Training in progress, step 200, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/global_step200/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step200/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step200/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step200/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step200/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step200/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step200/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step200/zero_pp_rank_3_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +739 -3
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13648688
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4e00c75665fad68257b13112a896f0e5bfe09116a5aa7d74e1a11ae36dfaf5ff
|
| 3 |
size 13648688
|
last-checkpoint/global_step200/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b28310f9cdeec28ac828f4b272b8422f969423fa5606b6a5280dcae9804a9906
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step200/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:336f8476bbfc2636bde52666b9fd678cb9ea4f3960967488f79a0e126589659b
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step200/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cdbd809b7cfdfc9e87ae66c687d52ef0e79bad9c1ffb974143c9207b719a8143
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step200/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2a3bd6c3c45c00ca288387009386aae02b69e0f61a21d8aa7f447ce9a0ac7b72
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step200/zero_pp_rank_0_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d11585c69f2eb7fc6ae25f8f7d341eb09bb43308fbb6c512ffd6a07473415410
|
| 3 |
+
size 152238
|
last-checkpoint/global_step200/zero_pp_rank_1_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:da48dbffc3ff23a696096bc8ab8bbd5b584bc72a5fe5c27c36305aa26dd9e47b
|
| 3 |
+
size 152238
|
last-checkpoint/global_step200/zero_pp_rank_2_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b2092bba3deb5b81e41ea98bc99a0f8aa82b9246b54ef2a0e9e7063d5f9aab6
|
| 3 |
+
size 152238
|
last-checkpoint/global_step200/zero_pp_rank_3_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:39ad00fd5370f273bbd2355e4bf561718800f099a2c153171eb9e0ff78a91ba3
|
| 3 |
+
size 152238
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step200
|
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 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2c24a39de75f23f0d84f98b720b05e4f552cdd0306626c901205b2d9690be33
|
| 3 |
size 14960
|
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 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:75baba3fff49778fb1f9915f06b5fd052daf5b241c6df48d63ce4cc2fd74ad52
|
| 3 |
size 14960
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43726b0b6816cd7dc4fb0c6379613d398bab9cf9069c5ee8ac83eea24f4fa621
|
| 3 |
size 14960
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c26a88d8955baf0e31629f8efe3e6d01a1336c21cf4cdd20becb43acefdfdd69
|
| 3 |
size 14960
|
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 sha256:863b742f77d90cf639cf492ec2d91dab7a9ebd0f58799b06186327b2c961991e
|
| 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": 40,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -743,6 +743,742 @@
|
|
| 743 |
"learning_rate": 1.9970093649572567e-05,
|
| 744 |
"loss": 0.2257,
|
| 745 |
"step": 100
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 746 |
}
|
| 747 |
],
|
| 748 |
"logging_steps": 1,
|
|
@@ -762,7 +1498,7 @@
|
|
| 762 |
"attributes": {}
|
| 763 |
}
|
| 764 |
},
|
| 765 |
-
"total_flos":
|
| 766 |
"train_batch_size": 4,
|
| 767 |
"trial_name": null,
|
| 768 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.24592683676606208,
|
| 5 |
"eval_steps": 40,
|
| 6 |
+
"global_step": 200,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 743 |
"learning_rate": 1.9970093649572567e-05,
|
| 744 |
"loss": 0.2257,
|
| 745 |
"step": 100
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.12419305256686136,
|
| 749 |
+
"grad_norm": 0.18210033905044673,
|
| 750 |
+
"learning_rate": 1.9966680315970647e-05,
|
| 751 |
+
"loss": 0.2629,
|
| 752 |
+
"step": 101
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.12542268675069168,
|
| 756 |
+
"grad_norm": 0.22735479010308166,
|
| 757 |
+
"learning_rate": 1.996308289887366e-05,
|
| 758 |
+
"loss": 0.3055,
|
| 759 |
+
"step": 102
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.12665232093452197,
|
| 763 |
+
"grad_norm": 0.21880883838789394,
|
| 764 |
+
"learning_rate": 1.9959301464725507e-05,
|
| 765 |
+
"loss": 0.2896,
|
| 766 |
+
"step": 103
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.1278819551183523,
|
| 770 |
+
"grad_norm": 0.1568188763941854,
|
| 771 |
+
"learning_rate": 1.995533608336886e-05,
|
| 772 |
+
"loss": 0.1856,
|
| 773 |
+
"step": 104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.1291115893021826,
|
| 777 |
+
"grad_norm": 0.21046557286925857,
|
| 778 |
+
"learning_rate": 1.995118682804388e-05,
|
| 779 |
+
"loss": 0.2328,
|
| 780 |
+
"step": 105
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.1303412234860129,
|
| 784 |
+
"grad_norm": 0.23586873079680978,
|
| 785 |
+
"learning_rate": 1.9946853775386857e-05,
|
| 786 |
+
"loss": 0.2608,
|
| 787 |
+
"step": 106
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.13157085766984322,
|
| 791 |
+
"grad_norm": 0.19655270590211088,
|
| 792 |
+
"learning_rate": 1.9942337005428805e-05,
|
| 793 |
+
"loss": 0.2982,
|
| 794 |
+
"step": 107
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.13280049185367354,
|
| 798 |
+
"grad_norm": 0.31014757888444056,
|
| 799 |
+
"learning_rate": 1.9937636601593965e-05,
|
| 800 |
+
"loss": 0.3041,
|
| 801 |
+
"step": 108
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.13403012603750383,
|
| 805 |
+
"grad_norm": 0.3337300802942654,
|
| 806 |
+
"learning_rate": 1.9932752650698285e-05,
|
| 807 |
+
"loss": 0.3085,
|
| 808 |
+
"step": 109
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.13525976022133415,
|
| 812 |
+
"grad_norm": 0.2637379526580471,
|
| 813 |
+
"learning_rate": 1.9927685242947804e-05,
|
| 814 |
+
"loss": 0.301,
|
| 815 |
+
"step": 110
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.13648939440516447,
|
| 819 |
+
"grad_norm": 0.22577400909500966,
|
| 820 |
+
"learning_rate": 1.9922434471936987e-05,
|
| 821 |
+
"loss": 0.2234,
|
| 822 |
+
"step": 111
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.13771902858899476,
|
| 826 |
+
"grad_norm": 0.21966929441456037,
|
| 827 |
+
"learning_rate": 1.9917000434647e-05,
|
| 828 |
+
"loss": 0.2743,
|
| 829 |
+
"step": 112
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.13894866277282508,
|
| 833 |
+
"grad_norm": 0.2540802763063313,
|
| 834 |
+
"learning_rate": 1.991138323144392e-05,
|
| 835 |
+
"loss": 0.2307,
|
| 836 |
+
"step": 113
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.1401782969566554,
|
| 840 |
+
"grad_norm": 0.3204592419266727,
|
| 841 |
+
"learning_rate": 1.990558296607687e-05,
|
| 842 |
+
"loss": 0.2156,
|
| 843 |
+
"step": 114
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.1414079311404857,
|
| 847 |
+
"grad_norm": 0.24090237444938584,
|
| 848 |
+
"learning_rate": 1.9899599745676123e-05,
|
| 849 |
+
"loss": 0.2809,
|
| 850 |
+
"step": 115
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.142637565324316,
|
| 854 |
+
"grad_norm": 0.2611220625407227,
|
| 855 |
+
"learning_rate": 1.9893433680751105e-05,
|
| 856 |
+
"loss": 0.2307,
|
| 857 |
+
"step": 116
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.14386719950814633,
|
| 861 |
+
"grad_norm": 0.37292140730199286,
|
| 862 |
+
"learning_rate": 1.9887084885188354e-05,
|
| 863 |
+
"loss": 0.2051,
|
| 864 |
+
"step": 117
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.14509683369197662,
|
| 868 |
+
"grad_norm": 0.21515044559789018,
|
| 869 |
+
"learning_rate": 1.9880553476249437e-05,
|
| 870 |
+
"loss": 0.2416,
|
| 871 |
+
"step": 118
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.14632646787580694,
|
| 875 |
+
"grad_norm": 0.27918682412366513,
|
| 876 |
+
"learning_rate": 1.9873839574568756e-05,
|
| 877 |
+
"loss": 0.2711,
|
| 878 |
+
"step": 119
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.14755610205963726,
|
| 882 |
+
"grad_norm": 0.23334267872599612,
|
| 883 |
+
"learning_rate": 1.9866943304151346e-05,
|
| 884 |
+
"loss": 0.305,
|
| 885 |
+
"step": 120
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.14755610205963726,
|
| 889 |
+
"eval_accuracy": 0.8021390374331551,
|
| 890 |
+
"eval_f1": 0.5316455696202531,
|
| 891 |
+
"eval_loss": 0.4852343797683716,
|
| 892 |
+
"eval_precision": 0.7241379310344828,
|
| 893 |
+
"eval_recall": 0.42,
|
| 894 |
+
"eval_runtime": 22.4489,
|
| 895 |
+
"eval_samples_per_second": 2.227,
|
| 896 |
+
"eval_steps_per_second": 0.178,
|
| 897 |
+
"step": 120
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.14878573624346758,
|
| 901 |
+
"grad_norm": 0.255894846030308,
|
| 902 |
+
"learning_rate": 1.9859864792370565e-05,
|
| 903 |
+
"loss": 0.2925,
|
| 904 |
+
"step": 121
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 0.15001537042729787,
|
| 908 |
+
"grad_norm": 0.2894368730807284,
|
| 909 |
+
"learning_rate": 1.985260416996575e-05,
|
| 910 |
+
"loss": 0.3092,
|
| 911 |
+
"step": 122
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.1512450046111282,
|
| 915 |
+
"grad_norm": 0.2819736938873003,
|
| 916 |
+
"learning_rate": 1.9845161571039805e-05,
|
| 917 |
+
"loss": 0.2827,
|
| 918 |
+
"step": 123
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.1524746387949585,
|
| 922 |
+
"grad_norm": 0.2849595623136817,
|
| 923 |
+
"learning_rate": 1.983753713305672e-05,
|
| 924 |
+
"loss": 0.3301,
|
| 925 |
+
"step": 124
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.1537042729787888,
|
| 929 |
+
"grad_norm": 0.36826872994416193,
|
| 930 |
+
"learning_rate": 1.982973099683902e-05,
|
| 931 |
+
"loss": 0.2657,
|
| 932 |
+
"step": 125
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.15493390716261912,
|
| 936 |
+
"grad_norm": 0.26692338354049955,
|
| 937 |
+
"learning_rate": 1.98217433065652e-05,
|
| 938 |
+
"loss": 0.1732,
|
| 939 |
+
"step": 126
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.15616354134644944,
|
| 943 |
+
"grad_norm": 0.2573459669567664,
|
| 944 |
+
"learning_rate": 1.9813574209767013e-05,
|
| 945 |
+
"loss": 0.2586,
|
| 946 |
+
"step": 127
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.15739317553027973,
|
| 950 |
+
"grad_norm": 0.3781515395257974,
|
| 951 |
+
"learning_rate": 1.9805223857326794e-05,
|
| 952 |
+
"loss": 0.2687,
|
| 953 |
+
"step": 128
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.15862280971411005,
|
| 957 |
+
"grad_norm": 0.2609213888795932,
|
| 958 |
+
"learning_rate": 1.9796692403474632e-05,
|
| 959 |
+
"loss": 0.243,
|
| 960 |
+
"step": 129
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.15985244389794037,
|
| 964 |
+
"grad_norm": 0.33148152190355884,
|
| 965 |
+
"learning_rate": 1.9787980005785553e-05,
|
| 966 |
+
"loss": 0.284,
|
| 967 |
+
"step": 130
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.16108207808177066,
|
| 971 |
+
"grad_norm": 0.24671664914766978,
|
| 972 |
+
"learning_rate": 1.977908682517658e-05,
|
| 973 |
+
"loss": 0.2743,
|
| 974 |
+
"step": 131
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.16231171226560098,
|
| 978 |
+
"grad_norm": 0.23136891800428114,
|
| 979 |
+
"learning_rate": 1.9770013025903797e-05,
|
| 980 |
+
"loss": 0.1988,
|
| 981 |
+
"step": 132
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.1635413464494313,
|
| 985 |
+
"grad_norm": 0.3625913614143042,
|
| 986 |
+
"learning_rate": 1.9760758775559275e-05,
|
| 987 |
+
"loss": 0.3571,
|
| 988 |
+
"step": 133
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 0.1647709806332616,
|
| 992 |
+
"grad_norm": 0.4654838486463352,
|
| 993 |
+
"learning_rate": 1.9751324245068008e-05,
|
| 994 |
+
"loss": 0.2483,
|
| 995 |
+
"step": 134
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.1660006148170919,
|
| 999 |
+
"grad_norm": 0.23389863720470436,
|
| 1000 |
+
"learning_rate": 1.974170960868474e-05,
|
| 1001 |
+
"loss": 0.2121,
|
| 1002 |
+
"step": 135
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.16723024900092223,
|
| 1006 |
+
"grad_norm": 0.26709730418277217,
|
| 1007 |
+
"learning_rate": 1.973191504399076e-05,
|
| 1008 |
+
"loss": 0.2277,
|
| 1009 |
+
"step": 136
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 0.16845988318475252,
|
| 1013 |
+
"grad_norm": 0.3065908377026066,
|
| 1014 |
+
"learning_rate": 1.97219407318906e-05,
|
| 1015 |
+
"loss": 0.2811,
|
| 1016 |
+
"step": 137
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.16968951736858284,
|
| 1020 |
+
"grad_norm": 0.2561047976359151,
|
| 1021 |
+
"learning_rate": 1.9711786856608714e-05,
|
| 1022 |
+
"loss": 0.2702,
|
| 1023 |
+
"step": 138
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.17091915155241316,
|
| 1027 |
+
"grad_norm": 0.3029014908375809,
|
| 1028 |
+
"learning_rate": 1.970145360568607e-05,
|
| 1029 |
+
"loss": 0.2703,
|
| 1030 |
+
"step": 139
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.17214878573624345,
|
| 1034 |
+
"grad_norm": 0.2928125726746547,
|
| 1035 |
+
"learning_rate": 1.969094116997668e-05,
|
| 1036 |
+
"loss": 0.2925,
|
| 1037 |
+
"step": 140
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 0.17337841992007377,
|
| 1041 |
+
"grad_norm": 0.417971783345298,
|
| 1042 |
+
"learning_rate": 1.968024974364408e-05,
|
| 1043 |
+
"loss": 0.2496,
|
| 1044 |
+
"step": 141
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 0.1746080541039041,
|
| 1048 |
+
"grad_norm": 0.2496018935199012,
|
| 1049 |
+
"learning_rate": 1.9669379524157755e-05,
|
| 1050 |
+
"loss": 0.2279,
|
| 1051 |
+
"step": 142
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 0.1758376882877344,
|
| 1055 |
+
"grad_norm": 0.3714257892665527,
|
| 1056 |
+
"learning_rate": 1.9658330712289456e-05,
|
| 1057 |
+
"loss": 0.295,
|
| 1058 |
+
"step": 143
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.1770673224715647,
|
| 1062 |
+
"grad_norm": 0.2580656993421598,
|
| 1063 |
+
"learning_rate": 1.9647103512109535e-05,
|
| 1064 |
+
"loss": 0.253,
|
| 1065 |
+
"step": 144
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.17829695665539502,
|
| 1069 |
+
"grad_norm": 0.2560142798427468,
|
| 1070 |
+
"learning_rate": 1.9635698130983153e-05,
|
| 1071 |
+
"loss": 0.251,
|
| 1072 |
+
"step": 145
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.17952659083922534,
|
| 1076 |
+
"grad_norm": 0.2607452518822491,
|
| 1077 |
+
"learning_rate": 1.962411477956645e-05,
|
| 1078 |
+
"loss": 0.2395,
|
| 1079 |
+
"step": 146
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.18075622502305563,
|
| 1083 |
+
"grad_norm": 0.25191931122780475,
|
| 1084 |
+
"learning_rate": 1.9612353671802658e-05,
|
| 1085 |
+
"loss": 0.2389,
|
| 1086 |
+
"step": 147
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.18198585920688595,
|
| 1090 |
+
"grad_norm": 0.2991928933187448,
|
| 1091 |
+
"learning_rate": 1.960041502491815e-05,
|
| 1092 |
+
"loss": 0.3016,
|
| 1093 |
+
"step": 148
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 0.18321549339071627,
|
| 1097 |
+
"grad_norm": 0.2775853628798545,
|
| 1098 |
+
"learning_rate": 1.9588299059418434e-05,
|
| 1099 |
+
"loss": 0.2558,
|
| 1100 |
+
"step": 149
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.18444512757454656,
|
| 1104 |
+
"grad_norm": 0.25029731861303234,
|
| 1105 |
+
"learning_rate": 1.957600599908406e-05,
|
| 1106 |
+
"loss": 0.221,
|
| 1107 |
+
"step": 150
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.18567476175837688,
|
| 1111 |
+
"grad_norm": 0.3473969966642877,
|
| 1112 |
+
"learning_rate": 1.9563536070966513e-05,
|
| 1113 |
+
"loss": 0.3639,
|
| 1114 |
+
"step": 151
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 0.1869043959422072,
|
| 1118 |
+
"grad_norm": 0.21700372590221412,
|
| 1119 |
+
"learning_rate": 1.9550889505383996e-05,
|
| 1120 |
+
"loss": 0.2122,
|
| 1121 |
+
"step": 152
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 0.1881340301260375,
|
| 1125 |
+
"grad_norm": 0.26057622594959473,
|
| 1126 |
+
"learning_rate": 1.9538066535917196e-05,
|
| 1127 |
+
"loss": 0.2631,
|
| 1128 |
+
"step": 153
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 0.1893636643098678,
|
| 1132 |
+
"grad_norm": 0.31462240972233546,
|
| 1133 |
+
"learning_rate": 1.952506739940496e-05,
|
| 1134 |
+
"loss": 0.2576,
|
| 1135 |
+
"step": 154
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 0.19059329849369813,
|
| 1139 |
+
"grad_norm": 0.26307248972641967,
|
| 1140 |
+
"learning_rate": 1.9511892335939904e-05,
|
| 1141 |
+
"loss": 0.2419,
|
| 1142 |
+
"step": 155
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.19182293267752842,
|
| 1146 |
+
"grad_norm": 0.3849365062506917,
|
| 1147 |
+
"learning_rate": 1.9498541588864022e-05,
|
| 1148 |
+
"loss": 0.2316,
|
| 1149 |
+
"step": 156
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 0.19305256686135874,
|
| 1153 |
+
"grad_norm": 0.2858572602811596,
|
| 1154 |
+
"learning_rate": 1.948501540476414e-05,
|
| 1155 |
+
"loss": 0.2242,
|
| 1156 |
+
"step": 157
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 0.19428220104518906,
|
| 1160 |
+
"grad_norm": 0.20651885971329154,
|
| 1161 |
+
"learning_rate": 1.9471314033467413e-05,
|
| 1162 |
+
"loss": 0.2597,
|
| 1163 |
+
"step": 158
|
| 1164 |
+
},
|
| 1165 |
+
{
|
| 1166 |
+
"epoch": 0.19551183522901935,
|
| 1167 |
+
"grad_norm": 0.24219070077131208,
|
| 1168 |
+
"learning_rate": 1.945743772803666e-05,
|
| 1169 |
+
"loss": 0.1932,
|
| 1170 |
+
"step": 159
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 0.19674146941284967,
|
| 1174 |
+
"grad_norm": 0.2870212902155916,
|
| 1175 |
+
"learning_rate": 1.9443386744765726e-05,
|
| 1176 |
+
"loss": 0.256,
|
| 1177 |
+
"step": 160
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 0.19674146941284967,
|
| 1181 |
+
"eval_accuracy": 0.8021390374331551,
|
| 1182 |
+
"eval_f1": 0.4931506849315068,
|
| 1183 |
+
"eval_loss": 0.43281251192092896,
|
| 1184 |
+
"eval_precision": 0.782608695652174,
|
| 1185 |
+
"eval_recall": 0.36,
|
| 1186 |
+
"eval_runtime": 23.7371,
|
| 1187 |
+
"eval_samples_per_second": 2.106,
|
| 1188 |
+
"eval_steps_per_second": 0.169,
|
| 1189 |
+
"step": 160
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.19797110359668,
|
| 1193 |
+
"grad_norm": 0.2614284036745401,
|
| 1194 |
+
"learning_rate": 1.942916134317473e-05,
|
| 1195 |
+
"loss": 0.2436,
|
| 1196 |
+
"step": 161
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.1992007377805103,
|
| 1200 |
+
"grad_norm": 0.26937001401772626,
|
| 1201 |
+
"learning_rate": 1.9414761786005293e-05,
|
| 1202 |
+
"loss": 0.1725,
|
| 1203 |
+
"step": 162
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.2004303719643406,
|
| 1207 |
+
"grad_norm": 0.28202676187309017,
|
| 1208 |
+
"learning_rate": 1.9400188339215657e-05,
|
| 1209 |
+
"loss": 0.2591,
|
| 1210 |
+
"step": 163
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.20166000614817092,
|
| 1214 |
+
"grad_norm": 0.27016058532381143,
|
| 1215 |
+
"learning_rate": 1.9385441271975786e-05,
|
| 1216 |
+
"loss": 0.2003,
|
| 1217 |
+
"step": 164
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 0.20288964033200124,
|
| 1221 |
+
"grad_norm": 0.20605906416711317,
|
| 1222 |
+
"learning_rate": 1.9370520856662406e-05,
|
| 1223 |
+
"loss": 0.1778,
|
| 1224 |
+
"step": 165
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.20411927451583153,
|
| 1228 |
+
"grad_norm": 0.21687941485697337,
|
| 1229 |
+
"learning_rate": 1.9355427368853946e-05,
|
| 1230 |
+
"loss": 0.2145,
|
| 1231 |
+
"step": 166
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 0.20534890869966185,
|
| 1235 |
+
"grad_norm": 0.4233260372619336,
|
| 1236 |
+
"learning_rate": 1.9340161087325483e-05,
|
| 1237 |
+
"loss": 0.1657,
|
| 1238 |
+
"step": 167
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.20657854288349217,
|
| 1242 |
+
"grad_norm": 0.26680222767798134,
|
| 1243 |
+
"learning_rate": 1.932472229404356e-05,
|
| 1244 |
+
"loss": 0.1846,
|
| 1245 |
+
"step": 168
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.20780817706732246,
|
| 1249 |
+
"grad_norm": 0.3380713604084801,
|
| 1250 |
+
"learning_rate": 1.9309111274161005e-05,
|
| 1251 |
+
"loss": 0.2896,
|
| 1252 |
+
"step": 169
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.20903781125115278,
|
| 1256 |
+
"grad_norm": 0.3308557221991788,
|
| 1257 |
+
"learning_rate": 1.9293328316011645e-05,
|
| 1258 |
+
"loss": 0.2199,
|
| 1259 |
+
"step": 170
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 0.2102674454349831,
|
| 1263 |
+
"grad_norm": 0.3136020181316013,
|
| 1264 |
+
"learning_rate": 1.927737371110499e-05,
|
| 1265 |
+
"loss": 0.213,
|
| 1266 |
+
"step": 171
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.2114970796188134,
|
| 1270 |
+
"grad_norm": 0.35343090913220965,
|
| 1271 |
+
"learning_rate": 1.9261247754120846e-05,
|
| 1272 |
+
"loss": 0.2322,
|
| 1273 |
+
"step": 172
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.2127267138026437,
|
| 1277 |
+
"grad_norm": 0.2824826725729394,
|
| 1278 |
+
"learning_rate": 1.924495074290388e-05,
|
| 1279 |
+
"loss": 0.2523,
|
| 1280 |
+
"step": 173
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.21395634798647403,
|
| 1284 |
+
"grad_norm": 0.34041228997887535,
|
| 1285 |
+
"learning_rate": 1.92284829784581e-05,
|
| 1286 |
+
"loss": 0.2854,
|
| 1287 |
+
"step": 174
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 0.21518598217030432,
|
| 1291 |
+
"grad_norm": 0.3318426940507861,
|
| 1292 |
+
"learning_rate": 1.9211844764941318e-05,
|
| 1293 |
+
"loss": 0.1669,
|
| 1294 |
+
"step": 175
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 0.21641561635413464,
|
| 1298 |
+
"grad_norm": 0.3090750522717658,
|
| 1299 |
+
"learning_rate": 1.919503640965951e-05,
|
| 1300 |
+
"loss": 0.1843,
|
| 1301 |
+
"step": 176
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.21764525053796496,
|
| 1305 |
+
"grad_norm": 0.297018364631407,
|
| 1306 |
+
"learning_rate": 1.917805822306117e-05,
|
| 1307 |
+
"loss": 0.2038,
|
| 1308 |
+
"step": 177
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.21887488472179525,
|
| 1312 |
+
"grad_norm": 0.2679923976809244,
|
| 1313 |
+
"learning_rate": 1.916091051873154e-05,
|
| 1314 |
+
"loss": 0.14,
|
| 1315 |
+
"step": 178
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 0.22010451890562557,
|
| 1319 |
+
"grad_norm": 0.3130518688020483,
|
| 1320 |
+
"learning_rate": 1.9143593613386845e-05,
|
| 1321 |
+
"loss": 0.1871,
|
| 1322 |
+
"step": 179
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.2213341530894559,
|
| 1326 |
+
"grad_norm": 0.38588801765002845,
|
| 1327 |
+
"learning_rate": 1.9126107826868436e-05,
|
| 1328 |
+
"loss": 0.275,
|
| 1329 |
+
"step": 180
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 0.22256378727328618,
|
| 1333 |
+
"grad_norm": 0.3157899959865116,
|
| 1334 |
+
"learning_rate": 1.9108453482136866e-05,
|
| 1335 |
+
"loss": 0.2098,
|
| 1336 |
+
"step": 181
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 0.2237934214571165,
|
| 1340 |
+
"grad_norm": 0.3517837295222883,
|
| 1341 |
+
"learning_rate": 1.9090630905265963e-05,
|
| 1342 |
+
"loss": 0.2855,
|
| 1343 |
+
"step": 182
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.22502305564094682,
|
| 1347 |
+
"grad_norm": 0.3092532190930959,
|
| 1348 |
+
"learning_rate": 1.9072640425436762e-05,
|
| 1349 |
+
"loss": 0.2278,
|
| 1350 |
+
"step": 183
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.22625268982477714,
|
| 1354 |
+
"grad_norm": 0.38543714174275906,
|
| 1355 |
+
"learning_rate": 1.905448237493147e-05,
|
| 1356 |
+
"loss": 0.289,
|
| 1357 |
+
"step": 184
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.22748232400860743,
|
| 1361 |
+
"grad_norm": 0.3569146643137087,
|
| 1362 |
+
"learning_rate": 1.9036157089127278e-05,
|
| 1363 |
+
"loss": 0.2716,
|
| 1364 |
+
"step": 185
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.22871195819243775,
|
| 1368 |
+
"grad_norm": 0.40683148192292634,
|
| 1369 |
+
"learning_rate": 1.901766490649022e-05,
|
| 1370 |
+
"loss": 0.2983,
|
| 1371 |
+
"step": 186
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.22994159237626807,
|
| 1375 |
+
"grad_norm": 0.4253940265898542,
|
| 1376 |
+
"learning_rate": 1.8999006168568883e-05,
|
| 1377 |
+
"loss": 0.2284,
|
| 1378 |
+
"step": 187
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.23117122656009836,
|
| 1382 |
+
"grad_norm": 0.26669631723987736,
|
| 1383 |
+
"learning_rate": 1.8980181219988117e-05,
|
| 1384 |
+
"loss": 0.1757,
|
| 1385 |
+
"step": 188
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.23240086074392868,
|
| 1389 |
+
"grad_norm": 0.2721222099304509,
|
| 1390 |
+
"learning_rate": 1.8961190408442662e-05,
|
| 1391 |
+
"loss": 0.2298,
|
| 1392 |
+
"step": 189
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.233630494927759,
|
| 1396 |
+
"grad_norm": 0.36072785713712757,
|
| 1397 |
+
"learning_rate": 1.8942034084690727e-05,
|
| 1398 |
+
"loss": 0.2847,
|
| 1399 |
+
"step": 190
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.2348601291115893,
|
| 1403 |
+
"grad_norm": 0.34669408899608534,
|
| 1404 |
+
"learning_rate": 1.8922712602547516e-05,
|
| 1405 |
+
"loss": 0.2453,
|
| 1406 |
+
"step": 191
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.2360897632954196,
|
| 1410 |
+
"grad_norm": 0.3295391150787403,
|
| 1411 |
+
"learning_rate": 1.89032263188787e-05,
|
| 1412 |
+
"loss": 0.2633,
|
| 1413 |
+
"step": 192
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.23731939747924993,
|
| 1417 |
+
"grad_norm": 0.2836153191531997,
|
| 1418 |
+
"learning_rate": 1.8883575593593793e-05,
|
| 1419 |
+
"loss": 0.2218,
|
| 1420 |
+
"step": 193
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 0.23854903166308022,
|
| 1424 |
+
"grad_norm": 0.2680039523838409,
|
| 1425 |
+
"learning_rate": 1.8863760789639548e-05,
|
| 1426 |
+
"loss": 0.2422,
|
| 1427 |
+
"step": 194
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 0.23977866584691054,
|
| 1431 |
+
"grad_norm": 0.18616539199723142,
|
| 1432 |
+
"learning_rate": 1.8843782272993225e-05,
|
| 1433 |
+
"loss": 0.1552,
|
| 1434 |
+
"step": 195
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.24100830003074086,
|
| 1438 |
+
"grad_norm": 0.2810251721018552,
|
| 1439 |
+
"learning_rate": 1.8823640412655844e-05,
|
| 1440 |
+
"loss": 0.1982,
|
| 1441 |
+
"step": 196
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.24223793421457115,
|
| 1445 |
+
"grad_norm": 0.2586309972920896,
|
| 1446 |
+
"learning_rate": 1.880333558064536e-05,
|
| 1447 |
+
"loss": 0.2115,
|
| 1448 |
+
"step": 197
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.24346756839840147,
|
| 1452 |
+
"grad_norm": 0.2529365933222149,
|
| 1453 |
+
"learning_rate": 1.878286815198979e-05,
|
| 1454 |
+
"loss": 0.2142,
|
| 1455 |
+
"step": 198
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 0.2446972025822318,
|
| 1459 |
+
"grad_norm": 0.376229567543972,
|
| 1460 |
+
"learning_rate": 1.876223850472032e-05,
|
| 1461 |
+
"loss": 0.2328,
|
| 1462 |
+
"step": 199
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 0.24592683676606208,
|
| 1466 |
+
"grad_norm": 0.23504541587554245,
|
| 1467 |
+
"learning_rate": 1.8741447019864263e-05,
|
| 1468 |
+
"loss": 0.2062,
|
| 1469 |
+
"step": 200
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.24592683676606208,
|
| 1473 |
+
"eval_accuracy": 0.786096256684492,
|
| 1474 |
+
"eval_f1": 0.42857142857142855,
|
| 1475 |
+
"eval_loss": 0.46992188692092896,
|
| 1476 |
+
"eval_precision": 0.75,
|
| 1477 |
+
"eval_recall": 0.3,
|
| 1478 |
+
"eval_runtime": 23.5849,
|
| 1479 |
+
"eval_samples_per_second": 2.12,
|
| 1480 |
+
"eval_steps_per_second": 0.17,
|
| 1481 |
+
"step": 200
|
| 1482 |
}
|
| 1483 |
],
|
| 1484 |
"logging_steps": 1,
|
|
|
|
| 1498 |
"attributes": {}
|
| 1499 |
}
|
| 1500 |
},
|
| 1501 |
+
"total_flos": 197328810967040.0,
|
| 1502 |
"train_batch_size": 4,
|
| 1503 |
"trial_name": null,
|
| 1504 |
"trial_params": null
|