Training in progress, step 5000, checkpoint
Browse files- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/optimizer.pt +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/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +783 -3
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 570000656
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31dda666ef4dd077b95b2dd599130970195a63112fb34ebe2cea847e829d411c
|
| 3 |
size 570000656
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1622091
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fc58d84e7ac21638657925d41422fb7d57079f769226403669f908cbea736d76
|
| 3 |
size 1622091
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99c87c9fc366387cf9786f9653243f9c44c717b64598437d407912326159dd35
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:feb59a4cd63d6473a172790b98d81273192b04584c89ada56fbb5e0345fb77f5
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:32c5b8df44b7620babbac136f8d60d3181b4cd84c7d8e94dad8faf113cb87f0e
|
| 3 |
size 14469
|
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 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccc3b91a9e9cb171e2bca9bd53a21ec9407b75b1887e08091f277959636c5675
|
| 3 |
size 14469
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3b73a86b0118832f0ed2d1762ada435144272d934c4253e44e24943429f32374
|
| 3 |
size 14469
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31bb6385598d89d4a03aff83885a5694f826c4b88813a1e38d655af9cd4337f3
|
| 3 |
size 14469
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ce362a133b33d0f2d7103dbe6e2283c6f1977e9d9ba0616cf26f73d1dde9339
|
| 3 |
size 14469
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14469
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f80a0503a9bd73c9f92b2afc522c0030f38431ef3c416c26f5095a53a96fe818
|
| 3 |
size 14469
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:40853ecc212e01e867ec0a5405df4309683b8c34e06986a13d1a856a47ba9534
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 100,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -3128,6 +3128,786 @@
|
|
| 3128 |
"eval_samples_per_second": 16.389,
|
| 3129 |
"eval_steps_per_second": 0.911,
|
| 3130 |
"step": 4000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3131 |
}
|
| 3132 |
],
|
| 3133 |
"logging_steps": 10,
|
|
@@ -3147,7 +3927,7 @@
|
|
| 3147 |
"attributes": {}
|
| 3148 |
}
|
| 3149 |
},
|
| 3150 |
-
"total_flos":
|
| 3151 |
"train_batch_size": 4,
|
| 3152 |
"trial_name": null,
|
| 3153 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.45454545454545453,
|
| 6 |
"eval_steps": 100,
|
| 7 |
+
"global_step": 5000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": false,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 3128 |
"eval_samples_per_second": 16.389,
|
| 3129 |
"eval_steps_per_second": 0.911,
|
| 3130 |
"step": 4000
|
| 3131 |
+
},
|
| 3132 |
+
{
|
| 3133 |
+
"epoch": 0.36454545454545456,
|
| 3134 |
+
"grad_norm": 526528664305664.0,
|
| 3135 |
+
"learning_rate": 7.065377653279787e-07,
|
| 3136 |
+
"loss": 26.51700439453125,
|
| 3137 |
+
"step": 4010
|
| 3138 |
+
},
|
| 3139 |
+
{
|
| 3140 |
+
"epoch": 0.3654545454545455,
|
| 3141 |
+
"grad_norm": 177814212968448.0,
|
| 3142 |
+
"learning_rate": 7.052364528176298e-07,
|
| 3143 |
+
"loss": 26.30553283691406,
|
| 3144 |
+
"step": 4020
|
| 3145 |
+
},
|
| 3146 |
+
{
|
| 3147 |
+
"epoch": 0.3663636363636364,
|
| 3148 |
+
"grad_norm": 95500158107648.0,
|
| 3149 |
+
"learning_rate": 7.03933466256685e-07,
|
| 3150 |
+
"loss": 26.378070068359374,
|
| 3151 |
+
"step": 4030
|
| 3152 |
+
},
|
| 3153 |
+
{
|
| 3154 |
+
"epoch": 0.36727272727272725,
|
| 3155 |
+
"grad_norm": 316484127555584.0,
|
| 3156 |
+
"learning_rate": 7.026288162732046e-07,
|
| 3157 |
+
"loss": 26.92432556152344,
|
| 3158 |
+
"step": 4040
|
| 3159 |
+
},
|
| 3160 |
+
{
|
| 3161 |
+
"epoch": 0.36818181818181817,
|
| 3162 |
+
"grad_norm": 296097897512960.0,
|
| 3163 |
+
"learning_rate": 7.013225135088171e-07,
|
| 3164 |
+
"loss": 26.409463500976564,
|
| 3165 |
+
"step": 4050
|
| 3166 |
+
},
|
| 3167 |
+
{
|
| 3168 |
+
"epoch": 0.3690909090909091,
|
| 3169 |
+
"grad_norm": 146435114795008.0,
|
| 3170 |
+
"learning_rate": 7.000145686186323e-07,
|
| 3171 |
+
"loss": 26.875320434570312,
|
| 3172 |
+
"step": 4060
|
| 3173 |
+
},
|
| 3174 |
+
{
|
| 3175 |
+
"epoch": 0.37,
|
| 3176 |
+
"grad_norm": 90155608178688.0,
|
| 3177 |
+
"learning_rate": 6.987049922711543e-07,
|
| 3178 |
+
"loss": 26.703628540039062,
|
| 3179 |
+
"step": 4070
|
| 3180 |
+
},
|
| 3181 |
+
{
|
| 3182 |
+
"epoch": 0.3709090909090909,
|
| 3183 |
+
"grad_norm": 169079994318848.0,
|
| 3184 |
+
"learning_rate": 6.973937951481943e-07,
|
| 3185 |
+
"loss": 26.59369812011719,
|
| 3186 |
+
"step": 4080
|
| 3187 |
+
},
|
| 3188 |
+
{
|
| 3189 |
+
"epoch": 0.3718181818181818,
|
| 3190 |
+
"grad_norm": 224013699776512.0,
|
| 3191 |
+
"learning_rate": 6.960809879447838e-07,
|
| 3192 |
+
"loss": 26.508026123046875,
|
| 3193 |
+
"step": 4090
|
| 3194 |
+
},
|
| 3195 |
+
{
|
| 3196 |
+
"epoch": 0.37272727272727274,
|
| 3197 |
+
"grad_norm": 368716332138496.0,
|
| 3198 |
+
"learning_rate": 6.947665813690871e-07,
|
| 3199 |
+
"loss": 26.604330444335936,
|
| 3200 |
+
"step": 4100
|
| 3201 |
+
},
|
| 3202 |
+
{
|
| 3203 |
+
"epoch": 0.37272727272727274,
|
| 3204 |
+
"eval_loss": 26.31926727294922,
|
| 3205 |
+
"eval_runtime": 2.1723,
|
| 3206 |
+
"eval_samples_per_second": 16.572,
|
| 3207 |
+
"eval_steps_per_second": 0.921,
|
| 3208 |
+
"step": 4100
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.37363636363636366,
|
| 3212 |
+
"grad_norm": 416285577969664.0,
|
| 3213 |
+
"learning_rate": 6.934505861423144e-07,
|
| 3214 |
+
"loss": 26.69903564453125,
|
| 3215 |
+
"step": 4110
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.37454545454545457,
|
| 3219 |
+
"grad_norm": 192259412721664.0,
|
| 3220 |
+
"learning_rate": 6.921330129986338e-07,
|
| 3221 |
+
"loss": 26.503787231445312,
|
| 3222 |
+
"step": 4120
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.37545454545454543,
|
| 3226 |
+
"grad_norm": 605233705123840.0,
|
| 3227 |
+
"learning_rate": 6.90813872685084e-07,
|
| 3228 |
+
"loss": 26.778439331054688,
|
| 3229 |
+
"step": 4130
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.37636363636363634,
|
| 3233 |
+
"grad_norm": 265230370209792.0,
|
| 3234 |
+
"learning_rate": 6.894931759614865e-07,
|
| 3235 |
+
"loss": 27.23222961425781,
|
| 3236 |
+
"step": 4140
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.37727272727272726,
|
| 3240 |
+
"grad_norm": 423306272440320.0,
|
| 3241 |
+
"learning_rate": 6.881709336003584e-07,
|
| 3242 |
+
"loss": 26.96778564453125,
|
| 3243 |
+
"step": 4150
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.3781818181818182,
|
| 3247 |
+
"grad_norm": 343047359430656.0,
|
| 3248 |
+
"learning_rate": 6.868471563868235e-07,
|
| 3249 |
+
"loss": 26.680606079101562,
|
| 3250 |
+
"step": 4160
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.3790909090909091,
|
| 3254 |
+
"grad_norm": 266783386763264.0,
|
| 3255 |
+
"learning_rate": 6.855218551185254e-07,
|
| 3256 |
+
"loss": 26.349310302734374,
|
| 3257 |
+
"step": 4170
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.38,
|
| 3261 |
+
"grad_norm": 207645310976000.0,
|
| 3262 |
+
"learning_rate": 6.841950406055389e-07,
|
| 3263 |
+
"loss": 26.872479248046876,
|
| 3264 |
+
"step": 4180
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.3809090909090909,
|
| 3268 |
+
"grad_norm": 227184543268864.0,
|
| 3269 |
+
"learning_rate": 6.828667236702815e-07,
|
| 3270 |
+
"loss": 26.167431640625,
|
| 3271 |
+
"step": 4190
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.38181818181818183,
|
| 3275 |
+
"grad_norm": 332900432084992.0,
|
| 3276 |
+
"learning_rate": 6.815369151474256e-07,
|
| 3277 |
+
"loss": 26.787005615234374,
|
| 3278 |
+
"step": 4200
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.38181818181818183,
|
| 3282 |
+
"eval_loss": 26.321739196777344,
|
| 3283 |
+
"eval_runtime": 2.1695,
|
| 3284 |
+
"eval_samples_per_second": 16.594,
|
| 3285 |
+
"eval_steps_per_second": 0.922,
|
| 3286 |
+
"step": 4200
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 0.38272727272727275,
|
| 3290 |
+
"grad_norm": 98059253972992.0,
|
| 3291 |
+
"learning_rate": 6.802056258838104e-07,
|
| 3292 |
+
"loss": 26.625421142578126,
|
| 3293 |
+
"step": 4210
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 0.3836363636363636,
|
| 3297 |
+
"grad_norm": 241705676701696.0,
|
| 3298 |
+
"learning_rate": 6.788728667383528e-07,
|
| 3299 |
+
"loss": 26.678131103515625,
|
| 3300 |
+
"step": 4220
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 0.3845454545454545,
|
| 3304 |
+
"grad_norm": 150195207667712.0,
|
| 3305 |
+
"learning_rate": 6.775386485819589e-07,
|
| 3306 |
+
"loss": 26.601348876953125,
|
| 3307 |
+
"step": 4230
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 0.38545454545454544,
|
| 3311 |
+
"grad_norm": 184757732245504.0,
|
| 3312 |
+
"learning_rate": 6.762029822974359e-07,
|
| 3313 |
+
"loss": 26.485671997070312,
|
| 3314 |
+
"step": 4240
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 0.38636363636363635,
|
| 3318 |
+
"grad_norm": 161859432873984.0,
|
| 3319 |
+
"learning_rate": 6.748658787794025e-07,
|
| 3320 |
+
"loss": 26.345684814453126,
|
| 3321 |
+
"step": 4250
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.38727272727272727,
|
| 3325 |
+
"grad_norm": 142884854562816.0,
|
| 3326 |
+
"learning_rate": 6.735273489342008e-07,
|
| 3327 |
+
"loss": 26.75887451171875,
|
| 3328 |
+
"step": 4260
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 0.3881818181818182,
|
| 3332 |
+
"grad_norm": 439033939361792.0,
|
| 3333 |
+
"learning_rate": 6.721874036798068e-07,
|
| 3334 |
+
"loss": 26.148809814453124,
|
| 3335 |
+
"step": 4270
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 0.3890909090909091,
|
| 3339 |
+
"grad_norm": 309939335593984.0,
|
| 3340 |
+
"learning_rate": 6.708460539457417e-07,
|
| 3341 |
+
"loss": 26.634625244140626,
|
| 3342 |
+
"step": 4280
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 0.39,
|
| 3346 |
+
"grad_norm": 120628896071680.0,
|
| 3347 |
+
"learning_rate": 6.695033106729824e-07,
|
| 3348 |
+
"loss": 26.592654418945312,
|
| 3349 |
+
"step": 4290
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 0.39090909090909093,
|
| 3353 |
+
"grad_norm": 194318715322368.0,
|
| 3354 |
+
"learning_rate": 6.681591848138731e-07,
|
| 3355 |
+
"loss": 26.4625732421875,
|
| 3356 |
+
"step": 4300
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 0.39090909090909093,
|
| 3360 |
+
"eval_loss": 26.325037002563477,
|
| 3361 |
+
"eval_runtime": 2.1684,
|
| 3362 |
+
"eval_samples_per_second": 16.602,
|
| 3363 |
+
"eval_steps_per_second": 0.922,
|
| 3364 |
+
"step": 4300
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 0.39181818181818184,
|
| 3368 |
+
"grad_norm": 399684388519936.0,
|
| 3369 |
+
"learning_rate": 6.668136873320344e-07,
|
| 3370 |
+
"loss": 26.953164672851564,
|
| 3371 |
+
"step": 4310
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.3927272727272727,
|
| 3375 |
+
"grad_norm": 109830022889472.0,
|
| 3376 |
+
"learning_rate": 6.654668292022754e-07,
|
| 3377 |
+
"loss": 26.49449768066406,
|
| 3378 |
+
"step": 4320
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.3936363636363636,
|
| 3382 |
+
"grad_norm": 92155989196800.0,
|
| 3383 |
+
"learning_rate": 6.641186214105033e-07,
|
| 3384 |
+
"loss": 26.64600524902344,
|
| 3385 |
+
"step": 4330
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.39454545454545453,
|
| 3389 |
+
"grad_norm": 276388091265024.0,
|
| 3390 |
+
"learning_rate": 6.627690749536346e-07,
|
| 3391 |
+
"loss": 26.74027099609375,
|
| 3392 |
+
"step": 4340
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.39545454545454545,
|
| 3396 |
+
"grad_norm": 151451183611904.0,
|
| 3397 |
+
"learning_rate": 6.614182008395042e-07,
|
| 3398 |
+
"loss": 26.55115661621094,
|
| 3399 |
+
"step": 4350
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.39636363636363636,
|
| 3403 |
+
"grad_norm": 375987342671872.0,
|
| 3404 |
+
"learning_rate": 6.600660100867765e-07,
|
| 3405 |
+
"loss": 26.10811767578125,
|
| 3406 |
+
"step": 4360
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.3972727272727273,
|
| 3410 |
+
"grad_norm": 559519281184768.0,
|
| 3411 |
+
"learning_rate": 6.587125137248558e-07,
|
| 3412 |
+
"loss": 26.605978393554686,
|
| 3413 |
+
"step": 4370
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.3981818181818182,
|
| 3417 |
+
"grad_norm": 363120157523968.0,
|
| 3418 |
+
"learning_rate": 6.573577227937951e-07,
|
| 3419 |
+
"loss": 26.561282348632812,
|
| 3420 |
+
"step": 4380
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 0.3990909090909091,
|
| 3424 |
+
"grad_norm": 394054760136704.0,
|
| 3425 |
+
"learning_rate": 6.560016483442075e-07,
|
| 3426 |
+
"loss": 26.416183471679688,
|
| 3427 |
+
"step": 4390
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 0.4,
|
| 3431 |
+
"grad_norm": 291861650472960.0,
|
| 3432 |
+
"learning_rate": 6.546443014371745e-07,
|
| 3433 |
+
"loss": 26.458407592773437,
|
| 3434 |
+
"step": 4400
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 0.4,
|
| 3438 |
+
"eval_loss": 26.32141876220703,
|
| 3439 |
+
"eval_runtime": 2.1831,
|
| 3440 |
+
"eval_samples_per_second": 16.49,
|
| 3441 |
+
"eval_steps_per_second": 0.916,
|
| 3442 |
+
"step": 4400
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 0.4009090909090909,
|
| 3446 |
+
"grad_norm": 84496426729472.0,
|
| 3447 |
+
"learning_rate": 6.53285693144158e-07,
|
| 3448 |
+
"loss": 26.468145751953124,
|
| 3449 |
+
"step": 4410
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 0.4018181818181818,
|
| 3453 |
+
"grad_norm": 190245425381376.0,
|
| 3454 |
+
"learning_rate": 6.51925834546907e-07,
|
| 3455 |
+
"loss": 26.321368408203124,
|
| 3456 |
+
"step": 4420
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 0.4027272727272727,
|
| 3460 |
+
"grad_norm": 88420181344256.0,
|
| 3461 |
+
"learning_rate": 6.5056473673737e-07,
|
| 3462 |
+
"loss": 26.454937744140626,
|
| 3463 |
+
"step": 4430
|
| 3464 |
+
},
|
| 3465 |
+
{
|
| 3466 |
+
"epoch": 0.4036363636363636,
|
| 3467 |
+
"grad_norm": 332825404375040.0,
|
| 3468 |
+
"learning_rate": 6.49202410817603e-07,
|
| 3469 |
+
"loss": 26.5259521484375,
|
| 3470 |
+
"step": 4440
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 0.40454545454545454,
|
| 3474 |
+
"grad_norm": 497012105543680.0,
|
| 3475 |
+
"learning_rate": 6.478388678996796e-07,
|
| 3476 |
+
"loss": 26.6334228515625,
|
| 3477 |
+
"step": 4450
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 0.40545454545454546,
|
| 3481 |
+
"grad_norm": 96112023175168.0,
|
| 3482 |
+
"learning_rate": 6.464741191055994e-07,
|
| 3483 |
+
"loss": 26.60364990234375,
|
| 3484 |
+
"step": 4460
|
| 3485 |
+
},
|
| 3486 |
+
{
|
| 3487 |
+
"epoch": 0.40636363636363637,
|
| 3488 |
+
"grad_norm": 149687059349504.0,
|
| 3489 |
+
"learning_rate": 6.451081755671985e-07,
|
| 3490 |
+
"loss": 26.84015808105469,
|
| 3491 |
+
"step": 4470
|
| 3492 |
+
},
|
| 3493 |
+
{
|
| 3494 |
+
"epoch": 0.4072727272727273,
|
| 3495 |
+
"grad_norm": 78247802961920.0,
|
| 3496 |
+
"learning_rate": 6.437410484260583e-07,
|
| 3497 |
+
"loss": 26.285125732421875,
|
| 3498 |
+
"step": 4480
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 0.4081818181818182,
|
| 3502 |
+
"grad_norm": 593727554846720.0,
|
| 3503 |
+
"learning_rate": 6.42372748833414e-07,
|
| 3504 |
+
"loss": 26.160516357421876,
|
| 3505 |
+
"step": 4490
|
| 3506 |
+
},
|
| 3507 |
+
{
|
| 3508 |
+
"epoch": 0.4090909090909091,
|
| 3509 |
+
"grad_norm": 153949931831296.0,
|
| 3510 |
+
"learning_rate": 6.410032879500646e-07,
|
| 3511 |
+
"loss": 26.74964294433594,
|
| 3512 |
+
"step": 4500
|
| 3513 |
+
},
|
| 3514 |
+
{
|
| 3515 |
+
"epoch": 0.4090909090909091,
|
| 3516 |
+
"eval_loss": 26.31894874572754,
|
| 3517 |
+
"eval_runtime": 2.1406,
|
| 3518 |
+
"eval_samples_per_second": 16.818,
|
| 3519 |
+
"eval_steps_per_second": 0.934,
|
| 3520 |
+
"step": 4500
|
| 3521 |
+
},
|
| 3522 |
+
{
|
| 3523 |
+
"epoch": 0.41,
|
| 3524 |
+
"grad_norm": 232447153274880.0,
|
| 3525 |
+
"learning_rate": 6.396326769462811e-07,
|
| 3526 |
+
"loss": 26.883270263671875,
|
| 3527 |
+
"step": 4510
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 0.4109090909090909,
|
| 3531 |
+
"grad_norm": 298445030031360.0,
|
| 3532 |
+
"learning_rate": 6.382609270017154e-07,
|
| 3533 |
+
"loss": 26.760006713867188,
|
| 3534 |
+
"step": 4520
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 0.4118181818181818,
|
| 3538 |
+
"grad_norm": 72679117815808.0,
|
| 3539 |
+
"learning_rate": 6.3688804930531e-07,
|
| 3540 |
+
"loss": 26.37138671875,
|
| 3541 |
+
"step": 4530
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 0.4127272727272727,
|
| 3545 |
+
"grad_norm": 141969456103424.0,
|
| 3546 |
+
"learning_rate": 6.355140550552058e-07,
|
| 3547 |
+
"loss": 26.522723388671874,
|
| 3548 |
+
"step": 4540
|
| 3549 |
+
},
|
| 3550 |
+
{
|
| 3551 |
+
"epoch": 0.41363636363636364,
|
| 3552 |
+
"grad_norm": 160081098309632.0,
|
| 3553 |
+
"learning_rate": 6.341389554586511e-07,
|
| 3554 |
+
"loss": 26.483316040039064,
|
| 3555 |
+
"step": 4550
|
| 3556 |
+
},
|
| 3557 |
+
{
|
| 3558 |
+
"epoch": 0.41454545454545455,
|
| 3559 |
+
"grad_norm": 324477833445376.0,
|
| 3560 |
+
"learning_rate": 6.327627617319103e-07,
|
| 3561 |
+
"loss": 26.508364868164062,
|
| 3562 |
+
"step": 4560
|
| 3563 |
+
},
|
| 3564 |
+
{
|
| 3565 |
+
"epoch": 0.41545454545454547,
|
| 3566 |
+
"grad_norm": 276562825969664.0,
|
| 3567 |
+
"learning_rate": 6.313854851001721e-07,
|
| 3568 |
+
"loss": 26.42707214355469,
|
| 3569 |
+
"step": 4570
|
| 3570 |
+
},
|
| 3571 |
+
{
|
| 3572 |
+
"epoch": 0.4163636363636364,
|
| 3573 |
+
"grad_norm": 307064727404544.0,
|
| 3574 |
+
"learning_rate": 6.300071367974582e-07,
|
| 3575 |
+
"loss": 26.318887329101564,
|
| 3576 |
+
"step": 4580
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 0.4172727272727273,
|
| 3580 |
+
"grad_norm": 178794975461376.0,
|
| 3581 |
+
"learning_rate": 6.286277280665315e-07,
|
| 3582 |
+
"loss": 26.655364990234375,
|
| 3583 |
+
"step": 4590
|
| 3584 |
+
},
|
| 3585 |
+
{
|
| 3586 |
+
"epoch": 0.41818181818181815,
|
| 3587 |
+
"grad_norm": 468974525480960.0,
|
| 3588 |
+
"learning_rate": 6.27247270158805e-07,
|
| 3589 |
+
"loss": 26.251632690429688,
|
| 3590 |
+
"step": 4600
|
| 3591 |
+
},
|
| 3592 |
+
{
|
| 3593 |
+
"epoch": 0.41818181818181815,
|
| 3594 |
+
"eval_loss": 26.31459617614746,
|
| 3595 |
+
"eval_runtime": 2.1711,
|
| 3596 |
+
"eval_samples_per_second": 16.581,
|
| 3597 |
+
"eval_steps_per_second": 0.921,
|
| 3598 |
+
"step": 4600
|
| 3599 |
+
},
|
| 3600 |
+
{
|
| 3601 |
+
"epoch": 0.41909090909090907,
|
| 3602 |
+
"grad_norm": 406509628424192.0,
|
| 3603 |
+
"learning_rate": 6.258657743342486e-07,
|
| 3604 |
+
"loss": 26.561279296875,
|
| 3605 |
+
"step": 4610
|
| 3606 |
+
},
|
| 3607 |
+
{
|
| 3608 |
+
"epoch": 0.42,
|
| 3609 |
+
"grad_norm": 152516016734208.0,
|
| 3610 |
+
"learning_rate": 6.244832518612989e-07,
|
| 3611 |
+
"loss": 26.661614990234376,
|
| 3612 |
+
"step": 4620
|
| 3613 |
+
},
|
| 3614 |
+
{
|
| 3615 |
+
"epoch": 0.4209090909090909,
|
| 3616 |
+
"grad_norm": 334286095908864.0,
|
| 3617 |
+
"learning_rate": 6.230997140167662e-07,
|
| 3618 |
+
"loss": 26.73835144042969,
|
| 3619 |
+
"step": 4630
|
| 3620 |
+
},
|
| 3621 |
+
{
|
| 3622 |
+
"epoch": 0.4218181818181818,
|
| 3623 |
+
"grad_norm": 109297497276416.0,
|
| 3624 |
+
"learning_rate": 6.217151720857432e-07,
|
| 3625 |
+
"loss": 26.554644775390624,
|
| 3626 |
+
"step": 4640
|
| 3627 |
+
},
|
| 3628 |
+
{
|
| 3629 |
+
"epoch": 0.42272727272727273,
|
| 3630 |
+
"grad_norm": 251784824094720.0,
|
| 3631 |
+
"learning_rate": 6.203296373615122e-07,
|
| 3632 |
+
"loss": 26.43360290527344,
|
| 3633 |
+
"step": 4650
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 0.42363636363636364,
|
| 3637 |
+
"grad_norm": 213711767732224.0,
|
| 3638 |
+
"learning_rate": 6.189431211454538e-07,
|
| 3639 |
+
"loss": 26.469012451171874,
|
| 3640 |
+
"step": 4660
|
| 3641 |
+
},
|
| 3642 |
+
{
|
| 3643 |
+
"epoch": 0.42454545454545456,
|
| 3644 |
+
"grad_norm": 119396852826112.0,
|
| 3645 |
+
"learning_rate": 6.175556347469541e-07,
|
| 3646 |
+
"loss": 26.383865356445312,
|
| 3647 |
+
"step": 4670
|
| 3648 |
+
},
|
| 3649 |
+
{
|
| 3650 |
+
"epoch": 0.4254545454545455,
|
| 3651 |
+
"grad_norm": 115268542005248.0,
|
| 3652 |
+
"learning_rate": 6.161671894833127e-07,
|
| 3653 |
+
"loss": 26.740725708007812,
|
| 3654 |
+
"step": 4680
|
| 3655 |
+
},
|
| 3656 |
+
{
|
| 3657 |
+
"epoch": 0.4263636363636364,
|
| 3658 |
+
"grad_norm": 311501025968128.0,
|
| 3659 |
+
"learning_rate": 6.147777966796505e-07,
|
| 3660 |
+
"loss": 26.425564575195313,
|
| 3661 |
+
"step": 4690
|
| 3662 |
+
},
|
| 3663 |
+
{
|
| 3664 |
+
"epoch": 0.42727272727272725,
|
| 3665 |
+
"grad_norm": 115545743556608.0,
|
| 3666 |
+
"learning_rate": 6.13387467668817e-07,
|
| 3667 |
+
"loss": 26.428509521484376,
|
| 3668 |
+
"step": 4700
|
| 3669 |
+
},
|
| 3670 |
+
{
|
| 3671 |
+
"epoch": 0.42727272727272725,
|
| 3672 |
+
"eval_loss": 26.31597137451172,
|
| 3673 |
+
"eval_runtime": 2.1641,
|
| 3674 |
+
"eval_samples_per_second": 16.635,
|
| 3675 |
+
"eval_steps_per_second": 0.924,
|
| 3676 |
+
"step": 4700
|
| 3677 |
+
},
|
| 3678 |
+
{
|
| 3679 |
+
"epoch": 0.42818181818181816,
|
| 3680 |
+
"grad_norm": 189478069075968.0,
|
| 3681 |
+
"learning_rate": 6.119962137912979e-07,
|
| 3682 |
+
"loss": 26.67225341796875,
|
| 3683 |
+
"step": 4710
|
| 3684 |
+
},
|
| 3685 |
+
{
|
| 3686 |
+
"epoch": 0.4290909090909091,
|
| 3687 |
+
"grad_norm": 335568076537856.0,
|
| 3688 |
+
"learning_rate": 6.106040463951236e-07,
|
| 3689 |
+
"loss": 26.434231567382813,
|
| 3690 |
+
"step": 4720
|
| 3691 |
+
},
|
| 3692 |
+
{
|
| 3693 |
+
"epoch": 0.43,
|
| 3694 |
+
"grad_norm": 208644343857152.0,
|
| 3695 |
+
"learning_rate": 6.092109768357747e-07,
|
| 3696 |
+
"loss": 26.637188720703126,
|
| 3697 |
+
"step": 4730
|
| 3698 |
+
},
|
| 3699 |
+
{
|
| 3700 |
+
"epoch": 0.4309090909090909,
|
| 3701 |
+
"grad_norm": 264374128214016.0,
|
| 3702 |
+
"learning_rate": 6.078170164760911e-07,
|
| 3703 |
+
"loss": 26.46937255859375,
|
| 3704 |
+
"step": 4740
|
| 3705 |
+
},
|
| 3706 |
+
{
|
| 3707 |
+
"epoch": 0.4318181818181818,
|
| 3708 |
+
"grad_norm": 281521348935680.0,
|
| 3709 |
+
"learning_rate": 6.064221766861785e-07,
|
| 3710 |
+
"loss": 26.524288940429688,
|
| 3711 |
+
"step": 4750
|
| 3712 |
+
},
|
| 3713 |
+
{
|
| 3714 |
+
"epoch": 0.43272727272727274,
|
| 3715 |
+
"grad_norm": 224812597248000.0,
|
| 3716 |
+
"learning_rate": 6.050264688433161e-07,
|
| 3717 |
+
"loss": 26.77779541015625,
|
| 3718 |
+
"step": 4760
|
| 3719 |
+
},
|
| 3720 |
+
{
|
| 3721 |
+
"epoch": 0.43363636363636365,
|
| 3722 |
+
"grad_norm": 225832148664320.0,
|
| 3723 |
+
"learning_rate": 6.036299043318631e-07,
|
| 3724 |
+
"loss": 26.691378784179687,
|
| 3725 |
+
"step": 4770
|
| 3726 |
+
},
|
| 3727 |
+
{
|
| 3728 |
+
"epoch": 0.43454545454545457,
|
| 3729 |
+
"grad_norm": 72125075423232.0,
|
| 3730 |
+
"learning_rate": 6.022324945431665e-07,
|
| 3731 |
+
"loss": 26.35833740234375,
|
| 3732 |
+
"step": 4780
|
| 3733 |
+
},
|
| 3734 |
+
{
|
| 3735 |
+
"epoch": 0.4354545454545454,
|
| 3736 |
+
"grad_norm": 279958349938688.0,
|
| 3737 |
+
"learning_rate": 6.008342508754682e-07,
|
| 3738 |
+
"loss": 26.597698974609376,
|
| 3739 |
+
"step": 4790
|
| 3740 |
+
},
|
| 3741 |
+
{
|
| 3742 |
+
"epoch": 0.43636363636363634,
|
| 3743 |
+
"grad_norm": 307064257642496.0,
|
| 3744 |
+
"learning_rate": 5.994351847338113e-07,
|
| 3745 |
+
"loss": 26.659085083007813,
|
| 3746 |
+
"step": 4800
|
| 3747 |
+
},
|
| 3748 |
+
{
|
| 3749 |
+
"epoch": 0.43636363636363634,
|
| 3750 |
+
"eval_loss": 26.316986083984375,
|
| 3751 |
+
"eval_runtime": 2.1561,
|
| 3752 |
+
"eval_samples_per_second": 16.696,
|
| 3753 |
+
"eval_steps_per_second": 0.928,
|
| 3754 |
+
"step": 4800
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.43727272727272726,
|
| 3758 |
+
"grad_norm": 171293076881408.0,
|
| 3759 |
+
"learning_rate": 5.980353075299481e-07,
|
| 3760 |
+
"loss": 26.330426025390626,
|
| 3761 |
+
"step": 4810
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.4381818181818182,
|
| 3765 |
+
"grad_norm": 248409114017792.0,
|
| 3766 |
+
"learning_rate": 5.966346306822458e-07,
|
| 3767 |
+
"loss": 26.47012939453125,
|
| 3768 |
+
"step": 4820
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.4390909090909091,
|
| 3772 |
+
"grad_norm": 204697034031104.0,
|
| 3773 |
+
"learning_rate": 5.95233165615595e-07,
|
| 3774 |
+
"loss": 26.68152770996094,
|
| 3775 |
+
"step": 4830
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.44,
|
| 3779 |
+
"grad_norm": 130607472443392.0,
|
| 3780 |
+
"learning_rate": 5.938309237613146e-07,
|
| 3781 |
+
"loss": 26.43096923828125,
|
| 3782 |
+
"step": 4840
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.4409090909090909,
|
| 3786 |
+
"grad_norm": 296476022407168.0,
|
| 3787 |
+
"learning_rate": 5.924279165570602e-07,
|
| 3788 |
+
"loss": 26.51325378417969,
|
| 3789 |
+
"step": 4850
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.44181818181818183,
|
| 3793 |
+
"grad_norm": 46552391352320.0,
|
| 3794 |
+
"learning_rate": 5.910241554467298e-07,
|
| 3795 |
+
"loss": 26.71526184082031,
|
| 3796 |
+
"step": 4860
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.44272727272727275,
|
| 3800 |
+
"grad_norm": 97249417756672.0,
|
| 3801 |
+
"learning_rate": 5.896196518803707e-07,
|
| 3802 |
+
"loss": 26.19861755371094,
|
| 3803 |
+
"step": 4870
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.44363636363636366,
|
| 3807 |
+
"grad_norm": 90017338753024.0,
|
| 3808 |
+
"learning_rate": 5.882144173140867e-07,
|
| 3809 |
+
"loss": 26.474087524414063,
|
| 3810 |
+
"step": 4880
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.4445454545454545,
|
| 3814 |
+
"grad_norm": 518103079321600.0,
|
| 3815 |
+
"learning_rate": 5.868084632099435e-07,
|
| 3816 |
+
"loss": 26.318057250976562,
|
| 3817 |
+
"step": 4890
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.44545454545454544,
|
| 3821 |
+
"grad_norm": 329086534680576.0,
|
| 3822 |
+
"learning_rate": 5.854018010358761e-07,
|
| 3823 |
+
"loss": 26.530551147460937,
|
| 3824 |
+
"step": 4900
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.44545454545454544,
|
| 3828 |
+
"eval_loss": 26.317190170288086,
|
| 3829 |
+
"eval_runtime": 2.1808,
|
| 3830 |
+
"eval_samples_per_second": 16.508,
|
| 3831 |
+
"eval_steps_per_second": 0.917,
|
| 3832 |
+
"step": 4900
|
| 3833 |
+
},
|
| 3834 |
+
{
|
| 3835 |
+
"epoch": 0.44636363636363635,
|
| 3836 |
+
"grad_norm": 272803269967872.0,
|
| 3837 |
+
"learning_rate": 5.839944422655952e-07,
|
| 3838 |
+
"loss": 26.631591796875,
|
| 3839 |
+
"step": 4910
|
| 3840 |
+
},
|
| 3841 |
+
{
|
| 3842 |
+
"epoch": 0.44727272727272727,
|
| 3843 |
+
"grad_norm": 622259425247232.0,
|
| 3844 |
+
"learning_rate": 5.825863983784931e-07,
|
| 3845 |
+
"loss": 26.726950073242186,
|
| 3846 |
+
"step": 4920
|
| 3847 |
+
},
|
| 3848 |
+
{
|
| 3849 |
+
"epoch": 0.4481818181818182,
|
| 3850 |
+
"grad_norm": 186363479588864.0,
|
| 3851 |
+
"learning_rate": 5.811776808595506e-07,
|
| 3852 |
+
"loss": 26.2813720703125,
|
| 3853 |
+
"step": 4930
|
| 3854 |
+
},
|
| 3855 |
+
{
|
| 3856 |
+
"epoch": 0.4490909090909091,
|
| 3857 |
+
"grad_norm": 155909409996800.0,
|
| 3858 |
+
"learning_rate": 5.797683011992432e-07,
|
| 3859 |
+
"loss": 26.543890380859374,
|
| 3860 |
+
"step": 4940
|
| 3861 |
+
},
|
| 3862 |
+
{
|
| 3863 |
+
"epoch": 0.45,
|
| 3864 |
+
"grad_norm": 232914314854400.0,
|
| 3865 |
+
"learning_rate": 5.783582708934468e-07,
|
| 3866 |
+
"loss": 26.611404418945312,
|
| 3867 |
+
"step": 4950
|
| 3868 |
+
},
|
| 3869 |
+
{
|
| 3870 |
+
"epoch": 0.4509090909090909,
|
| 3871 |
+
"grad_norm": 231428809818112.0,
|
| 3872 |
+
"learning_rate": 5.769476014433451e-07,
|
| 3873 |
+
"loss": 26.66617431640625,
|
| 3874 |
+
"step": 4960
|
| 3875 |
+
},
|
| 3876 |
+
{
|
| 3877 |
+
"epoch": 0.45181818181818184,
|
| 3878 |
+
"grad_norm": 342491630927872.0,
|
| 3879 |
+
"learning_rate": 5.755363043553346e-07,
|
| 3880 |
+
"loss": 26.529336547851564,
|
| 3881 |
+
"step": 4970
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 0.4527272727272727,
|
| 3885 |
+
"grad_norm": 133048012111872.0,
|
| 3886 |
+
"learning_rate": 5.741243911409314e-07,
|
| 3887 |
+
"loss": 26.72532958984375,
|
| 3888 |
+
"step": 4980
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 0.4536363636363636,
|
| 3892 |
+
"grad_norm": 286773523513344.0,
|
| 3893 |
+
"learning_rate": 5.72711873316677e-07,
|
| 3894 |
+
"loss": 26.638458251953125,
|
| 3895 |
+
"step": 4990
|
| 3896 |
+
},
|
| 3897 |
+
{
|
| 3898 |
+
"epoch": 0.45454545454545453,
|
| 3899 |
+
"grad_norm": 273551349252096.0,
|
| 3900 |
+
"learning_rate": 5.712987624040451e-07,
|
| 3901 |
+
"loss": 26.445010375976562,
|
| 3902 |
+
"step": 5000
|
| 3903 |
+
},
|
| 3904 |
+
{
|
| 3905 |
+
"epoch": 0.45454545454545453,
|
| 3906 |
+
"eval_loss": 26.315853118896484,
|
| 3907 |
+
"eval_runtime": 2.1781,
|
| 3908 |
+
"eval_samples_per_second": 16.528,
|
| 3909 |
+
"eval_steps_per_second": 0.918,
|
| 3910 |
+
"step": 5000
|
| 3911 |
}
|
| 3912 |
],
|
| 3913 |
"logging_steps": 10,
|
|
|
|
| 3927 |
"attributes": {}
|
| 3928 |
}
|
| 3929 |
},
|
| 3930 |
+
"total_flos": 4.529923817472e+16,
|
| 3931 |
"train_batch_size": 4,
|
| 3932 |
"trial_name": null,
|
| 3933 |
"trial_params": null
|