Training in progress, step 1800
Browse files- checkpoint-1800/optimizer.pt +3 -0
- checkpoint-1800/pytorch_model.bin +3 -0
- checkpoint-1800/rng_state.pth +3 -0
- checkpoint-1800/scaler.pt +3 -0
- checkpoint-1800/scheduler.pt +3 -0
- checkpoint-1800/trainer_state.json +2248 -0
- checkpoint-1800/training_args.bin +3 -0
- pytorch_model.bin +1 -1
- runs/May07_17-47-06_bc858596195a/events.out.tfevents.1683481716.bc858596195a.2495.0 +2 -2
checkpoint-1800/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c14817f4e4c4c80a6c4012220f1592b7ce8036240b6858d00b220b3b60a1b43
|
| 3 |
+
size 25279045
|
checkpoint-1800/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0034d605ad777cf2436a73d44b7e2d05cc74ba8454a92397ea8256b4d3e8438
|
| 3 |
+
size 12653389
|
checkpoint-1800/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f57f579be2693f3524a2c78014053b1c91642cd6832f17e6e0b163d0989b94d
|
| 3 |
+
size 14511
|
checkpoint-1800/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e372dbdefa78360aa0c173dac32ed07b6adf2d79936f267acae43d837649895
|
| 3 |
+
size 557
|
checkpoint-1800/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf441c9e7b547c1b494b9fb26d72ac74645602d2ba1e3b444d48f49c4fc3271a
|
| 3 |
+
size 627
|
checkpoint-1800/trainer_state.json
ADDED
|
@@ -0,0 +1,2248 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.157345175743103,
|
| 3 |
+
"best_model_checkpoint": "outputs/checkpoint-1800",
|
| 4 |
+
"epoch": 0.576,
|
| 5 |
+
"global_step": 1800,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 1e-05,
|
| 13 |
+
"loss": 1.8184,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.0,
|
| 18 |
+
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 1.7858,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.0,
|
| 24 |
+
"learning_rate": 3e-05,
|
| 25 |
+
"loss": 1.7218,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4e-05,
|
| 31 |
+
"loss": 1.7218,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 1.7009,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 6e-05,
|
| 43 |
+
"loss": 1.6714,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01,
|
| 48 |
+
"learning_rate": 7e-05,
|
| 49 |
+
"loss": 1.6582,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.01,
|
| 54 |
+
"learning_rate": 8e-05,
|
| 55 |
+
"loss": 1.5659,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.01,
|
| 60 |
+
"learning_rate": 9e-05,
|
| 61 |
+
"loss": 1.5343,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
+
"learning_rate": 0.0001,
|
| 67 |
+
"loss": 1.5252,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.02,
|
| 72 |
+
"learning_rate": 0.00011000000000000002,
|
| 73 |
+
"loss": 1.414,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.02,
|
| 78 |
+
"learning_rate": 0.00012,
|
| 79 |
+
"loss": 1.3919,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02,
|
| 84 |
+
"learning_rate": 0.00013000000000000002,
|
| 85 |
+
"loss": 1.352,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.02,
|
| 90 |
+
"learning_rate": 0.00014,
|
| 91 |
+
"loss": 1.3762,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.02,
|
| 96 |
+
"learning_rate": 0.00015000000000000001,
|
| 97 |
+
"loss": 1.3238,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
+
"learning_rate": 0.00016,
|
| 103 |
+
"loss": 1.3306,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.03,
|
| 108 |
+
"learning_rate": 0.00017,
|
| 109 |
+
"loss": 1.3423,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.03,
|
| 114 |
+
"learning_rate": 0.00018,
|
| 115 |
+
"loss": 1.3199,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.03,
|
| 120 |
+
"learning_rate": 0.00019,
|
| 121 |
+
"loss": 1.3506,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.03,
|
| 126 |
+
"learning_rate": 0.0002,
|
| 127 |
+
"loss": 1.3315,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.03,
|
| 132 |
+
"learning_rate": 0.0001998921832884097,
|
| 133 |
+
"loss": 1.3142,
|
| 134 |
+
"step": 105
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.04,
|
| 138 |
+
"learning_rate": 0.00019978436657681943,
|
| 139 |
+
"loss": 1.2994,
|
| 140 |
+
"step": 110
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.04,
|
| 144 |
+
"learning_rate": 0.00019967654986522912,
|
| 145 |
+
"loss": 1.2972,
|
| 146 |
+
"step": 115
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.04,
|
| 150 |
+
"learning_rate": 0.00019956873315363883,
|
| 151 |
+
"loss": 1.3135,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.04,
|
| 156 |
+
"learning_rate": 0.00019946091644204851,
|
| 157 |
+
"loss": 1.2903,
|
| 158 |
+
"step": 125
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.04,
|
| 162 |
+
"learning_rate": 0.00019935309973045823,
|
| 163 |
+
"loss": 1.2783,
|
| 164 |
+
"step": 130
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.04,
|
| 168 |
+
"learning_rate": 0.00019924528301886794,
|
| 169 |
+
"loss": 1.3044,
|
| 170 |
+
"step": 135
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.04,
|
| 174 |
+
"learning_rate": 0.00019913746630727762,
|
| 175 |
+
"loss": 1.2947,
|
| 176 |
+
"step": 140
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.05,
|
| 180 |
+
"learning_rate": 0.00019902964959568736,
|
| 181 |
+
"loss": 1.2759,
|
| 182 |
+
"step": 145
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.05,
|
| 186 |
+
"learning_rate": 0.00019892183288409705,
|
| 187 |
+
"loss": 1.2746,
|
| 188 |
+
"step": 150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.05,
|
| 192 |
+
"learning_rate": 0.00019881401617250676,
|
| 193 |
+
"loss": 1.2869,
|
| 194 |
+
"step": 155
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.05,
|
| 198 |
+
"learning_rate": 0.00019870619946091644,
|
| 199 |
+
"loss": 1.26,
|
| 200 |
+
"step": 160
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.05,
|
| 204 |
+
"learning_rate": 0.00019859838274932616,
|
| 205 |
+
"loss": 1.258,
|
| 206 |
+
"step": 165
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.05,
|
| 210 |
+
"learning_rate": 0.00019849056603773587,
|
| 211 |
+
"loss": 1.2649,
|
| 212 |
+
"step": 170
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.06,
|
| 216 |
+
"learning_rate": 0.00019838274932614555,
|
| 217 |
+
"loss": 1.2549,
|
| 218 |
+
"step": 175
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.06,
|
| 222 |
+
"learning_rate": 0.00019827493261455526,
|
| 223 |
+
"loss": 1.2859,
|
| 224 |
+
"step": 180
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.06,
|
| 228 |
+
"learning_rate": 0.00019816711590296498,
|
| 229 |
+
"loss": 1.2934,
|
| 230 |
+
"step": 185
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.06,
|
| 234 |
+
"learning_rate": 0.00019805929919137466,
|
| 235 |
+
"loss": 1.3011,
|
| 236 |
+
"step": 190
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.06,
|
| 240 |
+
"learning_rate": 0.00019795148247978437,
|
| 241 |
+
"loss": 1.2304,
|
| 242 |
+
"step": 195
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.06,
|
| 246 |
+
"learning_rate": 0.00019784366576819408,
|
| 247 |
+
"loss": 1.2849,
|
| 248 |
+
"step": 200
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.06,
|
| 252 |
+
"eval_loss": 1.2345943450927734,
|
| 253 |
+
"eval_runtime": 710.2534,
|
| 254 |
+
"eval_samples_per_second": 7.04,
|
| 255 |
+
"eval_steps_per_second": 0.88,
|
| 256 |
+
"step": 200
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.07,
|
| 260 |
+
"learning_rate": 0.0001977358490566038,
|
| 261 |
+
"loss": 1.2726,
|
| 262 |
+
"step": 205
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.07,
|
| 266 |
+
"learning_rate": 0.00019762803234501348,
|
| 267 |
+
"loss": 1.2996,
|
| 268 |
+
"step": 210
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.07,
|
| 272 |
+
"learning_rate": 0.0001975202156334232,
|
| 273 |
+
"loss": 1.2456,
|
| 274 |
+
"step": 215
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.07,
|
| 278 |
+
"learning_rate": 0.0001974123989218329,
|
| 279 |
+
"loss": 1.2794,
|
| 280 |
+
"step": 220
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.07,
|
| 284 |
+
"learning_rate": 0.0001973045822102426,
|
| 285 |
+
"loss": 1.2637,
|
| 286 |
+
"step": 225
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.07,
|
| 290 |
+
"learning_rate": 0.0001971967654986523,
|
| 291 |
+
"loss": 1.252,
|
| 292 |
+
"step": 230
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.08,
|
| 296 |
+
"learning_rate": 0.000197088948787062,
|
| 297 |
+
"loss": 1.2492,
|
| 298 |
+
"step": 235
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.08,
|
| 302 |
+
"learning_rate": 0.0001969811320754717,
|
| 303 |
+
"loss": 1.2972,
|
| 304 |
+
"step": 240
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.08,
|
| 308 |
+
"learning_rate": 0.0001968733153638814,
|
| 309 |
+
"loss": 1.3023,
|
| 310 |
+
"step": 245
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.08,
|
| 314 |
+
"learning_rate": 0.00019676549865229112,
|
| 315 |
+
"loss": 1.2682,
|
| 316 |
+
"step": 250
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.08,
|
| 320 |
+
"learning_rate": 0.00019665768194070083,
|
| 321 |
+
"loss": 1.2981,
|
| 322 |
+
"step": 255
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.08,
|
| 326 |
+
"learning_rate": 0.00019654986522911052,
|
| 327 |
+
"loss": 1.3084,
|
| 328 |
+
"step": 260
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.08,
|
| 332 |
+
"learning_rate": 0.00019644204851752023,
|
| 333 |
+
"loss": 1.2666,
|
| 334 |
+
"step": 265
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.09,
|
| 338 |
+
"learning_rate": 0.00019633423180592994,
|
| 339 |
+
"loss": 1.2623,
|
| 340 |
+
"step": 270
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.09,
|
| 344 |
+
"learning_rate": 0.00019622641509433963,
|
| 345 |
+
"loss": 1.2406,
|
| 346 |
+
"step": 275
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.09,
|
| 350 |
+
"learning_rate": 0.00019611859838274934,
|
| 351 |
+
"loss": 1.2623,
|
| 352 |
+
"step": 280
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.09,
|
| 356 |
+
"learning_rate": 0.00019601078167115905,
|
| 357 |
+
"loss": 1.2562,
|
| 358 |
+
"step": 285
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.09,
|
| 362 |
+
"learning_rate": 0.00019590296495956873,
|
| 363 |
+
"loss": 1.2166,
|
| 364 |
+
"step": 290
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.09,
|
| 368 |
+
"learning_rate": 0.00019579514824797845,
|
| 369 |
+
"loss": 1.2931,
|
| 370 |
+
"step": 295
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.1,
|
| 374 |
+
"learning_rate": 0.00019568733153638813,
|
| 375 |
+
"loss": 1.2354,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.1,
|
| 380 |
+
"learning_rate": 0.00019557951482479787,
|
| 381 |
+
"loss": 1.2246,
|
| 382 |
+
"step": 305
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.1,
|
| 386 |
+
"learning_rate": 0.00019547169811320755,
|
| 387 |
+
"loss": 1.2236,
|
| 388 |
+
"step": 310
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.1,
|
| 392 |
+
"learning_rate": 0.00019536388140161727,
|
| 393 |
+
"loss": 1.2442,
|
| 394 |
+
"step": 315
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.1,
|
| 398 |
+
"learning_rate": 0.00019525606469002698,
|
| 399 |
+
"loss": 1.2779,
|
| 400 |
+
"step": 320
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.1,
|
| 404 |
+
"learning_rate": 0.00019514824797843666,
|
| 405 |
+
"loss": 1.2285,
|
| 406 |
+
"step": 325
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.11,
|
| 410 |
+
"learning_rate": 0.00019504043126684637,
|
| 411 |
+
"loss": 1.2157,
|
| 412 |
+
"step": 330
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.11,
|
| 416 |
+
"learning_rate": 0.00019493261455525606,
|
| 417 |
+
"loss": 1.2405,
|
| 418 |
+
"step": 335
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.11,
|
| 422 |
+
"learning_rate": 0.0001948247978436658,
|
| 423 |
+
"loss": 1.2335,
|
| 424 |
+
"step": 340
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.11,
|
| 428 |
+
"learning_rate": 0.00019471698113207548,
|
| 429 |
+
"loss": 1.2401,
|
| 430 |
+
"step": 345
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.11,
|
| 434 |
+
"learning_rate": 0.00019460916442048517,
|
| 435 |
+
"loss": 1.2521,
|
| 436 |
+
"step": 350
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.11,
|
| 440 |
+
"learning_rate": 0.0001945013477088949,
|
| 441 |
+
"loss": 1.2386,
|
| 442 |
+
"step": 355
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.12,
|
| 446 |
+
"learning_rate": 0.0001943935309973046,
|
| 447 |
+
"loss": 1.2196,
|
| 448 |
+
"step": 360
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.12,
|
| 452 |
+
"learning_rate": 0.0001942857142857143,
|
| 453 |
+
"loss": 1.192,
|
| 454 |
+
"step": 365
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.12,
|
| 458 |
+
"learning_rate": 0.000194177897574124,
|
| 459 |
+
"loss": 1.1964,
|
| 460 |
+
"step": 370
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.12,
|
| 464 |
+
"learning_rate": 0.0001940700808625337,
|
| 465 |
+
"loss": 1.2876,
|
| 466 |
+
"step": 375
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.12,
|
| 470 |
+
"learning_rate": 0.0001939622641509434,
|
| 471 |
+
"loss": 1.2622,
|
| 472 |
+
"step": 380
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.12,
|
| 476 |
+
"learning_rate": 0.0001938544474393531,
|
| 477 |
+
"loss": 1.2183,
|
| 478 |
+
"step": 385
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.12,
|
| 482 |
+
"learning_rate": 0.00019374663072776284,
|
| 483 |
+
"loss": 1.206,
|
| 484 |
+
"step": 390
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.13,
|
| 488 |
+
"learning_rate": 0.00019363881401617252,
|
| 489 |
+
"loss": 1.2697,
|
| 490 |
+
"step": 395
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.13,
|
| 494 |
+
"learning_rate": 0.0001935309973045822,
|
| 495 |
+
"loss": 1.2515,
|
| 496 |
+
"step": 400
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.13,
|
| 500 |
+
"eval_loss": 1.207701563835144,
|
| 501 |
+
"eval_runtime": 709.7794,
|
| 502 |
+
"eval_samples_per_second": 7.044,
|
| 503 |
+
"eval_steps_per_second": 0.881,
|
| 504 |
+
"step": 400
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.13,
|
| 508 |
+
"learning_rate": 0.00019342318059299192,
|
| 509 |
+
"loss": 1.2596,
|
| 510 |
+
"step": 405
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.13,
|
| 514 |
+
"learning_rate": 0.00019331536388140163,
|
| 515 |
+
"loss": 1.2537,
|
| 516 |
+
"step": 410
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.13,
|
| 520 |
+
"learning_rate": 0.00019320754716981134,
|
| 521 |
+
"loss": 1.2545,
|
| 522 |
+
"step": 415
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.13,
|
| 526 |
+
"learning_rate": 0.00019309973045822103,
|
| 527 |
+
"loss": 1.2473,
|
| 528 |
+
"step": 420
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.14,
|
| 532 |
+
"learning_rate": 0.00019299191374663074,
|
| 533 |
+
"loss": 1.2308,
|
| 534 |
+
"step": 425
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.14,
|
| 538 |
+
"learning_rate": 0.00019288409703504045,
|
| 539 |
+
"loss": 1.2482,
|
| 540 |
+
"step": 430
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.14,
|
| 544 |
+
"learning_rate": 0.00019277628032345013,
|
| 545 |
+
"loss": 1.2342,
|
| 546 |
+
"step": 435
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 0.14,
|
| 550 |
+
"learning_rate": 0.00019266846361185985,
|
| 551 |
+
"loss": 1.2101,
|
| 552 |
+
"step": 440
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.14,
|
| 556 |
+
"learning_rate": 0.00019256064690026956,
|
| 557 |
+
"loss": 1.233,
|
| 558 |
+
"step": 445
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.14,
|
| 562 |
+
"learning_rate": 0.00019245283018867927,
|
| 563 |
+
"loss": 1.2401,
|
| 564 |
+
"step": 450
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.15,
|
| 568 |
+
"learning_rate": 0.00019234501347708895,
|
| 569 |
+
"loss": 1.2088,
|
| 570 |
+
"step": 455
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.15,
|
| 574 |
+
"learning_rate": 0.00019223719676549864,
|
| 575 |
+
"loss": 1.2205,
|
| 576 |
+
"step": 460
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.15,
|
| 580 |
+
"learning_rate": 0.00019212938005390838,
|
| 581 |
+
"loss": 1.218,
|
| 582 |
+
"step": 465
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.15,
|
| 586 |
+
"learning_rate": 0.00019202156334231806,
|
| 587 |
+
"loss": 1.2208,
|
| 588 |
+
"step": 470
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.15,
|
| 592 |
+
"learning_rate": 0.00019191374663072777,
|
| 593 |
+
"loss": 1.2783,
|
| 594 |
+
"step": 475
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.15,
|
| 598 |
+
"learning_rate": 0.00019180592991913749,
|
| 599 |
+
"loss": 1.1886,
|
| 600 |
+
"step": 480
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.16,
|
| 604 |
+
"learning_rate": 0.00019169811320754717,
|
| 605 |
+
"loss": 1.1939,
|
| 606 |
+
"step": 485
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.16,
|
| 610 |
+
"learning_rate": 0.00019159029649595688,
|
| 611 |
+
"loss": 1.2079,
|
| 612 |
+
"step": 490
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.16,
|
| 616 |
+
"learning_rate": 0.00019148247978436657,
|
| 617 |
+
"loss": 1.2473,
|
| 618 |
+
"step": 495
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.16,
|
| 622 |
+
"learning_rate": 0.0001913746630727763,
|
| 623 |
+
"loss": 1.2295,
|
| 624 |
+
"step": 500
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.16,
|
| 628 |
+
"learning_rate": 0.000191266846361186,
|
| 629 |
+
"loss": 1.2064,
|
| 630 |
+
"step": 505
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.16,
|
| 634 |
+
"learning_rate": 0.00019115902964959568,
|
| 635 |
+
"loss": 1.2138,
|
| 636 |
+
"step": 510
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.16,
|
| 640 |
+
"learning_rate": 0.00019105121293800541,
|
| 641 |
+
"loss": 1.2618,
|
| 642 |
+
"step": 515
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.17,
|
| 646 |
+
"learning_rate": 0.0001909433962264151,
|
| 647 |
+
"loss": 1.2495,
|
| 648 |
+
"step": 520
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.17,
|
| 652 |
+
"learning_rate": 0.0001908355795148248,
|
| 653 |
+
"loss": 1.2553,
|
| 654 |
+
"step": 525
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.17,
|
| 658 |
+
"learning_rate": 0.0001907277628032345,
|
| 659 |
+
"loss": 1.2334,
|
| 660 |
+
"step": 530
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.17,
|
| 664 |
+
"learning_rate": 0.0001906199460916442,
|
| 665 |
+
"loss": 1.2442,
|
| 666 |
+
"step": 535
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.17,
|
| 670 |
+
"learning_rate": 0.00019051212938005392,
|
| 671 |
+
"loss": 1.2301,
|
| 672 |
+
"step": 540
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.17,
|
| 676 |
+
"learning_rate": 0.0001904043126684636,
|
| 677 |
+
"loss": 1.2678,
|
| 678 |
+
"step": 545
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.18,
|
| 682 |
+
"learning_rate": 0.00019029649595687334,
|
| 683 |
+
"loss": 1.2116,
|
| 684 |
+
"step": 550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.18,
|
| 688 |
+
"learning_rate": 0.00019018867924528303,
|
| 689 |
+
"loss": 1.2556,
|
| 690 |
+
"step": 555
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.18,
|
| 694 |
+
"learning_rate": 0.00019008086253369274,
|
| 695 |
+
"loss": 1.2427,
|
| 696 |
+
"step": 560
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.18,
|
| 700 |
+
"learning_rate": 0.00018997304582210242,
|
| 701 |
+
"loss": 1.2138,
|
| 702 |
+
"step": 565
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.18,
|
| 706 |
+
"learning_rate": 0.00018986522911051214,
|
| 707 |
+
"loss": 1.248,
|
| 708 |
+
"step": 570
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.18,
|
| 712 |
+
"learning_rate": 0.00018975741239892185,
|
| 713 |
+
"loss": 1.2415,
|
| 714 |
+
"step": 575
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.19,
|
| 718 |
+
"learning_rate": 0.00018964959568733153,
|
| 719 |
+
"loss": 1.2303,
|
| 720 |
+
"step": 580
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.19,
|
| 724 |
+
"learning_rate": 0.00018954177897574125,
|
| 725 |
+
"loss": 1.1939,
|
| 726 |
+
"step": 585
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.19,
|
| 730 |
+
"learning_rate": 0.00018943396226415096,
|
| 731 |
+
"loss": 1.1927,
|
| 732 |
+
"step": 590
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.19,
|
| 736 |
+
"learning_rate": 0.00018932614555256064,
|
| 737 |
+
"loss": 1.1982,
|
| 738 |
+
"step": 595
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.19,
|
| 742 |
+
"learning_rate": 0.00018921832884097035,
|
| 743 |
+
"loss": 1.1859,
|
| 744 |
+
"step": 600
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.19,
|
| 748 |
+
"eval_loss": 1.1948214769363403,
|
| 749 |
+
"eval_runtime": 709.3061,
|
| 750 |
+
"eval_samples_per_second": 7.049,
|
| 751 |
+
"eval_steps_per_second": 0.881,
|
| 752 |
+
"step": 600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.19,
|
| 756 |
+
"learning_rate": 0.00018911051212938007,
|
| 757 |
+
"loss": 1.2265,
|
| 758 |
+
"step": 605
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.2,
|
| 762 |
+
"learning_rate": 0.00018900269541778978,
|
| 763 |
+
"loss": 1.2402,
|
| 764 |
+
"step": 610
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.2,
|
| 768 |
+
"learning_rate": 0.00018889487870619946,
|
| 769 |
+
"loss": 1.2261,
|
| 770 |
+
"step": 615
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.2,
|
| 774 |
+
"learning_rate": 0.00018878706199460917,
|
| 775 |
+
"loss": 1.2264,
|
| 776 |
+
"step": 620
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.2,
|
| 780 |
+
"learning_rate": 0.00018867924528301889,
|
| 781 |
+
"loss": 1.2044,
|
| 782 |
+
"step": 625
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.2,
|
| 786 |
+
"learning_rate": 0.00018857142857142857,
|
| 787 |
+
"loss": 1.2203,
|
| 788 |
+
"step": 630
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.2,
|
| 792 |
+
"learning_rate": 0.00018846361185983828,
|
| 793 |
+
"loss": 1.2258,
|
| 794 |
+
"step": 635
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.2,
|
| 798 |
+
"learning_rate": 0.000188355795148248,
|
| 799 |
+
"loss": 1.2475,
|
| 800 |
+
"step": 640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.21,
|
| 804 |
+
"learning_rate": 0.00018824797843665768,
|
| 805 |
+
"loss": 1.239,
|
| 806 |
+
"step": 645
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.21,
|
| 810 |
+
"learning_rate": 0.0001881401617250674,
|
| 811 |
+
"loss": 1.2374,
|
| 812 |
+
"step": 650
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.21,
|
| 816 |
+
"learning_rate": 0.0001880323450134771,
|
| 817 |
+
"loss": 1.2472,
|
| 818 |
+
"step": 655
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.21,
|
| 822 |
+
"learning_rate": 0.00018792452830188681,
|
| 823 |
+
"loss": 1.2411,
|
| 824 |
+
"step": 660
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.21,
|
| 828 |
+
"learning_rate": 0.0001878167115902965,
|
| 829 |
+
"loss": 1.1916,
|
| 830 |
+
"step": 665
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.21,
|
| 834 |
+
"learning_rate": 0.0001877088948787062,
|
| 835 |
+
"loss": 1.2043,
|
| 836 |
+
"step": 670
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.22,
|
| 840 |
+
"learning_rate": 0.00018760107816711592,
|
| 841 |
+
"loss": 1.2087,
|
| 842 |
+
"step": 675
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.22,
|
| 846 |
+
"learning_rate": 0.0001874932614555256,
|
| 847 |
+
"loss": 1.2147,
|
| 848 |
+
"step": 680
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.22,
|
| 852 |
+
"learning_rate": 0.00018738544474393532,
|
| 853 |
+
"loss": 1.1999,
|
| 854 |
+
"step": 685
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.22,
|
| 858 |
+
"learning_rate": 0.00018727762803234503,
|
| 859 |
+
"loss": 1.2031,
|
| 860 |
+
"step": 690
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.22,
|
| 864 |
+
"learning_rate": 0.00018716981132075472,
|
| 865 |
+
"loss": 1.268,
|
| 866 |
+
"step": 695
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.22,
|
| 870 |
+
"learning_rate": 0.00018706199460916443,
|
| 871 |
+
"loss": 1.2181,
|
| 872 |
+
"step": 700
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.23,
|
| 876 |
+
"learning_rate": 0.0001869541778975741,
|
| 877 |
+
"loss": 1.2187,
|
| 878 |
+
"step": 705
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.23,
|
| 882 |
+
"learning_rate": 0.00018684636118598385,
|
| 883 |
+
"loss": 1.2475,
|
| 884 |
+
"step": 710
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.23,
|
| 888 |
+
"learning_rate": 0.00018673854447439354,
|
| 889 |
+
"loss": 1.2043,
|
| 890 |
+
"step": 715
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.23,
|
| 894 |
+
"learning_rate": 0.00018663072776280325,
|
| 895 |
+
"loss": 1.1833,
|
| 896 |
+
"step": 720
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.23,
|
| 900 |
+
"learning_rate": 0.00018652291105121296,
|
| 901 |
+
"loss": 1.1899,
|
| 902 |
+
"step": 725
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.23,
|
| 906 |
+
"learning_rate": 0.00018641509433962264,
|
| 907 |
+
"loss": 1.232,
|
| 908 |
+
"step": 730
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.24,
|
| 912 |
+
"learning_rate": 0.00018630727762803236,
|
| 913 |
+
"loss": 1.1885,
|
| 914 |
+
"step": 735
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.24,
|
| 918 |
+
"learning_rate": 0.00018619946091644204,
|
| 919 |
+
"loss": 1.2087,
|
| 920 |
+
"step": 740
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.24,
|
| 924 |
+
"learning_rate": 0.00018609164420485178,
|
| 925 |
+
"loss": 1.2433,
|
| 926 |
+
"step": 745
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.24,
|
| 930 |
+
"learning_rate": 0.00018598382749326146,
|
| 931 |
+
"loss": 1.2233,
|
| 932 |
+
"step": 750
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.24,
|
| 936 |
+
"learning_rate": 0.00018587601078167115,
|
| 937 |
+
"loss": 1.2348,
|
| 938 |
+
"step": 755
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.24,
|
| 942 |
+
"learning_rate": 0.0001857681940700809,
|
| 943 |
+
"loss": 1.1667,
|
| 944 |
+
"step": 760
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.24,
|
| 948 |
+
"learning_rate": 0.00018566037735849057,
|
| 949 |
+
"loss": 1.2952,
|
| 950 |
+
"step": 765
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.25,
|
| 954 |
+
"learning_rate": 0.00018555256064690029,
|
| 955 |
+
"loss": 1.1939,
|
| 956 |
+
"step": 770
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.25,
|
| 960 |
+
"learning_rate": 0.00018544474393530997,
|
| 961 |
+
"loss": 1.1646,
|
| 962 |
+
"step": 775
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.25,
|
| 966 |
+
"learning_rate": 0.00018533692722371968,
|
| 967 |
+
"loss": 1.2055,
|
| 968 |
+
"step": 780
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.25,
|
| 972 |
+
"learning_rate": 0.0001852291105121294,
|
| 973 |
+
"loss": 1.183,
|
| 974 |
+
"step": 785
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.25,
|
| 978 |
+
"learning_rate": 0.00018512129380053908,
|
| 979 |
+
"loss": 1.2367,
|
| 980 |
+
"step": 790
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.25,
|
| 984 |
+
"learning_rate": 0.00018501347708894882,
|
| 985 |
+
"loss": 1.2265,
|
| 986 |
+
"step": 795
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.26,
|
| 990 |
+
"learning_rate": 0.0001849056603773585,
|
| 991 |
+
"loss": 1.2007,
|
| 992 |
+
"step": 800
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.26,
|
| 996 |
+
"eval_loss": 1.184471845626831,
|
| 997 |
+
"eval_runtime": 708.4336,
|
| 998 |
+
"eval_samples_per_second": 7.058,
|
| 999 |
+
"eval_steps_per_second": 0.882,
|
| 1000 |
+
"step": 800
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 0.26,
|
| 1004 |
+
"learning_rate": 0.0001847978436657682,
|
| 1005 |
+
"loss": 1.204,
|
| 1006 |
+
"step": 805
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.26,
|
| 1010 |
+
"learning_rate": 0.0001846900269541779,
|
| 1011 |
+
"loss": 1.2379,
|
| 1012 |
+
"step": 810
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.26,
|
| 1016 |
+
"learning_rate": 0.0001845822102425876,
|
| 1017 |
+
"loss": 1.157,
|
| 1018 |
+
"step": 815
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.26,
|
| 1022 |
+
"learning_rate": 0.00018447439353099732,
|
| 1023 |
+
"loss": 1.2072,
|
| 1024 |
+
"step": 820
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.26,
|
| 1028 |
+
"learning_rate": 0.000184366576819407,
|
| 1029 |
+
"loss": 1.1764,
|
| 1030 |
+
"step": 825
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.27,
|
| 1034 |
+
"learning_rate": 0.00018425876010781672,
|
| 1035 |
+
"loss": 1.21,
|
| 1036 |
+
"step": 830
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.27,
|
| 1040 |
+
"learning_rate": 0.00018415094339622643,
|
| 1041 |
+
"loss": 1.19,
|
| 1042 |
+
"step": 835
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 0.27,
|
| 1046 |
+
"learning_rate": 0.00018404312668463612,
|
| 1047 |
+
"loss": 1.1923,
|
| 1048 |
+
"step": 840
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.27,
|
| 1052 |
+
"learning_rate": 0.00018393530997304583,
|
| 1053 |
+
"loss": 1.2367,
|
| 1054 |
+
"step": 845
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.27,
|
| 1058 |
+
"learning_rate": 0.00018382749326145554,
|
| 1059 |
+
"loss": 1.2394,
|
| 1060 |
+
"step": 850
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.27,
|
| 1064 |
+
"learning_rate": 0.00018371967654986525,
|
| 1065 |
+
"loss": 1.2275,
|
| 1066 |
+
"step": 855
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.28,
|
| 1070 |
+
"learning_rate": 0.00018361185983827494,
|
| 1071 |
+
"loss": 1.2206,
|
| 1072 |
+
"step": 860
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.28,
|
| 1076 |
+
"learning_rate": 0.00018350404312668465,
|
| 1077 |
+
"loss": 1.2017,
|
| 1078 |
+
"step": 865
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 0.28,
|
| 1082 |
+
"learning_rate": 0.00018339622641509436,
|
| 1083 |
+
"loss": 1.2277,
|
| 1084 |
+
"step": 870
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.28,
|
| 1088 |
+
"learning_rate": 0.00018328840970350404,
|
| 1089 |
+
"loss": 1.1958,
|
| 1090 |
+
"step": 875
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.28,
|
| 1094 |
+
"learning_rate": 0.00018318059299191376,
|
| 1095 |
+
"loss": 1.2317,
|
| 1096 |
+
"step": 880
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.28,
|
| 1100 |
+
"learning_rate": 0.00018307277628032347,
|
| 1101 |
+
"loss": 1.2339,
|
| 1102 |
+
"step": 885
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.28,
|
| 1106 |
+
"learning_rate": 0.00018296495956873315,
|
| 1107 |
+
"loss": 1.2408,
|
| 1108 |
+
"step": 890
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.29,
|
| 1112 |
+
"learning_rate": 0.00018285714285714286,
|
| 1113 |
+
"loss": 1.1818,
|
| 1114 |
+
"step": 895
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 0.29,
|
| 1118 |
+
"learning_rate": 0.00018274932614555258,
|
| 1119 |
+
"loss": 1.2051,
|
| 1120 |
+
"step": 900
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 0.29,
|
| 1124 |
+
"learning_rate": 0.0001826415094339623,
|
| 1125 |
+
"loss": 1.2304,
|
| 1126 |
+
"step": 905
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.29,
|
| 1130 |
+
"learning_rate": 0.00018253369272237197,
|
| 1131 |
+
"loss": 1.1717,
|
| 1132 |
+
"step": 910
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.29,
|
| 1136 |
+
"learning_rate": 0.00018242587601078168,
|
| 1137 |
+
"loss": 1.1811,
|
| 1138 |
+
"step": 915
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.29,
|
| 1142 |
+
"learning_rate": 0.0001823180592991914,
|
| 1143 |
+
"loss": 1.2127,
|
| 1144 |
+
"step": 920
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.3,
|
| 1148 |
+
"learning_rate": 0.00018221024258760108,
|
| 1149 |
+
"loss": 1.2278,
|
| 1150 |
+
"step": 925
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.3,
|
| 1154 |
+
"learning_rate": 0.0001821024258760108,
|
| 1155 |
+
"loss": 1.223,
|
| 1156 |
+
"step": 930
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 0.3,
|
| 1160 |
+
"learning_rate": 0.0001819946091644205,
|
| 1161 |
+
"loss": 1.233,
|
| 1162 |
+
"step": 935
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 0.3,
|
| 1166 |
+
"learning_rate": 0.0001818867924528302,
|
| 1167 |
+
"loss": 1.2321,
|
| 1168 |
+
"step": 940
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.3,
|
| 1172 |
+
"learning_rate": 0.0001817789757412399,
|
| 1173 |
+
"loss": 1.1894,
|
| 1174 |
+
"step": 945
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.3,
|
| 1178 |
+
"learning_rate": 0.00018167115902964959,
|
| 1179 |
+
"loss": 1.2213,
|
| 1180 |
+
"step": 950
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.31,
|
| 1184 |
+
"learning_rate": 0.00018156334231805933,
|
| 1185 |
+
"loss": 1.2444,
|
| 1186 |
+
"step": 955
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.31,
|
| 1190 |
+
"learning_rate": 0.000181455525606469,
|
| 1191 |
+
"loss": 1.1988,
|
| 1192 |
+
"step": 960
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.31,
|
| 1196 |
+
"learning_rate": 0.00018134770889487872,
|
| 1197 |
+
"loss": 1.2039,
|
| 1198 |
+
"step": 965
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 0.31,
|
| 1202 |
+
"learning_rate": 0.00018123989218328843,
|
| 1203 |
+
"loss": 1.1972,
|
| 1204 |
+
"step": 970
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 0.31,
|
| 1208 |
+
"learning_rate": 0.00018113207547169812,
|
| 1209 |
+
"loss": 1.2146,
|
| 1210 |
+
"step": 975
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.31,
|
| 1214 |
+
"learning_rate": 0.00018102425876010783,
|
| 1215 |
+
"loss": 1.2346,
|
| 1216 |
+
"step": 980
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.32,
|
| 1220 |
+
"learning_rate": 0.00018091644204851751,
|
| 1221 |
+
"loss": 1.2283,
|
| 1222 |
+
"step": 985
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.32,
|
| 1226 |
+
"learning_rate": 0.00018080862533692723,
|
| 1227 |
+
"loss": 1.201,
|
| 1228 |
+
"step": 990
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.32,
|
| 1232 |
+
"learning_rate": 0.00018070080862533694,
|
| 1233 |
+
"loss": 1.2149,
|
| 1234 |
+
"step": 995
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.32,
|
| 1238 |
+
"learning_rate": 0.00018059299191374662,
|
| 1239 |
+
"loss": 1.2085,
|
| 1240 |
+
"step": 1000
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 0.32,
|
| 1244 |
+
"eval_loss": 1.176331877708435,
|
| 1245 |
+
"eval_runtime": 709.1565,
|
| 1246 |
+
"eval_samples_per_second": 7.051,
|
| 1247 |
+
"eval_steps_per_second": 0.881,
|
| 1248 |
+
"step": 1000
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.32,
|
| 1252 |
+
"learning_rate": 0.00018048517520215636,
|
| 1253 |
+
"loss": 1.2484,
|
| 1254 |
+
"step": 1005
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 0.32,
|
| 1258 |
+
"learning_rate": 0.00018037735849056605,
|
| 1259 |
+
"loss": 1.2193,
|
| 1260 |
+
"step": 1010
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 0.32,
|
| 1264 |
+
"learning_rate": 0.00018026954177897576,
|
| 1265 |
+
"loss": 1.1733,
|
| 1266 |
+
"step": 1015
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.33,
|
| 1270 |
+
"learning_rate": 0.00018016172506738544,
|
| 1271 |
+
"loss": 1.2152,
|
| 1272 |
+
"step": 1020
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.33,
|
| 1276 |
+
"learning_rate": 0.00018005390835579516,
|
| 1277 |
+
"loss": 1.1688,
|
| 1278 |
+
"step": 1025
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.33,
|
| 1282 |
+
"learning_rate": 0.00017994609164420487,
|
| 1283 |
+
"loss": 1.1893,
|
| 1284 |
+
"step": 1030
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.33,
|
| 1288 |
+
"learning_rate": 0.00017983827493261455,
|
| 1289 |
+
"loss": 1.1483,
|
| 1290 |
+
"step": 1035
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.33,
|
| 1294 |
+
"learning_rate": 0.0001797304582210243,
|
| 1295 |
+
"loss": 1.189,
|
| 1296 |
+
"step": 1040
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 0.33,
|
| 1300 |
+
"learning_rate": 0.00017962264150943398,
|
| 1301 |
+
"loss": 1.1873,
|
| 1302 |
+
"step": 1045
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 0.34,
|
| 1306 |
+
"learning_rate": 0.00017951482479784366,
|
| 1307 |
+
"loss": 1.1883,
|
| 1308 |
+
"step": 1050
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.34,
|
| 1312 |
+
"learning_rate": 0.00017940700808625337,
|
| 1313 |
+
"loss": 1.1741,
|
| 1314 |
+
"step": 1055
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.34,
|
| 1318 |
+
"learning_rate": 0.00017929919137466308,
|
| 1319 |
+
"loss": 1.2546,
|
| 1320 |
+
"step": 1060
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.34,
|
| 1324 |
+
"learning_rate": 0.0001791913746630728,
|
| 1325 |
+
"loss": 1.2194,
|
| 1326 |
+
"step": 1065
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.34,
|
| 1330 |
+
"learning_rate": 0.00017908355795148248,
|
| 1331 |
+
"loss": 1.1945,
|
| 1332 |
+
"step": 1070
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.34,
|
| 1336 |
+
"learning_rate": 0.0001789757412398922,
|
| 1337 |
+
"loss": 1.2011,
|
| 1338 |
+
"step": 1075
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.35,
|
| 1342 |
+
"learning_rate": 0.0001788679245283019,
|
| 1343 |
+
"loss": 1.1836,
|
| 1344 |
+
"step": 1080
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 0.35,
|
| 1348 |
+
"learning_rate": 0.0001787601078167116,
|
| 1349 |
+
"loss": 1.2116,
|
| 1350 |
+
"step": 1085
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.35,
|
| 1354 |
+
"learning_rate": 0.0001786522911051213,
|
| 1355 |
+
"loss": 1.1983,
|
| 1356 |
+
"step": 1090
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.35,
|
| 1360 |
+
"learning_rate": 0.000178544474393531,
|
| 1361 |
+
"loss": 1.2095,
|
| 1362 |
+
"step": 1095
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 0.35,
|
| 1366 |
+
"learning_rate": 0.0001784366576819407,
|
| 1367 |
+
"loss": 1.2007,
|
| 1368 |
+
"step": 1100
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.35,
|
| 1372 |
+
"learning_rate": 0.0001783288409703504,
|
| 1373 |
+
"loss": 1.1936,
|
| 1374 |
+
"step": 1105
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.36,
|
| 1378 |
+
"learning_rate": 0.0001782210242587601,
|
| 1379 |
+
"loss": 1.2207,
|
| 1380 |
+
"step": 1110
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.36,
|
| 1384 |
+
"learning_rate": 0.00017811320754716983,
|
| 1385 |
+
"loss": 1.2016,
|
| 1386 |
+
"step": 1115
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 0.36,
|
| 1390 |
+
"learning_rate": 0.00017800539083557952,
|
| 1391 |
+
"loss": 1.2013,
|
| 1392 |
+
"step": 1120
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.36,
|
| 1396 |
+
"learning_rate": 0.00017789757412398923,
|
| 1397 |
+
"loss": 1.1794,
|
| 1398 |
+
"step": 1125
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.36,
|
| 1402 |
+
"learning_rate": 0.00017778975741239894,
|
| 1403 |
+
"loss": 1.235,
|
| 1404 |
+
"step": 1130
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.36,
|
| 1408 |
+
"learning_rate": 0.00017768194070080863,
|
| 1409 |
+
"loss": 1.1865,
|
| 1410 |
+
"step": 1135
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.36,
|
| 1414 |
+
"learning_rate": 0.00017757412398921834,
|
| 1415 |
+
"loss": 1.2159,
|
| 1416 |
+
"step": 1140
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.37,
|
| 1420 |
+
"learning_rate": 0.00017746630727762802,
|
| 1421 |
+
"loss": 1.1887,
|
| 1422 |
+
"step": 1145
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 0.37,
|
| 1426 |
+
"learning_rate": 0.00017735849056603776,
|
| 1427 |
+
"loss": 1.1923,
|
| 1428 |
+
"step": 1150
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.37,
|
| 1432 |
+
"learning_rate": 0.00017725067385444745,
|
| 1433 |
+
"loss": 1.1894,
|
| 1434 |
+
"step": 1155
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.37,
|
| 1438 |
+
"learning_rate": 0.00017714285714285713,
|
| 1439 |
+
"loss": 1.1863,
|
| 1440 |
+
"step": 1160
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.37,
|
| 1444 |
+
"learning_rate": 0.00017703504043126687,
|
| 1445 |
+
"loss": 1.2128,
|
| 1446 |
+
"step": 1165
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.37,
|
| 1450 |
+
"learning_rate": 0.00017692722371967655,
|
| 1451 |
+
"loss": 1.214,
|
| 1452 |
+
"step": 1170
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.38,
|
| 1456 |
+
"learning_rate": 0.00017681940700808627,
|
| 1457 |
+
"loss": 1.2083,
|
| 1458 |
+
"step": 1175
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.38,
|
| 1462 |
+
"learning_rate": 0.00017671159029649595,
|
| 1463 |
+
"loss": 1.2252,
|
| 1464 |
+
"step": 1180
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 0.38,
|
| 1468 |
+
"learning_rate": 0.00017660377358490566,
|
| 1469 |
+
"loss": 1.1818,
|
| 1470 |
+
"step": 1185
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.38,
|
| 1474 |
+
"learning_rate": 0.00017649595687331538,
|
| 1475 |
+
"loss": 1.2334,
|
| 1476 |
+
"step": 1190
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.38,
|
| 1480 |
+
"learning_rate": 0.00017638814016172506,
|
| 1481 |
+
"loss": 1.1752,
|
| 1482 |
+
"step": 1195
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.38,
|
| 1486 |
+
"learning_rate": 0.0001762803234501348,
|
| 1487 |
+
"loss": 1.2036,
|
| 1488 |
+
"step": 1200
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.38,
|
| 1492 |
+
"eval_loss": 1.1710153818130493,
|
| 1493 |
+
"eval_runtime": 708.8537,
|
| 1494 |
+
"eval_samples_per_second": 7.054,
|
| 1495 |
+
"eval_steps_per_second": 0.882,
|
| 1496 |
+
"step": 1200
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.39,
|
| 1500 |
+
"learning_rate": 0.00017617250673854448,
|
| 1501 |
+
"loss": 1.2283,
|
| 1502 |
+
"step": 1205
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.39,
|
| 1506 |
+
"learning_rate": 0.0001760646900269542,
|
| 1507 |
+
"loss": 1.2464,
|
| 1508 |
+
"step": 1210
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 0.39,
|
| 1512 |
+
"learning_rate": 0.00017595687331536388,
|
| 1513 |
+
"loss": 1.2126,
|
| 1514 |
+
"step": 1215
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.39,
|
| 1518 |
+
"learning_rate": 0.0001758490566037736,
|
| 1519 |
+
"loss": 1.2217,
|
| 1520 |
+
"step": 1220
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.39,
|
| 1524 |
+
"learning_rate": 0.0001757412398921833,
|
| 1525 |
+
"loss": 1.2245,
|
| 1526 |
+
"step": 1225
|
| 1527 |
+
},
|
| 1528 |
+
{
|
| 1529 |
+
"epoch": 0.39,
|
| 1530 |
+
"learning_rate": 0.000175633423180593,
|
| 1531 |
+
"loss": 1.1941,
|
| 1532 |
+
"step": 1230
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 0.4,
|
| 1536 |
+
"learning_rate": 0.0001755256064690027,
|
| 1537 |
+
"loss": 1.2041,
|
| 1538 |
+
"step": 1235
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.4,
|
| 1542 |
+
"learning_rate": 0.0001754177897574124,
|
| 1543 |
+
"loss": 1.2076,
|
| 1544 |
+
"step": 1240
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.4,
|
| 1548 |
+
"learning_rate": 0.0001753099730458221,
|
| 1549 |
+
"loss": 1.2154,
|
| 1550 |
+
"step": 1245
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 0.4,
|
| 1554 |
+
"learning_rate": 0.0001752021563342318,
|
| 1555 |
+
"loss": 1.2247,
|
| 1556 |
+
"step": 1250
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.4,
|
| 1560 |
+
"learning_rate": 0.00017509433962264152,
|
| 1561 |
+
"loss": 1.1822,
|
| 1562 |
+
"step": 1255
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.4,
|
| 1566 |
+
"learning_rate": 0.00017498652291105123,
|
| 1567 |
+
"loss": 1.2015,
|
| 1568 |
+
"step": 1260
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.4,
|
| 1572 |
+
"learning_rate": 0.00017487870619946092,
|
| 1573 |
+
"loss": 1.2111,
|
| 1574 |
+
"step": 1265
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 0.41,
|
| 1578 |
+
"learning_rate": 0.00017477088948787063,
|
| 1579 |
+
"loss": 1.1946,
|
| 1580 |
+
"step": 1270
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.41,
|
| 1584 |
+
"learning_rate": 0.00017466307277628034,
|
| 1585 |
+
"loss": 1.1365,
|
| 1586 |
+
"step": 1275
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.41,
|
| 1590 |
+
"learning_rate": 0.00017455525606469003,
|
| 1591 |
+
"loss": 1.1742,
|
| 1592 |
+
"step": 1280
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 0.41,
|
| 1596 |
+
"learning_rate": 0.00017444743935309974,
|
| 1597 |
+
"loss": 1.2442,
|
| 1598 |
+
"step": 1285
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.41,
|
| 1602 |
+
"learning_rate": 0.00017433962264150945,
|
| 1603 |
+
"loss": 1.2095,
|
| 1604 |
+
"step": 1290
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 0.41,
|
| 1608 |
+
"learning_rate": 0.00017423180592991913,
|
| 1609 |
+
"loss": 1.1727,
|
| 1610 |
+
"step": 1295
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 0.42,
|
| 1614 |
+
"learning_rate": 0.00017412398921832885,
|
| 1615 |
+
"loss": 1.1948,
|
| 1616 |
+
"step": 1300
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.42,
|
| 1620 |
+
"learning_rate": 0.00017401617250673856,
|
| 1621 |
+
"loss": 1.1737,
|
| 1622 |
+
"step": 1305
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.42,
|
| 1626 |
+
"learning_rate": 0.00017390835579514827,
|
| 1627 |
+
"loss": 1.198,
|
| 1628 |
+
"step": 1310
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.42,
|
| 1632 |
+
"learning_rate": 0.00017380053908355795,
|
| 1633 |
+
"loss": 1.2152,
|
| 1634 |
+
"step": 1315
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 0.42,
|
| 1638 |
+
"learning_rate": 0.00017369272237196767,
|
| 1639 |
+
"loss": 1.2061,
|
| 1640 |
+
"step": 1320
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.42,
|
| 1644 |
+
"learning_rate": 0.00017358490566037738,
|
| 1645 |
+
"loss": 1.1925,
|
| 1646 |
+
"step": 1325
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 0.43,
|
| 1650 |
+
"learning_rate": 0.00017347708894878706,
|
| 1651 |
+
"loss": 1.2067,
|
| 1652 |
+
"step": 1330
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"epoch": 0.43,
|
| 1656 |
+
"learning_rate": 0.00017336927223719677,
|
| 1657 |
+
"loss": 1.1964,
|
| 1658 |
+
"step": 1335
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.43,
|
| 1662 |
+
"learning_rate": 0.0001732614555256065,
|
| 1663 |
+
"loss": 1.2062,
|
| 1664 |
+
"step": 1340
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.43,
|
| 1668 |
+
"learning_rate": 0.00017315363881401617,
|
| 1669 |
+
"loss": 1.2208,
|
| 1670 |
+
"step": 1345
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.43,
|
| 1674 |
+
"learning_rate": 0.00017304582210242588,
|
| 1675 |
+
"loss": 1.167,
|
| 1676 |
+
"step": 1350
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.43,
|
| 1680 |
+
"learning_rate": 0.00017293800539083557,
|
| 1681 |
+
"loss": 1.1802,
|
| 1682 |
+
"step": 1355
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.44,
|
| 1686 |
+
"learning_rate": 0.0001728301886792453,
|
| 1687 |
+
"loss": 1.2193,
|
| 1688 |
+
"step": 1360
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 0.44,
|
| 1692 |
+
"learning_rate": 0.000172722371967655,
|
| 1693 |
+
"loss": 1.2015,
|
| 1694 |
+
"step": 1365
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 0.44,
|
| 1698 |
+
"learning_rate": 0.0001726145552560647,
|
| 1699 |
+
"loss": 1.2183,
|
| 1700 |
+
"step": 1370
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.44,
|
| 1704 |
+
"learning_rate": 0.00017250673854447442,
|
| 1705 |
+
"loss": 1.1986,
|
| 1706 |
+
"step": 1375
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 0.44,
|
| 1710 |
+
"learning_rate": 0.0001723989218328841,
|
| 1711 |
+
"loss": 1.1575,
|
| 1712 |
+
"step": 1380
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.44,
|
| 1716 |
+
"learning_rate": 0.0001722911051212938,
|
| 1717 |
+
"loss": 1.1949,
|
| 1718 |
+
"step": 1385
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.44,
|
| 1722 |
+
"learning_rate": 0.0001721832884097035,
|
| 1723 |
+
"loss": 1.2103,
|
| 1724 |
+
"step": 1390
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.45,
|
| 1728 |
+
"learning_rate": 0.00017207547169811324,
|
| 1729 |
+
"loss": 1.2141,
|
| 1730 |
+
"step": 1395
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 0.45,
|
| 1734 |
+
"learning_rate": 0.00017196765498652292,
|
| 1735 |
+
"loss": 1.2134,
|
| 1736 |
+
"step": 1400
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 0.45,
|
| 1740 |
+
"eval_loss": 1.1655155420303345,
|
| 1741 |
+
"eval_runtime": 709.3424,
|
| 1742 |
+
"eval_samples_per_second": 7.049,
|
| 1743 |
+
"eval_steps_per_second": 0.881,
|
| 1744 |
+
"step": 1400
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 0.45,
|
| 1748 |
+
"learning_rate": 0.0001718598382749326,
|
| 1749 |
+
"loss": 1.1573,
|
| 1750 |
+
"step": 1405
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.45,
|
| 1754 |
+
"learning_rate": 0.00017175202156334234,
|
| 1755 |
+
"loss": 1.1694,
|
| 1756 |
+
"step": 1410
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.45,
|
| 1760 |
+
"learning_rate": 0.00017164420485175203,
|
| 1761 |
+
"loss": 1.226,
|
| 1762 |
+
"step": 1415
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.45,
|
| 1766 |
+
"learning_rate": 0.00017153638814016174,
|
| 1767 |
+
"loss": 1.1758,
|
| 1768 |
+
"step": 1420
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.46,
|
| 1772 |
+
"learning_rate": 0.00017142857142857143,
|
| 1773 |
+
"loss": 1.185,
|
| 1774 |
+
"step": 1425
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.46,
|
| 1778 |
+
"learning_rate": 0.00017132075471698114,
|
| 1779 |
+
"loss": 1.1878,
|
| 1780 |
+
"step": 1430
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.46,
|
| 1784 |
+
"learning_rate": 0.00017121293800539085,
|
| 1785 |
+
"loss": 1.2108,
|
| 1786 |
+
"step": 1435
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 0.46,
|
| 1790 |
+
"learning_rate": 0.00017110512129380053,
|
| 1791 |
+
"loss": 1.254,
|
| 1792 |
+
"step": 1440
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.46,
|
| 1796 |
+
"learning_rate": 0.00017099730458221027,
|
| 1797 |
+
"loss": 1.2071,
|
| 1798 |
+
"step": 1445
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 0.46,
|
| 1802 |
+
"learning_rate": 0.00017088948787061996,
|
| 1803 |
+
"loss": 1.1914,
|
| 1804 |
+
"step": 1450
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.47,
|
| 1808 |
+
"learning_rate": 0.00017078167115902964,
|
| 1809 |
+
"loss": 1.1936,
|
| 1810 |
+
"step": 1455
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.47,
|
| 1814 |
+
"learning_rate": 0.00017067385444743935,
|
| 1815 |
+
"loss": 1.2369,
|
| 1816 |
+
"step": 1460
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.47,
|
| 1820 |
+
"learning_rate": 0.00017056603773584907,
|
| 1821 |
+
"loss": 1.1627,
|
| 1822 |
+
"step": 1465
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.47,
|
| 1826 |
+
"learning_rate": 0.00017045822102425878,
|
| 1827 |
+
"loss": 1.216,
|
| 1828 |
+
"step": 1470
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 0.47,
|
| 1832 |
+
"learning_rate": 0.00017035040431266846,
|
| 1833 |
+
"loss": 1.1931,
|
| 1834 |
+
"step": 1475
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 0.47,
|
| 1838 |
+
"learning_rate": 0.00017024258760107817,
|
| 1839 |
+
"loss": 1.1859,
|
| 1840 |
+
"step": 1480
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 0.48,
|
| 1844 |
+
"learning_rate": 0.00017013477088948789,
|
| 1845 |
+
"loss": 1.1769,
|
| 1846 |
+
"step": 1485
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.48,
|
| 1850 |
+
"learning_rate": 0.00017002695417789757,
|
| 1851 |
+
"loss": 1.1376,
|
| 1852 |
+
"step": 1490
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.48,
|
| 1856 |
+
"learning_rate": 0.00016991913746630728,
|
| 1857 |
+
"loss": 1.1692,
|
| 1858 |
+
"step": 1495
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.48,
|
| 1862 |
+
"learning_rate": 0.000169811320754717,
|
| 1863 |
+
"loss": 1.1948,
|
| 1864 |
+
"step": 1500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.48,
|
| 1868 |
+
"learning_rate": 0.0001697035040431267,
|
| 1869 |
+
"loss": 1.183,
|
| 1870 |
+
"step": 1505
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 0.48,
|
| 1874 |
+
"learning_rate": 0.0001695956873315364,
|
| 1875 |
+
"loss": 1.1687,
|
| 1876 |
+
"step": 1510
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 0.48,
|
| 1880 |
+
"learning_rate": 0.0001694878706199461,
|
| 1881 |
+
"loss": 1.1754,
|
| 1882 |
+
"step": 1515
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 0.49,
|
| 1886 |
+
"learning_rate": 0.00016938005390835581,
|
| 1887 |
+
"loss": 1.2051,
|
| 1888 |
+
"step": 1520
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.49,
|
| 1892 |
+
"learning_rate": 0.0001692722371967655,
|
| 1893 |
+
"loss": 1.1862,
|
| 1894 |
+
"step": 1525
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 0.49,
|
| 1898 |
+
"learning_rate": 0.0001691644204851752,
|
| 1899 |
+
"loss": 1.2014,
|
| 1900 |
+
"step": 1530
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 0.49,
|
| 1904 |
+
"learning_rate": 0.00016905660377358492,
|
| 1905 |
+
"loss": 1.1992,
|
| 1906 |
+
"step": 1535
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.49,
|
| 1910 |
+
"learning_rate": 0.0001689487870619946,
|
| 1911 |
+
"loss": 1.2253,
|
| 1912 |
+
"step": 1540
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 0.49,
|
| 1916 |
+
"learning_rate": 0.00016884097035040432,
|
| 1917 |
+
"loss": 1.2125,
|
| 1918 |
+
"step": 1545
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 0.5,
|
| 1922 |
+
"learning_rate": 0.00016873315363881403,
|
| 1923 |
+
"loss": 1.1838,
|
| 1924 |
+
"step": 1550
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 0.5,
|
| 1928 |
+
"learning_rate": 0.00016862533692722374,
|
| 1929 |
+
"loss": 1.2048,
|
| 1930 |
+
"step": 1555
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.5,
|
| 1934 |
+
"learning_rate": 0.00016851752021563343,
|
| 1935 |
+
"loss": 1.1855,
|
| 1936 |
+
"step": 1560
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.5,
|
| 1940 |
+
"learning_rate": 0.0001684097035040431,
|
| 1941 |
+
"loss": 1.1646,
|
| 1942 |
+
"step": 1565
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.5,
|
| 1946 |
+
"learning_rate": 0.00016830188679245285,
|
| 1947 |
+
"loss": 1.2024,
|
| 1948 |
+
"step": 1570
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.5,
|
| 1952 |
+
"learning_rate": 0.00016819407008086254,
|
| 1953 |
+
"loss": 1.2162,
|
| 1954 |
+
"step": 1575
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 0.51,
|
| 1958 |
+
"learning_rate": 0.00016808625336927225,
|
| 1959 |
+
"loss": 1.1909,
|
| 1960 |
+
"step": 1580
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.51,
|
| 1964 |
+
"learning_rate": 0.00016797843665768196,
|
| 1965 |
+
"loss": 1.1807,
|
| 1966 |
+
"step": 1585
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 0.51,
|
| 1970 |
+
"learning_rate": 0.00016787061994609164,
|
| 1971 |
+
"loss": 1.1803,
|
| 1972 |
+
"step": 1590
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.51,
|
| 1976 |
+
"learning_rate": 0.00016776280323450136,
|
| 1977 |
+
"loss": 1.1894,
|
| 1978 |
+
"step": 1595
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.51,
|
| 1982 |
+
"learning_rate": 0.00016765498652291104,
|
| 1983 |
+
"loss": 1.1869,
|
| 1984 |
+
"step": 1600
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.51,
|
| 1988 |
+
"eval_loss": 1.1610257625579834,
|
| 1989 |
+
"eval_runtime": 709.4343,
|
| 1990 |
+
"eval_samples_per_second": 7.048,
|
| 1991 |
+
"eval_steps_per_second": 0.881,
|
| 1992 |
+
"step": 1600
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 0.51,
|
| 1996 |
+
"learning_rate": 0.00016754716981132078,
|
| 1997 |
+
"loss": 1.1832,
|
| 1998 |
+
"step": 1605
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 0.52,
|
| 2002 |
+
"learning_rate": 0.00016743935309973047,
|
| 2003 |
+
"loss": 1.1553,
|
| 2004 |
+
"step": 1610
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.52,
|
| 2008 |
+
"learning_rate": 0.00016733153638814018,
|
| 2009 |
+
"loss": 1.1788,
|
| 2010 |
+
"step": 1615
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 0.52,
|
| 2014 |
+
"learning_rate": 0.0001672237196765499,
|
| 2015 |
+
"loss": 1.1896,
|
| 2016 |
+
"step": 1620
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.52,
|
| 2020 |
+
"learning_rate": 0.00016711590296495957,
|
| 2021 |
+
"loss": 1.1839,
|
| 2022 |
+
"step": 1625
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 0.52,
|
| 2026 |
+
"learning_rate": 0.00016700808625336929,
|
| 2027 |
+
"loss": 1.2063,
|
| 2028 |
+
"step": 1630
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 0.52,
|
| 2032 |
+
"learning_rate": 0.00016690026954177897,
|
| 2033 |
+
"loss": 1.1756,
|
| 2034 |
+
"step": 1635
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 0.52,
|
| 2038 |
+
"learning_rate": 0.00016679245283018868,
|
| 2039 |
+
"loss": 1.1903,
|
| 2040 |
+
"step": 1640
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 0.53,
|
| 2044 |
+
"learning_rate": 0.0001666846361185984,
|
| 2045 |
+
"loss": 1.1901,
|
| 2046 |
+
"step": 1645
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.53,
|
| 2050 |
+
"learning_rate": 0.00016657681940700808,
|
| 2051 |
+
"loss": 1.1852,
|
| 2052 |
+
"step": 1650
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 0.53,
|
| 2056 |
+
"learning_rate": 0.00016646900269541782,
|
| 2057 |
+
"loss": 1.1824,
|
| 2058 |
+
"step": 1655
|
| 2059 |
+
},
|
| 2060 |
+
{
|
| 2061 |
+
"epoch": 0.53,
|
| 2062 |
+
"learning_rate": 0.0001663611859838275,
|
| 2063 |
+
"loss": 1.1514,
|
| 2064 |
+
"step": 1660
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 0.53,
|
| 2068 |
+
"learning_rate": 0.00016625336927223721,
|
| 2069 |
+
"loss": 1.1835,
|
| 2070 |
+
"step": 1665
|
| 2071 |
+
},
|
| 2072 |
+
{
|
| 2073 |
+
"epoch": 0.53,
|
| 2074 |
+
"learning_rate": 0.0001661455525606469,
|
| 2075 |
+
"loss": 1.1564,
|
| 2076 |
+
"step": 1670
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 0.54,
|
| 2080 |
+
"learning_rate": 0.0001660377358490566,
|
| 2081 |
+
"loss": 1.2315,
|
| 2082 |
+
"step": 1675
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.54,
|
| 2086 |
+
"learning_rate": 0.00016592991913746632,
|
| 2087 |
+
"loss": 1.2016,
|
| 2088 |
+
"step": 1680
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.54,
|
| 2092 |
+
"learning_rate": 0.000165822102425876,
|
| 2093 |
+
"loss": 1.2058,
|
| 2094 |
+
"step": 1685
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 0.54,
|
| 2098 |
+
"learning_rate": 0.00016571428571428575,
|
| 2099 |
+
"loss": 1.2109,
|
| 2100 |
+
"step": 1690
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.54,
|
| 2104 |
+
"learning_rate": 0.00016560646900269543,
|
| 2105 |
+
"loss": 1.2114,
|
| 2106 |
+
"step": 1695
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 0.54,
|
| 2110 |
+
"learning_rate": 0.00016549865229110512,
|
| 2111 |
+
"loss": 1.1754,
|
| 2112 |
+
"step": 1700
|
| 2113 |
+
},
|
| 2114 |
+
{
|
| 2115 |
+
"epoch": 0.55,
|
| 2116 |
+
"learning_rate": 0.00016539083557951483,
|
| 2117 |
+
"loss": 1.1659,
|
| 2118 |
+
"step": 1705
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 0.55,
|
| 2122 |
+
"learning_rate": 0.00016528301886792454,
|
| 2123 |
+
"loss": 1.2003,
|
| 2124 |
+
"step": 1710
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.55,
|
| 2128 |
+
"learning_rate": 0.00016517520215633425,
|
| 2129 |
+
"loss": 1.2209,
|
| 2130 |
+
"step": 1715
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.55,
|
| 2134 |
+
"learning_rate": 0.00016506738544474394,
|
| 2135 |
+
"loss": 1.173,
|
| 2136 |
+
"step": 1720
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 0.55,
|
| 2140 |
+
"learning_rate": 0.00016495956873315365,
|
| 2141 |
+
"loss": 1.2318,
|
| 2142 |
+
"step": 1725
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 0.55,
|
| 2146 |
+
"learning_rate": 0.00016485175202156336,
|
| 2147 |
+
"loss": 1.2312,
|
| 2148 |
+
"step": 1730
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 0.56,
|
| 2152 |
+
"learning_rate": 0.00016474393530997304,
|
| 2153 |
+
"loss": 1.1828,
|
| 2154 |
+
"step": 1735
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 0.56,
|
| 2158 |
+
"learning_rate": 0.00016463611859838276,
|
| 2159 |
+
"loss": 1.1994,
|
| 2160 |
+
"step": 1740
|
| 2161 |
+
},
|
| 2162 |
+
{
|
| 2163 |
+
"epoch": 0.56,
|
| 2164 |
+
"learning_rate": 0.00016452830188679247,
|
| 2165 |
+
"loss": 1.1953,
|
| 2166 |
+
"step": 1745
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.56,
|
| 2170 |
+
"learning_rate": 0.00016442048517520215,
|
| 2171 |
+
"loss": 1.1821,
|
| 2172 |
+
"step": 1750
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.56,
|
| 2176 |
+
"learning_rate": 0.00016431266846361186,
|
| 2177 |
+
"loss": 1.2484,
|
| 2178 |
+
"step": 1755
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 0.56,
|
| 2182 |
+
"learning_rate": 0.00016420485175202155,
|
| 2183 |
+
"loss": 1.1838,
|
| 2184 |
+
"step": 1760
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 0.56,
|
| 2188 |
+
"learning_rate": 0.0001640970350404313,
|
| 2189 |
+
"loss": 1.1851,
|
| 2190 |
+
"step": 1765
|
| 2191 |
+
},
|
| 2192 |
+
{
|
| 2193 |
+
"epoch": 0.57,
|
| 2194 |
+
"learning_rate": 0.00016398921832884097,
|
| 2195 |
+
"loss": 1.2036,
|
| 2196 |
+
"step": 1770
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 0.57,
|
| 2200 |
+
"learning_rate": 0.00016388140161725068,
|
| 2201 |
+
"loss": 1.2104,
|
| 2202 |
+
"step": 1775
|
| 2203 |
+
},
|
| 2204 |
+
{
|
| 2205 |
+
"epoch": 0.57,
|
| 2206 |
+
"learning_rate": 0.0001637735849056604,
|
| 2207 |
+
"loss": 1.1789,
|
| 2208 |
+
"step": 1780
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.57,
|
| 2212 |
+
"learning_rate": 0.00016366576819407008,
|
| 2213 |
+
"loss": 1.2007,
|
| 2214 |
+
"step": 1785
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.57,
|
| 2218 |
+
"learning_rate": 0.0001635579514824798,
|
| 2219 |
+
"loss": 1.1963,
|
| 2220 |
+
"step": 1790
|
| 2221 |
+
},
|
| 2222 |
+
{
|
| 2223 |
+
"epoch": 0.57,
|
| 2224 |
+
"learning_rate": 0.00016345013477088948,
|
| 2225 |
+
"loss": 1.1817,
|
| 2226 |
+
"step": 1795
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 0.58,
|
| 2230 |
+
"learning_rate": 0.00016334231805929922,
|
| 2231 |
+
"loss": 1.2017,
|
| 2232 |
+
"step": 1800
|
| 2233 |
+
},
|
| 2234 |
+
{
|
| 2235 |
+
"epoch": 0.58,
|
| 2236 |
+
"eval_loss": 1.157345175743103,
|
| 2237 |
+
"eval_runtime": 708.5342,
|
| 2238 |
+
"eval_samples_per_second": 7.057,
|
| 2239 |
+
"eval_steps_per_second": 0.882,
|
| 2240 |
+
"step": 1800
|
| 2241 |
+
}
|
| 2242 |
+
],
|
| 2243 |
+
"max_steps": 9375,
|
| 2244 |
+
"num_train_epochs": 3,
|
| 2245 |
+
"total_flos": 1.072077389955072e+17,
|
| 2246 |
+
"trial_name": null,
|
| 2247 |
+
"trial_params": null
|
| 2248 |
+
}
|
checkpoint-1800/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37558cbc88ba261791fbfa1f320a80e38b6c62c51b94e8ff0512ef16db3ba9ad
|
| 3 |
+
size 3963
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 12653389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d0034d605ad777cf2436a73d44b7e2d05cc74ba8454a92397ea8256b4d3e8438
|
| 3 |
size 12653389
|
runs/May07_17-47-06_bc858596195a/events.out.tfevents.1683481716.bc858596195a.2495.0
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 sha256:45d3b5aeea2df11c47af27cd5d256f6eb961b10a262528b9c66b7cd7cd8960b2
|
| 3 |
+
size 63139
|