Training in progress, step 583, 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 80013120
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:faa688cd98990073aff639ebd4a3b2b379792da6408f701658a084a6c9a67e68
|
| 3 |
size 80013120
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 41120084
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5eb1704df54be29043bc390d4c6a0a8bf765c611f32fed903b777d844b7ec767
|
| 3 |
size 41120084
|
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 sha256:15742b0303b779f7ba81698192a8dbd1ba4ab375f37877d7de83e6a31f9a7244
|
| 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 sha256:4477778303bed47c1d65922d6e7649fd04e01c218f6911cbe8e084a58d87e771
|
| 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,
|
|
@@ -3073,6 +3073,1029 @@
|
|
| 3073 |
"learning_rate": 1.4740874490528395e-05,
|
| 3074 |
"loss": 0.9391,
|
| 3075 |
"step": 438
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3076 |
}
|
| 3077 |
],
|
| 3078 |
"logging_steps": 1,
|
|
@@ -3087,12 +4110,12 @@
|
|
| 3087 |
"should_evaluate": false,
|
| 3088 |
"should_log": false,
|
| 3089 |
"should_save": true,
|
| 3090 |
-
"should_training_stop":
|
| 3091 |
},
|
| 3092 |
"attributes": {}
|
| 3093 |
}
|
| 3094 |
},
|
| 3095 |
-
"total_flos":
|
| 3096 |
"train_batch_size": 4,
|
| 3097 |
"trial_name": null,
|
| 3098 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0004290004290004,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 583,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3073 |
"learning_rate": 1.4740874490528395e-05,
|
| 3074 |
"loss": 0.9391,
|
| 3075 |
"step": 438
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 0.7533247533247533,
|
| 3079 |
+
"grad_norm": 0.8027117252349854,
|
| 3080 |
+
"learning_rate": 1.4548708510022824e-05,
|
| 3081 |
+
"loss": 0.9091,
|
| 3082 |
+
"step": 439
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.7550407550407551,
|
| 3086 |
+
"grad_norm": 0.8315735459327698,
|
| 3087 |
+
"learning_rate": 1.435758983905955e-05,
|
| 3088 |
+
"loss": 0.7665,
|
| 3089 |
+
"step": 440
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.7567567567567568,
|
| 3093 |
+
"grad_norm": 0.9047240614891052,
|
| 3094 |
+
"learning_rate": 1.4167524123706743e-05,
|
| 3095 |
+
"loss": 0.9747,
|
| 3096 |
+
"step": 441
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.7584727584727585,
|
| 3100 |
+
"grad_norm": 0.8648135662078857,
|
| 3101 |
+
"learning_rate": 1.397851697892592e-05,
|
| 3102 |
+
"loss": 0.9396,
|
| 3103 |
+
"step": 442
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 0.7601887601887601,
|
| 3107 |
+
"grad_norm": 0.8910363912582397,
|
| 3108 |
+
"learning_rate": 1.3790573988406074e-05,
|
| 3109 |
+
"loss": 0.8706,
|
| 3110 |
+
"step": 443
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 0.7619047619047619,
|
| 3114 |
+
"grad_norm": 0.9140577912330627,
|
| 3115 |
+
"learning_rate": 1.3603700704398741e-05,
|
| 3116 |
+
"loss": 0.9549,
|
| 3117 |
+
"step": 444
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 0.7636207636207636,
|
| 3121 |
+
"grad_norm": 1.0109758377075195,
|
| 3122 |
+
"learning_rate": 1.3417902647553948e-05,
|
| 3123 |
+
"loss": 0.961,
|
| 3124 |
+
"step": 445
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.7653367653367653,
|
| 3128 |
+
"grad_norm": 0.9864091873168945,
|
| 3129 |
+
"learning_rate": 1.3233185306757145e-05,
|
| 3130 |
+
"loss": 0.9942,
|
| 3131 |
+
"step": 446
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 0.767052767052767,
|
| 3135 |
+
"grad_norm": 1.0668342113494873,
|
| 3136 |
+
"learning_rate": 1.3049554138967051e-05,
|
| 3137 |
+
"loss": 0.9568,
|
| 3138 |
+
"step": 447
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 0.7687687687687688,
|
| 3142 |
+
"grad_norm": 1.0860650539398193,
|
| 3143 |
+
"learning_rate": 1.2867014569054387e-05,
|
| 3144 |
+
"loss": 0.9073,
|
| 3145 |
+
"step": 448
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.7704847704847705,
|
| 3149 |
+
"grad_norm": 1.0945936441421509,
|
| 3150 |
+
"learning_rate": 1.2685571989641698e-05,
|
| 3151 |
+
"loss": 0.9378,
|
| 3152 |
+
"step": 449
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 0.7722007722007722,
|
| 3156 |
+
"grad_norm": 1.3445020914077759,
|
| 3157 |
+
"learning_rate": 1.2505231760943998e-05,
|
| 3158 |
+
"loss": 1.0138,
|
| 3159 |
+
"step": 450
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.7739167739167739,
|
| 3163 |
+
"grad_norm": 0.3614446818828583,
|
| 3164 |
+
"learning_rate": 1.2325999210610423e-05,
|
| 3165 |
+
"loss": 1.0215,
|
| 3166 |
+
"step": 451
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.7756327756327757,
|
| 3170 |
+
"grad_norm": 0.41316771507263184,
|
| 3171 |
+
"learning_rate": 1.214787963356685e-05,
|
| 3172 |
+
"loss": 1.1453,
|
| 3173 |
+
"step": 452
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 0.7773487773487774,
|
| 3177 |
+
"grad_norm": 0.4298244118690491,
|
| 3178 |
+
"learning_rate": 1.1970878291859423e-05,
|
| 3179 |
+
"loss": 1.0364,
|
| 3180 |
+
"step": 453
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.7790647790647791,
|
| 3184 |
+
"grad_norm": 0.48577290773391724,
|
| 3185 |
+
"learning_rate": 1.1795000414499186e-05,
|
| 3186 |
+
"loss": 1.0668,
|
| 3187 |
+
"step": 454
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 0.7807807807807807,
|
| 3191 |
+
"grad_norm": 0.5088241696357727,
|
| 3192 |
+
"learning_rate": 1.1620251197307535e-05,
|
| 3193 |
+
"loss": 1.0152,
|
| 3194 |
+
"step": 455
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 0.7824967824967825,
|
| 3198 |
+
"grad_norm": 0.546314001083374,
|
| 3199 |
+
"learning_rate": 1.1446635802762795e-05,
|
| 3200 |
+
"loss": 1.0387,
|
| 3201 |
+
"step": 456
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.7842127842127842,
|
| 3205 |
+
"grad_norm": 0.5127102136611938,
|
| 3206 |
+
"learning_rate": 1.1274159359847591e-05,
|
| 3207 |
+
"loss": 1.0903,
|
| 3208 |
+
"step": 457
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.7859287859287859,
|
| 3212 |
+
"grad_norm": 0.5245983004570007,
|
| 3213 |
+
"learning_rate": 1.110282696389745e-05,
|
| 3214 |
+
"loss": 1.0062,
|
| 3215 |
+
"step": 458
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.7876447876447876,
|
| 3219 |
+
"grad_norm": 0.5490956902503967,
|
| 3220 |
+
"learning_rate": 1.0932643676450205e-05,
|
| 3221 |
+
"loss": 0.9529,
|
| 3222 |
+
"step": 459
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.7893607893607893,
|
| 3226 |
+
"grad_norm": 0.5959334969520569,
|
| 3227 |
+
"learning_rate": 1.07636145250965e-05,
|
| 3228 |
+
"loss": 1.0001,
|
| 3229 |
+
"step": 460
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.7910767910767911,
|
| 3233 |
+
"grad_norm": 0.5808343291282654,
|
| 3234 |
+
"learning_rate": 1.0595744503331207e-05,
|
| 3235 |
+
"loss": 1.0115,
|
| 3236 |
+
"step": 461
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.7927927927927928,
|
| 3240 |
+
"grad_norm": 0.5722229480743408,
|
| 3241 |
+
"learning_rate": 1.0429038570405975e-05,
|
| 3242 |
+
"loss": 1.0045,
|
| 3243 |
+
"step": 462
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.7945087945087945,
|
| 3247 |
+
"grad_norm": 0.5969129800796509,
|
| 3248 |
+
"learning_rate": 1.0263501651182706e-05,
|
| 3249 |
+
"loss": 0.9347,
|
| 3250 |
+
"step": 463
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.7962247962247962,
|
| 3254 |
+
"grad_norm": 0.553562343120575,
|
| 3255 |
+
"learning_rate": 1.0099138635988026e-05,
|
| 3256 |
+
"loss": 0.9087,
|
| 3257 |
+
"step": 464
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.797940797940798,
|
| 3261 |
+
"grad_norm": 0.6024061441421509,
|
| 3262 |
+
"learning_rate": 9.935954380468859e-06,
|
| 3263 |
+
"loss": 0.9291,
|
| 3264 |
+
"step": 465
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.7996567996567997,
|
| 3268 |
+
"grad_norm": 0.6290377378463745,
|
| 3269 |
+
"learning_rate": 9.773953705448952e-06,
|
| 3270 |
+
"loss": 0.9676,
|
| 3271 |
+
"step": 466
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.8013728013728014,
|
| 3275 |
+
"grad_norm": 0.6255950927734375,
|
| 3276 |
+
"learning_rate": 9.613141396786462e-06,
|
| 3277 |
+
"loss": 0.8735,
|
| 3278 |
+
"step": 467
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.803088803088803,
|
| 3282 |
+
"grad_norm": 0.6054927706718445,
|
| 3283 |
+
"learning_rate": 9.453522205232612e-06,
|
| 3284 |
+
"loss": 1.0264,
|
| 3285 |
+
"step": 468
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.8048048048048048,
|
| 3289 |
+
"grad_norm": 0.6100555658340454,
|
| 3290 |
+
"learning_rate": 9.295100846291238e-06,
|
| 3291 |
+
"loss": 0.9192,
|
| 3292 |
+
"step": 469
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 0.8065208065208065,
|
| 3296 |
+
"grad_norm": 0.6099107265472412,
|
| 3297 |
+
"learning_rate": 9.137882000079611e-06,
|
| 3298 |
+
"loss": 0.9077,
|
| 3299 |
+
"step": 470
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 0.8082368082368082,
|
| 3303 |
+
"grad_norm": 0.682094156742096,
|
| 3304 |
+
"learning_rate": 8.981870311190099e-06,
|
| 3305 |
+
"loss": 0.9293,
|
| 3306 |
+
"step": 471
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 0.8099528099528099,
|
| 3310 |
+
"grad_norm": 0.6415854096412659,
|
| 3311 |
+
"learning_rate": 8.827070388552976e-06,
|
| 3312 |
+
"loss": 0.9176,
|
| 3313 |
+
"step": 472
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.8116688116688117,
|
| 3317 |
+
"grad_norm": 0.6942915320396423,
|
| 3318 |
+
"learning_rate": 8.673486805300263e-06,
|
| 3319 |
+
"loss": 0.8685,
|
| 3320 |
+
"step": 473
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 0.8133848133848134,
|
| 3324 |
+
"grad_norm": 0.6811203360557556,
|
| 3325 |
+
"learning_rate": 8.521124098630601e-06,
|
| 3326 |
+
"loss": 0.889,
|
| 3327 |
+
"step": 474
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.8151008151008151,
|
| 3331 |
+
"grad_norm": 0.6820942163467407,
|
| 3332 |
+
"learning_rate": 8.369986769675269e-06,
|
| 3333 |
+
"loss": 0.9223,
|
| 3334 |
+
"step": 475
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 0.8168168168168168,
|
| 3338 |
+
"grad_norm": 0.7045984864234924,
|
| 3339 |
+
"learning_rate": 8.220079283365156e-06,
|
| 3340 |
+
"loss": 0.7944,
|
| 3341 |
+
"step": 476
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 0.8185328185328186,
|
| 3345 |
+
"grad_norm": 0.704068124294281,
|
| 3346 |
+
"learning_rate": 8.071406068298926e-06,
|
| 3347 |
+
"loss": 0.9811,
|
| 3348 |
+
"step": 477
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 0.8202488202488203,
|
| 3352 |
+
"grad_norm": 0.758382260799408,
|
| 3353 |
+
"learning_rate": 7.923971516612067e-06,
|
| 3354 |
+
"loss": 0.9227,
|
| 3355 |
+
"step": 478
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.821964821964822,
|
| 3359 |
+
"grad_norm": 0.736240804195404,
|
| 3360 |
+
"learning_rate": 7.77777998384726e-06,
|
| 3361 |
+
"loss": 0.8681,
|
| 3362 |
+
"step": 479
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.8236808236808236,
|
| 3366 |
+
"grad_norm": 0.7136363983154297,
|
| 3367 |
+
"learning_rate": 7.63283578882566e-06,
|
| 3368 |
+
"loss": 0.9201,
|
| 3369 |
+
"step": 480
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.8253968253968254,
|
| 3373 |
+
"grad_norm": 0.743965208530426,
|
| 3374 |
+
"learning_rate": 7.489143213519301e-06,
|
| 3375 |
+
"loss": 0.9445,
|
| 3376 |
+
"step": 481
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.8271128271128271,
|
| 3380 |
+
"grad_norm": 0.7171424031257629,
|
| 3381 |
+
"learning_rate": 7.346706502924572e-06,
|
| 3382 |
+
"loss": 0.8123,
|
| 3383 |
+
"step": 482
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.8288288288288288,
|
| 3387 |
+
"grad_norm": 0.774448573589325,
|
| 3388 |
+
"learning_rate": 7.205529864936883e-06,
|
| 3389 |
+
"loss": 0.9915,
|
| 3390 |
+
"step": 483
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.8305448305448305,
|
| 3394 |
+
"grad_norm": 0.820214033126831,
|
| 3395 |
+
"learning_rate": 7.065617470226299e-06,
|
| 3396 |
+
"loss": 0.9966,
|
| 3397 |
+
"step": 484
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.8322608322608323,
|
| 3401 |
+
"grad_norm": 0.7767578959465027,
|
| 3402 |
+
"learning_rate": 6.926973452114338e-06,
|
| 3403 |
+
"loss": 0.8903,
|
| 3404 |
+
"step": 485
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.833976833976834,
|
| 3408 |
+
"grad_norm": 0.8457949757575989,
|
| 3409 |
+
"learning_rate": 6.78960190645187e-06,
|
| 3410 |
+
"loss": 0.9167,
|
| 3411 |
+
"step": 486
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.8356928356928357,
|
| 3415 |
+
"grad_norm": 0.8684483170509338,
|
| 3416 |
+
"learning_rate": 6.653506891498118e-06,
|
| 3417 |
+
"loss": 0.9654,
|
| 3418 |
+
"step": 487
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.8374088374088374,
|
| 3422 |
+
"grad_norm": 0.8468960523605347,
|
| 3423 |
+
"learning_rate": 6.518692427800765e-06,
|
| 3424 |
+
"loss": 0.8618,
|
| 3425 |
+
"step": 488
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.8391248391248392,
|
| 3429 |
+
"grad_norm": 0.8221365213394165,
|
| 3430 |
+
"learning_rate": 6.3851624980771905e-06,
|
| 3431 |
+
"loss": 0.9196,
|
| 3432 |
+
"step": 489
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.8408408408408409,
|
| 3436 |
+
"grad_norm": 0.8817353844642639,
|
| 3437 |
+
"learning_rate": 6.2529210470967605e-06,
|
| 3438 |
+
"loss": 0.8776,
|
| 3439 |
+
"step": 490
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.8425568425568426,
|
| 3443 |
+
"grad_norm": 0.8569510579109192,
|
| 3444 |
+
"learning_rate": 6.121971981564367e-06,
|
| 3445 |
+
"loss": 0.89,
|
| 3446 |
+
"step": 491
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.8442728442728443,
|
| 3450 |
+
"grad_norm": 0.8733360767364502,
|
| 3451 |
+
"learning_rate": 5.992319170004967e-06,
|
| 3452 |
+
"loss": 0.8815,
|
| 3453 |
+
"step": 492
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.8459888459888459,
|
| 3457 |
+
"grad_norm": 0.9367774128913879,
|
| 3458 |
+
"learning_rate": 5.863966442649327e-06,
|
| 3459 |
+
"loss": 0.9849,
|
| 3460 |
+
"step": 493
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.8477048477048477,
|
| 3464 |
+
"grad_norm": 0.944500744342804,
|
| 3465 |
+
"learning_rate": 5.7369175913208e-06,
|
| 3466 |
+
"loss": 0.9588,
|
| 3467 |
+
"step": 494
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.8494208494208494,
|
| 3471 |
+
"grad_norm": 1.0418821573257446,
|
| 3472 |
+
"learning_rate": 5.611176369323412e-06,
|
| 3473 |
+
"loss": 1.0623,
|
| 3474 |
+
"step": 495
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.8511368511368511,
|
| 3478 |
+
"grad_norm": 1.1258809566497803,
|
| 3479 |
+
"learning_rate": 5.4867464913308965e-06,
|
| 3480 |
+
"loss": 0.9395,
|
| 3481 |
+
"step": 496
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.8528528528528528,
|
| 3485 |
+
"grad_norm": 1.159876823425293,
|
| 3486 |
+
"learning_rate": 5.363631633277006e-06,
|
| 3487 |
+
"loss": 1.0624,
|
| 3488 |
+
"step": 497
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.8545688545688546,
|
| 3492 |
+
"grad_norm": 1.2327461242675781,
|
| 3493 |
+
"learning_rate": 5.241835432246889e-06,
|
| 3494 |
+
"loss": 1.0205,
|
| 3495 |
+
"step": 498
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.8562848562848563,
|
| 3499 |
+
"grad_norm": 1.2912688255310059,
|
| 3500 |
+
"learning_rate": 5.121361486369625e-06,
|
| 3501 |
+
"loss": 0.9273,
|
| 3502 |
+
"step": 499
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.858000858000858,
|
| 3506 |
+
"grad_norm": 1.6289055347442627,
|
| 3507 |
+
"learning_rate": 5.002213354711987e-06,
|
| 3508 |
+
"loss": 0.9433,
|
| 3509 |
+
"step": 500
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 0.8597168597168597,
|
| 3513 |
+
"grad_norm": 0.3974000811576843,
|
| 3514 |
+
"learning_rate": 4.884394557173249e-06,
|
| 3515 |
+
"loss": 1.0532,
|
| 3516 |
+
"step": 501
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 0.8614328614328615,
|
| 3520 |
+
"grad_norm": 0.40964797139167786,
|
| 3521 |
+
"learning_rate": 4.7679085743812306e-06,
|
| 3522 |
+
"loss": 1.1004,
|
| 3523 |
+
"step": 502
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 0.8631488631488632,
|
| 3527 |
+
"grad_norm": 0.4464317560195923,
|
| 3528 |
+
"learning_rate": 4.652758847589416e-06,
|
| 3529 |
+
"loss": 0.9262,
|
| 3530 |
+
"step": 503
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.8648648648648649,
|
| 3534 |
+
"grad_norm": 0.44399040937423706,
|
| 3535 |
+
"learning_rate": 4.538948778575375e-06,
|
| 3536 |
+
"loss": 0.9454,
|
| 3537 |
+
"step": 504
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.8665808665808665,
|
| 3541 |
+
"grad_norm": 0.5176399350166321,
|
| 3542 |
+
"learning_rate": 4.426481729540205e-06,
|
| 3543 |
+
"loss": 1.0398,
|
| 3544 |
+
"step": 505
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 0.8682968682968683,
|
| 3548 |
+
"grad_norm": 0.515872597694397,
|
| 3549 |
+
"learning_rate": 4.315361023009229e-06,
|
| 3550 |
+
"loss": 1.0069,
|
| 3551 |
+
"step": 506
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 0.87001287001287,
|
| 3555 |
+
"grad_norm": 0.5160545110702515,
|
| 3556 |
+
"learning_rate": 4.205589941733834e-06,
|
| 3557 |
+
"loss": 0.9582,
|
| 3558 |
+
"step": 507
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 0.8717288717288717,
|
| 3562 |
+
"grad_norm": 0.5296468734741211,
|
| 3563 |
+
"learning_rate": 4.0971717285944954e-06,
|
| 3564 |
+
"loss": 0.9789,
|
| 3565 |
+
"step": 508
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 0.8734448734448734,
|
| 3569 |
+
"grad_norm": 0.5391986966133118,
|
| 3570 |
+
"learning_rate": 3.990109586504965e-06,
|
| 3571 |
+
"loss": 1.0041,
|
| 3572 |
+
"step": 509
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 0.8751608751608752,
|
| 3576 |
+
"grad_norm": 0.5538270473480225,
|
| 3577 |
+
"learning_rate": 3.88440667831767e-06,
|
| 3578 |
+
"loss": 0.9985,
|
| 3579 |
+
"step": 510
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 0.8768768768768769,
|
| 3583 |
+
"grad_norm": 0.5680475831031799,
|
| 3584 |
+
"learning_rate": 3.7800661267302417e-06,
|
| 3585 |
+
"loss": 0.9688,
|
| 3586 |
+
"step": 511
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 0.8785928785928786,
|
| 3590 |
+
"grad_norm": 0.5948978662490845,
|
| 3591 |
+
"learning_rate": 3.6770910141932956e-06,
|
| 3592 |
+
"loss": 0.8732,
|
| 3593 |
+
"step": 512
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.8803088803088803,
|
| 3597 |
+
"grad_norm": 0.5594288110733032,
|
| 3598 |
+
"learning_rate": 3.5754843828193716e-06,
|
| 3599 |
+
"loss": 0.9891,
|
| 3600 |
+
"step": 513
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 0.882024882024882,
|
| 3604 |
+
"grad_norm": 0.5742771625518799,
|
| 3605 |
+
"learning_rate": 3.4752492342930377e-06,
|
| 3606 |
+
"loss": 0.952,
|
| 3607 |
+
"step": 514
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 0.8837408837408838,
|
| 3611 |
+
"grad_norm": 0.6060066223144531,
|
| 3612 |
+
"learning_rate": 3.376388529782215e-06,
|
| 3613 |
+
"loss": 1.07,
|
| 3614 |
+
"step": 515
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 0.8854568854568855,
|
| 3618 |
+
"grad_norm": 0.6440708637237549,
|
| 3619 |
+
"learning_rate": 3.2789051898507228e-06,
|
| 3620 |
+
"loss": 0.9818,
|
| 3621 |
+
"step": 516
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 0.8871728871728872,
|
| 3625 |
+
"grad_norm": 0.5899673700332642,
|
| 3626 |
+
"learning_rate": 3.1828020943719894e-06,
|
| 3627 |
+
"loss": 0.9473,
|
| 3628 |
+
"step": 517
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 0.8888888888888888,
|
| 3632 |
+
"grad_norm": 0.6313216090202332,
|
| 3633 |
+
"learning_rate": 3.088082082443966e-06,
|
| 3634 |
+
"loss": 0.9787,
|
| 3635 |
+
"step": 518
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 0.8906048906048906,
|
| 3639 |
+
"grad_norm": 0.6282995939254761,
|
| 3640 |
+
"learning_rate": 2.9947479523052548e-06,
|
| 3641 |
+
"loss": 1.0242,
|
| 3642 |
+
"step": 519
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 0.8923208923208923,
|
| 3646 |
+
"grad_norm": 0.6684787273406982,
|
| 3647 |
+
"learning_rate": 2.9028024612524297e-06,
|
| 3648 |
+
"loss": 0.9753,
|
| 3649 |
+
"step": 520
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.894036894036894,
|
| 3653 |
+
"grad_norm": 0.6588143110275269,
|
| 3654 |
+
"learning_rate": 2.8122483255586252e-06,
|
| 3655 |
+
"loss": 0.9871,
|
| 3656 |
+
"step": 521
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 0.8957528957528957,
|
| 3660 |
+
"grad_norm": 0.6997570991516113,
|
| 3661 |
+
"learning_rate": 2.723088220393244e-06,
|
| 3662 |
+
"loss": 0.9316,
|
| 3663 |
+
"step": 522
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 0.8974688974688975,
|
| 3667 |
+
"grad_norm": 0.6888900399208069,
|
| 3668 |
+
"learning_rate": 2.6353247797429535e-06,
|
| 3669 |
+
"loss": 0.9034,
|
| 3670 |
+
"step": 523
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 0.8991848991848992,
|
| 3674 |
+
"grad_norm": 0.6676020622253418,
|
| 3675 |
+
"learning_rate": 2.5489605963338435e-06,
|
| 3676 |
+
"loss": 0.7746,
|
| 3677 |
+
"step": 524
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 0.9009009009009009,
|
| 3681 |
+
"grad_norm": 0.7101950645446777,
|
| 3682 |
+
"learning_rate": 2.463998221554875e-06,
|
| 3683 |
+
"loss": 0.9886,
|
| 3684 |
+
"step": 525
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 0.9026169026169026,
|
| 3688 |
+
"grad_norm": 0.7177839875221252,
|
| 3689 |
+
"learning_rate": 2.3804401653824693e-06,
|
| 3690 |
+
"loss": 0.8845,
|
| 3691 |
+
"step": 526
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 0.9043329043329044,
|
| 3695 |
+
"grad_norm": 0.7837296724319458,
|
| 3696 |
+
"learning_rate": 2.2982888963063774e-06,
|
| 3697 |
+
"loss": 0.9724,
|
| 3698 |
+
"step": 527
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 0.9060489060489061,
|
| 3702 |
+
"grad_norm": 0.7096304893493652,
|
| 3703 |
+
"learning_rate": 2.2175468412567403e-06,
|
| 3704 |
+
"loss": 0.9332,
|
| 3705 |
+
"step": 528
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.9077649077649078,
|
| 3709 |
+
"grad_norm": 0.7648287415504456,
|
| 3710 |
+
"learning_rate": 2.13821638553241e-06,
|
| 3711 |
+
"loss": 0.8184,
|
| 3712 |
+
"step": 529
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 0.9094809094809094,
|
| 3716 |
+
"grad_norm": 0.7152864336967468,
|
| 3717 |
+
"learning_rate": 2.060299872730459e-06,
|
| 3718 |
+
"loss": 0.8864,
|
| 3719 |
+
"step": 530
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 0.9111969111969112,
|
| 3723 |
+
"grad_norm": 0.7374707460403442,
|
| 3724 |
+
"learning_rate": 1.9837996046769837e-06,
|
| 3725 |
+
"loss": 0.9573,
|
| 3726 |
+
"step": 531
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 0.9129129129129129,
|
| 3730 |
+
"grad_norm": 0.7648698091506958,
|
| 3731 |
+
"learning_rate": 1.908717841359048e-06,
|
| 3732 |
+
"loss": 0.917,
|
| 3733 |
+
"step": 532
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 0.9146289146289146,
|
| 3737 |
+
"grad_norm": 0.7598331570625305,
|
| 3738 |
+
"learning_rate": 1.8350568008579705e-06,
|
| 3739 |
+
"loss": 0.8335,
|
| 3740 |
+
"step": 533
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 0.9163449163449163,
|
| 3744 |
+
"grad_norm": 0.7711589932441711,
|
| 3745 |
+
"learning_rate": 1.762818659283777e-06,
|
| 3746 |
+
"loss": 0.9328,
|
| 3747 |
+
"step": 534
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 0.918060918060918,
|
| 3751 |
+
"grad_norm": 0.7900502681732178,
|
| 3752 |
+
"learning_rate": 1.692005550710901e-06,
|
| 3753 |
+
"loss": 0.9006,
|
| 3754 |
+
"step": 535
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.9197769197769198,
|
| 3758 |
+
"grad_norm": 0.826158881187439,
|
| 3759 |
+
"learning_rate": 1.6226195671151523e-06,
|
| 3760 |
+
"loss": 1.0144,
|
| 3761 |
+
"step": 536
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.9214929214929215,
|
| 3765 |
+
"grad_norm": 0.8326795101165771,
|
| 3766 |
+
"learning_rate": 1.5546627583119088e-06,
|
| 3767 |
+
"loss": 0.9397,
|
| 3768 |
+
"step": 537
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.9232089232089232,
|
| 3772 |
+
"grad_norm": 0.8485985398292542,
|
| 3773 |
+
"learning_rate": 1.4881371318955894e-06,
|
| 3774 |
+
"loss": 0.9444,
|
| 3775 |
+
"step": 538
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.924924924924925,
|
| 3779 |
+
"grad_norm": 0.9028431177139282,
|
| 3780 |
+
"learning_rate": 1.4230446531803e-06,
|
| 3781 |
+
"loss": 1.002,
|
| 3782 |
+
"step": 539
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.9266409266409267,
|
| 3786 |
+
"grad_norm": 0.8997206687927246,
|
| 3787 |
+
"learning_rate": 1.3593872451417966e-06,
|
| 3788 |
+
"loss": 0.8769,
|
| 3789 |
+
"step": 540
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.9283569283569284,
|
| 3793 |
+
"grad_norm": 0.8881352543830872,
|
| 3794 |
+
"learning_rate": 1.2971667883606652e-06,
|
| 3795 |
+
"loss": 0.9384,
|
| 3796 |
+
"step": 541
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.9300729300729301,
|
| 3800 |
+
"grad_norm": 0.9173288941383362,
|
| 3801 |
+
"learning_rate": 1.2363851209667932e-06,
|
| 3802 |
+
"loss": 0.9652,
|
| 3803 |
+
"step": 542
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.9317889317889317,
|
| 3807 |
+
"grad_norm": 1.0023092031478882,
|
| 3808 |
+
"learning_rate": 1.1770440385850401e-06,
|
| 3809 |
+
"loss": 0.9053,
|
| 3810 |
+
"step": 543
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.9335049335049335,
|
| 3814 |
+
"grad_norm": 0.9619191884994507,
|
| 3815 |
+
"learning_rate": 1.1191452942821922e-06,
|
| 3816 |
+
"loss": 0.9865,
|
| 3817 |
+
"step": 544
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.9352209352209352,
|
| 3821 |
+
"grad_norm": 0.965644896030426,
|
| 3822 |
+
"learning_rate": 1.062690598515187e-06,
|
| 3823 |
+
"loss": 0.8441,
|
| 3824 |
+
"step": 545
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.9369369369369369,
|
| 3828 |
+
"grad_norm": 0.9868291020393372,
|
| 3829 |
+
"learning_rate": 1.0076816190805749e-06,
|
| 3830 |
+
"loss": 0.9169,
|
| 3831 |
+
"step": 546
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.9386529386529386,
|
| 3835 |
+
"grad_norm": 1.7749027013778687,
|
| 3836 |
+
"learning_rate": 9.54119981065238e-07,
|
| 3837 |
+
"loss": 0.9687,
|
| 3838 |
+
"step": 547
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 0.9403689403689404,
|
| 3842 |
+
"grad_norm": 1.1022123098373413,
|
| 3843 |
+
"learning_rate": 9.020072667984159e-07,
|
| 3844 |
+
"loss": 0.9759,
|
| 3845 |
+
"step": 548
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 0.9420849420849421,
|
| 3849 |
+
"grad_norm": 1.136970043182373,
|
| 3850 |
+
"learning_rate": 8.513450158049108e-07,
|
| 3851 |
+
"loss": 0.9549,
|
| 3852 |
+
"step": 549
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 0.9438009438009438,
|
| 3856 |
+
"grad_norm": 1.3609838485717773,
|
| 3857 |
+
"learning_rate": 8.021347247596511e-07,
|
| 3858 |
+
"loss": 0.9802,
|
| 3859 |
+
"step": 550
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 0.9455169455169455,
|
| 3863 |
+
"grad_norm": 0.3996894657611847,
|
| 3864 |
+
"learning_rate": 7.543778474434438e-07,
|
| 3865 |
+
"loss": 1.0867,
|
| 3866 |
+
"step": 551
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 0.9472329472329473,
|
| 3870 |
+
"grad_norm": 0.43260958790779114,
|
| 3871 |
+
"learning_rate": 7.080757947000582e-07,
|
| 3872 |
+
"loss": 1.0586,
|
| 3873 |
+
"step": 552
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.948948948948949,
|
| 3877 |
+
"grad_norm": 0.4378548860549927,
|
| 3878 |
+
"learning_rate": 6.632299343945103e-07,
|
| 3879 |
+
"loss": 1.0392,
|
| 3880 |
+
"step": 553
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 0.9506649506649507,
|
| 3884 |
+
"grad_norm": 0.5156790018081665,
|
| 3885 |
+
"learning_rate": 6.198415913726718e-07,
|
| 3886 |
+
"loss": 1.0315,
|
| 3887 |
+
"step": 554
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 0.9523809523809523,
|
| 3891 |
+
"grad_norm": 0.5243450999259949,
|
| 3892 |
+
"learning_rate": 5.779120474221522e-07,
|
| 3893 |
+
"loss": 0.9634,
|
| 3894 |
+
"step": 555
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 0.954096954096954,
|
| 3898 |
+
"grad_norm": 0.5240376591682434,
|
| 3899 |
+
"learning_rate": 5.374425412343898e-07,
|
| 3900 |
+
"loss": 0.9457,
|
| 3901 |
+
"step": 556
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 0.9558129558129558,
|
| 3905 |
+
"grad_norm": 0.5501764416694641,
|
| 3906 |
+
"learning_rate": 4.984342683680809e-07,
|
| 3907 |
+
"loss": 0.8232,
|
| 3908 |
+
"step": 557
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 0.9575289575289575,
|
| 3912 |
+
"grad_norm": 0.5496497750282288,
|
| 3913 |
+
"learning_rate": 4.608883812138698e-07,
|
| 3914 |
+
"loss": 0.9965,
|
| 3915 |
+
"step": 558
|
| 3916 |
+
},
|
| 3917 |
+
{
|
| 3918 |
+
"epoch": 0.9592449592449592,
|
| 3919 |
+
"grad_norm": 0.5370264649391174,
|
| 3920 |
+
"learning_rate": 4.2480598896028624e-07,
|
| 3921 |
+
"loss": 0.9317,
|
| 3922 |
+
"step": 559
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 0.960960960960961,
|
| 3926 |
+
"grad_norm": 0.5712248086929321,
|
| 3927 |
+
"learning_rate": 3.9018815756098893e-07,
|
| 3928 |
+
"loss": 0.9311,
|
| 3929 |
+
"step": 560
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 0.9626769626769627,
|
| 3933 |
+
"grad_norm": 0.6146517992019653,
|
| 3934 |
+
"learning_rate": 3.570359097032516e-07,
|
| 3935 |
+
"loss": 0.8991,
|
| 3936 |
+
"step": 561
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 0.9643929643929644,
|
| 3940 |
+
"grad_norm": 0.581969141960144,
|
| 3941 |
+
"learning_rate": 3.2535022477779844e-07,
|
| 3942 |
+
"loss": 0.8634,
|
| 3943 |
+
"step": 562
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 0.9661089661089661,
|
| 3947 |
+
"grad_norm": 0.6420133709907532,
|
| 3948 |
+
"learning_rate": 2.9513203884981577e-07,
|
| 3949 |
+
"loss": 0.9863,
|
| 3950 |
+
"step": 563
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 0.9678249678249679,
|
| 3954 |
+
"grad_norm": 0.6845783591270447,
|
| 3955 |
+
"learning_rate": 2.663822446313469e-07,
|
| 3956 |
+
"loss": 0.9438,
|
| 3957 |
+
"step": 564
|
| 3958 |
+
},
|
| 3959 |
+
{
|
| 3960 |
+
"epoch": 0.9695409695409696,
|
| 3961 |
+
"grad_norm": 0.6241647005081177,
|
| 3962 |
+
"learning_rate": 2.3910169145487936e-07,
|
| 3963 |
+
"loss": 0.7631,
|
| 3964 |
+
"step": 565
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 0.9712569712569713,
|
| 3968 |
+
"grad_norm": 0.648339033126831,
|
| 3969 |
+
"learning_rate": 2.1329118524827662e-07,
|
| 3970 |
+
"loss": 0.8195,
|
| 3971 |
+
"step": 566
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 0.972972972972973,
|
| 3975 |
+
"grad_norm": 0.6824851632118225,
|
| 3976 |
+
"learning_rate": 1.889514885109689e-07,
|
| 3977 |
+
"loss": 0.9853,
|
| 3978 |
+
"step": 567
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 0.9746889746889746,
|
| 3982 |
+
"grad_norm": 0.7320943474769592,
|
| 3983 |
+
"learning_rate": 1.6608332029141582e-07,
|
| 3984 |
+
"loss": 0.8781,
|
| 3985 |
+
"step": 568
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 0.9764049764049764,
|
| 3989 |
+
"grad_norm": 0.6736683249473572,
|
| 3990 |
+
"learning_rate": 1.4468735616587904e-07,
|
| 3991 |
+
"loss": 0.8924,
|
| 3992 |
+
"step": 569
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 0.9781209781209781,
|
| 3996 |
+
"grad_norm": 0.7399722933769226,
|
| 3997 |
+
"learning_rate": 1.2476422821844913e-07,
|
| 3998 |
+
"loss": 0.8944,
|
| 3999 |
+
"step": 570
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 0.9798369798369798,
|
| 4003 |
+
"grad_norm": 0.7640992403030396,
|
| 4004 |
+
"learning_rate": 1.0631452502237737e-07,
|
| 4005 |
+
"loss": 0.973,
|
| 4006 |
+
"step": 571
|
| 4007 |
+
},
|
| 4008 |
+
{
|
| 4009 |
+
"epoch": 0.9815529815529815,
|
| 4010 |
+
"grad_norm": 0.8096172213554382,
|
| 4011 |
+
"learning_rate": 8.933879162270065e-08,
|
| 4012 |
+
"loss": 0.8524,
|
| 4013 |
+
"step": 572
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 0.9832689832689833,
|
| 4017 |
+
"grad_norm": 0.7598038911819458,
|
| 4018 |
+
"learning_rate": 7.383752952010992e-08,
|
| 4019 |
+
"loss": 0.9069,
|
| 4020 |
+
"step": 573
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 0.984984984984985,
|
| 4024 |
+
"grad_norm": 0.7837955951690674,
|
| 4025 |
+
"learning_rate": 5.981119665617319e-08,
|
| 4026 |
+
"loss": 0.887,
|
| 4027 |
+
"step": 574
|
| 4028 |
+
},
|
| 4029 |
+
{
|
| 4030 |
+
"epoch": 0.9867009867009867,
|
| 4031 |
+
"grad_norm": 0.8999657034873962,
|
| 4032 |
+
"learning_rate": 4.7260207399774105e-08,
|
| 4033 |
+
"loss": 1.0129,
|
| 4034 |
+
"step": 575
|
| 4035 |
+
},
|
| 4036 |
+
{
|
| 4037 |
+
"epoch": 0.9884169884169884,
|
| 4038 |
+
"grad_norm": 0.8845619559288025,
|
| 4039 |
+
"learning_rate": 3.618493253489397e-08,
|
| 4040 |
+
"loss": 0.8977,
|
| 4041 |
+
"step": 576
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 0.9901329901329902,
|
| 4045 |
+
"grad_norm": 0.8886672854423523,
|
| 4046 |
+
"learning_rate": 2.6585699249642716e-08,
|
| 4047 |
+
"loss": 0.9296,
|
| 4048 |
+
"step": 577
|
| 4049 |
+
},
|
| 4050 |
+
{
|
| 4051 |
+
"epoch": 0.9918489918489919,
|
| 4052 |
+
"grad_norm": 0.9511525630950928,
|
| 4053 |
+
"learning_rate": 1.8462791126588886e-08,
|
| 4054 |
+
"loss": 0.8848,
|
| 4055 |
+
"step": 578
|
| 4056 |
+
},
|
| 4057 |
+
{
|
| 4058 |
+
"epoch": 0.9935649935649936,
|
| 4059 |
+
"grad_norm": 0.9490180611610413,
|
| 4060 |
+
"learning_rate": 1.181644813441074e-08,
|
| 4061 |
+
"loss": 0.9306,
|
| 4062 |
+
"step": 579
|
| 4063 |
+
},
|
| 4064 |
+
{
|
| 4065 |
+
"epoch": 0.9952809952809952,
|
| 4066 |
+
"grad_norm": 0.9856688380241394,
|
| 4067 |
+
"learning_rate": 6.646866620768633e-09,
|
| 4068 |
+
"loss": 0.9884,
|
| 4069 |
+
"step": 580
|
| 4070 |
+
},
|
| 4071 |
+
{
|
| 4072 |
+
"epoch": 0.996996996996997,
|
| 4073 |
+
"grad_norm": 1.101925015449524,
|
| 4074 |
+
"learning_rate": 2.9541993065373976e-09,
|
| 4075 |
+
"loss": 0.9484,
|
| 4076 |
+
"step": 581
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 0.9987129987129987,
|
| 4080 |
+
"grad_norm": 1.1237503290176392,
|
| 4081 |
+
"learning_rate": 7.385552812710917e-10,
|
| 4082 |
+
"loss": 0.8525,
|
| 4083 |
+
"step": 582
|
| 4084 |
+
},
|
| 4085 |
+
{
|
| 4086 |
+
"epoch": 0.9987129987129987,
|
| 4087 |
+
"eval_loss": 0.9280687570571899,
|
| 4088 |
+
"eval_runtime": 25.5571,
|
| 4089 |
+
"eval_samples_per_second": 19.212,
|
| 4090 |
+
"eval_steps_per_second": 4.813,
|
| 4091 |
+
"step": 582
|
| 4092 |
+
},
|
| 4093 |
+
{
|
| 4094 |
+
"epoch": 1.0004290004290004,
|
| 4095 |
+
"grad_norm": 1.515831470489502,
|
| 4096 |
+
"learning_rate": 0.0,
|
| 4097 |
+
"loss": 1.0995,
|
| 4098 |
+
"step": 583
|
| 4099 |
}
|
| 4100 |
],
|
| 4101 |
"logging_steps": 1,
|
|
|
|
| 4110 |
"should_evaluate": false,
|
| 4111 |
"should_log": false,
|
| 4112 |
"should_save": true,
|
| 4113 |
+
"should_training_stop": true
|
| 4114 |
},
|
| 4115 |
"attributes": {}
|
| 4116 |
}
|
| 4117 |
},
|
| 4118 |
+
"total_flos": 3.9366869180114534e+17,
|
| 4119 |
"train_batch_size": 4,
|
| 4120 |
"trial_name": null,
|
| 4121 |
"trial_params": null
|