Training in progress, step 423, 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:efedcd7712efe5df4242d40d0fc157567550dc57198de0fde11a067a253c3786
|
| 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 sha256:2a3093ef84d124bf4f3a388a3f58cedd89b5fbf3ec80a866e1189f65649a0f5e
|
| 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 sha256:9e52b4ddcd925a725a65812af6610fe4debc708c6e4fc1ee7e0e17160e2a6fc5
|
| 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 sha256:7d9aa8c4c4812086f9a0cd74c7d98dc727224f492c2c8deb8168a9fa04e2846e
|
| 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:b2bb049f58262ac24b66ea8e4bbb35c588cda72b0f20c7495d16197e65e5d114
|
| 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,
|
|
@@ -2005,6 +2005,1001 @@
|
|
| 2005 |
"eval_samples_per_second": 5.989,
|
| 2006 |
"eval_steps_per_second": 1.502,
|
| 2007 |
"step": 282
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2008 |
}
|
| 2009 |
],
|
| 2010 |
"logging_steps": 1,
|
|
@@ -2024,7 +3019,7 @@
|
|
| 2024 |
"attributes": {}
|
| 2025 |
}
|
| 2026 |
},
|
| 2027 |
-
"total_flos":
|
| 2028 |
"train_batch_size": 2,
|
| 2029 |
"trial_name": null,
|
| 2030 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.751165371809101,
|
| 5 |
"eval_steps": 141,
|
| 6 |
+
"global_step": 423,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2005 |
"eval_samples_per_second": 5.989,
|
| 2006 |
"eval_steps_per_second": 1.502,
|
| 2007 |
"step": 282
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.502552719200888,
|
| 2011 |
+
"grad_norm": 0.331061989068985,
|
| 2012 |
+
"learning_rate": 0.00011434168642236964,
|
| 2013 |
+
"loss": 0.8114,
|
| 2014 |
+
"step": 283
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.5043285238623751,
|
| 2018 |
+
"grad_norm": 0.3186919689178467,
|
| 2019 |
+
"learning_rate": 0.00011373535578184082,
|
| 2020 |
+
"loss": 0.7872,
|
| 2021 |
+
"step": 284
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 0.5061043285238623,
|
| 2025 |
+
"grad_norm": 0.3114188611507416,
|
| 2026 |
+
"learning_rate": 0.00011312851002705383,
|
| 2027 |
+
"loss": 0.7311,
|
| 2028 |
+
"step": 285
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.5078801331853496,
|
| 2032 |
+
"grad_norm": 0.3148879408836365,
|
| 2033 |
+
"learning_rate": 0.00011252117191642175,
|
| 2034 |
+
"loss": 0.7311,
|
| 2035 |
+
"step": 286
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.5096559378468368,
|
| 2039 |
+
"grad_norm": 0.3390887379646301,
|
| 2040 |
+
"learning_rate": 0.00011191336422682237,
|
| 2041 |
+
"loss": 0.7773,
|
| 2042 |
+
"step": 287
|
| 2043 |
+
},
|
| 2044 |
+
{
|
| 2045 |
+
"epoch": 0.5114317425083241,
|
| 2046 |
+
"grad_norm": 0.31982842087745667,
|
| 2047 |
+
"learning_rate": 0.00011130510975274409,
|
| 2048 |
+
"loss": 0.7474,
|
| 2049 |
+
"step": 288
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.5132075471698113,
|
| 2053 |
+
"grad_norm": 0.31643104553222656,
|
| 2054 |
+
"learning_rate": 0.00011069643130543084,
|
| 2055 |
+
"loss": 0.7375,
|
| 2056 |
+
"step": 289
|
| 2057 |
+
},
|
| 2058 |
+
{
|
| 2059 |
+
"epoch": 0.5149833518312985,
|
| 2060 |
+
"grad_norm": 0.33758479356765747,
|
| 2061 |
+
"learning_rate": 0.00011008735171202684,
|
| 2062 |
+
"loss": 0.7411,
|
| 2063 |
+
"step": 290
|
| 2064 |
+
},
|
| 2065 |
+
{
|
| 2066 |
+
"epoch": 0.5167591564927858,
|
| 2067 |
+
"grad_norm": 0.324556440114975,
|
| 2068 |
+
"learning_rate": 0.00010947789381472035,
|
| 2069 |
+
"loss": 0.7235,
|
| 2070 |
+
"step": 291
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.518534961154273,
|
| 2074 |
+
"grad_norm": 0.3768496513366699,
|
| 2075 |
+
"learning_rate": 0.00010886808046988717,
|
| 2076 |
+
"loss": 0.7618,
|
| 2077 |
+
"step": 292
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 0.5203107658157603,
|
| 2081 |
+
"grad_norm": 0.34034618735313416,
|
| 2082 |
+
"learning_rate": 0.00010825793454723325,
|
| 2083 |
+
"loss": 0.7426,
|
| 2084 |
+
"step": 293
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.5220865704772475,
|
| 2088 |
+
"grad_norm": 0.3409979045391083,
|
| 2089 |
+
"learning_rate": 0.00010764747892893723,
|
| 2090 |
+
"loss": 0.7327,
|
| 2091 |
+
"step": 294
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.5238623751387348,
|
| 2095 |
+
"grad_norm": 0.35839787125587463,
|
| 2096 |
+
"learning_rate": 0.00010703673650879218,
|
| 2097 |
+
"loss": 0.7057,
|
| 2098 |
+
"step": 295
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 0.525638179800222,
|
| 2102 |
+
"grad_norm": 0.3807874023914337,
|
| 2103 |
+
"learning_rate": 0.00010642573019134703,
|
| 2104 |
+
"loss": 0.7225,
|
| 2105 |
+
"step": 296
|
| 2106 |
+
},
|
| 2107 |
+
{
|
| 2108 |
+
"epoch": 0.5274139844617092,
|
| 2109 |
+
"grad_norm": 0.4682140648365021,
|
| 2110 |
+
"learning_rate": 0.00010581448289104758,
|
| 2111 |
+
"loss": 0.715,
|
| 2112 |
+
"step": 297
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.5291897891231965,
|
| 2116 |
+
"grad_norm": 0.4261273145675659,
|
| 2117 |
+
"learning_rate": 0.00010520301753137724,
|
| 2118 |
+
"loss": 0.7239,
|
| 2119 |
+
"step": 298
|
| 2120 |
+
},
|
| 2121 |
+
{
|
| 2122 |
+
"epoch": 0.5309655937846837,
|
| 2123 |
+
"grad_norm": 0.4854682981967926,
|
| 2124 |
+
"learning_rate": 0.00010459135704399718,
|
| 2125 |
+
"loss": 0.7304,
|
| 2126 |
+
"step": 299
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.532741398446171,
|
| 2130 |
+
"grad_norm": 0.6740989685058594,
|
| 2131 |
+
"learning_rate": 0.00010397952436788642,
|
| 2132 |
+
"loss": 0.8604,
|
| 2133 |
+
"step": 300
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.5345172031076582,
|
| 2137 |
+
"grad_norm": 0.2903907299041748,
|
| 2138 |
+
"learning_rate": 0.00010336754244848157,
|
| 2139 |
+
"loss": 0.8551,
|
| 2140 |
+
"step": 301
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 0.5362930077691453,
|
| 2144 |
+
"grad_norm": 0.28648582100868225,
|
| 2145 |
+
"learning_rate": 0.00010275543423681621,
|
| 2146 |
+
"loss": 0.7958,
|
| 2147 |
+
"step": 302
|
| 2148 |
+
},
|
| 2149 |
+
{
|
| 2150 |
+
"epoch": 0.5380688124306326,
|
| 2151 |
+
"grad_norm": 0.33123767375946045,
|
| 2152 |
+
"learning_rate": 0.00010214322268866032,
|
| 2153 |
+
"loss": 0.7853,
|
| 2154 |
+
"step": 303
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.5398446170921198,
|
| 2158 |
+
"grad_norm": 0.31327784061431885,
|
| 2159 |
+
"learning_rate": 0.00010153093076365923,
|
| 2160 |
+
"loss": 0.7856,
|
| 2161 |
+
"step": 304
|
| 2162 |
+
},
|
| 2163 |
+
{
|
| 2164 |
+
"epoch": 0.5416204217536071,
|
| 2165 |
+
"grad_norm": 0.3101854622364044,
|
| 2166 |
+
"learning_rate": 0.00010091858142447265,
|
| 2167 |
+
"loss": 0.7694,
|
| 2168 |
+
"step": 305
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.5433962264150943,
|
| 2172 |
+
"grad_norm": 0.3217926621437073,
|
| 2173 |
+
"learning_rate": 0.00010030619763591347,
|
| 2174 |
+
"loss": 0.7899,
|
| 2175 |
+
"step": 306
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.5451720310765816,
|
| 2179 |
+
"grad_norm": 0.33827194571495056,
|
| 2180 |
+
"learning_rate": 9.969380236408656e-05,
|
| 2181 |
+
"loss": 0.8088,
|
| 2182 |
+
"step": 307
|
| 2183 |
+
},
|
| 2184 |
+
{
|
| 2185 |
+
"epoch": 0.5469478357380688,
|
| 2186 |
+
"grad_norm": 0.32632124423980713,
|
| 2187 |
+
"learning_rate": 9.908141857552737e-05,
|
| 2188 |
+
"loss": 0.769,
|
| 2189 |
+
"step": 308
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 0.548723640399556,
|
| 2193 |
+
"grad_norm": 0.3152617812156677,
|
| 2194 |
+
"learning_rate": 9.846906923634079e-05,
|
| 2195 |
+
"loss": 0.7804,
|
| 2196 |
+
"step": 309
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.5504994450610433,
|
| 2200 |
+
"grad_norm": 0.33337536454200745,
|
| 2201 |
+
"learning_rate": 9.78567773113397e-05,
|
| 2202 |
+
"loss": 0.7379,
|
| 2203 |
+
"step": 310
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 0.5522752497225305,
|
| 2207 |
+
"grad_norm": 0.3020349144935608,
|
| 2208 |
+
"learning_rate": 9.724456576318381e-05,
|
| 2209 |
+
"loss": 0.7146,
|
| 2210 |
+
"step": 311
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.5540510543840178,
|
| 2214 |
+
"grad_norm": 0.34656378626823425,
|
| 2215 |
+
"learning_rate": 9.663245755151846e-05,
|
| 2216 |
+
"loss": 0.7437,
|
| 2217 |
+
"step": 312
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.555826859045505,
|
| 2221 |
+
"grad_norm": 0.3417186737060547,
|
| 2222 |
+
"learning_rate": 9.602047563211359e-05,
|
| 2223 |
+
"loss": 0.7472,
|
| 2224 |
+
"step": 313
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 0.5576026637069922,
|
| 2228 |
+
"grad_norm": 0.34442222118377686,
|
| 2229 |
+
"learning_rate": 9.540864295600283e-05,
|
| 2230 |
+
"loss": 0.7426,
|
| 2231 |
+
"step": 314
|
| 2232 |
+
},
|
| 2233 |
+
{
|
| 2234 |
+
"epoch": 0.5593784683684795,
|
| 2235 |
+
"grad_norm": 0.3521478772163391,
|
| 2236 |
+
"learning_rate": 9.479698246862276e-05,
|
| 2237 |
+
"loss": 0.7522,
|
| 2238 |
+
"step": 315
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 0.5611542730299667,
|
| 2242 |
+
"grad_norm": 0.3358227014541626,
|
| 2243 |
+
"learning_rate": 9.418551710895243e-05,
|
| 2244 |
+
"loss": 0.7454,
|
| 2245 |
+
"step": 316
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 0.562930077691454,
|
| 2249 |
+
"grad_norm": 0.343226820230484,
|
| 2250 |
+
"learning_rate": 9.357426980865301e-05,
|
| 2251 |
+
"loss": 0.7341,
|
| 2252 |
+
"step": 317
|
| 2253 |
+
},
|
| 2254 |
+
{
|
| 2255 |
+
"epoch": 0.5647058823529412,
|
| 2256 |
+
"grad_norm": 0.3432699739933014,
|
| 2257 |
+
"learning_rate": 9.296326349120785e-05,
|
| 2258 |
+
"loss": 0.6836,
|
| 2259 |
+
"step": 318
|
| 2260 |
+
},
|
| 2261 |
+
{
|
| 2262 |
+
"epoch": 0.5664816870144284,
|
| 2263 |
+
"grad_norm": 0.3710852265357971,
|
| 2264 |
+
"learning_rate": 9.235252107106279e-05,
|
| 2265 |
+
"loss": 0.6961,
|
| 2266 |
+
"step": 319
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.5682574916759157,
|
| 2270 |
+
"grad_norm": 0.351094514131546,
|
| 2271 |
+
"learning_rate": 9.174206545276677e-05,
|
| 2272 |
+
"loss": 0.6668,
|
| 2273 |
+
"step": 320
|
| 2274 |
+
},
|
| 2275 |
+
{
|
| 2276 |
+
"epoch": 0.5700332963374029,
|
| 2277 |
+
"grad_norm": 0.4484163224697113,
|
| 2278 |
+
"learning_rate": 9.113191953011287e-05,
|
| 2279 |
+
"loss": 0.7427,
|
| 2280 |
+
"step": 321
|
| 2281 |
+
},
|
| 2282 |
+
{
|
| 2283 |
+
"epoch": 0.5718091009988902,
|
| 2284 |
+
"grad_norm": 0.44636109471321106,
|
| 2285 |
+
"learning_rate": 9.052210618527966e-05,
|
| 2286 |
+
"loss": 0.8119,
|
| 2287 |
+
"step": 322
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.5735849056603773,
|
| 2291 |
+
"grad_norm": 0.43749314546585083,
|
| 2292 |
+
"learning_rate": 8.991264828797319e-05,
|
| 2293 |
+
"loss": 0.7846,
|
| 2294 |
+
"step": 323
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 0.5753607103218646,
|
| 2298 |
+
"grad_norm": 0.4471510350704193,
|
| 2299 |
+
"learning_rate": 8.930356869456919e-05,
|
| 2300 |
+
"loss": 0.7215,
|
| 2301 |
+
"step": 324
|
| 2302 |
+
},
|
| 2303 |
+
{
|
| 2304 |
+
"epoch": 0.5771365149833518,
|
| 2305 |
+
"grad_norm": 0.5141078233718872,
|
| 2306 |
+
"learning_rate": 8.869489024725595e-05,
|
| 2307 |
+
"loss": 0.7492,
|
| 2308 |
+
"step": 325
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 0.578912319644839,
|
| 2312 |
+
"grad_norm": 0.2640296518802643,
|
| 2313 |
+
"learning_rate": 8.808663577317764e-05,
|
| 2314 |
+
"loss": 0.8625,
|
| 2315 |
+
"step": 326
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 0.5806881243063263,
|
| 2319 |
+
"grad_norm": 0.28867048025131226,
|
| 2320 |
+
"learning_rate": 8.747882808357828e-05,
|
| 2321 |
+
"loss": 0.8352,
|
| 2322 |
+
"step": 327
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.5824639289678135,
|
| 2326 |
+
"grad_norm": 0.2925030589103699,
|
| 2327 |
+
"learning_rate": 8.687148997294621e-05,
|
| 2328 |
+
"loss": 0.8091,
|
| 2329 |
+
"step": 328
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.5842397336293008,
|
| 2333 |
+
"grad_norm": 0.28383681178092957,
|
| 2334 |
+
"learning_rate": 8.626464421815919e-05,
|
| 2335 |
+
"loss": 0.784,
|
| 2336 |
+
"step": 329
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.586015538290788,
|
| 2340 |
+
"grad_norm": 0.3055633306503296,
|
| 2341 |
+
"learning_rate": 8.565831357763039e-05,
|
| 2342 |
+
"loss": 0.79,
|
| 2343 |
+
"step": 330
|
| 2344 |
+
},
|
| 2345 |
+
{
|
| 2346 |
+
"epoch": 0.5877913429522752,
|
| 2347 |
+
"grad_norm": 0.30299943685531616,
|
| 2348 |
+
"learning_rate": 8.505252079045458e-05,
|
| 2349 |
+
"loss": 0.8105,
|
| 2350 |
+
"step": 331
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 0.5895671476137625,
|
| 2354 |
+
"grad_norm": 0.3154890239238739,
|
| 2355 |
+
"learning_rate": 8.444728857555572e-05,
|
| 2356 |
+
"loss": 0.7664,
|
| 2357 |
+
"step": 332
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.5913429522752497,
|
| 2361 |
+
"grad_norm": 0.31844133138656616,
|
| 2362 |
+
"learning_rate": 8.384263963083453e-05,
|
| 2363 |
+
"loss": 0.7709,
|
| 2364 |
+
"step": 333
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.593118756936737,
|
| 2368 |
+
"grad_norm": 0.31844353675842285,
|
| 2369 |
+
"learning_rate": 8.323859663231768e-05,
|
| 2370 |
+
"loss": 0.7426,
|
| 2371 |
+
"step": 334
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.5948945615982242,
|
| 2375 |
+
"grad_norm": 0.31527841091156006,
|
| 2376 |
+
"learning_rate": 8.263518223330697e-05,
|
| 2377 |
+
"loss": 0.7441,
|
| 2378 |
+
"step": 335
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.5966703662597115,
|
| 2382 |
+
"grad_norm": 0.32145699858665466,
|
| 2383 |
+
"learning_rate": 8.203241906353014e-05,
|
| 2384 |
+
"loss": 0.7333,
|
| 2385 |
+
"step": 336
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.5984461709211987,
|
| 2389 |
+
"grad_norm": 0.3175109922885895,
|
| 2390 |
+
"learning_rate": 8.143032972829183e-05,
|
| 2391 |
+
"loss": 0.7488,
|
| 2392 |
+
"step": 337
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.6002219755826859,
|
| 2396 |
+
"grad_norm": 0.3342651128768921,
|
| 2397 |
+
"learning_rate": 8.082893680762619e-05,
|
| 2398 |
+
"loss": 0.7265,
|
| 2399 |
+
"step": 338
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.6019977802441732,
|
| 2403 |
+
"grad_norm": 0.339743971824646,
|
| 2404 |
+
"learning_rate": 8.022826285544968e-05,
|
| 2405 |
+
"loss": 0.7005,
|
| 2406 |
+
"step": 339
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.6037735849056604,
|
| 2410 |
+
"grad_norm": 0.35757359862327576,
|
| 2411 |
+
"learning_rate": 7.96283303987156e-05,
|
| 2412 |
+
"loss": 0.7806,
|
| 2413 |
+
"step": 340
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.6055493895671477,
|
| 2417 |
+
"grad_norm": 0.4024328291416168,
|
| 2418 |
+
"learning_rate": 7.902916193656898e-05,
|
| 2419 |
+
"loss": 0.6895,
|
| 2420 |
+
"step": 341
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.6073251942286348,
|
| 2424 |
+
"grad_norm": 0.3628247380256653,
|
| 2425 |
+
"learning_rate": 7.843077993950302e-05,
|
| 2426 |
+
"loss": 0.7285,
|
| 2427 |
+
"step": 342
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.609100998890122,
|
| 2431 |
+
"grad_norm": 0.3793889582157135,
|
| 2432 |
+
"learning_rate": 7.783320684851614e-05,
|
| 2433 |
+
"loss": 0.729,
|
| 2434 |
+
"step": 343
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 0.6108768035516093,
|
| 2438 |
+
"grad_norm": 0.37614578008651733,
|
| 2439 |
+
"learning_rate": 7.72364650742707e-05,
|
| 2440 |
+
"loss": 0.6869,
|
| 2441 |
+
"step": 344
|
| 2442 |
+
},
|
| 2443 |
+
{
|
| 2444 |
+
"epoch": 0.6126526082130965,
|
| 2445 |
+
"grad_norm": 0.3737132251262665,
|
| 2446 |
+
"learning_rate": 7.664057699625214e-05,
|
| 2447 |
+
"loss": 0.7373,
|
| 2448 |
+
"step": 345
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.6144284128745838,
|
| 2452 |
+
"grad_norm": 0.40523961186408997,
|
| 2453 |
+
"learning_rate": 7.604556496193015e-05,
|
| 2454 |
+
"loss": 0.729,
|
| 2455 |
+
"step": 346
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 0.616204217536071,
|
| 2459 |
+
"grad_norm": 0.3903469145298004,
|
| 2460 |
+
"learning_rate": 7.54514512859201e-05,
|
| 2461 |
+
"loss": 0.7063,
|
| 2462 |
+
"step": 347
|
| 2463 |
+
},
|
| 2464 |
+
{
|
| 2465 |
+
"epoch": 0.6179800221975583,
|
| 2466 |
+
"grad_norm": 0.43782973289489746,
|
| 2467 |
+
"learning_rate": 7.485825824914659e-05,
|
| 2468 |
+
"loss": 0.6763,
|
| 2469 |
+
"step": 348
|
| 2470 |
+
},
|
| 2471 |
+
{
|
| 2472 |
+
"epoch": 0.6197558268590455,
|
| 2473 |
+
"grad_norm": 0.4907206594944,
|
| 2474 |
+
"learning_rate": 7.426600809800752e-05,
|
| 2475 |
+
"loss": 0.7405,
|
| 2476 |
+
"step": 349
|
| 2477 |
+
},
|
| 2478 |
+
{
|
| 2479 |
+
"epoch": 0.6215316315205327,
|
| 2480 |
+
"grad_norm": 0.5378274917602539,
|
| 2481 |
+
"learning_rate": 7.36747230435401e-05,
|
| 2482 |
+
"loss": 0.7417,
|
| 2483 |
+
"step": 350
|
| 2484 |
+
},
|
| 2485 |
+
{
|
| 2486 |
+
"epoch": 0.62330743618202,
|
| 2487 |
+
"grad_norm": 0.266481876373291,
|
| 2488 |
+
"learning_rate": 7.308442526058756e-05,
|
| 2489 |
+
"loss": 0.8434,
|
| 2490 |
+
"step": 351
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 0.6250832408435072,
|
| 2494 |
+
"grad_norm": 0.28670433163642883,
|
| 2495 |
+
"learning_rate": 7.249513688696786e-05,
|
| 2496 |
+
"loss": 0.8049,
|
| 2497 |
+
"step": 352
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.6268590455049945,
|
| 2501 |
+
"grad_norm": 0.29961690306663513,
|
| 2502 |
+
"learning_rate": 7.190688002264308e-05,
|
| 2503 |
+
"loss": 0.762,
|
| 2504 |
+
"step": 353
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 0.6286348501664817,
|
| 2508 |
+
"grad_norm": 0.2873949706554413,
|
| 2509 |
+
"learning_rate": 7.131967672889101e-05,
|
| 2510 |
+
"loss": 0.7389,
|
| 2511 |
+
"step": 354
|
| 2512 |
+
},
|
| 2513 |
+
{
|
| 2514 |
+
"epoch": 0.6304106548279689,
|
| 2515 |
+
"grad_norm": 0.3315136730670929,
|
| 2516 |
+
"learning_rate": 7.073354902747741e-05,
|
| 2517 |
+
"loss": 0.7719,
|
| 2518 |
+
"step": 355
|
| 2519 |
+
},
|
| 2520 |
+
{
|
| 2521 |
+
"epoch": 0.6321864594894562,
|
| 2522 |
+
"grad_norm": 0.31057095527648926,
|
| 2523 |
+
"learning_rate": 7.014851889983057e-05,
|
| 2524 |
+
"loss": 0.7407,
|
| 2525 |
+
"step": 356
|
| 2526 |
+
},
|
| 2527 |
+
{
|
| 2528 |
+
"epoch": 0.6339622641509434,
|
| 2529 |
+
"grad_norm": 0.345838725566864,
|
| 2530 |
+
"learning_rate": 6.95646082862164e-05,
|
| 2531 |
+
"loss": 0.7838,
|
| 2532 |
+
"step": 357
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 0.6357380688124307,
|
| 2536 |
+
"grad_norm": 0.31915196776390076,
|
| 2537 |
+
"learning_rate": 6.898183908491617e-05,
|
| 2538 |
+
"loss": 0.7591,
|
| 2539 |
+
"step": 358
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.6375138734739179,
|
| 2543 |
+
"grad_norm": 0.3124110698699951,
|
| 2544 |
+
"learning_rate": 6.840023315140475e-05,
|
| 2545 |
+
"loss": 0.7222,
|
| 2546 |
+
"step": 359
|
| 2547 |
+
},
|
| 2548 |
+
{
|
| 2549 |
+
"epoch": 0.6392896781354052,
|
| 2550 |
+
"grad_norm": 0.3307512104511261,
|
| 2551 |
+
"learning_rate": 6.781981229753145e-05,
|
| 2552 |
+
"loss": 0.7472,
|
| 2553 |
+
"step": 360
|
| 2554 |
+
},
|
| 2555 |
+
{
|
| 2556 |
+
"epoch": 0.6410654827968923,
|
| 2557 |
+
"grad_norm": 0.3425205945968628,
|
| 2558 |
+
"learning_rate": 6.724059829070158e-05,
|
| 2559 |
+
"loss": 0.764,
|
| 2560 |
+
"step": 361
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 0.6428412874583795,
|
| 2564 |
+
"grad_norm": 0.33861225843429565,
|
| 2565 |
+
"learning_rate": 6.666261285306047e-05,
|
| 2566 |
+
"loss": 0.7396,
|
| 2567 |
+
"step": 362
|
| 2568 |
+
},
|
| 2569 |
+
{
|
| 2570 |
+
"epoch": 0.6446170921198668,
|
| 2571 |
+
"grad_norm": 0.3248923420906067,
|
| 2572 |
+
"learning_rate": 6.608587766067852e-05,
|
| 2573 |
+
"loss": 0.7158,
|
| 2574 |
+
"step": 363
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 0.646392896781354,
|
| 2578 |
+
"grad_norm": 0.349185049533844,
|
| 2579 |
+
"learning_rate": 6.551041434273861e-05,
|
| 2580 |
+
"loss": 0.7415,
|
| 2581 |
+
"step": 364
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.6481687014428413,
|
| 2585 |
+
"grad_norm": 0.33934569358825684,
|
| 2586 |
+
"learning_rate": 6.493624448072457e-05,
|
| 2587 |
+
"loss": 0.744,
|
| 2588 |
+
"step": 365
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"epoch": 0.6499445061043285,
|
| 2592 |
+
"grad_norm": 0.3628052771091461,
|
| 2593 |
+
"learning_rate": 6.43633896076122e-05,
|
| 2594 |
+
"loss": 0.7328,
|
| 2595 |
+
"step": 366
|
| 2596 |
+
},
|
| 2597 |
+
{
|
| 2598 |
+
"epoch": 0.6517203107658157,
|
| 2599 |
+
"grad_norm": 0.348979115486145,
|
| 2600 |
+
"learning_rate": 6.379187120706138e-05,
|
| 2601 |
+
"loss": 0.6755,
|
| 2602 |
+
"step": 367
|
| 2603 |
+
},
|
| 2604 |
+
{
|
| 2605 |
+
"epoch": 0.653496115427303,
|
| 2606 |
+
"grad_norm": 0.38474076986312866,
|
| 2607 |
+
"learning_rate": 6.322171071261071e-05,
|
| 2608 |
+
"loss": 0.711,
|
| 2609 |
+
"step": 368
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 0.6552719200887902,
|
| 2613 |
+
"grad_norm": 0.34556257724761963,
|
| 2614 |
+
"learning_rate": 6.26529295068733e-05,
|
| 2615 |
+
"loss": 0.6995,
|
| 2616 |
+
"step": 369
|
| 2617 |
+
},
|
| 2618 |
+
{
|
| 2619 |
+
"epoch": 0.6570477247502775,
|
| 2620 |
+
"grad_norm": 0.4337230622768402,
|
| 2621 |
+
"learning_rate": 6.208554892073528e-05,
|
| 2622 |
+
"loss": 0.7412,
|
| 2623 |
+
"step": 370
|
| 2624 |
+
},
|
| 2625 |
+
{
|
| 2626 |
+
"epoch": 0.6588235294117647,
|
| 2627 |
+
"grad_norm": 0.37804853916168213,
|
| 2628 |
+
"learning_rate": 6.151959023255545e-05,
|
| 2629 |
+
"loss": 0.6724,
|
| 2630 |
+
"step": 371
|
| 2631 |
+
},
|
| 2632 |
+
{
|
| 2633 |
+
"epoch": 0.6605993340732519,
|
| 2634 |
+
"grad_norm": 0.40870919823646545,
|
| 2635 |
+
"learning_rate": 6.095507466736763e-05,
|
| 2636 |
+
"loss": 0.7243,
|
| 2637 |
+
"step": 372
|
| 2638 |
+
},
|
| 2639 |
+
{
|
| 2640 |
+
"epoch": 0.6623751387347392,
|
| 2641 |
+
"grad_norm": 0.45504140853881836,
|
| 2642 |
+
"learning_rate": 6.039202339608432e-05,
|
| 2643 |
+
"loss": 0.7373,
|
| 2644 |
+
"step": 373
|
| 2645 |
+
},
|
| 2646 |
+
{
|
| 2647 |
+
"epoch": 0.6641509433962264,
|
| 2648 |
+
"grad_norm": 0.46973538398742676,
|
| 2649 |
+
"learning_rate": 5.983045753470308e-05,
|
| 2650 |
+
"loss": 0.7101,
|
| 2651 |
+
"step": 374
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 0.6659267480577137,
|
| 2655 |
+
"grad_norm": 0.5572993755340576,
|
| 2656 |
+
"learning_rate": 5.927039814351426e-05,
|
| 2657 |
+
"loss": 0.7393,
|
| 2658 |
+
"step": 375
|
| 2659 |
+
},
|
| 2660 |
+
{
|
| 2661 |
+
"epoch": 0.6677025527192009,
|
| 2662 |
+
"grad_norm": 0.2691468596458435,
|
| 2663 |
+
"learning_rate": 5.8711866226311553e-05,
|
| 2664 |
+
"loss": 0.8102,
|
| 2665 |
+
"step": 376
|
| 2666 |
+
},
|
| 2667 |
+
{
|
| 2668 |
+
"epoch": 0.6694783573806882,
|
| 2669 |
+
"grad_norm": 0.2898322641849518,
|
| 2670 |
+
"learning_rate": 5.8154882729603876e-05,
|
| 2671 |
+
"loss": 0.7968,
|
| 2672 |
+
"step": 377
|
| 2673 |
+
},
|
| 2674 |
+
{
|
| 2675 |
+
"epoch": 0.6712541620421754,
|
| 2676 |
+
"grad_norm": 0.3048444092273712,
|
| 2677 |
+
"learning_rate": 5.7599468541830356e-05,
|
| 2678 |
+
"loss": 0.775,
|
| 2679 |
+
"step": 378
|
| 2680 |
+
},
|
| 2681 |
+
{
|
| 2682 |
+
"epoch": 0.6730299667036626,
|
| 2683 |
+
"grad_norm": 0.3111611604690552,
|
| 2684 |
+
"learning_rate": 5.7045644492576346e-05,
|
| 2685 |
+
"loss": 0.7742,
|
| 2686 |
+
"step": 379
|
| 2687 |
+
},
|
| 2688 |
+
{
|
| 2689 |
+
"epoch": 0.6748057713651499,
|
| 2690 |
+
"grad_norm": 0.31889772415161133,
|
| 2691 |
+
"learning_rate": 5.64934313517927e-05,
|
| 2692 |
+
"loss": 0.7304,
|
| 2693 |
+
"step": 380
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 0.676581576026637,
|
| 2697 |
+
"grad_norm": 0.3219664692878723,
|
| 2698 |
+
"learning_rate": 5.5942849829016695e-05,
|
| 2699 |
+
"loss": 0.7679,
|
| 2700 |
+
"step": 381
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.6783573806881243,
|
| 2704 |
+
"grad_norm": 0.30955034494400024,
|
| 2705 |
+
"learning_rate": 5.5393920572595356e-05,
|
| 2706 |
+
"loss": 0.7443,
|
| 2707 |
+
"step": 382
|
| 2708 |
+
},
|
| 2709 |
+
{
|
| 2710 |
+
"epoch": 0.6801331853496115,
|
| 2711 |
+
"grad_norm": 0.344043105840683,
|
| 2712 |
+
"learning_rate": 5.484666416891109e-05,
|
| 2713 |
+
"loss": 0.7272,
|
| 2714 |
+
"step": 383
|
| 2715 |
+
},
|
| 2716 |
+
{
|
| 2717 |
+
"epoch": 0.6819089900110987,
|
| 2718 |
+
"grad_norm": 0.33895599842071533,
|
| 2719 |
+
"learning_rate": 5.430110114160964e-05,
|
| 2720 |
+
"loss": 0.7585,
|
| 2721 |
+
"step": 384
|
| 2722 |
+
},
|
| 2723 |
+
{
|
| 2724 |
+
"epoch": 0.683684794672586,
|
| 2725 |
+
"grad_norm": 0.37816834449768066,
|
| 2726 |
+
"learning_rate": 5.375725195083046e-05,
|
| 2727 |
+
"loss": 0.7749,
|
| 2728 |
+
"step": 385
|
| 2729 |
+
},
|
| 2730 |
+
{
|
| 2731 |
+
"epoch": 0.6854605993340732,
|
| 2732 |
+
"grad_norm": 0.3477395176887512,
|
| 2733 |
+
"learning_rate": 5.321513699243924e-05,
|
| 2734 |
+
"loss": 0.7022,
|
| 2735 |
+
"step": 386
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 0.6872364039955605,
|
| 2739 |
+
"grad_norm": 0.3380398154258728,
|
| 2740 |
+
"learning_rate": 5.2674776597263186e-05,
|
| 2741 |
+
"loss": 0.7266,
|
| 2742 |
+
"step": 387
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 0.6890122086570477,
|
| 2746 |
+
"grad_norm": 0.35505762696266174,
|
| 2747 |
+
"learning_rate": 5.2136191030328455e-05,
|
| 2748 |
+
"loss": 0.7411,
|
| 2749 |
+
"step": 388
|
| 2750 |
+
},
|
| 2751 |
+
{
|
| 2752 |
+
"epoch": 0.690788013318535,
|
| 2753 |
+
"grad_norm": 0.38739171624183655,
|
| 2754 |
+
"learning_rate": 5.159940049010015e-05,
|
| 2755 |
+
"loss": 0.7666,
|
| 2756 |
+
"step": 389
|
| 2757 |
+
},
|
| 2758 |
+
{
|
| 2759 |
+
"epoch": 0.6925638179800222,
|
| 2760 |
+
"grad_norm": 0.38473132252693176,
|
| 2761 |
+
"learning_rate": 5.106442510772489e-05,
|
| 2762 |
+
"loss": 0.7038,
|
| 2763 |
+
"step": 390
|
| 2764 |
+
},
|
| 2765 |
+
{
|
| 2766 |
+
"epoch": 0.6943396226415094,
|
| 2767 |
+
"grad_norm": 0.37635302543640137,
|
| 2768 |
+
"learning_rate": 5.0531284946275784e-05,
|
| 2769 |
+
"loss": 0.7488,
|
| 2770 |
+
"step": 391
|
| 2771 |
+
},
|
| 2772 |
+
{
|
| 2773 |
+
"epoch": 0.6961154273029967,
|
| 2774 |
+
"grad_norm": 0.37422046065330505,
|
| 2775 |
+
"learning_rate": 5.000000000000002e-05,
|
| 2776 |
+
"loss": 0.693,
|
| 2777 |
+
"step": 392
|
| 2778 |
+
},
|
| 2779 |
+
{
|
| 2780 |
+
"epoch": 0.6978912319644839,
|
| 2781 |
+
"grad_norm": 0.3987278342247009,
|
| 2782 |
+
"learning_rate": 4.9470590193569044e-05,
|
| 2783 |
+
"loss": 0.6965,
|
| 2784 |
+
"step": 393
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 0.6996670366259712,
|
| 2788 |
+
"grad_norm": 0.34372609853744507,
|
| 2789 |
+
"learning_rate": 4.894307538133129e-05,
|
| 2790 |
+
"loss": 0.6632,
|
| 2791 |
+
"step": 394
|
| 2792 |
+
},
|
| 2793 |
+
{
|
| 2794 |
+
"epoch": 0.7014428412874584,
|
| 2795 |
+
"grad_norm": 0.4215118885040283,
|
| 2796 |
+
"learning_rate": 4.841747534656763e-05,
|
| 2797 |
+
"loss": 0.7081,
|
| 2798 |
+
"step": 395
|
| 2799 |
+
},
|
| 2800 |
+
{
|
| 2801 |
+
"epoch": 0.7032186459489456,
|
| 2802 |
+
"grad_norm": 0.4211183488368988,
|
| 2803 |
+
"learning_rate": 4.7893809800749403e-05,
|
| 2804 |
+
"loss": 0.687,
|
| 2805 |
+
"step": 396
|
| 2806 |
+
},
|
| 2807 |
+
{
|
| 2808 |
+
"epoch": 0.7049944506104329,
|
| 2809 |
+
"grad_norm": 0.44248080253601074,
|
| 2810 |
+
"learning_rate": 4.737209838279922e-05,
|
| 2811 |
+
"loss": 0.7118,
|
| 2812 |
+
"step": 397
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 0.7067702552719201,
|
| 2816 |
+
"grad_norm": 0.38100606203079224,
|
| 2817 |
+
"learning_rate": 4.685236065835443e-05,
|
| 2818 |
+
"loss": 0.6259,
|
| 2819 |
+
"step": 398
|
| 2820 |
+
},
|
| 2821 |
+
{
|
| 2822 |
+
"epoch": 0.7085460599334074,
|
| 2823 |
+
"grad_norm": 0.46482354402542114,
|
| 2824 |
+
"learning_rate": 4.6334616119033356e-05,
|
| 2825 |
+
"loss": 0.6668,
|
| 2826 |
+
"step": 399
|
| 2827 |
+
},
|
| 2828 |
+
{
|
| 2829 |
+
"epoch": 0.7103218645948945,
|
| 2830 |
+
"grad_norm": 0.5484885573387146,
|
| 2831 |
+
"learning_rate": 4.5818884181704294e-05,
|
| 2832 |
+
"loss": 0.7973,
|
| 2833 |
+
"step": 400
|
| 2834 |
+
},
|
| 2835 |
+
{
|
| 2836 |
+
"epoch": 0.7120976692563818,
|
| 2837 |
+
"grad_norm": 0.2660059928894043,
|
| 2838 |
+
"learning_rate": 4.530518418775733e-05,
|
| 2839 |
+
"loss": 0.7845,
|
| 2840 |
+
"step": 401
|
| 2841 |
+
},
|
| 2842 |
+
{
|
| 2843 |
+
"epoch": 0.713873473917869,
|
| 2844 |
+
"grad_norm": 0.30005505681037903,
|
| 2845 |
+
"learning_rate": 4.479353540237903e-05,
|
| 2846 |
+
"loss": 0.8141,
|
| 2847 |
+
"step": 402
|
| 2848 |
+
},
|
| 2849 |
+
{
|
| 2850 |
+
"epoch": 0.7156492785793562,
|
| 2851 |
+
"grad_norm": 0.3031437397003174,
|
| 2852 |
+
"learning_rate": 4.4283957013829846e-05,
|
| 2853 |
+
"loss": 0.7505,
|
| 2854 |
+
"step": 403
|
| 2855 |
+
},
|
| 2856 |
+
{
|
| 2857 |
+
"epoch": 0.7174250832408435,
|
| 2858 |
+
"grad_norm": 0.3152884542942047,
|
| 2859 |
+
"learning_rate": 4.3776468132724604e-05,
|
| 2860 |
+
"loss": 0.8191,
|
| 2861 |
+
"step": 404
|
| 2862 |
+
},
|
| 2863 |
+
{
|
| 2864 |
+
"epoch": 0.7192008879023307,
|
| 2865 |
+
"grad_norm": 0.3122805058956146,
|
| 2866 |
+
"learning_rate": 4.3271087791315734e-05,
|
| 2867 |
+
"loss": 0.7732,
|
| 2868 |
+
"step": 405
|
| 2869 |
+
},
|
| 2870 |
+
{
|
| 2871 |
+
"epoch": 0.720976692563818,
|
| 2872 |
+
"grad_norm": 0.3241139054298401,
|
| 2873 |
+
"learning_rate": 4.276783494277954e-05,
|
| 2874 |
+
"loss": 0.7652,
|
| 2875 |
+
"step": 406
|
| 2876 |
+
},
|
| 2877 |
+
{
|
| 2878 |
+
"epoch": 0.7227524972253052,
|
| 2879 |
+
"grad_norm": 0.3523857295513153,
|
| 2880 |
+
"learning_rate": 4.2266728460505375e-05,
|
| 2881 |
+
"loss": 0.7923,
|
| 2882 |
+
"step": 407
|
| 2883 |
+
},
|
| 2884 |
+
{
|
| 2885 |
+
"epoch": 0.7245283018867924,
|
| 2886 |
+
"grad_norm": 0.3518478274345398,
|
| 2887 |
+
"learning_rate": 4.176778713738787e-05,
|
| 2888 |
+
"loss": 0.8046,
|
| 2889 |
+
"step": 408
|
| 2890 |
+
},
|
| 2891 |
+
{
|
| 2892 |
+
"epoch": 0.7263041065482797,
|
| 2893 |
+
"grad_norm": 0.35740435123443604,
|
| 2894 |
+
"learning_rate": 4.127102968512214e-05,
|
| 2895 |
+
"loss": 0.741,
|
| 2896 |
+
"step": 409
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 0.7280799112097669,
|
| 2900 |
+
"grad_norm": 0.3561273217201233,
|
| 2901 |
+
"learning_rate": 4.077647473350201e-05,
|
| 2902 |
+
"loss": 0.7304,
|
| 2903 |
+
"step": 410
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 0.7298557158712542,
|
| 2907 |
+
"grad_norm": 0.3595544397830963,
|
| 2908 |
+
"learning_rate": 4.028414082972141e-05,
|
| 2909 |
+
"loss": 0.7601,
|
| 2910 |
+
"step": 411
|
| 2911 |
+
},
|
| 2912 |
+
{
|
| 2913 |
+
"epoch": 0.7316315205327414,
|
| 2914 |
+
"grad_norm": 0.38603028655052185,
|
| 2915 |
+
"learning_rate": 3.97940464376787e-05,
|
| 2916 |
+
"loss": 0.768,
|
| 2917 |
+
"step": 412
|
| 2918 |
+
},
|
| 2919 |
+
{
|
| 2920 |
+
"epoch": 0.7334073251942287,
|
| 2921 |
+
"grad_norm": 0.347781240940094,
|
| 2922 |
+
"learning_rate": 3.9306209937284346e-05,
|
| 2923 |
+
"loss": 0.7255,
|
| 2924 |
+
"step": 413
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 0.7351831298557159,
|
| 2928 |
+
"grad_norm": 0.3760242462158203,
|
| 2929 |
+
"learning_rate": 3.882064962377154e-05,
|
| 2930 |
+
"loss": 0.7371,
|
| 2931 |
+
"step": 414
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 0.7369589345172031,
|
| 2935 |
+
"grad_norm": 0.359371542930603,
|
| 2936 |
+
"learning_rate": 3.83373837070101e-05,
|
| 2937 |
+
"loss": 0.7422,
|
| 2938 |
+
"step": 415
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 0.7387347391786904,
|
| 2942 |
+
"grad_norm": 0.3574449419975281,
|
| 2943 |
+
"learning_rate": 3.7856430310823545e-05,
|
| 2944 |
+
"loss": 0.6915,
|
| 2945 |
+
"step": 416
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 0.7405105438401776,
|
| 2949 |
+
"grad_norm": 0.3730245530605316,
|
| 2950 |
+
"learning_rate": 3.737780747230941e-05,
|
| 2951 |
+
"loss": 0.7309,
|
| 2952 |
+
"step": 417
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 0.7422863485016649,
|
| 2956 |
+
"grad_norm": 0.36496400833129883,
|
| 2957 |
+
"learning_rate": 3.69015331411628e-05,
|
| 2958 |
+
"loss": 0.7245,
|
| 2959 |
+
"step": 418
|
| 2960 |
+
},
|
| 2961 |
+
{
|
| 2962 |
+
"epoch": 0.744062153163152,
|
| 2963 |
+
"grad_norm": 0.3593985140323639,
|
| 2964 |
+
"learning_rate": 3.642762517900322e-05,
|
| 2965 |
+
"loss": 0.6389,
|
| 2966 |
+
"step": 419
|
| 2967 |
+
},
|
| 2968 |
+
{
|
| 2969 |
+
"epoch": 0.7458379578246392,
|
| 2970 |
+
"grad_norm": 0.3603939116001129,
|
| 2971 |
+
"learning_rate": 3.595610135870472e-05,
|
| 2972 |
+
"loss": 0.703,
|
| 2973 |
+
"step": 420
|
| 2974 |
+
},
|
| 2975 |
+
{
|
| 2976 |
+
"epoch": 0.7476137624861265,
|
| 2977 |
+
"grad_norm": 0.397124320268631,
|
| 2978 |
+
"learning_rate": 3.548697936372937e-05,
|
| 2979 |
+
"loss": 0.7265,
|
| 2980 |
+
"step": 421
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 0.7493895671476137,
|
| 2984 |
+
"grad_norm": 0.4071907103061676,
|
| 2985 |
+
"learning_rate": 3.5020276787464056e-05,
|
| 2986 |
+
"loss": 0.6752,
|
| 2987 |
+
"step": 422
|
| 2988 |
+
},
|
| 2989 |
+
{
|
| 2990 |
+
"epoch": 0.751165371809101,
|
| 2991 |
+
"grad_norm": 0.3834024965763092,
|
| 2992 |
+
"learning_rate": 3.455601113256073e-05,
|
| 2993 |
+
"loss": 0.6297,
|
| 2994 |
+
"step": 423
|
| 2995 |
+
},
|
| 2996 |
+
{
|
| 2997 |
+
"epoch": 0.751165371809101,
|
| 2998 |
+
"eval_loss": 0.7374839186668396,
|
| 2999 |
+
"eval_runtime": 156.6123,
|
| 3000 |
+
"eval_samples_per_second": 6.06,
|
| 3001 |
+
"eval_steps_per_second": 1.52,
|
| 3002 |
+
"step": 423
|
| 3003 |
}
|
| 3004 |
],
|
| 3005 |
"logging_steps": 1,
|
|
|
|
| 3019 |
"attributes": {}
|
| 3020 |
}
|
| 3021 |
},
|
| 3022 |
+
"total_flos": 5.571234948741857e+17,
|
| 3023 |
"train_batch_size": 2,
|
| 3024 |
"trial_name": null,
|
| 3025 |
"trial_params": null
|