Training in progress, step 553, 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 125918320
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b677e4bf5c8e6273a99bda7d3d2c0afce4b991f97e4ce838664059bb222700c6
|
| 3 |
size 125918320
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 64684244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5fea8135af20d4c63d861a4c8774fe66918b6f8f8d4020335bf9dae0a8919ad8
|
| 3 |
size 64684244
|
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 sha256:74efbaaf0803a77ee24332433c97b4c4efa9d7113b5aa6975ff10a0c932be806
|
| 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:b4b0627147991988ca81f95580a67cfcca6614bdcbd8a39f73ebb79a57cc0418
|
| 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,
|
|
@@ -2926,6 +2926,966 @@
|
|
| 2926 |
"learning_rate": 1.444255676716637e-05,
|
| 2927 |
"loss": 3.0206,
|
| 2928 |
"step": 417
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2929 |
}
|
| 2930 |
],
|
| 2931 |
"logging_steps": 1,
|
|
@@ -2940,12 +3900,12 @@
|
|
| 2940 |
"should_evaluate": false,
|
| 2941 |
"should_log": false,
|
| 2942 |
"should_save": true,
|
| 2943 |
-
"should_training_stop":
|
| 2944 |
},
|
| 2945 |
"attributes": {}
|
| 2946 |
}
|
| 2947 |
},
|
| 2948 |
-
"total_flos":
|
| 2949 |
"train_batch_size": 4,
|
| 2950 |
"trial_name": null,
|
| 2951 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0009045680687472,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 553,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2926 |
"learning_rate": 1.444255676716637e-05,
|
| 2927 |
"loss": 3.0206,
|
| 2928 |
"step": 417
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 0.7562189054726368,
|
| 2932 |
+
"grad_norm": 6.555543422698975,
|
| 2933 |
+
"learning_rate": 1.424162119926336e-05,
|
| 2934 |
+
"loss": 3.9264,
|
| 2935 |
+
"step": 418
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 0.7580280416101312,
|
| 2939 |
+
"grad_norm": 6.464334487915039,
|
| 2940 |
+
"learning_rate": 1.40418608407689e-05,
|
| 2941 |
+
"loss": 3.7147,
|
| 2942 |
+
"step": 419
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 0.7598371777476255,
|
| 2946 |
+
"grad_norm": 6.810556411743164,
|
| 2947 |
+
"learning_rate": 1.3843282256864599e-05,
|
| 2948 |
+
"loss": 3.7548,
|
| 2949 |
+
"step": 420
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 0.7616463138851198,
|
| 2953 |
+
"grad_norm": 6.693656921386719,
|
| 2954 |
+
"learning_rate": 1.3645891973892772e-05,
|
| 2955 |
+
"loss": 3.8381,
|
| 2956 |
+
"step": 421
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 0.7634554500226142,
|
| 2960 |
+
"grad_norm": 6.932521820068359,
|
| 2961 |
+
"learning_rate": 1.3449696479141854e-05,
|
| 2962 |
+
"loss": 4.0131,
|
| 2963 |
+
"step": 422
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 0.7652645861601085,
|
| 2967 |
+
"grad_norm": 6.466965675354004,
|
| 2968 |
+
"learning_rate": 1.325470222063327e-05,
|
| 2969 |
+
"loss": 3.4764,
|
| 2970 |
+
"step": 423
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 0.7670737222976028,
|
| 2974 |
+
"grad_norm": 7.379868507385254,
|
| 2975 |
+
"learning_rate": 1.3060915606909413e-05,
|
| 2976 |
+
"loss": 3.9783,
|
| 2977 |
+
"step": 424
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 0.7688828584350973,
|
| 2981 |
+
"grad_norm": 7.2505083084106445,
|
| 2982 |
+
"learning_rate": 1.2868343006823114e-05,
|
| 2983 |
+
"loss": 3.2758,
|
| 2984 |
+
"step": 425
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 0.7706919945725916,
|
| 2988 |
+
"grad_norm": 6.697902202606201,
|
| 2989 |
+
"learning_rate": 1.2676990749328254e-05,
|
| 2990 |
+
"loss": 3.3263,
|
| 2991 |
+
"step": 426
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 0.772501130710086,
|
| 2995 |
+
"grad_norm": 7.053414821624756,
|
| 2996 |
+
"learning_rate": 1.2486865123271868e-05,
|
| 2997 |
+
"loss": 3.518,
|
| 2998 |
+
"step": 427
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 0.7743102668475803,
|
| 3002 |
+
"grad_norm": 7.461160182952881,
|
| 3003 |
+
"learning_rate": 1.2297972377187361e-05,
|
| 3004 |
+
"loss": 3.8753,
|
| 3005 |
+
"step": 428
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 0.7761194029850746,
|
| 3009 |
+
"grad_norm": 7.320030212402344,
|
| 3010 |
+
"learning_rate": 1.2110318719089158e-05,
|
| 3011 |
+
"loss": 3.9377,
|
| 3012 |
+
"step": 429
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 0.777928539122569,
|
| 3016 |
+
"grad_norm": 6.707526683807373,
|
| 3017 |
+
"learning_rate": 1.1923910316268782e-05,
|
| 3018 |
+
"loss": 3.577,
|
| 3019 |
+
"step": 430
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 0.7797376752600633,
|
| 3023 |
+
"grad_norm": 7.633321285247803,
|
| 3024 |
+
"learning_rate": 1.1738753295091986e-05,
|
| 3025 |
+
"loss": 3.8751,
|
| 3026 |
+
"step": 431
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 0.7815468113975577,
|
| 3030 |
+
"grad_norm": 7.908304691314697,
|
| 3031 |
+
"learning_rate": 1.1554853740797555e-05,
|
| 3032 |
+
"loss": 3.6976,
|
| 3033 |
+
"step": 432
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.783355947535052,
|
| 3037 |
+
"grad_norm": 7.982555866241455,
|
| 3038 |
+
"learning_rate": 1.1372217697297249e-05,
|
| 3039 |
+
"loss": 3.6444,
|
| 3040 |
+
"step": 433
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 0.7851650836725463,
|
| 3044 |
+
"grad_norm": 7.220830917358398,
|
| 3045 |
+
"learning_rate": 1.1190851166977217e-05,
|
| 3046 |
+
"loss": 3.5019,
|
| 3047 |
+
"step": 434
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 0.7869742198100407,
|
| 3051 |
+
"grad_norm": 6.941662311553955,
|
| 3052 |
+
"learning_rate": 1.101076011050065e-05,
|
| 3053 |
+
"loss": 3.5344,
|
| 3054 |
+
"step": 435
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 0.788783355947535,
|
| 3058 |
+
"grad_norm": 6.613343715667725,
|
| 3059 |
+
"learning_rate": 1.0831950446611949e-05,
|
| 3060 |
+
"loss": 3.4078,
|
| 3061 |
+
"step": 436
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 0.7905924920850294,
|
| 3065 |
+
"grad_norm": 7.152531147003174,
|
| 3066 |
+
"learning_rate": 1.065442805194214e-05,
|
| 3067 |
+
"loss": 3.7873,
|
| 3068 |
+
"step": 437
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 0.7924016282225237,
|
| 3072 |
+
"grad_norm": 6.536508083343506,
|
| 3073 |
+
"learning_rate": 1.0478198760815832e-05,
|
| 3074 |
+
"loss": 2.7608,
|
| 3075 |
+
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 0.7942107643600181,
|
| 3079 |
+
"grad_norm": 6.4156904220581055,
|
| 3080 |
+
"learning_rate": 1.0303268365059382e-05,
|
| 3081 |
+
"loss": 2.9791,
|
| 3082 |
+
"step": 439
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.7960199004975125,
|
| 3086 |
+
"grad_norm": 5.869282245635986,
|
| 3087 |
+
"learning_rate": 1.0129642613810576e-05,
|
| 3088 |
+
"loss": 3.5902,
|
| 3089 |
+
"step": 440
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.7978290366350068,
|
| 3093 |
+
"grad_norm": 6.527062892913818,
|
| 3094 |
+
"learning_rate": 9.957327213329687e-06,
|
| 3095 |
+
"loss": 5.1202,
|
| 3096 |
+
"step": 441
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.7996381727725012,
|
| 3100 |
+
"grad_norm": 9.36404800415039,
|
| 3101 |
+
"learning_rate": 9.786327826811942e-06,
|
| 3102 |
+
"loss": 5.934,
|
| 3103 |
+
"step": 442
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 0.8014473089099955,
|
| 3107 |
+
"grad_norm": 8.085516929626465,
|
| 3108 |
+
"learning_rate": 9.616650074201383e-06,
|
| 3109 |
+
"loss": 5.8671,
|
| 3110 |
+
"step": 443
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 0.8032564450474898,
|
| 3114 |
+
"grad_norm": 8.88477611541748,
|
| 3115 |
+
"learning_rate": 9.448299532006149e-06,
|
| 3116 |
+
"loss": 6.4374,
|
| 3117 |
+
"step": 444
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 0.8050655811849842,
|
| 3121 |
+
"grad_norm": 10.586947441101074,
|
| 3122 |
+
"learning_rate": 9.281281733115288e-06,
|
| 3123 |
+
"loss": 5.8352,
|
| 3124 |
+
"step": 445
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.8068747173224785,
|
| 3128 |
+
"grad_norm": 9.525203704833984,
|
| 3129 |
+
"learning_rate": 9.115602166616805e-06,
|
| 3130 |
+
"loss": 5.6282,
|
| 3131 |
+
"step": 446
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 0.8086838534599728,
|
| 3135 |
+
"grad_norm": 8.886335372924805,
|
| 3136 |
+
"learning_rate": 8.951266277617326e-06,
|
| 3137 |
+
"loss": 4.3723,
|
| 3138 |
+
"step": 447
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 0.8104929895974672,
|
| 3142 |
+
"grad_norm": 9.162277221679688,
|
| 3143 |
+
"learning_rate": 8.78827946706311e-06,
|
| 3144 |
+
"loss": 4.5638,
|
| 3145 |
+
"step": 448
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.8123021257349615,
|
| 3149 |
+
"grad_norm": 11.042312622070312,
|
| 3150 |
+
"learning_rate": 8.626647091562612e-06,
|
| 3151 |
+
"loss": 5.3548,
|
| 3152 |
+
"step": 449
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 0.8141112618724559,
|
| 3156 |
+
"grad_norm": 15.882914543151855,
|
| 3157 |
+
"learning_rate": 8.466374463210346e-06,
|
| 3158 |
+
"loss": 4.3994,
|
| 3159 |
+
"step": 450
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.8159203980099502,
|
| 3163 |
+
"grad_norm": 6.315022945404053,
|
| 3164 |
+
"learning_rate": 8.307466849412366e-06,
|
| 3165 |
+
"loss": 3.6107,
|
| 3166 |
+
"step": 451
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.8177295341474446,
|
| 3170 |
+
"grad_norm": 6.3623175621032715,
|
| 3171 |
+
"learning_rate": 8.149929472713125e-06,
|
| 3172 |
+
"loss": 3.3937,
|
| 3173 |
+
"step": 452
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 0.819538670284939,
|
| 3177 |
+
"grad_norm": 8.390593528747559,
|
| 3178 |
+
"learning_rate": 7.993767510623834e-06,
|
| 3179 |
+
"loss": 4.0698,
|
| 3180 |
+
"step": 453
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.8213478064224333,
|
| 3184 |
+
"grad_norm": 8.719853401184082,
|
| 3185 |
+
"learning_rate": 7.838986095452311e-06,
|
| 3186 |
+
"loss": 3.5636,
|
| 3187 |
+
"step": 454
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 0.8231569425599277,
|
| 3191 |
+
"grad_norm": 8.96474838256836,
|
| 3192 |
+
"learning_rate": 7.685590314134294e-06,
|
| 3193 |
+
"loss": 3.81,
|
| 3194 |
+
"step": 455
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 0.824966078697422,
|
| 3198 |
+
"grad_norm": 7.717404842376709,
|
| 3199 |
+
"learning_rate": 7.533585208066301e-06,
|
| 3200 |
+
"loss": 3.1148,
|
| 3201 |
+
"step": 456
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.8267752148349163,
|
| 3205 |
+
"grad_norm": 6.736077308654785,
|
| 3206 |
+
"learning_rate": 7.382975772939865e-06,
|
| 3207 |
+
"loss": 2.6002,
|
| 3208 |
+
"step": 457
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.8285843509724107,
|
| 3212 |
+
"grad_norm": 6.166501998901367,
|
| 3213 |
+
"learning_rate": 7.2337669585774205e-06,
|
| 3214 |
+
"loss": 2.667,
|
| 3215 |
+
"step": 458
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.830393487109905,
|
| 3219 |
+
"grad_norm": 6.2878851890563965,
|
| 3220 |
+
"learning_rate": 7.085963668769552e-06,
|
| 3221 |
+
"loss": 2.9261,
|
| 3222 |
+
"step": 459
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.8322026232473994,
|
| 3226 |
+
"grad_norm": 6.576757431030273,
|
| 3227 |
+
"learning_rate": 6.939570761113939e-06,
|
| 3228 |
+
"loss": 3.0651,
|
| 3229 |
+
"step": 460
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.8340117593848937,
|
| 3233 |
+
"grad_norm": 7.321299076080322,
|
| 3234 |
+
"learning_rate": 6.794593046855613e-06,
|
| 3235 |
+
"loss": 3.4093,
|
| 3236 |
+
"step": 461
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.835820895522388,
|
| 3240 |
+
"grad_norm": 6.695417404174805,
|
| 3241 |
+
"learning_rate": 6.651035290728858e-06,
|
| 3242 |
+
"loss": 3.0874,
|
| 3243 |
+
"step": 462
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.8376300316598824,
|
| 3247 |
+
"grad_norm": 6.7737956047058105,
|
| 3248 |
+
"learning_rate": 6.508902210800649e-06,
|
| 3249 |
+
"loss": 3.2408,
|
| 3250 |
+
"step": 463
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.8394391677973767,
|
| 3254 |
+
"grad_norm": 6.985015869140625,
|
| 3255 |
+
"learning_rate": 6.36819847831554e-06,
|
| 3256 |
+
"loss": 3.3483,
|
| 3257 |
+
"step": 464
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.841248303934871,
|
| 3261 |
+
"grad_norm": 7.275407791137695,
|
| 3262 |
+
"learning_rate": 6.228928717542204e-06,
|
| 3263 |
+
"loss": 3.6385,
|
| 3264 |
+
"step": 465
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.8430574400723655,
|
| 3268 |
+
"grad_norm": 7.622275352478027,
|
| 3269 |
+
"learning_rate": 6.091097505621374e-06,
|
| 3270 |
+
"loss": 3.5877,
|
| 3271 |
+
"step": 466
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.8448665762098598,
|
| 3275 |
+
"grad_norm": 7.0419840812683105,
|
| 3276 |
+
"learning_rate": 5.9547093724155235e-06,
|
| 3277 |
+
"loss": 3.6186,
|
| 3278 |
+
"step": 467
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.8466757123473542,
|
| 3282 |
+
"grad_norm": 6.253328800201416,
|
| 3283 |
+
"learning_rate": 5.8197688003598815e-06,
|
| 3284 |
+
"loss": 3.0429,
|
| 3285 |
+
"step": 468
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.8484848484848485,
|
| 3289 |
+
"grad_norm": 7.221704006195068,
|
| 3290 |
+
"learning_rate": 5.686280224315188e-06,
|
| 3291 |
+
"loss": 3.9403,
|
| 3292 |
+
"step": 469
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 0.8502939846223428,
|
| 3296 |
+
"grad_norm": 6.812623977661133,
|
| 3297 |
+
"learning_rate": 5.554248031421871e-06,
|
| 3298 |
+
"loss": 3.6381,
|
| 3299 |
+
"step": 470
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 0.8521031207598372,
|
| 3303 |
+
"grad_norm": 6.868687152862549,
|
| 3304 |
+
"learning_rate": 5.423676560955976e-06,
|
| 3305 |
+
"loss": 3.789,
|
| 3306 |
+
"step": 471
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 0.8539122568973315,
|
| 3310 |
+
"grad_norm": 6.682875156402588,
|
| 3311 |
+
"learning_rate": 5.294570104186436e-06,
|
| 3312 |
+
"loss": 3.7681,
|
| 3313 |
+
"step": 472
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.8557213930348259,
|
| 3317 |
+
"grad_norm": 6.816806793212891,
|
| 3318 |
+
"learning_rate": 5.166932904234101e-06,
|
| 3319 |
+
"loss": 3.6686,
|
| 3320 |
+
"step": 473
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 0.8575305291723202,
|
| 3324 |
+
"grad_norm": 6.691549301147461,
|
| 3325 |
+
"learning_rate": 5.040769155932284e-06,
|
| 3326 |
+
"loss": 3.9365,
|
| 3327 |
+
"step": 474
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.8593396653098145,
|
| 3331 |
+
"grad_norm": 7.533728122711182,
|
| 3332 |
+
"learning_rate": 4.916083005688865e-06,
|
| 3333 |
+
"loss": 3.7974,
|
| 3334 |
+
"step": 475
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 0.8611488014473089,
|
| 3338 |
+
"grad_norm": 6.8411993980407715,
|
| 3339 |
+
"learning_rate": 4.792878551350055e-06,
|
| 3340 |
+
"loss": 3.7125,
|
| 3341 |
+
"step": 476
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 0.8629579375848032,
|
| 3345 |
+
"grad_norm": 6.767866611480713,
|
| 3346 |
+
"learning_rate": 4.671159842065698e-06,
|
| 3347 |
+
"loss": 3.4574,
|
| 3348 |
+
"step": 477
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 0.8647670737222976,
|
| 3352 |
+
"grad_norm": 7.180441856384277,
|
| 3353 |
+
"learning_rate": 4.550930878156185e-06,
|
| 3354 |
+
"loss": 4.3838,
|
| 3355 |
+
"step": 478
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.866576209859792,
|
| 3359 |
+
"grad_norm": 6.1762237548828125,
|
| 3360 |
+
"learning_rate": 4.432195610981032e-06,
|
| 3361 |
+
"loss": 3.9066,
|
| 3362 |
+
"step": 479
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.8683853459972863,
|
| 3366 |
+
"grad_norm": 6.982791423797607,
|
| 3367 |
+
"learning_rate": 4.314957942808956e-06,
|
| 3368 |
+
"loss": 3.5071,
|
| 3369 |
+
"step": 480
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.8701944821347807,
|
| 3373 |
+
"grad_norm": 7.1507368087768555,
|
| 3374 |
+
"learning_rate": 4.199221726689634e-06,
|
| 3375 |
+
"loss": 4.2304,
|
| 3376 |
+
"step": 481
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.872003618272275,
|
| 3380 |
+
"grad_norm": 6.495567321777344,
|
| 3381 |
+
"learning_rate": 4.084990766327135e-06,
|
| 3382 |
+
"loss": 3.4303,
|
| 3383 |
+
"step": 482
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.8738127544097694,
|
| 3387 |
+
"grad_norm": 6.444892406463623,
|
| 3388 |
+
"learning_rate": 3.972268815954832e-06,
|
| 3389 |
+
"loss": 3.7368,
|
| 3390 |
+
"step": 483
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.8756218905472637,
|
| 3394 |
+
"grad_norm": 6.858259201049805,
|
| 3395 |
+
"learning_rate": 3.861059580212056e-06,
|
| 3396 |
+
"loss": 3.6686,
|
| 3397 |
+
"step": 484
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.877431026684758,
|
| 3401 |
+
"grad_norm": 6.297997951507568,
|
| 3402 |
+
"learning_rate": 3.7513667140223417e-06,
|
| 3403 |
+
"loss": 3.3313,
|
| 3404 |
+
"step": 485
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.8792401628222524,
|
| 3408 |
+
"grad_norm": 6.799905300140381,
|
| 3409 |
+
"learning_rate": 3.6431938224733008e-06,
|
| 3410 |
+
"loss": 3.4037,
|
| 3411 |
+
"step": 486
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.8810492989597467,
|
| 3415 |
+
"grad_norm": 7.7271318435668945,
|
| 3416 |
+
"learning_rate": 3.5365444606981435e-06,
|
| 3417 |
+
"loss": 3.579,
|
| 3418 |
+
"step": 487
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.882858435097241,
|
| 3422 |
+
"grad_norm": 6.616701126098633,
|
| 3423 |
+
"learning_rate": 3.4314221337588217e-06,
|
| 3424 |
+
"loss": 2.8946,
|
| 3425 |
+
"step": 488
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.8846675712347354,
|
| 3429 |
+
"grad_norm": 6.063207626342773,
|
| 3430 |
+
"learning_rate": 3.3278302965308596e-06,
|
| 3431 |
+
"loss": 3.6569,
|
| 3432 |
+
"step": 489
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.8864767073722297,
|
| 3436 |
+
"grad_norm": 6.46480131149292,
|
| 3437 |
+
"learning_rate": 3.2257723535898175e-06,
|
| 3438 |
+
"loss": 4.9097,
|
| 3439 |
+
"step": 490
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.8882858435097241,
|
| 3443 |
+
"grad_norm": 6.829686164855957,
|
| 3444 |
+
"learning_rate": 3.125251659099332e-06,
|
| 3445 |
+
"loss": 5.4115,
|
| 3446 |
+
"step": 491
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.8900949796472184,
|
| 3450 |
+
"grad_norm": 7.085306167602539,
|
| 3451 |
+
"learning_rate": 3.0262715167009458e-06,
|
| 3452 |
+
"loss": 5.6045,
|
| 3453 |
+
"step": 492
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.8919041157847128,
|
| 3457 |
+
"grad_norm": 7.94856071472168,
|
| 3458 |
+
"learning_rate": 2.928835179405548e-06,
|
| 3459 |
+
"loss": 5.6366,
|
| 3460 |
+
"step": 493
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.8937132519222072,
|
| 3464 |
+
"grad_norm": 8.26148796081543,
|
| 3465 |
+
"learning_rate": 2.8329458494863847e-06,
|
| 3466 |
+
"loss": 5.7356,
|
| 3467 |
+
"step": 494
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.8955223880597015,
|
| 3471 |
+
"grad_norm": 8.643485069274902,
|
| 3472 |
+
"learning_rate": 2.738606678373873e-06,
|
| 3473 |
+
"loss": 5.2438,
|
| 3474 |
+
"step": 495
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.8973315241971959,
|
| 3478 |
+
"grad_norm": 9.256296157836914,
|
| 3479 |
+
"learning_rate": 2.645820766552026e-06,
|
| 3480 |
+
"loss": 5.6607,
|
| 3481 |
+
"step": 496
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.8991406603346902,
|
| 3485 |
+
"grad_norm": 10.048086166381836,
|
| 3486 |
+
"learning_rate": 2.554591163456527e-06,
|
| 3487 |
+
"loss": 5.3024,
|
| 3488 |
+
"step": 497
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.9009497964721845,
|
| 3492 |
+
"grad_norm": 10.239492416381836,
|
| 3493 |
+
"learning_rate": 2.4649208673745316e-06,
|
| 3494 |
+
"loss": 4.6682,
|
| 3495 |
+
"step": 498
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.9027589326096789,
|
| 3499 |
+
"grad_norm": 12.140185356140137,
|
| 3500 |
+
"learning_rate": 2.3768128253461253e-06,
|
| 3501 |
+
"loss": 5.7405,
|
| 3502 |
+
"step": 499
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.9045680687471732,
|
| 3506 |
+
"grad_norm": 11.735483169555664,
|
| 3507 |
+
"learning_rate": 2.2902699330674573e-06,
|
| 3508 |
+
"loss": 4.4613,
|
| 3509 |
+
"step": 500
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 0.9063772048846676,
|
| 3513 |
+
"grad_norm": 5.878971099853516,
|
| 3514 |
+
"learning_rate": 2.205295034795596e-06,
|
| 3515 |
+
"loss": 3.6071,
|
| 3516 |
+
"step": 501
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 0.9081863410221619,
|
| 3520 |
+
"grad_norm": 7.937101364135742,
|
| 3521 |
+
"learning_rate": 2.1218909232550155e-06,
|
| 3522 |
+
"loss": 3.7888,
|
| 3523 |
+
"step": 502
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 0.9099954771596562,
|
| 3527 |
+
"grad_norm": 9.610072135925293,
|
| 3528 |
+
"learning_rate": 2.0400603395458407e-06,
|
| 3529 |
+
"loss": 3.6901,
|
| 3530 |
+
"step": 503
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.9118046132971506,
|
| 3534 |
+
"grad_norm": 7.69047737121582,
|
| 3535 |
+
"learning_rate": 1.9598059730537466e-06,
|
| 3536 |
+
"loss": 3.14,
|
| 3537 |
+
"step": 504
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.9136137494346449,
|
| 3541 |
+
"grad_norm": 7.912228107452393,
|
| 3542 |
+
"learning_rate": 1.8811304613615909e-06,
|
| 3543 |
+
"loss": 2.9848,
|
| 3544 |
+
"step": 505
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 0.9154228855721394,
|
| 3548 |
+
"grad_norm": 7.0957489013671875,
|
| 3549 |
+
"learning_rate": 1.8040363901627e-06,
|
| 3550 |
+
"loss": 2.4616,
|
| 3551 |
+
"step": 506
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 0.9172320217096337,
|
| 3555 |
+
"grad_norm": 6.868281364440918,
|
| 3556 |
+
"learning_rate": 1.7285262931759082e-06,
|
| 3557 |
+
"loss": 2.9809,
|
| 3558 |
+
"step": 507
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 0.919041157847128,
|
| 3562 |
+
"grad_norm": 7.486063480377197,
|
| 3563 |
+
"learning_rate": 1.6546026520622759e-06,
|
| 3564 |
+
"loss": 2.928,
|
| 3565 |
+
"step": 508
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 0.9208502939846224,
|
| 3569 |
+
"grad_norm": 7.200881481170654,
|
| 3570 |
+
"learning_rate": 1.5822678963435478e-06,
|
| 3571 |
+
"loss": 3.2972,
|
| 3572 |
+
"step": 509
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 0.9226594301221167,
|
| 3576 |
+
"grad_norm": 6.531789779663086,
|
| 3577 |
+
"learning_rate": 1.5115244033222731e-06,
|
| 3578 |
+
"loss": 2.9837,
|
| 3579 |
+
"step": 510
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 0.924468566259611,
|
| 3583 |
+
"grad_norm": 6.44420051574707,
|
| 3584 |
+
"learning_rate": 1.4423744980037068e-06,
|
| 3585 |
+
"loss": 3.4836,
|
| 3586 |
+
"step": 511
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 0.9262777023971054,
|
| 3590 |
+
"grad_norm": 6.985484600067139,
|
| 3591 |
+
"learning_rate": 1.3748204530193987e-06,
|
| 3592 |
+
"loss": 3.5155,
|
| 3593 |
+
"step": 512
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.9280868385345997,
|
| 3597 |
+
"grad_norm": 6.48081111907959,
|
| 3598 |
+
"learning_rate": 1.3088644885524637e-06,
|
| 3599 |
+
"loss": 2.7183,
|
| 3600 |
+
"step": 513
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 0.9298959746720941,
|
| 3604 |
+
"grad_norm": 6.328634738922119,
|
| 3605 |
+
"learning_rate": 1.2445087722646575e-06,
|
| 3606 |
+
"loss": 3.8345,
|
| 3607 |
+
"step": 514
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 0.9317051108095884,
|
| 3611 |
+
"grad_norm": 6.782224655151367,
|
| 3612 |
+
"learning_rate": 1.1817554192251e-06,
|
| 3613 |
+
"loss": 4.047,
|
| 3614 |
+
"step": 515
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 0.9335142469470827,
|
| 3618 |
+
"grad_norm": 7.253306865692139,
|
| 3619 |
+
"learning_rate": 1.1206064918408143e-06,
|
| 3620 |
+
"loss": 3.5562,
|
| 3621 |
+
"step": 516
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 0.9353233830845771,
|
| 3625 |
+
"grad_norm": 7.641778469085693,
|
| 3626 |
+
"learning_rate": 1.0610639997888916e-06,
|
| 3627 |
+
"loss": 3.4378,
|
| 3628 |
+
"step": 517
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 0.9371325192220714,
|
| 3632 |
+
"grad_norm": 8.075942039489746,
|
| 3633 |
+
"learning_rate": 1.0031298999504558e-06,
|
| 3634 |
+
"loss": 4.6599,
|
| 3635 |
+
"step": 518
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 0.9389416553595658,
|
| 3639 |
+
"grad_norm": 6.147914409637451,
|
| 3640 |
+
"learning_rate": 9.468060963463755e-07,
|
| 3641 |
+
"loss": 3.3544,
|
| 3642 |
+
"step": 519
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 0.9407507914970602,
|
| 3646 |
+
"grad_norm": 6.727973937988281,
|
| 3647 |
+
"learning_rate": 8.920944400746589e-07,
|
| 3648 |
+
"loss": 3.6667,
|
| 3649 |
+
"step": 520
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.9425599276345545,
|
| 3653 |
+
"grad_norm": 6.554884433746338,
|
| 3654 |
+
"learning_rate": 8.389967292496359e-07,
|
| 3655 |
+
"loss": 2.9346,
|
| 3656 |
+
"step": 521
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 0.9443690637720489,
|
| 3660 |
+
"grad_norm": 6.31584358215332,
|
| 3661 |
+
"learning_rate": 7.875147089428437e-07,
|
| 3662 |
+
"loss": 3.7725,
|
| 3663 |
+
"step": 522
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 0.9461781999095432,
|
| 3667 |
+
"grad_norm": 7.042573928833008,
|
| 3668 |
+
"learning_rate": 7.376500711257061e-07,
|
| 3669 |
+
"loss": 4.3886,
|
| 3670 |
+
"step": 523
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 0.9479873360470376,
|
| 3674 |
+
"grad_norm": 6.561697959899902,
|
| 3675 |
+
"learning_rate": 6.894044546138845e-07,
|
| 3676 |
+
"loss": 3.9783,
|
| 3677 |
+
"step": 524
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 0.9497964721845319,
|
| 3681 |
+
"grad_norm": 6.423247814178467,
|
| 3682 |
+
"learning_rate": 6.427794450134528e-07,
|
| 3683 |
+
"loss": 3.5429,
|
| 3684 |
+
"step": 525
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 0.9516056083220262,
|
| 3688 |
+
"grad_norm": 8.323712348937988,
|
| 3689 |
+
"learning_rate": 5.977765746687569e-07,
|
| 3690 |
+
"loss": 3.9779,
|
| 3691 |
+
"step": 526
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 0.9534147444595206,
|
| 3695 |
+
"grad_norm": 7.722854137420654,
|
| 3696 |
+
"learning_rate": 5.543973226120935e-07,
|
| 3697 |
+
"loss": 4.0921,
|
| 3698 |
+
"step": 527
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 0.9552238805970149,
|
| 3702 |
+
"grad_norm": 6.630421161651611,
|
| 3703 |
+
"learning_rate": 5.126431145150546e-07,
|
| 3704 |
+
"loss": 4.1686,
|
| 3705 |
+
"step": 528
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.9570330167345092,
|
| 3709 |
+
"grad_norm": 6.825472831726074,
|
| 3710 |
+
"learning_rate": 4.7251532264170895e-07,
|
| 3711 |
+
"loss": 4.5243,
|
| 3712 |
+
"step": 529
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 0.9588421528720036,
|
| 3716 |
+
"grad_norm": 6.3734965324401855,
|
| 3717 |
+
"learning_rate": 4.340152658034835e-07,
|
| 3718 |
+
"loss": 3.5058,
|
| 3719 |
+
"step": 530
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 0.9606512890094979,
|
| 3723 |
+
"grad_norm": 6.643232345581055,
|
| 3724 |
+
"learning_rate": 3.971442093158195e-07,
|
| 3725 |
+
"loss": 3.8592,
|
| 3726 |
+
"step": 531
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 0.9624604251469923,
|
| 3730 |
+
"grad_norm": 6.958840847015381,
|
| 3731 |
+
"learning_rate": 3.61903364956595e-07,
|
| 3732 |
+
"loss": 3.8099,
|
| 3733 |
+
"step": 532
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 0.9642695612844867,
|
| 3737 |
+
"grad_norm": 6.364724159240723,
|
| 3738 |
+
"learning_rate": 3.282938909263122e-07,
|
| 3739 |
+
"loss": 3.5475,
|
| 3740 |
+
"step": 533
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 0.966078697421981,
|
| 3744 |
+
"grad_norm": 6.911684513092041,
|
| 3745 |
+
"learning_rate": 2.9631689180999457e-07,
|
| 3746 |
+
"loss": 3.4872,
|
| 3747 |
+
"step": 534
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 0.9678878335594754,
|
| 3751 |
+
"grad_norm": 6.476950645446777,
|
| 3752 |
+
"learning_rate": 2.6597341854092684e-07,
|
| 3753 |
+
"loss": 3.4591,
|
| 3754 |
+
"step": 535
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.9696969696969697,
|
| 3758 |
+
"grad_norm": 7.094813823699951,
|
| 3759 |
+
"learning_rate": 2.3726446836608296e-07,
|
| 3760 |
+
"loss": 3.6974,
|
| 3761 |
+
"step": 536
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.9715061058344641,
|
| 3765 |
+
"grad_norm": 8.420538902282715,
|
| 3766 |
+
"learning_rate": 2.101909848133743e-07,
|
| 3767 |
+
"loss": 3.4565,
|
| 3768 |
+
"step": 537
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.9733152419719584,
|
| 3772 |
+
"grad_norm": 7.818291187286377,
|
| 3773 |
+
"learning_rate": 1.8475385766063003e-07,
|
| 3774 |
+
"loss": 3.5051,
|
| 3775 |
+
"step": 538
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.9751243781094527,
|
| 3779 |
+
"grad_norm": 7.150783538818359,
|
| 3780 |
+
"learning_rate": 1.6095392290635393e-07,
|
| 3781 |
+
"loss": 3.7192,
|
| 3782 |
+
"step": 539
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.9769335142469471,
|
| 3786 |
+
"grad_norm": 6.040256023406982,
|
| 3787 |
+
"learning_rate": 1.3879196274224627e-07,
|
| 3788 |
+
"loss": 3.8009,
|
| 3789 |
+
"step": 540
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.9787426503844414,
|
| 3793 |
+
"grad_norm": 6.803682327270508,
|
| 3794 |
+
"learning_rate": 1.1826870552749669e-07,
|
| 3795 |
+
"loss": 5.3465,
|
| 3796 |
+
"step": 541
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.9805517865219358,
|
| 3800 |
+
"grad_norm": 6.929571151733398,
|
| 3801 |
+
"learning_rate": 9.938482576487552e-08,
|
| 3802 |
+
"loss": 5.3678,
|
| 3803 |
+
"step": 542
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.9823609226594301,
|
| 3807 |
+
"grad_norm": 7.471601486206055,
|
| 3808 |
+
"learning_rate": 8.214094407851813e-08,
|
| 3809 |
+
"loss": 5.8603,
|
| 3810 |
+
"step": 543
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.9841700587969244,
|
| 3814 |
+
"grad_norm": 8.225199699401855,
|
| 3815 |
+
"learning_rate": 6.653762719355805e-08,
|
| 3816 |
+
"loss": 5.9743,
|
| 3817 |
+
"step": 544
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.9859791949344188,
|
| 3821 |
+
"grad_norm": 8.476972579956055,
|
| 3822 |
+
"learning_rate": 5.257538791749172e-08,
|
| 3823 |
+
"loss": 5.3998,
|
| 3824 |
+
"step": 545
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.9877883310719131,
|
| 3828 |
+
"grad_norm": 9.56104564666748,
|
| 3829 |
+
"learning_rate": 4.025468512333097e-08,
|
| 3830 |
+
"loss": 5.3685,
|
| 3831 |
+
"step": 546
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.9895974672094076,
|
| 3835 |
+
"grad_norm": 8.92081356048584,
|
| 3836 |
+
"learning_rate": 2.957592373452056e-08,
|
| 3837 |
+
"loss": 4.7335,
|
| 3838 |
+
"step": 547
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 0.9914066033469019,
|
| 3842 |
+
"grad_norm": 10.06761360168457,
|
| 3843 |
+
"learning_rate": 2.053945471162666e-08,
|
| 3844 |
+
"loss": 5.0433,
|
| 3845 |
+
"step": 548
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 0.9932157394843962,
|
| 3849 |
+
"grad_norm": 10.523520469665527,
|
| 3850 |
+
"learning_rate": 1.3145575040801606e-08,
|
| 3851 |
+
"loss": 5.8829,
|
| 3852 |
+
"step": 549
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 0.9950248756218906,
|
| 3856 |
+
"grad_norm": 11.902076721191406,
|
| 3857 |
+
"learning_rate": 7.394527724030598e-09,
|
| 3858 |
+
"loss": 4.1261,
|
| 3859 |
+
"step": 550
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 0.9968340117593849,
|
| 3863 |
+
"grad_norm": 7.296747207641602,
|
| 3864 |
+
"learning_rate": 3.286501771138095e-09,
|
| 3865 |
+
"loss": 3.5647,
|
| 3866 |
+
"step": 551
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 0.9986431478968792,
|
| 3870 |
+
"grad_norm": 6.893759250640869,
|
| 3871 |
+
"learning_rate": 8.216321935816674e-10,
|
| 3872 |
+
"loss": 3.7598,
|
| 3873 |
+
"step": 552
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.9986431478968792,
|
| 3877 |
+
"eval_loss": 0.9817073345184326,
|
| 3878 |
+
"eval_runtime": 38.8717,
|
| 3879 |
+
"eval_samples_per_second": 11.988,
|
| 3880 |
+
"eval_steps_per_second": 3.01,
|
| 3881 |
+
"step": 552
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 1.0009045680687472,
|
| 3885 |
+
"grad_norm": 7.406569957733154,
|
| 3886 |
+
"learning_rate": 0.0,
|
| 3887 |
+
"loss": 4.3257,
|
| 3888 |
+
"step": 553
|
| 3889 |
}
|
| 3890 |
],
|
| 3891 |
"logging_steps": 1,
|
|
|
|
| 3900 |
"should_evaluate": false,
|
| 3901 |
"should_log": false,
|
| 3902 |
"should_save": true,
|
| 3903 |
+
"should_training_stop": true
|
| 3904 |
},
|
| 3905 |
"attributes": {}
|
| 3906 |
}
|
| 3907 |
},
|
| 3908 |
+
"total_flos": 6.383309931864392e+17,
|
| 3909 |
"train_batch_size": 4,
|
| 3910 |
"trial_name": null,
|
| 3911 |
"trial_params": null
|