Training in progress, step 563, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 101752088
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8c595720a41e9384906f2d3e480d3aa304689f4e0612ea898704da5876df4ce
|
| 3 |
size 101752088
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 203713238
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa85267159afa7ee96961d8e086ff9125093d3784893802d3ad17d9a528ea772
|
| 3 |
size 203713238
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da5aca99bcde1ec8b0dc9a1dd61af6a832b8ca17d5e8974363414a00fe156561
|
| 3 |
size 14512
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14512
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6355acd2ed897b92e4c2ba4445c30b5ccb7ab5d77d60dc3141cf3e52bd674a29
|
| 3 |
size 14512
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1aa6a07fde8b7b9172b2dbbfa971ec598626e705bbfe8cea4899774e3eba905a
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 141,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3000,6 +3000,986 @@
|
|
| 3000 |
"eval_samples_per_second": 6.06,
|
| 3001 |
"eval_steps_per_second": 1.52,
|
| 3002 |
"step": 423
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3003 |
}
|
| 3004 |
],
|
| 3005 |
"logging_steps": 1,
|
|
@@ -3014,12 +3994,12 @@
|
|
| 3014 |
"should_evaluate": false,
|
| 3015 |
"should_log": false,
|
| 3016 |
"should_save": true,
|
| 3017 |
-
"should_training_stop":
|
| 3018 |
},
|
| 3019 |
"attributes": {}
|
| 3020 |
}
|
| 3021 |
},
|
| 3022 |
-
"total_flos":
|
| 3023 |
"train_batch_size": 2,
|
| 3024 |
"trial_name": null,
|
| 3025 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9997780244173141,
|
| 5 |
"eval_steps": 141,
|
| 6 |
+
"global_step": 563,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3000 |
"eval_samples_per_second": 6.06,
|
| 3001 |
"eval_steps_per_second": 1.52,
|
| 3002 |
"step": 423
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 0.7529411764705882,
|
| 3006 |
+
"grad_norm": 0.44857800006866455,
|
| 3007 |
+
"learning_rate": 3.4094199810279924e-05,
|
| 3008 |
+
"loss": 0.6288,
|
| 3009 |
+
"step": 424
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.7547169811320755,
|
| 3013 |
+
"grad_norm": 0.564264178276062,
|
| 3014 |
+
"learning_rate": 3.363486013983788e-05,
|
| 3015 |
+
"loss": 0.784,
|
| 3016 |
+
"step": 425
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 0.7564927857935627,
|
| 3020 |
+
"grad_norm": 0.26752522587776184,
|
| 3021 |
+
"learning_rate": 3.317800934775696e-05,
|
| 3022 |
+
"loss": 0.8435,
|
| 3023 |
+
"step": 426
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 0.7582685904550499,
|
| 3027 |
+
"grad_norm": 0.2690868377685547,
|
| 3028 |
+
"learning_rate": 3.2723664567219626e-05,
|
| 3029 |
+
"loss": 0.8125,
|
| 3030 |
+
"step": 427
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 0.7600443951165372,
|
| 3034 |
+
"grad_norm": 0.28497472405433655,
|
| 3035 |
+
"learning_rate": 3.227184283742591e-05,
|
| 3036 |
+
"loss": 0.7349,
|
| 3037 |
+
"step": 428
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 0.7618201997780244,
|
| 3041 |
+
"grad_norm": 0.3184243142604828,
|
| 3042 |
+
"learning_rate": 3.182256110295437e-05,
|
| 3043 |
+
"loss": 0.8037,
|
| 3044 |
+
"step": 429
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.7635960044395117,
|
| 3048 |
+
"grad_norm": 0.29851233959198,
|
| 3049 |
+
"learning_rate": 3.137583621312665e-05,
|
| 3050 |
+
"loss": 0.7631,
|
| 3051 |
+
"step": 430
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 0.7653718091009989,
|
| 3055 |
+
"grad_norm": 0.31429582834243774,
|
| 3056 |
+
"learning_rate": 3.093168492137557e-05,
|
| 3057 |
+
"loss": 0.7446,
|
| 3058 |
+
"step": 431
|
| 3059 |
+
},
|
| 3060 |
+
{
|
| 3061 |
+
"epoch": 0.7671476137624861,
|
| 3062 |
+
"grad_norm": 0.3191598653793335,
|
| 3063 |
+
"learning_rate": 3.0490123884616796e-05,
|
| 3064 |
+
"loss": 0.7301,
|
| 3065 |
+
"step": 432
|
| 3066 |
+
},
|
| 3067 |
+
{
|
| 3068 |
+
"epoch": 0.7689234184239734,
|
| 3069 |
+
"grad_norm": 0.32959750294685364,
|
| 3070 |
+
"learning_rate": 3.0051169662624225e-05,
|
| 3071 |
+
"loss": 0.7891,
|
| 3072 |
+
"step": 433
|
| 3073 |
+
},
|
| 3074 |
+
{
|
| 3075 |
+
"epoch": 0.7706992230854606,
|
| 3076 |
+
"grad_norm": 0.306445449590683,
|
| 3077 |
+
"learning_rate": 2.9614838717408867e-05,
|
| 3078 |
+
"loss": 0.7205,
|
| 3079 |
+
"step": 434
|
| 3080 |
+
},
|
| 3081 |
+
{
|
| 3082 |
+
"epoch": 0.7724750277469479,
|
| 3083 |
+
"grad_norm": 0.3152560889720917,
|
| 3084 |
+
"learning_rate": 2.9181147412601562e-05,
|
| 3085 |
+
"loss": 0.7436,
|
| 3086 |
+
"step": 435
|
| 3087 |
+
},
|
| 3088 |
+
{
|
| 3089 |
+
"epoch": 0.7742508324084351,
|
| 3090 |
+
"grad_norm": 0.3343624174594879,
|
| 3091 |
+
"learning_rate": 2.8750112012839214e-05,
|
| 3092 |
+
"loss": 0.7004,
|
| 3093 |
+
"step": 436
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 0.7760266370699223,
|
| 3097 |
+
"grad_norm": 0.3389514982700348,
|
| 3098 |
+
"learning_rate": 2.8321748683154893e-05,
|
| 3099 |
+
"loss": 0.7037,
|
| 3100 |
+
"step": 437
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 0.7778024417314096,
|
| 3104 |
+
"grad_norm": 0.36133134365081787,
|
| 3105 |
+
"learning_rate": 2.789607348837153e-05,
|
| 3106 |
+
"loss": 0.7648,
|
| 3107 |
+
"step": 438
|
| 3108 |
+
},
|
| 3109 |
+
{
|
| 3110 |
+
"epoch": 0.7795782463928967,
|
| 3111 |
+
"grad_norm": 0.3495211601257324,
|
| 3112 |
+
"learning_rate": 2.7473102392499518e-05,
|
| 3113 |
+
"loss": 0.7668,
|
| 3114 |
+
"step": 439
|
| 3115 |
+
},
|
| 3116 |
+
{
|
| 3117 |
+
"epoch": 0.781354051054384,
|
| 3118 |
+
"grad_norm": 0.3677636384963989,
|
| 3119 |
+
"learning_rate": 2.7052851258137935e-05,
|
| 3120 |
+
"loss": 0.7267,
|
| 3121 |
+
"step": 440
|
| 3122 |
+
},
|
| 3123 |
+
{
|
| 3124 |
+
"epoch": 0.7831298557158712,
|
| 3125 |
+
"grad_norm": 0.36717966198921204,
|
| 3126 |
+
"learning_rate": 2.6635335845879737e-05,
|
| 3127 |
+
"loss": 0.7577,
|
| 3128 |
+
"step": 441
|
| 3129 |
+
},
|
| 3130 |
+
{
|
| 3131 |
+
"epoch": 0.7849056603773585,
|
| 3132 |
+
"grad_norm": 0.38047298789024353,
|
| 3133 |
+
"learning_rate": 2.622057181372063e-05,
|
| 3134 |
+
"loss": 0.6682,
|
| 3135 |
+
"step": 442
|
| 3136 |
+
},
|
| 3137 |
+
{
|
| 3138 |
+
"epoch": 0.7866814650388457,
|
| 3139 |
+
"grad_norm": 0.37885257601737976,
|
| 3140 |
+
"learning_rate": 2.5808574716471856e-05,
|
| 3141 |
+
"loss": 0.7454,
|
| 3142 |
+
"step": 443
|
| 3143 |
+
},
|
| 3144 |
+
{
|
| 3145 |
+
"epoch": 0.7884572697003329,
|
| 3146 |
+
"grad_norm": 0.37553516030311584,
|
| 3147 |
+
"learning_rate": 2.5399360005176886e-05,
|
| 3148 |
+
"loss": 0.6721,
|
| 3149 |
+
"step": 444
|
| 3150 |
+
},
|
| 3151 |
+
{
|
| 3152 |
+
"epoch": 0.7902330743618202,
|
| 3153 |
+
"grad_norm": 0.42358729243278503,
|
| 3154 |
+
"learning_rate": 2.4992943026531935e-05,
|
| 3155 |
+
"loss": 0.7339,
|
| 3156 |
+
"step": 445
|
| 3157 |
+
},
|
| 3158 |
+
{
|
| 3159 |
+
"epoch": 0.7920088790233074,
|
| 3160 |
+
"grad_norm": 0.3923121392726898,
|
| 3161 |
+
"learning_rate": 2.4589339022310386e-05,
|
| 3162 |
+
"loss": 0.6952,
|
| 3163 |
+
"step": 446
|
| 3164 |
+
},
|
| 3165 |
+
{
|
| 3166 |
+
"epoch": 0.7937846836847947,
|
| 3167 |
+
"grad_norm": 0.39447784423828125,
|
| 3168 |
+
"learning_rate": 2.4188563128791254e-05,
|
| 3169 |
+
"loss": 0.6898,
|
| 3170 |
+
"step": 447
|
| 3171 |
+
},
|
| 3172 |
+
{
|
| 3173 |
+
"epoch": 0.7955604883462819,
|
| 3174 |
+
"grad_norm": 0.4486071467399597,
|
| 3175 |
+
"learning_rate": 2.379063037619146e-05,
|
| 3176 |
+
"loss": 0.6485,
|
| 3177 |
+
"step": 448
|
| 3178 |
+
},
|
| 3179 |
+
{
|
| 3180 |
+
"epoch": 0.7973362930077691,
|
| 3181 |
+
"grad_norm": 0.47466063499450684,
|
| 3182 |
+
"learning_rate": 2.339555568810221e-05,
|
| 3183 |
+
"loss": 0.6961,
|
| 3184 |
+
"step": 449
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 0.7991120976692564,
|
| 3188 |
+
"grad_norm": 0.5662741661071777,
|
| 3189 |
+
"learning_rate": 2.300335388092929e-05,
|
| 3190 |
+
"loss": 0.7295,
|
| 3191 |
+
"step": 450
|
| 3192 |
+
},
|
| 3193 |
+
{
|
| 3194 |
+
"epoch": 0.8008879023307436,
|
| 3195 |
+
"grad_norm": 0.28682488203048706,
|
| 3196 |
+
"learning_rate": 2.2614039663337417e-05,
|
| 3197 |
+
"loss": 0.8068,
|
| 3198 |
+
"step": 451
|
| 3199 |
+
},
|
| 3200 |
+
{
|
| 3201 |
+
"epoch": 0.8026637069922309,
|
| 3202 |
+
"grad_norm": 0.2780812680721283,
|
| 3203 |
+
"learning_rate": 2.222762763569862e-05,
|
| 3204 |
+
"loss": 0.8236,
|
| 3205 |
+
"step": 452
|
| 3206 |
+
},
|
| 3207 |
+
{
|
| 3208 |
+
"epoch": 0.8044395116537181,
|
| 3209 |
+
"grad_norm": 0.29560670256614685,
|
| 3210 |
+
"learning_rate": 2.184413228954468e-05,
|
| 3211 |
+
"loss": 0.7894,
|
| 3212 |
+
"step": 453
|
| 3213 |
+
},
|
| 3214 |
+
{
|
| 3215 |
+
"epoch": 0.8062153163152054,
|
| 3216 |
+
"grad_norm": 0.2896682918071747,
|
| 3217 |
+
"learning_rate": 2.1463568007023704e-05,
|
| 3218 |
+
"loss": 0.7534,
|
| 3219 |
+
"step": 454
|
| 3220 |
+
},
|
| 3221 |
+
{
|
| 3222 |
+
"epoch": 0.8079911209766926,
|
| 3223 |
+
"grad_norm": 0.2878231108188629,
|
| 3224 |
+
"learning_rate": 2.1085949060360654e-05,
|
| 3225 |
+
"loss": 0.7245,
|
| 3226 |
+
"step": 455
|
| 3227 |
+
},
|
| 3228 |
+
{
|
| 3229 |
+
"epoch": 0.8097669256381798,
|
| 3230 |
+
"grad_norm": 0.3214217722415924,
|
| 3231 |
+
"learning_rate": 2.0711289611322204e-05,
|
| 3232 |
+
"loss": 0.7731,
|
| 3233 |
+
"step": 456
|
| 3234 |
+
},
|
| 3235 |
+
{
|
| 3236 |
+
"epoch": 0.8115427302996671,
|
| 3237 |
+
"grad_norm": 0.3201158940792084,
|
| 3238 |
+
"learning_rate": 2.033960371068557e-05,
|
| 3239 |
+
"loss": 0.7475,
|
| 3240 |
+
"step": 457
|
| 3241 |
+
},
|
| 3242 |
+
{
|
| 3243 |
+
"epoch": 0.8133185349611542,
|
| 3244 |
+
"grad_norm": 0.33665937185287476,
|
| 3245 |
+
"learning_rate": 1.9970905297711606e-05,
|
| 3246 |
+
"loss": 0.721,
|
| 3247 |
+
"step": 458
|
| 3248 |
+
},
|
| 3249 |
+
{
|
| 3250 |
+
"epoch": 0.8150943396226416,
|
| 3251 |
+
"grad_norm": 0.3305956721305847,
|
| 3252 |
+
"learning_rate": 1.9605208199621995e-05,
|
| 3253 |
+
"loss": 0.7249,
|
| 3254 |
+
"step": 459
|
| 3255 |
+
},
|
| 3256 |
+
{
|
| 3257 |
+
"epoch": 0.8168701442841287,
|
| 3258 |
+
"grad_norm": 0.3384665548801422,
|
| 3259 |
+
"learning_rate": 1.924252613108073e-05,
|
| 3260 |
+
"loss": 0.724,
|
| 3261 |
+
"step": 460
|
| 3262 |
+
},
|
| 3263 |
+
{
|
| 3264 |
+
"epoch": 0.8186459489456159,
|
| 3265 |
+
"grad_norm": 0.359944224357605,
|
| 3266 |
+
"learning_rate": 1.888287269367979e-05,
|
| 3267 |
+
"loss": 0.7516,
|
| 3268 |
+
"step": 461
|
| 3269 |
+
},
|
| 3270 |
+
{
|
| 3271 |
+
"epoch": 0.8204217536071032,
|
| 3272 |
+
"grad_norm": 0.3357195556163788,
|
| 3273 |
+
"learning_rate": 1.8526261375428955e-05,
|
| 3274 |
+
"loss": 0.7327,
|
| 3275 |
+
"step": 462
|
| 3276 |
+
},
|
| 3277 |
+
{
|
| 3278 |
+
"epoch": 0.8221975582685904,
|
| 3279 |
+
"grad_norm": 0.34272313117980957,
|
| 3280 |
+
"learning_rate": 1.8172705550250092e-05,
|
| 3281 |
+
"loss": 0.7161,
|
| 3282 |
+
"step": 463
|
| 3283 |
+
},
|
| 3284 |
+
{
|
| 3285 |
+
"epoch": 0.8239733629300777,
|
| 3286 |
+
"grad_norm": 0.32507383823394775,
|
| 3287 |
+
"learning_rate": 1.7822218477475494e-05,
|
| 3288 |
+
"loss": 0.6392,
|
| 3289 |
+
"step": 464
|
| 3290 |
+
},
|
| 3291 |
+
{
|
| 3292 |
+
"epoch": 0.8257491675915649,
|
| 3293 |
+
"grad_norm": 0.37359514832496643,
|
| 3294 |
+
"learning_rate": 1.7474813301350666e-05,
|
| 3295 |
+
"loss": 0.7298,
|
| 3296 |
+
"step": 465
|
| 3297 |
+
},
|
| 3298 |
+
{
|
| 3299 |
+
"epoch": 0.8275249722530522,
|
| 3300 |
+
"grad_norm": 0.3348104655742645,
|
| 3301 |
+
"learning_rate": 1.7130503050541368e-05,
|
| 3302 |
+
"loss": 0.6568,
|
| 3303 |
+
"step": 466
|
| 3304 |
+
},
|
| 3305 |
+
{
|
| 3306 |
+
"epoch": 0.8293007769145394,
|
| 3307 |
+
"grad_norm": 0.3460003435611725,
|
| 3308 |
+
"learning_rate": 1.6789300637645e-05,
|
| 3309 |
+
"loss": 0.6742,
|
| 3310 |
+
"step": 467
|
| 3311 |
+
},
|
| 3312 |
+
{
|
| 3313 |
+
"epoch": 0.8310765815760266,
|
| 3314 |
+
"grad_norm": 0.3993259370326996,
|
| 3315 |
+
"learning_rate": 1.6451218858706374e-05,
|
| 3316 |
+
"loss": 0.7365,
|
| 3317 |
+
"step": 468
|
| 3318 |
+
},
|
| 3319 |
+
{
|
| 3320 |
+
"epoch": 0.8328523862375139,
|
| 3321 |
+
"grad_norm": 0.3710460960865021,
|
| 3322 |
+
"learning_rate": 1.6116270392737754e-05,
|
| 3323 |
+
"loss": 0.699,
|
| 3324 |
+
"step": 469
|
| 3325 |
+
},
|
| 3326 |
+
{
|
| 3327 |
+
"epoch": 0.8346281908990011,
|
| 3328 |
+
"grad_norm": 0.41104644536972046,
|
| 3329 |
+
"learning_rate": 1.578446780124344e-05,
|
| 3330 |
+
"loss": 0.7185,
|
| 3331 |
+
"step": 470
|
| 3332 |
+
},
|
| 3333 |
+
{
|
| 3334 |
+
"epoch": 0.8364039955604884,
|
| 3335 |
+
"grad_norm": 0.39822298288345337,
|
| 3336 |
+
"learning_rate": 1.5455823527748626e-05,
|
| 3337 |
+
"loss": 0.6968,
|
| 3338 |
+
"step": 471
|
| 3339 |
+
},
|
| 3340 |
+
{
|
| 3341 |
+
"epoch": 0.8381798002219756,
|
| 3342 |
+
"grad_norm": 0.3909063935279846,
|
| 3343 |
+
"learning_rate": 1.5130349897332763e-05,
|
| 3344 |
+
"loss": 0.6427,
|
| 3345 |
+
"step": 472
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 0.8399556048834628,
|
| 3349 |
+
"grad_norm": 0.39908355474472046,
|
| 3350 |
+
"learning_rate": 1.4808059116167305e-05,
|
| 3351 |
+
"loss": 0.6307,
|
| 3352 |
+
"step": 473
|
| 3353 |
+
},
|
| 3354 |
+
{
|
| 3355 |
+
"epoch": 0.8417314095449501,
|
| 3356 |
+
"grad_norm": 0.4725106954574585,
|
| 3357 |
+
"learning_rate": 1.4488963271057943e-05,
|
| 3358 |
+
"loss": 0.7274,
|
| 3359 |
+
"step": 474
|
| 3360 |
+
},
|
| 3361 |
+
{
|
| 3362 |
+
"epoch": 0.8435072142064373,
|
| 3363 |
+
"grad_norm": 0.58518385887146,
|
| 3364 |
+
"learning_rate": 1.4173074328991377e-05,
|
| 3365 |
+
"loss": 0.7112,
|
| 3366 |
+
"step": 475
|
| 3367 |
+
},
|
| 3368 |
+
{
|
| 3369 |
+
"epoch": 0.8452830188679246,
|
| 3370 |
+
"grad_norm": 0.2664174735546112,
|
| 3371 |
+
"learning_rate": 1.3860404136686411e-05,
|
| 3372 |
+
"loss": 0.8515,
|
| 3373 |
+
"step": 476
|
| 3374 |
+
},
|
| 3375 |
+
{
|
| 3376 |
+
"epoch": 0.8470588235294118,
|
| 3377 |
+
"grad_norm": 0.30460578203201294,
|
| 3378 |
+
"learning_rate": 1.355096442014977e-05,
|
| 3379 |
+
"loss": 0.8107,
|
| 3380 |
+
"step": 477
|
| 3381 |
+
},
|
| 3382 |
+
{
|
| 3383 |
+
"epoch": 0.848834628190899,
|
| 3384 |
+
"grad_norm": 0.28965044021606445,
|
| 3385 |
+
"learning_rate": 1.3244766784236307e-05,
|
| 3386 |
+
"loss": 0.7361,
|
| 3387 |
+
"step": 478
|
| 3388 |
+
},
|
| 3389 |
+
{
|
| 3390 |
+
"epoch": 0.8506104328523862,
|
| 3391 |
+
"grad_norm": 0.3329102396965027,
|
| 3392 |
+
"learning_rate": 1.294182271221377e-05,
|
| 3393 |
+
"loss": 0.7712,
|
| 3394 |
+
"step": 479
|
| 3395 |
+
},
|
| 3396 |
+
{
|
| 3397 |
+
"epoch": 0.8523862375138734,
|
| 3398 |
+
"grad_norm": 0.30333012342453003,
|
| 3399 |
+
"learning_rate": 1.2642143565332154e-05,
|
| 3400 |
+
"loss": 0.7245,
|
| 3401 |
+
"step": 480
|
| 3402 |
+
},
|
| 3403 |
+
{
|
| 3404 |
+
"epoch": 0.8541620421753607,
|
| 3405 |
+
"grad_norm": 0.328744500875473,
|
| 3406 |
+
"learning_rate": 1.2345740582397648e-05,
|
| 3407 |
+
"loss": 0.7557,
|
| 3408 |
+
"step": 481
|
| 3409 |
+
},
|
| 3410 |
+
{
|
| 3411 |
+
"epoch": 0.8559378468368479,
|
| 3412 |
+
"grad_norm": 0.33845219016075134,
|
| 3413 |
+
"learning_rate": 1.2052624879351104e-05,
|
| 3414 |
+
"loss": 0.7857,
|
| 3415 |
+
"step": 482
|
| 3416 |
+
},
|
| 3417 |
+
{
|
| 3418 |
+
"epoch": 0.8577136514983352,
|
| 3419 |
+
"grad_norm": 0.3305346667766571,
|
| 3420 |
+
"learning_rate": 1.176280744885121e-05,
|
| 3421 |
+
"loss": 0.7512,
|
| 3422 |
+
"step": 483
|
| 3423 |
+
},
|
| 3424 |
+
{
|
| 3425 |
+
"epoch": 0.8594894561598224,
|
| 3426 |
+
"grad_norm": 0.340707391500473,
|
| 3427 |
+
"learning_rate": 1.1476299159862203e-05,
|
| 3428 |
+
"loss": 0.7678,
|
| 3429 |
+
"step": 484
|
| 3430 |
+
},
|
| 3431 |
+
{
|
| 3432 |
+
"epoch": 0.8612652608213096,
|
| 3433 |
+
"grad_norm": 0.3646427392959595,
|
| 3434 |
+
"learning_rate": 1.119311075724625e-05,
|
| 3435 |
+
"loss": 0.7473,
|
| 3436 |
+
"step": 485
|
| 3437 |
+
},
|
| 3438 |
+
{
|
| 3439 |
+
"epoch": 0.8630410654827969,
|
| 3440 |
+
"grad_norm": 0.35034942626953125,
|
| 3441 |
+
"learning_rate": 1.09132528613605e-05,
|
| 3442 |
+
"loss": 0.7679,
|
| 3443 |
+
"step": 486
|
| 3444 |
+
},
|
| 3445 |
+
{
|
| 3446 |
+
"epoch": 0.8648168701442841,
|
| 3447 |
+
"grad_norm": 0.35471147298812866,
|
| 3448 |
+
"learning_rate": 1.0636735967658784e-05,
|
| 3449 |
+
"loss": 0.7416,
|
| 3450 |
+
"step": 487
|
| 3451 |
+
},
|
| 3452 |
+
{
|
| 3453 |
+
"epoch": 0.8665926748057714,
|
| 3454 |
+
"grad_norm": 0.3597537875175476,
|
| 3455 |
+
"learning_rate": 1.0363570446297999e-05,
|
| 3456 |
+
"loss": 0.7125,
|
| 3457 |
+
"step": 488
|
| 3458 |
+
},
|
| 3459 |
+
{
|
| 3460 |
+
"epoch": 0.8683684794672586,
|
| 3461 |
+
"grad_norm": 0.35092103481292725,
|
| 3462 |
+
"learning_rate": 1.0093766541749205e-05,
|
| 3463 |
+
"loss": 0.692,
|
| 3464 |
+
"step": 489
|
| 3465 |
+
},
|
| 3466 |
+
{
|
| 3467 |
+
"epoch": 0.8701442841287459,
|
| 3468 |
+
"grad_norm": 0.35275334119796753,
|
| 3469 |
+
"learning_rate": 9.827334372413444e-06,
|
| 3470 |
+
"loss": 0.6683,
|
| 3471 |
+
"step": 490
|
| 3472 |
+
},
|
| 3473 |
+
{
|
| 3474 |
+
"epoch": 0.8719200887902331,
|
| 3475 |
+
"grad_norm": 0.3727843463420868,
|
| 3476 |
+
"learning_rate": 9.564283930242257e-06,
|
| 3477 |
+
"loss": 0.665,
|
| 3478 |
+
"step": 491
|
| 3479 |
+
},
|
| 3480 |
+
{
|
| 3481 |
+
"epoch": 0.8736958934517203,
|
| 3482 |
+
"grad_norm": 0.3570787310600281,
|
| 3483 |
+
"learning_rate": 9.30462508036294e-06,
|
| 3484 |
+
"loss": 0.6736,
|
| 3485 |
+
"step": 492
|
| 3486 |
+
},
|
| 3487 |
+
{
|
| 3488 |
+
"epoch": 0.8754716981132076,
|
| 3489 |
+
"grad_norm": 0.39428988099098206,
|
| 3490 |
+
"learning_rate": 9.048367560708604e-06,
|
| 3491 |
+
"loss": 0.7076,
|
| 3492 |
+
"step": 493
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 0.8772475027746948,
|
| 3496 |
+
"grad_norm": 0.3717636168003082,
|
| 3497 |
+
"learning_rate": 8.795520981652961e-06,
|
| 3498 |
+
"loss": 0.6807,
|
| 3499 |
+
"step": 494
|
| 3500 |
+
},
|
| 3501 |
+
{
|
| 3502 |
+
"epoch": 0.8790233074361821,
|
| 3503 |
+
"grad_norm": 0.4105593264102936,
|
| 3504 |
+
"learning_rate": 8.546094825649908e-06,
|
| 3505 |
+
"loss": 0.7068,
|
| 3506 |
+
"step": 495
|
| 3507 |
+
},
|
| 3508 |
+
{
|
| 3509 |
+
"epoch": 0.8807991120976693,
|
| 3510 |
+
"grad_norm": 0.45720747113227844,
|
| 3511 |
+
"learning_rate": 8.300098446877923e-06,
|
| 3512 |
+
"loss": 0.7189,
|
| 3513 |
+
"step": 496
|
| 3514 |
+
},
|
| 3515 |
+
{
|
| 3516 |
+
"epoch": 0.8825749167591564,
|
| 3517 |
+
"grad_norm": 0.44911620020866394,
|
| 3518 |
+
"learning_rate": 8.05754107088923e-06,
|
| 3519 |
+
"loss": 0.6891,
|
| 3520 |
+
"step": 497
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 0.8843507214206437,
|
| 3524 |
+
"grad_norm": 0.4414433240890503,
|
| 3525 |
+
"learning_rate": 7.818431794263836e-06,
|
| 3526 |
+
"loss": 0.7167,
|
| 3527 |
+
"step": 498
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 0.8861265260821309,
|
| 3531 |
+
"grad_norm": 0.49208080768585205,
|
| 3532 |
+
"learning_rate": 7.582779584268373e-06,
|
| 3533 |
+
"loss": 0.7084,
|
| 3534 |
+
"step": 499
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 0.8879023307436182,
|
| 3538 |
+
"grad_norm": 0.5831857323646545,
|
| 3539 |
+
"learning_rate": 7.350593278519824e-06,
|
| 3540 |
+
"loss": 0.7877,
|
| 3541 |
+
"step": 500
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 0.8896781354051054,
|
| 3545 |
+
"grad_norm": 0.25118499994277954,
|
| 3546 |
+
"learning_rate": 7.121881584654056e-06,
|
| 3547 |
+
"loss": 0.8006,
|
| 3548 |
+
"step": 501
|
| 3549 |
+
},
|
| 3550 |
+
{
|
| 3551 |
+
"epoch": 0.8914539400665926,
|
| 3552 |
+
"grad_norm": 0.2842087745666504,
|
| 3553 |
+
"learning_rate": 6.896653079999249e-06,
|
| 3554 |
+
"loss": 0.7796,
|
| 3555 |
+
"step": 502
|
| 3556 |
+
},
|
| 3557 |
+
{
|
| 3558 |
+
"epoch": 0.8932297447280799,
|
| 3559 |
+
"grad_norm": 0.2935945391654968,
|
| 3560 |
+
"learning_rate": 6.674916211254289e-06,
|
| 3561 |
+
"loss": 0.7614,
|
| 3562 |
+
"step": 503
|
| 3563 |
+
},
|
| 3564 |
+
{
|
| 3565 |
+
"epoch": 0.8950055493895671,
|
| 3566 |
+
"grad_norm": 0.3194078505039215,
|
| 3567 |
+
"learning_rate": 6.45667929417193e-06,
|
| 3568 |
+
"loss": 0.7537,
|
| 3569 |
+
"step": 504
|
| 3570 |
+
},
|
| 3571 |
+
{
|
| 3572 |
+
"epoch": 0.8967813540510544,
|
| 3573 |
+
"grad_norm": 0.32085007429122925,
|
| 3574 |
+
"learning_rate": 6.2419505132469305e-06,
|
| 3575 |
+
"loss": 0.7843,
|
| 3576 |
+
"step": 505
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 0.8985571587125416,
|
| 3580 |
+
"grad_norm": 0.32116949558258057,
|
| 3581 |
+
"learning_rate": 6.030737921409169e-06,
|
| 3582 |
+
"loss": 0.736,
|
| 3583 |
+
"step": 506
|
| 3584 |
+
},
|
| 3585 |
+
{
|
| 3586 |
+
"epoch": 0.9003329633740289,
|
| 3587 |
+
"grad_norm": 0.32136133313179016,
|
| 3588 |
+
"learning_rate": 5.823049439721561e-06,
|
| 3589 |
+
"loss": 0.7388,
|
| 3590 |
+
"step": 507
|
| 3591 |
+
},
|
| 3592 |
+
{
|
| 3593 |
+
"epoch": 0.9021087680355161,
|
| 3594 |
+
"grad_norm": 0.33068105578422546,
|
| 3595 |
+
"learning_rate": 5.618892857083069e-06,
|
| 3596 |
+
"loss": 0.6994,
|
| 3597 |
+
"step": 508
|
| 3598 |
+
},
|
| 3599 |
+
{
|
| 3600 |
+
"epoch": 0.9038845726970033,
|
| 3601 |
+
"grad_norm": 0.3484742343425751,
|
| 3602 |
+
"learning_rate": 5.418275829936537e-06,
|
| 3603 |
+
"loss": 0.7431,
|
| 3604 |
+
"step": 509
|
| 3605 |
+
},
|
| 3606 |
+
{
|
| 3607 |
+
"epoch": 0.9056603773584906,
|
| 3608 |
+
"grad_norm": 0.35299912095069885,
|
| 3609 |
+
"learning_rate": 5.221205881981595e-06,
|
| 3610 |
+
"loss": 0.7568,
|
| 3611 |
+
"step": 510
|
| 3612 |
+
},
|
| 3613 |
+
{
|
| 3614 |
+
"epoch": 0.9074361820199778,
|
| 3615 |
+
"grad_norm": 0.34243056178092957,
|
| 3616 |
+
"learning_rate": 5.02769040389246e-06,
|
| 3617 |
+
"loss": 0.6817,
|
| 3618 |
+
"step": 511
|
| 3619 |
+
},
|
| 3620 |
+
{
|
| 3621 |
+
"epoch": 0.9092119866814651,
|
| 3622 |
+
"grad_norm": 0.38018926978111267,
|
| 3623 |
+
"learning_rate": 4.8377366530408254e-06,
|
| 3624 |
+
"loss": 0.759,
|
| 3625 |
+
"step": 512
|
| 3626 |
+
},
|
| 3627 |
+
{
|
| 3628 |
+
"epoch": 0.9109877913429523,
|
| 3629 |
+
"grad_norm": 0.36619412899017334,
|
| 3630 |
+
"learning_rate": 4.65135175322361e-06,
|
| 3631 |
+
"loss": 0.7202,
|
| 3632 |
+
"step": 513
|
| 3633 |
+
},
|
| 3634 |
+
{
|
| 3635 |
+
"epoch": 0.9127635960044395,
|
| 3636 |
+
"grad_norm": 0.38696765899658203,
|
| 3637 |
+
"learning_rate": 4.468542694395861e-06,
|
| 3638 |
+
"loss": 0.7202,
|
| 3639 |
+
"step": 514
|
| 3640 |
+
},
|
| 3641 |
+
{
|
| 3642 |
+
"epoch": 0.9145394006659268,
|
| 3643 |
+
"grad_norm": 0.39391985535621643,
|
| 3644 |
+
"learning_rate": 4.2893163324085885e-06,
|
| 3645 |
+
"loss": 0.7091,
|
| 3646 |
+
"step": 515
|
| 3647 |
+
},
|
| 3648 |
+
{
|
| 3649 |
+
"epoch": 0.916315205327414,
|
| 3650 |
+
"grad_norm": 0.37037429213523865,
|
| 3651 |
+
"learning_rate": 4.1136793887516345e-06,
|
| 3652 |
+
"loss": 0.6974,
|
| 3653 |
+
"step": 516
|
| 3654 |
+
},
|
| 3655 |
+
{
|
| 3656 |
+
"epoch": 0.9180910099889013,
|
| 3657 |
+
"grad_norm": 0.39087411761283875,
|
| 3658 |
+
"learning_rate": 3.941638450301644e-06,
|
| 3659 |
+
"loss": 0.7328,
|
| 3660 |
+
"step": 517
|
| 3661 |
+
},
|
| 3662 |
+
{
|
| 3663 |
+
"epoch": 0.9198668146503884,
|
| 3664 |
+
"grad_norm": 0.38174766302108765,
|
| 3665 |
+
"learning_rate": 3.7731999690749585e-06,
|
| 3666 |
+
"loss": 0.7184,
|
| 3667 |
+
"step": 518
|
| 3668 |
+
},
|
| 3669 |
+
{
|
| 3670 |
+
"epoch": 0.9216426193118757,
|
| 3671 |
+
"grad_norm": 0.3953598737716675,
|
| 3672 |
+
"learning_rate": 3.6083702619857605e-06,
|
| 3673 |
+
"loss": 0.7121,
|
| 3674 |
+
"step": 519
|
| 3675 |
+
},
|
| 3676 |
+
{
|
| 3677 |
+
"epoch": 0.9234184239733629,
|
| 3678 |
+
"grad_norm": 0.39504143595695496,
|
| 3679 |
+
"learning_rate": 3.447155510609057e-06,
|
| 3680 |
+
"loss": 0.6665,
|
| 3681 |
+
"step": 520
|
| 3682 |
+
},
|
| 3683 |
+
{
|
| 3684 |
+
"epoch": 0.9251942286348501,
|
| 3685 |
+
"grad_norm": 0.4065762162208557,
|
| 3686 |
+
"learning_rate": 3.2895617609489336e-06,
|
| 3687 |
+
"loss": 0.7019,
|
| 3688 |
+
"step": 521
|
| 3689 |
+
},
|
| 3690 |
+
{
|
| 3691 |
+
"epoch": 0.9269700332963374,
|
| 3692 |
+
"grad_norm": 0.39577242732048035,
|
| 3693 |
+
"learning_rate": 3.135594923211771e-06,
|
| 3694 |
+
"loss": 0.6444,
|
| 3695 |
+
"step": 522
|
| 3696 |
+
},
|
| 3697 |
+
{
|
| 3698 |
+
"epoch": 0.9287458379578246,
|
| 3699 |
+
"grad_norm": 0.4378613531589508,
|
| 3700 |
+
"learning_rate": 2.9852607715846193e-06,
|
| 3701 |
+
"loss": 0.7066,
|
| 3702 |
+
"step": 523
|
| 3703 |
+
},
|
| 3704 |
+
{
|
| 3705 |
+
"epoch": 0.9305216426193119,
|
| 3706 |
+
"grad_norm": 0.4423007369041443,
|
| 3707 |
+
"learning_rate": 2.838564944018618e-06,
|
| 3708 |
+
"loss": 0.6555,
|
| 3709 |
+
"step": 524
|
| 3710 |
+
},
|
| 3711 |
+
{
|
| 3712 |
+
"epoch": 0.9322974472807991,
|
| 3713 |
+
"grad_norm": 0.5693342089653015,
|
| 3714 |
+
"learning_rate": 2.6955129420176196e-06,
|
| 3715 |
+
"loss": 0.8152,
|
| 3716 |
+
"step": 525
|
| 3717 |
+
},
|
| 3718 |
+
{
|
| 3719 |
+
"epoch": 0.9340732519422863,
|
| 3720 |
+
"grad_norm": 0.2593931555747986,
|
| 3721 |
+
"learning_rate": 2.556110130431788e-06,
|
| 3722 |
+
"loss": 0.813,
|
| 3723 |
+
"step": 526
|
| 3724 |
+
},
|
| 3725 |
+
{
|
| 3726 |
+
"epoch": 0.9358490566037736,
|
| 3727 |
+
"grad_norm": 0.30711933970451355,
|
| 3728 |
+
"learning_rate": 2.420361737256438e-06,
|
| 3729 |
+
"loss": 0.7564,
|
| 3730 |
+
"step": 527
|
| 3731 |
+
},
|
| 3732 |
+
{
|
| 3733 |
+
"epoch": 0.9376248612652608,
|
| 3734 |
+
"grad_norm": 0.29575708508491516,
|
| 3735 |
+
"learning_rate": 2.288272853436013e-06,
|
| 3736 |
+
"loss": 0.7813,
|
| 3737 |
+
"step": 528
|
| 3738 |
+
},
|
| 3739 |
+
{
|
| 3740 |
+
"epoch": 0.9394006659267481,
|
| 3741 |
+
"grad_norm": 0.3270512521266937,
|
| 3742 |
+
"learning_rate": 2.1598484326730837e-06,
|
| 3743 |
+
"loss": 0.7658,
|
| 3744 |
+
"step": 529
|
| 3745 |
+
},
|
| 3746 |
+
{
|
| 3747 |
+
"epoch": 0.9411764705882353,
|
| 3748 |
+
"grad_norm": 0.3134397268295288,
|
| 3749 |
+
"learning_rate": 2.035093291242607e-06,
|
| 3750 |
+
"loss": 0.7335,
|
| 3751 |
+
"step": 530
|
| 3752 |
+
},
|
| 3753 |
+
{
|
| 3754 |
+
"epoch": 0.9429522752497226,
|
| 3755 |
+
"grad_norm": 0.34165453910827637,
|
| 3756 |
+
"learning_rate": 1.914012107811336e-06,
|
| 3757 |
+
"loss": 0.798,
|
| 3758 |
+
"step": 531
|
| 3759 |
+
},
|
| 3760 |
+
{
|
| 3761 |
+
"epoch": 0.9447280799112098,
|
| 3762 |
+
"grad_norm": 0.33172738552093506,
|
| 3763 |
+
"learning_rate": 1.7966094232622855e-06,
|
| 3764 |
+
"loss": 0.7516,
|
| 3765 |
+
"step": 532
|
| 3766 |
+
},
|
| 3767 |
+
{
|
| 3768 |
+
"epoch": 0.946503884572697,
|
| 3769 |
+
"grad_norm": 0.35185980796813965,
|
| 3770 |
+
"learning_rate": 1.6828896405244988e-06,
|
| 3771 |
+
"loss": 0.7745,
|
| 3772 |
+
"step": 533
|
| 3773 |
+
},
|
| 3774 |
+
{
|
| 3775 |
+
"epoch": 0.9482796892341843,
|
| 3776 |
+
"grad_norm": 0.3368275761604309,
|
| 3777 |
+
"learning_rate": 1.572857024407881e-06,
|
| 3778 |
+
"loss": 0.749,
|
| 3779 |
+
"step": 534
|
| 3780 |
+
},
|
| 3781 |
+
{
|
| 3782 |
+
"epoch": 0.9500554938956715,
|
| 3783 |
+
"grad_norm": 0.3556804656982422,
|
| 3784 |
+
"learning_rate": 1.466515701443294e-06,
|
| 3785 |
+
"loss": 0.737,
|
| 3786 |
+
"step": 535
|
| 3787 |
+
},
|
| 3788 |
+
{
|
| 3789 |
+
"epoch": 0.9518312985571588,
|
| 3790 |
+
"grad_norm": 0.35035377740859985,
|
| 3791 |
+
"learning_rate": 1.3638696597277679e-06,
|
| 3792 |
+
"loss": 0.716,
|
| 3793 |
+
"step": 536
|
| 3794 |
+
},
|
| 3795 |
+
{
|
| 3796 |
+
"epoch": 0.953607103218646,
|
| 3797 |
+
"grad_norm": 0.356507807970047,
|
| 3798 |
+
"learning_rate": 1.2649227487749548e-06,
|
| 3799 |
+
"loss": 0.7292,
|
| 3800 |
+
"step": 537
|
| 3801 |
+
},
|
| 3802 |
+
{
|
| 3803 |
+
"epoch": 0.9553829078801331,
|
| 3804 |
+
"grad_norm": 0.3645875155925751,
|
| 3805 |
+
"learning_rate": 1.1696786793707781e-06,
|
| 3806 |
+
"loss": 0.7325,
|
| 3807 |
+
"step": 538
|
| 3808 |
+
},
|
| 3809 |
+
{
|
| 3810 |
+
"epoch": 0.9571587125416204,
|
| 3811 |
+
"grad_norm": 0.35591599345207214,
|
| 3812 |
+
"learning_rate": 1.0781410234342094e-06,
|
| 3813 |
+
"loss": 0.7203,
|
| 3814 |
+
"step": 539
|
| 3815 |
+
},
|
| 3816 |
+
{
|
| 3817 |
+
"epoch": 0.9589345172031076,
|
| 3818 |
+
"grad_norm": 0.35885104537010193,
|
| 3819 |
+
"learning_rate": 9.90313213883376e-07,
|
| 3820 |
+
"loss": 0.665,
|
| 3821 |
+
"step": 540
|
| 3822 |
+
},
|
| 3823 |
+
{
|
| 3824 |
+
"epoch": 0.9607103218645949,
|
| 3825 |
+
"grad_norm": 0.38247016072273254,
|
| 3826 |
+
"learning_rate": 9.061985445067756e-07,
|
| 3827 |
+
"loss": 0.6885,
|
| 3828 |
+
"step": 541
|
| 3829 |
+
},
|
| 3830 |
+
{
|
| 3831 |
+
"epoch": 0.9624861265260821,
|
| 3832 |
+
"grad_norm": 0.38850679993629456,
|
| 3833 |
+
"learning_rate": 8.258001698397744e-07,
|
| 3834 |
+
"loss": 0.707,
|
| 3835 |
+
"step": 542
|
| 3836 |
+
},
|
| 3837 |
+
{
|
| 3838 |
+
"epoch": 0.9642619311875694,
|
| 3839 |
+
"grad_norm": 0.3912898004055023,
|
| 3840 |
+
"learning_rate": 7.491211050462798e-07,
|
| 3841 |
+
"loss": 0.6818,
|
| 3842 |
+
"step": 543
|
| 3843 |
+
},
|
| 3844 |
+
{
|
| 3845 |
+
"epoch": 0.9660377358490566,
|
| 3846 |
+
"grad_norm": 0.3983571529388428,
|
| 3847 |
+
"learning_rate": 6.761642258056978e-07,
|
| 3848 |
+
"loss": 0.6786,
|
| 3849 |
+
"step": 544
|
| 3850 |
+
},
|
| 3851 |
+
{
|
| 3852 |
+
"epoch": 0.9678135405105438,
|
| 3853 |
+
"grad_norm": 0.4406982660293579,
|
| 3854 |
+
"learning_rate": 6.069322682050516e-07,
|
| 3855 |
+
"loss": 0.6564,
|
| 3856 |
+
"step": 545
|
| 3857 |
+
},
|
| 3858 |
+
{
|
| 3859 |
+
"epoch": 0.9695893451720311,
|
| 3860 |
+
"grad_norm": 0.38563239574432373,
|
| 3861 |
+
"learning_rate": 5.414278286363761e-07,
|
| 3862 |
+
"loss": 0.5921,
|
| 3863 |
+
"step": 546
|
| 3864 |
+
},
|
| 3865 |
+
{
|
| 3866 |
+
"epoch": 0.9713651498335183,
|
| 3867 |
+
"grad_norm": 0.4474928081035614,
|
| 3868 |
+
"learning_rate": 4.796533636993727e-07,
|
| 3869 |
+
"loss": 0.681,
|
| 3870 |
+
"step": 547
|
| 3871 |
+
},
|
| 3872 |
+
{
|
| 3873 |
+
"epoch": 0.9731409544950056,
|
| 3874 |
+
"grad_norm": 0.43392056226730347,
|
| 3875 |
+
"learning_rate": 4.216111901092501e-07,
|
| 3876 |
+
"loss": 0.6673,
|
| 3877 |
+
"step": 548
|
| 3878 |
+
},
|
| 3879 |
+
{
|
| 3880 |
+
"epoch": 0.9749167591564928,
|
| 3881 |
+
"grad_norm": 0.5074283480644226,
|
| 3882 |
+
"learning_rate": 3.6730348460985996e-07,
|
| 3883 |
+
"loss": 0.7363,
|
| 3884 |
+
"step": 549
|
| 3885 |
+
},
|
| 3886 |
+
{
|
| 3887 |
+
"epoch": 0.97669256381798,
|
| 3888 |
+
"grad_norm": 0.5351345539093018,
|
| 3889 |
+
"learning_rate": 3.1673228389204055e-07,
|
| 3890 |
+
"loss": 0.6898,
|
| 3891 |
+
"step": 550
|
| 3892 |
+
},
|
| 3893 |
+
{
|
| 3894 |
+
"epoch": 0.9784683684794673,
|
| 3895 |
+
"grad_norm": 0.2640553414821625,
|
| 3896 |
+
"learning_rate": 2.6989948451726643e-07,
|
| 3897 |
+
"loss": 0.7556,
|
| 3898 |
+
"step": 551
|
| 3899 |
+
},
|
| 3900 |
+
{
|
| 3901 |
+
"epoch": 0.9802441731409545,
|
| 3902 |
+
"grad_norm": 0.28559839725494385,
|
| 3903 |
+
"learning_rate": 2.2680684284650533e-07,
|
| 3904 |
+
"loss": 0.7636,
|
| 3905 |
+
"step": 552
|
| 3906 |
+
},
|
| 3907 |
+
{
|
| 3908 |
+
"epoch": 0.9820199778024418,
|
| 3909 |
+
"grad_norm": 0.3131345510482788,
|
| 3910 |
+
"learning_rate": 1.8745597497433765e-07,
|
| 3911 |
+
"loss": 0.7366,
|
| 3912 |
+
"step": 553
|
| 3913 |
+
},
|
| 3914 |
+
{
|
| 3915 |
+
"epoch": 0.983795782463929,
|
| 3916 |
+
"grad_norm": 0.3156748116016388,
|
| 3917 |
+
"learning_rate": 1.518483566683826e-07,
|
| 3918 |
+
"loss": 0.7377,
|
| 3919 |
+
"step": 554
|
| 3920 |
+
},
|
| 3921 |
+
{
|
| 3922 |
+
"epoch": 0.9855715871254163,
|
| 3923 |
+
"grad_norm": 0.34887486696243286,
|
| 3924 |
+
"learning_rate": 1.199853233138981e-07,
|
| 3925 |
+
"loss": 0.7588,
|
| 3926 |
+
"step": 555
|
| 3927 |
+
},
|
| 3928 |
+
{
|
| 3929 |
+
"epoch": 0.9873473917869034,
|
| 3930 |
+
"grad_norm": 0.3328061103820801,
|
| 3931 |
+
"learning_rate": 9.186806986376529e-08,
|
| 3932 |
+
"loss": 0.744,
|
| 3933 |
+
"step": 556
|
| 3934 |
+
},
|
| 3935 |
+
{
|
| 3936 |
+
"epoch": 0.9891231964483906,
|
| 3937 |
+
"grad_norm": 0.36429402232170105,
|
| 3938 |
+
"learning_rate": 6.749765079363534e-08,
|
| 3939 |
+
"loss": 0.7261,
|
| 3940 |
+
"step": 557
|
| 3941 |
+
},
|
| 3942 |
+
{
|
| 3943 |
+
"epoch": 0.9908990011098779,
|
| 3944 |
+
"grad_norm": 0.36065101623535156,
|
| 3945 |
+
"learning_rate": 4.687498006236135e-08,
|
| 3946 |
+
"loss": 0.6815,
|
| 3947 |
+
"step": 558
|
| 3948 |
+
},
|
| 3949 |
+
{
|
| 3950 |
+
"epoch": 0.9926748057713651,
|
| 3951 |
+
"grad_norm": 0.3841758966445923,
|
| 3952 |
+
"learning_rate": 3.000083107780327e-08,
|
| 3953 |
+
"loss": 0.7204,
|
| 3954 |
+
"step": 559
|
| 3955 |
+
},
|
| 3956 |
+
{
|
| 3957 |
+
"epoch": 0.9944506104328524,
|
| 3958 |
+
"grad_norm": 0.43167850375175476,
|
| 3959 |
+
"learning_rate": 1.687583666772907e-08,
|
| 3960 |
+
"loss": 0.725,
|
| 3961 |
+
"step": 560
|
| 3962 |
+
},
|
| 3963 |
+
{
|
| 3964 |
+
"epoch": 0.9962264150943396,
|
| 3965 |
+
"grad_norm": 0.4131225347518921,
|
| 3966 |
+
"learning_rate": 7.500489056133652e-09,
|
| 3967 |
+
"loss": 0.6763,
|
| 3968 |
+
"step": 561
|
| 3969 |
+
},
|
| 3970 |
+
{
|
| 3971 |
+
"epoch": 0.9980022197558268,
|
| 3972 |
+
"grad_norm": 0.4707167446613312,
|
| 3973 |
+
"learning_rate": 1.8751398447758306e-09,
|
| 3974 |
+
"loss": 0.6854,
|
| 3975 |
+
"step": 562
|
| 3976 |
+
},
|
| 3977 |
+
{
|
| 3978 |
+
"epoch": 0.9997780244173141,
|
| 3979 |
+
"grad_norm": 0.4748234152793884,
|
| 3980 |
+
"learning_rate": 0.0,
|
| 3981 |
+
"loss": 0.6381,
|
| 3982 |
+
"step": 563
|
| 3983 |
}
|
| 3984 |
],
|
| 3985 |
"logging_steps": 1,
|
|
|
|
| 3994 |
"should_evaluate": false,
|
| 3995 |
"should_log": false,
|
| 3996 |
"should_save": true,
|
| 3997 |
+
"should_training_stop": true
|
| 3998 |
},
|
| 3999 |
"attributes": {}
|
| 4000 |
}
|
| 4001 |
},
|
| 4002 |
+
"total_flos": 7.415142496788808e+17,
|
| 4003 |
"train_batch_size": 2,
|
| 4004 |
"trial_name": null,
|
| 4005 |
"trial_params": null
|