate": 1.7456384255323922e-06,
|
| 3690 |
+
"loss": 5.8682,
|
| 3691 |
+
"step": 526
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 0.9185185185185185,
|
| 3695 |
+
"grad_norm": 2.7636468410491943,
|
| 3696 |
+
"learning_rate": 1.6740655698493313e-06,
|
| 3697 |
+
"loss": 6.0393,
|
| 3698 |
+
"step": 527
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 0.9202614379084967,
|
| 3702 |
+
"grad_norm": 2.768242835998535,
|
| 3703 |
+
"learning_rate": 1.6039658892661181e-06,
|
| 3704 |
+
"loss": 6.1033,
|
| 3705 |
+
"step": 528
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.922004357298475,
|
| 3709 |
+
"grad_norm": 2.9885339736938477,
|
| 3710 |
+
"learning_rate": 1.535341520712008e-06,
|
| 3711 |
+
"loss": 5.7918,
|
| 3712 |
+
"step": 529
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 0.9237472766884531,
|
| 3716 |
+
"grad_norm": 2.9355344772338867,
|
| 3717 |
+
"learning_rate": 1.4681945561426547e-06,
|
| 3718 |
+
"loss": 5.739,
|
| 3719 |
+
"step": 530
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 0.9254901960784314,
|
| 3723 |
+
"grad_norm": 3.1985771656036377,
|
| 3724 |
+
"learning_rate": 1.402527042476276e-06,
|
| 3725 |
+
"loss": 5.9063,
|
| 3726 |
+
"step": 531
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 0.9272331154684096,
|
| 3730 |
+
"grad_norm": 3.049199342727661,
|
| 3731 |
+
"learning_rate": 1.3383409815313108e-06,
|
| 3732 |
+
"loss": 5.8753,
|
| 3733 |
+
"step": 532
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 0.9289760348583878,
|
| 3737 |
+
"grad_norm": 3.181462526321411,
|
| 3738 |
+
"learning_rate": 1.2756383299653452e-06,
|
| 3739 |
+
"loss": 6.3117,
|
| 3740 |
+
"step": 533
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 0.930718954248366,
|
| 3744 |
+
"grad_norm": 3.082789182662964,
|
| 3745 |
+
"learning_rate": 1.2144209992155043e-06,
|
| 3746 |
+
"loss": 5.4386,
|
| 3747 |
+
"step": 534
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 0.9324618736383442,
|
| 3751 |
+
"grad_norm": 3.15106201171875,
|
| 3752 |
+
"learning_rate": 1.1546908554401659e-06,
|
| 3753 |
+
"loss": 5.7718,
|
| 3754 |
+
"step": 535
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.9342047930283225,
|
| 3758 |
+
"grad_norm": 3.4478657245635986,
|
| 3759 |
+
"learning_rate": 1.0964497194620726e-06,
|
| 3760 |
+
"loss": 5.9466,
|
| 3761 |
+
"step": 536
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.9359477124183007,
|
| 3765 |
+
"grad_norm": 3.367365837097168,
|
| 3766 |
+
"learning_rate": 1.0396993667128318e-06,
|
| 3767 |
+
"loss": 6.1564,
|
| 3768 |
+
"step": 537
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.9376906318082788,
|
| 3772 |
+
"grad_norm": 3.604461908340454,
|
| 3773 |
+
"learning_rate": 9.84441527178781e-07,
|
| 3774 |
+
"loss": 6.2624,
|
| 3775 |
+
"step": 538
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.9394335511982571,
|
| 3779 |
+
"grad_norm": 3.753371000289917,
|
| 3780 |
+
"learning_rate": 9.306778853482745e-07,
|
| 3781 |
+
"loss": 6.473,
|
| 3782 |
+
"step": 539
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.9411764705882353,
|
| 3786 |
+
"grad_norm": 3.6445400714874268,
|
| 3787 |
+
"learning_rate": 8.784100801602912e-07,
|
| 3788 |
+
"loss": 5.285,
|
| 3789 |
+
"step": 540
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.9429193899782136,
|
| 3793 |
+
"grad_norm": 4.095015048980713,
|
| 3794 |
+
"learning_rate": 8.276397049545359e-07,
|
| 3795 |
+
"loss": 5.9943,
|
| 3796 |
+
"step": 541
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.9446623093681917,
|
| 3800 |
+
"grad_norm": 3.9936633110046387,
|
| 3801 |
+
"learning_rate": 7.783683074227943e-07,
|
| 3802 |
+
"loss": 5.6086,
|
| 3803 |
+
"step": 542
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.9464052287581699,
|
| 3807 |
+
"grad_norm": 4.380242347717285,
|
| 3808 |
+
"learning_rate": 7.305973895618157e-07,
|
| 3809 |
+
"loss": 5.7241,
|
| 3810 |
+
"step": 543
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.9481481481481482,
|
| 3814 |
+
"grad_norm": 4.292196750640869,
|
| 3815 |
+
"learning_rate": 6.84328407627477e-07,
|
| 3816 |
+
"loss": 5.2435,
|
| 3817 |
+
"step": 544
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.9498910675381264,
|
| 3821 |
+
"grad_norm": 4.3875346183776855,
|
| 3822 |
+
"learning_rate": 6.395627720904518e-07,
|
| 3823 |
+
"loss": 5.469,
|
| 3824 |
+
"step": 545
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.9516339869281045,
|
| 3828 |
+
"grad_norm": 4.875105857849121,
|
| 3829 |
+
"learning_rate": 5.963018475931281e-07,
|
| 3830 |
+
"loss": 6.1045,
|
| 3831 |
+
"step": 546
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.9533769063180828,
|
| 3835 |
+
"grad_norm": 5.184041500091553,
|
| 3836 |
+
"learning_rate": 5.54546952908086e-07,
|
| 3837 |
+
"loss": 5.6123,
|
| 3838 |
+
"step": 547
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 0.955119825708061,
|
| 3842 |
+
"grad_norm": 5.260396957397461,
|
| 3843 |
+
"learning_rate": 5.14299360897852e-07,
|
| 3844 |
+
"loss": 5.7216,
|
| 3845 |
+
"step": 548
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 0.9568627450980393,
|
| 3849 |
+
"grad_norm": 5.843105316162109,
|
| 3850 |
+
"learning_rate": 4.7556029847611317e-07,
|
| 3851 |
+
"loss": 5.7342,
|
| 3852 |
+
"step": 549
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 0.9586056644880174,
|
| 3856 |
+
"grad_norm": 8.344443321228027,
|
| 3857 |
+
"learning_rate": 4.383309465703145e-07,
|
| 3858 |
+
"loss": 6.5191,
|
| 3859 |
+
"step": 550
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 0.9603485838779956,
|
| 3863 |
+
"grad_norm": 2.1281239986419678,
|
| 3864 |
+
"learning_rate": 4.0261244008564794e-07,
|
| 3865 |
+
"loss": 5.9559,
|
| 3866 |
+
"step": 551
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 0.9620915032679739,
|
| 3870 |
+
"grad_norm": 2.288827896118164,
|
| 3871 |
+
"learning_rate": 3.684058678704805e-07,
|
| 3872 |
+
"loss": 6.0769,
|
| 3873 |
+
"step": 552
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.963834422657952,
|
| 3877 |
+
"grad_norm": 2.3264894485473633,
|
| 3878 |
+
"learning_rate": 3.357122726831252e-07,
|
| 3879 |
+
"loss": 6.2131,
|
| 3880 |
+
"step": 553
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 0.9655773420479303,
|
| 3884 |
+
"grad_norm": 2.297492027282715,
|
| 3885 |
+
"learning_rate": 3.045326511600999e-07,
|
| 3886 |
+
"loss": 6.7654,
|
| 3887 |
+
"step": 554
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 0.9673202614379085,
|
| 3891 |
+
"grad_norm": 2.309363842010498,
|
| 3892 |
+
"learning_rate": 2.748679537857013e-07,
|
| 3893 |
+
"loss": 6.0648,
|
| 3894 |
+
"step": 555
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 0.9690631808278867,
|
| 3898 |
+
"grad_norm": 2.221372365951538,
|
| 3899 |
+
"learning_rate": 2.4671908486305075e-07,
|
| 3900 |
+
"loss": 6.0033,
|
| 3901 |
+
"step": 556
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 0.970806100217865,
|
| 3905 |
+
"grad_norm": 2.2333123683929443,
|
| 3906 |
+
"learning_rate": 2.2008690248653262e-07,
|
| 3907 |
+
"loss": 5.9659,
|
| 3908 |
+
"step": 557
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 0.9725490196078431,
|
| 3912 |
+
"grad_norm": 2.4658637046813965,
|
| 3913 |
+
"learning_rate": 1.9497221851562663e-07,
|
| 3914 |
+
"loss": 6.1484,
|
| 3915 |
+
"step": 558
|
| 3916 |
+
},
|
| 3917 |
+
{
|
| 3918 |
+
"epoch": 0.9742919389978214,
|
| 3919 |
+
"grad_norm": 2.4412317276000977,
|
| 3920 |
+
"learning_rate": 1.7137579855016073e-07,
|
| 3921 |
+
"loss": 5.6514,
|
| 3922 |
+
"step": 559
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 0.9760348583877996,
|
| 3926 |
+
"grad_norm": 2.5595967769622803,
|
| 3927 |
+
"learning_rate": 1.4929836190696323e-07,
|
| 3928 |
+
"loss": 6.1126,
|
| 3929 |
+
"step": 560
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 0.9777777777777777,
|
| 3933 |
+
"grad_norm": 2.9756367206573486,
|
| 3934 |
+
"learning_rate": 1.2874058159796364e-07,
|
| 3935 |
+
"loss": 6.9962,
|
| 3936 |
+
"step": 561
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 0.979520697167756,
|
| 3940 |
+
"grad_norm": 2.626969814300537,
|
| 3941 |
+
"learning_rate": 1.0970308430964803e-07,
|
| 3942 |
+
"loss": 6.007,
|
| 3943 |
+
"step": 562
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 0.9812636165577342,
|
| 3947 |
+
"grad_norm": 2.7828261852264404,
|
| 3948 |
+
"learning_rate": 9.218645038396311e-08,
|
| 3949 |
+
"loss": 5.8444,
|
| 3950 |
+
"step": 563
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 0.9830065359477124,
|
| 3954 |
+
"grad_norm": 3.111481189727783,
|
| 3955 |
+
"learning_rate": 7.619121380063043e-08,
|
| 3956 |
+
"loss": 6.6361,
|
| 3957 |
+
"step": 564
|
| 3958 |
+
},
|
| 3959 |
+
{
|
| 3960 |
+
"epoch": 0.9847494553376906,
|
| 3961 |
+
"grad_norm": 2.865208625793457,
|
| 3962 |
+
"learning_rate": 6.171786216085939e-08,
|
| 3963 |
+
"loss": 5.9192,
|
| 3964 |
+
"step": 565
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 0.9864923747276688,
|
| 3968 |
+
"grad_norm": 3.054203510284424,
|
| 3969 |
+
"learning_rate": 4.876683667249804e-08,
|
| 3970 |
+
"loss": 6.1105,
|
| 3971 |
+
"step": 566
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 0.9882352941176471,
|
| 3975 |
+
"grad_norm": 3.3111000061035156,
|
| 3976 |
+
"learning_rate": 3.733853213656047e-08,
|
| 3977 |
+
"loss": 5.8512,
|
| 3978 |
+
"step": 567
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 0.9899782135076253,
|
| 3982 |
+
"grad_norm": 3.1707890033721924,
|
| 3983 |
+
"learning_rate": 2.7433296935219787e-08,
|
| 3984 |
+
"loss": 5.2856,
|
| 3985 |
+
"step": 568
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 0.9917211328976034,
|
| 3989 |
+
"grad_norm": 3.5281925201416016,
|
| 3990 |
+
"learning_rate": 1.905143302116108e-08,
|
| 3991 |
+
"loss": 5.6356,
|
| 3992 |
+
"step": 569
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 0.9934640522875817,
|
| 3996 |
+
"grad_norm": 3.746992588043213,
|
| 3997 |
+
"learning_rate": 1.2193195908388743e-08,
|
| 3998 |
+
"loss": 6.2382,
|
| 3999 |
+
"step": 570
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 0.9952069716775599,
|
| 4003 |
+
"grad_norm": 4.109386444091797,
|
| 4004 |
+
"learning_rate": 6.8587946644493864e-09,
|
| 4005 |
+
"loss": 5.4891,
|
| 4006 |
+
"step": 571
|
| 4007 |
+
},
|
| 4008 |
+
{
|
| 4009 |
+
"epoch": 0.9969498910675382,
|
| 4010 |
+
"grad_norm": 4.418205261230469,
|
| 4011 |
+
"learning_rate": 3.048391904031389e-09,
|
| 4012 |
+
"loss": 5.5848,
|
| 4013 |
+
"step": 572
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 0.9986928104575163,
|
| 4017 |
+
"grad_norm": 4.8515520095825195,
|
| 4018 |
+
"learning_rate": 7.621037840299661e-10,
|
| 4019 |
+
"loss": 6.073,
|
| 4020 |
+
"step": 573
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 0.9986928104575163,
|
| 4024 |
+
"eval_loss": 1.503620982170105,
|
| 4025 |
+
"eval_runtime": 41.0454,
|
| 4026 |
+
"eval_samples_per_second": 11.792,
|
| 4027 |
+
"eval_steps_per_second": 2.948,
|
| 4028 |
+
"step": 573
|
| 4029 |
+
},
|
| 4030 |
+
{
|
| 4031 |
+
"epoch": 1.0004357298474946,
|
| 4032 |
+
"grad_norm": 5.828014850616455,
|
| 4033 |
+
"learning_rate": 0.0,
|
| 4034 |
+
"loss": 6.2816,
|
| 4035 |
+
"step": 574
|
| 4036 |
}
|
| 4037 |
],
|
| 4038 |
"logging_steps": 1,
|
|
|
|
| 4047 |
"should_evaluate": false,
|
| 4048 |
"should_log": false,
|
| 4049 |
"should_save": true,
|
| 4050 |
+
"should_training_stop": true
|
| 4051 |
},
|
| 4052 |
"attributes": {}
|
| 4053 |
}
|
| 4054 |
},
|
| 4055 |
+
"total_flos": 8.344500518151782e+17,
|
| 4056 |
"train_batch_size": 4,
|
| 4057 |
"trial_name": null,
|
| 4058 |
"trial_params": null
|