Training in progress, step 100, checkpoint
Browse files- .gitattributes +8 -0
- last-checkpoint/optimizer_0/.metadata +0 -0
- last-checkpoint/optimizer_0/__0_0.distcp +3 -0
- last-checkpoint/optimizer_0/__1_0.distcp +3 -0
- last-checkpoint/optimizer_0/__2_0.distcp +3 -0
- last-checkpoint/optimizer_0/__3_0.distcp +3 -0
- last-checkpoint/pytorch_model_fsdp_0/.metadata +0 -0
- last-checkpoint/pytorch_model_fsdp_0/__0_0.distcp +3 -0
- last-checkpoint/pytorch_model_fsdp_0/__1_0.distcp +3 -0
- last-checkpoint/pytorch_model_fsdp_0/__2_0.distcp +3 -0
- last-checkpoint/pytorch_model_fsdp_0/__3_0.distcp +3 -0
- last-checkpoint/rng_state_0.pth +3 -0
- last-checkpoint/rng_state_1.pth +3 -0
- last-checkpoint/rng_state_2.pth +3 -0
- last-checkpoint/rng_state_3.pth +3 -0
- last-checkpoint/scheduler.pt +3 -0
- last-checkpoint/trainer_state.json +805 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,11 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
last-checkpoint/optimizer_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
last-checkpoint/optimizer_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
last-checkpoint/optimizer_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
last-checkpoint/optimizer_0/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 40 |
+
last-checkpoint/pytorch_model_fsdp_0/__0_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 41 |
+
last-checkpoint/pytorch_model_fsdp_0/__1_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
last-checkpoint/pytorch_model_fsdp_0/__2_0.distcp filter=lfs diff=lfs merge=lfs -text
|
| 43 |
+
last-checkpoint/pytorch_model_fsdp_0/__3_0.distcp filter=lfs diff=lfs merge=lfs -text
|
last-checkpoint/optimizer_0/.metadata
ADDED
|
Binary file (369 kB). View file
|
|
|
last-checkpoint/optimizer_0/__0_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dcb63ef647999d891aca611972191a7ef63cdd6b5a72d1f1cd9faccf092dc513
|
| 3 |
+
size 13934748
|
last-checkpoint/optimizer_0/__1_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa268dd80acfc5cde4bf3bcfa8a42fa9530951eb3f81d2c9b9693fc7f3af6bb5
|
| 3 |
+
size 13999412
|
last-checkpoint/optimizer_0/__2_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d0ea4a1565c360d58ed9f6dadf07e5a29d241213d79057486f9b6220f49daa6
|
| 3 |
+
size 13990904
|
last-checkpoint/optimizer_0/__3_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2fcb110bd43be3db5fbdb14fd66c8025f1c9ad9a279e51f670b913c453f9d08
|
| 3 |
+
size 13990904
|
last-checkpoint/pytorch_model_fsdp_0/.metadata
ADDED
|
Binary file (135 kB). View file
|
|
|
last-checkpoint/pytorch_model_fsdp_0/__0_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c45130d1af0412bda308db66a53c5b638e90386eb566b3c13a9a88bfa9bfb806
|
| 3 |
+
size 6966784
|
last-checkpoint/pytorch_model_fsdp_0/__1_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4c6f046a24d1b6cbffea1980486eabd340f42ae4b962f6b446ea5ac3c7b6a697
|
| 3 |
+
size 6966784
|
last-checkpoint/pytorch_model_fsdp_0/__2_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:727be209ce9e6c45a7b6feda354228088df7ce27204df9a9348752498be8452d
|
| 3 |
+
size 6966784
|
last-checkpoint/pytorch_model_fsdp_0/__3_0.distcp
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:794e99357a2fbead1aaca358574aaaecfdc7cb895b787cb9d96c264ac72bb0c0
|
| 3 |
+
size 6966784
|
last-checkpoint/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33a0773f4c98251681ed846731287836f124116c5c095f6034e6f777ff3d2294
|
| 3 |
+
size 14960
|
last-checkpoint/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:05e4871c2d84ecbc1efac64854d25ee3f52b310cdc9e7aa704123b1be1e82dc1
|
| 3 |
+
size 14960
|
last-checkpoint/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2224282bdf68451bf16fa617356ee085cc44369ceb8aeadcfdd169e4f1c53cda
|
| 3 |
+
size 14960
|
last-checkpoint/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d400bf947081cbe3449eeae2e2bf5d09fede9253207279d467443269b9ffad31
|
| 3 |
+
size 14960
|
last-checkpoint/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:586f94c580fada001a98a596617db52634c95811e53cfca9a69e4db4d223a891
|
| 3 |
+
size 1064
|
last-checkpoint/trainer_state.json
ADDED
|
@@ -0,0 +1,805 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.26595744680851063,
|
| 5 |
+
"eval_steps": 20,
|
| 6 |
+
"global_step": 100,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0,
|
| 13 |
+
"eval_accuracy": 0.7505882352941177,
|
| 14 |
+
"eval_f1": 0.12396694214876033,
|
| 15 |
+
"eval_loss": 0.6111783981323242,
|
| 16 |
+
"eval_precision": 0.36585365853658536,
|
| 17 |
+
"eval_recall": 0.07462686567164178,
|
| 18 |
+
"eval_runtime": 34.3787,
|
| 19 |
+
"eval_samples_per_second": 6.516,
|
| 20 |
+
"eval_steps_per_second": 0.204,
|
| 21 |
+
"step": 0
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.0026595744680851063,
|
| 25 |
+
"grad_norm": 3.310136556625366,
|
| 26 |
+
"learning_rate": 5.263157894736843e-07,
|
| 27 |
+
"loss": 0.6542,
|
| 28 |
+
"step": 1
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0.005319148936170213,
|
| 32 |
+
"grad_norm": 2.5591301918029785,
|
| 33 |
+
"learning_rate": 1.0526315789473685e-06,
|
| 34 |
+
"loss": 0.6609,
|
| 35 |
+
"step": 2
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.007978723404255319,
|
| 39 |
+
"grad_norm": 2.7341604232788086,
|
| 40 |
+
"learning_rate": 1.5789473684210526e-06,
|
| 41 |
+
"loss": 0.6752,
|
| 42 |
+
"step": 3
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.010638297872340425,
|
| 46 |
+
"grad_norm": 2.8091554641723633,
|
| 47 |
+
"learning_rate": 2.105263157894737e-06,
|
| 48 |
+
"loss": 0.618,
|
| 49 |
+
"step": 4
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.013297872340425532,
|
| 53 |
+
"grad_norm": 2.5653722286224365,
|
| 54 |
+
"learning_rate": 2.631578947368421e-06,
|
| 55 |
+
"loss": 0.6752,
|
| 56 |
+
"step": 5
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.015957446808510637,
|
| 60 |
+
"grad_norm": 3.4398417472839355,
|
| 61 |
+
"learning_rate": 3.157894736842105e-06,
|
| 62 |
+
"loss": 0.6783,
|
| 63 |
+
"step": 6
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.018617021276595744,
|
| 67 |
+
"grad_norm": 2.5178332328796387,
|
| 68 |
+
"learning_rate": 3.6842105263157896e-06,
|
| 69 |
+
"loss": 0.631,
|
| 70 |
+
"step": 7
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.02127659574468085,
|
| 74 |
+
"grad_norm": 2.8207452297210693,
|
| 75 |
+
"learning_rate": 4.210526315789474e-06,
|
| 76 |
+
"loss": 0.6361,
|
| 77 |
+
"step": 8
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.023936170212765957,
|
| 81 |
+
"grad_norm": 2.5525949001312256,
|
| 82 |
+
"learning_rate": 4.736842105263158e-06,
|
| 83 |
+
"loss": 0.6088,
|
| 84 |
+
"step": 9
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.026595744680851064,
|
| 88 |
+
"grad_norm": 2.415248155593872,
|
| 89 |
+
"learning_rate": 5.263157894736842e-06,
|
| 90 |
+
"loss": 0.6556,
|
| 91 |
+
"step": 10
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.02925531914893617,
|
| 95 |
+
"grad_norm": 3.0247888565063477,
|
| 96 |
+
"learning_rate": 5.789473684210527e-06,
|
| 97 |
+
"loss": 0.6039,
|
| 98 |
+
"step": 11
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.031914893617021274,
|
| 102 |
+
"grad_norm": 2.5060417652130127,
|
| 103 |
+
"learning_rate": 6.31578947368421e-06,
|
| 104 |
+
"loss": 0.6012,
|
| 105 |
+
"step": 12
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.034574468085106384,
|
| 109 |
+
"grad_norm": 2.770636558532715,
|
| 110 |
+
"learning_rate": 6.842105263157896e-06,
|
| 111 |
+
"loss": 0.6775,
|
| 112 |
+
"step": 13
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.03723404255319149,
|
| 116 |
+
"grad_norm": 2.582097291946411,
|
| 117 |
+
"learning_rate": 7.368421052631579e-06,
|
| 118 |
+
"loss": 0.5993,
|
| 119 |
+
"step": 14
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 0.0398936170212766,
|
| 123 |
+
"grad_norm": 2.5408666133880615,
|
| 124 |
+
"learning_rate": 7.894736842105265e-06,
|
| 125 |
+
"loss": 0.596,
|
| 126 |
+
"step": 15
|
| 127 |
+
},
|
| 128 |
+
{
|
| 129 |
+
"epoch": 0.0425531914893617,
|
| 130 |
+
"grad_norm": 2.478731870651245,
|
| 131 |
+
"learning_rate": 8.421052631578948e-06,
|
| 132 |
+
"loss": 0.5919,
|
| 133 |
+
"step": 16
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 0.04521276595744681,
|
| 137 |
+
"grad_norm": 2.5440561771392822,
|
| 138 |
+
"learning_rate": 8.947368421052632e-06,
|
| 139 |
+
"loss": 0.5223,
|
| 140 |
+
"step": 17
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.047872340425531915,
|
| 144 |
+
"grad_norm": 2.1620945930480957,
|
| 145 |
+
"learning_rate": 9.473684210526315e-06,
|
| 146 |
+
"loss": 0.5412,
|
| 147 |
+
"step": 18
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.05053191489361702,
|
| 151 |
+
"grad_norm": 2.2398860454559326,
|
| 152 |
+
"learning_rate": 1e-05,
|
| 153 |
+
"loss": 0.5847,
|
| 154 |
+
"step": 19
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.05319148936170213,
|
| 158 |
+
"grad_norm": 2.1689343452453613,
|
| 159 |
+
"learning_rate": 1.0526315789473684e-05,
|
| 160 |
+
"loss": 0.5265,
|
| 161 |
+
"step": 20
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.05319148936170213,
|
| 165 |
+
"eval_accuracy": 0.7647058823529411,
|
| 166 |
+
"eval_f1": 0.05660377358490566,
|
| 167 |
+
"eval_loss": 0.6081312894821167,
|
| 168 |
+
"eval_precision": 0.5454545454545454,
|
| 169 |
+
"eval_recall": 0.029850746268656716,
|
| 170 |
+
"eval_runtime": 34.9953,
|
| 171 |
+
"eval_samples_per_second": 6.401,
|
| 172 |
+
"eval_steps_per_second": 0.2,
|
| 173 |
+
"step": 20
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.05585106382978723,
|
| 177 |
+
"grad_norm": 1.8998128175735474,
|
| 178 |
+
"learning_rate": 1.105263157894737e-05,
|
| 179 |
+
"loss": 0.5347,
|
| 180 |
+
"step": 21
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.05851063829787234,
|
| 184 |
+
"grad_norm": 2.079780340194702,
|
| 185 |
+
"learning_rate": 1.1578947368421053e-05,
|
| 186 |
+
"loss": 0.4814,
|
| 187 |
+
"step": 22
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.061170212765957445,
|
| 191 |
+
"grad_norm": 1.8792980909347534,
|
| 192 |
+
"learning_rate": 1.2105263157894737e-05,
|
| 193 |
+
"loss": 0.5084,
|
| 194 |
+
"step": 23
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.06382978723404255,
|
| 198 |
+
"grad_norm": 1.9132519960403442,
|
| 199 |
+
"learning_rate": 1.263157894736842e-05,
|
| 200 |
+
"loss": 0.5027,
|
| 201 |
+
"step": 24
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.06648936170212766,
|
| 205 |
+
"grad_norm": 1.3962018489837646,
|
| 206 |
+
"learning_rate": 1.3157894736842108e-05,
|
| 207 |
+
"loss": 0.5136,
|
| 208 |
+
"step": 25
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.06914893617021277,
|
| 212 |
+
"grad_norm": 1.4877433776855469,
|
| 213 |
+
"learning_rate": 1.3684210526315791e-05,
|
| 214 |
+
"loss": 0.4567,
|
| 215 |
+
"step": 26
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.07180851063829788,
|
| 219 |
+
"grad_norm": 1.5485683679580688,
|
| 220 |
+
"learning_rate": 1.4210526315789475e-05,
|
| 221 |
+
"loss": 0.4365,
|
| 222 |
+
"step": 27
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.07446808510638298,
|
| 226 |
+
"grad_norm": 1.164844036102295,
|
| 227 |
+
"learning_rate": 1.4736842105263159e-05,
|
| 228 |
+
"loss": 0.4142,
|
| 229 |
+
"step": 28
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.07712765957446809,
|
| 233 |
+
"grad_norm": 1.354490876197815,
|
| 234 |
+
"learning_rate": 1.5263157894736846e-05,
|
| 235 |
+
"loss": 0.4492,
|
| 236 |
+
"step": 29
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.0797872340425532,
|
| 240 |
+
"grad_norm": 1.067051649093628,
|
| 241 |
+
"learning_rate": 1.578947368421053e-05,
|
| 242 |
+
"loss": 0.4294,
|
| 243 |
+
"step": 30
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.08244680851063829,
|
| 247 |
+
"grad_norm": 1.3097209930419922,
|
| 248 |
+
"learning_rate": 1.6315789473684213e-05,
|
| 249 |
+
"loss": 0.452,
|
| 250 |
+
"step": 31
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.0851063829787234,
|
| 254 |
+
"grad_norm": 0.9226462244987488,
|
| 255 |
+
"learning_rate": 1.6842105263157896e-05,
|
| 256 |
+
"loss": 0.3848,
|
| 257 |
+
"step": 32
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.08776595744680851,
|
| 261 |
+
"grad_norm": 1.1755656003952026,
|
| 262 |
+
"learning_rate": 1.736842105263158e-05,
|
| 263 |
+
"loss": 0.4307,
|
| 264 |
+
"step": 33
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.09042553191489362,
|
| 268 |
+
"grad_norm": 1.2210921049118042,
|
| 269 |
+
"learning_rate": 1.7894736842105264e-05,
|
| 270 |
+
"loss": 0.4232,
|
| 271 |
+
"step": 34
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.09308510638297872,
|
| 275 |
+
"grad_norm": 0.9078745245933533,
|
| 276 |
+
"learning_rate": 1.8421052631578947e-05,
|
| 277 |
+
"loss": 0.3752,
|
| 278 |
+
"step": 35
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.09574468085106383,
|
| 282 |
+
"grad_norm": 0.936310350894928,
|
| 283 |
+
"learning_rate": 1.894736842105263e-05,
|
| 284 |
+
"loss": 0.3655,
|
| 285 |
+
"step": 36
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.09840425531914894,
|
| 289 |
+
"grad_norm": 1.5738509893417358,
|
| 290 |
+
"learning_rate": 1.9473684210526318e-05,
|
| 291 |
+
"loss": 0.4547,
|
| 292 |
+
"step": 37
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.10106382978723404,
|
| 296 |
+
"grad_norm": 1.1838228702545166,
|
| 297 |
+
"learning_rate": 2e-05,
|
| 298 |
+
"loss": 0.4347,
|
| 299 |
+
"step": 38
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.10372340425531915,
|
| 303 |
+
"grad_norm": 1.4948188066482544,
|
| 304 |
+
"learning_rate": 1.9999568050254373e-05,
|
| 305 |
+
"loss": 0.4135,
|
| 306 |
+
"step": 39
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.10638297872340426,
|
| 310 |
+
"grad_norm": 1.1098586320877075,
|
| 311 |
+
"learning_rate": 1.9998272238333606e-05,
|
| 312 |
+
"loss": 0.4127,
|
| 313 |
+
"step": 40
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.10638297872340426,
|
| 317 |
+
"eval_accuracy": 0.768235294117647,
|
| 318 |
+
"eval_f1": 0.06635071090047394,
|
| 319 |
+
"eval_loss": 0.5411638021469116,
|
| 320 |
+
"eval_precision": 0.7,
|
| 321 |
+
"eval_recall": 0.03482587064676617,
|
| 322 |
+
"eval_runtime": 34.6313,
|
| 323 |
+
"eval_samples_per_second": 6.468,
|
| 324 |
+
"eval_steps_per_second": 0.202,
|
| 325 |
+
"step": 40
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.10904255319148937,
|
| 329 |
+
"grad_norm": 0.779417872428894,
|
| 330 |
+
"learning_rate": 1.999611267618283e-05,
|
| 331 |
+
"loss": 0.3893,
|
| 332 |
+
"step": 41
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.11170212765957446,
|
| 336 |
+
"grad_norm": 1.021106481552124,
|
| 337 |
+
"learning_rate": 1.99930895503665e-05,
|
| 338 |
+
"loss": 0.3104,
|
| 339 |
+
"step": 42
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.11436170212765957,
|
| 343 |
+
"grad_norm": 1.174867868423462,
|
| 344 |
+
"learning_rate": 1.998920312205231e-05,
|
| 345 |
+
"loss": 0.4124,
|
| 346 |
+
"step": 43
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.11702127659574468,
|
| 350 |
+
"grad_norm": 0.8697633743286133,
|
| 351 |
+
"learning_rate": 1.99844537269886e-05,
|
| 352 |
+
"loss": 0.3785,
|
| 353 |
+
"step": 44
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.1196808510638298,
|
| 357 |
+
"grad_norm": 0.9835452437400818,
|
| 358 |
+
"learning_rate": 1.9978841775475368e-05,
|
| 359 |
+
"loss": 0.4014,
|
| 360 |
+
"step": 45
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.12234042553191489,
|
| 364 |
+
"grad_norm": 1.1810511350631714,
|
| 365 |
+
"learning_rate": 1.9972367752328824e-05,
|
| 366 |
+
"loss": 0.3518,
|
| 367 |
+
"step": 46
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.125,
|
| 371 |
+
"grad_norm": 0.9265549778938293,
|
| 372 |
+
"learning_rate": 1.9965032216839493e-05,
|
| 373 |
+
"loss": 0.4024,
|
| 374 |
+
"step": 47
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.1276595744680851,
|
| 378 |
+
"grad_norm": 1.186259150505066,
|
| 379 |
+
"learning_rate": 1.9956835802723916e-05,
|
| 380 |
+
"loss": 0.3599,
|
| 381 |
+
"step": 48
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.13031914893617022,
|
| 385 |
+
"grad_norm": 1.2196171283721924,
|
| 386 |
+
"learning_rate": 1.994777921806989e-05,
|
| 387 |
+
"loss": 0.3411,
|
| 388 |
+
"step": 49
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.13297872340425532,
|
| 392 |
+
"grad_norm": 1.1862437725067139,
|
| 393 |
+
"learning_rate": 1.9937863245275303e-05,
|
| 394 |
+
"loss": 0.362,
|
| 395 |
+
"step": 50
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.1356382978723404,
|
| 399 |
+
"grad_norm": 1.319501280784607,
|
| 400 |
+
"learning_rate": 1.992708874098054e-05,
|
| 401 |
+
"loss": 0.4189,
|
| 402 |
+
"step": 51
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.13829787234042554,
|
| 406 |
+
"grad_norm": 0.9766789674758911,
|
| 407 |
+
"learning_rate": 1.991545663599448e-05,
|
| 408 |
+
"loss": 0.358,
|
| 409 |
+
"step": 52
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.14095744680851063,
|
| 413 |
+
"grad_norm": 1.0482966899871826,
|
| 414 |
+
"learning_rate": 1.990296793521408e-05,
|
| 415 |
+
"loss": 0.3736,
|
| 416 |
+
"step": 53
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.14361702127659576,
|
| 420 |
+
"grad_norm": 0.8634902834892273,
|
| 421 |
+
"learning_rate": 1.9889623717537564e-05,
|
| 422 |
+
"loss": 0.3582,
|
| 423 |
+
"step": 54
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.14627659574468085,
|
| 427 |
+
"grad_norm": 0.9416165947914124,
|
| 428 |
+
"learning_rate": 1.987542513577122e-05,
|
| 429 |
+
"loss": 0.3495,
|
| 430 |
+
"step": 55
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.14893617021276595,
|
| 434 |
+
"grad_norm": 0.9823614358901978,
|
| 435 |
+
"learning_rate": 1.9860373416529804e-05,
|
| 436 |
+
"loss": 0.3446,
|
| 437 |
+
"step": 56
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.15159574468085107,
|
| 441 |
+
"grad_norm": 0.8403105139732361,
|
| 442 |
+
"learning_rate": 1.984446986013057e-05,
|
| 443 |
+
"loss": 0.3177,
|
| 444 |
+
"step": 57
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.15425531914893617,
|
| 448 |
+
"grad_norm": 1.0707823038101196,
|
| 449 |
+
"learning_rate": 1.9827715840480962e-05,
|
| 450 |
+
"loss": 0.323,
|
| 451 |
+
"step": 58
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.15691489361702127,
|
| 455 |
+
"grad_norm": 0.933045506477356,
|
| 456 |
+
"learning_rate": 1.9810112804959867e-05,
|
| 457 |
+
"loss": 0.3123,
|
| 458 |
+
"step": 59
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.1595744680851064,
|
| 462 |
+
"grad_norm": 0.9361464977264404,
|
| 463 |
+
"learning_rate": 1.9791662274292638e-05,
|
| 464 |
+
"loss": 0.3347,
|
| 465 |
+
"step": 60
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.1595744680851064,
|
| 469 |
+
"eval_accuracy": 0.7741176470588236,
|
| 470 |
+
"eval_f1": 0.12727272727272726,
|
| 471 |
+
"eval_loss": 0.5019634962081909,
|
| 472 |
+
"eval_precision": 0.7368421052631579,
|
| 473 |
+
"eval_recall": 0.06965174129353234,
|
| 474 |
+
"eval_runtime": 34.7325,
|
| 475 |
+
"eval_samples_per_second": 6.449,
|
| 476 |
+
"eval_steps_per_second": 0.202,
|
| 477 |
+
"step": 60
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.1622340425531915,
|
| 481 |
+
"grad_norm": 0.8992587327957153,
|
| 482 |
+
"learning_rate": 1.977236584241968e-05,
|
| 483 |
+
"loss": 0.3457,
|
| 484 |
+
"step": 61
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.16489361702127658,
|
| 488 |
+
"grad_norm": 1.282809853553772,
|
| 489 |
+
"learning_rate": 1.9752225176358757e-05,
|
| 490 |
+
"loss": 0.3226,
|
| 491 |
+
"step": 62
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.1675531914893617,
|
| 495 |
+
"grad_norm": 2.4324252605438232,
|
| 496 |
+
"learning_rate": 1.9731242016060985e-05,
|
| 497 |
+
"loss": 0.4227,
|
| 498 |
+
"step": 63
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 0.1702127659574468,
|
| 502 |
+
"grad_norm": 1.0456701517105103,
|
| 503 |
+
"learning_rate": 1.9709418174260523e-05,
|
| 504 |
+
"loss": 0.3102,
|
| 505 |
+
"step": 64
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.17287234042553193,
|
| 509 |
+
"grad_norm": 1.2882471084594727,
|
| 510 |
+
"learning_rate": 1.9686755536317945e-05,
|
| 511 |
+
"loss": 0.3145,
|
| 512 |
+
"step": 65
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.17553191489361702,
|
| 516 |
+
"grad_norm": 1.1312603950500488,
|
| 517 |
+
"learning_rate": 1.9663256060057395e-05,
|
| 518 |
+
"loss": 0.3353,
|
| 519 |
+
"step": 66
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.17819148936170212,
|
| 523 |
+
"grad_norm": 1.0174272060394287,
|
| 524 |
+
"learning_rate": 1.9638921775597428e-05,
|
| 525 |
+
"loss": 0.2845,
|
| 526 |
+
"step": 67
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.18085106382978725,
|
| 530 |
+
"grad_norm": 1.241572380065918,
|
| 531 |
+
"learning_rate": 1.961375478517564e-05,
|
| 532 |
+
"loss": 0.3015,
|
| 533 |
+
"step": 68
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.18351063829787234,
|
| 537 |
+
"grad_norm": 1.3726611137390137,
|
| 538 |
+
"learning_rate": 1.958775726296706e-05,
|
| 539 |
+
"loss": 0.3671,
|
| 540 |
+
"step": 69
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.18617021276595744,
|
| 544 |
+
"grad_norm": 1.2311499118804932,
|
| 545 |
+
"learning_rate": 1.95609314548963e-05,
|
| 546 |
+
"loss": 0.2902,
|
| 547 |
+
"step": 70
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.18882978723404256,
|
| 551 |
+
"grad_norm": 1.3199646472930908,
|
| 552 |
+
"learning_rate": 1.953327967844356e-05,
|
| 553 |
+
"loss": 0.3594,
|
| 554 |
+
"step": 71
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.19148936170212766,
|
| 558 |
+
"grad_norm": 1.6513502597808838,
|
| 559 |
+
"learning_rate": 1.95048043224444e-05,
|
| 560 |
+
"loss": 0.2831,
|
| 561 |
+
"step": 72
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.19414893617021275,
|
| 565 |
+
"grad_norm": 1.763235330581665,
|
| 566 |
+
"learning_rate": 1.9475507846883377e-05,
|
| 567 |
+
"loss": 0.3675,
|
| 568 |
+
"step": 73
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.19680851063829788,
|
| 572 |
+
"grad_norm": 1.8195736408233643,
|
| 573 |
+
"learning_rate": 1.9445392782681523e-05,
|
| 574 |
+
"loss": 0.398,
|
| 575 |
+
"step": 74
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.19946808510638298,
|
| 579 |
+
"grad_norm": 1.9659175872802734,
|
| 580 |
+
"learning_rate": 1.94144617314777e-05,
|
| 581 |
+
"loss": 0.353,
|
| 582 |
+
"step": 75
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.20212765957446807,
|
| 586 |
+
"grad_norm": 1.60419762134552,
|
| 587 |
+
"learning_rate": 1.9382717365403854e-05,
|
| 588 |
+
"loss": 0.3565,
|
| 589 |
+
"step": 76
|
| 590 |
+
},
|
| 591 |
+
{
|
| 592 |
+
"epoch": 0.2047872340425532,
|
| 593 |
+
"grad_norm": 1.5443696975708008,
|
| 594 |
+
"learning_rate": 1.9350162426854152e-05,
|
| 595 |
+
"loss": 0.3246,
|
| 596 |
+
"step": 77
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.2074468085106383,
|
| 600 |
+
"grad_norm": 1.8536072969436646,
|
| 601 |
+
"learning_rate": 1.9316799728248074e-05,
|
| 602 |
+
"loss": 0.3491,
|
| 603 |
+
"step": 78
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 0.21010638297872342,
|
| 607 |
+
"grad_norm": 2.2563788890838623,
|
| 608 |
+
"learning_rate": 1.9282632151787462e-05,
|
| 609 |
+
"loss": 0.4211,
|
| 610 |
+
"step": 79
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.2127659574468085,
|
| 614 |
+
"grad_norm": 1.3425776958465576,
|
| 615 |
+
"learning_rate": 1.924766264920751e-05,
|
| 616 |
+
"loss": 0.3077,
|
| 617 |
+
"step": 80
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.2127659574468085,
|
| 621 |
+
"eval_accuracy": 0.7964705882352942,
|
| 622 |
+
"eval_f1": 0.3663003663003663,
|
| 623 |
+
"eval_loss": 0.4462856650352478,
|
| 624 |
+
"eval_precision": 0.6944444444444444,
|
| 625 |
+
"eval_recall": 0.24875621890547264,
|
| 626 |
+
"eval_runtime": 34.8097,
|
| 627 |
+
"eval_samples_per_second": 6.435,
|
| 628 |
+
"eval_steps_per_second": 0.201,
|
| 629 |
+
"step": 80
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 0.2154255319148936,
|
| 633 |
+
"grad_norm": 2.540194272994995,
|
| 634 |
+
"learning_rate": 1.9211894241521757e-05,
|
| 635 |
+
"loss": 0.3127,
|
| 636 |
+
"step": 81
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.21808510638297873,
|
| 640 |
+
"grad_norm": 1.8769720792770386,
|
| 641 |
+
"learning_rate": 1.917533001876113e-05,
|
| 642 |
+
"loss": 0.2998,
|
| 643 |
+
"step": 82
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.22074468085106383,
|
| 647 |
+
"grad_norm": 1.4883919954299927,
|
| 648 |
+
"learning_rate": 1.9137973139706973e-05,
|
| 649 |
+
"loss": 0.3245,
|
| 650 |
+
"step": 83
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.22340425531914893,
|
| 654 |
+
"grad_norm": 1.6703698635101318,
|
| 655 |
+
"learning_rate": 1.9099826831618168e-05,
|
| 656 |
+
"loss": 0.3199,
|
| 657 |
+
"step": 84
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.22606382978723405,
|
| 661 |
+
"grad_norm": 2.8918988704681396,
|
| 662 |
+
"learning_rate": 1.9060894389952328e-05,
|
| 663 |
+
"loss": 0.2825,
|
| 664 |
+
"step": 85
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 0.22872340425531915,
|
| 668 |
+
"grad_norm": 1.5494073629379272,
|
| 669 |
+
"learning_rate": 1.9021179178081107e-05,
|
| 670 |
+
"loss": 0.3213,
|
| 671 |
+
"step": 86
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 0.23138297872340424,
|
| 675 |
+
"grad_norm": 1.331063151359558,
|
| 676 |
+
"learning_rate": 1.898068462699964e-05,
|
| 677 |
+
"loss": 0.2572,
|
| 678 |
+
"step": 87
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.23404255319148937,
|
| 682 |
+
"grad_norm": 1.5478427410125732,
|
| 683 |
+
"learning_rate": 1.8939414235030137e-05,
|
| 684 |
+
"loss": 0.3001,
|
| 685 |
+
"step": 88
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.23670212765957446,
|
| 689 |
+
"grad_norm": 2.469545602798462,
|
| 690 |
+
"learning_rate": 1.889737156751965e-05,
|
| 691 |
+
"loss": 0.3199,
|
| 692 |
+
"step": 89
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.2393617021276596,
|
| 696 |
+
"grad_norm": 2.134981155395508,
|
| 697 |
+
"learning_rate": 1.8854560256532098e-05,
|
| 698 |
+
"loss": 0.3192,
|
| 699 |
+
"step": 90
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.24202127659574468,
|
| 703 |
+
"grad_norm": 1.727616548538208,
|
| 704 |
+
"learning_rate": 1.8810984000534457e-05,
|
| 705 |
+
"loss": 0.3072,
|
| 706 |
+
"step": 91
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.24468085106382978,
|
| 710 |
+
"grad_norm": 2.0483591556549072,
|
| 711 |
+
"learning_rate": 1.8766646564077265e-05,
|
| 712 |
+
"loss": 0.297,
|
| 713 |
+
"step": 92
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.2473404255319149,
|
| 717 |
+
"grad_norm": 1.8018875122070312,
|
| 718 |
+
"learning_rate": 1.8721551777469397e-05,
|
| 719 |
+
"loss": 0.2798,
|
| 720 |
+
"step": 93
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.25,
|
| 724 |
+
"grad_norm": 1.7120018005371094,
|
| 725 |
+
"learning_rate": 1.8675703536447178e-05,
|
| 726 |
+
"loss": 0.2438,
|
| 727 |
+
"step": 94
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.2526595744680851,
|
| 731 |
+
"grad_norm": 1.8456470966339111,
|
| 732 |
+
"learning_rate": 1.862910580183782e-05,
|
| 733 |
+
"loss": 0.333,
|
| 734 |
+
"step": 95
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.2553191489361702,
|
| 738 |
+
"grad_norm": 2.701077461242676,
|
| 739 |
+
"learning_rate": 1.858176259921724e-05,
|
| 740 |
+
"loss": 0.3214,
|
| 741 |
+
"step": 96
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 0.2579787234042553,
|
| 745 |
+
"grad_norm": 1.6109999418258667,
|
| 746 |
+
"learning_rate": 1.853367801856231e-05,
|
| 747 |
+
"loss": 0.2701,
|
| 748 |
+
"step": 97
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.26063829787234044,
|
| 752 |
+
"grad_norm": 1.524688482284546,
|
| 753 |
+
"learning_rate": 1.8484856213897496e-05,
|
| 754 |
+
"loss": 0.2455,
|
| 755 |
+
"step": 98
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 0.2632978723404255,
|
| 759 |
+
"grad_norm": 1.8650296926498413,
|
| 760 |
+
"learning_rate": 1.843530140293603e-05,
|
| 761 |
+
"loss": 0.273,
|
| 762 |
+
"step": 99
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 0.26595744680851063,
|
| 766 |
+
"grad_norm": 1.5295664072036743,
|
| 767 |
+
"learning_rate": 1.8385017866715507e-05,
|
| 768 |
+
"loss": 0.307,
|
| 769 |
+
"step": 100
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.26595744680851063,
|
| 773 |
+
"eval_accuracy": 0.8,
|
| 774 |
+
"eval_f1": 0.4097222222222222,
|
| 775 |
+
"eval_loss": 0.44977959990501404,
|
| 776 |
+
"eval_precision": 0.6781609195402298,
|
| 777 |
+
"eval_recall": 0.2935323383084577,
|
| 778 |
+
"eval_runtime": 34.7061,
|
| 779 |
+
"eval_samples_per_second": 6.454,
|
| 780 |
+
"eval_steps_per_second": 0.202,
|
| 781 |
+
"step": 100
|
| 782 |
+
}
|
| 783 |
+
],
|
| 784 |
+
"logging_steps": 1,
|
| 785 |
+
"max_steps": 376,
|
| 786 |
+
"num_input_tokens_seen": 0,
|
| 787 |
+
"num_train_epochs": 1,
|
| 788 |
+
"save_steps": 100,
|
| 789 |
+
"stateful_callbacks": {
|
| 790 |
+
"TrainerControl": {
|
| 791 |
+
"args": {
|
| 792 |
+
"should_epoch_stop": false,
|
| 793 |
+
"should_evaluate": false,
|
| 794 |
+
"should_log": false,
|
| 795 |
+
"should_save": true,
|
| 796 |
+
"should_training_stop": false
|
| 797 |
+
},
|
| 798 |
+
"attributes": {}
|
| 799 |
+
}
|
| 800 |
+
},
|
| 801 |
+
"total_flos": 3.2170466684698624e+16,
|
| 802 |
+
"train_batch_size": 8,
|
| 803 |
+
"trial_name": null,
|
| 804 |
+
"trial_params": null
|
| 805 |
+
}
|