Training in progress, epoch 0
Browse files
all_results.json
CHANGED
|
@@ -1,7 +1,13 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"eval_accuracy": 0.11342436602234439,
|
| 3 |
"eval_loss": 2.3934237957000732,
|
| 4 |
"eval_runtime": 514.1383,
|
| 5 |
"eval_samples_per_second": 54.839,
|
| 6 |
-
"eval_steps_per_second": 0.858
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 9.993390614672835,
|
| 3 |
"eval_accuracy": 0.11342436602234439,
|
| 4 |
"eval_loss": 2.3934237957000732,
|
| 5 |
"eval_runtime": 514.1383,
|
| 6 |
"eval_samples_per_second": 54.839,
|
| 7 |
+
"eval_steps_per_second": 0.858,
|
| 8 |
+
"total_flos": 7.511248143138256e+19,
|
| 9 |
+
"train_loss": 1.6334815527396227,
|
| 10 |
+
"train_runtime": 44633.0248,
|
| 11 |
+
"train_samples_per_second": 21.683,
|
| 12 |
+
"train_steps_per_second": 0.085
|
| 13 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 343248584
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97016281036b19c1e77a456c81f3be4700ff5dede23cca6cf00773d5a3e3c5b4
|
| 3 |
size 343248584
|
runs/May16_00-21-59_e2e-66-39/events.out.tfevents.1715799126.e2e-66-39.276148.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9b05809ba1e54448989a1b06fb1af5af0a32b1ee87e3a06ebf31a14cc7eb3575
|
| 3 |
+
size 11804
|
runs/May16_10-48-15_e2e-66-39/events.out.tfevents.1715836698.e2e-66-39.330728.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:83fe818f0397f5e7a86f8a91dee8b8baa1ab77fce182fe738a447375694da3eb
|
| 3 |
+
size 13072
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 9.993390614672835,
|
| 3 |
+
"total_flos": 7.511248143138256e+19,
|
| 4 |
+
"train_loss": 1.6334815527396227,
|
| 5 |
+
"train_runtime": 44633.0248,
|
| 6 |
+
"train_samples_per_second": 21.683,
|
| 7 |
+
"train_steps_per_second": 0.085
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2766 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.32094342968611456,
|
| 3 |
+
"best_model_checkpoint": "Mrohit01/cards-vit-base-patch16-224-finetuned-v1/checkpoint-3404",
|
| 4 |
+
"epoch": 9.993390614672835,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 3780,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.026437541308658295,
|
| 13 |
+
"grad_norm": 1.8425437211990356,
|
| 14 |
+
"learning_rate": 1.3227513227513228e-06,
|
| 15 |
+
"loss": 1.9993,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05287508261731659,
|
| 20 |
+
"grad_norm": 1.9051584005355835,
|
| 21 |
+
"learning_rate": 2.6455026455026455e-06,
|
| 22 |
+
"loss": 2.0327,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.07931262392597488,
|
| 27 |
+
"grad_norm": 1.936784267425537,
|
| 28 |
+
"learning_rate": 3.968253968253968e-06,
|
| 29 |
+
"loss": 2.0055,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.10575016523463318,
|
| 34 |
+
"grad_norm": 1.9875764846801758,
|
| 35 |
+
"learning_rate": 5.291005291005291e-06,
|
| 36 |
+
"loss": 1.9807,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.13218770654329148,
|
| 41 |
+
"grad_norm": 2.0385992527008057,
|
| 42 |
+
"learning_rate": 6.613756613756614e-06,
|
| 43 |
+
"loss": 1.9866,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.15862524785194976,
|
| 48 |
+
"grad_norm": 1.9623510837554932,
|
| 49 |
+
"learning_rate": 7.936507936507936e-06,
|
| 50 |
+
"loss": 1.9612,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.18506278916060806,
|
| 55 |
+
"grad_norm": 1.8609628677368164,
|
| 56 |
+
"learning_rate": 9.259259259259259e-06,
|
| 57 |
+
"loss": 1.9499,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.21150033046926636,
|
| 62 |
+
"grad_norm": 2.279550552368164,
|
| 63 |
+
"learning_rate": 1.0582010582010582e-05,
|
| 64 |
+
"loss": 1.9277,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.23793787177792466,
|
| 69 |
+
"grad_norm": 2.0092594623565674,
|
| 70 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 71 |
+
"loss": 1.9246,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.26437541308658297,
|
| 76 |
+
"grad_norm": 1.8894586563110352,
|
| 77 |
+
"learning_rate": 1.3227513227513228e-05,
|
| 78 |
+
"loss": 1.9187,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.29081295439524124,
|
| 83 |
+
"grad_norm": 1.841611623764038,
|
| 84 |
+
"learning_rate": 1.455026455026455e-05,
|
| 85 |
+
"loss": 1.9279,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3172504957038995,
|
| 90 |
+
"grad_norm": 1.9487497806549072,
|
| 91 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 92 |
+
"loss": 1.8989,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.34368803701255785,
|
| 97 |
+
"grad_norm": 1.8522062301635742,
|
| 98 |
+
"learning_rate": 1.7195767195767195e-05,
|
| 99 |
+
"loss": 1.9041,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.3701255783212161,
|
| 104 |
+
"grad_norm": 1.6527143716812134,
|
| 105 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 106 |
+
"loss": 1.9107,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3965631196298744,
|
| 111 |
+
"grad_norm": 1.6777476072311401,
|
| 112 |
+
"learning_rate": 1.984126984126984e-05,
|
| 113 |
+
"loss": 1.8915,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.4230006609385327,
|
| 118 |
+
"grad_norm": 2.1537675857543945,
|
| 119 |
+
"learning_rate": 2.1164021164021164e-05,
|
| 120 |
+
"loss": 1.8814,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.449438202247191,
|
| 125 |
+
"grad_norm": 2.119074583053589,
|
| 126 |
+
"learning_rate": 2.2486772486772487e-05,
|
| 127 |
+
"loss": 1.8767,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.47587574355584933,
|
| 132 |
+
"grad_norm": 2.027825355529785,
|
| 133 |
+
"learning_rate": 2.380952380952381e-05,
|
| 134 |
+
"loss": 1.8542,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.5023132848645075,
|
| 139 |
+
"grad_norm": 1.9427117109298706,
|
| 140 |
+
"learning_rate": 2.5132275132275137e-05,
|
| 141 |
+
"loss": 1.8614,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.5287508261731659,
|
| 146 |
+
"grad_norm": 2.057250499725342,
|
| 147 |
+
"learning_rate": 2.6455026455026456e-05,
|
| 148 |
+
"loss": 1.8578,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.5551883674818242,
|
| 153 |
+
"grad_norm": 1.760541558265686,
|
| 154 |
+
"learning_rate": 2.777777777777778e-05,
|
| 155 |
+
"loss": 1.8507,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.5816259087904825,
|
| 160 |
+
"grad_norm": 2.5953147411346436,
|
| 161 |
+
"learning_rate": 2.91005291005291e-05,
|
| 162 |
+
"loss": 1.8325,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.6080634500991408,
|
| 167 |
+
"grad_norm": 2.2065205574035645,
|
| 168 |
+
"learning_rate": 3.0423280423280425e-05,
|
| 169 |
+
"loss": 1.8418,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.634500991407799,
|
| 174 |
+
"grad_norm": 1.930132269859314,
|
| 175 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 176 |
+
"loss": 1.8295,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.6609385327164574,
|
| 181 |
+
"grad_norm": 2.2985873222351074,
|
| 182 |
+
"learning_rate": 3.306878306878307e-05,
|
| 183 |
+
"loss": 1.8367,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.6873760740251157,
|
| 188 |
+
"grad_norm": 2.229893207550049,
|
| 189 |
+
"learning_rate": 3.439153439153439e-05,
|
| 190 |
+
"loss": 1.8298,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.713813615333774,
|
| 195 |
+
"grad_norm": 2.627453088760376,
|
| 196 |
+
"learning_rate": 3.571428571428572e-05,
|
| 197 |
+
"loss": 1.8321,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.7402511566424322,
|
| 202 |
+
"grad_norm": 2.1309328079223633,
|
| 203 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 204 |
+
"loss": 1.8377,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.7666886979510905,
|
| 209 |
+
"grad_norm": 1.6566647291183472,
|
| 210 |
+
"learning_rate": 3.835978835978836e-05,
|
| 211 |
+
"loss": 1.8391,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.7931262392597488,
|
| 216 |
+
"grad_norm": 3.22412109375,
|
| 217 |
+
"learning_rate": 3.968253968253968e-05,
|
| 218 |
+
"loss": 1.825,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.8195637805684072,
|
| 223 |
+
"grad_norm": 2.1069350242614746,
|
| 224 |
+
"learning_rate": 4.100529100529101e-05,
|
| 225 |
+
"loss": 1.8074,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.8460013218770654,
|
| 230 |
+
"grad_norm": 2.3734729290008545,
|
| 231 |
+
"learning_rate": 4.232804232804233e-05,
|
| 232 |
+
"loss": 1.8281,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.8724388631857237,
|
| 237 |
+
"grad_norm": 1.7937417030334473,
|
| 238 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 239 |
+
"loss": 1.8065,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.898876404494382,
|
| 244 |
+
"grad_norm": 2.0130672454833984,
|
| 245 |
+
"learning_rate": 4.4973544973544974e-05,
|
| 246 |
+
"loss": 1.8133,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.9253139458030403,
|
| 251 |
+
"grad_norm": 1.8942523002624512,
|
| 252 |
+
"learning_rate": 4.62962962962963e-05,
|
| 253 |
+
"loss": 1.8097,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.9517514871116987,
|
| 258 |
+
"grad_norm": 2.3066070079803467,
|
| 259 |
+
"learning_rate": 4.761904761904762e-05,
|
| 260 |
+
"loss": 1.8325,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.9781890284203569,
|
| 265 |
+
"grad_norm": 1.563308596611023,
|
| 266 |
+
"learning_rate": 4.894179894179895e-05,
|
| 267 |
+
"loss": 1.7866,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.9993390614672836,
|
| 272 |
+
"eval_accuracy": 0.2745522255719099,
|
| 273 |
+
"eval_loss": 1.7676516771316528,
|
| 274 |
+
"eval_runtime": 664.7291,
|
| 275 |
+
"eval_samples_per_second": 42.416,
|
| 276 |
+
"eval_steps_per_second": 0.663,
|
| 277 |
+
"step": 378
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.004626569729015,
|
| 281 |
+
"grad_norm": 1.7081835269927979,
|
| 282 |
+
"learning_rate": 4.9970605526161084e-05,
|
| 283 |
+
"loss": 1.7884,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.0310641110376735,
|
| 288 |
+
"grad_norm": 2.7630808353424072,
|
| 289 |
+
"learning_rate": 4.982363315696649e-05,
|
| 290 |
+
"loss": 1.7739,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.0575016523463319,
|
| 295 |
+
"grad_norm": 3.5567917823791504,
|
| 296 |
+
"learning_rate": 4.96766607877719e-05,
|
| 297 |
+
"loss": 1.7764,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.08393919365499,
|
| 302 |
+
"grad_norm": 2.181006669998169,
|
| 303 |
+
"learning_rate": 4.952968841857731e-05,
|
| 304 |
+
"loss": 1.8154,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.1103767349636484,
|
| 309 |
+
"grad_norm": 2.285545587539673,
|
| 310 |
+
"learning_rate": 4.938271604938271e-05,
|
| 311 |
+
"loss": 1.7998,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.1368142762723066,
|
| 316 |
+
"grad_norm": 2.4093916416168213,
|
| 317 |
+
"learning_rate": 4.923574368018813e-05,
|
| 318 |
+
"loss": 1.7643,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.163251817580965,
|
| 323 |
+
"grad_norm": 1.6129083633422852,
|
| 324 |
+
"learning_rate": 4.908877131099354e-05,
|
| 325 |
+
"loss": 1.7932,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.1896893588896233,
|
| 330 |
+
"grad_norm": 1.6316031217575073,
|
| 331 |
+
"learning_rate": 4.894179894179895e-05,
|
| 332 |
+
"loss": 1.7851,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.2161269001982815,
|
| 337 |
+
"grad_norm": 1.6007944345474243,
|
| 338 |
+
"learning_rate": 4.879482657260435e-05,
|
| 339 |
+
"loss": 1.795,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.24256444150694,
|
| 344 |
+
"grad_norm": 1.69334077835083,
|
| 345 |
+
"learning_rate": 4.864785420340976e-05,
|
| 346 |
+
"loss": 1.7523,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.269001982815598,
|
| 351 |
+
"grad_norm": 1.9214173555374146,
|
| 352 |
+
"learning_rate": 4.850088183421517e-05,
|
| 353 |
+
"loss": 1.7595,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.2954395241242564,
|
| 358 |
+
"grad_norm": 1.5973252058029175,
|
| 359 |
+
"learning_rate": 4.835390946502058e-05,
|
| 360 |
+
"loss": 1.7542,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.3218770654329148,
|
| 365 |
+
"grad_norm": 1.8921444416046143,
|
| 366 |
+
"learning_rate": 4.820693709582599e-05,
|
| 367 |
+
"loss": 1.7466,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.348314606741573,
|
| 372 |
+
"grad_norm": 1.84861159324646,
|
| 373 |
+
"learning_rate": 4.8059964726631394e-05,
|
| 374 |
+
"loss": 1.7763,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.3747521480502314,
|
| 379 |
+
"grad_norm": 2.476146936416626,
|
| 380 |
+
"learning_rate": 4.79129923574368e-05,
|
| 381 |
+
"loss": 1.7577,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.4011896893588895,
|
| 386 |
+
"grad_norm": 1.5669139623641968,
|
| 387 |
+
"learning_rate": 4.776601998824221e-05,
|
| 388 |
+
"loss": 1.7697,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.427627230667548,
|
| 393 |
+
"grad_norm": 2.1005804538726807,
|
| 394 |
+
"learning_rate": 4.761904761904762e-05,
|
| 395 |
+
"loss": 1.7737,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.4540647719762063,
|
| 400 |
+
"grad_norm": 1.8329302072525024,
|
| 401 |
+
"learning_rate": 4.747207524985303e-05,
|
| 402 |
+
"loss": 1.7476,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.4805023132848645,
|
| 407 |
+
"grad_norm": 1.7490884065628052,
|
| 408 |
+
"learning_rate": 4.732510288065844e-05,
|
| 409 |
+
"loss": 1.7695,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.5069398545935226,
|
| 414 |
+
"grad_norm": 1.7790248394012451,
|
| 415 |
+
"learning_rate": 4.717813051146385e-05,
|
| 416 |
+
"loss": 1.7696,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.533377395902181,
|
| 421 |
+
"grad_norm": 1.4908485412597656,
|
| 422 |
+
"learning_rate": 4.7031158142269256e-05,
|
| 423 |
+
"loss": 1.7809,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.5598149372108394,
|
| 428 |
+
"grad_norm": 1.4122120141983032,
|
| 429 |
+
"learning_rate": 4.6884185773074665e-05,
|
| 430 |
+
"loss": 1.7281,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.5862524785194978,
|
| 435 |
+
"grad_norm": 1.5507616996765137,
|
| 436 |
+
"learning_rate": 4.673721340388007e-05,
|
| 437 |
+
"loss": 1.7906,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.612690019828156,
|
| 442 |
+
"grad_norm": 1.7312538623809814,
|
| 443 |
+
"learning_rate": 4.659024103468548e-05,
|
| 444 |
+
"loss": 1.7863,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.6391275611368141,
|
| 449 |
+
"grad_norm": 1.9535393714904785,
|
| 450 |
+
"learning_rate": 4.644326866549089e-05,
|
| 451 |
+
"loss": 1.7963,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.6655651024454725,
|
| 456 |
+
"grad_norm": 2.0952374935150146,
|
| 457 |
+
"learning_rate": 4.62962962962963e-05,
|
| 458 |
+
"loss": 1.784,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.692002643754131,
|
| 463 |
+
"grad_norm": 1.8771756887435913,
|
| 464 |
+
"learning_rate": 4.614932392710171e-05,
|
| 465 |
+
"loss": 1.7717,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.7184401850627893,
|
| 470 |
+
"grad_norm": 1.9628280401229858,
|
| 471 |
+
"learning_rate": 4.600235155790711e-05,
|
| 472 |
+
"loss": 1.7156,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.7448777263714474,
|
| 477 |
+
"grad_norm": 1.7401697635650635,
|
| 478 |
+
"learning_rate": 4.585537918871252e-05,
|
| 479 |
+
"loss": 1.7748,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.7713152676801056,
|
| 484 |
+
"grad_norm": 1.9188202619552612,
|
| 485 |
+
"learning_rate": 4.5708406819517937e-05,
|
| 486 |
+
"loss": 1.7759,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.797752808988764,
|
| 491 |
+
"grad_norm": 1.7092291116714478,
|
| 492 |
+
"learning_rate": 4.5561434450323345e-05,
|
| 493 |
+
"loss": 1.7368,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.8241903502974224,
|
| 498 |
+
"grad_norm": 1.9781934022903442,
|
| 499 |
+
"learning_rate": 4.541446208112875e-05,
|
| 500 |
+
"loss": 1.7743,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.8506278916060808,
|
| 505 |
+
"grad_norm": 2.159959077835083,
|
| 506 |
+
"learning_rate": 4.5267489711934157e-05,
|
| 507 |
+
"loss": 1.7723,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.877065432914739,
|
| 512 |
+
"grad_norm": 1.9522385597229004,
|
| 513 |
+
"learning_rate": 4.5120517342739565e-05,
|
| 514 |
+
"loss": 1.7669,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.903502974223397,
|
| 519 |
+
"grad_norm": 2.2479653358459473,
|
| 520 |
+
"learning_rate": 4.4973544973544974e-05,
|
| 521 |
+
"loss": 1.7479,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.9299405155320555,
|
| 526 |
+
"grad_norm": 2.117433547973633,
|
| 527 |
+
"learning_rate": 4.482657260435038e-05,
|
| 528 |
+
"loss": 1.7363,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.9563780568407139,
|
| 533 |
+
"grad_norm": 1.7201601266860962,
|
| 534 |
+
"learning_rate": 4.467960023515579e-05,
|
| 535 |
+
"loss": 1.7551,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.9828155981493722,
|
| 540 |
+
"grad_norm": 1.7983427047729492,
|
| 541 |
+
"learning_rate": 4.45326278659612e-05,
|
| 542 |
+
"loss": 1.7457,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.998678122934567,
|
| 547 |
+
"eval_accuracy": 0.29902464976059584,
|
| 548 |
+
"eval_loss": 1.7163423299789429,
|
| 549 |
+
"eval_runtime": 618.2518,
|
| 550 |
+
"eval_samples_per_second": 45.604,
|
| 551 |
+
"eval_steps_per_second": 0.713,
|
| 552 |
+
"step": 756
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.00925313945803,
|
| 556 |
+
"grad_norm": 1.3950998783111572,
|
| 557 |
+
"learning_rate": 4.438565549676661e-05,
|
| 558 |
+
"loss": 1.7106,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.0356906807666886,
|
| 563 |
+
"grad_norm": 1.764172077178955,
|
| 564 |
+
"learning_rate": 4.423868312757202e-05,
|
| 565 |
+
"loss": 1.7251,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.062128222075347,
|
| 570 |
+
"grad_norm": 1.5473984479904175,
|
| 571 |
+
"learning_rate": 4.409171075837743e-05,
|
| 572 |
+
"loss": 1.7107,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.0885657633840053,
|
| 577 |
+
"grad_norm": 1.851131558418274,
|
| 578 |
+
"learning_rate": 4.394473838918284e-05,
|
| 579 |
+
"loss": 1.7104,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.1150033046926637,
|
| 584 |
+
"grad_norm": 1.5341060161590576,
|
| 585 |
+
"learning_rate": 4.3797766019988246e-05,
|
| 586 |
+
"loss": 1.7051,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.1414408460013217,
|
| 591 |
+
"grad_norm": 1.7939796447753906,
|
| 592 |
+
"learning_rate": 4.3650793650793655e-05,
|
| 593 |
+
"loss": 1.7393,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.16787838730998,
|
| 598 |
+
"grad_norm": 1.713605284690857,
|
| 599 |
+
"learning_rate": 4.3503821281599064e-05,
|
| 600 |
+
"loss": 1.6997,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.1943159286186384,
|
| 605 |
+
"grad_norm": 1.7437527179718018,
|
| 606 |
+
"learning_rate": 4.3356848912404466e-05,
|
| 607 |
+
"loss": 1.7004,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.220753469927297,
|
| 612 |
+
"grad_norm": 1.8171709775924683,
|
| 613 |
+
"learning_rate": 4.3209876543209875e-05,
|
| 614 |
+
"loss": 1.7275,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.247191011235955,
|
| 619 |
+
"grad_norm": 1.628079891204834,
|
| 620 |
+
"learning_rate": 4.306290417401529e-05,
|
| 621 |
+
"loss": 1.7378,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.273628552544613,
|
| 626 |
+
"grad_norm": 1.7622318267822266,
|
| 627 |
+
"learning_rate": 4.29159318048207e-05,
|
| 628 |
+
"loss": 1.7247,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.3000660938532715,
|
| 633 |
+
"grad_norm": 1.4866081476211548,
|
| 634 |
+
"learning_rate": 4.27689594356261e-05,
|
| 635 |
+
"loss": 1.7166,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.32650363516193,
|
| 640 |
+
"grad_norm": 1.7897216081619263,
|
| 641 |
+
"learning_rate": 4.262198706643151e-05,
|
| 642 |
+
"loss": 1.7096,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.3529411764705883,
|
| 647 |
+
"grad_norm": 1.7703666687011719,
|
| 648 |
+
"learning_rate": 4.247501469723692e-05,
|
| 649 |
+
"loss": 1.7096,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.3793787177792467,
|
| 654 |
+
"grad_norm": 1.9003839492797852,
|
| 655 |
+
"learning_rate": 4.232804232804233e-05,
|
| 656 |
+
"loss": 1.7172,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.4058162590879046,
|
| 661 |
+
"grad_norm": 2.338798999786377,
|
| 662 |
+
"learning_rate": 4.2181069958847744e-05,
|
| 663 |
+
"loss": 1.7209,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.432253800396563,
|
| 668 |
+
"grad_norm": 2.1922528743743896,
|
| 669 |
+
"learning_rate": 4.2034097589653146e-05,
|
| 670 |
+
"loss": 1.7195,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.4586913417052214,
|
| 675 |
+
"grad_norm": 1.7596076726913452,
|
| 676 |
+
"learning_rate": 4.1887125220458555e-05,
|
| 677 |
+
"loss": 1.7316,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.48512888301388,
|
| 682 |
+
"grad_norm": 1.8625160455703735,
|
| 683 |
+
"learning_rate": 4.1740152851263964e-05,
|
| 684 |
+
"loss": 1.7224,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 2.511566424322538,
|
| 689 |
+
"grad_norm": 2.2473673820495605,
|
| 690 |
+
"learning_rate": 4.159318048206937e-05,
|
| 691 |
+
"loss": 1.7198,
|
| 692 |
+
"step": 950
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.538003965631196,
|
| 696 |
+
"grad_norm": 1.8248813152313232,
|
| 697 |
+
"learning_rate": 4.144620811287478e-05,
|
| 698 |
+
"loss": 1.7246,
|
| 699 |
+
"step": 960
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 2.5644415069398545,
|
| 703 |
+
"grad_norm": 1.8088594675064087,
|
| 704 |
+
"learning_rate": 4.129923574368019e-05,
|
| 705 |
+
"loss": 1.707,
|
| 706 |
+
"step": 970
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.590879048248513,
|
| 710 |
+
"grad_norm": 1.818827748298645,
|
| 711 |
+
"learning_rate": 4.11522633744856e-05,
|
| 712 |
+
"loss": 1.7038,
|
| 713 |
+
"step": 980
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 2.6173165895571713,
|
| 717 |
+
"grad_norm": 1.716060757637024,
|
| 718 |
+
"learning_rate": 4.100529100529101e-05,
|
| 719 |
+
"loss": 1.7066,
|
| 720 |
+
"step": 990
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 2.6437541308658297,
|
| 724 |
+
"grad_norm": 1.7306545972824097,
|
| 725 |
+
"learning_rate": 4.085831863609642e-05,
|
| 726 |
+
"loss": 1.6767,
|
| 727 |
+
"step": 1000
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 2.6701916721744876,
|
| 731 |
+
"grad_norm": 1.7304011583328247,
|
| 732 |
+
"learning_rate": 4.071134626690182e-05,
|
| 733 |
+
"loss": 1.6969,
|
| 734 |
+
"step": 1010
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 2.696629213483146,
|
| 738 |
+
"grad_norm": 2.024599552154541,
|
| 739 |
+
"learning_rate": 4.056437389770723e-05,
|
| 740 |
+
"loss": 1.7181,
|
| 741 |
+
"step": 1020
|
| 742 |
+
},
|
| 743 |
+
{
|
| 744 |
+
"epoch": 2.7230667547918044,
|
| 745 |
+
"grad_norm": 2.146998643875122,
|
| 746 |
+
"learning_rate": 4.0417401528512645e-05,
|
| 747 |
+
"loss": 1.7311,
|
| 748 |
+
"step": 1030
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 2.7495042961004628,
|
| 752 |
+
"grad_norm": 1.935433030128479,
|
| 753 |
+
"learning_rate": 4.0270429159318054e-05,
|
| 754 |
+
"loss": 1.7194,
|
| 755 |
+
"step": 1040
|
| 756 |
+
},
|
| 757 |
+
{
|
| 758 |
+
"epoch": 2.775941837409121,
|
| 759 |
+
"grad_norm": 1.6351971626281738,
|
| 760 |
+
"learning_rate": 4.012345679012346e-05,
|
| 761 |
+
"loss": 1.7519,
|
| 762 |
+
"step": 1050
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 2.802379378717779,
|
| 766 |
+
"grad_norm": 1.971442461013794,
|
| 767 |
+
"learning_rate": 3.9976484420928865e-05,
|
| 768 |
+
"loss": 1.7377,
|
| 769 |
+
"step": 1060
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.8288169200264375,
|
| 773 |
+
"grad_norm": 1.8716827630996704,
|
| 774 |
+
"learning_rate": 3.9829512051734274e-05,
|
| 775 |
+
"loss": 1.7066,
|
| 776 |
+
"step": 1070
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.855254461335096,
|
| 780 |
+
"grad_norm": 1.6349095106124878,
|
| 781 |
+
"learning_rate": 3.968253968253968e-05,
|
| 782 |
+
"loss": 1.7,
|
| 783 |
+
"step": 1080
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 2.8816920026437542,
|
| 787 |
+
"grad_norm": 1.7663626670837402,
|
| 788 |
+
"learning_rate": 3.95355673133451e-05,
|
| 789 |
+
"loss": 1.6906,
|
| 790 |
+
"step": 1090
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.9081295439524126,
|
| 794 |
+
"grad_norm": 2.1080522537231445,
|
| 795 |
+
"learning_rate": 3.93885949441505e-05,
|
| 796 |
+
"loss": 1.7245,
|
| 797 |
+
"step": 1100
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 2.9345670852610706,
|
| 801 |
+
"grad_norm": 1.7439042329788208,
|
| 802 |
+
"learning_rate": 3.924162257495591e-05,
|
| 803 |
+
"loss": 1.6977,
|
| 804 |
+
"step": 1110
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.961004626569729,
|
| 808 |
+
"grad_norm": 1.6347275972366333,
|
| 809 |
+
"learning_rate": 3.909465020576132e-05,
|
| 810 |
+
"loss": 1.7225,
|
| 811 |
+
"step": 1120
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.9874421678783873,
|
| 815 |
+
"grad_norm": 2.7075507640838623,
|
| 816 |
+
"learning_rate": 3.894767783656673e-05,
|
| 817 |
+
"loss": 1.7123,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.9980171844018506,
|
| 822 |
+
"eval_accuracy": 0.30069161198794114,
|
| 823 |
+
"eval_loss": 1.6861532926559448,
|
| 824 |
+
"eval_runtime": 594.266,
|
| 825 |
+
"eval_samples_per_second": 47.445,
|
| 826 |
+
"eval_steps_per_second": 0.742,
|
| 827 |
+
"step": 1134
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.0138797091870457,
|
| 831 |
+
"grad_norm": 2.001418113708496,
|
| 832 |
+
"learning_rate": 3.8800705467372136e-05,
|
| 833 |
+
"loss": 1.7138,
|
| 834 |
+
"step": 1140
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.040317250495704,
|
| 838 |
+
"grad_norm": 1.5465753078460693,
|
| 839 |
+
"learning_rate": 3.8653733098177545e-05,
|
| 840 |
+
"loss": 1.6694,
|
| 841 |
+
"step": 1150
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.066754791804362,
|
| 845 |
+
"grad_norm": 2.0662121772766113,
|
| 846 |
+
"learning_rate": 3.8506760728982954e-05,
|
| 847 |
+
"loss": 1.6635,
|
| 848 |
+
"step": 1160
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.0931923331130204,
|
| 852 |
+
"grad_norm": 1.6004307270050049,
|
| 853 |
+
"learning_rate": 3.835978835978836e-05,
|
| 854 |
+
"loss": 1.6928,
|
| 855 |
+
"step": 1170
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.119629874421679,
|
| 859 |
+
"grad_norm": 2.330245018005371,
|
| 860 |
+
"learning_rate": 3.821281599059377e-05,
|
| 861 |
+
"loss": 1.6512,
|
| 862 |
+
"step": 1180
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.146067415730337,
|
| 866 |
+
"grad_norm": 2.421938896179199,
|
| 867 |
+
"learning_rate": 3.806584362139918e-05,
|
| 868 |
+
"loss": 1.6904,
|
| 869 |
+
"step": 1190
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.1725049570389956,
|
| 873 |
+
"grad_norm": 1.943880319595337,
|
| 874 |
+
"learning_rate": 3.791887125220458e-05,
|
| 875 |
+
"loss": 1.6807,
|
| 876 |
+
"step": 1200
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.1989424983476535,
|
| 880 |
+
"grad_norm": 1.9737601280212402,
|
| 881 |
+
"learning_rate": 3.777189888301e-05,
|
| 882 |
+
"loss": 1.6763,
|
| 883 |
+
"step": 1210
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.225380039656312,
|
| 887 |
+
"grad_norm": 1.5722044706344604,
|
| 888 |
+
"learning_rate": 3.762492651381541e-05,
|
| 889 |
+
"loss": 1.6449,
|
| 890 |
+
"step": 1220
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.2518175809649703,
|
| 894 |
+
"grad_norm": 1.8357861042022705,
|
| 895 |
+
"learning_rate": 3.7477954144620817e-05,
|
| 896 |
+
"loss": 1.6637,
|
| 897 |
+
"step": 1230
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.2782551222736287,
|
| 901 |
+
"grad_norm": 2.0803420543670654,
|
| 902 |
+
"learning_rate": 3.733098177542622e-05,
|
| 903 |
+
"loss": 1.6715,
|
| 904 |
+
"step": 1240
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 3.304692663582287,
|
| 908 |
+
"grad_norm": 1.9865127801895142,
|
| 909 |
+
"learning_rate": 3.718400940623163e-05,
|
| 910 |
+
"loss": 1.6841,
|
| 911 |
+
"step": 1250
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.331130204890945,
|
| 915 |
+
"grad_norm": 1.953385353088379,
|
| 916 |
+
"learning_rate": 3.7037037037037037e-05,
|
| 917 |
+
"loss": 1.6699,
|
| 918 |
+
"step": 1260
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 3.3575677461996034,
|
| 922 |
+
"grad_norm": 4.0031867027282715,
|
| 923 |
+
"learning_rate": 3.689006466784245e-05,
|
| 924 |
+
"loss": 1.6795,
|
| 925 |
+
"step": 1270
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 3.384005287508262,
|
| 929 |
+
"grad_norm": 2.2489538192749023,
|
| 930 |
+
"learning_rate": 3.6743092298647854e-05,
|
| 931 |
+
"loss": 1.6705,
|
| 932 |
+
"step": 1280
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 3.41044282881692,
|
| 936 |
+
"grad_norm": 1.7907918691635132,
|
| 937 |
+
"learning_rate": 3.659611992945326e-05,
|
| 938 |
+
"loss": 1.6754,
|
| 939 |
+
"step": 1290
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 3.4368803701255786,
|
| 943 |
+
"grad_norm": 1.7390344142913818,
|
| 944 |
+
"learning_rate": 3.644914756025867e-05,
|
| 945 |
+
"loss": 1.6621,
|
| 946 |
+
"step": 1300
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 3.4633179114342365,
|
| 950 |
+
"grad_norm": 1.6002901792526245,
|
| 951 |
+
"learning_rate": 3.630217519106408e-05,
|
| 952 |
+
"loss": 1.679,
|
| 953 |
+
"step": 1310
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 3.489755452742895,
|
| 957 |
+
"grad_norm": 1.9788098335266113,
|
| 958 |
+
"learning_rate": 3.615520282186949e-05,
|
| 959 |
+
"loss": 1.6633,
|
| 960 |
+
"step": 1320
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 3.5161929940515533,
|
| 964 |
+
"grad_norm": 2.1224050521850586,
|
| 965 |
+
"learning_rate": 3.60082304526749e-05,
|
| 966 |
+
"loss": 1.679,
|
| 967 |
+
"step": 1330
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 3.5426305353602117,
|
| 971 |
+
"grad_norm": 2.0570385456085205,
|
| 972 |
+
"learning_rate": 3.586125808348031e-05,
|
| 973 |
+
"loss": 1.6752,
|
| 974 |
+
"step": 1340
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 3.56906807666887,
|
| 978 |
+
"grad_norm": 2.158527374267578,
|
| 979 |
+
"learning_rate": 3.571428571428572e-05,
|
| 980 |
+
"loss": 1.6624,
|
| 981 |
+
"step": 1350
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 3.595505617977528,
|
| 985 |
+
"grad_norm": 2.440808057785034,
|
| 986 |
+
"learning_rate": 3.5567313345091126e-05,
|
| 987 |
+
"loss": 1.66,
|
| 988 |
+
"step": 1360
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 3.6219431592861864,
|
| 992 |
+
"grad_norm": 2.058363914489746,
|
| 993 |
+
"learning_rate": 3.5420340975896535e-05,
|
| 994 |
+
"loss": 1.6985,
|
| 995 |
+
"step": 1370
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 3.6483807005948448,
|
| 999 |
+
"grad_norm": 1.795248031616211,
|
| 1000 |
+
"learning_rate": 3.527336860670194e-05,
|
| 1001 |
+
"loss": 1.7037,
|
| 1002 |
+
"step": 1380
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 3.6748182419035027,
|
| 1006 |
+
"grad_norm": 1.7252378463745117,
|
| 1007 |
+
"learning_rate": 3.512639623750735e-05,
|
| 1008 |
+
"loss": 1.6633,
|
| 1009 |
+
"step": 1390
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 3.7012557832121615,
|
| 1013 |
+
"grad_norm": 1.8798094987869263,
|
| 1014 |
+
"learning_rate": 3.497942386831276e-05,
|
| 1015 |
+
"loss": 1.6881,
|
| 1016 |
+
"step": 1400
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 3.7276933245208195,
|
| 1020 |
+
"grad_norm": 2.317443609237671,
|
| 1021 |
+
"learning_rate": 3.483245149911817e-05,
|
| 1022 |
+
"loss": 1.6916,
|
| 1023 |
+
"step": 1410
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 3.754130865829478,
|
| 1027 |
+
"grad_norm": 2.2357330322265625,
|
| 1028 |
+
"learning_rate": 3.468547912992357e-05,
|
| 1029 |
+
"loss": 1.665,
|
| 1030 |
+
"step": 1420
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 3.7805684071381362,
|
| 1034 |
+
"grad_norm": 1.8255904912948608,
|
| 1035 |
+
"learning_rate": 3.453850676072898e-05,
|
| 1036 |
+
"loss": 1.6839,
|
| 1037 |
+
"step": 1430
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 3.807005948446794,
|
| 1041 |
+
"grad_norm": 1.95132315158844,
|
| 1042 |
+
"learning_rate": 3.439153439153439e-05,
|
| 1043 |
+
"loss": 1.6977,
|
| 1044 |
+
"step": 1440
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 3.833443489755453,
|
| 1048 |
+
"grad_norm": 1.942299723625183,
|
| 1049 |
+
"learning_rate": 3.4244562022339806e-05,
|
| 1050 |
+
"loss": 1.6648,
|
| 1051 |
+
"step": 1450
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 3.859881031064111,
|
| 1055 |
+
"grad_norm": 2.280670642852783,
|
| 1056 |
+
"learning_rate": 3.4097589653145215e-05,
|
| 1057 |
+
"loss": 1.6621,
|
| 1058 |
+
"step": 1460
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 3.8863185723727693,
|
| 1062 |
+
"grad_norm": 2.0500681400299072,
|
| 1063 |
+
"learning_rate": 3.395061728395062e-05,
|
| 1064 |
+
"loss": 1.6395,
|
| 1065 |
+
"step": 1470
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 3.9127561136814277,
|
| 1069 |
+
"grad_norm": 1.9235775470733643,
|
| 1070 |
+
"learning_rate": 3.3803644914756026e-05,
|
| 1071 |
+
"loss": 1.6907,
|
| 1072 |
+
"step": 1480
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 3.9391936549900857,
|
| 1076 |
+
"grad_norm": 1.8194037675857544,
|
| 1077 |
+
"learning_rate": 3.3656672545561435e-05,
|
| 1078 |
+
"loss": 1.6793,
|
| 1079 |
+
"step": 1490
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 3.965631196298744,
|
| 1083 |
+
"grad_norm": 2.047736167907715,
|
| 1084 |
+
"learning_rate": 3.3509700176366844e-05,
|
| 1085 |
+
"loss": 1.682,
|
| 1086 |
+
"step": 1500
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 3.9920687376074024,
|
| 1090 |
+
"grad_norm": 2.2198452949523926,
|
| 1091 |
+
"learning_rate": 3.336272780717225e-05,
|
| 1092 |
+
"loss": 1.6607,
|
| 1093 |
+
"step": 1510
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 4.0,
|
| 1097 |
+
"eval_accuracy": 0.3080688065259798,
|
| 1098 |
+
"eval_loss": 1.6822971105575562,
|
| 1099 |
+
"eval_runtime": 603.9647,
|
| 1100 |
+
"eval_samples_per_second": 46.683,
|
| 1101 |
+
"eval_steps_per_second": 0.73,
|
| 1102 |
+
"step": 1513
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 4.01850627891606,
|
| 1106 |
+
"grad_norm": 1.416684627532959,
|
| 1107 |
+
"learning_rate": 3.321575543797766e-05,
|
| 1108 |
+
"loss": 1.6447,
|
| 1109 |
+
"step": 1520
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 4.044943820224719,
|
| 1113 |
+
"grad_norm": 2.129450798034668,
|
| 1114 |
+
"learning_rate": 3.306878306878307e-05,
|
| 1115 |
+
"loss": 1.6303,
|
| 1116 |
+
"step": 1530
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 4.071381361533377,
|
| 1120 |
+
"grad_norm": 1.7213213443756104,
|
| 1121 |
+
"learning_rate": 3.292181069958848e-05,
|
| 1122 |
+
"loss": 1.6126,
|
| 1123 |
+
"step": 1540
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 4.097818902842036,
|
| 1127 |
+
"grad_norm": 2.0465450286865234,
|
| 1128 |
+
"learning_rate": 3.277483833039389e-05,
|
| 1129 |
+
"loss": 1.6351,
|
| 1130 |
+
"step": 1550
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 4.124256444150694,
|
| 1134 |
+
"grad_norm": 2.259282350540161,
|
| 1135 |
+
"learning_rate": 3.262786596119929e-05,
|
| 1136 |
+
"loss": 1.6437,
|
| 1137 |
+
"step": 1560
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.150693985459352,
|
| 1141 |
+
"grad_norm": 2.2626054286956787,
|
| 1142 |
+
"learning_rate": 3.24808935920047e-05,
|
| 1143 |
+
"loss": 1.6523,
|
| 1144 |
+
"step": 1570
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 4.177131526768011,
|
| 1148 |
+
"grad_norm": 2.269216537475586,
|
| 1149 |
+
"learning_rate": 3.2333921222810116e-05,
|
| 1150 |
+
"loss": 1.6369,
|
| 1151 |
+
"step": 1580
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 4.203569068076669,
|
| 1155 |
+
"grad_norm": 2.1457483768463135,
|
| 1156 |
+
"learning_rate": 3.2186948853615525e-05,
|
| 1157 |
+
"loss": 1.6211,
|
| 1158 |
+
"step": 1590
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 4.2300066093853275,
|
| 1162 |
+
"grad_norm": 1.9659942388534546,
|
| 1163 |
+
"learning_rate": 3.2039976484420934e-05,
|
| 1164 |
+
"loss": 1.639,
|
| 1165 |
+
"step": 1600
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 4.256444150693985,
|
| 1169 |
+
"grad_norm": 2.447136402130127,
|
| 1170 |
+
"learning_rate": 3.1893004115226336e-05,
|
| 1171 |
+
"loss": 1.6548,
|
| 1172 |
+
"step": 1610
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 4.282881692002643,
|
| 1176 |
+
"grad_norm": 2.2011499404907227,
|
| 1177 |
+
"learning_rate": 3.1746031746031745e-05,
|
| 1178 |
+
"loss": 1.6042,
|
| 1179 |
+
"step": 1620
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 4.309319233311302,
|
| 1183 |
+
"grad_norm": 1.9171946048736572,
|
| 1184 |
+
"learning_rate": 3.1599059376837154e-05,
|
| 1185 |
+
"loss": 1.6273,
|
| 1186 |
+
"step": 1630
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 4.33575677461996,
|
| 1190 |
+
"grad_norm": 1.9522180557250977,
|
| 1191 |
+
"learning_rate": 3.145208700764257e-05,
|
| 1192 |
+
"loss": 1.6479,
|
| 1193 |
+
"step": 1640
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 4.362194315928619,
|
| 1197 |
+
"grad_norm": 1.8249017000198364,
|
| 1198 |
+
"learning_rate": 3.130511463844797e-05,
|
| 1199 |
+
"loss": 1.633,
|
| 1200 |
+
"step": 1650
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 4.388631857237277,
|
| 1204 |
+
"grad_norm": 2.093296766281128,
|
| 1205 |
+
"learning_rate": 3.115814226925338e-05,
|
| 1206 |
+
"loss": 1.6519,
|
| 1207 |
+
"step": 1660
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 4.415069398545935,
|
| 1211 |
+
"grad_norm": 2.218247652053833,
|
| 1212 |
+
"learning_rate": 3.101116990005879e-05,
|
| 1213 |
+
"loss": 1.6478,
|
| 1214 |
+
"step": 1670
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 4.441506939854594,
|
| 1218 |
+
"grad_norm": 1.9361261129379272,
|
| 1219 |
+
"learning_rate": 3.08641975308642e-05,
|
| 1220 |
+
"loss": 1.6473,
|
| 1221 |
+
"step": 1680
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 4.467944481163252,
|
| 1225 |
+
"grad_norm": 1.9687649011611938,
|
| 1226 |
+
"learning_rate": 3.071722516166961e-05,
|
| 1227 |
+
"loss": 1.6263,
|
| 1228 |
+
"step": 1690
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 4.49438202247191,
|
| 1232 |
+
"grad_norm": 1.8882747888565063,
|
| 1233 |
+
"learning_rate": 3.0570252792475016e-05,
|
| 1234 |
+
"loss": 1.6531,
|
| 1235 |
+
"step": 1700
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 4.520819563780568,
|
| 1239 |
+
"grad_norm": 2.0464086532592773,
|
| 1240 |
+
"learning_rate": 3.0423280423280425e-05,
|
| 1241 |
+
"loss": 1.6481,
|
| 1242 |
+
"step": 1710
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 4.547257105089226,
|
| 1246 |
+
"grad_norm": 2.1452088356018066,
|
| 1247 |
+
"learning_rate": 3.0276308054085834e-05,
|
| 1248 |
+
"loss": 1.6126,
|
| 1249 |
+
"step": 1720
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 4.573694646397885,
|
| 1253 |
+
"grad_norm": 1.7108988761901855,
|
| 1254 |
+
"learning_rate": 3.012933568489124e-05,
|
| 1255 |
+
"loss": 1.6365,
|
| 1256 |
+
"step": 1730
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 4.600132187706543,
|
| 1260 |
+
"grad_norm": 1.7739413976669312,
|
| 1261 |
+
"learning_rate": 2.998236331569665e-05,
|
| 1262 |
+
"loss": 1.6373,
|
| 1263 |
+
"step": 1740
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 4.626569729015202,
|
| 1267 |
+
"grad_norm": 1.9384492635726929,
|
| 1268 |
+
"learning_rate": 2.9835390946502057e-05,
|
| 1269 |
+
"loss": 1.6212,
|
| 1270 |
+
"step": 1750
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 4.65300727032386,
|
| 1274 |
+
"grad_norm": 2.0029444694519043,
|
| 1275 |
+
"learning_rate": 2.968841857730747e-05,
|
| 1276 |
+
"loss": 1.6185,
|
| 1277 |
+
"step": 1760
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 4.679444811632518,
|
| 1281 |
+
"grad_norm": 2.2173917293548584,
|
| 1282 |
+
"learning_rate": 2.954144620811288e-05,
|
| 1283 |
+
"loss": 1.6441,
|
| 1284 |
+
"step": 1770
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 4.705882352941177,
|
| 1288 |
+
"grad_norm": 2.4079232215881348,
|
| 1289 |
+
"learning_rate": 2.9394473838918284e-05,
|
| 1290 |
+
"loss": 1.6441,
|
| 1291 |
+
"step": 1780
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 4.732319894249835,
|
| 1295 |
+
"grad_norm": 2.1316912174224854,
|
| 1296 |
+
"learning_rate": 2.9247501469723693e-05,
|
| 1297 |
+
"loss": 1.6464,
|
| 1298 |
+
"step": 1790
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 4.758757435558493,
|
| 1302 |
+
"grad_norm": 2.2552201747894287,
|
| 1303 |
+
"learning_rate": 2.91005291005291e-05,
|
| 1304 |
+
"loss": 1.6321,
|
| 1305 |
+
"step": 1800
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 4.785194976867151,
|
| 1309 |
+
"grad_norm": 2.0181891918182373,
|
| 1310 |
+
"learning_rate": 2.8953556731334508e-05,
|
| 1311 |
+
"loss": 1.6584,
|
| 1312 |
+
"step": 1810
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 4.811632518175809,
|
| 1316 |
+
"grad_norm": 1.912116527557373,
|
| 1317 |
+
"learning_rate": 2.880658436213992e-05,
|
| 1318 |
+
"loss": 1.6618,
|
| 1319 |
+
"step": 1820
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 4.838070059484468,
|
| 1323 |
+
"grad_norm": 1.916869878768921,
|
| 1324 |
+
"learning_rate": 2.865961199294533e-05,
|
| 1325 |
+
"loss": 1.6719,
|
| 1326 |
+
"step": 1830
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 4.864507600793126,
|
| 1330 |
+
"grad_norm": 2.0317203998565674,
|
| 1331 |
+
"learning_rate": 2.8512639623750738e-05,
|
| 1332 |
+
"loss": 1.6258,
|
| 1333 |
+
"step": 1840
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 4.890945142101785,
|
| 1337 |
+
"grad_norm": 2.080411195755005,
|
| 1338 |
+
"learning_rate": 2.8365667254556143e-05,
|
| 1339 |
+
"loss": 1.6518,
|
| 1340 |
+
"step": 1850
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 4.917382683410443,
|
| 1344 |
+
"grad_norm": 1.9478580951690674,
|
| 1345 |
+
"learning_rate": 2.8218694885361552e-05,
|
| 1346 |
+
"loss": 1.614,
|
| 1347 |
+
"step": 1860
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 4.943820224719101,
|
| 1351 |
+
"grad_norm": 1.889710545539856,
|
| 1352 |
+
"learning_rate": 2.8071722516166958e-05,
|
| 1353 |
+
"loss": 1.6221,
|
| 1354 |
+
"step": 1870
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 4.97025776602776,
|
| 1358 |
+
"grad_norm": 2.255192756652832,
|
| 1359 |
+
"learning_rate": 2.7924750146972374e-05,
|
| 1360 |
+
"loss": 1.6271,
|
| 1361 |
+
"step": 1880
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 4.9966953073364175,
|
| 1365 |
+
"grad_norm": 2.026352643966675,
|
| 1366 |
+
"learning_rate": 2.777777777777778e-05,
|
| 1367 |
+
"loss": 1.6188,
|
| 1368 |
+
"step": 1890
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 4.999339061467284,
|
| 1372 |
+
"eval_accuracy": 0.3107997871963114,
|
| 1373 |
+
"eval_loss": 1.6907494068145752,
|
| 1374 |
+
"eval_runtime": 587.6675,
|
| 1375 |
+
"eval_samples_per_second": 47.978,
|
| 1376 |
+
"eval_steps_per_second": 0.75,
|
| 1377 |
+
"step": 1891
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 5.023132848645076,
|
| 1381 |
+
"grad_norm": 1.7152390480041504,
|
| 1382 |
+
"learning_rate": 2.7630805408583188e-05,
|
| 1383 |
+
"loss": 1.6156,
|
| 1384 |
+
"step": 1900
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 5.049570389953734,
|
| 1388 |
+
"grad_norm": 2.2232937812805176,
|
| 1389 |
+
"learning_rate": 2.7483833039388597e-05,
|
| 1390 |
+
"loss": 1.618,
|
| 1391 |
+
"step": 1910
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 5.076007931262392,
|
| 1395 |
+
"grad_norm": 1.9228154420852661,
|
| 1396 |
+
"learning_rate": 2.7336860670194003e-05,
|
| 1397 |
+
"loss": 1.6137,
|
| 1398 |
+
"step": 1920
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 5.102445472571051,
|
| 1402 |
+
"grad_norm": 1.9937798976898193,
|
| 1403 |
+
"learning_rate": 2.718988830099941e-05,
|
| 1404 |
+
"loss": 1.6139,
|
| 1405 |
+
"step": 1930
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 5.128883013879709,
|
| 1409 |
+
"grad_norm": 2.4573240280151367,
|
| 1410 |
+
"learning_rate": 2.7042915931804824e-05,
|
| 1411 |
+
"loss": 1.5976,
|
| 1412 |
+
"step": 1940
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 5.155320555188368,
|
| 1416 |
+
"grad_norm": 2.107956647872925,
|
| 1417 |
+
"learning_rate": 2.6895943562610233e-05,
|
| 1418 |
+
"loss": 1.6046,
|
| 1419 |
+
"step": 1950
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 5.181758096497026,
|
| 1423 |
+
"grad_norm": 1.9604800939559937,
|
| 1424 |
+
"learning_rate": 2.6748971193415638e-05,
|
| 1425 |
+
"loss": 1.6001,
|
| 1426 |
+
"step": 1960
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 5.208195637805684,
|
| 1430 |
+
"grad_norm": 2.3065364360809326,
|
| 1431 |
+
"learning_rate": 2.6601998824221047e-05,
|
| 1432 |
+
"loss": 1.5818,
|
| 1433 |
+
"step": 1970
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 5.234633179114343,
|
| 1437 |
+
"grad_norm": 2.3068606853485107,
|
| 1438 |
+
"learning_rate": 2.6455026455026456e-05,
|
| 1439 |
+
"loss": 1.629,
|
| 1440 |
+
"step": 1980
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 5.2610707204230005,
|
| 1444 |
+
"grad_norm": 1.9791010618209839,
|
| 1445 |
+
"learning_rate": 2.630805408583186e-05,
|
| 1446 |
+
"loss": 1.5831,
|
| 1447 |
+
"step": 1990
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 5.287508261731659,
|
| 1451 |
+
"grad_norm": 2.128002643585205,
|
| 1452 |
+
"learning_rate": 2.6161081716637274e-05,
|
| 1453 |
+
"loss": 1.6208,
|
| 1454 |
+
"step": 2000
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 5.313945803040317,
|
| 1458 |
+
"grad_norm": 1.999395728111267,
|
| 1459 |
+
"learning_rate": 2.6014109347442683e-05,
|
| 1460 |
+
"loss": 1.5791,
|
| 1461 |
+
"step": 2010
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 5.340383344348975,
|
| 1465 |
+
"grad_norm": 2.4271738529205322,
|
| 1466 |
+
"learning_rate": 2.5867136978248092e-05,
|
| 1467 |
+
"loss": 1.6145,
|
| 1468 |
+
"step": 2020
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 5.366820885657634,
|
| 1472 |
+
"grad_norm": 2.0685036182403564,
|
| 1473 |
+
"learning_rate": 2.5720164609053497e-05,
|
| 1474 |
+
"loss": 1.6013,
|
| 1475 |
+
"step": 2030
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 5.393258426966292,
|
| 1479 |
+
"grad_norm": 2.3582699298858643,
|
| 1480 |
+
"learning_rate": 2.5573192239858906e-05,
|
| 1481 |
+
"loss": 1.5878,
|
| 1482 |
+
"step": 2040
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 5.419695968274951,
|
| 1486 |
+
"grad_norm": 2.0049946308135986,
|
| 1487 |
+
"learning_rate": 2.5426219870664315e-05,
|
| 1488 |
+
"loss": 1.6114,
|
| 1489 |
+
"step": 2050
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 5.446133509583609,
|
| 1493 |
+
"grad_norm": 1.902218222618103,
|
| 1494 |
+
"learning_rate": 2.5279247501469728e-05,
|
| 1495 |
+
"loss": 1.5934,
|
| 1496 |
+
"step": 2060
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 5.472571050892267,
|
| 1500 |
+
"grad_norm": 2.1240365505218506,
|
| 1501 |
+
"learning_rate": 2.5132275132275137e-05,
|
| 1502 |
+
"loss": 1.561,
|
| 1503 |
+
"step": 2070
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 5.4990085922009255,
|
| 1507 |
+
"grad_norm": 2.0185890197753906,
|
| 1508 |
+
"learning_rate": 2.4985302763080542e-05,
|
| 1509 |
+
"loss": 1.5912,
|
| 1510 |
+
"step": 2080
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 5.5254461335095835,
|
| 1514 |
+
"grad_norm": 2.5916292667388916,
|
| 1515 |
+
"learning_rate": 2.483833039388595e-05,
|
| 1516 |
+
"loss": 1.6176,
|
| 1517 |
+
"step": 2090
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 5.551883674818242,
|
| 1521 |
+
"grad_norm": 2.4469070434570312,
|
| 1522 |
+
"learning_rate": 2.4691358024691357e-05,
|
| 1523 |
+
"loss": 1.6373,
|
| 1524 |
+
"step": 2100
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 5.5783212161269,
|
| 1528 |
+
"grad_norm": 1.9966007471084595,
|
| 1529 |
+
"learning_rate": 2.454438565549677e-05,
|
| 1530 |
+
"loss": 1.6049,
|
| 1531 |
+
"step": 2110
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 5.604758757435558,
|
| 1535 |
+
"grad_norm": 2.5291779041290283,
|
| 1536 |
+
"learning_rate": 2.4397413286302174e-05,
|
| 1537 |
+
"loss": 1.595,
|
| 1538 |
+
"step": 2120
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 5.631196298744217,
|
| 1542 |
+
"grad_norm": 2.031622886657715,
|
| 1543 |
+
"learning_rate": 2.4250440917107583e-05,
|
| 1544 |
+
"loss": 1.6069,
|
| 1545 |
+
"step": 2130
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 5.657633840052875,
|
| 1549 |
+
"grad_norm": 1.9951566457748413,
|
| 1550 |
+
"learning_rate": 2.4103468547912996e-05,
|
| 1551 |
+
"loss": 1.583,
|
| 1552 |
+
"step": 2140
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 5.684071381361534,
|
| 1556 |
+
"grad_norm": 2.609140634536743,
|
| 1557 |
+
"learning_rate": 2.39564961787184e-05,
|
| 1558 |
+
"loss": 1.5821,
|
| 1559 |
+
"step": 2150
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 5.710508922670192,
|
| 1563 |
+
"grad_norm": 2.389698028564453,
|
| 1564 |
+
"learning_rate": 2.380952380952381e-05,
|
| 1565 |
+
"loss": 1.6114,
|
| 1566 |
+
"step": 2160
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 5.73694646397885,
|
| 1570 |
+
"grad_norm": 2.4471077919006348,
|
| 1571 |
+
"learning_rate": 2.366255144032922e-05,
|
| 1572 |
+
"loss": 1.5919,
|
| 1573 |
+
"step": 2170
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 5.7633840052875085,
|
| 1577 |
+
"grad_norm": 2.0070338249206543,
|
| 1578 |
+
"learning_rate": 2.3515579071134628e-05,
|
| 1579 |
+
"loss": 1.5963,
|
| 1580 |
+
"step": 2180
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 5.789821546596166,
|
| 1584 |
+
"grad_norm": 2.1658718585968018,
|
| 1585 |
+
"learning_rate": 2.3368606701940034e-05,
|
| 1586 |
+
"loss": 1.6038,
|
| 1587 |
+
"step": 2190
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 5.816259087904825,
|
| 1591 |
+
"grad_norm": 2.08571457862854,
|
| 1592 |
+
"learning_rate": 2.3221634332745446e-05,
|
| 1593 |
+
"loss": 1.5962,
|
| 1594 |
+
"step": 2200
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 5.842696629213483,
|
| 1598 |
+
"grad_norm": 2.118481397628784,
|
| 1599 |
+
"learning_rate": 2.3074661963550855e-05,
|
| 1600 |
+
"loss": 1.5945,
|
| 1601 |
+
"step": 2210
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 5.869134170522141,
|
| 1605 |
+
"grad_norm": 2.135956048965454,
|
| 1606 |
+
"learning_rate": 2.292768959435626e-05,
|
| 1607 |
+
"loss": 1.5785,
|
| 1608 |
+
"step": 2220
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 5.8955717118308,
|
| 1612 |
+
"grad_norm": 2.1876909732818604,
|
| 1613 |
+
"learning_rate": 2.2780717225161673e-05,
|
| 1614 |
+
"loss": 1.5769,
|
| 1615 |
+
"step": 2230
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 5.922009253139458,
|
| 1619 |
+
"grad_norm": 2.121159315109253,
|
| 1620 |
+
"learning_rate": 2.2633744855967078e-05,
|
| 1621 |
+
"loss": 1.5869,
|
| 1622 |
+
"step": 2240
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 5.948446794448117,
|
| 1626 |
+
"grad_norm": 2.0962424278259277,
|
| 1627 |
+
"learning_rate": 2.2486772486772487e-05,
|
| 1628 |
+
"loss": 1.6203,
|
| 1629 |
+
"step": 2250
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 5.974884335756775,
|
| 1633 |
+
"grad_norm": 1.9238280057907104,
|
| 1634 |
+
"learning_rate": 2.2339800117577896e-05,
|
| 1635 |
+
"loss": 1.6009,
|
| 1636 |
+
"step": 2260
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 5.998678122934567,
|
| 1640 |
+
"eval_accuracy": 0.3150203936868239,
|
| 1641 |
+
"eval_loss": 1.677284598350525,
|
| 1642 |
+
"eval_runtime": 825.1974,
|
| 1643 |
+
"eval_samples_per_second": 34.168,
|
| 1644 |
+
"eval_steps_per_second": 0.534,
|
| 1645 |
+
"step": 2269
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 6.001321877065433,
|
| 1649 |
+
"grad_norm": 3.6250228881835938,
|
| 1650 |
+
"learning_rate": 2.2192827748383305e-05,
|
| 1651 |
+
"loss": 1.6085,
|
| 1652 |
+
"step": 2270
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 6.0277594183740915,
|
| 1656 |
+
"grad_norm": 2.6421561241149902,
|
| 1657 |
+
"learning_rate": 2.2045855379188714e-05,
|
| 1658 |
+
"loss": 1.5612,
|
| 1659 |
+
"step": 2280
|
| 1660 |
+
},
|
| 1661 |
+
{
|
| 1662 |
+
"epoch": 6.054196959682749,
|
| 1663 |
+
"grad_norm": 1.9773502349853516,
|
| 1664 |
+
"learning_rate": 2.1898883009994123e-05,
|
| 1665 |
+
"loss": 1.558,
|
| 1666 |
+
"step": 2290
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 6.080634500991408,
|
| 1670 |
+
"grad_norm": 2.200873374938965,
|
| 1671 |
+
"learning_rate": 2.1751910640799532e-05,
|
| 1672 |
+
"loss": 1.5701,
|
| 1673 |
+
"step": 2300
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 6.107072042300066,
|
| 1677 |
+
"grad_norm": 2.346799373626709,
|
| 1678 |
+
"learning_rate": 2.1604938271604937e-05,
|
| 1679 |
+
"loss": 1.5605,
|
| 1680 |
+
"step": 2310
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 6.133509583608724,
|
| 1684 |
+
"grad_norm": 2.3931312561035156,
|
| 1685 |
+
"learning_rate": 2.145796590241035e-05,
|
| 1686 |
+
"loss": 1.5498,
|
| 1687 |
+
"step": 2320
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 6.159947124917383,
|
| 1691 |
+
"grad_norm": 2.662827968597412,
|
| 1692 |
+
"learning_rate": 2.1310993533215755e-05,
|
| 1693 |
+
"loss": 1.5782,
|
| 1694 |
+
"step": 2330
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 6.186384666226041,
|
| 1698 |
+
"grad_norm": 2.2345123291015625,
|
| 1699 |
+
"learning_rate": 2.1164021164021164e-05,
|
| 1700 |
+
"loss": 1.5672,
|
| 1701 |
+
"step": 2340
|
| 1702 |
+
},
|
| 1703 |
+
{
|
| 1704 |
+
"epoch": 6.2128222075347,
|
| 1705 |
+
"grad_norm": 2.2579140663146973,
|
| 1706 |
+
"learning_rate": 2.1017048794826573e-05,
|
| 1707 |
+
"loss": 1.5323,
|
| 1708 |
+
"step": 2350
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 6.239259748843358,
|
| 1712 |
+
"grad_norm": 2.2183802127838135,
|
| 1713 |
+
"learning_rate": 2.0870076425631982e-05,
|
| 1714 |
+
"loss": 1.5499,
|
| 1715 |
+
"step": 2360
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 6.265697290152016,
|
| 1719 |
+
"grad_norm": 2.227612257003784,
|
| 1720 |
+
"learning_rate": 2.072310405643739e-05,
|
| 1721 |
+
"loss": 1.5522,
|
| 1722 |
+
"step": 2370
|
| 1723 |
+
},
|
| 1724 |
+
{
|
| 1725 |
+
"epoch": 6.292134831460674,
|
| 1726 |
+
"grad_norm": 2.7729430198669434,
|
| 1727 |
+
"learning_rate": 2.05761316872428e-05,
|
| 1728 |
+
"loss": 1.5614,
|
| 1729 |
+
"step": 2380
|
| 1730 |
+
},
|
| 1731 |
+
{
|
| 1732 |
+
"epoch": 6.318572372769332,
|
| 1733 |
+
"grad_norm": 2.166227340698242,
|
| 1734 |
+
"learning_rate": 2.042915931804821e-05,
|
| 1735 |
+
"loss": 1.5358,
|
| 1736 |
+
"step": 2390
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 6.345009914077991,
|
| 1740 |
+
"grad_norm": 2.1141440868377686,
|
| 1741 |
+
"learning_rate": 2.0282186948853614e-05,
|
| 1742 |
+
"loss": 1.554,
|
| 1743 |
+
"step": 2400
|
| 1744 |
+
},
|
| 1745 |
+
{
|
| 1746 |
+
"epoch": 6.371447455386649,
|
| 1747 |
+
"grad_norm": 2.1744093894958496,
|
| 1748 |
+
"learning_rate": 2.0135214579659027e-05,
|
| 1749 |
+
"loss": 1.5583,
|
| 1750 |
+
"step": 2410
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 6.397884996695307,
|
| 1754 |
+
"grad_norm": 2.0957577228546143,
|
| 1755 |
+
"learning_rate": 1.9988242210464432e-05,
|
| 1756 |
+
"loss": 1.5573,
|
| 1757 |
+
"step": 2420
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 6.424322538003966,
|
| 1761 |
+
"grad_norm": 2.6140244007110596,
|
| 1762 |
+
"learning_rate": 1.984126984126984e-05,
|
| 1763 |
+
"loss": 1.5597,
|
| 1764 |
+
"step": 2430
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 6.450760079312624,
|
| 1768 |
+
"grad_norm": 2.0459976196289062,
|
| 1769 |
+
"learning_rate": 1.969429747207525e-05,
|
| 1770 |
+
"loss": 1.5439,
|
| 1771 |
+
"step": 2440
|
| 1772 |
+
},
|
| 1773 |
+
{
|
| 1774 |
+
"epoch": 6.477197620621283,
|
| 1775 |
+
"grad_norm": 2.2509853839874268,
|
| 1776 |
+
"learning_rate": 1.954732510288066e-05,
|
| 1777 |
+
"loss": 1.5478,
|
| 1778 |
+
"step": 2450
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"epoch": 6.503635161929941,
|
| 1782 |
+
"grad_norm": 2.0103518962860107,
|
| 1783 |
+
"learning_rate": 1.9400352733686068e-05,
|
| 1784 |
+
"loss": 1.555,
|
| 1785 |
+
"step": 2460
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 6.530072703238599,
|
| 1789 |
+
"grad_norm": 2.5956430435180664,
|
| 1790 |
+
"learning_rate": 1.9253380364491477e-05,
|
| 1791 |
+
"loss": 1.5604,
|
| 1792 |
+
"step": 2470
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 6.556510244547257,
|
| 1796 |
+
"grad_norm": 2.4196279048919678,
|
| 1797 |
+
"learning_rate": 1.9106407995296886e-05,
|
| 1798 |
+
"loss": 1.554,
|
| 1799 |
+
"step": 2480
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 6.582947785855915,
|
| 1803 |
+
"grad_norm": 2.2963523864746094,
|
| 1804 |
+
"learning_rate": 1.895943562610229e-05,
|
| 1805 |
+
"loss": 1.5345,
|
| 1806 |
+
"step": 2490
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 6.609385327164574,
|
| 1810 |
+
"grad_norm": 2.4753236770629883,
|
| 1811 |
+
"learning_rate": 1.8812463256907704e-05,
|
| 1812 |
+
"loss": 1.5698,
|
| 1813 |
+
"step": 2500
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 6.635822868473232,
|
| 1817 |
+
"grad_norm": 2.251317024230957,
|
| 1818 |
+
"learning_rate": 1.866549088771311e-05,
|
| 1819 |
+
"loss": 1.5625,
|
| 1820 |
+
"step": 2510
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 6.66226040978189,
|
| 1824 |
+
"grad_norm": 2.4385430812835693,
|
| 1825 |
+
"learning_rate": 1.8518518518518518e-05,
|
| 1826 |
+
"loss": 1.5649,
|
| 1827 |
+
"step": 2520
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 6.688697951090549,
|
| 1831 |
+
"grad_norm": 2.1294140815734863,
|
| 1832 |
+
"learning_rate": 1.8371546149323927e-05,
|
| 1833 |
+
"loss": 1.5214,
|
| 1834 |
+
"step": 2530
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 6.715135492399207,
|
| 1838 |
+
"grad_norm": 2.234025478363037,
|
| 1839 |
+
"learning_rate": 1.8224573780129336e-05,
|
| 1840 |
+
"loss": 1.5928,
|
| 1841 |
+
"step": 2540
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 6.741573033707866,
|
| 1845 |
+
"grad_norm": 2.6647934913635254,
|
| 1846 |
+
"learning_rate": 1.8077601410934745e-05,
|
| 1847 |
+
"loss": 1.5519,
|
| 1848 |
+
"step": 2550
|
| 1849 |
+
},
|
| 1850 |
+
{
|
| 1851 |
+
"epoch": 6.768010575016524,
|
| 1852 |
+
"grad_norm": 2.2540123462677,
|
| 1853 |
+
"learning_rate": 1.7930629041740154e-05,
|
| 1854 |
+
"loss": 1.5642,
|
| 1855 |
+
"step": 2560
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 6.7944481163251815,
|
| 1859 |
+
"grad_norm": 2.701176404953003,
|
| 1860 |
+
"learning_rate": 1.7783656672545563e-05,
|
| 1861 |
+
"loss": 1.5733,
|
| 1862 |
+
"step": 2570
|
| 1863 |
+
},
|
| 1864 |
+
{
|
| 1865 |
+
"epoch": 6.82088565763384,
|
| 1866 |
+
"grad_norm": 2.6911590099334717,
|
| 1867 |
+
"learning_rate": 1.763668430335097e-05,
|
| 1868 |
+
"loss": 1.5651,
|
| 1869 |
+
"step": 2580
|
| 1870 |
+
},
|
| 1871 |
+
{
|
| 1872 |
+
"epoch": 6.847323198942498,
|
| 1873 |
+
"grad_norm": 2.3102941513061523,
|
| 1874 |
+
"learning_rate": 1.748971193415638e-05,
|
| 1875 |
+
"loss": 1.5716,
|
| 1876 |
+
"step": 2590
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 6.873760740251157,
|
| 1880 |
+
"grad_norm": 2.4161031246185303,
|
| 1881 |
+
"learning_rate": 1.7342739564961786e-05,
|
| 1882 |
+
"loss": 1.5558,
|
| 1883 |
+
"step": 2600
|
| 1884 |
+
},
|
| 1885 |
+
{
|
| 1886 |
+
"epoch": 6.900198281559815,
|
| 1887 |
+
"grad_norm": 2.4778518676757812,
|
| 1888 |
+
"learning_rate": 1.7195767195767195e-05,
|
| 1889 |
+
"loss": 1.5766,
|
| 1890 |
+
"step": 2610
|
| 1891 |
+
},
|
| 1892 |
+
{
|
| 1893 |
+
"epoch": 6.926635822868473,
|
| 1894 |
+
"grad_norm": 2.188422918319702,
|
| 1895 |
+
"learning_rate": 1.7048794826572608e-05,
|
| 1896 |
+
"loss": 1.5593,
|
| 1897 |
+
"step": 2620
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 6.953073364177132,
|
| 1901 |
+
"grad_norm": 2.146476984024048,
|
| 1902 |
+
"learning_rate": 1.6901822457378013e-05,
|
| 1903 |
+
"loss": 1.6021,
|
| 1904 |
+
"step": 2630
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 6.97951090548579,
|
| 1908 |
+
"grad_norm": 2.322460174560547,
|
| 1909 |
+
"learning_rate": 1.6754850088183422e-05,
|
| 1910 |
+
"loss": 1.5485,
|
| 1911 |
+
"step": 2640
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 6.998017184401851,
|
| 1915 |
+
"eval_accuracy": 0.31984394396169535,
|
| 1916 |
+
"eval_loss": 1.6719752550125122,
|
| 1917 |
+
"eval_runtime": 687.5385,
|
| 1918 |
+
"eval_samples_per_second": 41.009,
|
| 1919 |
+
"eval_steps_per_second": 0.641,
|
| 1920 |
+
"step": 2647
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 7.005948446794448,
|
| 1924 |
+
"grad_norm": 2.46777606010437,
|
| 1925 |
+
"learning_rate": 1.660787771898883e-05,
|
| 1926 |
+
"loss": 1.5634,
|
| 1927 |
+
"step": 2650
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 7.0323859881031066,
|
| 1931 |
+
"grad_norm": 2.2213032245635986,
|
| 1932 |
+
"learning_rate": 1.646090534979424e-05,
|
| 1933 |
+
"loss": 1.5148,
|
| 1934 |
+
"step": 2660
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 7.0588235294117645,
|
| 1938 |
+
"grad_norm": 2.3514413833618164,
|
| 1939 |
+
"learning_rate": 1.6313932980599646e-05,
|
| 1940 |
+
"loss": 1.4979,
|
| 1941 |
+
"step": 2670
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 7.085261070720423,
|
| 1945 |
+
"grad_norm": 2.6042823791503906,
|
| 1946 |
+
"learning_rate": 1.6166960611405058e-05,
|
| 1947 |
+
"loss": 1.5066,
|
| 1948 |
+
"step": 2680
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 7.111698612029081,
|
| 1952 |
+
"grad_norm": 2.6137919425964355,
|
| 1953 |
+
"learning_rate": 1.6019988242210467e-05,
|
| 1954 |
+
"loss": 1.5394,
|
| 1955 |
+
"step": 2690
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 7.138136153337739,
|
| 1959 |
+
"grad_norm": 2.8564794063568115,
|
| 1960 |
+
"learning_rate": 1.5873015873015872e-05,
|
| 1961 |
+
"loss": 1.5452,
|
| 1962 |
+
"step": 2700
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 7.164573694646398,
|
| 1966 |
+
"grad_norm": 2.685819149017334,
|
| 1967 |
+
"learning_rate": 1.5726043503821285e-05,
|
| 1968 |
+
"loss": 1.5514,
|
| 1969 |
+
"step": 2710
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 7.191011235955056,
|
| 1973 |
+
"grad_norm": 2.6131200790405273,
|
| 1974 |
+
"learning_rate": 1.557907113462669e-05,
|
| 1975 |
+
"loss": 1.5208,
|
| 1976 |
+
"step": 2720
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 7.217448777263715,
|
| 1980 |
+
"grad_norm": 2.341998338699341,
|
| 1981 |
+
"learning_rate": 1.54320987654321e-05,
|
| 1982 |
+
"loss": 1.525,
|
| 1983 |
+
"step": 2730
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 7.243886318572373,
|
| 1987 |
+
"grad_norm": 2.3170387744903564,
|
| 1988 |
+
"learning_rate": 1.5285126396237508e-05,
|
| 1989 |
+
"loss": 1.5363,
|
| 1990 |
+
"step": 2740
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 7.270323859881031,
|
| 1994 |
+
"grad_norm": 2.702817440032959,
|
| 1995 |
+
"learning_rate": 1.5138154027042917e-05,
|
| 1996 |
+
"loss": 1.5114,
|
| 1997 |
+
"step": 2750
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 7.2967614011896895,
|
| 2001 |
+
"grad_norm": 2.3331921100616455,
|
| 2002 |
+
"learning_rate": 1.4991181657848324e-05,
|
| 2003 |
+
"loss": 1.5078,
|
| 2004 |
+
"step": 2760
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 7.3231989424983475,
|
| 2008 |
+
"grad_norm": 3.1344053745269775,
|
| 2009 |
+
"learning_rate": 1.4844209288653735e-05,
|
| 2010 |
+
"loss": 1.5251,
|
| 2011 |
+
"step": 2770
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 7.349636483807006,
|
| 2015 |
+
"grad_norm": 2.3613967895507812,
|
| 2016 |
+
"learning_rate": 1.4697236919459142e-05,
|
| 2017 |
+
"loss": 1.5444,
|
| 2018 |
+
"step": 2780
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 7.376074025115664,
|
| 2022 |
+
"grad_norm": 2.8878893852233887,
|
| 2023 |
+
"learning_rate": 1.455026455026455e-05,
|
| 2024 |
+
"loss": 1.5078,
|
| 2025 |
+
"step": 2790
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 7.402511566424322,
|
| 2029 |
+
"grad_norm": 2.3623602390289307,
|
| 2030 |
+
"learning_rate": 1.440329218106996e-05,
|
| 2031 |
+
"loss": 1.5497,
|
| 2032 |
+
"step": 2800
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 7.428949107732981,
|
| 2036 |
+
"grad_norm": 2.708502769470215,
|
| 2037 |
+
"learning_rate": 1.4256319811875369e-05,
|
| 2038 |
+
"loss": 1.5249,
|
| 2039 |
+
"step": 2810
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 7.455386649041639,
|
| 2043 |
+
"grad_norm": 2.521937847137451,
|
| 2044 |
+
"learning_rate": 1.4109347442680776e-05,
|
| 2045 |
+
"loss": 1.5556,
|
| 2046 |
+
"step": 2820
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 7.481824190350298,
|
| 2050 |
+
"grad_norm": 2.3380672931671143,
|
| 2051 |
+
"learning_rate": 1.3962375073486187e-05,
|
| 2052 |
+
"loss": 1.5295,
|
| 2053 |
+
"step": 2830
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 7.508261731658956,
|
| 2057 |
+
"grad_norm": 2.609221935272217,
|
| 2058 |
+
"learning_rate": 1.3815402704291594e-05,
|
| 2059 |
+
"loss": 1.5173,
|
| 2060 |
+
"step": 2840
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 7.534699272967614,
|
| 2064 |
+
"grad_norm": 2.692725658416748,
|
| 2065 |
+
"learning_rate": 1.3668430335097001e-05,
|
| 2066 |
+
"loss": 1.5256,
|
| 2067 |
+
"step": 2850
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 7.5611368142762725,
|
| 2071 |
+
"grad_norm": 2.8412363529205322,
|
| 2072 |
+
"learning_rate": 1.3521457965902412e-05,
|
| 2073 |
+
"loss": 1.5213,
|
| 2074 |
+
"step": 2860
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 7.58757435558493,
|
| 2078 |
+
"grad_norm": 2.349609375,
|
| 2079 |
+
"learning_rate": 1.3374485596707819e-05,
|
| 2080 |
+
"loss": 1.5245,
|
| 2081 |
+
"step": 2870
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 7.614011896893589,
|
| 2085 |
+
"grad_norm": 2.3491156101226807,
|
| 2086 |
+
"learning_rate": 1.3227513227513228e-05,
|
| 2087 |
+
"loss": 1.5397,
|
| 2088 |
+
"step": 2880
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 7.640449438202247,
|
| 2092 |
+
"grad_norm": 2.2094242572784424,
|
| 2093 |
+
"learning_rate": 1.3080540858318637e-05,
|
| 2094 |
+
"loss": 1.5225,
|
| 2095 |
+
"step": 2890
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.666886979510905,
|
| 2099 |
+
"grad_norm": 2.7071151733398438,
|
| 2100 |
+
"learning_rate": 1.2933568489124046e-05,
|
| 2101 |
+
"loss": 1.5397,
|
| 2102 |
+
"step": 2900
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 7.693324520819564,
|
| 2106 |
+
"grad_norm": 2.4890148639678955,
|
| 2107 |
+
"learning_rate": 1.2786596119929453e-05,
|
| 2108 |
+
"loss": 1.5188,
|
| 2109 |
+
"step": 2910
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 7.719762062128222,
|
| 2113 |
+
"grad_norm": 2.607391595840454,
|
| 2114 |
+
"learning_rate": 1.2639623750734864e-05,
|
| 2115 |
+
"loss": 1.5243,
|
| 2116 |
+
"step": 2920
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 7.746199603436881,
|
| 2120 |
+
"grad_norm": 2.255218505859375,
|
| 2121 |
+
"learning_rate": 1.2492651381540271e-05,
|
| 2122 |
+
"loss": 1.4923,
|
| 2123 |
+
"step": 2930
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 7.772637144745539,
|
| 2127 |
+
"grad_norm": 2.095276355743408,
|
| 2128 |
+
"learning_rate": 1.2345679012345678e-05,
|
| 2129 |
+
"loss": 1.5154,
|
| 2130 |
+
"step": 2940
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 7.799074686054197,
|
| 2134 |
+
"grad_norm": 2.3163740634918213,
|
| 2135 |
+
"learning_rate": 1.2198706643151087e-05,
|
| 2136 |
+
"loss": 1.5273,
|
| 2137 |
+
"step": 2950
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 7.8255122273628555,
|
| 2141 |
+
"grad_norm": 2.356632709503174,
|
| 2142 |
+
"learning_rate": 1.2051734273956498e-05,
|
| 2143 |
+
"loss": 1.5434,
|
| 2144 |
+
"step": 2960
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 7.851949768671513,
|
| 2148 |
+
"grad_norm": 2.6374380588531494,
|
| 2149 |
+
"learning_rate": 1.1904761904761905e-05,
|
| 2150 |
+
"loss": 1.5443,
|
| 2151 |
+
"step": 2970
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 7.878387309980172,
|
| 2155 |
+
"grad_norm": 2.324978828430176,
|
| 2156 |
+
"learning_rate": 1.1757789535567314e-05,
|
| 2157 |
+
"loss": 1.5277,
|
| 2158 |
+
"step": 2980
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 7.90482485128883,
|
| 2162 |
+
"grad_norm": 2.326624870300293,
|
| 2163 |
+
"learning_rate": 1.1610817166372723e-05,
|
| 2164 |
+
"loss": 1.5392,
|
| 2165 |
+
"step": 2990
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 7.931262392597488,
|
| 2169 |
+
"grad_norm": 2.3055641651153564,
|
| 2170 |
+
"learning_rate": 1.146384479717813e-05,
|
| 2171 |
+
"loss": 1.5304,
|
| 2172 |
+
"step": 3000
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 7.957699933906147,
|
| 2176 |
+
"grad_norm": 2.652165174484253,
|
| 2177 |
+
"learning_rate": 1.1316872427983539e-05,
|
| 2178 |
+
"loss": 1.4925,
|
| 2179 |
+
"step": 3010
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 7.984137475214805,
|
| 2183 |
+
"grad_norm": 2.6190438270568848,
|
| 2184 |
+
"learning_rate": 1.1169900058788948e-05,
|
| 2185 |
+
"loss": 1.5133,
|
| 2186 |
+
"step": 3020
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 8.0,
|
| 2190 |
+
"eval_accuracy": 0.3199148785245611,
|
| 2191 |
+
"eval_loss": 1.6810516119003296,
|
| 2192 |
+
"eval_runtime": 655.8832,
|
| 2193 |
+
"eval_samples_per_second": 42.988,
|
| 2194 |
+
"eval_steps_per_second": 0.672,
|
| 2195 |
+
"step": 3026
|
| 2196 |
+
},
|
| 2197 |
+
{
|
| 2198 |
+
"epoch": 8.010575016523463,
|
| 2199 |
+
"grad_norm": 2.4034554958343506,
|
| 2200 |
+
"learning_rate": 1.1022927689594357e-05,
|
| 2201 |
+
"loss": 1.4946,
|
| 2202 |
+
"step": 3030
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 8.03701255783212,
|
| 2206 |
+
"grad_norm": 2.3963074684143066,
|
| 2207 |
+
"learning_rate": 1.0875955320399766e-05,
|
| 2208 |
+
"loss": 1.4916,
|
| 2209 |
+
"step": 3040
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 8.06345009914078,
|
| 2213 |
+
"grad_norm": 2.228837728500366,
|
| 2214 |
+
"learning_rate": 1.0728982951205175e-05,
|
| 2215 |
+
"loss": 1.4966,
|
| 2216 |
+
"step": 3050
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 8.089887640449438,
|
| 2220 |
+
"grad_norm": 2.434828042984009,
|
| 2221 |
+
"learning_rate": 1.0582010582010582e-05,
|
| 2222 |
+
"loss": 1.476,
|
| 2223 |
+
"step": 3060
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 8.116325181758096,
|
| 2227 |
+
"grad_norm": 3.0598857402801514,
|
| 2228 |
+
"learning_rate": 1.0435038212815991e-05,
|
| 2229 |
+
"loss": 1.509,
|
| 2230 |
+
"step": 3070
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 8.142762723066754,
|
| 2234 |
+
"grad_norm": 2.3611867427825928,
|
| 2235 |
+
"learning_rate": 1.02880658436214e-05,
|
| 2236 |
+
"loss": 1.5034,
|
| 2237 |
+
"step": 3080
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 8.169200264375412,
|
| 2241 |
+
"grad_norm": 2.5760085582733154,
|
| 2242 |
+
"learning_rate": 1.0141093474426807e-05,
|
| 2243 |
+
"loss": 1.4722,
|
| 2244 |
+
"step": 3090
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 8.195637805684072,
|
| 2248 |
+
"grad_norm": 2.364185333251953,
|
| 2249 |
+
"learning_rate": 9.994121105232216e-06,
|
| 2250 |
+
"loss": 1.503,
|
| 2251 |
+
"step": 3100
|
| 2252 |
+
},
|
| 2253 |
+
{
|
| 2254 |
+
"epoch": 8.22207534699273,
|
| 2255 |
+
"grad_norm": 2.9226784706115723,
|
| 2256 |
+
"learning_rate": 9.847148736037625e-06,
|
| 2257 |
+
"loss": 1.5095,
|
| 2258 |
+
"step": 3110
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 8.248512888301388,
|
| 2262 |
+
"grad_norm": 2.599740505218506,
|
| 2263 |
+
"learning_rate": 9.700176366843034e-06,
|
| 2264 |
+
"loss": 1.5113,
|
| 2265 |
+
"step": 3120
|
| 2266 |
+
},
|
| 2267 |
+
{
|
| 2268 |
+
"epoch": 8.274950429610046,
|
| 2269 |
+
"grad_norm": 2.4345662593841553,
|
| 2270 |
+
"learning_rate": 9.553203997648443e-06,
|
| 2271 |
+
"loss": 1.4932,
|
| 2272 |
+
"step": 3130
|
| 2273 |
+
},
|
| 2274 |
+
{
|
| 2275 |
+
"epoch": 8.301387970918704,
|
| 2276 |
+
"grad_norm": 2.821202516555786,
|
| 2277 |
+
"learning_rate": 9.406231628453852e-06,
|
| 2278 |
+
"loss": 1.4873,
|
| 2279 |
+
"step": 3140
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 8.327825512227363,
|
| 2283 |
+
"grad_norm": 2.3733325004577637,
|
| 2284 |
+
"learning_rate": 9.259259259259259e-06,
|
| 2285 |
+
"loss": 1.4714,
|
| 2286 |
+
"step": 3150
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 8.354263053536021,
|
| 2290 |
+
"grad_norm": 2.5024616718292236,
|
| 2291 |
+
"learning_rate": 9.112286890064668e-06,
|
| 2292 |
+
"loss": 1.4943,
|
| 2293 |
+
"step": 3160
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 8.38070059484468,
|
| 2297 |
+
"grad_norm": 2.6110079288482666,
|
| 2298 |
+
"learning_rate": 8.965314520870077e-06,
|
| 2299 |
+
"loss": 1.4889,
|
| 2300 |
+
"step": 3170
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 8.407138136153337,
|
| 2304 |
+
"grad_norm": 2.432016611099243,
|
| 2305 |
+
"learning_rate": 8.818342151675484e-06,
|
| 2306 |
+
"loss": 1.498,
|
| 2307 |
+
"step": 3180
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 8.433575677461995,
|
| 2311 |
+
"grad_norm": 2.565856695175171,
|
| 2312 |
+
"learning_rate": 8.671369782480893e-06,
|
| 2313 |
+
"loss": 1.5063,
|
| 2314 |
+
"step": 3190
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 8.460013218770655,
|
| 2318 |
+
"grad_norm": 2.3665668964385986,
|
| 2319 |
+
"learning_rate": 8.524397413286304e-06,
|
| 2320 |
+
"loss": 1.4984,
|
| 2321 |
+
"step": 3200
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 8.486450760079313,
|
| 2325 |
+
"grad_norm": 2.3754770755767822,
|
| 2326 |
+
"learning_rate": 8.377425044091711e-06,
|
| 2327 |
+
"loss": 1.4884,
|
| 2328 |
+
"step": 3210
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 8.51288830138797,
|
| 2332 |
+
"grad_norm": 2.3668394088745117,
|
| 2333 |
+
"learning_rate": 8.23045267489712e-06,
|
| 2334 |
+
"loss": 1.5134,
|
| 2335 |
+
"step": 3220
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 8.539325842696629,
|
| 2339 |
+
"grad_norm": 2.5414023399353027,
|
| 2340 |
+
"learning_rate": 8.083480305702529e-06,
|
| 2341 |
+
"loss": 1.4838,
|
| 2342 |
+
"step": 3230
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 8.565763384005287,
|
| 2346 |
+
"grad_norm": 3.5699350833892822,
|
| 2347 |
+
"learning_rate": 7.936507936507936e-06,
|
| 2348 |
+
"loss": 1.5143,
|
| 2349 |
+
"step": 3240
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 8.592200925313946,
|
| 2353 |
+
"grad_norm": 2.8063623905181885,
|
| 2354 |
+
"learning_rate": 7.789535567313345e-06,
|
| 2355 |
+
"loss": 1.496,
|
| 2356 |
+
"step": 3250
|
| 2357 |
+
},
|
| 2358 |
+
{
|
| 2359 |
+
"epoch": 8.618638466622604,
|
| 2360 |
+
"grad_norm": 2.588224172592163,
|
| 2361 |
+
"learning_rate": 7.642563198118754e-06,
|
| 2362 |
+
"loss": 1.4791,
|
| 2363 |
+
"step": 3260
|
| 2364 |
+
},
|
| 2365 |
+
{
|
| 2366 |
+
"epoch": 8.645076007931262,
|
| 2367 |
+
"grad_norm": 2.4756646156311035,
|
| 2368 |
+
"learning_rate": 7.495590828924162e-06,
|
| 2369 |
+
"loss": 1.5166,
|
| 2370 |
+
"step": 3270
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 8.67151354923992,
|
| 2374 |
+
"grad_norm": 2.575335741043091,
|
| 2375 |
+
"learning_rate": 7.348618459729571e-06,
|
| 2376 |
+
"loss": 1.4806,
|
| 2377 |
+
"step": 3280
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 8.697951090548578,
|
| 2381 |
+
"grad_norm": 2.5106089115142822,
|
| 2382 |
+
"learning_rate": 7.20164609053498e-06,
|
| 2383 |
+
"loss": 1.517,
|
| 2384 |
+
"step": 3290
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 8.724388631857238,
|
| 2388 |
+
"grad_norm": 2.419461727142334,
|
| 2389 |
+
"learning_rate": 7.054673721340388e-06,
|
| 2390 |
+
"loss": 1.5046,
|
| 2391 |
+
"step": 3300
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 8.750826173165896,
|
| 2395 |
+
"grad_norm": 2.6011338233947754,
|
| 2396 |
+
"learning_rate": 6.907701352145797e-06,
|
| 2397 |
+
"loss": 1.51,
|
| 2398 |
+
"step": 3310
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 8.777263714474554,
|
| 2402 |
+
"grad_norm": 3.790093183517456,
|
| 2403 |
+
"learning_rate": 6.760728982951206e-06,
|
| 2404 |
+
"loss": 1.4779,
|
| 2405 |
+
"step": 3320
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 8.803701255783212,
|
| 2409 |
+
"grad_norm": 2.2048144340515137,
|
| 2410 |
+
"learning_rate": 6.613756613756614e-06,
|
| 2411 |
+
"loss": 1.5019,
|
| 2412 |
+
"step": 3330
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 8.83013879709187,
|
| 2416 |
+
"grad_norm": 2.383060932159424,
|
| 2417 |
+
"learning_rate": 6.466784244562023e-06,
|
| 2418 |
+
"loss": 1.4841,
|
| 2419 |
+
"step": 3340
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 8.85657633840053,
|
| 2423 |
+
"grad_norm": 2.647857666015625,
|
| 2424 |
+
"learning_rate": 6.319811875367432e-06,
|
| 2425 |
+
"loss": 1.486,
|
| 2426 |
+
"step": 3350
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 8.883013879709187,
|
| 2430 |
+
"grad_norm": 2.314814805984497,
|
| 2431 |
+
"learning_rate": 6.172839506172839e-06,
|
| 2432 |
+
"loss": 1.4847,
|
| 2433 |
+
"step": 3360
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 8.909451421017845,
|
| 2437 |
+
"grad_norm": 2.521620988845825,
|
| 2438 |
+
"learning_rate": 6.025867136978249e-06,
|
| 2439 |
+
"loss": 1.4983,
|
| 2440 |
+
"step": 3370
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 8.935888962326503,
|
| 2444 |
+
"grad_norm": 2.53841495513916,
|
| 2445 |
+
"learning_rate": 5.878894767783657e-06,
|
| 2446 |
+
"loss": 1.5085,
|
| 2447 |
+
"step": 3380
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 8.962326503635161,
|
| 2451 |
+
"grad_norm": 3.0474183559417725,
|
| 2452 |
+
"learning_rate": 5.731922398589065e-06,
|
| 2453 |
+
"loss": 1.5159,
|
| 2454 |
+
"step": 3390
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 8.98876404494382,
|
| 2458 |
+
"grad_norm": 2.3993120193481445,
|
| 2459 |
+
"learning_rate": 5.584950029394474e-06,
|
| 2460 |
+
"loss": 1.5001,
|
| 2461 |
+
"step": 3400
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 8.999339061467284,
|
| 2465 |
+
"eval_accuracy": 0.32094342968611456,
|
| 2466 |
+
"eval_loss": 1.6820588111877441,
|
| 2467 |
+
"eval_runtime": 644.8259,
|
| 2468 |
+
"eval_samples_per_second": 43.725,
|
| 2469 |
+
"eval_steps_per_second": 0.684,
|
| 2470 |
+
"step": 3404
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 9.015201586252479,
|
| 2474 |
+
"grad_norm": 2.3832859992980957,
|
| 2475 |
+
"learning_rate": 5.437977660199883e-06,
|
| 2476 |
+
"loss": 1.4755,
|
| 2477 |
+
"step": 3410
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 9.041639127561137,
|
| 2481 |
+
"grad_norm": 2.3615214824676514,
|
| 2482 |
+
"learning_rate": 5.291005291005291e-06,
|
| 2483 |
+
"loss": 1.4527,
|
| 2484 |
+
"step": 3420
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 9.068076668869795,
|
| 2488 |
+
"grad_norm": 2.408276319503784,
|
| 2489 |
+
"learning_rate": 5.1440329218107e-06,
|
| 2490 |
+
"loss": 1.5007,
|
| 2491 |
+
"step": 3430
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 9.094514210178453,
|
| 2495 |
+
"grad_norm": 3.026998281478882,
|
| 2496 |
+
"learning_rate": 4.997060552616108e-06,
|
| 2497 |
+
"loss": 1.4759,
|
| 2498 |
+
"step": 3440
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 9.120951751487112,
|
| 2502 |
+
"grad_norm": 2.7928555011749268,
|
| 2503 |
+
"learning_rate": 4.850088183421517e-06,
|
| 2504 |
+
"loss": 1.5078,
|
| 2505 |
+
"step": 3450
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 9.14738929279577,
|
| 2509 |
+
"grad_norm": 2.3969271183013916,
|
| 2510 |
+
"learning_rate": 4.703115814226926e-06,
|
| 2511 |
+
"loss": 1.4786,
|
| 2512 |
+
"step": 3460
|
| 2513 |
+
},
|
| 2514 |
+
{
|
| 2515 |
+
"epoch": 9.173826834104428,
|
| 2516 |
+
"grad_norm": 2.5589447021484375,
|
| 2517 |
+
"learning_rate": 4.556143445032334e-06,
|
| 2518 |
+
"loss": 1.4857,
|
| 2519 |
+
"step": 3470
|
| 2520 |
+
},
|
| 2521 |
+
{
|
| 2522 |
+
"epoch": 9.200264375413086,
|
| 2523 |
+
"grad_norm": 2.3872599601745605,
|
| 2524 |
+
"learning_rate": 4.409171075837742e-06,
|
| 2525 |
+
"loss": 1.5131,
|
| 2526 |
+
"step": 3480
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"epoch": 9.226701916721744,
|
| 2530 |
+
"grad_norm": 3.470109462738037,
|
| 2531 |
+
"learning_rate": 4.262198706643152e-06,
|
| 2532 |
+
"loss": 1.4878,
|
| 2533 |
+
"step": 3490
|
| 2534 |
+
},
|
| 2535 |
+
{
|
| 2536 |
+
"epoch": 9.253139458030404,
|
| 2537 |
+
"grad_norm": 2.545947790145874,
|
| 2538 |
+
"learning_rate": 4.11522633744856e-06,
|
| 2539 |
+
"loss": 1.4973,
|
| 2540 |
+
"step": 3500
|
| 2541 |
+
},
|
| 2542 |
+
{
|
| 2543 |
+
"epoch": 9.279576999339062,
|
| 2544 |
+
"grad_norm": 2.713286876678467,
|
| 2545 |
+
"learning_rate": 3.968253968253968e-06,
|
| 2546 |
+
"loss": 1.4771,
|
| 2547 |
+
"step": 3510
|
| 2548 |
+
},
|
| 2549 |
+
{
|
| 2550 |
+
"epoch": 9.30601454064772,
|
| 2551 |
+
"grad_norm": 2.5866544246673584,
|
| 2552 |
+
"learning_rate": 3.821281599059377e-06,
|
| 2553 |
+
"loss": 1.4987,
|
| 2554 |
+
"step": 3520
|
| 2555 |
+
},
|
| 2556 |
+
{
|
| 2557 |
+
"epoch": 9.332452081956378,
|
| 2558 |
+
"grad_norm": 2.608093023300171,
|
| 2559 |
+
"learning_rate": 3.6743092298647855e-06,
|
| 2560 |
+
"loss": 1.4855,
|
| 2561 |
+
"step": 3530
|
| 2562 |
+
},
|
| 2563 |
+
{
|
| 2564 |
+
"epoch": 9.358889623265036,
|
| 2565 |
+
"grad_norm": 2.363469362258911,
|
| 2566 |
+
"learning_rate": 3.527336860670194e-06,
|
| 2567 |
+
"loss": 1.4903,
|
| 2568 |
+
"step": 3540
|
| 2569 |
+
},
|
| 2570 |
+
{
|
| 2571 |
+
"epoch": 9.385327164573695,
|
| 2572 |
+
"grad_norm": 2.4162986278533936,
|
| 2573 |
+
"learning_rate": 3.380364491475603e-06,
|
| 2574 |
+
"loss": 1.4944,
|
| 2575 |
+
"step": 3550
|
| 2576 |
+
},
|
| 2577 |
+
{
|
| 2578 |
+
"epoch": 9.411764705882353,
|
| 2579 |
+
"grad_norm": 2.531153440475464,
|
| 2580 |
+
"learning_rate": 3.2333921222810115e-06,
|
| 2581 |
+
"loss": 1.4974,
|
| 2582 |
+
"step": 3560
|
| 2583 |
+
},
|
| 2584 |
+
{
|
| 2585 |
+
"epoch": 9.438202247191011,
|
| 2586 |
+
"grad_norm": 2.674409866333008,
|
| 2587 |
+
"learning_rate": 3.0864197530864196e-06,
|
| 2588 |
+
"loss": 1.4544,
|
| 2589 |
+
"step": 3570
|
| 2590 |
+
},
|
| 2591 |
+
{
|
| 2592 |
+
"epoch": 9.46463978849967,
|
| 2593 |
+
"grad_norm": 2.8534953594207764,
|
| 2594 |
+
"learning_rate": 2.9394473838918285e-06,
|
| 2595 |
+
"loss": 1.5179,
|
| 2596 |
+
"step": 3580
|
| 2597 |
+
},
|
| 2598 |
+
{
|
| 2599 |
+
"epoch": 9.491077329808327,
|
| 2600 |
+
"grad_norm": 2.373920202255249,
|
| 2601 |
+
"learning_rate": 2.792475014697237e-06,
|
| 2602 |
+
"loss": 1.487,
|
| 2603 |
+
"step": 3590
|
| 2604 |
+
},
|
| 2605 |
+
{
|
| 2606 |
+
"epoch": 9.517514871116987,
|
| 2607 |
+
"grad_norm": 2.327130079269409,
|
| 2608 |
+
"learning_rate": 2.6455026455026455e-06,
|
| 2609 |
+
"loss": 1.464,
|
| 2610 |
+
"step": 3600
|
| 2611 |
+
},
|
| 2612 |
+
{
|
| 2613 |
+
"epoch": 9.543952412425645,
|
| 2614 |
+
"grad_norm": 2.580538749694824,
|
| 2615 |
+
"learning_rate": 2.498530276308054e-06,
|
| 2616 |
+
"loss": 1.474,
|
| 2617 |
+
"step": 3610
|
| 2618 |
+
},
|
| 2619 |
+
{
|
| 2620 |
+
"epoch": 9.570389953734303,
|
| 2621 |
+
"grad_norm": 2.5735127925872803,
|
| 2622 |
+
"learning_rate": 2.351557907113463e-06,
|
| 2623 |
+
"loss": 1.4443,
|
| 2624 |
+
"step": 3620
|
| 2625 |
+
},
|
| 2626 |
+
{
|
| 2627 |
+
"epoch": 9.59682749504296,
|
| 2628 |
+
"grad_norm": 2.6845741271972656,
|
| 2629 |
+
"learning_rate": 2.204585537918871e-06,
|
| 2630 |
+
"loss": 1.4728,
|
| 2631 |
+
"step": 3630
|
| 2632 |
+
},
|
| 2633 |
+
{
|
| 2634 |
+
"epoch": 9.623265036351619,
|
| 2635 |
+
"grad_norm": 2.7224180698394775,
|
| 2636 |
+
"learning_rate": 2.05761316872428e-06,
|
| 2637 |
+
"loss": 1.4737,
|
| 2638 |
+
"step": 3640
|
| 2639 |
+
},
|
| 2640 |
+
{
|
| 2641 |
+
"epoch": 9.649702577660278,
|
| 2642 |
+
"grad_norm": 2.6456120014190674,
|
| 2643 |
+
"learning_rate": 1.9106407995296885e-06,
|
| 2644 |
+
"loss": 1.4572,
|
| 2645 |
+
"step": 3650
|
| 2646 |
+
},
|
| 2647 |
+
{
|
| 2648 |
+
"epoch": 9.676140118968936,
|
| 2649 |
+
"grad_norm": 2.3827290534973145,
|
| 2650 |
+
"learning_rate": 1.763668430335097e-06,
|
| 2651 |
+
"loss": 1.4906,
|
| 2652 |
+
"step": 3660
|
| 2653 |
+
},
|
| 2654 |
+
{
|
| 2655 |
+
"epoch": 9.702577660277594,
|
| 2656 |
+
"grad_norm": 2.7378439903259277,
|
| 2657 |
+
"learning_rate": 1.6166960611405057e-06,
|
| 2658 |
+
"loss": 1.4844,
|
| 2659 |
+
"step": 3670
|
| 2660 |
+
},
|
| 2661 |
+
{
|
| 2662 |
+
"epoch": 9.729015201586252,
|
| 2663 |
+
"grad_norm": 2.330042839050293,
|
| 2664 |
+
"learning_rate": 1.4697236919459143e-06,
|
| 2665 |
+
"loss": 1.4552,
|
| 2666 |
+
"step": 3680
|
| 2667 |
+
},
|
| 2668 |
+
{
|
| 2669 |
+
"epoch": 9.75545274289491,
|
| 2670 |
+
"grad_norm": 2.9327447414398193,
|
| 2671 |
+
"learning_rate": 1.3227513227513228e-06,
|
| 2672 |
+
"loss": 1.4699,
|
| 2673 |
+
"step": 3690
|
| 2674 |
+
},
|
| 2675 |
+
{
|
| 2676 |
+
"epoch": 9.78189028420357,
|
| 2677 |
+
"grad_norm": 3.050356149673462,
|
| 2678 |
+
"learning_rate": 1.1757789535567315e-06,
|
| 2679 |
+
"loss": 1.4831,
|
| 2680 |
+
"step": 3700
|
| 2681 |
+
},
|
| 2682 |
+
{
|
| 2683 |
+
"epoch": 9.808327825512228,
|
| 2684 |
+
"grad_norm": 2.769808053970337,
|
| 2685 |
+
"learning_rate": 1.02880658436214e-06,
|
| 2686 |
+
"loss": 1.4886,
|
| 2687 |
+
"step": 3710
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"epoch": 9.834765366820886,
|
| 2691 |
+
"grad_norm": 2.5177111625671387,
|
| 2692 |
+
"learning_rate": 8.818342151675485e-07,
|
| 2693 |
+
"loss": 1.5074,
|
| 2694 |
+
"step": 3720
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 9.861202908129544,
|
| 2698 |
+
"grad_norm": 2.4207041263580322,
|
| 2699 |
+
"learning_rate": 7.348618459729571e-07,
|
| 2700 |
+
"loss": 1.4792,
|
| 2701 |
+
"step": 3730
|
| 2702 |
+
},
|
| 2703 |
+
{
|
| 2704 |
+
"epoch": 9.887640449438202,
|
| 2705 |
+
"grad_norm": 2.842231512069702,
|
| 2706 |
+
"learning_rate": 5.878894767783657e-07,
|
| 2707 |
+
"loss": 1.4696,
|
| 2708 |
+
"step": 3740
|
| 2709 |
+
},
|
| 2710 |
+
{
|
| 2711 |
+
"epoch": 9.914077990746861,
|
| 2712 |
+
"grad_norm": 2.7768394947052,
|
| 2713 |
+
"learning_rate": 4.4091710758377425e-07,
|
| 2714 |
+
"loss": 1.4642,
|
| 2715 |
+
"step": 3750
|
| 2716 |
+
},
|
| 2717 |
+
{
|
| 2718 |
+
"epoch": 9.94051553205552,
|
| 2719 |
+
"grad_norm": 2.7544069290161133,
|
| 2720 |
+
"learning_rate": 2.9394473838918287e-07,
|
| 2721 |
+
"loss": 1.4824,
|
| 2722 |
+
"step": 3760
|
| 2723 |
+
},
|
| 2724 |
+
{
|
| 2725 |
+
"epoch": 9.966953073364177,
|
| 2726 |
+
"grad_norm": 2.5182807445526123,
|
| 2727 |
+
"learning_rate": 1.4697236919459144e-07,
|
| 2728 |
+
"loss": 1.4768,
|
| 2729 |
+
"step": 3770
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 9.993390614672835,
|
| 2733 |
+
"grad_norm": 2.31466007232666,
|
| 2734 |
+
"learning_rate": 0.0,
|
| 2735 |
+
"loss": 1.4303,
|
| 2736 |
+
"step": 3780
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 9.993390614672835,
|
| 2740 |
+
"eval_accuracy": 0.3192410001773364,
|
| 2741 |
+
"eval_loss": 1.6835154294967651,
|
| 2742 |
+
"eval_runtime": 626.4861,
|
| 2743 |
+
"eval_samples_per_second": 45.005,
|
| 2744 |
+
"eval_steps_per_second": 0.704,
|
| 2745 |
+
"step": 3780
|
| 2746 |
+
},
|
| 2747 |
+
{
|
| 2748 |
+
"epoch": 9.993390614672835,
|
| 2749 |
+
"step": 3780,
|
| 2750 |
+
"total_flos": 7.511248143138256e+19,
|
| 2751 |
+
"train_loss": 1.6334815527396227,
|
| 2752 |
+
"train_runtime": 44633.0248,
|
| 2753 |
+
"train_samples_per_second": 21.683,
|
| 2754 |
+
"train_steps_per_second": 0.085
|
| 2755 |
+
}
|
| 2756 |
+
],
|
| 2757 |
+
"logging_steps": 10,
|
| 2758 |
+
"max_steps": 3780,
|
| 2759 |
+
"num_input_tokens_seen": 0,
|
| 2760 |
+
"num_train_epochs": 10,
|
| 2761 |
+
"save_steps": 500,
|
| 2762 |
+
"total_flos": 7.511248143138256e+19,
|
| 2763 |
+
"train_batch_size": 64,
|
| 2764 |
+
"trial_name": null,
|
| 2765 |
+
"trial_params": null
|
| 2766 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:20a69f58503433efa0516ad7a50800dcc65855d77ad90ea2cf82f5696584f14e
|
| 3 |
+
size 4603
|