Upload folder using huggingface_hub
Browse files- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/ema.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/model.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/optimizer.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/rng_state.pth +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/scaler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/scheduler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/trainer_state.json +1147 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/training_args.bin +3 -0
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:327f0412b3f020b86a9cccd7cdefbd305c15bb222b5000ef01e6649a5c70370c
|
| 3 |
+
size 550088
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80382ec03eaf0a12d28216919382670b75328bb12d7589af7fe3963e1c9c4c50
|
| 3 |
+
size 19331448
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2abb8f2b58e2badd6ef7a62f9c83c87348dafe00dbe24f9f481a95bb7aa5127
|
| 3 |
+
size 1175115
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5e46d1e941b1e6770d70eb593b1d56f6673c57d4816ab6f614779a84f1a34c0
|
| 3 |
+
size 14645
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8d65a18be8ad144f1452a4ee733f0265cea6c029232afcd4bec76d99776d19a6
|
| 3 |
+
size 1383
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d4857f67bc6bf15cb08e1beb481bc3347e36ec24fc500bcddd2fd4f7d346731
|
| 3 |
+
size 1465
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/trainer_state.json
ADDED
|
@@ -0,0 +1,1147 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 21651,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 0.8327222466468811,
|
| 15 |
+
"learning_rate": 0.000490234375,
|
| 16 |
+
"loss": 2.8200738430023193,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.3061302900314331,
|
| 22 |
+
"learning_rate": 0.000990234375,
|
| 23 |
+
"loss": 1.868248462677002,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.39593741297721863,
|
| 29 |
+
"learning_rate": 0.0009996521692610218,
|
| 30 |
+
"loss": 1.6698616743087769,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.5593295097351074,
|
| 36 |
+
"learning_rate": 0.0009985813302292343,
|
| 37 |
+
"loss": 1.608977198600769,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.2846408728595194,
|
| 43 |
+
"eval_loss": 1.5251963065095144,
|
| 44 |
+
"eval_mse_loss": 0.9559145632399816,
|
| 45 |
+
"flow/cos_sim": 0.7153591378910901,
|
| 46 |
+
"flow/improvement_ratio": 0.47883502166020814,
|
| 47 |
+
"flow/mag_ratio_mean": 0.7032401169816108,
|
| 48 |
+
"flow/mag_ratio_std": 0.1494562604281728,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.2846408728595194,
|
| 54 |
+
"eval_loss": 1.5251963065095144,
|
| 55 |
+
"eval_mse_loss": 0.9559145632399816,
|
| 56 |
+
"eval_runtime": 38.0007,
|
| 57 |
+
"eval_samples_per_second": 736.644,
|
| 58 |
+
"eval_steps_per_second": 11.526,
|
| 59 |
+
"flow/cos_sim": 0.7153591378910901,
|
| 60 |
+
"flow/improvement_ratio": 0.47883502166020814,
|
| 61 |
+
"flow/mag_ratio_mean": 0.7032401169816108,
|
| 62 |
+
"flow/mag_ratio_std": 0.1494562604281728,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.19968454539775848,
|
| 68 |
+
"learning_rate": 0.0009967888946875527,
|
| 69 |
+
"loss": 1.5696274042129517,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.3655131757259369,
|
| 75 |
+
"learning_rate": 0.0009942774568270404,
|
| 76 |
+
"loss": 1.5295122861862183,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.4029371440410614,
|
| 82 |
+
"learning_rate": 0.0009910506514504561,
|
| 83 |
+
"loss": 1.4991682767868042,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.2498495876789093,
|
| 89 |
+
"learning_rate": 0.0009871131487116083,
|
| 90 |
+
"loss": 1.4787085056304932,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.25528003949008576,
|
| 96 |
+
"eval_loss": 1.3903144954546403,
|
| 97 |
+
"eval_mse_loss": 0.879754417563138,
|
| 98 |
+
"flow/cos_sim": 0.7447199556109023,
|
| 99 |
+
"flow/improvement_ratio": 0.47933868948183106,
|
| 100 |
+
"flow/mag_ratio_mean": 0.7272915401959528,
|
| 101 |
+
"flow/mag_ratio_std": 0.1570750494151627,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.25528003949008576,
|
| 107 |
+
"eval_loss": 1.3903144954546403,
|
| 108 |
+
"eval_mse_loss": 0.879754417563138,
|
| 109 |
+
"eval_runtime": 37.9969,
|
| 110 |
+
"eval_samples_per_second": 736.717,
|
| 111 |
+
"eval_steps_per_second": 11.527,
|
| 112 |
+
"flow/cos_sim": 0.7447199556109023,
|
| 113 |
+
"flow/improvement_ratio": 0.47933868948183106,
|
| 114 |
+
"flow/mag_ratio_mean": 0.7272915401959528,
|
| 115 |
+
"flow/mag_ratio_std": 0.1570750494151627,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.940779447555542,
|
| 121 |
+
"learning_rate": 0.0009824706473562416,
|
| 122 |
+
"loss": 1.4612160921096802,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.9774911999702454,
|
| 128 |
+
"learning_rate": 0.0009771298664742456,
|
| 129 |
+
"loss": 1.4491184949874878,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.2685326337814331,
|
| 135 |
+
"learning_rate": 0.00097109853577512,
|
| 136 |
+
"loss": 1.4366470575332642,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.577179491519928,
|
| 142 |
+
"learning_rate": 0.0009643853844007692,
|
| 143 |
+
"loss": 1.423431158065796,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.24748063815510982,
|
| 149 |
+
"eval_loss": 1.3509440585358503,
|
| 150 |
+
"eval_mse_loss": 0.8559827820895469,
|
| 151 |
+
"flow/cos_sim": 0.7525193725274578,
|
| 152 |
+
"flow/improvement_ratio": 0.48529497612284744,
|
| 153 |
+
"flow/mag_ratio_mean": 0.742133051429165,
|
| 154 |
+
"flow/mag_ratio_std": 0.15954837645297726,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.24748063815510982,
|
| 160 |
+
"eval_loss": 1.3509440585358503,
|
| 161 |
+
"eval_mse_loss": 0.8559827820895469,
|
| 162 |
+
"eval_runtime": 38.3549,
|
| 163 |
+
"eval_samples_per_second": 729.842,
|
| 164 |
+
"eval_steps_per_second": 11.42,
|
| 165 |
+
"flow/cos_sim": 0.7525193725274578,
|
| 166 |
+
"flow/improvement_ratio": 0.48529497612284744,
|
| 167 |
+
"flow/mag_ratio_mean": 0.742133051429165,
|
| 168 |
+
"flow/mag_ratio_std": 0.15954837645297726,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.23250754177570343,
|
| 174 |
+
"learning_rate": 0.0009570001282918202,
|
| 175 |
+
"loss": 1.4150209426879883,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 1.1195614337921143,
|
| 181 |
+
"learning_rate": 0.0009489534561257458,
|
| 182 |
+
"loss": 1.4101005792617798,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.6475523710250854,
|
| 188 |
+
"learning_rate": 0.0009402570138471477,
|
| 189 |
+
"loss": 1.3949507474899292,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 0.3041413426399231,
|
| 195 |
+
"learning_rate": 0.0009309233878125867,
|
| 196 |
+
"loss": 1.398905634880066,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.2384823898202208,
|
| 202 |
+
"eval_loss": 1.3082205555210376,
|
| 203 |
+
"eval_mse_loss": 0.8312557756084286,
|
| 204 |
+
"flow/cos_sim": 0.7615176290954084,
|
| 205 |
+
"flow/improvement_ratio": 0.4767781115694133,
|
| 206 |
+
"flow/mag_ratio_mean": 0.7431108893052628,
|
| 207 |
+
"flow/mag_ratio_std": 0.16565074574321373,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.2384823898202208,
|
| 213 |
+
"eval_loss": 1.3082205555210376,
|
| 214 |
+
"eval_mse_loss": 0.8312557756084286,
|
| 215 |
+
"eval_runtime": 38.1625,
|
| 216 |
+
"eval_samples_per_second": 733.521,
|
| 217 |
+
"eval_steps_per_second": 11.477,
|
| 218 |
+
"flow/cos_sim": 0.7615176290954084,
|
| 219 |
+
"flow/improvement_ratio": 0.4767781115694133,
|
| 220 |
+
"flow/mag_ratio_mean": 0.7431108893052628,
|
| 221 |
+
"flow/mag_ratio_std": 0.16565074574321373,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.37201106548309326,
|
| 227 |
+
"learning_rate": 0.0009209660865743546,
|
| 228 |
+
"loss": 1.3921161890029907,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.47654396295547485,
|
| 234 |
+
"learning_rate": 0.0009103995213295538,
|
| 235 |
+
"loss": 1.3823015689849854,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.24332450330257416,
|
| 241 |
+
"learning_rate": 0.0008992389850627786,
|
| 242 |
+
"loss": 1.3821202516555786,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.6045255661010742,
|
| 248 |
+
"learning_rate": 0.0008875006304125866,
|
| 249 |
+
"loss": 1.368703842163086,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.2360525783066336,
|
| 255 |
+
"eval_loss": 1.2967310523333615,
|
| 256 |
+
"eval_mse_loss": 0.8246258939510067,
|
| 257 |
+
"flow/cos_sim": 0.7639474261841274,
|
| 258 |
+
"flow/improvement_ratio": 0.4758605305870918,
|
| 259 |
+
"flow/mag_ratio_mean": 0.751331170672151,
|
| 260 |
+
"flow/mag_ratio_std": 0.1651398052363635,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.2360525783066336,
|
| 266 |
+
"eval_loss": 1.2967310523333615,
|
| 267 |
+
"eval_mse_loss": 0.8246258939510067,
|
| 268 |
+
"eval_runtime": 37.8667,
|
| 269 |
+
"eval_samples_per_second": 739.25,
|
| 270 |
+
"eval_steps_per_second": 11.567,
|
| 271 |
+
"flow/cos_sim": 0.7639474261841274,
|
| 272 |
+
"flow/improvement_ratio": 0.4758605305870918,
|
| 273 |
+
"flow/mag_ratio_mean": 0.751331170672151,
|
| 274 |
+
"flow/mag_ratio_std": 0.1651398052363635,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 1.2842298746109009,
|
| 280 |
+
"learning_rate": 0.0008752014462937921,
|
| 281 |
+
"loss": 1.3652393817901611,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 1.3890047073364258,
|
| 287 |
+
"learning_rate": 0.0008623592333094195,
|
| 288 |
+
"loss": 1.3665603399276733,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 1.1705548763275146,
|
| 294 |
+
"learning_rate": 0.0008489925779878984,
|
| 295 |
+
"loss": 1.3597829341888428,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.5006002187728882,
|
| 301 |
+
"learning_rate": 0.0008351208258827912,
|
| 302 |
+
"loss": 1.3598123788833618,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.237595906710788,
|
| 308 |
+
"eval_loss": 1.302836928465595,
|
| 309 |
+
"eval_mse_loss": 0.8276451136151405,
|
| 310 |
+
"flow/cos_sim": 0.7624041227173043,
|
| 311 |
+
"flow/improvement_ratio": 0.48030529528448024,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7537709809329411,
|
| 313 |
+
"flow/mag_ratio_std": 0.1697745155798246,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.237595906710788,
|
| 319 |
+
"eval_loss": 1.302836928465595,
|
| 320 |
+
"eval_mse_loss": 0.8276451136151405,
|
| 321 |
+
"eval_runtime": 37.8433,
|
| 322 |
+
"eval_samples_per_second": 739.708,
|
| 323 |
+
"eval_steps_per_second": 11.574,
|
| 324 |
+
"flow/cos_sim": 0.7624041227173043,
|
| 325 |
+
"flow/improvement_ratio": 0.48030529528448024,
|
| 326 |
+
"flow/mag_ratio_mean": 0.7537709809329411,
|
| 327 |
+
"flow/mag_ratio_std": 0.1697745155798246,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 1.9024178981781006,
|
| 333 |
+
"learning_rate": 0.0008207640535739817,
|
| 334 |
+
"loss": 1.3518272638320923,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 1.5868515968322754,
|
| 340 |
+
"learning_rate": 0.0008059430396108527,
|
| 341 |
+
"loss": 1.3528562784194946,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 2.2589821815490723,
|
| 347 |
+
"learning_rate": 0.0007906792344395009,
|
| 348 |
+
"loss": 1.3518447875976562,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.36628156900405884,
|
| 354 |
+
"learning_rate": 0.000774994729357517,
|
| 355 |
+
"loss": 1.3518491983413696,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.23038810593624637,
|
| 361 |
+
"eval_loss": 1.2705201430407833,
|
| 362 |
+
"eval_mse_loss": 0.8097439307600396,
|
| 363 |
+
"flow/cos_sim": 0.769611899166891,
|
| 364 |
+
"flow/improvement_ratio": 0.48191102304959405,
|
| 365 |
+
"flow/mag_ratio_mean": 0.7535693184153675,
|
| 366 |
+
"flow/mag_ratio_std": 0.17145597665957665,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.23038810593624637,
|
| 372 |
+
"eval_loss": 1.2705201430407833,
|
| 373 |
+
"eval_mse_loss": 0.8097439307600396,
|
| 374 |
+
"eval_runtime": 38.3324,
|
| 375 |
+
"eval_samples_per_second": 730.27,
|
| 376 |
+
"eval_steps_per_second": 11.426,
|
| 377 |
+
"flow/cos_sim": 0.769611899166891,
|
| 378 |
+
"flow/improvement_ratio": 0.48191102304959405,
|
| 379 |
+
"flow/mag_ratio_mean": 0.7535693184153675,
|
| 380 |
+
"flow/mag_ratio_std": 0.17145597665957665,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.30675724148750305,
|
| 386 |
+
"learning_rate": 0.000758912224541262,
|
| 387 |
+
"loss": 1.3507543802261353,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.6440773606300354,
|
| 393 |
+
"learning_rate": 0.000742454996191914,
|
| 394 |
+
"loss": 1.3481366634368896,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.9887242317199707,
|
| 400 |
+
"learning_rate": 0.0007256468628478321,
|
| 401 |
+
"loss": 1.346606969833374,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.4718647003173828,
|
| 407 |
+
"learning_rate": 0.0007085121509119993,
|
| 408 |
+
"loss": 1.3463300466537476,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.2299874077240626,
|
| 414 |
+
"eval_loss": 1.2689139097248583,
|
| 415 |
+
"eval_mse_loss": 0.8089390935963148,
|
| 416 |
+
"flow/cos_sim": 0.7700126173833733,
|
| 417 |
+
"flow/improvement_ratio": 0.4802750931482881,
|
| 418 |
+
"flow/mag_ratio_mean": 0.7505737072256602,
|
| 419 |
+
"flow/mag_ratio_std": 0.17273653373462425,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.2299874077240626,
|
| 425 |
+
"eval_loss": 1.2689139097248583,
|
| 426 |
+
"eval_mse_loss": 0.8089390935963148,
|
| 427 |
+
"eval_runtime": 38.2412,
|
| 428 |
+
"eval_samples_per_second": 732.012,
|
| 429 |
+
"eval_steps_per_second": 11.454,
|
| 430 |
+
"flow/cos_sim": 0.7700126173833733,
|
| 431 |
+
"flow/improvement_ratio": 0.4802750931482881,
|
| 432 |
+
"flow/mag_ratio_mean": 0.7505737072256602,
|
| 433 |
+
"flow/mag_ratio_std": 0.17273653373462425,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.5701487064361572,
|
| 439 |
+
"learning_rate": 0.000691075659444431,
|
| 440 |
+
"loss": 1.3397765159606934,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.2903459966182709,
|
| 446 |
+
"learning_rate": 0.0006733626242705082,
|
| 447 |
+
"loss": 1.3429921865463257,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 1.0731102228164673,
|
| 453 |
+
"learning_rate": 0.0006553986814571795,
|
| 454 |
+
"loss": 1.3373149633407593,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 2.3944308757781982,
|
| 460 |
+
"learning_rate": 0.0006372098302098944,
|
| 461 |
+
"loss": 1.3349826335906982,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.22907086808795799,
|
| 467 |
+
"eval_loss": 1.263627827984013,
|
| 468 |
+
"eval_mse_loss": 0.8054860928287245,
|
| 469 |
+
"flow/cos_sim": 0.7709291401791246,
|
| 470 |
+
"flow/improvement_ratio": 0.4783504515478056,
|
| 471 |
+
"flow/mag_ratio_mean": 0.7555886954749556,
|
| 472 |
+
"flow/mag_ratio_std": 0.17359172498390554,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.22907086808795799,
|
| 478 |
+
"eval_loss": 1.263627827984013,
|
| 479 |
+
"eval_mse_loss": 0.8054860928287245,
|
| 480 |
+
"eval_runtime": 38.1725,
|
| 481 |
+
"eval_samples_per_second": 733.329,
|
| 482 |
+
"eval_steps_per_second": 11.474,
|
| 483 |
+
"flow/cos_sim": 0.7709291401791246,
|
| 484 |
+
"flow/improvement_ratio": 0.4783504515478056,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7555886954749556,
|
| 486 |
+
"flow/mag_ratio_std": 0.17359172498390554,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.8406338691711426,
|
| 492 |
+
"learning_rate": 0.0006188223952439671,
|
| 493 |
+
"loss": 1.3365044593811035,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.4198502004146576,
|
| 499 |
+
"learning_rate": 0.0006002629886848276,
|
| 500 |
+
"loss": 1.3363875150680542,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.25390586256980896,
|
| 506 |
+
"learning_rate": 0.0005815584715523073,
|
| 507 |
+
"loss": 1.3315783739089966,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 1.6368358135223389,
|
| 513 |
+
"learning_rate": 0.0005627359148846979,
|
| 514 |
+
"loss": 1.3290125131607056,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.22714132124974848,
|
| 520 |
+
"eval_loss": 1.2561225123601416,
|
| 521 |
+
"eval_mse_loss": 0.8018398724462343,
|
| 522 |
+
"flow/cos_sim": 0.7728586847379327,
|
| 523 |
+
"flow/improvement_ratio": 0.4823647101311923,
|
| 524 |
+
"flow/mag_ratio_mean": 0.7559084860973706,
|
| 525 |
+
"flow/mag_ratio_std": 0.17520164426195023,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.22714132124974848,
|
| 531 |
+
"eval_loss": 1.2561225123601416,
|
| 532 |
+
"eval_mse_loss": 0.8018398724462343,
|
| 533 |
+
"eval_runtime": 38.2497,
|
| 534 |
+
"eval_samples_per_second": 731.848,
|
| 535 |
+
"eval_steps_per_second": 11.451,
|
| 536 |
+
"flow/cos_sim": 0.7728586847379327,
|
| 537 |
+
"flow/improvement_ratio": 0.4823647101311923,
|
| 538 |
+
"flow/mag_ratio_mean": 0.7559084860973706,
|
| 539 |
+
"flow/mag_ratio_std": 0.17520164426195023,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.28762465715408325,
|
| 545 |
+
"learning_rate": 0.0005438225605588509,
|
| 546 |
+
"loss": 1.3272242546081543,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 1.8320599794387817,
|
| 552 |
+
"learning_rate": 0.0005248457818630231,
|
| 553 |
+
"loss": 1.3262661695480347,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 1.9320905208587646,
|
| 559 |
+
"learning_rate": 0.0005058330438795307,
|
| 560 |
+
"loss": 1.3232632875442505,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.5674813985824585,
|
| 566 |
+
"learning_rate": 0.0004868118637345485,
|
| 567 |
+
"loss": 1.3243262767791748,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.2250166033228783,
|
| 573 |
+
"eval_loss": 1.2463049975704386,
|
| 574 |
+
"eval_mse_loss": 0.7962717901081799,
|
| 575 |
+
"flow/cos_sim": 0.7749834096050698,
|
| 576 |
+
"flow/improvement_ratio": 0.4846222329901778,
|
| 577 |
+
"flow/mag_ratio_mean": 0.7608566733255778,
|
| 578 |
+
"flow/mag_ratio_std": 0.1756929681725698,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.2250166033228783,
|
| 584 |
+
"eval_loss": 1.2463049975704386,
|
| 585 |
+
"eval_mse_loss": 0.7962717901081799,
|
| 586 |
+
"eval_runtime": 38.6094,
|
| 587 |
+
"eval_samples_per_second": 725.031,
|
| 588 |
+
"eval_steps_per_second": 11.344,
|
| 589 |
+
"flow/cos_sim": 0.7749834096050698,
|
| 590 |
+
"flow/improvement_ratio": 0.4846222329901778,
|
| 591 |
+
"flow/mag_ratio_mean": 0.7608566733255778,
|
| 592 |
+
"flow/mag_ratio_std": 0.1756929681725698,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.5362160205841064,
|
| 598 |
+
"learning_rate": 0.0004678097707725879,
|
| 599 |
+
"loss": 1.3213114738464355,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.891897439956665,
|
| 605 |
+
"learning_rate": 0.00044885426671328947,
|
| 606 |
+
"loss": 1.3259453773498535,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 0.7019096612930298,
|
| 612 |
+
"learning_rate": 0.00043004636220449254,
|
| 613 |
+
"loss": 1.3259971141815186,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.5158335566520691,
|
| 619 |
+
"learning_rate": 0.0004112657828038095,
|
| 620 |
+
"loss": 1.3186417818069458,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.22481494250635034,
|
| 626 |
+
"eval_loss": 1.2442510468230399,
|
| 627 |
+
"eval_mse_loss": 0.794621159905168,
|
| 628 |
+
"flow/cos_sim": 0.7751850770216554,
|
| 629 |
+
"flow/improvement_ratio": 0.4843696723926013,
|
| 630 |
+
"flow/mag_ratio_mean": 0.7540967676737537,
|
| 631 |
+
"flow/mag_ratio_std": 0.17127465946625356,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.22481494250635034,
|
| 637 |
+
"eval_loss": 1.2442510468230399,
|
| 638 |
+
"eval_mse_loss": 0.794621159905168,
|
| 639 |
+
"eval_runtime": 38.4289,
|
| 640 |
+
"eval_samples_per_second": 728.436,
|
| 641 |
+
"eval_steps_per_second": 11.398,
|
| 642 |
+
"flow/cos_sim": 0.7751850770216554,
|
| 643 |
+
"flow/improvement_ratio": 0.4843696723926013,
|
| 644 |
+
"flow/mag_ratio_mean": 0.7540967676737537,
|
| 645 |
+
"flow/mag_ratio_std": 0.17127465946625356,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 1.6348507404327393,
|
| 651 |
+
"learning_rate": 0.0003926136283912977,
|
| 652 |
+
"loss": 1.3202893733978271,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.9154382348060608,
|
| 658 |
+
"learning_rate": 0.0003741168942207097,
|
| 659 |
+
"loss": 1.3224493265151978,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.36584168672561646,
|
| 665 |
+
"learning_rate": 0.00035580235060616054,
|
| 666 |
+
"loss": 1.3221790790557861,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.7461158633232117,
|
| 672 |
+
"learning_rate": 0.0003376965041774639,
|
| 673 |
+
"loss": 1.3157620429992676,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.22370701786726033,
|
| 679 |
+
"eval_loss": 1.2403812901070128,
|
| 680 |
+
"eval_mse_loss": 0.7929672547127014,
|
| 681 |
+
"flow/cos_sim": 0.77629301074433,
|
| 682 |
+
"flow/improvement_ratio": 0.47866371961216947,
|
| 683 |
+
"flow/mag_ratio_mean": 0.7604604626082938,
|
| 684 |
+
"flow/mag_ratio_std": 0.176010451462443,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.22370701786726033,
|
| 690 |
+
"eval_loss": 1.2403812901070128,
|
| 691 |
+
"eval_mse_loss": 0.7929672547127014,
|
| 692 |
+
"eval_runtime": 38.1923,
|
| 693 |
+
"eval_samples_per_second": 732.949,
|
| 694 |
+
"eval_steps_per_second": 11.468,
|
| 695 |
+
"flow/cos_sim": 0.77629301074433,
|
| 696 |
+
"flow/improvement_ratio": 0.47866371961216947,
|
| 697 |
+
"flow/mag_ratio_mean": 0.7604604626082938,
|
| 698 |
+
"flow/mag_ratio_std": 0.176010451462443,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.5198074579238892,
|
| 704 |
+
"learning_rate": 0.0003198255595171013,
|
| 705 |
+
"loss": 1.311034083366394,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.20684662461280823,
|
| 711 |
+
"learning_rate": 0.000302215381234342,
|
| 712 |
+
"loss": 1.3143961429595947,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.2593955099582672,
|
| 718 |
+
"learning_rate": 0.0002849585386095228,
|
| 719 |
+
"loss": 1.3148876428604126,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.7760673761367798,
|
| 725 |
+
"learning_rate": 0.0002679446760672468,
|
| 726 |
+
"loss": 1.3120242357254028,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.22288279759285112,
|
| 732 |
+
"eval_loss": 1.23468789390233,
|
| 733 |
+
"eval_mse_loss": 0.7889222980362095,
|
| 734 |
+
"flow/cos_sim": 0.7771172157433479,
|
| 735 |
+
"flow/improvement_ratio": 0.4807521503537757,
|
| 736 |
+
"flow/mag_ratio_mean": 0.7600993149356755,
|
| 737 |
+
"flow/mag_ratio_std": 0.17531098554667818,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.22288279759285112,
|
| 743 |
+
"eval_loss": 1.23468789390233,
|
| 744 |
+
"eval_mse_loss": 0.7889222980362095,
|
| 745 |
+
"eval_runtime": 38.2466,
|
| 746 |
+
"eval_samples_per_second": 731.908,
|
| 747 |
+
"eval_steps_per_second": 11.452,
|
| 748 |
+
"flow/cos_sim": 0.7771172157433479,
|
| 749 |
+
"flow/improvement_ratio": 0.4807521503537757,
|
| 750 |
+
"flow/mag_ratio_mean": 0.7600993149356755,
|
| 751 |
+
"flow/mag_ratio_std": 0.17531098554667818,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.7773881554603577,
|
| 757 |
+
"learning_rate": 0.0002512666670792449,
|
| 758 |
+
"loss": 1.3133234977722168,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 0.33675631880760193,
|
| 764 |
+
"learning_rate": 0.00023494864971958463,
|
| 765 |
+
"loss": 1.3156976699829102,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.39987826347351074,
|
| 771 |
+
"learning_rate": 0.0002190142410466297,
|
| 772 |
+
"loss": 1.312076210975647,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 1.2741974592208862,
|
| 778 |
+
"learning_rate": 0.00020348650292208355,
|
| 779 |
+
"loss": 1.3096755743026733,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.22364824656482157,
|
| 785 |
+
"eval_loss": 1.2394148476047604,
|
| 786 |
+
"eval_mse_loss": 0.7921183559720375,
|
| 787 |
+
"flow/cos_sim": 0.7763517757528993,
|
| 788 |
+
"flow/improvement_ratio": 0.48859692844626024,
|
| 789 |
+
"flow/mag_ratio_mean": 0.7602596438094361,
|
| 790 |
+
"flow/mag_ratio_std": 0.1753684315643354,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.22364824656482157,
|
| 796 |
+
"eval_loss": 1.2394148476047604,
|
| 797 |
+
"eval_mse_loss": 0.7921183559720375,
|
| 798 |
+
"eval_runtime": 38.2191,
|
| 799 |
+
"eval_samples_per_second": 732.436,
|
| 800 |
+
"eval_steps_per_second": 11.46,
|
| 801 |
+
"flow/cos_sim": 0.7763517757528993,
|
| 802 |
+
"flow/improvement_ratio": 0.48859692844626024,
|
| 803 |
+
"flow/mag_ratio_mean": 0.7602596438094361,
|
| 804 |
+
"flow/mag_ratio_std": 0.1753684315643354,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.91872239112854,
|
| 810 |
+
"learning_rate": 0.0001883879086335682,
|
| 811 |
+
"loss": 1.3155561685562134,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.7235752940177917,
|
| 817 |
+
"learning_rate": 0.00017374031036904646,
|
| 818 |
+
"loss": 1.3148996829986572,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 1.5837082862854004,
|
| 824 |
+
"learning_rate": 0.00015956490759016107,
|
| 825 |
+
"loss": 1.307360291481018,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.2564609944820404,
|
| 831 |
+
"learning_rate": 0.00014588221635026462,
|
| 832 |
+
"loss": 1.307948112487793,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.22264472640132252,
|
| 838 |
+
"eval_loss": 1.234629833262805,
|
| 839 |
+
"eval_mse_loss": 0.7893403812086202,
|
| 840 |
+
"flow/cos_sim": 0.7773553015978913,
|
| 841 |
+
"flow/improvement_ratio": 0.48276225923131044,
|
| 842 |
+
"flow/mag_ratio_mean": 0.7610696509276351,
|
| 843 |
+
"flow/mag_ratio_std": 0.17544000585601754,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.22264472640132252,
|
| 849 |
+
"eval_loss": 1.234629833262805,
|
| 850 |
+
"eval_mse_loss": 0.7893403812086202,
|
| 851 |
+
"eval_runtime": 38.0825,
|
| 852 |
+
"eval_samples_per_second": 735.063,
|
| 853 |
+
"eval_steps_per_second": 11.501,
|
| 854 |
+
"flow/cos_sim": 0.7773553015978913,
|
| 855 |
+
"flow/improvement_ratio": 0.48276225923131044,
|
| 856 |
+
"flow/mag_ratio_mean": 0.7610696509276351,
|
| 857 |
+
"flow/mag_ratio_std": 0.17544000585601754,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.2357015311717987,
|
| 863 |
+
"learning_rate": 0.00013271203960154576,
|
| 864 |
+
"loss": 1.3056586980819702,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.6545789837837219,
|
| 870 |
+
"learning_rate": 0.0001200734385342262,
|
| 871 |
+
"loss": 1.3069326877593994,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.5553286075592041,
|
| 877 |
+
"learning_rate": 0.00010798470498931034,
|
| 878 |
+
"loss": 1.3104249238967896,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.9544392824172974,
|
| 884 |
+
"learning_rate": 9.646333498481152e-05,
|
| 885 |
+
"loss": 1.3092049360275269,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.2226330109927208,
|
| 891 |
+
"eval_loss": 1.2352391797113635,
|
| 892 |
+
"eval_mse_loss": 0.7899731571815874,
|
| 893 |
+
"flow/cos_sim": 0.7773670049290679,
|
| 894 |
+
"flow/improvement_ratio": 0.48120235573483383,
|
| 895 |
+
"flow/mag_ratio_mean": 0.7611706157525381,
|
| 896 |
+
"flow/mag_ratio_std": 0.17672617512342592,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.2226330109927208,
|
| 902 |
+
"eval_loss": 1.2352391797113635,
|
| 903 |
+
"eval_mse_loss": 0.7899731571815874,
|
| 904 |
+
"eval_runtime": 38.1071,
|
| 905 |
+
"eval_samples_per_second": 734.588,
|
| 906 |
+
"eval_steps_per_second": 11.494,
|
| 907 |
+
"flow/cos_sim": 0.7773670049290679,
|
| 908 |
+
"flow/improvement_ratio": 0.48120235573483383,
|
| 909 |
+
"flow/mag_ratio_mean": 0.7611706157525381,
|
| 910 |
+
"flow/mag_ratio_std": 0.17672617512342592,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.365252286195755,
|
| 916 |
+
"learning_rate": 8.552600339377569e-05,
|
| 917 |
+
"loss": 1.3102151155471802,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.32710692286491394,
|
| 923 |
+
"learning_rate": 7.518853981074431e-05,
|
| 924 |
+
"loss": 1.3044898509979248,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.3619934022426605,
|
| 930 |
+
"learning_rate": 6.550267002511972e-05,
|
| 931 |
+
"loss": 1.305891513824463,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.40305039286613464,
|
| 937 |
+
"learning_rate": 5.6406453984677306e-05,
|
| 938 |
+
"loss": 1.305964469909668,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.2226418528246553,
|
| 944 |
+
"eval_loss": 1.2338021887491828,
|
| 945 |
+
"eval_mse_loss": 0.7885184851411271,
|
| 946 |
+
"flow/cos_sim": 0.7773581763652906,
|
| 947 |
+
"flow/improvement_ratio": 0.4792883807461556,
|
| 948 |
+
"flow/mag_ratio_mean": 0.7609404158102323,
|
| 949 |
+
"flow/mag_ratio_std": 0.17583323059152794,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.2226418528246553,
|
| 955 |
+
"eval_loss": 1.2338021887491828,
|
| 956 |
+
"eval_mse_loss": 0.7885184851411271,
|
| 957 |
+
"eval_runtime": 38.3022,
|
| 958 |
+
"eval_samples_per_second": 730.846,
|
| 959 |
+
"eval_steps_per_second": 11.435,
|
| 960 |
+
"flow/cos_sim": 0.7773581763652906,
|
| 961 |
+
"flow/improvement_ratio": 0.4792883807461556,
|
| 962 |
+
"flow/mag_ratio_mean": 0.7609404158102323,
|
| 963 |
+
"flow/mag_ratio_std": 0.17583323059152794,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.27111852169036865,
|
| 969 |
+
"learning_rate": 4.7952250661536243e-05,
|
| 970 |
+
"loss": 1.3054128885269165,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.2560623586177826,
|
| 976 |
+
"learning_rate": 4.0152295819938015e-05,
|
| 977 |
+
"loss": 1.2992324829101562,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.3480774462223053,
|
| 983 |
+
"learning_rate": 3.3017878330624194e-05,
|
| 984 |
+
"loss": 1.3068798780441284,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 1.4495835304260254,
|
| 990 |
+
"learning_rate": 2.655932383245879e-05,
|
| 991 |
+
"loss": 1.309512972831726,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.22318284805506877,
|
| 997 |
+
"eval_loss": 1.2365219271890648,
|
| 998 |
+
"eval_mse_loss": 0.7901562346171026,
|
| 999 |
+
"flow/cos_sim": 0.7768171618790387,
|
| 1000 |
+
"flow/improvement_ratio": 0.48550828644946287,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.759889675874144,
|
| 1002 |
+
"flow/mag_ratio_std": 0.1753429954522821,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.22318284805506877,
|
| 1008 |
+
"eval_loss": 1.2365219271890648,
|
| 1009 |
+
"eval_mse_loss": 0.7901562346171026,
|
| 1010 |
+
"eval_runtime": 37.9487,
|
| 1011 |
+
"eval_samples_per_second": 737.654,
|
| 1012 |
+
"eval_steps_per_second": 11.542,
|
| 1013 |
+
"flow/cos_sim": 0.7768171618790387,
|
| 1014 |
+
"flow/improvement_ratio": 0.48550828644946287,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.759889675874144,
|
| 1016 |
+
"flow/mag_ratio_std": 0.1753429954522821,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.2634163200855255,
|
| 1022 |
+
"learning_rate": 2.078597978813729e-05,
|
| 1023 |
+
"loss": 1.3040359020233154,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.3606112003326416,
|
| 1029 |
+
"learning_rate": 1.57062019556074e-05,
|
| 1030 |
+
"loss": 1.303109049797058,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.28589293360710144,
|
| 1036 |
+
"learning_rate": 1.1343075046968187e-05,
|
| 1037 |
+
"loss": 1.3101311922073364,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.3070381283760071,
|
| 1043 |
+
"learning_rate": 7.668697354415943e-06,
|
| 1044 |
+
"loss": 1.31185781955719,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.22114521217400626,
|
| 1050 |
+
"eval_loss": 1.227875959927633,
|
| 1051 |
+
"eval_mse_loss": 0.7855855340827002,
|
| 1052 |
+
"flow/cos_sim": 0.7788548091230871,
|
| 1053 |
+
"flow/improvement_ratio": 0.47847176986197903,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.7624764865664042,
|
| 1055 |
+
"flow/mag_ratio_std": 0.17646009566849225,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.22114521217400626,
|
| 1061 |
+
"eval_loss": 1.227875959927633,
|
| 1062 |
+
"eval_mse_loss": 0.7855855340827002,
|
| 1063 |
+
"eval_runtime": 38.0239,
|
| 1064 |
+
"eval_samples_per_second": 736.196,
|
| 1065 |
+
"eval_steps_per_second": 11.519,
|
| 1066 |
+
"flow/cos_sim": 0.7788548091230871,
|
| 1067 |
+
"flow/improvement_ratio": 0.47847176986197903,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.7624764865664042,
|
| 1069 |
+
"flow/mag_ratio_std": 0.17646009566849225,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.30065974593162537,
|
| 1075 |
+
"learning_rate": 4.706870509961658e-06,
|
| 1076 |
+
"loss": 1.3032283782958984,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.5164745450019836,
|
| 1082 |
+
"learning_rate": 2.461881164116053e-06,
|
| 1083 |
+
"loss": 1.3059771060943604,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.1629684418439865,
|
| 1089 |
+
"learning_rate": 9.369784888367261e-07,
|
| 1090 |
+
"loss": 1.3061639070510864,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.3277168869972229,
|
| 1096 |
+
"learning_rate": 1.3436947500289342e-07,
|
| 1097 |
+
"loss": 1.3064048290252686,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.22108487941358732,
|
| 1103 |
+
"eval_loss": 1.22712274991214,
|
| 1104 |
+
"eval_mse_loss": 0.7849529917653837,
|
| 1105 |
+
"flow/cos_sim": 0.7789151420876316,
|
| 1106 |
+
"flow/improvement_ratio": 0.4868433005325326,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.7618113677251285,
|
| 1108 |
+
"flow/mag_ratio_std": 0.17511140633391464,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.22108487941358732,
|
| 1114 |
+
"eval_loss": 1.22712274991214,
|
| 1115 |
+
"eval_mse_loss": 0.7849529917653837,
|
| 1116 |
+
"eval_runtime": 38.1855,
|
| 1117 |
+
"eval_samples_per_second": 733.079,
|
| 1118 |
+
"eval_steps_per_second": 11.47,
|
| 1119 |
+
"flow/cos_sim": 0.7789151420876316,
|
| 1120 |
+
"flow/improvement_ratio": 0.4868433005325326,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.7618113677251285,
|
| 1122 |
+
"flow/mag_ratio_std": 0.17511140633391464,
|
| 1123 |
+
"step": 21504
|
| 1124 |
+
}
|
| 1125 |
+
],
|
| 1126 |
+
"logging_steps": 256,
|
| 1127 |
+
"max_steps": 21651,
|
| 1128 |
+
"num_input_tokens_seen": 0,
|
| 1129 |
+
"num_train_epochs": 1,
|
| 1130 |
+
"save_steps": 1024,
|
| 1131 |
+
"stateful_callbacks": {
|
| 1132 |
+
"TrainerControl": {
|
| 1133 |
+
"args": {
|
| 1134 |
+
"should_epoch_stop": false,
|
| 1135 |
+
"should_evaluate": false,
|
| 1136 |
+
"should_log": false,
|
| 1137 |
+
"should_save": true,
|
| 1138 |
+
"should_training_stop": true
|
| 1139 |
+
},
|
| 1140 |
+
"attributes": {}
|
| 1141 |
+
}
|
| 1142 |
+
},
|
| 1143 |
+
"total_flos": 0.0,
|
| 1144 |
+
"train_batch_size": 64,
|
| 1145 |
+
"trial_name": null,
|
| 1146 |
+
"trial_params": null
|
| 1147 |
+
}
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-a/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03a8c5803c6e5663005c9dc7a4c98c213c8267649ccec42581475db390cc8017
|
| 3 |
+
size 5137
|