Upload folder using huggingface_hub
Browse files- checkpoints-v4.4/checkpoint-21504/ema.safetensors +3 -0
- checkpoints-v4.4/checkpoint-21504/model.safetensors +3 -0
- checkpoints-v4.4/checkpoint-21504/optimizer.pt +3 -0
- checkpoints-v4.4/checkpoint-21504/rng_state.pth +3 -0
- checkpoints-v4.4/checkpoint-21504/scaler.pt +3 -0
- checkpoints-v4.4/checkpoint-21504/scheduler.pt +3 -0
- checkpoints-v4.4/checkpoint-21504/trainer_state.json +1231 -0
- checkpoints-v4.4/checkpoint-21504/training_args.bin +3 -0
checkpoints-v4.4/checkpoint-21504/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ea1bcccabc28045e23de027c6f4e8cd708517cb783df5803c51fe53653a7579
|
| 3 |
+
size 24001976
|
checkpoints-v4.4/checkpoint-21504/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0953acd6a33be53ac3f23d4e1a08b76292593319eb2fe7d61cb11c5dad6fafe
|
| 3 |
+
size 24002008
|
checkpoints-v4.4/checkpoint-21504/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5e9d346fc916f047416c561e4b8fb002ce37eaede57d768bb8e46a7916550b77
|
| 3 |
+
size 48104011
|
checkpoints-v4.4/checkpoint-21504/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72e51026cf4a03a61656846384c555d5eb1ac026dd9131926cd9a7171f3c9fdf
|
| 3 |
+
size 14645
|
checkpoints-v4.4/checkpoint-21504/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12bb12075bfb1cfb108ce3c67b386c09566f641017ceab56d870955b4da75239
|
| 3 |
+
size 1383
|
checkpoints-v4.4/checkpoint-21504/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8d0f20f63f45b1b1775e9f97d9c6a1a2904fa04b7d5a41f30f11b9aef0479eb
|
| 3 |
+
size 1465
|
checkpoints-v4.4/checkpoint-21504/trainer_state.json
ADDED
|
@@ -0,0 +1,1231 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9932104752667313,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 21504,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 0.8164597153663635,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 9.018741607666016,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.7249562740325928,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 4.788314342498779,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 1.273499608039856,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 3.9647223949432373,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.7055355310440063,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 3.6936378479003906,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_bleu": 0.5077422295273598,
|
| 43 |
+
"eval_ce_loss": 2.198924214600428,
|
| 44 |
+
"eval_cos_loss": 0.4548168092557829,
|
| 45 |
+
"eval_loss": 3.5576477192308262,
|
| 46 |
+
"eval_mse_loss": 1.3587235043582306,
|
| 47 |
+
"flow/cos_sim": 0.5451831924452629,
|
| 48 |
+
"flow/improvement_ratio": 0.9661927943088148,
|
| 49 |
+
"flow/mag_ratio_mean": 0.5132733742668204,
|
| 50 |
+
"flow/mag_ratio_std": 0.23296540431236024,
|
| 51 |
+
"step": 1024
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.047295736917463395,
|
| 55 |
+
"eval_bleu": 0.5077422295273598,
|
| 56 |
+
"eval_ce_loss": 2.198924214600428,
|
| 57 |
+
"eval_cos_loss": 0.4548168092557829,
|
| 58 |
+
"eval_loss": 3.5576477192308262,
|
| 59 |
+
"eval_mse_loss": 1.3587235043582306,
|
| 60 |
+
"eval_runtime": 114.937,
|
| 61 |
+
"eval_samples_per_second": 243.551,
|
| 62 |
+
"eval_steps_per_second": 3.811,
|
| 63 |
+
"flow/cos_sim": 0.5451831924452629,
|
| 64 |
+
"flow/improvement_ratio": 0.9661927943088148,
|
| 65 |
+
"flow/mag_ratio_mean": 0.5132733742668204,
|
| 66 |
+
"flow/mag_ratio_std": 0.23296540431236024,
|
| 67 |
+
"step": 1024
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.05911967114682925,
|
| 71 |
+
"grad_norm": 0.9717119336128235,
|
| 72 |
+
"learning_rate": 0.0009967551747861387,
|
| 73 |
+
"loss": 3.523725748062134,
|
| 74 |
+
"step": 1280
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.0709436053761951,
|
| 78 |
+
"grad_norm": 0.9234328866004944,
|
| 79 |
+
"learning_rate": 0.000994232528651847,
|
| 80 |
+
"loss": 3.442486047744751,
|
| 81 |
+
"step": 1536
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.08276753960556095,
|
| 85 |
+
"grad_norm": 0.9983293414115906,
|
| 86 |
+
"learning_rate": 0.0009909945800260092,
|
| 87 |
+
"loss": 3.3732779026031494,
|
| 88 |
+
"step": 1792
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.09459147383492679,
|
| 92 |
+
"grad_norm": 0.9987860918045044,
|
| 93 |
+
"learning_rate": 0.0009870460151900522,
|
| 94 |
+
"loss": 3.3335657119750977,
|
| 95 |
+
"step": 2048
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.09459147383492679,
|
| 99 |
+
"eval_bleu": 0.5583098662491817,
|
| 100 |
+
"eval_ce_loss": 1.9428779066425481,
|
| 101 |
+
"eval_cos_loss": 0.4089699534520711,
|
| 102 |
+
"eval_loss": 3.2713746801358923,
|
| 103 |
+
"eval_mse_loss": 1.3284967718603404,
|
| 104 |
+
"flow/cos_sim": 0.5910300453231759,
|
| 105 |
+
"flow/improvement_ratio": 0.9727095938436517,
|
| 106 |
+
"flow/mag_ratio_mean": 0.5550923214791572,
|
| 107 |
+
"flow/mag_ratio_std": 0.2607482071608713,
|
| 108 |
+
"step": 2048
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.09459147383492679,
|
| 112 |
+
"eval_bleu": 0.5583098662491817,
|
| 113 |
+
"eval_ce_loss": 1.9428779066425481,
|
| 114 |
+
"eval_cos_loss": 0.4089699534520711,
|
| 115 |
+
"eval_loss": 3.2713746801358923,
|
| 116 |
+
"eval_mse_loss": 1.3284967718603404,
|
| 117 |
+
"eval_runtime": 112.8459,
|
| 118 |
+
"eval_samples_per_second": 248.064,
|
| 119 |
+
"eval_steps_per_second": 3.881,
|
| 120 |
+
"flow/cos_sim": 0.5910300453231759,
|
| 121 |
+
"flow/improvement_ratio": 0.9727095938436517,
|
| 122 |
+
"flow/mag_ratio_mean": 0.5550923214791572,
|
| 123 |
+
"flow/mag_ratio_std": 0.2607482071608713,
|
| 124 |
+
"step": 2048
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.10641540806429264,
|
| 128 |
+
"grad_norm": 0.8213550448417664,
|
| 129 |
+
"learning_rate": 0.0009823925488998885,
|
| 130 |
+
"loss": 3.3129796981811523,
|
| 131 |
+
"step": 2304
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.1182393422936585,
|
| 135 |
+
"grad_norm": 0.973803699016571,
|
| 136 |
+
"learning_rate": 0.0009770409161149525,
|
| 137 |
+
"loss": 3.2495813369750977,
|
| 138 |
+
"step": 2560
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.13006327652302435,
|
| 142 |
+
"grad_norm": 1.5047074556350708,
|
| 143 |
+
"learning_rate": 0.0009709988622506973,
|
| 144 |
+
"loss": 3.1962387561798096,
|
| 145 |
+
"step": 2816
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.1418872107523902,
|
| 149 |
+
"grad_norm": 0.9115265011787415,
|
| 150 |
+
"learning_rate": 0.000964275131968659,
|
| 151 |
+
"loss": 3.1230273246765137,
|
| 152 |
+
"step": 3072
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.1418872107523902,
|
| 156 |
+
"eval_bleu": 0.5612210129930764,
|
| 157 |
+
"eval_ce_loss": 1.8150666712625931,
|
| 158 |
+
"eval_cos_loss": 0.38208584085991393,
|
| 159 |
+
"eval_loss": 3.1303975484142565,
|
| 160 |
+
"eval_mse_loss": 1.3153308676258069,
|
| 161 |
+
"flow/cos_sim": 0.6179141614535083,
|
| 162 |
+
"flow/improvement_ratio": 0.9767382955986615,
|
| 163 |
+
"flow/mag_ratio_mean": 0.5887199336807477,
|
| 164 |
+
"flow/mag_ratio_std": 0.2682194593426299,
|
| 165 |
+
"step": 3072
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.1418872107523902,
|
| 169 |
+
"eval_bleu": 0.5612210129930764,
|
| 170 |
+
"eval_ce_loss": 1.8150666712625931,
|
| 171 |
+
"eval_cos_loss": 0.38208584085991393,
|
| 172 |
+
"eval_loss": 3.1303975484142565,
|
| 173 |
+
"eval_mse_loss": 1.3153308676258069,
|
| 174 |
+
"eval_runtime": 112.4286,
|
| 175 |
+
"eval_samples_per_second": 248.985,
|
| 176 |
+
"eval_steps_per_second": 3.896,
|
| 177 |
+
"flow/cos_sim": 0.6179141614535083,
|
| 178 |
+
"flow/improvement_ratio": 0.9767382955986615,
|
| 179 |
+
"flow/mag_ratio_mean": 0.5887199336807477,
|
| 180 |
+
"flow/mag_ratio_std": 0.2682194593426299,
|
| 181 |
+
"step": 3072
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 0.15371114498175603,
|
| 185 |
+
"grad_norm": 1.048264980316162,
|
| 186 |
+
"learning_rate": 0.0009568794565203123,
|
| 187 |
+
"loss": 3.161437749862671,
|
| 188 |
+
"step": 3328
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.1655350792111219,
|
| 192 |
+
"grad_norm": 0.8199158906936646,
|
| 193 |
+
"learning_rate": 0.0009488225396630347,
|
| 194 |
+
"loss": 3.10720157623291,
|
| 195 |
+
"step": 3584
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.17735901344048774,
|
| 199 |
+
"grad_norm": 1.1004403829574585,
|
| 200 |
+
"learning_rate": 0.0009401160421685646,
|
| 201 |
+
"loss": 3.081758975982666,
|
| 202 |
+
"step": 3840
|
| 203 |
+
},
|
| 204 |
+
{
|
| 205 |
+
"epoch": 0.18918294766985358,
|
| 206 |
+
"grad_norm": 0.8042743802070618,
|
| 207 |
+
"learning_rate": 0.0009307725649463714,
|
| 208 |
+
"loss": 3.0546185970306396,
|
| 209 |
+
"step": 4096
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.18918294766985358,
|
| 213 |
+
"eval_bleu": 0.6482649669737099,
|
| 214 |
+
"eval_ce_loss": 1.7232403153698195,
|
| 215 |
+
"eval_cos_loss": 0.3577009423138344,
|
| 216 |
+
"eval_loss": 3.022103031476339,
|
| 217 |
+
"eval_mse_loss": 1.2988627180116905,
|
| 218 |
+
"flow/cos_sim": 0.6422990591830859,
|
| 219 |
+
"flow/improvement_ratio": 0.9803378113872929,
|
| 220 |
+
"flow/mag_ratio_mean": 0.6228148394251523,
|
| 221 |
+
"flow/mag_ratio_std": 0.2741846599399227,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.18918294766985358,
|
| 226 |
+
"eval_bleu": 0.6482649669737099,
|
| 227 |
+
"eval_ce_loss": 1.7232403153698195,
|
| 228 |
+
"eval_cos_loss": 0.3577009423138344,
|
| 229 |
+
"eval_loss": 3.022103031476339,
|
| 230 |
+
"eval_mse_loss": 1.2988627180116905,
|
| 231 |
+
"eval_runtime": 114.314,
|
| 232 |
+
"eval_samples_per_second": 244.878,
|
| 233 |
+
"eval_steps_per_second": 3.832,
|
| 234 |
+
"flow/cos_sim": 0.6422990591830859,
|
| 235 |
+
"flow/improvement_ratio": 0.9803378113872929,
|
| 236 |
+
"flow/mag_ratio_mean": 0.6228148394251523,
|
| 237 |
+
"flow/mag_ratio_std": 0.2741846599399227,
|
| 238 |
+
"step": 4096
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.20100688189921945,
|
| 242 |
+
"grad_norm": 1.2677597999572754,
|
| 243 |
+
"learning_rate": 0.0009208056308063659,
|
| 244 |
+
"loss": 3.0482373237609863,
|
| 245 |
+
"step": 4352
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.2128308161285853,
|
| 249 |
+
"grad_norm": 1.2085199356079102,
|
| 250 |
+
"learning_rate": 0.0009102296648873445,
|
| 251 |
+
"loss": 3.047755241394043,
|
| 252 |
+
"step": 4608
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.22465475035795113,
|
| 256 |
+
"grad_norm": 1.0152058601379395,
|
| 257 |
+
"learning_rate": 0.0008990599737794927,
|
| 258 |
+
"loss": 2.990638494491577,
|
| 259 |
+
"step": 4864
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.236478684587317,
|
| 263 |
+
"grad_norm": 1.1251702308654785,
|
| 264 |
+
"learning_rate": 0.0008873127233711644,
|
| 265 |
+
"loss": 3.0137414932250977,
|
| 266 |
+
"step": 5120
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.236478684587317,
|
| 270 |
+
"eval_bleu": 0.5795940220557444,
|
| 271 |
+
"eval_ce_loss": 1.7088007153985707,
|
| 272 |
+
"eval_cos_loss": 0.3385225797080558,
|
| 273 |
+
"eval_loss": 2.9921626473126346,
|
| 274 |
+
"eval_mse_loss": 1.2833619313697293,
|
| 275 |
+
"flow/cos_sim": 0.6614774217888645,
|
| 276 |
+
"flow/improvement_ratio": 0.9821179653411587,
|
| 277 |
+
"flow/mag_ratio_mean": 0.637896000794624,
|
| 278 |
+
"flow/mag_ratio_std": 0.27109318641631025,
|
| 279 |
+
"step": 5120
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.236478684587317,
|
| 283 |
+
"eval_bleu": 0.5795940220557444,
|
| 284 |
+
"eval_ce_loss": 1.7088007153985707,
|
| 285 |
+
"eval_cos_loss": 0.3385225797080558,
|
| 286 |
+
"eval_loss": 2.9921626473126346,
|
| 287 |
+
"eval_mse_loss": 1.2833619313697293,
|
| 288 |
+
"eval_runtime": 117.7245,
|
| 289 |
+
"eval_samples_per_second": 237.784,
|
| 290 |
+
"eval_steps_per_second": 3.721,
|
| 291 |
+
"flow/cos_sim": 0.6614774217888645,
|
| 292 |
+
"flow/improvement_ratio": 0.9821179653411587,
|
| 293 |
+
"flow/mag_ratio_mean": 0.637896000794624,
|
| 294 |
+
"flow/mag_ratio_std": 0.27109318641631025,
|
| 295 |
+
"step": 5120
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.24830261881668284,
|
| 299 |
+
"grad_norm": 1.3458431959152222,
|
| 300 |
+
"learning_rate": 0.0008750049154520011,
|
| 301 |
+
"loss": 2.954760789871216,
|
| 302 |
+
"step": 5376
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.2601265530460487,
|
| 306 |
+
"grad_norm": 0.9947335124015808,
|
| 307 |
+
"learning_rate": 0.0008621543631062487,
|
| 308 |
+
"loss": 2.982330322265625,
|
| 309 |
+
"step": 5632
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 0.27195048727541454,
|
| 313 |
+
"grad_norm": 0.9248947501182556,
|
| 314 |
+
"learning_rate": 0.0008487796649318904,
|
| 315 |
+
"loss": 2.9677340984344482,
|
| 316 |
+
"step": 5888
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.2837744215047804,
|
| 320 |
+
"grad_norm": 1.0004583597183228,
|
| 321 |
+
"learning_rate": 0.0008349001781229053,
|
| 322 |
+
"loss": 2.931910753250122,
|
| 323 |
+
"step": 6144
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 0.2837744215047804,
|
| 327 |
+
"eval_bleu": 0.5966938923470444,
|
| 328 |
+
"eval_ce_loss": 1.6562061356082898,
|
| 329 |
+
"eval_cos_loss": 0.32249008207560675,
|
| 330 |
+
"eval_loss": 2.9227684172865462,
|
| 331 |
+
"eval_mse_loss": 1.2665622860329335,
|
| 332 |
+
"flow/cos_sim": 0.6775099191491463,
|
| 333 |
+
"flow/improvement_ratio": 0.9837169728866996,
|
| 334 |
+
"flow/mag_ratio_mean": 0.643968923587233,
|
| 335 |
+
"flow/mag_ratio_std": 0.2556537515632638,
|
| 336 |
+
"step": 6144
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.2837744215047804,
|
| 340 |
+
"eval_bleu": 0.5966938923470444,
|
| 341 |
+
"eval_ce_loss": 1.6562061356082898,
|
| 342 |
+
"eval_cos_loss": 0.32249008207560675,
|
| 343 |
+
"eval_loss": 2.9227684172865462,
|
| 344 |
+
"eval_mse_loss": 1.2665622860329335,
|
| 345 |
+
"eval_runtime": 117.1936,
|
| 346 |
+
"eval_samples_per_second": 238.861,
|
| 347 |
+
"eval_steps_per_second": 3.737,
|
| 348 |
+
"flow/cos_sim": 0.6775099191491463,
|
| 349 |
+
"flow/improvement_ratio": 0.9837169728866996,
|
| 350 |
+
"flow/mag_ratio_mean": 0.643968923587233,
|
| 351 |
+
"flow/mag_ratio_std": 0.2556537515632638,
|
| 352 |
+
"step": 6144
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.2955983557341462,
|
| 356 |
+
"grad_norm": 1.0061293840408325,
|
| 357 |
+
"learning_rate": 0.0008205359904536107,
|
| 358 |
+
"loss": 2.9485950469970703,
|
| 359 |
+
"step": 6400
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.30742228996351206,
|
| 363 |
+
"grad_norm": 1.0945794582366943,
|
| 364 |
+
"learning_rate": 0.0008057078912056363,
|
| 365 |
+
"loss": 2.9102377891540527,
|
| 366 |
+
"step": 6656
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3192462241928779,
|
| 370 |
+
"grad_norm": 0.8923175930976868,
|
| 371 |
+
"learning_rate": 0.0007904373410796086,
|
| 372 |
+
"loss": 2.9341015815734863,
|
| 373 |
+
"step": 6912
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.3310701584222438,
|
| 377 |
+
"grad_norm": 1.009055733680725,
|
| 378 |
+
"learning_rate": 0.0007747464411350876,
|
| 379 |
+
"loss": 2.9085001945495605,
|
| 380 |
+
"step": 7168
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3310701584222438,
|
| 384 |
+
"eval_bleu": 0.5952239839816952,
|
| 385 |
+
"eval_ce_loss": 1.5963479481875624,
|
| 386 |
+
"eval_cos_loss": 0.30743498025147337,
|
| 387 |
+
"eval_loss": 2.836080116224071,
|
| 388 |
+
"eval_mse_loss": 1.2397321745685246,
|
| 389 |
+
"flow/cos_sim": 0.6925650225382417,
|
| 390 |
+
"flow/improvement_ratio": 0.9844940459891541,
|
| 391 |
+
"flow/mag_ratio_mean": 0.6763553567672973,
|
| 392 |
+
"flow/mag_ratio_std": 0.26531474469187055,
|
| 393 |
+
"step": 7168
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.3310701584222438,
|
| 397 |
+
"eval_bleu": 0.5952239839816952,
|
| 398 |
+
"eval_ce_loss": 1.5963479481875624,
|
| 399 |
+
"eval_cos_loss": 0.30743498025147337,
|
| 400 |
+
"eval_loss": 2.836080116224071,
|
| 401 |
+
"eval_mse_loss": 1.2397321745685246,
|
| 402 |
+
"eval_runtime": 114.8272,
|
| 403 |
+
"eval_samples_per_second": 243.784,
|
| 404 |
+
"eval_steps_per_second": 3.814,
|
| 405 |
+
"flow/cos_sim": 0.6925650225382417,
|
| 406 |
+
"flow/improvement_ratio": 0.9844940459891541,
|
| 407 |
+
"flow/mag_ratio_mean": 0.6763553567672973,
|
| 408 |
+
"flow/mag_ratio_std": 0.26531474469187055,
|
| 409 |
+
"step": 7168
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.34289409265160964,
|
| 413 |
+
"grad_norm": 1.058870792388916,
|
| 414 |
+
"learning_rate": 0.000758657900803716,
|
| 415 |
+
"loss": 2.8502509593963623,
|
| 416 |
+
"step": 7424
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.3547180268809755,
|
| 420 |
+
"grad_norm": 1.6279115676879883,
|
| 421 |
+
"learning_rate": 0.000742195005021869,
|
| 422 |
+
"loss": 2.8677358627319336,
|
| 423 |
+
"step": 7680
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.3665419611103413,
|
| 427 |
+
"grad_norm": 0.719463586807251,
|
| 428 |
+
"learning_rate": 0.0007253815805303786,
|
| 429 |
+
"loss": 2.8696165084838867,
|
| 430 |
+
"step": 7936
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.37836589533970716,
|
| 434 |
+
"grad_norm": 1.2832756042480469,
|
| 435 |
+
"learning_rate": 0.0007082419613901028,
|
| 436 |
+
"loss": 2.8660786151885986,
|
| 437 |
+
"step": 8192
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.37836589533970716,
|
| 441 |
+
"eval_bleu": 0.5924882946880239,
|
| 442 |
+
"eval_ce_loss": 1.5985924006869259,
|
| 443 |
+
"eval_cos_loss": 0.298040063073646,
|
| 444 |
+
"eval_loss": 2.8347644716092986,
|
| 445 |
+
"eval_mse_loss": 1.2361720814008146,
|
| 446 |
+
"flow/cos_sim": 0.7019599421655751,
|
| 447 |
+
"flow/improvement_ratio": 0.985667058049816,
|
| 448 |
+
"flow/mag_ratio_mean": 0.676283470844025,
|
| 449 |
+
"flow/mag_ratio_std": 0.2555381012697742,
|
| 450 |
+
"step": 8192
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.37836589533970716,
|
| 454 |
+
"eval_bleu": 0.5924882946880239,
|
| 455 |
+
"eval_ce_loss": 1.5985924006869259,
|
| 456 |
+
"eval_cos_loss": 0.298040063073646,
|
| 457 |
+
"eval_loss": 2.8347644716092986,
|
| 458 |
+
"eval_mse_loss": 1.2361720814008146,
|
| 459 |
+
"eval_runtime": 111.5666,
|
| 460 |
+
"eval_samples_per_second": 250.908,
|
| 461 |
+
"eval_steps_per_second": 3.926,
|
| 462 |
+
"flow/cos_sim": 0.7019599421655751,
|
| 463 |
+
"flow/improvement_ratio": 0.985667058049816,
|
| 464 |
+
"flow/mag_ratio_mean": 0.676283470844025,
|
| 465 |
+
"flow/mag_ratio_std": 0.2555381012697742,
|
| 466 |
+
"step": 8192
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.390189829569073,
|
| 470 |
+
"grad_norm": 1.0193252563476562,
|
| 471 |
+
"learning_rate": 0.0006908009537632514,
|
| 472 |
+
"loss": 2.831209897994995,
|
| 473 |
+
"step": 8448
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4020137637984389,
|
| 477 |
+
"grad_norm": 0.9136300086975098,
|
| 478 |
+
"learning_rate": 0.0006730838000114403,
|
| 479 |
+
"loss": 2.838761329650879,
|
| 480 |
+
"step": 8704
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.41383769802780473,
|
| 484 |
+
"grad_norm": 1.4517526626586914,
|
| 485 |
+
"learning_rate": 0.0006551161421624341,
|
| 486 |
+
"loss": 2.8446762561798096,
|
| 487 |
+
"step": 8960
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4256616322571706,
|
| 491 |
+
"grad_norm": 1.3003419637680054,
|
| 492 |
+
"learning_rate": 0.0006369239847984517,
|
| 493 |
+
"loss": 2.815347909927368,
|
| 494 |
+
"step": 9216
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.4256616322571706,
|
| 498 |
+
"eval_bleu": 0.5888005150326592,
|
| 499 |
+
"eval_ce_loss": 1.5748343447299853,
|
| 500 |
+
"eval_cos_loss": 0.28894776742071865,
|
| 501 |
+
"eval_loss": 2.7988699277242026,
|
| 502 |
+
"eval_mse_loss": 1.2240355853076395,
|
| 503 |
+
"flow/cos_sim": 0.7110522344504318,
|
| 504 |
+
"flow/improvement_ratio": 0.9867022692068527,
|
| 505 |
+
"flow/mag_ratio_mean": 0.6932177976386188,
|
| 506 |
+
"flow/mag_ratio_std": 0.25633022032644104,
|
| 507 |
+
"step": 9216
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.4256616322571706,
|
| 511 |
+
"eval_bleu": 0.5888005150326592,
|
| 512 |
+
"eval_ce_loss": 1.5748343447299853,
|
| 513 |
+
"eval_cos_loss": 0.28894776742071865,
|
| 514 |
+
"eval_loss": 2.7988699277242026,
|
| 515 |
+
"eval_mse_loss": 1.2240355853076395,
|
| 516 |
+
"eval_runtime": 116.9194,
|
| 517 |
+
"eval_samples_per_second": 239.421,
|
| 518 |
+
"eval_steps_per_second": 3.746,
|
| 519 |
+
"flow/cos_sim": 0.7110522344504318,
|
| 520 |
+
"flow/improvement_ratio": 0.9867022692068527,
|
| 521 |
+
"flow/mag_ratio_mean": 0.6932177976386188,
|
| 522 |
+
"flow/mag_ratio_std": 0.25633022032644104,
|
| 523 |
+
"step": 9216
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.4374855664865364,
|
| 527 |
+
"grad_norm": 1.0110423564910889,
|
| 528 |
+
"learning_rate": 0.0006185336574197479,
|
| 529 |
+
"loss": 2.7822299003601074,
|
| 530 |
+
"step": 9472
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.44930950071590225,
|
| 534 |
+
"grad_norm": 0.8770432472229004,
|
| 535 |
+
"learning_rate": 0.0005999717763379407,
|
| 536 |
+
"loss": 2.7894957065582275,
|
| 537 |
+
"step": 9728
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.4611334349452681,
|
| 541 |
+
"grad_norm": 1.0478971004486084,
|
| 542 |
+
"learning_rate": 0.0005812652061542363,
|
| 543 |
+
"loss": 2.7769644260406494,
|
| 544 |
+
"step": 9984
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.472957369174634,
|
| 548 |
+
"grad_norm": 0.9663735032081604,
|
| 549 |
+
"learning_rate": 0.0005624410208783071,
|
| 550 |
+
"loss": 2.770944595336914,
|
| 551 |
+
"step": 10240
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.472957369174634,
|
| 555 |
+
"eval_bleu": 0.615012894123598,
|
| 556 |
+
"eval_ce_loss": 1.5484308441207832,
|
| 557 |
+
"eval_cos_loss": 0.28064475538523775,
|
| 558 |
+
"eval_loss": 2.7590326413716357,
|
| 559 |
+
"eval_mse_loss": 1.2106017988838562,
|
| 560 |
+
"flow/cos_sim": 0.7193552492416069,
|
| 561 |
+
"flow/improvement_ratio": 0.9875411418474973,
|
| 562 |
+
"flow/mag_ratio_mean": 0.7026482544261027,
|
| 563 |
+
"flow/mag_ratio_std": 0.24951966559369815,
|
| 564 |
+
"step": 10240
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.472957369174634,
|
| 568 |
+
"eval_bleu": 0.615012894123598,
|
| 569 |
+
"eval_ce_loss": 1.5484308441207832,
|
| 570 |
+
"eval_cos_loss": 0.28064475538523775,
|
| 571 |
+
"eval_loss": 2.7590326413716357,
|
| 572 |
+
"eval_mse_loss": 1.2106017988838562,
|
| 573 |
+
"eval_runtime": 116.4465,
|
| 574 |
+
"eval_samples_per_second": 240.394,
|
| 575 |
+
"eval_steps_per_second": 3.761,
|
| 576 |
+
"flow/cos_sim": 0.7193552492416069,
|
| 577 |
+
"flow/improvement_ratio": 0.9875411418474973,
|
| 578 |
+
"flow/mag_ratio_mean": 0.7026482544261027,
|
| 579 |
+
"flow/mag_ratio_std": 0.24951966559369815,
|
| 580 |
+
"step": 10240
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 0.48478130340399983,
|
| 584 |
+
"grad_norm": 1.2616784572601318,
|
| 585 |
+
"learning_rate": 0.0005435264647440881,
|
| 586 |
+
"loss": 2.7979249954223633,
|
| 587 |
+
"step": 10496
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 0.49660523763336567,
|
| 591 |
+
"grad_norm": 0.906895101070404,
|
| 592 |
+
"learning_rate": 0.000524548912779213,
|
| 593 |
+
"loss": 2.790898561477661,
|
| 594 |
+
"step": 10752
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.5084291718627315,
|
| 598 |
+
"grad_norm": 1.108110785484314,
|
| 599 |
+
"learning_rate": 0.0005055358311851499,
|
| 600 |
+
"loss": 2.7871291637420654,
|
| 601 |
+
"step": 11008
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.5202531060920974,
|
| 605 |
+
"grad_norm": 1.0619992017745972,
|
| 606 |
+
"learning_rate": 0.0004865147375853812,
|
| 607 |
+
"loss": 2.774048089981079,
|
| 608 |
+
"step": 11264
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.5202531060920974,
|
| 612 |
+
"eval_bleu": 0.6019788667659534,
|
| 613 |
+
"eval_ce_loss": 1.5569582646840239,
|
| 614 |
+
"eval_cos_loss": 0.2751268859119176,
|
| 615 |
+
"eval_loss": 2.763672876575766,
|
| 616 |
+
"eval_mse_loss": 1.2067146095783199,
|
| 617 |
+
"flow/cos_sim": 0.7248731139860197,
|
| 618 |
+
"flow/improvement_ratio": 0.9875338717138387,
|
| 619 |
+
"flow/mag_ratio_mean": 0.7146132612881595,
|
| 620 |
+
"flow/mag_ratio_std": 0.2529917611447099,
|
| 621 |
+
"step": 11264
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5202531060920974,
|
| 625 |
+
"eval_bleu": 0.6019788667659534,
|
| 626 |
+
"eval_ce_loss": 1.5569582646840239,
|
| 627 |
+
"eval_cos_loss": 0.2751268859119176,
|
| 628 |
+
"eval_loss": 2.763672876575766,
|
| 629 |
+
"eval_mse_loss": 1.2067146095783199,
|
| 630 |
+
"eval_runtime": 118.9502,
|
| 631 |
+
"eval_samples_per_second": 235.334,
|
| 632 |
+
"eval_steps_per_second": 3.682,
|
| 633 |
+
"flow/cos_sim": 0.7248731139860197,
|
| 634 |
+
"flow/improvement_ratio": 0.9875338717138387,
|
| 635 |
+
"flow/mag_ratio_mean": 0.7146132612881595,
|
| 636 |
+
"flow/mag_ratio_std": 0.2529917611447099,
|
| 637 |
+
"step": 11264
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 0.5320770403214632,
|
| 641 |
+
"grad_norm": 1.5396530628204346,
|
| 642 |
+
"learning_rate": 0.0004675131611991607,
|
| 643 |
+
"loss": 2.753552198410034,
|
| 644 |
+
"step": 11520
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.5439009745508291,
|
| 648 |
+
"grad_norm": 0.9192457795143127,
|
| 649 |
+
"learning_rate": 0.0004485586029984899,
|
| 650 |
+
"loss": 2.7478880882263184,
|
| 651 |
+
"step": 11776
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.5557249087801949,
|
| 655 |
+
"grad_norm": 1.0996524095535278,
|
| 656 |
+
"learning_rate": 0.00042967849590597266,
|
| 657 |
+
"loss": 2.73809552192688,
|
| 658 |
+
"step": 12032
|
| 659 |
+
},
|
| 660 |
+
{
|
| 661 |
+
"epoch": 0.5675488430095608,
|
| 662 |
+
"grad_norm": 0.9251135587692261,
|
| 663 |
+
"learning_rate": 0.0004109001650911621,
|
| 664 |
+
"loss": 2.757793664932251,
|
| 665 |
+
"step": 12288
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.5675488430095608,
|
| 669 |
+
"eval_bleu": 0.6123695589769809,
|
| 670 |
+
"eval_ce_loss": 1.540999772069661,
|
| 671 |
+
"eval_cos_loss": 0.27151700275945884,
|
| 672 |
+
"eval_loss": 2.744947803891413,
|
| 673 |
+
"eval_mse_loss": 1.203948032910421,
|
| 674 |
+
"flow/cos_sim": 0.7284829948590771,
|
| 675 |
+
"flow/improvement_ratio": 0.9874945781002306,
|
| 676 |
+
"flow/mag_ratio_mean": 0.7196056571997465,
|
| 677 |
+
"flow/mag_ratio_std": 0.25078553958162325,
|
| 678 |
+
"step": 12288
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.5675488430095608,
|
| 682 |
+
"eval_bleu": 0.6123695589769809,
|
| 683 |
+
"eval_ce_loss": 1.540999772069661,
|
| 684 |
+
"eval_cos_loss": 0.27151700275945884,
|
| 685 |
+
"eval_loss": 2.744947803891413,
|
| 686 |
+
"eval_mse_loss": 1.203948032910421,
|
| 687 |
+
"eval_runtime": 116.6151,
|
| 688 |
+
"eval_samples_per_second": 240.046,
|
| 689 |
+
"eval_steps_per_second": 3.756,
|
| 690 |
+
"flow/cos_sim": 0.7284829948590771,
|
| 691 |
+
"flow/improvement_ratio": 0.9874945781002306,
|
| 692 |
+
"flow/mag_ratio_mean": 0.7196056571997465,
|
| 693 |
+
"flow/mag_ratio_std": 0.25078553958162325,
|
| 694 |
+
"step": 12288
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 0.5793727772389267,
|
| 698 |
+
"grad_norm": 0.7880428433418274,
|
| 699 |
+
"learning_rate": 0.0003922507884228551,
|
| 700 |
+
"loss": 2.7514264583587646,
|
| 701 |
+
"step": 12544
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.5911967114682924,
|
| 705 |
+
"grad_norm": 1.0263415575027466,
|
| 706 |
+
"learning_rate": 0.00037375735713457723,
|
| 707 |
+
"loss": 2.728660821914673,
|
| 708 |
+
"step": 12800
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.6030206456976583,
|
| 712 |
+
"grad_norm": 1.3691045045852661,
|
| 713 |
+
"learning_rate": 0.00035544663676018276,
|
| 714 |
+
"loss": 2.7708632946014404,
|
| 715 |
+
"step": 13056
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.6148445799270241,
|
| 719 |
+
"grad_norm": 1.1120959520339966,
|
| 720 |
+
"learning_rate": 0.00033734512839611255,
|
| 721 |
+
"loss": 2.7620890140533447,
|
| 722 |
+
"step": 13312
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.6148445799270241,
|
| 726 |
+
"eval_bleu": 0.6120219445702544,
|
| 727 |
+
"eval_ce_loss": 1.4990101543735697,
|
| 728 |
+
"eval_cos_loss": 0.26497433351599464,
|
| 729 |
+
"eval_loss": 2.6835129650216123,
|
| 730 |
+
"eval_mse_loss": 1.1845028109202103,
|
| 731 |
+
"flow/cos_sim": 0.7350256655314197,
|
| 732 |
+
"flow/improvement_ratio": 0.9874911498805704,
|
| 733 |
+
"flow/mag_ratio_mean": 0.7202309272060655,
|
| 734 |
+
"flow/mag_ratio_std": 0.25011629386715695,
|
| 735 |
+
"step": 13312
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.6148445799270241,
|
| 739 |
+
"eval_bleu": 0.6120219445702544,
|
| 740 |
+
"eval_ce_loss": 1.4990101543735697,
|
| 741 |
+
"eval_cos_loss": 0.26497433351599464,
|
| 742 |
+
"eval_loss": 2.6835129650216123,
|
| 743 |
+
"eval_mse_loss": 1.1845028109202103,
|
| 744 |
+
"eval_runtime": 115.7143,
|
| 745 |
+
"eval_samples_per_second": 241.915,
|
| 746 |
+
"eval_steps_per_second": 3.785,
|
| 747 |
+
"flow/cos_sim": 0.7350256655314197,
|
| 748 |
+
"flow/improvement_ratio": 0.9874911498805704,
|
| 749 |
+
"flow/mag_ratio_mean": 0.7202309272060655,
|
| 750 |
+
"flow/mag_ratio_std": 0.25011629386715695,
|
| 751 |
+
"step": 13312
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.62666851415639,
|
| 755 |
+
"grad_norm": 1.2354153394699097,
|
| 756 |
+
"learning_rate": 0.0003194790303463687,
|
| 757 |
+
"loss": 2.736755847930908,
|
| 758 |
+
"step": 13568
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.6384924483857558,
|
| 762 |
+
"grad_norm": 1.015051245689392,
|
| 763 |
+
"learning_rate": 0.00030187420020572406,
|
| 764 |
+
"loss": 2.7210988998413086,
|
| 765 |
+
"step": 13824
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.6503163826151217,
|
| 769 |
+
"grad_norm": 0.7805538773536682,
|
| 770 |
+
"learning_rate": 0.00028455611743603626,
|
| 771 |
+
"loss": 2.7070095539093018,
|
| 772 |
+
"step": 14080
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.6621403168444876,
|
| 776 |
+
"grad_norm": 1.3120993375778198,
|
| 777 |
+
"learning_rate": 0.0002675498464898373,
|
| 778 |
+
"loss": 2.7322652339935303,
|
| 779 |
+
"step": 14336
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.6621403168444876,
|
| 783 |
+
"eval_bleu": 0.5479981354239509,
|
| 784 |
+
"eval_ce_loss": 1.5251544244485358,
|
| 785 |
+
"eval_cos_loss": 0.2640067928185746,
|
| 786 |
+
"eval_loss": 2.7153855798995656,
|
| 787 |
+
"eval_mse_loss": 1.1902311546345279,
|
| 788 |
+
"flow/cos_sim": 0.7359932074535928,
|
| 789 |
+
"flow/improvement_ratio": 0.9878783013722668,
|
| 790 |
+
"flow/mag_ratio_mean": 0.7280373928481585,
|
| 791 |
+
"flow/mag_ratio_std": 0.25223626611439603,
|
| 792 |
+
"step": 14336
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.6621403168444876,
|
| 796 |
+
"eval_bleu": 0.5479981354239509,
|
| 797 |
+
"eval_ce_loss": 1.5251544244485358,
|
| 798 |
+
"eval_cos_loss": 0.2640067928185746,
|
| 799 |
+
"eval_loss": 2.7153855798995656,
|
| 800 |
+
"eval_mse_loss": 1.1902311546345279,
|
| 801 |
+
"eval_runtime": 119.0169,
|
| 802 |
+
"eval_samples_per_second": 235.202,
|
| 803 |
+
"eval_steps_per_second": 3.68,
|
| 804 |
+
"flow/cos_sim": 0.7359932074535928,
|
| 805 |
+
"flow/improvement_ratio": 0.9878783013722668,
|
| 806 |
+
"flow/mag_ratio_mean": 0.7280373928481585,
|
| 807 |
+
"flow/mag_ratio_std": 0.25223626611439603,
|
| 808 |
+
"step": 14336
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.6739642510738534,
|
| 812 |
+
"grad_norm": 1.2459907531738281,
|
| 813 |
+
"learning_rate": 0.0002508800005345623,
|
| 814 |
+
"loss": 2.746056079864502,
|
| 815 |
+
"step": 14592
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.6857881853032193,
|
| 819 |
+
"grad_norm": 1.0967656373977661,
|
| 820 |
+
"learning_rate": 0.00023457070582992562,
|
| 821 |
+
"loss": 2.6974754333496094,
|
| 822 |
+
"step": 14848
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 0.6976121195325851,
|
| 826 |
+
"grad_norm": 1.05544114112854,
|
| 827 |
+
"learning_rate": 0.00021864556680999692,
|
| 828 |
+
"loss": 2.7050912380218506,
|
| 829 |
+
"step": 15104
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.709436053761951,
|
| 833 |
+
"grad_norm": 1.2396007776260376,
|
| 834 |
+
"learning_rate": 0.0002031276319205152,
|
| 835 |
+
"loss": 2.699166774749756,
|
| 836 |
+
"step": 15360
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.709436053761951,
|
| 840 |
+
"eval_bleu": 0.60911048337298,
|
| 841 |
+
"eval_ce_loss": 1.4985592939537955,
|
| 842 |
+
"eval_cos_loss": 0.26051041347795423,
|
| 843 |
+
"eval_loss": 2.6781036116216823,
|
| 844 |
+
"eval_mse_loss": 1.1795443144018791,
|
| 845 |
+
"flow/cos_sim": 0.7394895870663803,
|
| 846 |
+
"flow/improvement_ratio": 0.9873927095437158,
|
| 847 |
+
"flow/mag_ratio_mean": 0.7277950643132266,
|
| 848 |
+
"flow/mag_ratio_std": 0.24724318032667517,
|
| 849 |
+
"step": 15360
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.709436053761951,
|
| 853 |
+
"eval_bleu": 0.60911048337298,
|
| 854 |
+
"eval_ce_loss": 1.4985592939537955,
|
| 855 |
+
"eval_cos_loss": 0.26051041347795423,
|
| 856 |
+
"eval_loss": 2.6781036116216823,
|
| 857 |
+
"eval_mse_loss": 1.1795443144018791,
|
| 858 |
+
"eval_runtime": 116.4844,
|
| 859 |
+
"eval_samples_per_second": 240.315,
|
| 860 |
+
"eval_steps_per_second": 3.76,
|
| 861 |
+
"flow/cos_sim": 0.7394895870663803,
|
| 862 |
+
"flow/improvement_ratio": 0.9873927095437158,
|
| 863 |
+
"flow/mag_ratio_mean": 0.7277950643132266,
|
| 864 |
+
"flow/mag_ratio_std": 0.24724318032667517,
|
| 865 |
+
"step": 15360
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.7212599879913169,
|
| 869 |
+
"grad_norm": 1.060723900794983,
|
| 870 |
+
"learning_rate": 0.00018803936026088542,
|
| 871 |
+
"loss": 2.698255777359009,
|
| 872 |
+
"step": 15616
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7330839222206826,
|
| 876 |
+
"grad_norm": 1.1871165037155151,
|
| 877 |
+
"learning_rate": 0.00017340258907913464,
|
| 878 |
+
"loss": 2.6915080547332764,
|
| 879 |
+
"step": 15872
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.7449078564500485,
|
| 883 |
+
"grad_norm": 0.9529109597206116,
|
| 884 |
+
"learning_rate": 0.0001592385021668743,
|
| 885 |
+
"loss": 2.6958580017089844,
|
| 886 |
+
"step": 16128
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.7567317906794143,
|
| 890 |
+
"grad_norm": 1.0608938932418823,
|
| 891 |
+
"learning_rate": 0.0001455675992000087,
|
| 892 |
+
"loss": 2.685572862625122,
|
| 893 |
+
"step": 16384
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.7567317906794143,
|
| 897 |
+
"eval_bleu": 0.6033715700217472,
|
| 898 |
+
"eval_ce_loss": 1.5173689544745232,
|
| 899 |
+
"eval_cos_loss": 0.26027777413391084,
|
| 900 |
+
"eval_loss": 2.701112314990666,
|
| 901 |
+
"eval_mse_loss": 1.1837433556171313,
|
| 902 |
+
"flow/cos_sim": 0.7397222264444447,
|
| 903 |
+
"flow/improvement_ratio": 0.9886682670410365,
|
| 904 |
+
"flow/mag_ratio_mean": 0.7254251331499179,
|
| 905 |
+
"flow/mag_ratio_std": 0.24604364749911714,
|
| 906 |
+
"step": 16384
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.7567317906794143,
|
| 910 |
+
"eval_bleu": 0.6033715700217472,
|
| 911 |
+
"eval_ce_loss": 1.5173689544745232,
|
| 912 |
+
"eval_cos_loss": 0.26027777413391084,
|
| 913 |
+
"eval_loss": 2.701112314990666,
|
| 914 |
+
"eval_mse_loss": 1.1837433556171313,
|
| 915 |
+
"eval_runtime": 116.0079,
|
| 916 |
+
"eval_samples_per_second": 241.302,
|
| 917 |
+
"eval_steps_per_second": 3.776,
|
| 918 |
+
"flow/cos_sim": 0.7397222264444447,
|
| 919 |
+
"flow/improvement_ratio": 0.9886682670410365,
|
| 920 |
+
"flow/mag_ratio_mean": 0.7254251331499179,
|
| 921 |
+
"flow/mag_ratio_std": 0.24604364749911714,
|
| 922 |
+
"step": 16384
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 0.7685557249087802,
|
| 926 |
+
"grad_norm": 0.7993431687355042,
|
| 927 |
+
"learning_rate": 0.000132409666069565,
|
| 928 |
+
"loss": 2.687924861907959,
|
| 929 |
+
"step": 16640
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.780379659138146,
|
| 933 |
+
"grad_norm": 0.7456022500991821,
|
| 934 |
+
"learning_rate": 0.0001197837462455823,
|
| 935 |
+
"loss": 2.69299054145813,
|
| 936 |
+
"step": 16896
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.7922035933675119,
|
| 940 |
+
"grad_norm": 0.9615710377693176,
|
| 941 |
+
"learning_rate": 0.00010770811321550749,
|
| 942 |
+
"loss": 2.700143337249756,
|
| 943 |
+
"step": 17152
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 0.8040275275968778,
|
| 947 |
+
"grad_norm": 0.9432339072227478,
|
| 948 |
+
"learning_rate": 9.620024403698591e-05,
|
| 949 |
+
"loss": 2.7163219451904297,
|
| 950 |
+
"step": 17408
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8040275275968778,
|
| 954 |
+
"eval_bleu": 0.548308169220463,
|
| 955 |
+
"eval_ce_loss": 1.5058127282416984,
|
| 956 |
+
"eval_cos_loss": 0.25919875846197615,
|
| 957 |
+
"eval_loss": 2.6880075305564217,
|
| 958 |
+
"eval_mse_loss": 1.1821947968713769,
|
| 959 |
+
"flow/cos_sim": 0.7408012430689651,
|
| 960 |
+
"flow/improvement_ratio": 0.9881188005345053,
|
| 961 |
+
"flow/mag_ratio_mean": 0.7311118910573933,
|
| 962 |
+
"flow/mag_ratio_std": 0.24796967136941545,
|
| 963 |
+
"step": 17408
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.8040275275968778,
|
| 967 |
+
"eval_bleu": 0.548308169220463,
|
| 968 |
+
"eval_ce_loss": 1.5058127282416984,
|
| 969 |
+
"eval_cos_loss": 0.25919875846197615,
|
| 970 |
+
"eval_loss": 2.6880075305564217,
|
| 971 |
+
"eval_mse_loss": 1.1821947968713769,
|
| 972 |
+
"eval_runtime": 117.7488,
|
| 973 |
+
"eval_samples_per_second": 237.735,
|
| 974 |
+
"eval_steps_per_second": 3.72,
|
| 975 |
+
"flow/cos_sim": 0.7408012430689651,
|
| 976 |
+
"flow/improvement_ratio": 0.9881188005345053,
|
| 977 |
+
"flow/mag_ratio_mean": 0.7311118910573933,
|
| 978 |
+
"flow/mag_ratio_std": 0.24796967136941545,
|
| 979 |
+
"step": 17408
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 0.8158514618262436,
|
| 983 |
+
"grad_norm": 0.8737886548042297,
|
| 984 |
+
"learning_rate": 8.527679404332429e-05,
|
| 985 |
+
"loss": 2.6902272701263428,
|
| 986 |
+
"step": 17664
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.8276753960556095,
|
| 990 |
+
"grad_norm": 0.8381933569908142,
|
| 991 |
+
"learning_rate": 7.495357273823544e-05,
|
| 992 |
+
"loss": 2.678302764892578,
|
| 993 |
+
"step": 17920
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.8394993302849753,
|
| 997 |
+
"grad_norm": 1.1766976118087769,
|
| 998 |
+
"learning_rate": 6.524552091475183e-05,
|
| 999 |
+
"loss": 2.6962482929229736,
|
| 1000 |
+
"step": 18176
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 0.8513232645143411,
|
| 1004 |
+
"grad_norm": 1.1689873933792114,
|
| 1005 |
+
"learning_rate": 5.6166689031422024e-05,
|
| 1006 |
+
"loss": 2.6863315105438232,
|
| 1007 |
+
"step": 18432
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 0.8513232645143411,
|
| 1011 |
+
"eval_bleu": 0.545809467957221,
|
| 1012 |
+
"eval_ce_loss": 1.5024473004417331,
|
| 1013 |
+
"eval_cos_loss": 0.2577548390722166,
|
| 1014 |
+
"eval_loss": 2.6791613673510617,
|
| 1015 |
+
"eval_mse_loss": 1.1767140667732447,
|
| 1016 |
+
"flow/cos_sim": 0.742245161097888,
|
| 1017 |
+
"flow/improvement_ratio": 0.9881984092057023,
|
| 1018 |
+
"flow/mag_ratio_mean": 0.7322580480956595,
|
| 1019 |
+
"flow/mag_ratio_std": 0.24721906929527787,
|
| 1020 |
+
"step": 18432
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.8513232645143411,
|
| 1024 |
+
"eval_bleu": 0.545809467957221,
|
| 1025 |
+
"eval_ce_loss": 1.5024473004417331,
|
| 1026 |
+
"eval_cos_loss": 0.2577548390722166,
|
| 1027 |
+
"eval_loss": 2.6791613673510617,
|
| 1028 |
+
"eval_mse_loss": 1.1767140667732447,
|
| 1029 |
+
"eval_runtime": 117.9266,
|
| 1030 |
+
"eval_samples_per_second": 237.377,
|
| 1031 |
+
"eval_steps_per_second": 3.714,
|
| 1032 |
+
"flow/cos_sim": 0.742245161097888,
|
| 1033 |
+
"flow/improvement_ratio": 0.9881984092057023,
|
| 1034 |
+
"flow/mag_ratio_mean": 0.7322580480956595,
|
| 1035 |
+
"flow/mag_ratio_std": 0.24721906929527787,
|
| 1036 |
+
"step": 18432
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.8631471987437069,
|
| 1040 |
+
"grad_norm": 1.2247194051742554,
|
| 1041 |
+
"learning_rate": 4.773021687709067e-05,
|
| 1042 |
+
"loss": 2.6804656982421875,
|
| 1043 |
+
"step": 18688
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.8749711329730728,
|
| 1047 |
+
"grad_norm": 0.9656364917755127,
|
| 1048 |
+
"learning_rate": 3.994831455368719e-05,
|
| 1049 |
+
"loss": 2.6799473762512207,
|
| 1050 |
+
"step": 18944
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.8867950672024387,
|
| 1054 |
+
"grad_norm": 1.1167409420013428,
|
| 1055 |
+
"learning_rate": 3.283224480455282e-05,
|
| 1056 |
+
"loss": 2.6692895889282227,
|
| 1057 |
+
"step": 19200
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.8986190014318045,
|
| 1061 |
+
"grad_norm": 0.8527817726135254,
|
| 1062 |
+
"learning_rate": 2.639230671387627e-05,
|
| 1063 |
+
"loss": 2.6880292892456055,
|
| 1064 |
+
"step": 19456
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.8986190014318045,
|
| 1068 |
+
"eval_bleu": 0.5496708764507602,
|
| 1069 |
+
"eval_ce_loss": 1.4893978382354458,
|
| 1070 |
+
"eval_cos_loss": 0.2560059982075539,
|
| 1071 |
+
"eval_loss": 2.6582647809699247,
|
| 1072 |
+
"eval_mse_loss": 1.1688669447757338,
|
| 1073 |
+
"flow/cos_sim": 0.7439940057388724,
|
| 1074 |
+
"flow/improvement_ratio": 0.9888677304466021,
|
| 1075 |
+
"flow/mag_ratio_mean": 0.7347901796913583,
|
| 1076 |
+
"flow/mag_ratio_std": 0.2447861627288605,
|
| 1077 |
+
"step": 19456
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.8986190014318045,
|
| 1081 |
+
"eval_bleu": 0.5496708764507602,
|
| 1082 |
+
"eval_ce_loss": 1.4893978382354458,
|
| 1083 |
+
"eval_cos_loss": 0.2560059982075539,
|
| 1084 |
+
"eval_loss": 2.6582647809699247,
|
| 1085 |
+
"eval_mse_loss": 1.1688669447757338,
|
| 1086 |
+
"eval_runtime": 117.3021,
|
| 1087 |
+
"eval_samples_per_second": 238.64,
|
| 1088 |
+
"eval_steps_per_second": 3.734,
|
| 1089 |
+
"flow/cos_sim": 0.7439940057388724,
|
| 1090 |
+
"flow/improvement_ratio": 0.9888677304466021,
|
| 1091 |
+
"flow/mag_ratio_mean": 0.7347901796913583,
|
| 1092 |
+
"flow/mag_ratio_std": 0.2447861627288605,
|
| 1093 |
+
"step": 19456
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"epoch": 0.9104429356611704,
|
| 1097 |
+
"grad_norm": 0.5750219225883484,
|
| 1098 |
+
"learning_rate": 2.063782080083576e-05,
|
| 1099 |
+
"loss": 2.677205801010132,
|
| 1100 |
+
"step": 19712
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.9222668698905362,
|
| 1104 |
+
"grad_norm": 1.096099615097046,
|
| 1105 |
+
"learning_rate": 1.557711553001523e-05,
|
| 1106 |
+
"loss": 2.6789114475250244,
|
| 1107 |
+
"step": 19968
|
| 1108 |
+
},
|
| 1109 |
+
{
|
| 1110 |
+
"epoch": 0.9340908041199021,
|
| 1111 |
+
"grad_norm": 0.9802102446556091,
|
| 1112 |
+
"learning_rate": 1.1217515257622269e-05,
|
| 1113 |
+
"loss": 2.662231922149658,
|
| 1114 |
+
"step": 20224
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 0.945914738349268,
|
| 1118 |
+
"grad_norm": 0.7547997236251831,
|
| 1119 |
+
"learning_rate": 7.565329630950746e-06,
|
| 1120 |
+
"loss": 2.676410436630249,
|
| 1121 |
+
"step": 20480
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 0.945914738349268,
|
| 1125 |
+
"eval_bleu": 0.5464821752660199,
|
| 1126 |
+
"eval_ce_loss": 1.4990644795165213,
|
| 1127 |
+
"eval_cos_loss": 0.25722625508019914,
|
| 1128 |
+
"eval_loss": 2.67388042249636,
|
| 1129 |
+
"eval_mse_loss": 1.1748159417550859,
|
| 1130 |
+
"flow/cos_sim": 0.7427737461105329,
|
| 1131 |
+
"flow/improvement_ratio": 0.9887178831448838,
|
| 1132 |
+
"flow/mag_ratio_mean": 0.7328161248877713,
|
| 1133 |
+
"flow/mag_ratio_std": 0.2466802823965408,
|
| 1134 |
+
"step": 20480
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 0.945914738349268,
|
| 1138 |
+
"eval_bleu": 0.5464821752660199,
|
| 1139 |
+
"eval_ce_loss": 1.4990644795165213,
|
| 1140 |
+
"eval_cos_loss": 0.25722625508019914,
|
| 1141 |
+
"eval_loss": 2.67388042249636,
|
| 1142 |
+
"eval_mse_loss": 1.1748159417550859,
|
| 1143 |
+
"eval_runtime": 116.8427,
|
| 1144 |
+
"eval_samples_per_second": 239.578,
|
| 1145 |
+
"eval_steps_per_second": 3.749,
|
| 1146 |
+
"flow/cos_sim": 0.7427737461105329,
|
| 1147 |
+
"flow/improvement_ratio": 0.9887178831448838,
|
| 1148 |
+
"flow/mag_ratio_mean": 0.7328161248877713,
|
| 1149 |
+
"flow/mag_ratio_std": 0.2466802823965408,
|
| 1150 |
+
"step": 20480
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.9577386725786338,
|
| 1154 |
+
"grad_norm": 0.9208499193191528,
|
| 1155 |
+
"learning_rate": 4.62584445643166e-06,
|
| 1156 |
+
"loss": 2.667086362838745,
|
| 1157 |
+
"step": 20736
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.9695626068079997,
|
| 1161 |
+
"grad_norm": 0.6824806332588196,
|
| 1162 |
+
"learning_rate": 2.40331404948807e-06,
|
| 1163 |
+
"loss": 2.655881404876709,
|
| 1164 |
+
"step": 20992
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.9813865410373654,
|
| 1168 |
+
"grad_norm": 0.7420426607131958,
|
| 1169 |
+
"learning_rate": 9.009550772663965e-07,
|
| 1170 |
+
"loss": 2.674077033996582,
|
| 1171 |
+
"step": 21248
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.9932104752667313,
|
| 1175 |
+
"grad_norm": 0.6535339951515198,
|
| 1176 |
+
"learning_rate": 1.2094190315575791e-07,
|
| 1177 |
+
"loss": 2.6826326847076416,
|
| 1178 |
+
"step": 21504
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.9932104752667313,
|
| 1182 |
+
"eval_bleu": 0.5500433951110925,
|
| 1183 |
+
"eval_ce_loss": 1.4851820075620799,
|
| 1184 |
+
"eval_cos_loss": 0.25606720632614066,
|
| 1185 |
+
"eval_loss": 2.6555024687558006,
|
| 1186 |
+
"eval_mse_loss": 1.170320453709119,
|
| 1187 |
+
"flow/cos_sim": 0.7439327969398672,
|
| 1188 |
+
"flow/improvement_ratio": 0.9888444078839533,
|
| 1189 |
+
"flow/mag_ratio_mean": 0.7340902883440392,
|
| 1190 |
+
"flow/mag_ratio_std": 0.24632065518669885,
|
| 1191 |
+
"step": 21504
|
| 1192 |
+
},
|
| 1193 |
+
{
|
| 1194 |
+
"epoch": 0.9932104752667313,
|
| 1195 |
+
"eval_bleu": 0.5500433951110925,
|
| 1196 |
+
"eval_ce_loss": 1.4851820075620799,
|
| 1197 |
+
"eval_cos_loss": 0.25606720632614066,
|
| 1198 |
+
"eval_loss": 2.6555024687558006,
|
| 1199 |
+
"eval_mse_loss": 1.170320453709119,
|
| 1200 |
+
"eval_runtime": 115.8505,
|
| 1201 |
+
"eval_samples_per_second": 241.63,
|
| 1202 |
+
"eval_steps_per_second": 3.781,
|
| 1203 |
+
"flow/cos_sim": 0.7439327969398672,
|
| 1204 |
+
"flow/improvement_ratio": 0.9888444078839533,
|
| 1205 |
+
"flow/mag_ratio_mean": 0.7340902883440392,
|
| 1206 |
+
"flow/mag_ratio_std": 0.24632065518669885,
|
| 1207 |
+
"step": 21504
|
| 1208 |
+
}
|
| 1209 |
+
],
|
| 1210 |
+
"logging_steps": 256,
|
| 1211 |
+
"max_steps": 21651,
|
| 1212 |
+
"num_input_tokens_seen": 0,
|
| 1213 |
+
"num_train_epochs": 1,
|
| 1214 |
+
"save_steps": 1024,
|
| 1215 |
+
"stateful_callbacks": {
|
| 1216 |
+
"TrainerControl": {
|
| 1217 |
+
"args": {
|
| 1218 |
+
"should_epoch_stop": false,
|
| 1219 |
+
"should_evaluate": false,
|
| 1220 |
+
"should_log": false,
|
| 1221 |
+
"should_save": true,
|
| 1222 |
+
"should_training_stop": false
|
| 1223 |
+
},
|
| 1224 |
+
"attributes": {}
|
| 1225 |
+
}
|
| 1226 |
+
},
|
| 1227 |
+
"total_flos": 0.0,
|
| 1228 |
+
"train_batch_size": 64,
|
| 1229 |
+
"trial_name": null,
|
| 1230 |
+
"trial_params": null
|
| 1231 |
+
}
|
checkpoints-v4.4/checkpoint-21504/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ee9cb7ebeac20164f12c8a5bf77bbe01722a7871151494aba288a393bf7de081
|
| 3 |
+
size 5137
|