Upload folder using huggingface_hub
Browse files- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/eval_state.json +0 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/model.safetensors +3 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/optimizer.pt +3 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/rng_state.pth +3 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/scaler.pt +3 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/scheduler.pt +3 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/trainer_state.json +1538 -0
- checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/training_args.bin +3 -0
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/eval_state.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3ac77a1b34a00219f2941325e39db51e9af3764e1cda97026f63d89d9bc9c22b
|
| 3 |
+
size 15277472
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:efb42aca8a24f33181d1d4487457c8fa9caa1619df479c2a0032e7b4ed93f938
|
| 3 |
+
size 13820091
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f2c2d484f771e659bef2d8981d72c8f9967eb5a8999b2e590052563a415996ac
|
| 3 |
+
size 14645
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:12618c8ab208a7959c31d41d88aa9332181b0bd796209e3996b198705d02de7e
|
| 3 |
+
size 1383
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97c95ccf96168112b15bef5a867a0320af9b38c712db8ce53a3900f61a37b6ea
|
| 3 |
+
size 1465
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/trainer_state.json
ADDED
|
@@ -0,0 +1,1538 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2048.0,
|
| 6 |
+
"eval_steps": 64,
|
| 7 |
+
"global_step": 2048,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 16.0,
|
| 14 |
+
"grad_norm": 3.148576259613037,
|
| 15 |
+
"learning_rate": 0.000234375,
|
| 16 |
+
"loss": 0.23784185945987701,
|
| 17 |
+
"step": 16
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 32.0,
|
| 21 |
+
"grad_norm": 5.380302906036377,
|
| 22 |
+
"learning_rate": 0.000484375,
|
| 23 |
+
"loss": 0.25058040022850037,
|
| 24 |
+
"step": 32
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 48.0,
|
| 28 |
+
"grad_norm": 4.840417385101318,
|
| 29 |
+
"learning_rate": 0.000734375,
|
| 30 |
+
"loss": 0.2624973952770233,
|
| 31 |
+
"step": 48
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 64.0,
|
| 35 |
+
"grad_norm": 7.160272598266602,
|
| 36 |
+
"learning_rate": 0.000984375,
|
| 37 |
+
"loss": 0.2839544117450714,
|
| 38 |
+
"step": 64
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 64.0,
|
| 42 |
+
"eval_bleu": 0.577453197312378,
|
| 43 |
+
"eval_cos_loss": 0.06502640619874,
|
| 44 |
+
"eval_loss": 0.27716290950775146,
|
| 45 |
+
"eval_mse_loss": 0.27716290950775146,
|
| 46 |
+
"step": 64
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 64.0,
|
| 50 |
+
"eval_bleu": 0.577453197312378,
|
| 51 |
+
"eval_cos_loss": 0.06502640619874,
|
| 52 |
+
"eval_loss": 0.27716290950775146,
|
| 53 |
+
"eval_mse_loss": 0.27716290950775146,
|
| 54 |
+
"eval_runtime": 1.1951,
|
| 55 |
+
"eval_samples_per_second": 107.107,
|
| 56 |
+
"eval_steps_per_second": 1.674,
|
| 57 |
+
"step": 64
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 80.0,
|
| 61 |
+
"grad_norm": 6.120606422424316,
|
| 62 |
+
"learning_rate": 0.0009998589677226062,
|
| 63 |
+
"loss": 0.28671935200691223,
|
| 64 |
+
"step": 80
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 96.0,
|
| 68 |
+
"grad_norm": 7.4773030281066895,
|
| 69 |
+
"learning_rate": 0.0009993977281025862,
|
| 70 |
+
"loss": 0.2957626283168793,
|
| 71 |
+
"step": 96
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 112.0,
|
| 75 |
+
"grad_norm": 5.387135982513428,
|
| 76 |
+
"learning_rate": 0.000998615950346857,
|
| 77 |
+
"loss": 0.31511762738227844,
|
| 78 |
+
"step": 112
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 128.0,
|
| 82 |
+
"grad_norm": 5.680504322052002,
|
| 83 |
+
"learning_rate": 0.0009975141362390078,
|
| 84 |
+
"loss": 0.2962453067302704,
|
| 85 |
+
"step": 128
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 128.0,
|
| 89 |
+
"eval_bleu": 0.48555007255420557,
|
| 90 |
+
"eval_cos_loss": 0.07183961942791939,
|
| 91 |
+
"eval_loss": 0.3057769536972046,
|
| 92 |
+
"eval_mse_loss": 0.3057769536972046,
|
| 93 |
+
"step": 128
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 128.0,
|
| 97 |
+
"eval_bleu": 0.48555007255420557,
|
| 98 |
+
"eval_cos_loss": 0.07183961942791939,
|
| 99 |
+
"eval_loss": 0.3057769536972046,
|
| 100 |
+
"eval_mse_loss": 0.3057769536972046,
|
| 101 |
+
"eval_runtime": 0.8745,
|
| 102 |
+
"eval_samples_per_second": 146.366,
|
| 103 |
+
"eval_steps_per_second": 2.287,
|
| 104 |
+
"step": 128
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 144.0,
|
| 108 |
+
"grad_norm": 4.385462284088135,
|
| 109 |
+
"learning_rate": 0.0009960929929777704,
|
| 110 |
+
"loss": 0.29686614871025085,
|
| 111 |
+
"step": 144
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 160.0,
|
| 115 |
+
"grad_norm": 7.011136531829834,
|
| 116 |
+
"learning_rate": 0.0009943534327231042,
|
| 117 |
+
"loss": 0.2880735695362091,
|
| 118 |
+
"step": 160
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 176.0,
|
| 122 |
+
"grad_norm": 4.367685317993164,
|
| 123 |
+
"learning_rate": 0.0009922965720107278,
|
| 124 |
+
"loss": 0.2833232879638672,
|
| 125 |
+
"step": 176
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 192.0,
|
| 129 |
+
"grad_norm": 6.261765003204346,
|
| 130 |
+
"learning_rate": 0.0009899237310354695,
|
| 131 |
+
"loss": 0.29143810272216797,
|
| 132 |
+
"step": 192
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 192.0,
|
| 136 |
+
"eval_bleu": 0.5524037807785399,
|
| 137 |
+
"eval_cos_loss": 0.06608359143137932,
|
| 138 |
+
"eval_loss": 0.2805209159851074,
|
| 139 |
+
"eval_mse_loss": 0.2805209159851074,
|
| 140 |
+
"step": 192
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 192.0,
|
| 144 |
+
"eval_bleu": 0.5524037807785399,
|
| 145 |
+
"eval_cos_loss": 0.06608359143137932,
|
| 146 |
+
"eval_loss": 0.2805209159851074,
|
| 147 |
+
"eval_mse_loss": 0.2805209159851074,
|
| 148 |
+
"eval_runtime": 1.1374,
|
| 149 |
+
"eval_samples_per_second": 112.536,
|
| 150 |
+
"eval_steps_per_second": 1.758,
|
| 151 |
+
"step": 192
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 208.0,
|
| 155 |
+
"grad_norm": 6.616072654724121,
|
| 156 |
+
"learning_rate": 0.000987236432803903,
|
| 157 |
+
"loss": 0.27807363867759705,
|
| 158 |
+
"step": 208
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 224.0,
|
| 162 |
+
"grad_norm": 5.070626258850098,
|
| 163 |
+
"learning_rate": 0.0009842364021568047,
|
| 164 |
+
"loss": 0.28033965826034546,
|
| 165 |
+
"step": 224
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 240.0,
|
| 169 |
+
"grad_norm": 4.556347846984863,
|
| 170 |
+
"learning_rate": 0.000980925564662067,
|
| 171 |
+
"loss": 0.2816319465637207,
|
| 172 |
+
"step": 240
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 256.0,
|
| 176 |
+
"grad_norm": 5.806378364562988,
|
| 177 |
+
"learning_rate": 0.000977306045378773,
|
| 178 |
+
"loss": 0.27801281213760376,
|
| 179 |
+
"step": 256
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 256.0,
|
| 183 |
+
"eval_bleu": 0.5841216106014171,
|
| 184 |
+
"eval_cos_loss": 0.06353012472391129,
|
| 185 |
+
"eval_loss": 0.2689193934202194,
|
| 186 |
+
"eval_mse_loss": 0.2689193934202194,
|
| 187 |
+
"step": 256
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 256.0,
|
| 191 |
+
"eval_bleu": 0.5841216106014171,
|
| 192 |
+
"eval_cos_loss": 0.06353012472391129,
|
| 193 |
+
"eval_loss": 0.2689193934202194,
|
| 194 |
+
"eval_mse_loss": 0.2689193934202194,
|
| 195 |
+
"eval_runtime": 0.8786,
|
| 196 |
+
"eval_samples_per_second": 145.682,
|
| 197 |
+
"eval_steps_per_second": 2.276,
|
| 198 |
+
"step": 256
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 272.0,
|
| 202 |
+
"grad_norm": 6.300445079803467,
|
| 203 |
+
"learning_rate": 0.0009733801674932287,
|
| 204 |
+
"loss": 0.2830119729042053,
|
| 205 |
+
"step": 272
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 288.0,
|
| 209 |
+
"grad_norm": 5.797390937805176,
|
| 210 |
+
"learning_rate": 0.0009691504508278278,
|
| 211 |
+
"loss": 0.2906477153301239,
|
| 212 |
+
"step": 288
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 304.0,
|
| 216 |
+
"grad_norm": 3.824460744857788,
|
| 217 |
+
"learning_rate": 0.0009646196102237037,
|
| 218 |
+
"loss": 0.2848377227783203,
|
| 219 |
+
"step": 304
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 320.0,
|
| 223 |
+
"grad_norm": 4.844005584716797,
|
| 224 |
+
"learning_rate": 0.0009597905537982126,
|
| 225 |
+
"loss": 0.3038112223148346,
|
| 226 |
+
"step": 320
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 320.0,
|
| 230 |
+
"eval_bleu": 0.50877170912826,
|
| 231 |
+
"eval_cos_loss": 0.06883906945586205,
|
| 232 |
+
"eval_loss": 0.29213041067123413,
|
| 233 |
+
"eval_mse_loss": 0.29213041067123413,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 320.0,
|
| 238 |
+
"eval_bleu": 0.50877170912826,
|
| 239 |
+
"eval_cos_loss": 0.06883906945586205,
|
| 240 |
+
"eval_loss": 0.29213041067123413,
|
| 241 |
+
"eval_mse_loss": 0.29213041067123413,
|
| 242 |
+
"eval_runtime": 1.0869,
|
| 243 |
+
"eval_samples_per_second": 117.762,
|
| 244 |
+
"eval_steps_per_second": 1.84,
|
| 245 |
+
"step": 320
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 336.0,
|
| 249 |
+
"grad_norm": 4.888055801391602,
|
| 250 |
+
"learning_rate": 0.0009546663810783581,
|
| 251 |
+
"loss": 0.28091496229171753,
|
| 252 |
+
"step": 336
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 352.0,
|
| 256 |
+
"grad_norm": 4.821228504180908,
|
| 257 |
+
"learning_rate": 0.0009492503810113622,
|
| 258 |
+
"loss": 0.28097987174987793,
|
| 259 |
+
"step": 352
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 368.0,
|
| 263 |
+
"grad_norm": 4.503347873687744,
|
| 264 |
+
"learning_rate": 0.0009435460298536574,
|
| 265 |
+
"loss": 0.27521052956581116,
|
| 266 |
+
"step": 368
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 384.0,
|
| 270 |
+
"grad_norm": 3.8951001167297363,
|
| 271 |
+
"learning_rate": 0.000937556988939652,
|
| 272 |
+
"loss": 0.2731606066226959,
|
| 273 |
+
"step": 384
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 384.0,
|
| 277 |
+
"eval_bleu": 0.5467535012310109,
|
| 278 |
+
"eval_cos_loss": 0.06577872112393379,
|
| 279 |
+
"eval_loss": 0.2795831710100174,
|
| 280 |
+
"eval_mse_loss": 0.2795831710100174,
|
| 281 |
+
"step": 384
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 384.0,
|
| 285 |
+
"eval_bleu": 0.5467535012310109,
|
| 286 |
+
"eval_cos_loss": 0.06577872112393379,
|
| 287 |
+
"eval_loss": 0.2795831710100174,
|
| 288 |
+
"eval_mse_loss": 0.2795831710100174,
|
| 289 |
+
"eval_runtime": 0.8707,
|
| 290 |
+
"eval_samples_per_second": 147.005,
|
| 291 |
+
"eval_steps_per_second": 2.297,
|
| 292 |
+
"step": 384
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 400.0,
|
| 296 |
+
"grad_norm": 4.886165618896484,
|
| 297 |
+
"learning_rate": 0.0009312871023317064,
|
| 298 |
+
"loss": 0.27139905095100403,
|
| 299 |
+
"step": 400
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 416.0,
|
| 303 |
+
"grad_norm": 4.521392822265625,
|
| 304 |
+
"learning_rate": 0.0009247403943528224,
|
| 305 |
+
"loss": 0.2826443910598755,
|
| 306 |
+
"step": 416
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 432.0,
|
| 310 |
+
"grad_norm": 4.5247931480407715,
|
| 311 |
+
"learning_rate": 0.0009179210670036351,
|
| 312 |
+
"loss": 0.2796178162097931,
|
| 313 |
+
"step": 432
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 448.0,
|
| 317 |
+
"grad_norm": 6.002585411071777,
|
| 318 |
+
"learning_rate": 0.0009108334972653607,
|
| 319 |
+
"loss": 0.2687041163444519,
|
| 320 |
+
"step": 448
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 448.0,
|
| 324 |
+
"eval_bleu": 0.6011129343982055,
|
| 325 |
+
"eval_cos_loss": 0.06211286224424839,
|
| 326 |
+
"eval_loss": 0.26566024124622345,
|
| 327 |
+
"eval_mse_loss": 0.26566024124622345,
|
| 328 |
+
"step": 448
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 448.0,
|
| 332 |
+
"eval_bleu": 0.6011129343982055,
|
| 333 |
+
"eval_cos_loss": 0.06211286224424839,
|
| 334 |
+
"eval_loss": 0.26566024124622345,
|
| 335 |
+
"eval_mse_loss": 0.26566024124622345,
|
| 336 |
+
"eval_runtime": 0.8986,
|
| 337 |
+
"eval_samples_per_second": 142.45,
|
| 338 |
+
"eval_steps_per_second": 2.226,
|
| 339 |
+
"step": 448
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 464.0,
|
| 343 |
+
"grad_norm": 3.798197031021118,
|
| 344 |
+
"learning_rate": 0.0009034822342904337,
|
| 345 |
+
"loss": 0.2710013687610626,
|
| 346 |
+
"step": 464
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 480.0,
|
| 350 |
+
"grad_norm": 5.265723705291748,
|
| 351 |
+
"learning_rate": 0.0008958719964826361,
|
| 352 |
+
"loss": 0.28253716230392456,
|
| 353 |
+
"step": 480
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 496.0,
|
| 357 |
+
"grad_norm": 4.383965492248535,
|
| 358 |
+
"learning_rate": 0.0008880076684685934,
|
| 359 |
+
"loss": 0.2749737501144409,
|
| 360 |
+
"step": 496
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 512.0,
|
| 364 |
+
"grad_norm": 3.641213893890381,
|
| 365 |
+
"learning_rate": 0.0008798942979625785,
|
| 366 |
+
"loss": 0.2674626111984253,
|
| 367 |
+
"step": 512
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 512.0,
|
| 371 |
+
"eval_bleu": 0.580950657997831,
|
| 372 |
+
"eval_cos_loss": 0.06297661177814007,
|
| 373 |
+
"eval_loss": 0.2675176411867142,
|
| 374 |
+
"eval_mse_loss": 0.2675176411867142,
|
| 375 |
+
"step": 512
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 512.0,
|
| 379 |
+
"eval_bleu": 0.580950657997831,
|
| 380 |
+
"eval_cos_loss": 0.06297661177814007,
|
| 381 |
+
"eval_loss": 0.2675176411867142,
|
| 382 |
+
"eval_mse_loss": 0.2675176411867142,
|
| 383 |
+
"eval_runtime": 0.9024,
|
| 384 |
+
"eval_samples_per_second": 141.838,
|
| 385 |
+
"eval_steps_per_second": 2.216,
|
| 386 |
+
"step": 512
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 528.0,
|
| 390 |
+
"grad_norm": 4.270537376403809,
|
| 391 |
+
"learning_rate": 0.0008715370925266411,
|
| 392 |
+
"loss": 0.2660686671733856,
|
| 393 |
+
"step": 528
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 544.0,
|
| 397 |
+
"grad_norm": 4.056329727172852,
|
| 398 |
+
"learning_rate": 0.0008629414162281353,
|
| 399 |
+
"loss": 0.26294267177581787,
|
| 400 |
+
"step": 544
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 560.0,
|
| 404 |
+
"grad_norm": 5.651641368865967,
|
| 405 |
+
"learning_rate": 0.0008541127861967972,
|
| 406 |
+
"loss": 0.260432630777359,
|
| 407 |
+
"step": 560
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 576.0,
|
| 411 |
+
"grad_norm": 7.250360488891602,
|
| 412 |
+
"learning_rate": 0.0008450568690835785,
|
| 413 |
+
"loss": 0.27461671829223633,
|
| 414 |
+
"step": 576
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 576.0,
|
| 418 |
+
"eval_bleu": 0.5774127981545529,
|
| 419 |
+
"eval_cos_loss": 0.06340659596025944,
|
| 420 |
+
"eval_loss": 0.26926228404045105,
|
| 421 |
+
"eval_mse_loss": 0.26926228404045105,
|
| 422 |
+
"step": 576
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 576.0,
|
| 426 |
+
"eval_bleu": 0.5774127981545529,
|
| 427 |
+
"eval_cos_loss": 0.06340659596025944,
|
| 428 |
+
"eval_loss": 0.26926228404045105,
|
| 429 |
+
"eval_mse_loss": 0.26926228404045105,
|
| 430 |
+
"eval_runtime": 0.8733,
|
| 431 |
+
"eval_samples_per_second": 146.571,
|
| 432 |
+
"eval_steps_per_second": 2.29,
|
| 433 |
+
"step": 576
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 592.0,
|
| 437 |
+
"grad_norm": 3.2957684993743896,
|
| 438 |
+
"learning_rate": 0.0008357794774235093,
|
| 439 |
+
"loss": 0.2616701126098633,
|
| 440 |
+
"step": 592
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 608.0,
|
| 444 |
+
"grad_norm": 3.3370862007141113,
|
| 445 |
+
"learning_rate": 0.0008262865659049262,
|
| 446 |
+
"loss": 0.2559007406234741,
|
| 447 |
+
"step": 608
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 624.0,
|
| 451 |
+
"grad_norm": 4.537160396575928,
|
| 452 |
+
"learning_rate": 0.0008165842275474598,
|
| 453 |
+
"loss": 0.25834769010543823,
|
| 454 |
+
"step": 624
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 640.0,
|
| 458 |
+
"grad_norm": 6.2513580322265625,
|
| 459 |
+
"learning_rate": 0.0008066786897912324,
|
| 460 |
+
"loss": 0.2518532872200012,
|
| 461 |
+
"step": 640
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 640.0,
|
| 465 |
+
"eval_bleu": 0.6343472675167803,
|
| 466 |
+
"eval_cos_loss": 0.059439605101943016,
|
| 467 |
+
"eval_loss": 0.25308310985565186,
|
| 468 |
+
"eval_mse_loss": 0.25308310985565186,
|
| 469 |
+
"step": 640
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 640.0,
|
| 473 |
+
"eval_bleu": 0.6343472675167803,
|
| 474 |
+
"eval_cos_loss": 0.059439605101943016,
|
| 475 |
+
"eval_loss": 0.25308310985565186,
|
| 476 |
+
"eval_mse_loss": 0.25308310985565186,
|
| 477 |
+
"eval_runtime": 0.8725,
|
| 478 |
+
"eval_samples_per_second": 146.698,
|
| 479 |
+
"eval_steps_per_second": 2.292,
|
| 480 |
+
"step": 640
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 656.0,
|
| 484 |
+
"grad_norm": 4.446962833404541,
|
| 485 |
+
"learning_rate": 0.0007965763104997804,
|
| 486 |
+
"loss": 0.25539430975914,
|
| 487 |
+
"step": 656
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 672.0,
|
| 491 |
+
"grad_norm": 5.684469699859619,
|
| 492 |
+
"learning_rate": 0.0007862835738792626,
|
| 493 |
+
"loss": 0.2542829215526581,
|
| 494 |
+
"step": 672
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 688.0,
|
| 498 |
+
"grad_norm": 4.0447468757629395,
|
| 499 |
+
"learning_rate": 0.0007758070863165769,
|
| 500 |
+
"loss": 0.2549206018447876,
|
| 501 |
+
"step": 688
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 704.0,
|
| 505 |
+
"grad_norm": 5.396855354309082,
|
| 506 |
+
"learning_rate": 0.0007651535721390541,
|
| 507 |
+
"loss": 0.2627556622028351,
|
| 508 |
+
"step": 704
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 704.0,
|
| 512 |
+
"eval_bleu": 0.5813120020271406,
|
| 513 |
+
"eval_cos_loss": 0.06533811613917351,
|
| 514 |
+
"eval_loss": 0.27893321216106415,
|
| 515 |
+
"eval_mse_loss": 0.27893321216106415,
|
| 516 |
+
"step": 704
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 704.0,
|
| 520 |
+
"eval_bleu": 0.5813120020271406,
|
| 521 |
+
"eval_cos_loss": 0.06533811613917351,
|
| 522 |
+
"eval_loss": 0.27893321216106415,
|
| 523 |
+
"eval_mse_loss": 0.27893321216106415,
|
| 524 |
+
"eval_runtime": 0.8693,
|
| 525 |
+
"eval_samples_per_second": 147.241,
|
| 526 |
+
"eval_steps_per_second": 2.301,
|
| 527 |
+
"step": 704
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 720.0,
|
| 531 |
+
"grad_norm": 4.269003391265869,
|
| 532 |
+
"learning_rate": 0.0007543298692984531,
|
| 533 |
+
"loss": 0.2587549090385437,
|
| 534 |
+
"step": 720
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 736.0,
|
| 538 |
+
"grad_norm": 3.9153501987457275,
|
| 539 |
+
"learning_rate": 0.0007433429249820251,
|
| 540 |
+
"loss": 0.2565257251262665,
|
| 541 |
+
"step": 736
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 752.0,
|
| 545 |
+
"grad_norm": 5.764571666717529,
|
| 546 |
+
"learning_rate": 0.0007321997911534659,
|
| 547 |
+
"loss": 0.2554490566253662,
|
| 548 |
+
"step": 752
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 768.0,
|
| 552 |
+
"grad_norm": 4.181169509887695,
|
| 553 |
+
"learning_rate": 0.0007209076200266171,
|
| 554 |
+
"loss": 0.26280924677848816,
|
| 555 |
+
"step": 768
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 768.0,
|
| 559 |
+
"eval_bleu": 0.661364608412488,
|
| 560 |
+
"eval_cos_loss": 0.057915227487683296,
|
| 561 |
+
"eval_loss": 0.2471897453069687,
|
| 562 |
+
"eval_mse_loss": 0.2471897453069687,
|
| 563 |
+
"step": 768
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 768.0,
|
| 567 |
+
"eval_bleu": 0.661364608412488,
|
| 568 |
+
"eval_cos_loss": 0.057915227487683296,
|
| 569 |
+
"eval_loss": 0.2471897453069687,
|
| 570 |
+
"eval_mse_loss": 0.2471897453069687,
|
| 571 |
+
"eval_runtime": 0.9255,
|
| 572 |
+
"eval_samples_per_second": 138.3,
|
| 573 |
+
"eval_steps_per_second": 2.161,
|
| 574 |
+
"step": 768
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 784.0,
|
| 578 |
+
"grad_norm": 3.2729103565216064,
|
| 579 |
+
"learning_rate": 0.0007094736594748218,
|
| 580 |
+
"loss": 0.24987968802452087,
|
| 581 |
+
"step": 784
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 800.0,
|
| 585 |
+
"grad_norm": 3.1368510723114014,
|
| 586 |
+
"learning_rate": 0.0006979052483788812,
|
| 587 |
+
"loss": 0.24448727071285248,
|
| 588 |
+
"step": 800
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 816.0,
|
| 592 |
+
"grad_norm": 4.103678226470947,
|
| 593 |
+
"learning_rate": 0.0006862098119165992,
|
| 594 |
+
"loss": 0.2484358549118042,
|
| 595 |
+
"step": 816
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 832.0,
|
| 599 |
+
"grad_norm": 3.3490488529205322,
|
| 600 |
+
"learning_rate": 0.0006743948567969348,
|
| 601 |
+
"loss": 0.2439013123512268,
|
| 602 |
+
"step": 832
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 832.0,
|
| 606 |
+
"eval_bleu": 0.670957871478861,
|
| 607 |
+
"eval_cos_loss": 0.056928446516394615,
|
| 608 |
+
"eval_loss": 0.2426728531718254,
|
| 609 |
+
"eval_mse_loss": 0.2426728531718254,
|
| 610 |
+
"step": 832
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 832.0,
|
| 614 |
+
"eval_bleu": 0.670957871478861,
|
| 615 |
+
"eval_cos_loss": 0.056928446516394615,
|
| 616 |
+
"eval_loss": 0.2426728531718254,
|
| 617 |
+
"eval_mse_loss": 0.2426728531718254,
|
| 618 |
+
"eval_runtime": 0.8912,
|
| 619 |
+
"eval_samples_per_second": 143.626,
|
| 620 |
+
"eval_steps_per_second": 2.244,
|
| 621 |
+
"step": 832
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 848.0,
|
| 625 |
+
"grad_norm": 4.892444133758545,
|
| 626 |
+
"learning_rate": 0.0006624679664418274,
|
| 627 |
+
"loss": 0.2417246252298355,
|
| 628 |
+
"step": 848
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 864.0,
|
| 632 |
+
"grad_norm": 5.472821235656738,
|
| 633 |
+
"learning_rate": 0.0006504367961187803,
|
| 634 |
+
"loss": 0.24449890851974487,
|
| 635 |
+
"step": 864
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 880.0,
|
| 639 |
+
"grad_norm": 6.239707946777344,
|
| 640 |
+
"learning_rate": 0.0006383090680273322,
|
| 641 |
+
"loss": 0.2528042495250702,
|
| 642 |
+
"step": 880
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 896.0,
|
| 646 |
+
"grad_norm": 3.964495897293091,
|
| 647 |
+
"learning_rate": 0.000626092566342569,
|
| 648 |
+
"loss": 0.2635759711265564,
|
| 649 |
+
"step": 896
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 896.0,
|
| 653 |
+
"eval_bleu": 0.6793754732736595,
|
| 654 |
+
"eval_cos_loss": 0.05773136951029301,
|
| 655 |
+
"eval_loss": 0.24844638258218765,
|
| 656 |
+
"eval_mse_loss": 0.24844638258218765,
|
| 657 |
+
"step": 896
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 896.0,
|
| 661 |
+
"eval_bleu": 0.6793754732736595,
|
| 662 |
+
"eval_cos_loss": 0.05773136951029301,
|
| 663 |
+
"eval_loss": 0.24844638258218765,
|
| 664 |
+
"eval_mse_loss": 0.24844638258218765,
|
| 665 |
+
"eval_runtime": 1.1311,
|
| 666 |
+
"eval_samples_per_second": 113.16,
|
| 667 |
+
"eval_steps_per_second": 1.768,
|
| 668 |
+
"step": 896
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 912.0,
|
| 672 |
+
"grad_norm": 3.434793472290039,
|
| 673 |
+
"learning_rate": 0.0006137951322188541,
|
| 674 |
+
"loss": 0.2414565533399582,
|
| 675 |
+
"step": 912
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 928.0,
|
| 679 |
+
"grad_norm": 3.8701789379119873,
|
| 680 |
+
"learning_rate": 0.0006014246587569892,
|
| 681 |
+
"loss": 0.24117989838123322,
|
| 682 |
+
"step": 928
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 944.0,
|
| 686 |
+
"grad_norm": 4.297995567321777,
|
| 687 |
+
"learning_rate": 0.000588989085938032,
|
| 688 |
+
"loss": 0.2399703860282898,
|
| 689 |
+
"step": 944
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 960.0,
|
| 693 |
+
"grad_norm": 3.6679482460021973,
|
| 694 |
+
"learning_rate": 0.0005764963955270235,
|
| 695 |
+
"loss": 0.233713760972023,
|
| 696 |
+
"step": 960
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 960.0,
|
| 700 |
+
"eval_bleu": 0.696953364230658,
|
| 701 |
+
"eval_cos_loss": 0.054812436923384666,
|
| 702 |
+
"eval_loss": 0.23471949249505997,
|
| 703 |
+
"eval_mse_loss": 0.23471949249505997,
|
| 704 |
+
"step": 960
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 960.0,
|
| 708 |
+
"eval_bleu": 0.696953364230658,
|
| 709 |
+
"eval_cos_loss": 0.054812436923384666,
|
| 710 |
+
"eval_loss": 0.23471949249505997,
|
| 711 |
+
"eval_mse_loss": 0.23471949249505997,
|
| 712 |
+
"eval_runtime": 0.8756,
|
| 713 |
+
"eval_samples_per_second": 146.184,
|
| 714 |
+
"eval_steps_per_second": 2.284,
|
| 715 |
+
"step": 960
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 976.0,
|
| 719 |
+
"grad_norm": 4.1816887855529785,
|
| 720 |
+
"learning_rate": 0.0005639546059498979,
|
| 721 |
+
"loss": 0.2341586947441101,
|
| 722 |
+
"step": 976
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 992.0,
|
| 726 |
+
"grad_norm": 4.410841464996338,
|
| 727 |
+
"learning_rate": 0.0005513717671468594,
|
| 728 |
+
"loss": 0.23269318044185638,
|
| 729 |
+
"step": 992
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 1008.0,
|
| 733 |
+
"grad_norm": 3.832925796508789,
|
| 734 |
+
"learning_rate": 0.000538755955405534,
|
| 735 |
+
"loss": 0.2395596206188202,
|
| 736 |
+
"step": 1008
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 1024.0,
|
| 740 |
+
"grad_norm": 4.132075786590576,
|
| 741 |
+
"learning_rate": 0.0005261152681772086,
|
| 742 |
+
"loss": 0.23560243844985962,
|
| 743 |
+
"step": 1024
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 1024.0,
|
| 747 |
+
"eval_bleu": 0.7079378653405652,
|
| 748 |
+
"eval_cos_loss": 0.054174987599253654,
|
| 749 |
+
"eval_loss": 0.2315773442387581,
|
| 750 |
+
"eval_mse_loss": 0.2315773442387581,
|
| 751 |
+
"step": 1024
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1024.0,
|
| 755 |
+
"eval_bleu": 0.7079378653405652,
|
| 756 |
+
"eval_cos_loss": 0.054174987599253654,
|
| 757 |
+
"eval_loss": 0.2315773442387581,
|
| 758 |
+
"eval_mse_loss": 0.2315773442387581,
|
| 759 |
+
"eval_runtime": 1.1438,
|
| 760 |
+
"eval_samples_per_second": 111.912,
|
| 761 |
+
"eval_steps_per_second": 1.749,
|
| 762 |
+
"step": 1024
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 1040.0,
|
| 766 |
+
"grad_norm": 4.8190460205078125,
|
| 767 |
+
"learning_rate": 0.0005134578188794861,
|
| 768 |
+
"loss": 0.22934700548648834,
|
| 769 |
+
"step": 1040
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 1056.0,
|
| 773 |
+
"grad_norm": 5.013860702514648,
|
| 774 |
+
"learning_rate": 0.0005007917316886947,
|
| 775 |
+
"loss": 0.23661410808563232,
|
| 776 |
+
"step": 1056
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1072.0,
|
| 780 |
+
"grad_norm": 3.288074254989624,
|
| 781 |
+
"learning_rate": 0.00048812513632538816,
|
| 782 |
+
"loss": 0.22888951003551483,
|
| 783 |
+
"step": 1072
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 1088.0,
|
| 787 |
+
"grad_norm": 3.181110143661499,
|
| 788 |
+
"learning_rate": 0.000475466162836291,
|
| 789 |
+
"loss": 0.22868837416172028,
|
| 790 |
+
"step": 1088
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1088.0,
|
| 794 |
+
"eval_bleu": 0.7288450225502836,
|
| 795 |
+
"eval_cos_loss": 0.053315335884690285,
|
| 796 |
+
"eval_loss": 0.2283542901277542,
|
| 797 |
+
"eval_mse_loss": 0.2283542901277542,
|
| 798 |
+
"step": 1088
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 1088.0,
|
| 802 |
+
"eval_bleu": 0.7288450225502836,
|
| 803 |
+
"eval_cos_loss": 0.053315335884690285,
|
| 804 |
+
"eval_loss": 0.2283542901277542,
|
| 805 |
+
"eval_mse_loss": 0.2283542901277542,
|
| 806 |
+
"eval_runtime": 0.8923,
|
| 807 |
+
"eval_samples_per_second": 143.456,
|
| 808 |
+
"eval_steps_per_second": 2.241,
|
| 809 |
+
"step": 1088
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1104.0,
|
| 813 |
+
"grad_norm": 4.1082048416137695,
|
| 814 |
+
"learning_rate": 0.00046282293637603237,
|
| 815 |
+
"loss": 0.22575391829013824,
|
| 816 |
+
"step": 1104
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1120.0,
|
| 820 |
+
"grad_norm": 2.8534390926361084,
|
| 821 |
+
"learning_rate": 0.00045020357199201976,
|
| 822 |
+
"loss": 0.22352606058120728,
|
| 823 |
+
"step": 1120
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1136.0,
|
| 827 |
+
"grad_norm": 3.7242393493652344,
|
| 828 |
+
"learning_rate": 0.00043761616941580185,
|
| 829 |
+
"loss": 0.22356313467025757,
|
| 830 |
+
"step": 1136
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1152.0,
|
| 834 |
+
"grad_norm": 3.8674352169036865,
|
| 835 |
+
"learning_rate": 0.0004250688078642595,
|
| 836 |
+
"loss": 0.22347410023212433,
|
| 837 |
+
"step": 1152
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1152.0,
|
| 841 |
+
"eval_bleu": 0.7490331059842212,
|
| 842 |
+
"eval_cos_loss": 0.051677852869033813,
|
| 843 |
+
"eval_loss": 0.22193267196416855,
|
| 844 |
+
"eval_mse_loss": 0.22193267196416855,
|
| 845 |
+
"step": 1152
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 1152.0,
|
| 849 |
+
"eval_bleu": 0.7490331059842212,
|
| 850 |
+
"eval_cos_loss": 0.051677852869033813,
|
| 851 |
+
"eval_loss": 0.22193267196416855,
|
| 852 |
+
"eval_mse_loss": 0.22193267196416855,
|
| 853 |
+
"eval_runtime": 1.0773,
|
| 854 |
+
"eval_samples_per_second": 118.814,
|
| 855 |
+
"eval_steps_per_second": 1.856,
|
| 856 |
+
"step": 1152
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1168.0,
|
| 860 |
+
"grad_norm": 4.9404144287109375,
|
| 861 |
+
"learning_rate": 0.0004125695408539655,
|
| 862 |
+
"loss": 0.22273239493370056,
|
| 863 |
+
"step": 1168
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1184.0,
|
| 867 |
+
"grad_norm": 2.8793396949768066,
|
| 868 |
+
"learning_rate": 0.00040012639103204046,
|
| 869 |
+
"loss": 0.22370247542858124,
|
| 870 |
+
"step": 1184
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1200.0,
|
| 874 |
+
"grad_norm": 3.381091356277466,
|
| 875 |
+
"learning_rate": 0.0003877473450268202,
|
| 876 |
+
"loss": 0.2196229100227356,
|
| 877 |
+
"step": 1200
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1216.0,
|
| 881 |
+
"grad_norm": 3.1946024894714355,
|
| 882 |
+
"learning_rate": 0.0003754403483216441,
|
| 883 |
+
"loss": 0.2189682424068451,
|
| 884 |
+
"step": 1216
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1216.0,
|
| 888 |
+
"eval_bleu": 0.7228057743017877,
|
| 889 |
+
"eval_cos_loss": 0.05280923470854759,
|
| 890 |
+
"eval_loss": 0.22687970846891403,
|
| 891 |
+
"eval_mse_loss": 0.22687970846891403,
|
| 892 |
+
"step": 1216
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1216.0,
|
| 896 |
+
"eval_bleu": 0.7228057743017877,
|
| 897 |
+
"eval_cos_loss": 0.05280923470854759,
|
| 898 |
+
"eval_loss": 0.22687970846891403,
|
| 899 |
+
"eval_mse_loss": 0.22687970846891403,
|
| 900 |
+
"eval_runtime": 0.8713,
|
| 901 |
+
"eval_samples_per_second": 146.907,
|
| 902 |
+
"eval_steps_per_second": 2.295,
|
| 903 |
+
"step": 1216
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 1232.0,
|
| 907 |
+
"grad_norm": 2.8960962295532227,
|
| 908 |
+
"learning_rate": 0.0003632133001550515,
|
| 909 |
+
"loss": 0.21955418586730957,
|
| 910 |
+
"step": 1232
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 1248.0,
|
| 914 |
+
"grad_norm": 4.168556213378906,
|
| 915 |
+
"learning_rate": 0.0003510740484506588,
|
| 916 |
+
"loss": 0.21865320205688477,
|
| 917 |
+
"step": 1248
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 1264.0,
|
| 921 |
+
"grad_norm": 4.716127395629883,
|
| 922 |
+
"learning_rate": 0.0003390303847799755,
|
| 923 |
+
"loss": 0.22268880903720856,
|
| 924 |
+
"step": 1264
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 1280.0,
|
| 928 |
+
"grad_norm": 3.4384660720825195,
|
| 929 |
+
"learning_rate": 0.00032709003936138876,
|
| 930 |
+
"loss": 0.21789687871932983,
|
| 931 |
+
"step": 1280
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 1280.0,
|
| 935 |
+
"eval_bleu": 0.7668112675703869,
|
| 936 |
+
"eval_cos_loss": 0.05025000870227814,
|
| 937 |
+
"eval_loss": 0.21647433191537857,
|
| 938 |
+
"eval_mse_loss": 0.21647433191537857,
|
| 939 |
+
"step": 1280
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 1280.0,
|
| 943 |
+
"eval_bleu": 0.7668112675703869,
|
| 944 |
+
"eval_cos_loss": 0.05025000870227814,
|
| 945 |
+
"eval_loss": 0.21647433191537857,
|
| 946 |
+
"eval_mse_loss": 0.21647433191537857,
|
| 947 |
+
"eval_runtime": 0.8694,
|
| 948 |
+
"eval_samples_per_second": 147.223,
|
| 949 |
+
"eval_steps_per_second": 2.3,
|
| 950 |
+
"step": 1280
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 1296.0,
|
| 954 |
+
"grad_norm": 3.1409592628479004,
|
| 955 |
+
"learning_rate": 0.00031526067609852616,
|
| 956 |
+
"loss": 0.2181306630373001,
|
| 957 |
+
"step": 1296
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 1312.0,
|
| 961 |
+
"grad_norm": 2.3424947261810303,
|
| 962 |
+
"learning_rate": 0.00030354988766118513,
|
| 963 |
+
"loss": 0.21611051261425018,
|
| 964 |
+
"step": 1312
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 1328.0,
|
| 968 |
+
"grad_norm": 3.1680362224578857,
|
| 969 |
+
"learning_rate": 0.000291965190611981,
|
| 970 |
+
"loss": 0.2149377018213272,
|
| 971 |
+
"step": 1328
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1344.0,
|
| 975 |
+
"grad_norm": 3.9129462242126465,
|
| 976 |
+
"learning_rate": 0.00028051402058184704,
|
| 977 |
+
"loss": 0.2180047184228897,
|
| 978 |
+
"step": 1344
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1344.0,
|
| 982 |
+
"eval_bleu": 0.7675157336477649,
|
| 983 |
+
"eval_cos_loss": 0.050463948398828506,
|
| 984 |
+
"eval_loss": 0.21725857257843018,
|
| 985 |
+
"eval_mse_loss": 0.21725857257843018,
|
| 986 |
+
"step": 1344
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1344.0,
|
| 990 |
+
"eval_bleu": 0.7675157336477649,
|
| 991 |
+
"eval_cos_loss": 0.050463948398828506,
|
| 992 |
+
"eval_loss": 0.21725857257843018,
|
| 993 |
+
"eval_mse_loss": 0.21725857257843018,
|
| 994 |
+
"eval_runtime": 0.8788,
|
| 995 |
+
"eval_samples_per_second": 145.656,
|
| 996 |
+
"eval_steps_per_second": 2.276,
|
| 997 |
+
"step": 1344
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1360.0,
|
| 1001 |
+
"grad_norm": 2.6250803470611572,
|
| 1002 |
+
"learning_rate": 0.00026920372749747766,
|
| 1003 |
+
"loss": 0.2142384946346283,
|
| 1004 |
+
"step": 1360
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1376.0,
|
| 1008 |
+
"grad_norm": 2.543652057647705,
|
| 1009 |
+
"learning_rate": 0.0002580415708637841,
|
| 1010 |
+
"loss": 0.2133433222770691,
|
| 1011 |
+
"step": 1376
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1392.0,
|
| 1015 |
+
"grad_norm": 3.0058720111846924,
|
| 1016 |
+
"learning_rate": 0.0002470347151043839,
|
| 1017 |
+
"loss": 0.21205700933933258,
|
| 1018 |
+
"step": 1392
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1408.0,
|
| 1022 |
+
"grad_norm": 3.2701632976531982,
|
| 1023 |
+
"learning_rate": 0.0002361902249631207,
|
| 1024 |
+
"loss": 0.21385891735553741,
|
| 1025 |
+
"step": 1408
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1408.0,
|
| 1029 |
+
"eval_bleu": 0.7880750098541518,
|
| 1030 |
+
"eval_cos_loss": 0.049234725534915924,
|
| 1031 |
+
"eval_loss": 0.21210666000843048,
|
| 1032 |
+
"eval_mse_loss": 0.21210666000843048,
|
| 1033 |
+
"step": 1408
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1408.0,
|
| 1037 |
+
"eval_bleu": 0.7880750098541518,
|
| 1038 |
+
"eval_cos_loss": 0.049234725534915924,
|
| 1039 |
+
"eval_loss": 0.21210666000843048,
|
| 1040 |
+
"eval_mse_loss": 0.21210666000843048,
|
| 1041 |
+
"eval_runtime": 0.8893,
|
| 1042 |
+
"eval_samples_per_second": 143.935,
|
| 1043 |
+
"eval_steps_per_second": 2.249,
|
| 1044 |
+
"step": 1408
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 1424.0,
|
| 1048 |
+
"grad_norm": 3.444427967071533,
|
| 1049 |
+
"learning_rate": 0.00022551506096956222,
|
| 1050 |
+
"loss": 0.211622416973114,
|
| 1051 |
+
"step": 1424
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 1440.0,
|
| 1055 |
+
"grad_norm": 2.470618963241577,
|
| 1056 |
+
"learning_rate": 0.0002150160749713891,
|
| 1057 |
+
"loss": 0.20997348427772522,
|
| 1058 |
+
"step": 1440
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1456.0,
|
| 1062 |
+
"grad_norm": 1.7038332223892212,
|
| 1063 |
+
"learning_rate": 0.00020470000573653973,
|
| 1064 |
+
"loss": 0.20952679216861725,
|
| 1065 |
+
"step": 1456
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 1472.0,
|
| 1069 |
+
"grad_norm": 2.426316976547241,
|
| 1070 |
+
"learning_rate": 0.0001945734746279364,
|
| 1071 |
+
"loss": 0.21008668839931488,
|
| 1072 |
+
"step": 1472
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 1472.0,
|
| 1076 |
+
"eval_bleu": 0.7873788127854696,
|
| 1077 |
+
"eval_cos_loss": 0.04861448146402836,
|
| 1078 |
+
"eval_loss": 0.2095613032579422,
|
| 1079 |
+
"eval_mse_loss": 0.2095613032579422,
|
| 1080 |
+
"step": 1472
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1472.0,
|
| 1084 |
+
"eval_bleu": 0.7873788127854696,
|
| 1085 |
+
"eval_cos_loss": 0.04861448146402836,
|
| 1086 |
+
"eval_loss": 0.2095613032579422,
|
| 1087 |
+
"eval_mse_loss": 0.2095613032579422,
|
| 1088 |
+
"eval_runtime": 0.877,
|
| 1089 |
+
"eval_samples_per_second": 145.953,
|
| 1090 |
+
"eval_steps_per_second": 2.281,
|
| 1091 |
+
"step": 1472
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1488.0,
|
| 1095 |
+
"grad_norm": 2.561044216156006,
|
| 1096 |
+
"learning_rate": 0.00018464298135356777,
|
| 1097 |
+
"loss": 0.2087576985359192,
|
| 1098 |
+
"step": 1488
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 1504.0,
|
| 1102 |
+
"grad_norm": 2.8240408897399902,
|
| 1103 |
+
"learning_rate": 0.00017491489979465447,
|
| 1104 |
+
"loss": 0.2086133509874344,
|
| 1105 |
+
"step": 1504
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1520.0,
|
| 1109 |
+
"grad_norm": 2.6217775344848633,
|
| 1110 |
+
"learning_rate": 0.0001653954739145775,
|
| 1111 |
+
"loss": 0.20786544680595398,
|
| 1112 |
+
"step": 1520
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 1536.0,
|
| 1116 |
+
"grad_norm": 2.312251091003418,
|
| 1117 |
+
"learning_rate": 0.00015609081375119466,
|
| 1118 |
+
"loss": 0.20743107795715332,
|
| 1119 |
+
"step": 1536
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1536.0,
|
| 1123 |
+
"eval_bleu": 0.8002604686170025,
|
| 1124 |
+
"eval_cos_loss": 0.047877587378025055,
|
| 1125 |
+
"eval_loss": 0.2066633701324463,
|
| 1126 |
+
"eval_mse_loss": 0.2066633701324463,
|
| 1127 |
+
"step": 1536
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 1536.0,
|
| 1131 |
+
"eval_bleu": 0.8002604686170025,
|
| 1132 |
+
"eval_cos_loss": 0.047877587378025055,
|
| 1133 |
+
"eval_loss": 0.2066633701324463,
|
| 1134 |
+
"eval_mse_loss": 0.2066633701324463,
|
| 1135 |
+
"eval_runtime": 0.892,
|
| 1136 |
+
"eval_samples_per_second": 143.493,
|
| 1137 |
+
"eval_steps_per_second": 2.242,
|
| 1138 |
+
"step": 1536
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1552.0,
|
| 1142 |
+
"grad_norm": 2.6982738971710205,
|
| 1143 |
+
"learning_rate": 0.0001470068914951152,
|
| 1144 |
+
"loss": 0.2065943479537964,
|
| 1145 |
+
"step": 1552
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1568.0,
|
| 1149 |
+
"grad_norm": 2.9958996772766113,
|
| 1150 |
+
"learning_rate": 0.00013814953765645382,
|
| 1151 |
+
"loss": 0.20652760565280914,
|
| 1152 |
+
"step": 1568
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1584.0,
|
| 1156 |
+
"grad_norm": 1.8058233261108398,
|
| 1157 |
+
"learning_rate": 0.00012952443732252057,
|
| 1158 |
+
"loss": 0.20597848296165466,
|
| 1159 |
+
"step": 1584
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1600.0,
|
| 1163 |
+
"grad_norm": 1.8224921226501465,
|
| 1164 |
+
"learning_rate": 0.00012113712650885117,
|
| 1165 |
+
"loss": 0.20530889928340912,
|
| 1166 |
+
"step": 1600
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1600.0,
|
| 1170 |
+
"eval_bleu": 0.8031781855221607,
|
| 1171 |
+
"eval_cos_loss": 0.04770847223699093,
|
| 1172 |
+
"eval_loss": 0.20582260191440582,
|
| 1173 |
+
"eval_mse_loss": 0.20582260191440582,
|
| 1174 |
+
"step": 1600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 1600.0,
|
| 1178 |
+
"eval_bleu": 0.8031781855221607,
|
| 1179 |
+
"eval_cos_loss": 0.04770847223699093,
|
| 1180 |
+
"eval_loss": 0.20582260191440582,
|
| 1181 |
+
"eval_mse_loss": 0.20582260191440582,
|
| 1182 |
+
"eval_runtime": 0.9548,
|
| 1183 |
+
"eval_samples_per_second": 134.057,
|
| 1184 |
+
"eval_steps_per_second": 2.095,
|
| 1185 |
+
"step": 1600
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1616.0,
|
| 1189 |
+
"grad_norm": 2.6983439922332764,
|
| 1190 |
+
"learning_rate": 0.00011299298860591917,
|
| 1191 |
+
"loss": 0.20517539978027344,
|
| 1192 |
+
"step": 1616
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1632.0,
|
| 1196 |
+
"grad_norm": 2.2038075923919678,
|
| 1197 |
+
"learning_rate": 0.00010509725092380906,
|
| 1198 |
+
"loss": 0.2047526240348816,
|
| 1199 |
+
"step": 1632
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1648.0,
|
| 1203 |
+
"grad_norm": 1.9047884941101074,
|
| 1204 |
+
"learning_rate": 9.745498133707109e-05,
|
| 1205 |
+
"loss": 0.20422129333019257,
|
| 1206 |
+
"step": 1648
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1664.0,
|
| 1210 |
+
"grad_norm": 1.9464013576507568,
|
| 1211 |
+
"learning_rate": 9.007108503190864e-05,
|
| 1212 |
+
"loss": 0.20397356152534485,
|
| 1213 |
+
"step": 1664
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1664.0,
|
| 1217 |
+
"eval_bleu": 0.8100469239619555,
|
| 1218 |
+
"eval_cos_loss": 0.04725760594010353,
|
| 1219 |
+
"eval_loss": 0.2041708454489708,
|
| 1220 |
+
"eval_mse_loss": 0.2041708454489708,
|
| 1221 |
+
"step": 1664
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 1664.0,
|
| 1225 |
+
"eval_bleu": 0.8100469239619555,
|
| 1226 |
+
"eval_cos_loss": 0.04725760594010353,
|
| 1227 |
+
"eval_loss": 0.2041708454489708,
|
| 1228 |
+
"eval_mse_loss": 0.2041708454489708,
|
| 1229 |
+
"eval_runtime": 0.8953,
|
| 1230 |
+
"eval_samples_per_second": 142.964,
|
| 1231 |
+
"eval_steps_per_second": 2.234,
|
| 1232 |
+
"step": 1664
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 1680.0,
|
| 1236 |
+
"grad_norm": 1.9454442262649536,
|
| 1237 |
+
"learning_rate": 8.295030135778742e-05,
|
| 1238 |
+
"loss": 0.20367543399333954,
|
| 1239 |
+
"step": 1680
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 1696.0,
|
| 1243 |
+
"grad_norm": 2.6915903091430664,
|
| 1244 |
+
"learning_rate": 7.609720078548737e-05,
|
| 1245 |
+
"loss": 0.2033531665802002,
|
| 1246 |
+
"step": 1696
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 1712.0,
|
| 1250 |
+
"grad_norm": 1.5794174671173096,
|
| 1251 |
+
"learning_rate": 6.951618197354753e-05,
|
| 1252 |
+
"loss": 0.20306910574436188,
|
| 1253 |
+
"step": 1712
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 1728.0,
|
| 1257 |
+
"grad_norm": 2.828120470046997,
|
| 1258 |
+
"learning_rate": 6.321146894499102e-05,
|
| 1259 |
+
"loss": 0.2030172348022461,
|
| 1260 |
+
"step": 1728
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 1728.0,
|
| 1264 |
+
"eval_bleu": 0.8147237401837912,
|
| 1265 |
+
"eval_cos_loss": 0.046787988394498825,
|
| 1266 |
+
"eval_loss": 0.20282745361328125,
|
| 1267 |
+
"eval_mse_loss": 0.20282745361328125,
|
| 1268 |
+
"step": 1728
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 1728.0,
|
| 1272 |
+
"eval_bleu": 0.8147237401837912,
|
| 1273 |
+
"eval_cos_loss": 0.046787988394498825,
|
| 1274 |
+
"eval_loss": 0.20282745361328125,
|
| 1275 |
+
"eval_mse_loss": 0.20282745361328125,
|
| 1276 |
+
"eval_runtime": 1.1261,
|
| 1277 |
+
"eval_samples_per_second": 113.666,
|
| 1278 |
+
"eval_steps_per_second": 1.776,
|
| 1279 |
+
"step": 1728
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 1744.0,
|
| 1283 |
+
"grad_norm": 1.4509916305541992,
|
| 1284 |
+
"learning_rate": 5.7187108376137274e-05,
|
| 1285 |
+
"loss": 0.2027032971382141,
|
| 1286 |
+
"step": 1744
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 1760.0,
|
| 1290 |
+
"grad_norm": 2.41015887260437,
|
| 1291 |
+
"learning_rate": 5.144696699924578e-05,
|
| 1292 |
+
"loss": 0.2023259550333023,
|
| 1293 |
+
"step": 1760
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 1776.0,
|
| 1297 |
+
"grad_norm": 2.604257822036743,
|
| 1298 |
+
"learning_rate": 4.599472912065611e-05,
|
| 1299 |
+
"loss": 0.2021774798631668,
|
| 1300 |
+
"step": 1776
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1792.0,
|
| 1304 |
+
"grad_norm": 2.3466601371765137,
|
| 1305 |
+
"learning_rate": 4.083389425601836e-05,
|
| 1306 |
+
"loss": 0.20197318494319916,
|
| 1307 |
+
"step": 1792
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 1792.0,
|
| 1311 |
+
"eval_bleu": 0.8147862350993915,
|
| 1312 |
+
"eval_cos_loss": 0.04671955108642578,
|
| 1313 |
+
"eval_loss": 0.20188624411821365,
|
| 1314 |
+
"eval_mse_loss": 0.20188624411821365,
|
| 1315 |
+
"step": 1792
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1792.0,
|
| 1319 |
+
"eval_bleu": 0.8147862350993915,
|
| 1320 |
+
"eval_cos_loss": 0.04671955108642578,
|
| 1321 |
+
"eval_loss": 0.20188624411821365,
|
| 1322 |
+
"eval_mse_loss": 0.20188624411821365,
|
| 1323 |
+
"eval_runtime": 0.878,
|
| 1324 |
+
"eval_samples_per_second": 145.79,
|
| 1325 |
+
"eval_steps_per_second": 2.278,
|
| 1326 |
+
"step": 1792
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 1808.0,
|
| 1330 |
+
"grad_norm": 2.410977840423584,
|
| 1331 |
+
"learning_rate": 3.596777488413072e-05,
|
| 1332 |
+
"loss": 0.20183902978897095,
|
| 1333 |
+
"step": 1808
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1824.0,
|
| 1337 |
+
"grad_norm": 1.510903239250183,
|
| 1338 |
+
"learning_rate": 3.139949432082712e-05,
|
| 1339 |
+
"loss": 0.20134302973747253,
|
| 1340 |
+
"step": 1824
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 1840.0,
|
| 1344 |
+
"grad_norm": 1.8241870403289795,
|
| 1345 |
+
"learning_rate": 2.7131984714278636e-05,
|
| 1346 |
+
"loss": 0.20113661885261536,
|
| 1347 |
+
"step": 1840
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 1856.0,
|
| 1351 |
+
"grad_norm": 1.8893063068389893,
|
| 1352 |
+
"learning_rate": 2.316798516299623e-05,
|
| 1353 |
+
"loss": 0.2010827660560608,
|
| 1354 |
+
"step": 1856
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 1856.0,
|
| 1358 |
+
"eval_bleu": 0.8230810373941457,
|
| 1359 |
+
"eval_cos_loss": 0.04663713276386261,
|
| 1360 |
+
"eval_loss": 0.2013688012957573,
|
| 1361 |
+
"eval_mse_loss": 0.2013688012957573,
|
| 1362 |
+
"step": 1856
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1856.0,
|
| 1366 |
+
"eval_bleu": 0.8230810373941457,
|
| 1367 |
+
"eval_cos_loss": 0.04663713276386261,
|
| 1368 |
+
"eval_loss": 0.2013688012957573,
|
| 1369 |
+
"eval_mse_loss": 0.2013688012957573,
|
| 1370 |
+
"eval_runtime": 1.1304,
|
| 1371 |
+
"eval_samples_per_second": 113.233,
|
| 1372 |
+
"eval_steps_per_second": 1.769,
|
| 1373 |
+
"step": 1856
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 1872.0,
|
| 1377 |
+
"grad_norm": 1.3174251317977905,
|
| 1378 |
+
"learning_rate": 1.9510039957741866e-05,
|
| 1379 |
+
"loss": 0.20096558332443237,
|
| 1380 |
+
"step": 1872
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 1888.0,
|
| 1384 |
+
"grad_norm": 2.0146000385284424,
|
| 1385 |
+
"learning_rate": 1.6160496948476877e-05,
|
| 1386 |
+
"loss": 0.20089873671531677,
|
| 1387 |
+
"step": 1888
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 1904.0,
|
| 1391 |
+
"grad_norm": 1.5292428731918335,
|
| 1392 |
+
"learning_rate": 1.3121506037395925e-05,
|
| 1393 |
+
"loss": 0.20046915113925934,
|
| 1394 |
+
"step": 1904
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 1920.0,
|
| 1398 |
+
"grad_norm": 1.2913480997085571,
|
| 1399 |
+
"learning_rate": 1.0395017799013529e-05,
|
| 1400 |
+
"loss": 0.2005997598171234,
|
| 1401 |
+
"step": 1920
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 1920.0,
|
| 1405 |
+
"eval_bleu": 0.8245367073756386,
|
| 1406 |
+
"eval_cos_loss": 0.046230655163526535,
|
| 1407 |
+
"eval_loss": 0.19999407976865768,
|
| 1408 |
+
"eval_mse_loss": 0.19999407976865768,
|
| 1409 |
+
"step": 1920
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1920.0,
|
| 1413 |
+
"eval_bleu": 0.8245367073756386,
|
| 1414 |
+
"eval_cos_loss": 0.046230655163526535,
|
| 1415 |
+
"eval_loss": 0.19999407976865768,
|
| 1416 |
+
"eval_mse_loss": 0.19999407976865768,
|
| 1417 |
+
"eval_runtime": 0.8695,
|
| 1418 |
+
"eval_samples_per_second": 147.203,
|
| 1419 |
+
"eval_steps_per_second": 2.3,
|
| 1420 |
+
"step": 1920
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1936.0,
|
| 1424 |
+
"grad_norm": 1.2993054389953613,
|
| 1425 |
+
"learning_rate": 7.982782228189156e-06,
|
| 1426 |
+
"loss": 0.20030871033668518,
|
| 1427 |
+
"step": 1936
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1952.0,
|
| 1431 |
+
"grad_norm": 1.5305612087249756,
|
| 1432 |
+
"learning_rate": 5.886347616893772e-06,
|
| 1433 |
+
"loss": 0.2002657800912857,
|
| 1434 |
+
"step": 1952
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1968.0,
|
| 1438 |
+
"grad_norm": 1.21235990524292,
|
| 1439 |
+
"learning_rate": 4.107059560439718e-06,
|
| 1440 |
+
"loss": 0.2001296430826187,
|
| 1441 |
+
"step": 1968
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 1984.0,
|
| 1445 |
+
"grad_norm": 1.1770583391189575,
|
| 1446 |
+
"learning_rate": 2.646060093811187e-06,
|
| 1447 |
+
"loss": 0.20003925263881683,
|
| 1448 |
+
"step": 1984
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1984.0,
|
| 1452 |
+
"eval_bleu": 0.8271353650064853,
|
| 1453 |
+
"eval_cos_loss": 0.046304671093821526,
|
| 1454 |
+
"eval_loss": 0.19988874346017838,
|
| 1455 |
+
"eval_mse_loss": 0.19988874346017838,
|
| 1456 |
+
"step": 1984
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 1984.0,
|
| 1460 |
+
"eval_bleu": 0.8271353650064853,
|
| 1461 |
+
"eval_cos_loss": 0.046304671093821526,
|
| 1462 |
+
"eval_loss": 0.19988874346017838,
|
| 1463 |
+
"eval_mse_loss": 0.19988874346017838,
|
| 1464 |
+
"eval_runtime": 1.0168,
|
| 1465 |
+
"eval_samples_per_second": 125.883,
|
| 1466 |
+
"eval_steps_per_second": 1.967,
|
| 1467 |
+
"step": 1984
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 2000.0,
|
| 1471 |
+
"grad_norm": 1.1745353937149048,
|
| 1472 |
+
"learning_rate": 1.5042869586496811e-06,
|
| 1473 |
+
"loss": 0.19980217516422272,
|
| 1474 |
+
"step": 2000
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 2016.0,
|
| 1478 |
+
"grad_norm": 1.0710855722427368,
|
| 1479 |
+
"learning_rate": 6.824730013650915e-07,
|
| 1480 |
+
"loss": 0.19994820654392242,
|
| 1481 |
+
"step": 2016
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 2032.0,
|
| 1485 |
+
"grad_norm": 0.8578296303749084,
|
| 1486 |
+
"learning_rate": 1.8114570275867826e-07,
|
| 1487 |
+
"loss": 0.1998877078294754,
|
| 1488 |
+
"step": 2032
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 2048.0,
|
| 1492 |
+
"grad_norm": 0.7908188700675964,
|
| 1493 |
+
"learning_rate": 6.268394598385463e-10,
|
| 1494 |
+
"loss": 0.19961218535900116,
|
| 1495 |
+
"step": 2048
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 2048.0,
|
| 1499 |
+
"eval_bleu": 0.826127241114683,
|
| 1500 |
+
"eval_cos_loss": 0.04609876126050949,
|
| 1501 |
+
"eval_loss": 0.19921576976776123,
|
| 1502 |
+
"eval_mse_loss": 0.19921576976776123,
|
| 1503 |
+
"step": 2048
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 2048.0,
|
| 1507 |
+
"eval_bleu": 0.826127241114683,
|
| 1508 |
+
"eval_cos_loss": 0.04609876126050949,
|
| 1509 |
+
"eval_loss": 0.19921576976776123,
|
| 1510 |
+
"eval_mse_loss": 0.19921576976776123,
|
| 1511 |
+
"eval_runtime": 0.8731,
|
| 1512 |
+
"eval_samples_per_second": 146.603,
|
| 1513 |
+
"eval_steps_per_second": 2.291,
|
| 1514 |
+
"step": 2048
|
| 1515 |
+
}
|
| 1516 |
+
],
|
| 1517 |
+
"logging_steps": 16,
|
| 1518 |
+
"max_steps": 2048,
|
| 1519 |
+
"num_input_tokens_seen": 0,
|
| 1520 |
+
"num_train_epochs": 2048,
|
| 1521 |
+
"save_steps": 256,
|
| 1522 |
+
"stateful_callbacks": {
|
| 1523 |
+
"TrainerControl": {
|
| 1524 |
+
"args": {
|
| 1525 |
+
"should_epoch_stop": false,
|
| 1526 |
+
"should_evaluate": false,
|
| 1527 |
+
"should_log": false,
|
| 1528 |
+
"should_save": true,
|
| 1529 |
+
"should_training_stop": true
|
| 1530 |
+
},
|
| 1531 |
+
"attributes": {}
|
| 1532 |
+
}
|
| 1533 |
+
},
|
| 1534 |
+
"total_flos": 0.0,
|
| 1535 |
+
"train_batch_size": 64,
|
| 1536 |
+
"trial_name": null,
|
| 1537 |
+
"trial_params": null
|
| 1538 |
+
}
|
checkpoints-v2.0a-o-discrete-conditional/checkpoint-2048/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5937d08ef509a4ce944d3b2258b2f532be998f7a8e828dbfa14fe040f2d8bf93
|
| 3 |
+
size 5201
|