Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-d1.3/checkpoint-21504/ema.safetensors +3 -0
- checkpoints-d1.3/checkpoint-21504/eval_state.json +3 -0
- checkpoints-d1.3/checkpoint-21504/model.safetensors +3 -0
- checkpoints-d1.3/checkpoint-21504/optimizer.pt +3 -0
- checkpoints-d1.3/checkpoint-21504/rng_state.pth +3 -0
- checkpoints-d1.3/checkpoint-21504/scaler.pt +3 -0
- checkpoints-d1.3/checkpoint-21504/scheduler.pt +3 -0
- checkpoints-d1.3/checkpoint-21504/trainer_state.json +1357 -0
- checkpoints-d1.3/checkpoint-21504/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -80,3 +80,4 @@ checkpoints-v1.1-d/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lf
|
|
| 80 |
checkpoints-d2.0/checkpoint-7168/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 81 |
checkpoints-d2.0/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 82 |
checkpoints-d2.0-t1/checkpoint-17408/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 80 |
checkpoints-d2.0/checkpoint-7168/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 81 |
checkpoints-d2.0/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 82 |
checkpoints-d2.0-t1/checkpoint-17408/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 83 |
+
checkpoints-d1.3/checkpoint-21504/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-d1.3/checkpoint-21504/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:730148607533901f7fa56ad57e6365ee68a1b101f874cb7ab2e316c87913b4b7
|
| 3 |
+
size 1002792
|
checkpoints-d1.3/checkpoint-21504/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e53a85116f243a2dbf83694229fdd81647db5e9d8bd6706d734bdfae610a09dd
|
| 3 |
+
size 57313375
|
checkpoints-d1.3/checkpoint-21504/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:22a31cf25d10089e12182300ecf53c873e83c7806598bcc0299eb6a93b8520e8
|
| 3 |
+
size 24258096
|
checkpoints-d1.3/checkpoint-21504/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0bdac7203e11037a2c6325b6755540d1e80e9197c7edba95bdfda33bd4cffb3a
|
| 3 |
+
size 2083403
|
checkpoints-d1.3/checkpoint-21504/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a02a54f3812c80b81e316c43bf3d32db451c6e6118cec8ec0ecc3c0ecb21221b
|
| 3 |
+
size 14645
|
checkpoints-d1.3/checkpoint-21504/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6e8388a38d277d79f5c926801061f088da0cde6073d41fbc9f83c5bb77ecf0b
|
| 3 |
+
size 1383
|
checkpoints-d1.3/checkpoint-21504/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a42d7267ed49870bb055562ea7eea030ccd1c9e6b7e550f3ac684318037fde41
|
| 3 |
+
size 1465
|
checkpoints-d1.3/checkpoint-21504/trainer_state.json
ADDED
|
@@ -0,0 +1,1357 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.9267567392850217,
|
| 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": 0.011032818324821687,
|
| 14 |
+
"grad_norm": 11.799630165100098,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 104.40645599365234,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.022065636649643373,
|
| 21 |
+
"grad_norm": 24.033342361450195,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 2.2942047119140625,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03309845497446506,
|
| 28 |
+
"grad_norm": 21.89491081237793,
|
| 29 |
+
"learning_rate": 0.000999688448778502,
|
| 30 |
+
"loss": 2.301217555999756,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.04413127329928675,
|
| 35 |
+
"grad_norm": 24.77373695373535,
|
| 36 |
+
"learning_rate": 0.0009987492950653055,
|
| 37 |
+
"loss": 2.2358601093292236,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.04413127329928675,
|
| 42 |
+
"eval_bleu": 0.9163079254973637,
|
| 43 |
+
"eval_cos_loss": 0.8463371174930255,
|
| 44 |
+
"eval_dec_loss": 0.15604417777455437,
|
| 45 |
+
"eval_loss": 2.1683921397113597,
|
| 46 |
+
"eval_mse2_loss": 0.2577589998430789,
|
| 47 |
+
"eval_mse_loss": 1.9106331385020763,
|
| 48 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 49 |
+
"eval_var_loss": 0.01723895594080501,
|
| 50 |
+
"flow/cos_sim": 0.1536628818556444,
|
| 51 |
+
"flow/improvement_ratio": 0.5908637553898256,
|
| 52 |
+
"flow/mag_ratio_mean": 0.2672089617262517,
|
| 53 |
+
"flow/mag_ratio_std": 0.1647362987052149,
|
| 54 |
+
"step": 1024
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.04413127329928675,
|
| 58 |
+
"eval_bleu": 0.9163079254973637,
|
| 59 |
+
"eval_cos_loss": 0.8463371174930255,
|
| 60 |
+
"eval_dec_loss": 0.15604417777455437,
|
| 61 |
+
"eval_loss": 2.1683921397113597,
|
| 62 |
+
"eval_mse2_loss": 0.2577589998430789,
|
| 63 |
+
"eval_mse_loss": 1.9106331385020763,
|
| 64 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 65 |
+
"eval_runtime": 151.0019,
|
| 66 |
+
"eval_samples_per_second": 198.673,
|
| 67 |
+
"eval_steps_per_second": 3.106,
|
| 68 |
+
"eval_var_loss": 0.01723895594080501,
|
| 69 |
+
"flow/cos_sim": 0.1536628818556444,
|
| 70 |
+
"flow/improvement_ratio": 0.5908637553898256,
|
| 71 |
+
"flow/mag_ratio_mean": 0.2672089617262517,
|
| 72 |
+
"flow/mag_ratio_std": 0.1647362987052149,
|
| 73 |
+
"step": 1024
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.05516409162410843,
|
| 77 |
+
"grad_norm": 20.877023696899414,
|
| 78 |
+
"learning_rate": 0.0009971837136430763,
|
| 79 |
+
"loss": 2.162022352218628,
|
| 80 |
+
"step": 1280
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.06619690994893011,
|
| 84 |
+
"grad_norm": 16.984447479248047,
|
| 85 |
+
"learning_rate": 0.0009949936708776692,
|
| 86 |
+
"loss": 1.9889535903930664,
|
| 87 |
+
"step": 1536
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.07722972827375181,
|
| 91 |
+
"grad_norm": 18.61001205444336,
|
| 92 |
+
"learning_rate": 0.0009921819174566252,
|
| 93 |
+
"loss": 1.8391456604003906,
|
| 94 |
+
"step": 1792
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.0882625465985735,
|
| 98 |
+
"grad_norm": 9.276031494140625,
|
| 99 |
+
"learning_rate": 0.000988751984934317,
|
| 100 |
+
"loss": 1.736743688583374,
|
| 101 |
+
"step": 2048
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.0882625465985735,
|
| 105 |
+
"eval_bleu": 0.9323255434649543,
|
| 106 |
+
"eval_cos_loss": 0.5421419106821007,
|
| 107 |
+
"eval_dec_loss": 0.12311239358681098,
|
| 108 |
+
"eval_loss": 1.6836593265472444,
|
| 109 |
+
"eval_mse2_loss": 0.19289209935138982,
|
| 110 |
+
"eval_mse_loss": 1.4907672278154125,
|
| 111 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 112 |
+
"eval_var_loss": 0.01723895594080501,
|
| 113 |
+
"flow/cos_sim": 0.4578580862677682,
|
| 114 |
+
"flow/improvement_ratio": 0.8665720767049647,
|
| 115 |
+
"flow/mag_ratio_mean": 0.49134117085288076,
|
| 116 |
+
"flow/mag_ratio_std": 0.21147727505611713,
|
| 117 |
+
"step": 2048
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.0882625465985735,
|
| 121 |
+
"eval_bleu": 0.9323255434649543,
|
| 122 |
+
"eval_cos_loss": 0.5421419106821007,
|
| 123 |
+
"eval_dec_loss": 0.12311239358681098,
|
| 124 |
+
"eval_loss": 1.6836593265472444,
|
| 125 |
+
"eval_mse2_loss": 0.19289209935138982,
|
| 126 |
+
"eval_mse_loss": 1.4907672278154125,
|
| 127 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 128 |
+
"eval_runtime": 148.8988,
|
| 129 |
+
"eval_samples_per_second": 201.479,
|
| 130 |
+
"eval_steps_per_second": 3.15,
|
| 131 |
+
"eval_var_loss": 0.01723895594080501,
|
| 132 |
+
"flow/cos_sim": 0.4578580862677682,
|
| 133 |
+
"flow/improvement_ratio": 0.8665720767049647,
|
| 134 |
+
"flow/mag_ratio_mean": 0.49134117085288076,
|
| 135 |
+
"flow/mag_ratio_std": 0.21147727505611713,
|
| 136 |
+
"step": 2048
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.09929536492339518,
|
| 140 |
+
"grad_norm": 7.179141044616699,
|
| 141 |
+
"learning_rate": 0.0009847081812963268,
|
| 142 |
+
"loss": 1.6638906002044678,
|
| 143 |
+
"step": 2304
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.11032818324821686,
|
| 147 |
+
"grad_norm": 5.170185089111328,
|
| 148 |
+
"learning_rate": 0.0009800555855486275,
|
| 149 |
+
"loss": 1.6207516193389893,
|
| 150 |
+
"step": 2560
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.12136100157303854,
|
| 154 |
+
"grad_norm": 5.038082122802734,
|
| 155 |
+
"learning_rate": 0.0009748000413383664,
|
| 156 |
+
"loss": 1.5859862565994263,
|
| 157 |
+
"step": 2816
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.13239381989786023,
|
| 161 |
+
"grad_norm": 3.0588953495025635,
|
| 162 |
+
"learning_rate": 0.0009689481496142604,
|
| 163 |
+
"loss": 1.5566623210906982,
|
| 164 |
+
"step": 3072
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.13239381989786023,
|
| 168 |
+
"eval_bleu": 0.9326258807918764,
|
| 169 |
+
"eval_cos_loss": 0.48918826811349214,
|
| 170 |
+
"eval_dec_loss": 0.12837675790082037,
|
| 171 |
+
"eval_loss": 1.5384274639808801,
|
| 172 |
+
"eval_mse2_loss": 0.17169118979211045,
|
| 173 |
+
"eval_mse_loss": 1.3667362735215536,
|
| 174 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 175 |
+
"eval_var_loss": 0.01723895594080501,
|
| 176 |
+
"flow/cos_sim": 0.5108117323313186,
|
| 177 |
+
"flow/improvement_ratio": 0.8871580227605824,
|
| 178 |
+
"flow/mag_ratio_mean": 0.5184402796886624,
|
| 179 |
+
"flow/mag_ratio_std": 0.24528262540221468,
|
| 180 |
+
"step": 3072
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.13239381989786023,
|
| 184 |
+
"eval_bleu": 0.9326258807918764,
|
| 185 |
+
"eval_cos_loss": 0.48918826811349214,
|
| 186 |
+
"eval_dec_loss": 0.12837675790082037,
|
| 187 |
+
"eval_loss": 1.5384274639808801,
|
| 188 |
+
"eval_mse2_loss": 0.17169118979211045,
|
| 189 |
+
"eval_mse_loss": 1.3667362735215536,
|
| 190 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 191 |
+
"eval_runtime": 149.1668,
|
| 192 |
+
"eval_samples_per_second": 201.117,
|
| 193 |
+
"eval_steps_per_second": 3.144,
|
| 194 |
+
"eval_var_loss": 0.01723895594080501,
|
| 195 |
+
"flow/cos_sim": 0.5108117323313186,
|
| 196 |
+
"flow/improvement_ratio": 0.8871580227605824,
|
| 197 |
+
"flow/mag_ratio_mean": 0.5184402796886624,
|
| 198 |
+
"flow/mag_ratio_std": 0.24528262540221468,
|
| 199 |
+
"step": 3072
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.14342663822268192,
|
| 203 |
+
"grad_norm": 2.7404696941375732,
|
| 204 |
+
"learning_rate": 0.0009625072603358231,
|
| 205 |
+
"loss": 1.5345162153244019,
|
| 206 |
+
"step": 3328
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.15445945654750362,
|
| 210 |
+
"grad_norm": 2.8666281700134277,
|
| 211 |
+
"learning_rate": 0.0009554854632418371,
|
| 212 |
+
"loss": 1.518531084060669,
|
| 213 |
+
"step": 3584
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.1654922748723253,
|
| 217 |
+
"grad_norm": 1.7333565950393677,
|
| 218 |
+
"learning_rate": 0.000947891577689663,
|
| 219 |
+
"loss": 1.5072828531265259,
|
| 220 |
+
"step": 3840
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.176525093197147,
|
| 224 |
+
"grad_norm": 2.1371614933013916,
|
| 225 |
+
"learning_rate": 0.0009397351415781539,
|
| 226 |
+
"loss": 1.4991329908370972,
|
| 227 |
+
"step": 4096
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.176525093197147,
|
| 231 |
+
"eval_bleu": 0.9359739466567191,
|
| 232 |
+
"eval_cos_loss": 0.47476763149568524,
|
| 233 |
+
"eval_dec_loss": 0.11913440864024831,
|
| 234 |
+
"eval_loss": 1.4943630281033546,
|
| 235 |
+
"eval_mse2_loss": 0.16356163808722485,
|
| 236 |
+
"eval_mse_loss": 1.3308013893647996,
|
| 237 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 238 |
+
"eval_var_loss": 0.01723895594080501,
|
| 239 |
+
"flow/cos_sim": 0.5252323670427936,
|
| 240 |
+
"flow/improvement_ratio": 0.8878522331018184,
|
| 241 |
+
"flow/mag_ratio_mean": 0.5341137046498785,
|
| 242 |
+
"flow/mag_ratio_std": 0.25290533919324243,
|
| 243 |
+
"step": 4096
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.176525093197147,
|
| 247 |
+
"eval_bleu": 0.9359739466567191,
|
| 248 |
+
"eval_cos_loss": 0.47476763149568524,
|
| 249 |
+
"eval_dec_loss": 0.11913440864024831,
|
| 250 |
+
"eval_loss": 1.4943630281033546,
|
| 251 |
+
"eval_mse2_loss": 0.16356163808722485,
|
| 252 |
+
"eval_mse_loss": 1.3308013893647996,
|
| 253 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 254 |
+
"eval_runtime": 149.7517,
|
| 255 |
+
"eval_samples_per_second": 200.332,
|
| 256 |
+
"eval_steps_per_second": 3.132,
|
| 257 |
+
"eval_var_loss": 0.01723895594080501,
|
| 258 |
+
"flow/cos_sim": 0.5252323670427936,
|
| 259 |
+
"flow/improvement_ratio": 0.8878522331018184,
|
| 260 |
+
"flow/mag_ratio_mean": 0.5341137046498785,
|
| 261 |
+
"flow/mag_ratio_std": 0.25290533919324243,
|
| 262 |
+
"step": 4096
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.18755791152196866,
|
| 266 |
+
"grad_norm": 1.6938560009002686,
|
| 267 |
+
"learning_rate": 0.000931026399368079,
|
| 268 |
+
"loss": 1.4890222549438477,
|
| 269 |
+
"step": 4352
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.19859072984679035,
|
| 273 |
+
"grad_norm": 1.565247893333435,
|
| 274 |
+
"learning_rate": 0.0009217762892151117,
|
| 275 |
+
"loss": 1.4860106706619263,
|
| 276 |
+
"step": 4608
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.20962354817161205,
|
| 280 |
+
"grad_norm": 1.3487786054611206,
|
| 281 |
+
"learning_rate": 0.0009119964292315354,
|
| 282 |
+
"loss": 1.4800715446472168,
|
| 283 |
+
"step": 4864
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.22065636649643372,
|
| 287 |
+
"grad_norm": 1.212808609008789,
|
| 288 |
+
"learning_rate": 0.0009016991028939279,
|
| 289 |
+
"loss": 1.4720630645751953,
|
| 290 |
+
"step": 5120
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.22065636649643372,
|
| 294 |
+
"eval_bleu": 0.9371513693794589,
|
| 295 |
+
"eval_cos_loss": 0.46655392564181836,
|
| 296 |
+
"eval_dec_loss": 0.11585041659592248,
|
| 297 |
+
"eval_loss": 1.4702420382103178,
|
| 298 |
+
"eval_mse2_loss": 0.15908411601141317,
|
| 299 |
+
"eval_mse_loss": 1.3111579230090957,
|
| 300 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 301 |
+
"eval_var_loss": 0.01723895594080501,
|
| 302 |
+
"flow/cos_sim": 0.5334460733414713,
|
| 303 |
+
"flow/improvement_ratio": 0.8902226530158444,
|
| 304 |
+
"flow/mag_ratio_mean": 0.5489529184441069,
|
| 305 |
+
"flow/mag_ratio_std": 0.2660856873178279,
|
| 306 |
+
"step": 5120
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.22065636649643372,
|
| 310 |
+
"eval_bleu": 0.9371513693794589,
|
| 311 |
+
"eval_cos_loss": 0.46655392564181836,
|
| 312 |
+
"eval_dec_loss": 0.11585041659592248,
|
| 313 |
+
"eval_loss": 1.4702420382103178,
|
| 314 |
+
"eval_mse2_loss": 0.15908411601141317,
|
| 315 |
+
"eval_mse_loss": 1.3111579230090957,
|
| 316 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 317 |
+
"eval_runtime": 153.0228,
|
| 318 |
+
"eval_samples_per_second": 196.049,
|
| 319 |
+
"eval_steps_per_second": 3.065,
|
| 320 |
+
"eval_var_loss": 0.01723895594080501,
|
| 321 |
+
"flow/cos_sim": 0.5334460733414713,
|
| 322 |
+
"flow/improvement_ratio": 0.8902226530158444,
|
| 323 |
+
"flow/mag_ratio_mean": 0.5489529184441069,
|
| 324 |
+
"flow/mag_ratio_std": 0.2660856873178279,
|
| 325 |
+
"step": 5120
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.23168918482125542,
|
| 329 |
+
"grad_norm": 1.2820218801498413,
|
| 330 |
+
"learning_rate": 0.0008908972436151494,
|
| 331 |
+
"loss": 1.4665789604187012,
|
| 332 |
+
"step": 5376
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.2427220031460771,
|
| 336 |
+
"grad_norm": 1.428142786026001,
|
| 337 |
+
"learning_rate": 0.0008796044185000127,
|
| 338 |
+
"loss": 1.4617829322814941,
|
| 339 |
+
"step": 5632
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2537548214708988,
|
| 343 |
+
"grad_norm": 1.7342983484268188,
|
| 344 |
+
"learning_rate": 0.0008678348113050368,
|
| 345 |
+
"loss": 1.454346776008606,
|
| 346 |
+
"step": 5888
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.26478763979572045,
|
| 350 |
+
"grad_norm": 1.012923240661621,
|
| 351 |
+
"learning_rate": 0.0008556032046236897,
|
| 352 |
+
"loss": 1.4543688297271729,
|
| 353 |
+
"step": 6144
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.26478763979572045,
|
| 357 |
+
"eval_bleu": 0.9360335622230687,
|
| 358 |
+
"eval_cos_loss": 0.4603669429892924,
|
| 359 |
+
"eval_dec_loss": 0.11954856799669199,
|
| 360 |
+
"eval_loss": 1.452896745982709,
|
| 361 |
+
"eval_mse2_loss": 0.15859206668984915,
|
| 362 |
+
"eval_mse_loss": 1.2943046784350105,
|
| 363 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 364 |
+
"eval_var_loss": 0.01723895594080501,
|
| 365 |
+
"flow/cos_sim": 0.5396330592347615,
|
| 366 |
+
"flow/improvement_ratio": 0.8903370405565193,
|
| 367 |
+
"flow/mag_ratio_mean": 0.5503829778003286,
|
| 368 |
+
"flow/mag_ratio_std": 0.26988969053795087,
|
| 369 |
+
"step": 6144
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.26478763979572045,
|
| 373 |
+
"eval_bleu": 0.9360335622230687,
|
| 374 |
+
"eval_cos_loss": 0.4603669429892924,
|
| 375 |
+
"eval_dec_loss": 0.11954856799669199,
|
| 376 |
+
"eval_loss": 1.452896745982709,
|
| 377 |
+
"eval_mse2_loss": 0.15859206668984915,
|
| 378 |
+
"eval_mse_loss": 1.2943046784350105,
|
| 379 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 380 |
+
"eval_runtime": 152.3001,
|
| 381 |
+
"eval_samples_per_second": 196.979,
|
| 382 |
+
"eval_steps_per_second": 3.079,
|
| 383 |
+
"eval_var_loss": 0.01723895594080501,
|
| 384 |
+
"flow/cos_sim": 0.5396330592347615,
|
| 385 |
+
"flow/improvement_ratio": 0.8903370405565193,
|
| 386 |
+
"flow/mag_ratio_mean": 0.5503829778003286,
|
| 387 |
+
"flow/mag_ratio_std": 0.26988969053795087,
|
| 388 |
+
"step": 6144
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.2758204581205422,
|
| 392 |
+
"grad_norm": 1.2126809358596802,
|
| 393 |
+
"learning_rate": 0.000842924961319492,
|
| 394 |
+
"loss": 1.4489810466766357,
|
| 395 |
+
"step": 6400
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.28685327644536385,
|
| 399 |
+
"grad_norm": 1.6154913902282715,
|
| 400 |
+
"learning_rate": 0.0008298160052303045,
|
| 401 |
+
"loss": 1.4455806016921997,
|
| 402 |
+
"step": 6656
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2978860947701855,
|
| 406 |
+
"grad_norm": 1.1049504280090332,
|
| 407 |
+
"learning_rate": 0.0008162928011680314,
|
| 408 |
+
"loss": 1.4431121349334717,
|
| 409 |
+
"step": 6912
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.30891891309500724,
|
| 413 |
+
"grad_norm": 1.161877989768982,
|
| 414 |
+
"learning_rate": 0.000802372334238864,
|
| 415 |
+
"loss": 1.440521478652954,
|
| 416 |
+
"step": 7168
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.30891891309500724,
|
| 420 |
+
"eval_bleu": 0.9368360477187275,
|
| 421 |
+
"eval_cos_loss": 0.4548988695592006,
|
| 422 |
+
"eval_dec_loss": 0.11761775160673013,
|
| 423 |
+
"eval_loss": 1.4366158300371312,
|
| 424 |
+
"eval_mse2_loss": 0.1551807067636996,
|
| 425 |
+
"eval_mse_loss": 1.2814351221137463,
|
| 426 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 427 |
+
"eval_var_loss": 0.01723895594080501,
|
| 428 |
+
"flow/cos_sim": 0.545101128788645,
|
| 429 |
+
"flow/improvement_ratio": 0.8925937072300454,
|
| 430 |
+
"flow/mag_ratio_mean": 0.5560947322387939,
|
| 431 |
+
"flow/mag_ratio_std": 0.26709621954065904,
|
| 432 |
+
"step": 7168
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.30891891309500724,
|
| 436 |
+
"eval_bleu": 0.9368360477187275,
|
| 437 |
+
"eval_cos_loss": 0.4548988695592006,
|
| 438 |
+
"eval_dec_loss": 0.11761775160673013,
|
| 439 |
+
"eval_loss": 1.4366158300371312,
|
| 440 |
+
"eval_mse2_loss": 0.1551807067636996,
|
| 441 |
+
"eval_mse_loss": 1.2814351221137463,
|
| 442 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 443 |
+
"eval_runtime": 150.401,
|
| 444 |
+
"eval_samples_per_second": 199.467,
|
| 445 |
+
"eval_steps_per_second": 3.118,
|
| 446 |
+
"eval_var_loss": 0.01723895594080501,
|
| 447 |
+
"flow/cos_sim": 0.545101128788645,
|
| 448 |
+
"flow/improvement_ratio": 0.8925937072300454,
|
| 449 |
+
"flow/mag_ratio_mean": 0.5560947322387939,
|
| 450 |
+
"flow/mag_ratio_std": 0.26709621954065904,
|
| 451 |
+
"step": 7168
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.3199517314198289,
|
| 455 |
+
"grad_norm": 1.3750261068344116,
|
| 456 |
+
"learning_rate": 0.0007880720885100349,
|
| 457 |
+
"loss": 1.4386309385299683,
|
| 458 |
+
"step": 7424
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3309845497446506,
|
| 462 |
+
"grad_norm": 0.9077029228210449,
|
| 463 |
+
"learning_rate": 0.0007734100250498788,
|
| 464 |
+
"loss": 1.4343377351760864,
|
| 465 |
+
"step": 7680
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3420173680694723,
|
| 469 |
+
"grad_norm": 1.2454605102539062,
|
| 470 |
+
"learning_rate": 0.000758404559368781,
|
| 471 |
+
"loss": 1.4340428113937378,
|
| 472 |
+
"step": 7936
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.353050186394294,
|
| 476 |
+
"grad_norm": 1.4264285564422607,
|
| 477 |
+
"learning_rate": 0.0007430745382893488,
|
| 478 |
+
"loss": 1.4275590181350708,
|
| 479 |
+
"step": 8192
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.353050186394294,
|
| 483 |
+
"eval_bleu": 0.9352995295168319,
|
| 484 |
+
"eval_cos_loss": 0.4525942064043301,
|
| 485 |
+
"eval_dec_loss": 0.12027457510150953,
|
| 486 |
+
"eval_loss": 1.4305989170379476,
|
| 487 |
+
"eval_mse2_loss": 0.15546214358130497,
|
| 488 |
+
"eval_mse_loss": 1.2751367725034766,
|
| 489 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 490 |
+
"eval_var_loss": 0.01723895594080501,
|
| 491 |
+
"flow/cos_sim": 0.547405791689338,
|
| 492 |
+
"flow/improvement_ratio": 0.8944777580720784,
|
| 493 |
+
"flow/mag_ratio_mean": 0.5549619147009941,
|
| 494 |
+
"flow/mag_ratio_std": 0.2656229132973055,
|
| 495 |
+
"step": 8192
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.353050186394294,
|
| 499 |
+
"eval_bleu": 0.9352995295168319,
|
| 500 |
+
"eval_cos_loss": 0.4525942064043301,
|
| 501 |
+
"eval_dec_loss": 0.12027457510150953,
|
| 502 |
+
"eval_loss": 1.4305989170379476,
|
| 503 |
+
"eval_mse2_loss": 0.15546214358130497,
|
| 504 |
+
"eval_mse_loss": 1.2751367725034766,
|
| 505 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 506 |
+
"eval_runtime": 153.0138,
|
| 507 |
+
"eval_samples_per_second": 196.061,
|
| 508 |
+
"eval_steps_per_second": 3.065,
|
| 509 |
+
"eval_var_loss": 0.01723895594080501,
|
| 510 |
+
"flow/cos_sim": 0.547405791689338,
|
| 511 |
+
"flow/improvement_ratio": 0.8944777580720784,
|
| 512 |
+
"flow/mag_ratio_mean": 0.5549619147009941,
|
| 513 |
+
"flow/mag_ratio_std": 0.2656229132973055,
|
| 514 |
+
"step": 8192
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.36408300471911564,
|
| 518 |
+
"grad_norm": 1.2861735820770264,
|
| 519 |
+
"learning_rate": 0.0007274392162748551,
|
| 520 |
+
"loss": 1.4297358989715576,
|
| 521 |
+
"step": 8448
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3751158230439373,
|
| 525 |
+
"grad_norm": 0.9941537380218506,
|
| 526 |
+
"learning_rate": 0.000711518231245687,
|
| 527 |
+
"loss": 1.4256483316421509,
|
| 528 |
+
"step": 8704
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.38614864136875904,
|
| 532 |
+
"grad_norm": 1.144028663635254,
|
| 533 |
+
"learning_rate": 0.0006953315799141723,
|
| 534 |
+
"loss": 1.4252721071243286,
|
| 535 |
+
"step": 8960
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.3971814596935807,
|
| 539 |
+
"grad_norm": 1.3483773469924927,
|
| 540 |
+
"learning_rate": 0.0006788995926687669,
|
| 541 |
+
"loss": 1.4217591285705566,
|
| 542 |
+
"step": 9216
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.3971814596935807,
|
| 546 |
+
"eval_bleu": 0.9351012282354647,
|
| 547 |
+
"eval_cos_loss": 0.44966854437836196,
|
| 548 |
+
"eval_dec_loss": 0.12245929105155694,
|
| 549 |
+
"eval_loss": 1.4224536047815515,
|
| 550 |
+
"eval_mse2_loss": 0.15468242016237682,
|
| 551 |
+
"eval_mse_loss": 1.2677711858424043,
|
| 552 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 553 |
+
"eval_var_loss": 0.01723895594080501,
|
| 554 |
+
"flow/cos_sim": 0.5503314502203642,
|
| 555 |
+
"flow/improvement_ratio": 0.892241062640127,
|
| 556 |
+
"flow/mag_ratio_mean": 0.5635909434320576,
|
| 557 |
+
"flow/mag_ratio_std": 0.27180960243825975,
|
| 558 |
+
"step": 9216
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.3971814596935807,
|
| 562 |
+
"eval_bleu": 0.9351012282354647,
|
| 563 |
+
"eval_cos_loss": 0.44966854437836196,
|
| 564 |
+
"eval_dec_loss": 0.12245929105155694,
|
| 565 |
+
"eval_loss": 1.4224536047815515,
|
| 566 |
+
"eval_mse2_loss": 0.15468242016237682,
|
| 567 |
+
"eval_mse_loss": 1.2677711858424043,
|
| 568 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 569 |
+
"eval_runtime": 149.98,
|
| 570 |
+
"eval_samples_per_second": 200.027,
|
| 571 |
+
"eval_steps_per_second": 3.127,
|
| 572 |
+
"eval_var_loss": 0.01723895594080501,
|
| 573 |
+
"flow/cos_sim": 0.5503314502203642,
|
| 574 |
+
"flow/improvement_ratio": 0.892241062640127,
|
| 575 |
+
"flow/mag_ratio_mean": 0.5635909434320576,
|
| 576 |
+
"flow/mag_ratio_std": 0.27180960243825975,
|
| 577 |
+
"step": 9216
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4082142780184024,
|
| 581 |
+
"grad_norm": 1.1067237854003906,
|
| 582 |
+
"learning_rate": 0.0006622429080391422,
|
| 583 |
+
"loss": 1.4252803325653076,
|
| 584 |
+
"step": 9472
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.4192470963432241,
|
| 588 |
+
"grad_norm": 0.9428814649581909,
|
| 589 |
+
"learning_rate": 0.0006453824467742515,
|
| 590 |
+
"loss": 1.4204916954040527,
|
| 591 |
+
"step": 9728
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.43027991466804577,
|
| 595 |
+
"grad_norm": 0.9498366713523865,
|
| 596 |
+
"learning_rate": 0.0006283393855659275,
|
| 597 |
+
"loss": 1.421370506286621,
|
| 598 |
+
"step": 9984
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.44131273299286744,
|
| 602 |
+
"grad_norm": 1.0918892621994019,
|
| 603 |
+
"learning_rate": 0.0006111351304510173,
|
| 604 |
+
"loss": 1.4164854288101196,
|
| 605 |
+
"step": 10240
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.44131273299286744,
|
| 609 |
+
"eval_bleu": 0.9359500949598805,
|
| 610 |
+
"eval_cos_loss": 0.4479471450802614,
|
| 611 |
+
"eval_dec_loss": 0.11968307011623754,
|
| 612 |
+
"eval_loss": 1.4188302267334862,
|
| 613 |
+
"eval_mse2_loss": 0.1544050017653752,
|
| 614 |
+
"eval_mse_loss": 1.2644252238243119,
|
| 615 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 616 |
+
"eval_var_loss": 0.01723895594080501,
|
| 617 |
+
"flow/cos_sim": 0.5520528566354374,
|
| 618 |
+
"flow/improvement_ratio": 0.8973728770386181,
|
| 619 |
+
"flow/mag_ratio_mean": 0.5616697732573633,
|
| 620 |
+
"flow/mag_ratio_std": 0.26565590021071406,
|
| 621 |
+
"step": 10240
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.44131273299286744,
|
| 625 |
+
"eval_bleu": 0.9359500949598805,
|
| 626 |
+
"eval_cos_loss": 0.4479471450802614,
|
| 627 |
+
"eval_dec_loss": 0.11968307011623754,
|
| 628 |
+
"eval_loss": 1.4188302267334862,
|
| 629 |
+
"eval_mse2_loss": 0.1544050017653752,
|
| 630 |
+
"eval_mse_loss": 1.2644252238243119,
|
| 631 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 632 |
+
"eval_runtime": 154.1693,
|
| 633 |
+
"eval_samples_per_second": 194.591,
|
| 634 |
+
"eval_steps_per_second": 3.042,
|
| 635 |
+
"eval_var_loss": 0.01723895594080501,
|
| 636 |
+
"flow/cos_sim": 0.5520528566354374,
|
| 637 |
+
"flow/improvement_ratio": 0.8973728770386181,
|
| 638 |
+
"flow/mag_ratio_mean": 0.5616697732573633,
|
| 639 |
+
"flow/mag_ratio_std": 0.26565590021071406,
|
| 640 |
+
"step": 10240
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.45234555131768917,
|
| 644 |
+
"grad_norm": 1.4444184303283691,
|
| 645 |
+
"learning_rate": 0.0005937912899254605,
|
| 646 |
+
"loss": 1.4161497354507446,
|
| 647 |
+
"step": 10496
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.46337836964251083,
|
| 651 |
+
"grad_norm": 1.1204206943511963,
|
| 652 |
+
"learning_rate": 0.0005763296478040787,
|
| 653 |
+
"loss": 1.4160338640213013,
|
| 654 |
+
"step": 10752
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4744111879673325,
|
| 658 |
+
"grad_norm": 1.2254257202148438,
|
| 659 |
+
"learning_rate": 0.0005587721358601663,
|
| 660 |
+
"loss": 1.4160186052322388,
|
| 661 |
+
"step": 11008
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4854440062921542,
|
| 665 |
+
"grad_norm": 1.4265046119689941,
|
| 666 |
+
"learning_rate": 0.0005411408062792448,
|
| 667 |
+
"loss": 1.4131176471710205,
|
| 668 |
+
"step": 11264
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.4854440062921542,
|
| 672 |
+
"eval_bleu": 0.9355559828150384,
|
| 673 |
+
"eval_cos_loss": 0.4478943133150845,
|
| 674 |
+
"eval_dec_loss": 0.1184240544337962,
|
| 675 |
+
"eval_loss": 1.416160212396813,
|
| 676 |
+
"eval_mse2_loss": 0.15195046149210126,
|
| 677 |
+
"eval_mse_loss": 1.2642097521438274,
|
| 678 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 679 |
+
"eval_var_loss": 0.01723895594080501,
|
| 680 |
+
"flow/cos_sim": 0.5521056849056724,
|
| 681 |
+
"flow/improvement_ratio": 0.8935815333557535,
|
| 682 |
+
"flow/mag_ratio_mean": 0.5588731608156965,
|
| 683 |
+
"flow/mag_ratio_std": 0.2649187071364063,
|
| 684 |
+
"step": 11264
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.4854440062921542,
|
| 688 |
+
"eval_bleu": 0.9355559828150384,
|
| 689 |
+
"eval_cos_loss": 0.4478943133150845,
|
| 690 |
+
"eval_dec_loss": 0.1184240544337962,
|
| 691 |
+
"eval_loss": 1.416160212396813,
|
| 692 |
+
"eval_mse2_loss": 0.15195046149210126,
|
| 693 |
+
"eval_mse_loss": 1.2642097521438274,
|
| 694 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 695 |
+
"eval_runtime": 153.2221,
|
| 696 |
+
"eval_samples_per_second": 195.794,
|
| 697 |
+
"eval_steps_per_second": 3.061,
|
| 698 |
+
"eval_var_loss": 0.01723895594080501,
|
| 699 |
+
"flow/cos_sim": 0.5521056849056724,
|
| 700 |
+
"flow/improvement_ratio": 0.8935815333557535,
|
| 701 |
+
"flow/mag_ratio_mean": 0.5588731608156965,
|
| 702 |
+
"flow/mag_ratio_std": 0.2649187071364063,
|
| 703 |
+
"step": 11264
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.4964768246169759,
|
| 707 |
+
"grad_norm": 1.052992582321167,
|
| 708 |
+
"learning_rate": 0.0005234578039615789,
|
| 709 |
+
"loss": 1.4113584756851196,
|
| 710 |
+
"step": 11520
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5075096429417976,
|
| 714 |
+
"grad_norm": 1.5359535217285156,
|
| 715 |
+
"learning_rate": 0.0005057453387082458,
|
| 716 |
+
"loss": 1.410513997077942,
|
| 717 |
+
"step": 11776
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.5185424612666193,
|
| 721 |
+
"grad_norm": 1.2274341583251953,
|
| 722 |
+
"learning_rate": 0.0004880256573256866,
|
| 723 |
+
"loss": 1.4113372564315796,
|
| 724 |
+
"step": 12032
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.5295752795914409,
|
| 728 |
+
"grad_norm": 0.7585653066635132,
|
| 729 |
+
"learning_rate": 0.0004703210156837805,
|
| 730 |
+
"loss": 1.4105942249298096,
|
| 731 |
+
"step": 12288
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.5295752795914409,
|
| 735 |
+
"eval_bleu": 0.9365001737473854,
|
| 736 |
+
"eval_cos_loss": 0.4468586768295719,
|
| 737 |
+
"eval_dec_loss": 0.11578587821519959,
|
| 738 |
+
"eval_loss": 1.4128359007174527,
|
| 739 |
+
"eval_mse2_loss": 0.1518495731802383,
|
| 740 |
+
"eval_mse_loss": 1.2609863276166449,
|
| 741 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 742 |
+
"eval_var_loss": 0.01723895594080501,
|
| 743 |
+
"flow/cos_sim": 0.5531413205651078,
|
| 744 |
+
"flow/improvement_ratio": 0.8957801499346426,
|
| 745 |
+
"flow/mag_ratio_mean": 0.558759307556315,
|
| 746 |
+
"flow/mag_ratio_std": 0.2658895003770206,
|
| 747 |
+
"step": 12288
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.5295752795914409,
|
| 751 |
+
"eval_bleu": 0.9365001737473854,
|
| 752 |
+
"eval_cos_loss": 0.4468586768295719,
|
| 753 |
+
"eval_dec_loss": 0.11578587821519959,
|
| 754 |
+
"eval_loss": 1.4128359007174527,
|
| 755 |
+
"eval_mse2_loss": 0.1518495731802383,
|
| 756 |
+
"eval_mse_loss": 1.2609863276166449,
|
| 757 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 758 |
+
"eval_runtime": 152.5586,
|
| 759 |
+
"eval_samples_per_second": 196.646,
|
| 760 |
+
"eval_steps_per_second": 3.074,
|
| 761 |
+
"eval_var_loss": 0.01723895594080501,
|
| 762 |
+
"flow/cos_sim": 0.5531413205651078,
|
| 763 |
+
"flow/improvement_ratio": 0.8957801499346426,
|
| 764 |
+
"flow/mag_ratio_mean": 0.558759307556315,
|
| 765 |
+
"flow/mag_ratio_std": 0.2658895003770206,
|
| 766 |
+
"step": 12288
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.5406080979162626,
|
| 770 |
+
"grad_norm": 0.7903220653533936,
|
| 771 |
+
"learning_rate": 0.0004526536507625343,
|
| 772 |
+
"loss": 1.4081207513809204,
|
| 773 |
+
"step": 12544
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.5516409162410844,
|
| 777 |
+
"grad_norm": 1.299485445022583,
|
| 778 |
+
"learning_rate": 0.00043504575272249973,
|
| 779 |
+
"loss": 1.4094181060791016,
|
| 780 |
+
"step": 12800
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.562673734565906,
|
| 784 |
+
"grad_norm": 1.1680424213409424,
|
| 785 |
+
"learning_rate": 0.0004175194370339921,
|
| 786 |
+
"loss": 1.4080712795257568,
|
| 787 |
+
"step": 13056
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.5737065528907277,
|
| 791 |
+
"grad_norm": 1.1761500835418701,
|
| 792 |
+
"learning_rate": 0.0004000967167001243,
|
| 793 |
+
"loss": 1.406380534172058,
|
| 794 |
+
"step": 13312
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5737065528907277,
|
| 798 |
+
"eval_bleu": 0.9375331490996615,
|
| 799 |
+
"eval_cos_loss": 0.4439614661721024,
|
| 800 |
+
"eval_dec_loss": 0.11194793060858811,
|
| 801 |
+
"eval_loss": 1.4055488673862873,
|
| 802 |
+
"eval_mse2_loss": 0.1498908800547565,
|
| 803 |
+
"eval_mse_loss": 1.2556579890790016,
|
| 804 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 805 |
+
"eval_var_loss": 0.01723895594080501,
|
| 806 |
+
"flow/cos_sim": 0.5560385347175192,
|
| 807 |
+
"flow/improvement_ratio": 0.8952096234252458,
|
| 808 |
+
"flow/mag_ratio_mean": 0.5663808779930001,
|
| 809 |
+
"flow/mag_ratio_std": 0.2659331502627208,
|
| 810 |
+
"step": 13312
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.5737065528907277,
|
| 814 |
+
"eval_bleu": 0.9375331490996615,
|
| 815 |
+
"eval_cos_loss": 0.4439614661721024,
|
| 816 |
+
"eval_dec_loss": 0.11194793060858811,
|
| 817 |
+
"eval_loss": 1.4055488673862873,
|
| 818 |
+
"eval_mse2_loss": 0.1498908800547565,
|
| 819 |
+
"eval_mse_loss": 1.2556579890790016,
|
| 820 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 821 |
+
"eval_runtime": 151.6081,
|
| 822 |
+
"eval_samples_per_second": 197.879,
|
| 823 |
+
"eval_steps_per_second": 3.094,
|
| 824 |
+
"eval_var_loss": 0.01723895594080501,
|
| 825 |
+
"flow/cos_sim": 0.5560385347175192,
|
| 826 |
+
"flow/improvement_ratio": 0.8952096234252458,
|
| 827 |
+
"flow/mag_ratio_mean": 0.5663808779930001,
|
| 828 |
+
"flow/mag_ratio_std": 0.2659331502627208,
|
| 829 |
+
"step": 13312
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.5847393712155494,
|
| 833 |
+
"grad_norm": 0.7058362364768982,
|
| 834 |
+
"learning_rate": 0.00038279947460853446,
|
| 835 |
+
"loss": 1.4039076566696167,
|
| 836 |
+
"step": 13568
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.595772189540371,
|
| 840 |
+
"grad_norm": 1.867743730545044,
|
| 841 |
+
"learning_rate": 0.00036564943604654345,
|
| 842 |
+
"loss": 1.4034653902053833,
|
| 843 |
+
"step": 13824
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 0.6068050078651928,
|
| 847 |
+
"grad_norm": 0.9271458983421326,
|
| 848 |
+
"learning_rate": 0.00034866814141425254,
|
| 849 |
+
"loss": 1.4046720266342163,
|
| 850 |
+
"step": 14080
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.6178378261900145,
|
| 854 |
+
"grad_norm": 0.8186226487159729,
|
| 855 |
+
"learning_rate": 0.0003318769191698637,
|
| 856 |
+
"loss": 1.4014854431152344,
|
| 857 |
+
"step": 14336
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.6178378261900145,
|
| 861 |
+
"eval_bleu": 0.9378049095613451,
|
| 862 |
+
"eval_cos_loss": 0.44325206020493496,
|
| 863 |
+
"eval_dec_loss": 0.11280616647454658,
|
| 864 |
+
"eval_loss": 1.402791700637671,
|
| 865 |
+
"eval_mse2_loss": 0.14996937808515168,
|
| 866 |
+
"eval_mse_loss": 1.2528223221235946,
|
| 867 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 868 |
+
"eval_var_loss": 0.01723895594080501,
|
| 869 |
+
"flow/cos_sim": 0.5567479390325323,
|
| 870 |
+
"flow/improvement_ratio": 0.8963780609021055,
|
| 871 |
+
"flow/mag_ratio_mean": 0.5675655131909385,
|
| 872 |
+
"flow/mag_ratio_std": 0.2700714123274472,
|
| 873 |
+
"step": 14336
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.6178378261900145,
|
| 877 |
+
"eval_bleu": 0.9378049095613451,
|
| 878 |
+
"eval_cos_loss": 0.44325206020493496,
|
| 879 |
+
"eval_dec_loss": 0.11280616647454658,
|
| 880 |
+
"eval_loss": 1.402791700637671,
|
| 881 |
+
"eval_mse2_loss": 0.14996937808515168,
|
| 882 |
+
"eval_mse_loss": 1.2528223221235946,
|
| 883 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 884 |
+
"eval_runtime": 152.2222,
|
| 885 |
+
"eval_samples_per_second": 197.08,
|
| 886 |
+
"eval_steps_per_second": 3.081,
|
| 887 |
+
"eval_var_loss": 0.01723895594080501,
|
| 888 |
+
"flow/cos_sim": 0.5567479390325323,
|
| 889 |
+
"flow/improvement_ratio": 0.8963780609021055,
|
| 890 |
+
"flow/mag_ratio_mean": 0.5675655131909385,
|
| 891 |
+
"flow/mag_ratio_std": 0.2700714123274472,
|
| 892 |
+
"step": 14336
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.6288706445148361,
|
| 896 |
+
"grad_norm": 1.4290482997894287,
|
| 897 |
+
"learning_rate": 0.00031529685904119485,
|
| 898 |
+
"loss": 1.4003877639770508,
|
| 899 |
+
"step": 14592
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.6399034628396578,
|
| 903 |
+
"grad_norm": 0.92575603723526,
|
| 904 |
+
"learning_rate": 0.0002989487855370421,
|
| 905 |
+
"loss": 1.4029277563095093,
|
| 906 |
+
"step": 14848
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 0.6509362811644795,
|
| 910 |
+
"grad_norm": 0.9941800236701965,
|
| 911 |
+
"learning_rate": 0.00028285323179165424,
|
| 912 |
+
"loss": 1.401476263999939,
|
| 913 |
+
"step": 15104
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.6619690994893012,
|
| 917 |
+
"grad_norm": 0.5996092557907104,
|
| 918 |
+
"learning_rate": 0.0002670304137751759,
|
| 919 |
+
"loss": 1.40423583984375,
|
| 920 |
+
"step": 15360
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.6619690994893012,
|
| 924 |
+
"eval_bleu": 0.9359784532013938,
|
| 925 |
+
"eval_cos_loss": 0.44205584766259837,
|
| 926 |
+
"eval_dec_loss": 0.12092216288460407,
|
| 927 |
+
"eval_loss": 1.4025819934507422,
|
| 928 |
+
"eval_mse2_loss": 0.15172467312451873,
|
| 929 |
+
"eval_mse_loss": 1.2508573216924281,
|
| 930 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 931 |
+
"eval_var_loss": 0.01723895594080501,
|
| 932 |
+
"flow/cos_sim": 0.5579441513206913,
|
| 933 |
+
"flow/improvement_ratio": 0.893738461964166,
|
| 934 |
+
"flow/mag_ratio_mean": 0.5637592339058166,
|
| 935 |
+
"flow/mag_ratio_std": 0.2644218521903573,
|
| 936 |
+
"step": 15360
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 0.6619690994893012,
|
| 940 |
+
"eval_bleu": 0.9359784532013938,
|
| 941 |
+
"eval_cos_loss": 0.44205584766259837,
|
| 942 |
+
"eval_dec_loss": 0.12092216288460407,
|
| 943 |
+
"eval_loss": 1.4025819934507422,
|
| 944 |
+
"eval_mse2_loss": 0.15172467312451873,
|
| 945 |
+
"eval_mse_loss": 1.2508573216924281,
|
| 946 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 947 |
+
"eval_runtime": 150.8251,
|
| 948 |
+
"eval_samples_per_second": 198.906,
|
| 949 |
+
"eval_steps_per_second": 3.11,
|
| 950 |
+
"eval_var_loss": 0.01723895594080501,
|
| 951 |
+
"flow/cos_sim": 0.5579441513206913,
|
| 952 |
+
"flow/improvement_ratio": 0.893738461964166,
|
| 953 |
+
"flow/mag_ratio_mean": 0.5637592339058166,
|
| 954 |
+
"flow/mag_ratio_std": 0.2644218521903573,
|
| 955 |
+
"step": 15360
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.6730019178141229,
|
| 959 |
+
"grad_norm": 1.2950412034988403,
|
| 960 |
+
"learning_rate": 0.0002515002049024435,
|
| 961 |
+
"loss": 1.4007450342178345,
|
| 962 |
+
"step": 15616
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.6840347361389446,
|
| 966 |
+
"grad_norm": 0.9485428333282471,
|
| 967 |
+
"learning_rate": 0.00023628211107203429,
|
| 968 |
+
"loss": 1.4006445407867432,
|
| 969 |
+
"step": 15872
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.6950675544637662,
|
| 973 |
+
"grad_norm": 0.7694559693336487,
|
| 974 |
+
"learning_rate": 0.00022139524616691188,
|
| 975 |
+
"loss": 1.3995271921157837,
|
| 976 |
+
"step": 16128
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.706100372788588,
|
| 980 |
+
"grad_norm": 1.0373378992080688,
|
| 981 |
+
"learning_rate": 0.000206858308047443,
|
| 982 |
+
"loss": 1.396484136581421,
|
| 983 |
+
"step": 16384
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 0.706100372788588,
|
| 987 |
+
"eval_bleu": 0.9382976118785468,
|
| 988 |
+
"eval_cos_loss": 0.4408198703072473,
|
| 989 |
+
"eval_dec_loss": 0.11163298516814261,
|
| 990 |
+
"eval_loss": 1.3957897457740962,
|
| 991 |
+
"eval_mse2_loss": 0.14787080404219596,
|
| 992 |
+
"eval_mse_loss": 1.247918940556329,
|
| 993 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 994 |
+
"eval_var_loss": 0.01723895594080501,
|
| 995 |
+
"flow/cos_sim": 0.5591801271509769,
|
| 996 |
+
"flow/improvement_ratio": 0.8951915981672974,
|
| 997 |
+
"flow/mag_ratio_mean": 0.5686852338471646,
|
| 998 |
+
"flow/mag_ratio_std": 0.26723379195371927,
|
| 999 |
+
"step": 16384
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.706100372788588,
|
| 1003 |
+
"eval_bleu": 0.9382976118785468,
|
| 1004 |
+
"eval_cos_loss": 0.4408198703072473,
|
| 1005 |
+
"eval_dec_loss": 0.11163298516814261,
|
| 1006 |
+
"eval_loss": 1.3957897457740962,
|
| 1007 |
+
"eval_mse2_loss": 0.14787080404219596,
|
| 1008 |
+
"eval_mse_loss": 1.247918940556329,
|
| 1009 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1010 |
+
"eval_runtime": 149.6268,
|
| 1011 |
+
"eval_samples_per_second": 200.499,
|
| 1012 |
+
"eval_steps_per_second": 3.134,
|
| 1013 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1014 |
+
"flow/cos_sim": 0.5591801271509769,
|
| 1015 |
+
"flow/improvement_ratio": 0.8951915981672974,
|
| 1016 |
+
"flow/mag_ratio_mean": 0.5686852338471646,
|
| 1017 |
+
"flow/mag_ratio_std": 0.26723379195371927,
|
| 1018 |
+
"step": 16384
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.7171331911134096,
|
| 1022 |
+
"grad_norm": 0.7638829946517944,
|
| 1023 |
+
"learning_rate": 0.00019268955506693798,
|
| 1024 |
+
"loss": 1.3981057405471802,
|
| 1025 |
+
"step": 16640
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.7281660094382313,
|
| 1029 |
+
"grad_norm": 1.2486227750778198,
|
| 1030 |
+
"learning_rate": 0.00017890678313921,
|
| 1031 |
+
"loss": 1.396310806274414,
|
| 1032 |
+
"step": 16896
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 0.739198827763053,
|
| 1036 |
+
"grad_norm": 1.3649706840515137,
|
| 1037 |
+
"learning_rate": 0.00016552730338695792,
|
| 1038 |
+
"loss": 1.395443320274353,
|
| 1039 |
+
"step": 17152
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 0.7502316460878746,
|
| 1043 |
+
"grad_norm": 0.7208181023597717,
|
| 1044 |
+
"learning_rate": 0.00015256792039904465,
|
| 1045 |
+
"loss": 1.3944549560546875,
|
| 1046 |
+
"step": 17408
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.7502316460878746,
|
| 1050 |
+
"eval_bleu": 0.9394802954217433,
|
| 1051 |
+
"eval_cos_loss": 0.43953821437953633,
|
| 1052 |
+
"eval_dec_loss": 0.11125741161501357,
|
| 1053 |
+
"eval_loss": 1.390636357925594,
|
| 1054 |
+
"eval_mse2_loss": 0.14740569860950462,
|
| 1055 |
+
"eval_mse_loss": 1.243230658299379,
|
| 1056 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1057 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1058 |
+
"flow/cos_sim": 0.5604617825703326,
|
| 1059 |
+
"flow/improvement_ratio": 0.8963484140093139,
|
| 1060 |
+
"flow/mag_ratio_mean": 0.5701523471488628,
|
| 1061 |
+
"flow/mag_ratio_std": 0.2696803689701979,
|
| 1062 |
+
"step": 17408
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.7502316460878746,
|
| 1066 |
+
"eval_bleu": 0.9394802954217433,
|
| 1067 |
+
"eval_cos_loss": 0.43953821437953633,
|
| 1068 |
+
"eval_dec_loss": 0.11125741161501357,
|
| 1069 |
+
"eval_loss": 1.390636357925594,
|
| 1070 |
+
"eval_mse2_loss": 0.14740569860950462,
|
| 1071 |
+
"eval_mse_loss": 1.243230658299379,
|
| 1072 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1073 |
+
"eval_runtime": 150.0541,
|
| 1074 |
+
"eval_samples_per_second": 199.928,
|
| 1075 |
+
"eval_steps_per_second": 3.126,
|
| 1076 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1077 |
+
"flow/cos_sim": 0.5604617825703326,
|
| 1078 |
+
"flow/improvement_ratio": 0.8963484140093139,
|
| 1079 |
+
"flow/mag_ratio_mean": 0.5701523471488628,
|
| 1080 |
+
"flow/mag_ratio_std": 0.2696803689701979,
|
| 1081 |
+
"step": 17408
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.7612644644126964,
|
| 1085 |
+
"grad_norm": 0.7441651821136475,
|
| 1086 |
+
"learning_rate": 0.00014004491112398103,
|
| 1087 |
+
"loss": 1.3953999280929565,
|
| 1088 |
+
"step": 17664
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.7722972827375181,
|
| 1092 |
+
"grad_norm": 0.6111723780632019,
|
| 1093 |
+
"learning_rate": 0.00012797400442612433,
|
| 1094 |
+
"loss": 1.3906549215316772,
|
| 1095 |
+
"step": 17920
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 0.7833301010623397,
|
| 1099 |
+
"grad_norm": 0.7702776789665222,
|
| 1100 |
+
"learning_rate": 0.00011637036133026895,
|
| 1101 |
+
"loss": 1.393618106842041,
|
| 1102 |
+
"step": 18176
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.7943629193871614,
|
| 1106 |
+
"grad_norm": 1.200044870376587,
|
| 1107 |
+
"learning_rate": 0.00010524855597944216,
|
| 1108 |
+
"loss": 1.3919320106506348,
|
| 1109 |
+
"step": 18432
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.7943629193871614,
|
| 1113 |
+
"eval_bleu": 0.9383138215986374,
|
| 1114 |
+
"eval_cos_loss": 0.44080494201259574,
|
| 1115 |
+
"eval_dec_loss": 0.11339724150253956,
|
| 1116 |
+
"eval_loss": 1.3953827009526396,
|
| 1117 |
+
"eval_mse2_loss": 0.14798823779008027,
|
| 1118 |
+
"eval_mse_loss": 1.247394463400851,
|
| 1119 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1120 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1121 |
+
"flow/cos_sim": 0.5591950585593038,
|
| 1122 |
+
"flow/improvement_ratio": 0.8969450570118707,
|
| 1123 |
+
"flow/mag_ratio_mean": 0.5669063877449361,
|
| 1124 |
+
"flow/mag_ratio_std": 0.26610757334273,
|
| 1125 |
+
"step": 18432
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.7943629193871614,
|
| 1129 |
+
"eval_bleu": 0.9383138215986374,
|
| 1130 |
+
"eval_cos_loss": 0.44080494201259574,
|
| 1131 |
+
"eval_dec_loss": 0.11339724150253956,
|
| 1132 |
+
"eval_loss": 1.3953827009526396,
|
| 1133 |
+
"eval_mse2_loss": 0.14798823779008027,
|
| 1134 |
+
"eval_mse_loss": 1.247394463400851,
|
| 1135 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1136 |
+
"eval_runtime": 148.9298,
|
| 1137 |
+
"eval_samples_per_second": 201.437,
|
| 1138 |
+
"eval_steps_per_second": 3.149,
|
| 1139 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1140 |
+
"flow/cos_sim": 0.5591950585593038,
|
| 1141 |
+
"flow/improvement_ratio": 0.8969450570118707,
|
| 1142 |
+
"flow/mag_ratio_mean": 0.5669063877449361,
|
| 1143 |
+
"flow/mag_ratio_std": 0.26610757334273,
|
| 1144 |
+
"step": 18432
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.8053957377119831,
|
| 1148 |
+
"grad_norm": 1.2096973657608032,
|
| 1149 |
+
"learning_rate": 9.462255732982089e-05,
|
| 1150 |
+
"loss": 1.3907443284988403,
|
| 1151 |
+
"step": 18688
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.8164285560368048,
|
| 1155 |
+
"grad_norm": 0.7669377326965332,
|
| 1156 |
+
"learning_rate": 8.450571160576348e-05,
|
| 1157 |
+
"loss": 1.3931423425674438,
|
| 1158 |
+
"step": 18944
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.8274613743616265,
|
| 1162 |
+
"grad_norm": 0.6018663644790649,
|
| 1163 |
+
"learning_rate": 7.491072553698764e-05,
|
| 1164 |
+
"loss": 1.3906762599945068,
|
| 1165 |
+
"step": 19200
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.8384941926864482,
|
| 1169 |
+
"grad_norm": 0.5031753182411194,
|
| 1170 |
+
"learning_rate": 6.584965039895586e-05,
|
| 1171 |
+
"loss": 1.387755274772644,
|
| 1172 |
+
"step": 19456
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.8384941926864482,
|
| 1176 |
+
"eval_bleu": 0.9374315797458206,
|
| 1177 |
+
"eval_cos_loss": 0.4402599502474006,
|
| 1178 |
+
"eval_dec_loss": 0.11256631666909593,
|
| 1179 |
+
"eval_loss": 1.3931237858241554,
|
| 1180 |
+
"eval_mse2_loss": 0.14820685115323137,
|
| 1181 |
+
"eval_mse_loss": 1.2449169349568738,
|
| 1182 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1183 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1184 |
+
"flow/cos_sim": 0.5597400505151322,
|
| 1185 |
+
"flow/improvement_ratio": 0.8949816590433182,
|
| 1186 |
+
"flow/mag_ratio_mean": 0.5695700204448659,
|
| 1187 |
+
"flow/mag_ratio_std": 0.2675396429895084,
|
| 1188 |
+
"step": 19456
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.8384941926864482,
|
| 1192 |
+
"eval_bleu": 0.9374315797458206,
|
| 1193 |
+
"eval_cos_loss": 0.4402599502474006,
|
| 1194 |
+
"eval_dec_loss": 0.11256631666909593,
|
| 1195 |
+
"eval_loss": 1.3931237858241554,
|
| 1196 |
+
"eval_mse2_loss": 0.14820685115323137,
|
| 1197 |
+
"eval_mse_loss": 1.2449169349568738,
|
| 1198 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1199 |
+
"eval_runtime": 151.1209,
|
| 1200 |
+
"eval_samples_per_second": 198.517,
|
| 1201 |
+
"eval_steps_per_second": 3.103,
|
| 1202 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1203 |
+
"flow/cos_sim": 0.5597400505151322,
|
| 1204 |
+
"flow/improvement_ratio": 0.8949816590433182,
|
| 1205 |
+
"flow/mag_ratio_mean": 0.5695700204448659,
|
| 1206 |
+
"flow/mag_ratio_std": 0.2675396429895084,
|
| 1207 |
+
"step": 19456
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 0.8495270110112698,
|
| 1211 |
+
"grad_norm": 0.5668984055519104,
|
| 1212 |
+
"learning_rate": 5.73338668765051e-05,
|
| 1213 |
+
"loss": 1.3939695358276367,
|
| 1214 |
+
"step": 19712
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.8605598293360915,
|
| 1218 |
+
"grad_norm": 0.5667815208435059,
|
| 1219 |
+
"learning_rate": 4.9374070769740984e-05,
|
| 1220 |
+
"loss": 1.392187237739563,
|
| 1221 |
+
"step": 19968
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.8715926476609133,
|
| 1225 |
+
"grad_norm": 0.8259353637695312,
|
| 1226 |
+
"learning_rate": 4.198025956014095e-05,
|
| 1227 |
+
"loss": 1.3923636674880981,
|
| 1228 |
+
"step": 20224
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.8826254659857349,
|
| 1232 |
+
"grad_norm": 1.0611343383789062,
|
| 1233 |
+
"learning_rate": 3.516171985374755e-05,
|
| 1234 |
+
"loss": 1.3925312757492065,
|
| 1235 |
+
"step": 20480
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 0.8826254659857349,
|
| 1239 |
+
"eval_bleu": 0.9374427943864152,
|
| 1240 |
+
"eval_cos_loss": 0.44028231228338377,
|
| 1241 |
+
"eval_dec_loss": 0.11508091816158374,
|
| 1242 |
+
"eval_loss": 1.3936327850894887,
|
| 1243 |
+
"eval_mse2_loss": 0.14858852614408363,
|
| 1244 |
+
"eval_mse_loss": 1.2450442581034418,
|
| 1245 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1246 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1247 |
+
"flow/cos_sim": 0.5597176894958593,
|
| 1248 |
+
"flow/improvement_ratio": 0.8979350520349515,
|
| 1249 |
+
"flow/mag_ratio_mean": 0.5660728585999658,
|
| 1250 |
+
"flow/mag_ratio_std": 0.2672899484253133,
|
| 1251 |
+
"step": 20480
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.8826254659857349,
|
| 1255 |
+
"eval_bleu": 0.9374427943864152,
|
| 1256 |
+
"eval_cos_loss": 0.44028231228338377,
|
| 1257 |
+
"eval_dec_loss": 0.11508091816158374,
|
| 1258 |
+
"eval_loss": 1.3936327850894887,
|
| 1259 |
+
"eval_mse2_loss": 0.14858852614408363,
|
| 1260 |
+
"eval_mse_loss": 1.2450442581034418,
|
| 1261 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1262 |
+
"eval_runtime": 153.7462,
|
| 1263 |
+
"eval_samples_per_second": 195.127,
|
| 1264 |
+
"eval_steps_per_second": 3.05,
|
| 1265 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1266 |
+
"flow/cos_sim": 0.5597176894958593,
|
| 1267 |
+
"flow/improvement_ratio": 0.8979350520349515,
|
| 1268 |
+
"flow/mag_ratio_mean": 0.5660728585999658,
|
| 1269 |
+
"flow/mag_ratio_std": 0.2672899484253133,
|
| 1270 |
+
"step": 20480
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 0.8936582843105566,
|
| 1274 |
+
"grad_norm": 0.3865830898284912,
|
| 1275 |
+
"learning_rate": 2.8927015717215733e-05,
|
| 1276 |
+
"loss": 1.3925455808639526,
|
| 1277 |
+
"step": 20736
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.9046911026353783,
|
| 1281 |
+
"grad_norm": 0.5185984373092651,
|
| 1282 |
+
"learning_rate": 2.3283977921370547e-05,
|
| 1283 |
+
"loss": 1.391736388206482,
|
| 1284 |
+
"step": 20992
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.9157239209601999,
|
| 1288 |
+
"grad_norm": 0.5432521104812622,
|
| 1289 |
+
"learning_rate": 1.8239694105780413e-05,
|
| 1290 |
+
"loss": 1.3918673992156982,
|
| 1291 |
+
"step": 21248
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 0.9267567392850217,
|
| 1295 |
+
"grad_norm": 0.3220946788787842,
|
| 1296 |
+
"learning_rate": 1.3800499876701955e-05,
|
| 1297 |
+
"loss": 1.3899786472320557,
|
| 1298 |
+
"step": 21504
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.9267567392850217,
|
| 1302 |
+
"eval_bleu": 0.9376202589977451,
|
| 1303 |
+
"eval_cos_loss": 0.4390181550211998,
|
| 1304 |
+
"eval_dec_loss": 0.11783787643691815,
|
| 1305 |
+
"eval_loss": 1.3900849003273288,
|
| 1306 |
+
"eval_mse2_loss": 0.14892718683618472,
|
| 1307 |
+
"eval_mse_loss": 1.2411577106793044,
|
| 1308 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1309 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1310 |
+
"flow/cos_sim": 0.5609818455506997,
|
| 1311 |
+
"flow/improvement_ratio": 0.8963389490713188,
|
| 1312 |
+
"flow/mag_ratio_mean": 0.5705793747769744,
|
| 1313 |
+
"flow/mag_ratio_std": 0.27008461628133046,
|
| 1314 |
+
"step": 21504
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.9267567392850217,
|
| 1318 |
+
"eval_bleu": 0.9376202589977451,
|
| 1319 |
+
"eval_cos_loss": 0.4390181550211998,
|
| 1320 |
+
"eval_dec_loss": 0.11783787643691815,
|
| 1321 |
+
"eval_loss": 1.3900849003273288,
|
| 1322 |
+
"eval_mse2_loss": 0.14892718683618472,
|
| 1323 |
+
"eval_mse_loss": 1.2411577106793044,
|
| 1324 |
+
"eval_rec_loss": 0.047009017791098624,
|
| 1325 |
+
"eval_runtime": 150.1598,
|
| 1326 |
+
"eval_samples_per_second": 199.787,
|
| 1327 |
+
"eval_steps_per_second": 3.123,
|
| 1328 |
+
"eval_var_loss": 0.01723895594080501,
|
| 1329 |
+
"flow/cos_sim": 0.5609818455506997,
|
| 1330 |
+
"flow/improvement_ratio": 0.8963389490713188,
|
| 1331 |
+
"flow/mag_ratio_mean": 0.5705793747769744,
|
| 1332 |
+
"flow/mag_ratio_std": 0.27008461628133046,
|
| 1333 |
+
"step": 21504
|
| 1334 |
+
}
|
| 1335 |
+
],
|
| 1336 |
+
"logging_steps": 256,
|
| 1337 |
+
"max_steps": 23204,
|
| 1338 |
+
"num_input_tokens_seen": 0,
|
| 1339 |
+
"num_train_epochs": 1,
|
| 1340 |
+
"save_steps": 1024,
|
| 1341 |
+
"stateful_callbacks": {
|
| 1342 |
+
"TrainerControl": {
|
| 1343 |
+
"args": {
|
| 1344 |
+
"should_epoch_stop": false,
|
| 1345 |
+
"should_evaluate": false,
|
| 1346 |
+
"should_log": false,
|
| 1347 |
+
"should_save": true,
|
| 1348 |
+
"should_training_stop": false
|
| 1349 |
+
},
|
| 1350 |
+
"attributes": {}
|
| 1351 |
+
}
|
| 1352 |
+
},
|
| 1353 |
+
"total_flos": 0.0,
|
| 1354 |
+
"train_batch_size": 64,
|
| 1355 |
+
"trial_name": null,
|
| 1356 |
+
"trial_params": null
|
| 1357 |
+
}
|
checkpoints-d1.3/checkpoint-21504/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61a16bb839f687414b8e48611327c4b9cfddeefe38c031ca70808f9a97c476b7
|
| 3 |
+
size 5137
|