Training in progress, epoch 58, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13132
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:593f45feb57b6cf6b72deb4c2b645da0e09ef0560bb117ead2f8e48a01ba523c
|
| 3 |
size 13132
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 29970
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a83b182f0c5a15cc7cc60e05979466978862b18bd2943c1c557801b2452f1335
|
| 3 |
size 29970
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d6dbaea400350134c87609f2a6e7d568a19190ab54bccf06fc0c89f270eb6d2
|
| 3 |
size 14244
|
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 sha256:5815cf5dbd0e8e9b0a43084f1040b0b93a4a2d0c9a55323787cef2a001436293
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,10 +1,10 @@
|
|
| 1 |
{
|
| 2 |
-
"best_global_step":
|
| 3 |
-
"best_metric": 0.
|
| 4 |
-
"best_model_checkpoint": "./results/checkpoint-
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -23,6 +23,861 @@
|
|
| 23 |
"eval_samples_per_second": 644.821,
|
| 24 |
"eval_steps_per_second": 128.964,
|
| 25 |
"step": 1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
}
|
| 27 |
],
|
| 28 |
"logging_steps": 1,
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_global_step": 58,
|
| 3 |
+
"best_metric": 0.32613325119018555,
|
| 4 |
+
"best_model_checkpoint": "./results/checkpoint-58",
|
| 5 |
+
"epoch": 58.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 58,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 23 |
"eval_samples_per_second": 644.821,
|
| 24 |
"eval_steps_per_second": 128.964,
|
| 25 |
"step": 1
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 2.0,
|
| 29 |
+
"grad_norm": 0.8384326100349426,
|
| 30 |
+
"learning_rate": 9.9e-06,
|
| 31 |
+
"loss": 0.3286,
|
| 32 |
+
"step": 2
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_loss": 0.3285800814628601,
|
| 37 |
+
"eval_runtime": 0.0447,
|
| 38 |
+
"eval_samples_per_second": 223.596,
|
| 39 |
+
"eval_steps_per_second": 44.719,
|
| 40 |
+
"step": 2
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 3.0,
|
| 44 |
+
"grad_norm": 0.8382521867752075,
|
| 45 |
+
"learning_rate": 9.800000000000001e-06,
|
| 46 |
+
"loss": 0.3286,
|
| 47 |
+
"step": 3
|
| 48 |
+
},
|
| 49 |
+
{
|
| 50 |
+
"epoch": 3.0,
|
| 51 |
+
"eval_loss": 0.3285190165042877,
|
| 52 |
+
"eval_runtime": 0.0147,
|
| 53 |
+
"eval_samples_per_second": 681.358,
|
| 54 |
+
"eval_steps_per_second": 136.272,
|
| 55 |
+
"step": 3
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 4.0,
|
| 59 |
+
"grad_norm": 0.8380736112594604,
|
| 60 |
+
"learning_rate": 9.7e-06,
|
| 61 |
+
"loss": 0.3285,
|
| 62 |
+
"step": 4
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.0,
|
| 66 |
+
"eval_loss": 0.3284585475921631,
|
| 67 |
+
"eval_runtime": 0.015,
|
| 68 |
+
"eval_samples_per_second": 664.929,
|
| 69 |
+
"eval_steps_per_second": 132.986,
|
| 70 |
+
"step": 4
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 5.0,
|
| 74 |
+
"grad_norm": 0.8378969430923462,
|
| 75 |
+
"learning_rate": 9.600000000000001e-06,
|
| 76 |
+
"loss": 0.3285,
|
| 77 |
+
"step": 5
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 5.0,
|
| 81 |
+
"eval_loss": 0.32839876413345337,
|
| 82 |
+
"eval_runtime": 0.0146,
|
| 83 |
+
"eval_samples_per_second": 682.967,
|
| 84 |
+
"eval_steps_per_second": 136.593,
|
| 85 |
+
"step": 5
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 6.0,
|
| 89 |
+
"grad_norm": 0.8377220034599304,
|
| 90 |
+
"learning_rate": 9.5e-06,
|
| 91 |
+
"loss": 0.3284,
|
| 92 |
+
"step": 6
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 6.0,
|
| 96 |
+
"eval_loss": 0.3283396065235138,
|
| 97 |
+
"eval_runtime": 0.0171,
|
| 98 |
+
"eval_samples_per_second": 584.458,
|
| 99 |
+
"eval_steps_per_second": 116.892,
|
| 100 |
+
"step": 6
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 7.0,
|
| 104 |
+
"grad_norm": 0.8375489711761475,
|
| 105 |
+
"learning_rate": 9.4e-06,
|
| 106 |
+
"loss": 0.3283,
|
| 107 |
+
"step": 7
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 7.0,
|
| 111 |
+
"eval_loss": 0.328281044960022,
|
| 112 |
+
"eval_runtime": 0.0139,
|
| 113 |
+
"eval_samples_per_second": 719.089,
|
| 114 |
+
"eval_steps_per_second": 143.818,
|
| 115 |
+
"step": 7
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 8.0,
|
| 119 |
+
"grad_norm": 0.8373778462409973,
|
| 120 |
+
"learning_rate": 9.3e-06,
|
| 121 |
+
"loss": 0.3283,
|
| 122 |
+
"step": 8
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 8.0,
|
| 126 |
+
"eval_loss": 0.3282231390476227,
|
| 127 |
+
"eval_runtime": 0.0142,
|
| 128 |
+
"eval_samples_per_second": 705.138,
|
| 129 |
+
"eval_steps_per_second": 141.028,
|
| 130 |
+
"step": 8
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 9.0,
|
| 134 |
+
"grad_norm": 0.8372084498405457,
|
| 135 |
+
"learning_rate": 9.200000000000002e-06,
|
| 136 |
+
"loss": 0.3282,
|
| 137 |
+
"step": 9
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 9.0,
|
| 141 |
+
"eval_loss": 0.32816585898399353,
|
| 142 |
+
"eval_runtime": 0.0144,
|
| 143 |
+
"eval_samples_per_second": 695.135,
|
| 144 |
+
"eval_steps_per_second": 139.027,
|
| 145 |
+
"step": 9
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 10.0,
|
| 149 |
+
"grad_norm": 0.837040901184082,
|
| 150 |
+
"learning_rate": 9.100000000000001e-06,
|
| 151 |
+
"loss": 0.3282,
|
| 152 |
+
"step": 10
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 10.0,
|
| 156 |
+
"eval_loss": 0.328109472990036,
|
| 157 |
+
"eval_runtime": 0.0164,
|
| 158 |
+
"eval_samples_per_second": 609.921,
|
| 159 |
+
"eval_steps_per_second": 121.984,
|
| 160 |
+
"step": 10
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 11.0,
|
| 164 |
+
"grad_norm": 0.8362537622451782,
|
| 165 |
+
"learning_rate": 9e-06,
|
| 166 |
+
"loss": 0.3281,
|
| 167 |
+
"step": 11
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 11.0,
|
| 171 |
+
"eval_loss": 0.3280538320541382,
|
| 172 |
+
"eval_runtime": 0.0139,
|
| 173 |
+
"eval_samples_per_second": 720.72,
|
| 174 |
+
"eval_steps_per_second": 144.144,
|
| 175 |
+
"step": 11
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 12.0,
|
| 179 |
+
"grad_norm": 0.8360907435417175,
|
| 180 |
+
"learning_rate": 8.900000000000001e-06,
|
| 181 |
+
"loss": 0.3281,
|
| 182 |
+
"step": 12
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 12.0,
|
| 186 |
+
"eval_loss": 0.3279987871646881,
|
| 187 |
+
"eval_runtime": 0.0164,
|
| 188 |
+
"eval_samples_per_second": 611.219,
|
| 189 |
+
"eval_steps_per_second": 122.244,
|
| 190 |
+
"step": 12
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 13.0,
|
| 194 |
+
"grad_norm": 0.8359295129776001,
|
| 195 |
+
"learning_rate": 8.8e-06,
|
| 196 |
+
"loss": 0.328,
|
| 197 |
+
"step": 13
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 13.0,
|
| 201 |
+
"eval_loss": 0.32794439792633057,
|
| 202 |
+
"eval_runtime": 0.0145,
|
| 203 |
+
"eval_samples_per_second": 689.977,
|
| 204 |
+
"eval_steps_per_second": 137.995,
|
| 205 |
+
"step": 13
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 14.0,
|
| 209 |
+
"grad_norm": 0.8357701301574707,
|
| 210 |
+
"learning_rate": 8.700000000000001e-06,
|
| 211 |
+
"loss": 0.3279,
|
| 212 |
+
"step": 14
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 14.0,
|
| 216 |
+
"eval_loss": 0.3278906047344208,
|
| 217 |
+
"eval_runtime": 0.0146,
|
| 218 |
+
"eval_samples_per_second": 686.398,
|
| 219 |
+
"eval_steps_per_second": 137.28,
|
| 220 |
+
"step": 14
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 15.0,
|
| 224 |
+
"grad_norm": 0.8356127142906189,
|
| 225 |
+
"learning_rate": 8.6e-06,
|
| 226 |
+
"loss": 0.3279,
|
| 227 |
+
"step": 15
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 15.0,
|
| 231 |
+
"eval_loss": 0.3278374969959259,
|
| 232 |
+
"eval_runtime": 0.0164,
|
| 233 |
+
"eval_samples_per_second": 610.347,
|
| 234 |
+
"eval_steps_per_second": 122.069,
|
| 235 |
+
"step": 15
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 16.0,
|
| 239 |
+
"grad_norm": 0.8354570269584656,
|
| 240 |
+
"learning_rate": 8.5e-06,
|
| 241 |
+
"loss": 0.3278,
|
| 242 |
+
"step": 16
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 16.0,
|
| 246 |
+
"eval_loss": 0.3277849853038788,
|
| 247 |
+
"eval_runtime": 0.0141,
|
| 248 |
+
"eval_samples_per_second": 708.402,
|
| 249 |
+
"eval_steps_per_second": 141.68,
|
| 250 |
+
"step": 16
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 17.0,
|
| 254 |
+
"grad_norm": 0.8353032469749451,
|
| 255 |
+
"learning_rate": 8.400000000000001e-06,
|
| 256 |
+
"loss": 0.3278,
|
| 257 |
+
"step": 17
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 17.0,
|
| 261 |
+
"eval_loss": 0.3277330994606018,
|
| 262 |
+
"eval_runtime": 0.0172,
|
| 263 |
+
"eval_samples_per_second": 581.516,
|
| 264 |
+
"eval_steps_per_second": 116.303,
|
| 265 |
+
"step": 17
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 18.0,
|
| 269 |
+
"grad_norm": 0.8351512551307678,
|
| 270 |
+
"learning_rate": 8.3e-06,
|
| 271 |
+
"loss": 0.3277,
|
| 272 |
+
"step": 18
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 18.0,
|
| 276 |
+
"eval_loss": 0.32768189907073975,
|
| 277 |
+
"eval_runtime": 0.0189,
|
| 278 |
+
"eval_samples_per_second": 529.918,
|
| 279 |
+
"eval_steps_per_second": 105.984,
|
| 280 |
+
"step": 18
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 19.0,
|
| 284 |
+
"grad_norm": 0.8350011706352234,
|
| 285 |
+
"learning_rate": 8.2e-06,
|
| 286 |
+
"loss": 0.3277,
|
| 287 |
+
"step": 19
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 19.0,
|
| 291 |
+
"eval_loss": 0.3276313245296478,
|
| 292 |
+
"eval_runtime": 0.0147,
|
| 293 |
+
"eval_samples_per_second": 679.691,
|
| 294 |
+
"eval_steps_per_second": 135.938,
|
| 295 |
+
"step": 19
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 20.0,
|
| 299 |
+
"grad_norm": 0.8348528742790222,
|
| 300 |
+
"learning_rate": 8.1e-06,
|
| 301 |
+
"loss": 0.3276,
|
| 302 |
+
"step": 20
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 20.0,
|
| 306 |
+
"eval_loss": 0.32758134603500366,
|
| 307 |
+
"eval_runtime": 0.0141,
|
| 308 |
+
"eval_samples_per_second": 707.672,
|
| 309 |
+
"eval_steps_per_second": 141.534,
|
| 310 |
+
"step": 20
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 21.0,
|
| 314 |
+
"grad_norm": 0.8347064852714539,
|
| 315 |
+
"learning_rate": 8.000000000000001e-06,
|
| 316 |
+
"loss": 0.3276,
|
| 317 |
+
"step": 21
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 21.0,
|
| 321 |
+
"eval_loss": 0.32753199338912964,
|
| 322 |
+
"eval_runtime": 0.0192,
|
| 323 |
+
"eval_samples_per_second": 522.167,
|
| 324 |
+
"eval_steps_per_second": 104.433,
|
| 325 |
+
"step": 21
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 22.0,
|
| 329 |
+
"grad_norm": 0.834561824798584,
|
| 330 |
+
"learning_rate": 7.9e-06,
|
| 331 |
+
"loss": 0.3275,
|
| 332 |
+
"step": 22
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 22.0,
|
| 336 |
+
"eval_loss": 0.3274833559989929,
|
| 337 |
+
"eval_runtime": 0.0135,
|
| 338 |
+
"eval_samples_per_second": 738.109,
|
| 339 |
+
"eval_steps_per_second": 147.622,
|
| 340 |
+
"step": 22
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 23.0,
|
| 344 |
+
"grad_norm": 0.8344190716743469,
|
| 345 |
+
"learning_rate": 7.800000000000002e-06,
|
| 346 |
+
"loss": 0.3275,
|
| 347 |
+
"step": 23
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 23.0,
|
| 351 |
+
"eval_loss": 0.3274352252483368,
|
| 352 |
+
"eval_runtime": 0.0144,
|
| 353 |
+
"eval_samples_per_second": 694.812,
|
| 354 |
+
"eval_steps_per_second": 138.962,
|
| 355 |
+
"step": 23
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 24.0,
|
| 359 |
+
"grad_norm": 0.8342781662940979,
|
| 360 |
+
"learning_rate": 7.7e-06,
|
| 361 |
+
"loss": 0.3274,
|
| 362 |
+
"step": 24
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 24.0,
|
| 366 |
+
"eval_loss": 0.32738780975341797,
|
| 367 |
+
"eval_runtime": 0.0141,
|
| 368 |
+
"eval_samples_per_second": 710.779,
|
| 369 |
+
"eval_steps_per_second": 142.156,
|
| 370 |
+
"step": 24
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 25.0,
|
| 374 |
+
"grad_norm": 0.8341390490531921,
|
| 375 |
+
"learning_rate": 7.600000000000001e-06,
|
| 376 |
+
"loss": 0.3274,
|
| 377 |
+
"step": 25
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 25.0,
|
| 381 |
+
"eval_loss": 0.3273409605026245,
|
| 382 |
+
"eval_runtime": 0.0141,
|
| 383 |
+
"eval_samples_per_second": 709.072,
|
| 384 |
+
"eval_steps_per_second": 141.814,
|
| 385 |
+
"step": 25
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 26.0,
|
| 389 |
+
"grad_norm": 0.8340017795562744,
|
| 390 |
+
"learning_rate": 7.500000000000001e-06,
|
| 391 |
+
"loss": 0.3273,
|
| 392 |
+
"step": 26
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 26.0,
|
| 396 |
+
"eval_loss": 0.32729482650756836,
|
| 397 |
+
"eval_runtime": 0.0158,
|
| 398 |
+
"eval_samples_per_second": 634.194,
|
| 399 |
+
"eval_steps_per_second": 126.839,
|
| 400 |
+
"step": 26
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 27.0,
|
| 404 |
+
"grad_norm": 0.8338663578033447,
|
| 405 |
+
"learning_rate": 7.4e-06,
|
| 406 |
+
"loss": 0.3273,
|
| 407 |
+
"step": 27
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 27.0,
|
| 411 |
+
"eval_loss": 0.3272492289543152,
|
| 412 |
+
"eval_runtime": 0.0137,
|
| 413 |
+
"eval_samples_per_second": 729.495,
|
| 414 |
+
"eval_steps_per_second": 145.899,
|
| 415 |
+
"step": 27
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 28.0,
|
| 419 |
+
"grad_norm": 0.8337326645851135,
|
| 420 |
+
"learning_rate": 7.3e-06,
|
| 421 |
+
"loss": 0.3272,
|
| 422 |
+
"step": 28
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 28.0,
|
| 426 |
+
"eval_loss": 0.32720428705215454,
|
| 427 |
+
"eval_runtime": 0.0156,
|
| 428 |
+
"eval_samples_per_second": 639.834,
|
| 429 |
+
"eval_steps_per_second": 127.967,
|
| 430 |
+
"step": 28
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 29.0,
|
| 434 |
+
"grad_norm": 0.8336009383201599,
|
| 435 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 436 |
+
"loss": 0.3272,
|
| 437 |
+
"step": 29
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 29.0,
|
| 441 |
+
"eval_loss": 0.32715997099876404,
|
| 442 |
+
"eval_runtime": 0.0135,
|
| 443 |
+
"eval_samples_per_second": 741.24,
|
| 444 |
+
"eval_steps_per_second": 148.248,
|
| 445 |
+
"step": 29
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 30.0,
|
| 449 |
+
"grad_norm": 0.8334709405899048,
|
| 450 |
+
"learning_rate": 7.100000000000001e-06,
|
| 451 |
+
"loss": 0.3272,
|
| 452 |
+
"step": 30
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 30.0,
|
| 456 |
+
"eval_loss": 0.3271161913871765,
|
| 457 |
+
"eval_runtime": 0.0157,
|
| 458 |
+
"eval_samples_per_second": 638.908,
|
| 459 |
+
"eval_steps_per_second": 127.782,
|
| 460 |
+
"step": 30
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 31.0,
|
| 464 |
+
"grad_norm": 0.834907054901123,
|
| 465 |
+
"learning_rate": 7e-06,
|
| 466 |
+
"loss": 0.3271,
|
| 467 |
+
"step": 31
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 31.0,
|
| 471 |
+
"eval_loss": 0.32707276940345764,
|
| 472 |
+
"eval_runtime": 0.0147,
|
| 473 |
+
"eval_samples_per_second": 678.679,
|
| 474 |
+
"eval_steps_per_second": 135.736,
|
| 475 |
+
"step": 31
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 32.0,
|
| 479 |
+
"grad_norm": 0.8347804546356201,
|
| 480 |
+
"learning_rate": 6.9e-06,
|
| 481 |
+
"loss": 0.3271,
|
| 482 |
+
"step": 32
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 32.0,
|
| 486 |
+
"eval_loss": 0.32702988386154175,
|
| 487 |
+
"eval_runtime": 0.0137,
|
| 488 |
+
"eval_samples_per_second": 731.48,
|
| 489 |
+
"eval_steps_per_second": 146.296,
|
| 490 |
+
"step": 32
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 33.0,
|
| 494 |
+
"grad_norm": 0.8346555829048157,
|
| 495 |
+
"learning_rate": 6.800000000000001e-06,
|
| 496 |
+
"loss": 0.327,
|
| 497 |
+
"step": 33
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 33.0,
|
| 501 |
+
"eval_loss": 0.326987624168396,
|
| 502 |
+
"eval_runtime": 0.0154,
|
| 503 |
+
"eval_samples_per_second": 647.809,
|
| 504 |
+
"eval_steps_per_second": 129.562,
|
| 505 |
+
"step": 33
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 34.0,
|
| 509 |
+
"grad_norm": 0.8345323801040649,
|
| 510 |
+
"learning_rate": 6.700000000000001e-06,
|
| 511 |
+
"loss": 0.327,
|
| 512 |
+
"step": 34
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 34.0,
|
| 516 |
+
"eval_loss": 0.326945960521698,
|
| 517 |
+
"eval_runtime": 0.015,
|
| 518 |
+
"eval_samples_per_second": 668.532,
|
| 519 |
+
"eval_steps_per_second": 133.706,
|
| 520 |
+
"step": 34
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 35.0,
|
| 524 |
+
"grad_norm": 0.8344109654426575,
|
| 525 |
+
"learning_rate": 6.600000000000001e-06,
|
| 526 |
+
"loss": 0.3269,
|
| 527 |
+
"step": 35
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 35.0,
|
| 531 |
+
"eval_loss": 0.32690495252609253,
|
| 532 |
+
"eval_runtime": 0.0141,
|
| 533 |
+
"eval_samples_per_second": 711.152,
|
| 534 |
+
"eval_steps_per_second": 142.23,
|
| 535 |
+
"step": 35
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 36.0,
|
| 539 |
+
"grad_norm": 0.8342913389205933,
|
| 540 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 541 |
+
"loss": 0.3269,
|
| 542 |
+
"step": 36
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 36.0,
|
| 546 |
+
"eval_loss": 0.3268645405769348,
|
| 547 |
+
"eval_runtime": 0.0148,
|
| 548 |
+
"eval_samples_per_second": 676.653,
|
| 549 |
+
"eval_steps_per_second": 135.331,
|
| 550 |
+
"step": 36
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 37.0,
|
| 554 |
+
"grad_norm": 0.8341735005378723,
|
| 555 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 556 |
+
"loss": 0.3269,
|
| 557 |
+
"step": 37
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 37.0,
|
| 561 |
+
"eval_loss": 0.32682472467422485,
|
| 562 |
+
"eval_runtime": 0.0184,
|
| 563 |
+
"eval_samples_per_second": 544.856,
|
| 564 |
+
"eval_steps_per_second": 108.971,
|
| 565 |
+
"step": 37
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 38.0,
|
| 569 |
+
"grad_norm": 0.8340575098991394,
|
| 570 |
+
"learning_rate": 6.300000000000001e-06,
|
| 571 |
+
"loss": 0.3268,
|
| 572 |
+
"step": 38
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 38.0,
|
| 576 |
+
"eval_loss": 0.32678553462028503,
|
| 577 |
+
"eval_runtime": 0.0135,
|
| 578 |
+
"eval_samples_per_second": 738.577,
|
| 579 |
+
"eval_steps_per_second": 147.715,
|
| 580 |
+
"step": 38
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 39.0,
|
| 584 |
+
"grad_norm": 0.8339433073997498,
|
| 585 |
+
"learning_rate": 6.200000000000001e-06,
|
| 586 |
+
"loss": 0.3268,
|
| 587 |
+
"step": 39
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 39.0,
|
| 591 |
+
"eval_loss": 0.32674694061279297,
|
| 592 |
+
"eval_runtime": 0.0139,
|
| 593 |
+
"eval_samples_per_second": 720.337,
|
| 594 |
+
"eval_steps_per_second": 144.067,
|
| 595 |
+
"step": 39
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 40.0,
|
| 599 |
+
"grad_norm": 0.8338308334350586,
|
| 600 |
+
"learning_rate": 6.1e-06,
|
| 601 |
+
"loss": 0.3267,
|
| 602 |
+
"step": 40
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 40.0,
|
| 606 |
+
"eval_loss": 0.3267090320587158,
|
| 607 |
+
"eval_runtime": 0.0146,
|
| 608 |
+
"eval_samples_per_second": 685.064,
|
| 609 |
+
"eval_steps_per_second": 137.013,
|
| 610 |
+
"step": 40
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 41.0,
|
| 614 |
+
"grad_norm": 0.8337202668190002,
|
| 615 |
+
"learning_rate": 6e-06,
|
| 616 |
+
"loss": 0.3267,
|
| 617 |
+
"step": 41
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 41.0,
|
| 621 |
+
"eval_loss": 0.32667168974876404,
|
| 622 |
+
"eval_runtime": 0.0137,
|
| 623 |
+
"eval_samples_per_second": 729.584,
|
| 624 |
+
"eval_steps_per_second": 145.917,
|
| 625 |
+
"step": 41
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 42.0,
|
| 629 |
+
"grad_norm": 0.8336114883422852,
|
| 630 |
+
"learning_rate": 5.9e-06,
|
| 631 |
+
"loss": 0.3267,
|
| 632 |
+
"step": 42
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 42.0,
|
| 636 |
+
"eval_loss": 0.32663506269454956,
|
| 637 |
+
"eval_runtime": 0.0139,
|
| 638 |
+
"eval_samples_per_second": 721.154,
|
| 639 |
+
"eval_steps_per_second": 144.231,
|
| 640 |
+
"step": 42
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 43.0,
|
| 644 |
+
"grad_norm": 0.8335044384002686,
|
| 645 |
+
"learning_rate": 5.8e-06,
|
| 646 |
+
"loss": 0.3266,
|
| 647 |
+
"step": 43
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 43.0,
|
| 651 |
+
"eval_loss": 0.32659897208213806,
|
| 652 |
+
"eval_runtime": 0.0163,
|
| 653 |
+
"eval_samples_per_second": 613.75,
|
| 654 |
+
"eval_steps_per_second": 122.75,
|
| 655 |
+
"step": 43
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 44.0,
|
| 659 |
+
"grad_norm": 0.8333994150161743,
|
| 660 |
+
"learning_rate": 5.7e-06,
|
| 661 |
+
"loss": 0.3266,
|
| 662 |
+
"step": 44
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 44.0,
|
| 666 |
+
"eval_loss": 0.3265635371208191,
|
| 667 |
+
"eval_runtime": 0.0137,
|
| 668 |
+
"eval_samples_per_second": 730.842,
|
| 669 |
+
"eval_steps_per_second": 146.168,
|
| 670 |
+
"step": 44
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 45.0,
|
| 674 |
+
"grad_norm": 0.8332960605621338,
|
| 675 |
+
"learning_rate": 5.600000000000001e-06,
|
| 676 |
+
"loss": 0.3266,
|
| 677 |
+
"step": 45
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 45.0,
|
| 681 |
+
"eval_loss": 0.32652872800827026,
|
| 682 |
+
"eval_runtime": 0.015,
|
| 683 |
+
"eval_samples_per_second": 667.819,
|
| 684 |
+
"eval_steps_per_second": 133.564,
|
| 685 |
+
"step": 45
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 46.0,
|
| 689 |
+
"grad_norm": 0.8331945538520813,
|
| 690 |
+
"learning_rate": 5.500000000000001e-06,
|
| 691 |
+
"loss": 0.3265,
|
| 692 |
+
"step": 46
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 46.0,
|
| 696 |
+
"eval_loss": 0.3264945149421692,
|
| 697 |
+
"eval_runtime": 0.0148,
|
| 698 |
+
"eval_samples_per_second": 674.141,
|
| 699 |
+
"eval_steps_per_second": 134.828,
|
| 700 |
+
"step": 46
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 47.0,
|
| 704 |
+
"grad_norm": 0.8330948948860168,
|
| 705 |
+
"learning_rate": 5.400000000000001e-06,
|
| 706 |
+
"loss": 0.3265,
|
| 707 |
+
"step": 47
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 47.0,
|
| 711 |
+
"eval_loss": 0.32646098732948303,
|
| 712 |
+
"eval_runtime": 0.0147,
|
| 713 |
+
"eval_samples_per_second": 678.086,
|
| 714 |
+
"eval_steps_per_second": 135.617,
|
| 715 |
+
"step": 47
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 48.0,
|
| 719 |
+
"grad_norm": 0.8329970240592957,
|
| 720 |
+
"learning_rate": 5.300000000000001e-06,
|
| 721 |
+
"loss": 0.3265,
|
| 722 |
+
"step": 48
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 48.0,
|
| 726 |
+
"eval_loss": 0.32642805576324463,
|
| 727 |
+
"eval_runtime": 0.0151,
|
| 728 |
+
"eval_samples_per_second": 663.53,
|
| 729 |
+
"eval_steps_per_second": 132.706,
|
| 730 |
+
"step": 48
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 49.0,
|
| 734 |
+
"grad_norm": 0.8329010009765625,
|
| 735 |
+
"learning_rate": 5.2e-06,
|
| 736 |
+
"loss": 0.3264,
|
| 737 |
+
"step": 49
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 49.0,
|
| 741 |
+
"eval_loss": 0.32639575004577637,
|
| 742 |
+
"eval_runtime": 0.0141,
|
| 743 |
+
"eval_samples_per_second": 707.1,
|
| 744 |
+
"eval_steps_per_second": 141.42,
|
| 745 |
+
"step": 49
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 50.0,
|
| 749 |
+
"grad_norm": 0.8328068852424622,
|
| 750 |
+
"learning_rate": 5.1e-06,
|
| 751 |
+
"loss": 0.3264,
|
| 752 |
+
"step": 50
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 50.0,
|
| 756 |
+
"eval_loss": 0.32636409997940063,
|
| 757 |
+
"eval_runtime": 0.0154,
|
| 758 |
+
"eval_samples_per_second": 650.905,
|
| 759 |
+
"eval_steps_per_second": 130.181,
|
| 760 |
+
"step": 50
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 51.0,
|
| 764 |
+
"grad_norm": 0.8327144384384155,
|
| 765 |
+
"learning_rate": 5e-06,
|
| 766 |
+
"loss": 0.3264,
|
| 767 |
+
"step": 51
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 51.0,
|
| 771 |
+
"eval_loss": 0.3263329863548279,
|
| 772 |
+
"eval_runtime": 0.0156,
|
| 773 |
+
"eval_samples_per_second": 639.142,
|
| 774 |
+
"eval_steps_per_second": 127.828,
|
| 775 |
+
"step": 51
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 52.0,
|
| 779 |
+
"grad_norm": 0.8326238989830017,
|
| 780 |
+
"learning_rate": 4.9000000000000005e-06,
|
| 781 |
+
"loss": 0.3263,
|
| 782 |
+
"step": 52
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 52.0,
|
| 786 |
+
"eval_loss": 0.32630258798599243,
|
| 787 |
+
"eval_runtime": 0.0155,
|
| 788 |
+
"eval_samples_per_second": 645.069,
|
| 789 |
+
"eval_steps_per_second": 129.014,
|
| 790 |
+
"step": 52
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 53.0,
|
| 794 |
+
"grad_norm": 0.8325351476669312,
|
| 795 |
+
"learning_rate": 4.800000000000001e-06,
|
| 796 |
+
"loss": 0.3263,
|
| 797 |
+
"step": 53
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 53.0,
|
| 801 |
+
"eval_loss": 0.32627278566360474,
|
| 802 |
+
"eval_runtime": 0.0138,
|
| 803 |
+
"eval_samples_per_second": 724.205,
|
| 804 |
+
"eval_steps_per_second": 144.841,
|
| 805 |
+
"step": 53
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 54.0,
|
| 809 |
+
"grad_norm": 0.8324483036994934,
|
| 810 |
+
"learning_rate": 4.7e-06,
|
| 811 |
+
"loss": 0.3263,
|
| 812 |
+
"step": 54
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 54.0,
|
| 816 |
+
"eval_loss": 0.32624363899230957,
|
| 817 |
+
"eval_runtime": 0.0156,
|
| 818 |
+
"eval_samples_per_second": 642.933,
|
| 819 |
+
"eval_steps_per_second": 128.587,
|
| 820 |
+
"step": 54
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 55.0,
|
| 824 |
+
"grad_norm": 0.8323632478713989,
|
| 825 |
+
"learning_rate": 4.600000000000001e-06,
|
| 826 |
+
"loss": 0.3262,
|
| 827 |
+
"step": 55
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 55.0,
|
| 831 |
+
"eval_loss": 0.32621514797210693,
|
| 832 |
+
"eval_runtime": 0.0137,
|
| 833 |
+
"eval_samples_per_second": 728.203,
|
| 834 |
+
"eval_steps_per_second": 145.641,
|
| 835 |
+
"step": 55
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 56.0,
|
| 839 |
+
"grad_norm": 0.8322799205780029,
|
| 840 |
+
"learning_rate": 4.5e-06,
|
| 841 |
+
"loss": 0.3262,
|
| 842 |
+
"step": 56
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 56.0,
|
| 846 |
+
"eval_loss": 0.3261871933937073,
|
| 847 |
+
"eval_runtime": 0.0148,
|
| 848 |
+
"eval_samples_per_second": 674.466,
|
| 849 |
+
"eval_steps_per_second": 134.893,
|
| 850 |
+
"step": 56
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 57.0,
|
| 854 |
+
"grad_norm": 0.8321985602378845,
|
| 855 |
+
"learning_rate": 4.4e-06,
|
| 856 |
+
"loss": 0.3262,
|
| 857 |
+
"step": 57
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 57.0,
|
| 861 |
+
"eval_loss": 0.32615989446640015,
|
| 862 |
+
"eval_runtime": 0.0139,
|
| 863 |
+
"eval_samples_per_second": 718.24,
|
| 864 |
+
"eval_steps_per_second": 143.648,
|
| 865 |
+
"step": 57
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 58.0,
|
| 869 |
+
"grad_norm": 0.8321189880371094,
|
| 870 |
+
"learning_rate": 4.3e-06,
|
| 871 |
+
"loss": 0.3262,
|
| 872 |
+
"step": 58
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 58.0,
|
| 876 |
+
"eval_loss": 0.32613325119018555,
|
| 877 |
+
"eval_runtime": 0.015,
|
| 878 |
+
"eval_samples_per_second": 664.865,
|
| 879 |
+
"eval_steps_per_second": 132.973,
|
| 880 |
+
"step": 58
|
| 881 |
}
|
| 882 |
],
|
| 883 |
"logging_steps": 1,
|