Upload folder using huggingface_hub
Browse files- checkpoints-v2.6-b-whitened/checkpoint-21651/ema.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/model.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/optimizer.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/rng_state.pth +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/scaler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/scheduler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/trainer_state.json +1147 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651/training_args.bin +3 -0
checkpoints-v2.6-b-whitened/checkpoint-21651/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:37343d7ae125c4672414806f020f132d4a9e4139492839d9b160f4ba48e588a5
|
| 3 |
+
size 550088
|
checkpoints-v2.6-b-whitened/checkpoint-21651/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb49b550e57bd7bd927330b42a26755e6e45f19110c163d0ca79010ea7c655a2
|
| 3 |
+
size 19331448
|
checkpoints-v2.6-b-whitened/checkpoint-21651/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae2217dc15bc22e5d6d3b3865f6d699a53bccd5efd50c84f3af1a0bb346b0d21
|
| 3 |
+
size 1175115
|
checkpoints-v2.6-b-whitened/checkpoint-21651/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/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5262330afa87a9beb2e187fc62a9e9163e6a9b0f556d0056631fbd5274bab19a
|
| 3 |
+
size 1383
|
checkpoints-v2.6-b-whitened/checkpoint-21651/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7471abe87daee5812e0143e98b883deb1b8470ee6335c3211aed51c21b3e9b2
|
| 3 |
+
size 1465
|
checkpoints-v2.6-b-whitened/checkpoint-21651/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": 0.22174665331840515,
|
| 15 |
+
"learning_rate": 0.000498046875,
|
| 16 |
+
"loss": 1.7768769264221191,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.18746864795684814,
|
| 22 |
+
"learning_rate": 0.000998046875,
|
| 23 |
+
"loss": 1.1502758264541626,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.08876891434192657,
|
| 29 |
+
"learning_rate": 0.000999640996023194,
|
| 30 |
+
"loss": 1.076206922531128,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.1368323266506195,
|
| 36 |
+
"learning_rate": 0.0009985588674043958,
|
| 37 |
+
"loss": 1.0463787317276,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.30891107049979033,
|
| 43 |
+
"eval_loss": 1.0069173309084487,
|
| 44 |
+
"eval_mse_loss": 1.0069173309084487,
|
| 45 |
+
"flow/cos_sim": 0.6910889456261239,
|
| 46 |
+
"flow/improvement_ratio": 0.4776086096894251,
|
| 47 |
+
"flow/mag_ratio_mean": 0.6849441453474298,
|
| 48 |
+
"flow/mag_ratio_std": 0.15538107210846797,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.30891107049979033,
|
| 54 |
+
"eval_loss": 1.0069173309084487,
|
| 55 |
+
"eval_mse_loss": 1.0069173309084487,
|
| 56 |
+
"eval_runtime": 37.995,
|
| 57 |
+
"eval_samples_per_second": 736.756,
|
| 58 |
+
"eval_steps_per_second": 11.528,
|
| 59 |
+
"flow/cos_sim": 0.6910889456261239,
|
| 60 |
+
"flow/improvement_ratio": 0.4776086096894251,
|
| 61 |
+
"flow/mag_ratio_mean": 0.6849441453474298,
|
| 62 |
+
"flow/mag_ratio_std": 0.15538107210846797,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.16789202392101288,
|
| 68 |
+
"learning_rate": 0.0009967551747861387,
|
| 69 |
+
"loss": 1.0251458883285522,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.18474483489990234,
|
| 75 |
+
"learning_rate": 0.000994232528651847,
|
| 76 |
+
"loss": 1.0035444498062134,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.2826478183269501,
|
| 82 |
+
"learning_rate": 0.0009909945800260092,
|
| 83 |
+
"loss": 0.9828508496284485,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.27789613604545593,
|
| 89 |
+
"learning_rate": 0.0009870460151900522,
|
| 90 |
+
"loss": 0.9670260548591614,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.2706769869820168,
|
| 96 |
+
"eval_loss": 0.9161460058057689,
|
| 97 |
+
"eval_mse_loss": 0.9161460058057689,
|
| 98 |
+
"flow/cos_sim": 0.7293230125076695,
|
| 99 |
+
"flow/improvement_ratio": 0.47896241032641773,
|
| 100 |
+
"flow/mag_ratio_mean": 0.7157350781301385,
|
| 101 |
+
"flow/mag_ratio_std": 0.16307487988580852,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.2706769869820168,
|
| 107 |
+
"eval_loss": 0.9161460058057689,
|
| 108 |
+
"eval_mse_loss": 0.9161460058057689,
|
| 109 |
+
"eval_runtime": 37.5228,
|
| 110 |
+
"eval_samples_per_second": 746.025,
|
| 111 |
+
"eval_steps_per_second": 11.673,
|
| 112 |
+
"flow/cos_sim": 0.7293230125076695,
|
| 113 |
+
"flow/improvement_ratio": 0.47896241032641773,
|
| 114 |
+
"flow/mag_ratio_mean": 0.7157350781301385,
|
| 115 |
+
"flow/mag_ratio_std": 0.16307487988580852,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.403795063495636,
|
| 121 |
+
"learning_rate": 0.0009823925488998885,
|
| 122 |
+
"loss": 0.9500002264976501,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.14074541628360748,
|
| 128 |
+
"learning_rate": 0.0009770409161149525,
|
| 129 |
+
"loss": 0.9403261542320251,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.2946110665798187,
|
| 135 |
+
"learning_rate": 0.0009709988622506973,
|
| 136 |
+
"loss": 0.9271752238273621,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.18222713470458984,
|
| 142 |
+
"learning_rate": 0.000964275131968659,
|
| 143 |
+
"loss": 0.9190171360969543,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.25661825895581614,
|
| 149 |
+
"eval_loss": 0.8770867594300884,
|
| 150 |
+
"eval_mse_loss": 0.8770867594300884,
|
| 151 |
+
"flow/cos_sim": 0.7433817604361059,
|
| 152 |
+
"flow/improvement_ratio": 0.4856133287491864,
|
| 153 |
+
"flow/mag_ratio_mean": 0.7246256326431553,
|
| 154 |
+
"flow/mag_ratio_std": 0.16845453622406475,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.25661825895581614,
|
| 160 |
+
"eval_loss": 0.8770867594300884,
|
| 161 |
+
"eval_mse_loss": 0.8770867594300884,
|
| 162 |
+
"eval_runtime": 37.6061,
|
| 163 |
+
"eval_samples_per_second": 744.374,
|
| 164 |
+
"eval_steps_per_second": 11.647,
|
| 165 |
+
"flow/cos_sim": 0.7433817604361059,
|
| 166 |
+
"flow/improvement_ratio": 0.4856133287491864,
|
| 167 |
+
"flow/mag_ratio_mean": 0.7246256326431553,
|
| 168 |
+
"flow/mag_ratio_std": 0.16845453622406475,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.7066600322723389,
|
| 174 |
+
"learning_rate": 0.0009568794565203123,
|
| 175 |
+
"loss": 0.9131532907485962,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.1416773498058319,
|
| 181 |
+
"learning_rate": 0.0009488225396630347,
|
| 182 |
+
"loss": 0.9086329936981201,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.5534916520118713,
|
| 188 |
+
"learning_rate": 0.0009401160421685646,
|
| 189 |
+
"loss": 0.8983782529830933,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 0.3552296459674835,
|
| 195 |
+
"learning_rate": 0.0009307725649463714,
|
| 196 |
+
"loss": 0.9001545906066895,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.24766322664202076,
|
| 202 |
+
"eval_loss": 0.8496752049280628,
|
| 203 |
+
"eval_mse_loss": 0.8496752049280628,
|
| 204 |
+
"flow/cos_sim": 0.7523367976489133,
|
| 205 |
+
"flow/improvement_ratio": 0.47609518638484555,
|
| 206 |
+
"flow/mag_ratio_mean": 0.7459522199957338,
|
| 207 |
+
"flow/mag_ratio_std": 0.17555569415087025,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.24766322664202076,
|
| 213 |
+
"eval_loss": 0.8496752049280628,
|
| 214 |
+
"eval_mse_loss": 0.8496752049280628,
|
| 215 |
+
"eval_runtime": 37.6011,
|
| 216 |
+
"eval_samples_per_second": 744.472,
|
| 217 |
+
"eval_steps_per_second": 11.649,
|
| 218 |
+
"flow/cos_sim": 0.7523367976489133,
|
| 219 |
+
"flow/improvement_ratio": 0.47609518638484555,
|
| 220 |
+
"flow/mag_ratio_mean": 0.7459522199957338,
|
| 221 |
+
"flow/mag_ratio_std": 0.17555569415087025,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.6213570237159729,
|
| 227 |
+
"learning_rate": 0.0009208056308063659,
|
| 228 |
+
"loss": 0.8938103914260864,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.14793087542057037,
|
| 234 |
+
"learning_rate": 0.0009102296648873445,
|
| 235 |
+
"loss": 0.8882678151130676,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.3995091915130615,
|
| 241 |
+
"learning_rate": 0.0008990599737794927,
|
| 242 |
+
"loss": 0.8866235017776489,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.24713630974292755,
|
| 248 |
+
"learning_rate": 0.0008873127233711644,
|
| 249 |
+
"loss": 0.8769825100898743,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.24238179573049284,
|
| 255 |
+
"eval_loss": 0.8358934846915067,
|
| 256 |
+
"eval_mse_loss": 0.8358934846915067,
|
| 257 |
+
"flow/cos_sim": 0.7576182183881873,
|
| 258 |
+
"flow/improvement_ratio": 0.47563578297286274,
|
| 259 |
+
"flow/mag_ratio_mean": 0.7517019887220914,
|
| 260 |
+
"flow/mag_ratio_std": 0.17310417188357,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.24238179573049284,
|
| 266 |
+
"eval_loss": 0.8358934846915067,
|
| 267 |
+
"eval_mse_loss": 0.8358934846915067,
|
| 268 |
+
"eval_runtime": 37.601,
|
| 269 |
+
"eval_samples_per_second": 744.476,
|
| 270 |
+
"eval_steps_per_second": 11.649,
|
| 271 |
+
"flow/cos_sim": 0.7576182183881873,
|
| 272 |
+
"flow/improvement_ratio": 0.47563578297286274,
|
| 273 |
+
"flow/mag_ratio_mean": 0.7517019887220914,
|
| 274 |
+
"flow/mag_ratio_std": 0.17310417188357,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.7627041339874268,
|
| 280 |
+
"learning_rate": 0.0008750049154520011,
|
| 281 |
+
"loss": 0.8750782012939453,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.9538213014602661,
|
| 287 |
+
"learning_rate": 0.0008621543631062487,
|
| 288 |
+
"loss": 0.8753933310508728,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.2025260329246521,
|
| 294 |
+
"learning_rate": 0.0008487796649318904,
|
| 295 |
+
"loss": 0.8711850047111511,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.21802696585655212,
|
| 301 |
+
"learning_rate": 0.0008349553511611836,
|
| 302 |
+
"loss": 0.8702585101127625,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.2416084907846908,
|
| 308 |
+
"eval_loss": 0.831794443740148,
|
| 309 |
+
"eval_mse_loss": 0.831794443740148,
|
| 310 |
+
"flow/cos_sim": 0.7583915245587423,
|
| 311 |
+
"flow/improvement_ratio": 0.4801776420035863,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7564767071101219,
|
| 313 |
+
"flow/mag_ratio_std": 0.18214766653977574,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.2416084907846908,
|
| 319 |
+
"eval_loss": 0.831794443740148,
|
| 320 |
+
"eval_mse_loss": 0.831794443740148,
|
| 321 |
+
"eval_runtime": 37.6215,
|
| 322 |
+
"eval_samples_per_second": 744.068,
|
| 323 |
+
"eval_steps_per_second": 11.642,
|
| 324 |
+
"flow/cos_sim": 0.7583915245587423,
|
| 325 |
+
"flow/improvement_ratio": 0.4801776420035863,
|
| 326 |
+
"flow/mag_ratio_mean": 0.7564767071101219,
|
| 327 |
+
"flow/mag_ratio_std": 0.18214766653977574,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.24702104926109314,
|
| 333 |
+
"learning_rate": 0.0008205930168562264,
|
| 334 |
+
"loss": 0.8623026609420776,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 1.5572556257247925,
|
| 340 |
+
"learning_rate": 0.0008057666884383055,
|
| 341 |
+
"loss": 0.8629586100578308,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.9861658215522766,
|
| 347 |
+
"learning_rate": 0.0007904978240451256,
|
| 348 |
+
"loss": 0.8615624904632568,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 1.0924731492996216,
|
| 354 |
+
"learning_rate": 0.0007748085222964866,
|
| 355 |
+
"loss": 0.8608123660087585,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.23790057833608427,
|
| 361 |
+
"eval_loss": 0.8215727942174973,
|
| 362 |
+
"eval_mse_loss": 0.8215727942174973,
|
| 363 |
+
"flow/cos_sim": 0.7620994245897145,
|
| 364 |
+
"flow/improvement_ratio": 0.4812065729675772,
|
| 365 |
+
"flow/mag_ratio_mean": 0.7490190734329834,
|
| 366 |
+
"flow/mag_ratio_std": 0.18066592055232558,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.23790057833608427,
|
| 372 |
+
"eval_loss": 0.8215727942174973,
|
| 373 |
+
"eval_mse_loss": 0.8215727942174973,
|
| 374 |
+
"eval_runtime": 37.6105,
|
| 375 |
+
"eval_samples_per_second": 744.286,
|
| 376 |
+
"eval_steps_per_second": 11.646,
|
| 377 |
+
"flow/cos_sim": 0.7620994245897145,
|
| 378 |
+
"flow/improvement_ratio": 0.4812065729675772,
|
| 379 |
+
"flow/mag_ratio_mean": 0.7490190734329834,
|
| 380 |
+
"flow/mag_ratio_std": 0.18066592055232558,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.22789114713668823,
|
| 386 |
+
"learning_rate": 0.0007587214903109633,
|
| 387 |
+
"loss": 0.8596727848052979,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 1.1323052644729614,
|
| 393 |
+
"learning_rate": 0.0007422600108419029,
|
| 394 |
+
"loss": 0.857304036617279,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.6124768853187561,
|
| 400 |
+
"learning_rate": 0.0007254479085803078,
|
| 401 |
+
"loss": 0.8568376898765564,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.21999727189540863,
|
| 407 |
+
"learning_rate": 0.0007083770653557752,
|
| 408 |
+
"loss": 0.8547202944755554,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.23593882117505488,
|
| 414 |
+
"eval_loss": 0.8155627318713219,
|
| 415 |
+
"eval_mse_loss": 0.8155627318713219,
|
| 416 |
+
"flow/cos_sim": 0.7640611918549559,
|
| 417 |
+
"flow/improvement_ratio": 0.479555664337389,
|
| 418 |
+
"flow/mag_ratio_mean": 0.7553564744180741,
|
| 419 |
+
"flow/mag_ratio_std": 0.17947759054022838,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.23593882117505488,
|
| 425 |
+
"eval_loss": 0.8155627318713219,
|
| 426 |
+
"eval_mse_loss": 0.8155627318713219,
|
| 427 |
+
"eval_runtime": 37.7057,
|
| 428 |
+
"eval_samples_per_second": 742.408,
|
| 429 |
+
"eval_steps_per_second": 11.616,
|
| 430 |
+
"flow/cos_sim": 0.7640611918549559,
|
| 431 |
+
"flow/improvement_ratio": 0.479555664337389,
|
| 432 |
+
"flow/mag_ratio_mean": 0.7553564744180741,
|
| 433 |
+
"flow/mag_ratio_std": 0.17947759054022838,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.24838890135288239,
|
| 439 |
+
"learning_rate": 0.0006909383150382365,
|
| 440 |
+
"loss": 0.8514482378959656,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 1.0127853155136108,
|
| 446 |
+
"learning_rate": 0.0006732232197928334,
|
| 447 |
+
"loss": 0.8534135818481445,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.329195111989975,
|
| 453 |
+
"learning_rate": 0.0006553280517779596,
|
| 454 |
+
"loss": 0.8500664830207825,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 1.0210610628128052,
|
| 460 |
+
"learning_rate": 0.000637138373398857,
|
| 461 |
+
"loss": 0.8482282757759094,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.23423708308505142,
|
| 467 |
+
"eval_loss": 0.8108278319715909,
|
| 468 |
+
"eval_mse_loss": 0.8108278319715909,
|
| 469 |
+
"flow/cos_sim": 0.7657629441997232,
|
| 470 |
+
"flow/improvement_ratio": 0.4783103846387776,
|
| 471 |
+
"flow/mag_ratio_mean": 0.7577394862153214,
|
| 472 |
+
"flow/mag_ratio_std": 0.1817886676282099,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.23423708308505142,
|
| 478 |
+
"eval_loss": 0.8108278319715909,
|
| 479 |
+
"eval_mse_loss": 0.8108278319715909,
|
| 480 |
+
"eval_runtime": 37.6611,
|
| 481 |
+
"eval_samples_per_second": 743.286,
|
| 482 |
+
"eval_steps_per_second": 11.63,
|
| 483 |
+
"flow/cos_sim": 0.7657629441997232,
|
| 484 |
+
"flow/improvement_ratio": 0.4783103846387776,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7577394862153214,
|
| 486 |
+
"flow/mag_ratio_std": 0.1817886676282099,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.23793469369411469,
|
| 492 |
+
"learning_rate": 0.0006187502147205189,
|
| 493 |
+
"loss": 0.8497642278671265,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.3985717296600342,
|
| 499 |
+
"learning_rate": 0.000600190188915804,
|
| 500 |
+
"loss": 0.8469876050949097,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.8007420301437378,
|
| 506 |
+
"learning_rate": 0.0005814851579007778,
|
| 507 |
+
"loss": 0.8458417057991028,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.679559588432312,
|
| 513 |
+
"learning_rate": 0.0005626621934574736,
|
| 514 |
+
"loss": 0.843183696269989,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.23854120157352865,
|
| 520 |
+
"eval_loss": 0.82157666675032,
|
| 521 |
+
"eval_mse_loss": 0.82157666675032,
|
| 522 |
+
"flow/cos_sim": 0.7614588316989271,
|
| 523 |
+
"flow/improvement_ratio": 0.4812564908369491,
|
| 524 |
+
"flow/mag_ratio_mean": 0.7500045074991983,
|
| 525 |
+
"flow/mag_ratio_std": 0.18137900323764375,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.23854120157352865,
|
| 531 |
+
"eval_loss": 0.82157666675032,
|
| 532 |
+
"eval_mse_loss": 0.82157666675032,
|
| 533 |
+
"eval_runtime": 37.7032,
|
| 534 |
+
"eval_samples_per_second": 742.457,
|
| 535 |
+
"eval_steps_per_second": 11.617,
|
| 536 |
+
"flow/cos_sim": 0.7614588316989271,
|
| 537 |
+
"flow/improvement_ratio": 0.4812564908369491,
|
| 538 |
+
"flow/mag_ratio_mean": 0.7500045074991983,
|
| 539 |
+
"flow/mag_ratio_std": 0.18137900323764375,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.3322647511959076,
|
| 545 |
+
"learning_rate": 0.0005437485380529165,
|
| 546 |
+
"loss": 0.8423036932945251,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 0.7933261394500732,
|
| 552 |
+
"learning_rate": 0.0005247715654111146,
|
| 553 |
+
"loss": 0.840004563331604,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.28959041833877563,
|
| 559 |
+
"learning_rate": 0.0005057587408950816,
|
| 560 |
+
"loss": 0.8403760194778442,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.2990068793296814,
|
| 566 |
+
"learning_rate": 0.00048673758175623157,
|
| 567 |
+
"loss": 0.8397681713104248,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.23300612986632133,
|
| 573 |
+
"eval_loss": 0.8080156621051161,
|
| 574 |
+
"eval_mse_loss": 0.8080156621051161,
|
| 575 |
+
"flow/cos_sim": 0.7669938810884136,
|
| 576 |
+
"flow/improvement_ratio": 0.48327972294260924,
|
| 577 |
+
"flow/mag_ratio_mean": 0.7643517768818494,
|
| 578 |
+
"flow/mag_ratio_std": 0.18017336787426308,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.23300612986632133,
|
| 584 |
+
"eval_loss": 0.8080156621051161,
|
| 585 |
+
"eval_mse_loss": 0.8080156621051161,
|
| 586 |
+
"eval_runtime": 37.5083,
|
| 587 |
+
"eval_samples_per_second": 746.315,
|
| 588 |
+
"eval_steps_per_second": 11.677,
|
| 589 |
+
"flow/cos_sim": 0.7669938810884136,
|
| 590 |
+
"flow/improvement_ratio": 0.48327972294260924,
|
| 591 |
+
"flow/mag_ratio_mean": 0.7643517768818494,
|
| 592 |
+
"flow/mag_ratio_std": 0.18017336787426308,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.32387539744377136,
|
| 598 |
+
"learning_rate": 0.0004678097707725879,
|
| 599 |
+
"loss": 0.8380001783370972,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.6504921317100525,
|
| 605 |
+
"learning_rate": 0.00044885426671328947,
|
| 606 |
+
"loss": 0.8419417142868042,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 0.23372496664524078,
|
| 612 |
+
"learning_rate": 0.0004299727858481981,
|
| 613 |
+
"loss": 0.840018093585968,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.26540830731391907,
|
| 619 |
+
"learning_rate": 0.00041119265533512737,
|
| 620 |
+
"loss": 0.8361135721206665,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.2324118216832479,
|
| 626 |
+
"eval_loss": 0.8050629013477395,
|
| 627 |
+
"eval_mse_loss": 0.8050629013477395,
|
| 628 |
+
"flow/cos_sim": 0.7675881983210507,
|
| 629 |
+
"flow/improvement_ratio": 0.4834722017860848,
|
| 630 |
+
"flow/mag_ratio_mean": 0.7585079777458487,
|
| 631 |
+
"flow/mag_ratio_std": 0.18149165839773335,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.2324118216832479,
|
| 637 |
+
"eval_loss": 0.8050629013477395,
|
| 638 |
+
"eval_mse_loss": 0.8050629013477395,
|
| 639 |
+
"eval_runtime": 37.5615,
|
| 640 |
+
"eval_samples_per_second": 745.258,
|
| 641 |
+
"eval_steps_per_second": 11.661,
|
| 642 |
+
"flow/cos_sim": 0.7675881983210507,
|
| 643 |
+
"flow/improvement_ratio": 0.4834722017860848,
|
| 644 |
+
"flow/mag_ratio_mean": 0.7585079777458487,
|
| 645 |
+
"flow/mag_ratio_std": 0.18149165839773335,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.2836131751537323,
|
| 651 |
+
"learning_rate": 0.0003925410556475763,
|
| 652 |
+
"loss": 0.8373591303825378,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.6104307174682617,
|
| 658 |
+
"learning_rate": 0.00037404498123644504,
|
| 659 |
+
"loss": 0.8390231132507324,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.1676436811685562,
|
| 665 |
+
"learning_rate": 0.0003557312014609784,
|
| 666 |
+
"loss": 0.8399800062179565,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.64168781042099,
|
| 672 |
+
"learning_rate": 0.00033762622184548686,
|
| 673 |
+
"loss": 0.8341987729072571,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.23086259510691307,
|
| 679 |
+
"eval_loss": 0.8013574742291072,
|
| 680 |
+
"eval_mse_loss": 0.8013574742291072,
|
| 681 |
+
"flow/cos_sim": 0.7691374256458456,
|
| 682 |
+
"flow/improvement_ratio": 0.4776277557083461,
|
| 683 |
+
"flow/mag_ratio_mean": 0.7624699862580321,
|
| 684 |
+
"flow/mag_ratio_std": 0.1821992572856276,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.23086259510691307,
|
| 690 |
+
"eval_loss": 0.8013574742291072,
|
| 691 |
+
"eval_mse_loss": 0.8013574742291072,
|
| 692 |
+
"eval_runtime": 37.5545,
|
| 693 |
+
"eval_samples_per_second": 745.398,
|
| 694 |
+
"eval_steps_per_second": 11.663,
|
| 695 |
+
"flow/cos_sim": 0.7691374256458456,
|
| 696 |
+
"flow/improvement_ratio": 0.4776277557083461,
|
| 697 |
+
"flow/mag_ratio_mean": 0.7624699862580321,
|
| 698 |
+
"flow/mag_ratio_std": 0.1821992572856276,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.5916406512260437,
|
| 704 |
+
"learning_rate": 0.00031975624571791317,
|
| 705 |
+
"loss": 0.8317926526069641,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.1671808809041977,
|
| 711 |
+
"learning_rate": 0.00030214713628576984,
|
| 712 |
+
"loss": 0.8340721726417542,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.30089712142944336,
|
| 718 |
+
"learning_rate": 0.000284891456531408,
|
| 719 |
+
"loss": 0.8345981240272522,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.5909346342086792,
|
| 725 |
+
"learning_rate": 0.00026787885831585745,
|
| 726 |
+
"loss": 0.8323626518249512,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.23025590746097913,
|
| 732 |
+
"eval_loss": 0.7985243684352805,
|
| 733 |
+
"eval_mse_loss": 0.7985243684352805,
|
| 734 |
+
"flow/cos_sim": 0.7697441032215885,
|
| 735 |
+
"flow/improvement_ratio": 0.47972988646868703,
|
| 736 |
+
"flow/mag_ratio_mean": 0.7682525152212953,
|
| 737 |
+
"flow/mag_ratio_std": 0.18057268694655537,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.23025590746097913,
|
| 743 |
+
"eval_loss": 0.7985243684352805,
|
| 744 |
+
"eval_mse_loss": 0.7985243684352805,
|
| 745 |
+
"eval_runtime": 37.4989,
|
| 746 |
+
"eval_samples_per_second": 746.502,
|
| 747 |
+
"eval_steps_per_second": 11.68,
|
| 748 |
+
"flow/cos_sim": 0.7697441032215885,
|
| 749 |
+
"flow/improvement_ratio": 0.47972988646868703,
|
| 750 |
+
"flow/mag_ratio_mean": 0.7682525152212953,
|
| 751 |
+
"flow/mag_ratio_std": 0.18057268694655537,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.6134583950042725,
|
| 757 |
+
"learning_rate": 0.0002512022089125798,
|
| 758 |
+
"loss": 0.8331765532493591,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 0.4425458312034607,
|
| 764 |
+
"learning_rate": 0.00023488564442791606,
|
| 765 |
+
"loss": 0.8346396088600159,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.6286107301712036,
|
| 771 |
+
"learning_rate": 0.0002189527798174848,
|
| 772 |
+
"loss": 0.8328061699867249,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 0.10555960983037949,
|
| 778 |
+
"learning_rate": 0.00020342667470826854,
|
| 779 |
+
"loss": 0.8316177725791931,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.23124641037150606,
|
| 785 |
+
"eval_loss": 0.8023071732695244,
|
| 786 |
+
"eval_mse_loss": 0.8023071732695244,
|
| 787 |
+
"flow/cos_sim": 0.7687535997667269,
|
| 788 |
+
"flow/improvement_ratio": 0.48764961331946666,
|
| 789 |
+
"flow/mag_ratio_mean": 0.7629323857560005,
|
| 790 |
+
"flow/mag_ratio_std": 0.1813373495250532,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.23124641037150606,
|
| 796 |
+
"eval_loss": 0.8023071732695244,
|
| 797 |
+
"eval_mse_loss": 0.8023071732695244,
|
| 798 |
+
"eval_runtime": 37.5306,
|
| 799 |
+
"eval_samples_per_second": 745.872,
|
| 800 |
+
"eval_steps_per_second": 11.67,
|
| 801 |
+
"flow/cos_sim": 0.7687535997667269,
|
| 802 |
+
"flow/improvement_ratio": 0.48764961331946666,
|
| 803 |
+
"flow/mag_ratio_mean": 0.7629323857560005,
|
| 804 |
+
"flow/mag_ratio_std": 0.1813373495250532,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.48595598340034485,
|
| 810 |
+
"learning_rate": 0.00018832980002442701,
|
| 811 |
+
"loss": 0.834376335144043,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.31518030166625977,
|
| 817 |
+
"learning_rate": 0.00017368400546514002,
|
| 818 |
+
"loss": 0.8339245915412903,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 0.1571056842803955,
|
| 824 |
+
"learning_rate": 0.00015951048788154866,
|
| 825 |
+
"loss": 0.829267144203186,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.30787667632102966,
|
| 831 |
+
"learning_rate": 0.0001458297605985633,
|
| 832 |
+
"loss": 0.8302126526832581,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.23091397094263877,
|
| 838 |
+
"eval_loss": 0.801251276715161,
|
| 839 |
+
"eval_mse_loss": 0.801251276715161,
|
| 840 |
+
"flow/cos_sim": 0.7690860558318221,
|
| 841 |
+
"flow/improvement_ratio": 0.48135468649537594,
|
| 842 |
+
"flow/mag_ratio_mean": 0.7656754217191374,
|
| 843 |
+
"flow/mag_ratio_std": 0.18127168340770078,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.23091397094263877,
|
| 849 |
+
"eval_loss": 0.801251276715161,
|
| 850 |
+
"eval_mse_loss": 0.801251276715161,
|
| 851 |
+
"eval_runtime": 37.531,
|
| 852 |
+
"eval_samples_per_second": 745.863,
|
| 853 |
+
"eval_steps_per_second": 11.67,
|
| 854 |
+
"flow/cos_sim": 0.7690860558318221,
|
| 855 |
+
"flow/improvement_ratio": 0.48135468649537594,
|
| 856 |
+
"flow/mag_ratio_mean": 0.7656754217191374,
|
| 857 |
+
"flow/mag_ratio_std": 0.18127168340770078,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.1216726079583168,
|
| 863 |
+
"learning_rate": 0.00013266162372593872,
|
| 864 |
+
"loss": 0.8286399841308594,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.18312956392765045,
|
| 870 |
+
"learning_rate": 0.00012002513550158511,
|
| 871 |
+
"loss": 0.8293663263320923,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.49640128016471863,
|
| 877 |
+
"learning_rate": 0.00010798470498931034,
|
| 878 |
+
"loss": 0.8313955664634705,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.31345558166503906,
|
| 884 |
+
"learning_rate": 9.646333498481152e-05,
|
| 885 |
+
"loss": 0.830700695514679,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.23020775735242183,
|
| 891 |
+
"eval_loss": 0.7999282186706317,
|
| 892 |
+
"eval_mse_loss": 0.7999282186706317,
|
| 893 |
+
"flow/cos_sim": 0.7697922684014116,
|
| 894 |
+
"flow/improvement_ratio": 0.4801355227487817,
|
| 895 |
+
"flow/mag_ratio_mean": 0.7621459820771326,
|
| 896 |
+
"flow/mag_ratio_std": 0.18215228709047787,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.23020775735242183,
|
| 902 |
+
"eval_loss": 0.7999282186706317,
|
| 903 |
+
"eval_mse_loss": 0.7999282186706317,
|
| 904 |
+
"eval_runtime": 37.4254,
|
| 905 |
+
"eval_samples_per_second": 747.967,
|
| 906 |
+
"eval_steps_per_second": 11.703,
|
| 907 |
+
"flow/cos_sim": 0.7697922684014116,
|
| 908 |
+
"flow/improvement_ratio": 0.4801355227487817,
|
| 909 |
+
"flow/mag_ratio_mean": 0.7621459820771326,
|
| 910 |
+
"flow/mag_ratio_std": 0.18215228709047787,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.28336694836616516,
|
| 916 |
+
"learning_rate": 8.552600339377569e-05,
|
| 917 |
+
"loss": 0.8312592506408691,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.40511423349380493,
|
| 923 |
+
"learning_rate": 7.518853981074431e-05,
|
| 924 |
+
"loss": 0.8279933333396912,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.18280020356178284,
|
| 930 |
+
"learning_rate": 6.546590564159156e-05,
|
| 931 |
+
"loss": 0.8287937045097351,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.3500230014324188,
|
| 937 |
+
"learning_rate": 5.637217244988779e-05,
|
| 938 |
+
"loss": 0.8283677101135254,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.23032897584922782,
|
| 944 |
+
"eval_loss": 0.799097118040198,
|
| 945 |
+
"eval_mse_loss": 0.799097118040198,
|
| 946 |
+
"flow/cos_sim": 0.7696710277910102,
|
| 947 |
+
"flow/improvement_ratio": 0.47835244762298723,
|
| 948 |
+
"flow/mag_ratio_mean": 0.7640095663397279,
|
| 949 |
+
"flow/mag_ratio_std": 0.18172353131858182,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.23032897584922782,
|
| 955 |
+
"eval_loss": 0.799097118040198,
|
| 956 |
+
"eval_mse_loss": 0.799097118040198,
|
| 957 |
+
"eval_runtime": 38.9231,
|
| 958 |
+
"eval_samples_per_second": 719.188,
|
| 959 |
+
"eval_steps_per_second": 11.253,
|
| 960 |
+
"flow/cos_sim": 0.7696710277910102,
|
| 961 |
+
"flow/improvement_ratio": 0.47835244762298723,
|
| 962 |
+
"flow/mag_ratio_mean": 0.7640095663397279,
|
| 963 |
+
"flow/mag_ratio_std": 0.18172353131858182,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.2998882234096527,
|
| 969 |
+
"learning_rate": 4.792050159113337e-05,
|
| 970 |
+
"loss": 0.8285696506500244,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.1577506959438324,
|
| 976 |
+
"learning_rate": 4.012312516433581e-05,
|
| 977 |
+
"loss": 0.824936032295227,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.1853588968515396,
|
| 983 |
+
"learning_rate": 3.2991328308497545e-05,
|
| 984 |
+
"loss": 0.8292096257209778,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 0.5599551796913147,
|
| 990 |
+
"learning_rate": 2.653543286964183e-05,
|
| 991 |
+
"loss": 0.831167995929718,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.2312095735103028,
|
| 997 |
+
"eval_loss": 0.8017649225992699,
|
| 998 |
+
"eval_mse_loss": 0.8017649225992699,
|
| 999 |
+
"flow/cos_sim": 0.7687904472764768,
|
| 1000 |
+
"flow/improvement_ratio": 0.48452607518461743,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.7627046746478233,
|
| 1002 |
+
"flow/mag_ratio_std": 0.18138871069776413,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.2312095735103028,
|
| 1008 |
+
"eval_loss": 0.8017649225992699,
|
| 1009 |
+
"eval_mse_loss": 0.8017649225992699,
|
| 1010 |
+
"eval_runtime": 38.1085,
|
| 1011 |
+
"eval_samples_per_second": 734.561,
|
| 1012 |
+
"eval_steps_per_second": 11.494,
|
| 1013 |
+
"flow/cos_sim": 0.7687904472764768,
|
| 1014 |
+
"flow/improvement_ratio": 0.48452607518461743,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.7627046746478233,
|
| 1016 |
+
"flow/mag_ratio_std": 0.18138871069776413,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.14706018567085266,
|
| 1022 |
+
"learning_rate": 2.076478246200819e-05,
|
| 1023 |
+
"loss": 0.8276491165161133,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.35119906067848206,
|
| 1029 |
+
"learning_rate": 1.5687728945045944e-05,
|
| 1030 |
+
"loss": 0.8274345993995667,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.12262684106826782,
|
| 1036 |
+
"learning_rate": 1.1311620335770879e-05,
|
| 1037 |
+
"loss": 0.8314203023910522,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.22295571863651276,
|
| 1043 |
+
"learning_rate": 7.642790173984836e-06,
|
| 1044 |
+
"loss": 0.8324977159500122,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.22896028242018668,
|
| 1050 |
+
"eval_loss": 0.7961706479662629,
|
| 1051 |
+
"eval_mse_loss": 0.7961706479662629,
|
| 1052 |
+
"flow/cos_sim": 0.7710397344201667,
|
| 1053 |
+
"flow/improvement_ratio": 0.47750673419264356,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.7662278138883581,
|
| 1055 |
+
"flow/mag_ratio_std": 0.18190687843789793,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.22896028242018668,
|
| 1061 |
+
"eval_loss": 0.7961706479662629,
|
| 1062 |
+
"eval_mse_loss": 0.7961706479662629,
|
| 1063 |
+
"eval_runtime": 38.0955,
|
| 1064 |
+
"eval_samples_per_second": 734.812,
|
| 1065 |
+
"eval_steps_per_second": 11.497,
|
| 1066 |
+
"flow/cos_sim": 0.7710397344201667,
|
| 1067 |
+
"flow/improvement_ratio": 0.47750673419264356,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.7662278138883581,
|
| 1069 |
+
"flow/mag_ratio_std": 0.18190687843789793,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.8932636976242065,
|
| 1075 |
+
"learning_rate": 4.686548355746001e-06,
|
| 1076 |
+
"loss": 0.8278123140335083,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.23970206081867218,
|
| 1082 |
+
"learning_rate": 2.447173448359541e-06,
|
| 1083 |
+
"loss": 0.828546404838562,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.1667780578136444,
|
| 1089 |
+
"learning_rate": 9.279064980089457e-07,
|
| 1090 |
+
"loss": 0.8292500376701355,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.12882362306118011,
|
| 1096 |
+
"learning_rate": 1.3094633899163322e-07,
|
| 1097 |
+
"loss": 0.8290307521820068,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.22907172470060114,
|
| 1103 |
+
"eval_loss": 0.7964604407140653,
|
| 1104 |
+
"eval_mse_loss": 0.7964604407140653,
|
| 1105 |
+
"flow/cos_sim": 0.7709282877510542,
|
| 1106 |
+
"flow/improvement_ratio": 0.4860201631778996,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.7650731019505627,
|
| 1108 |
+
"flow/mag_ratio_std": 0.1804714862886629,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.22907172470060114,
|
| 1114 |
+
"eval_loss": 0.7964604407140653,
|
| 1115 |
+
"eval_mse_loss": 0.7964604407140653,
|
| 1116 |
+
"eval_runtime": 37.8942,
|
| 1117 |
+
"eval_samples_per_second": 738.715,
|
| 1118 |
+
"eval_steps_per_second": 11.558,
|
| 1119 |
+
"flow/cos_sim": 0.7709282877510542,
|
| 1120 |
+
"flow/improvement_ratio": 0.4860201631778996,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.7650731019505627,
|
| 1122 |
+
"flow/mag_ratio_std": 0.1804714862886629,
|
| 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/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03a8c5803c6e5663005c9dc7a4c98c213c8267649ccec42581475db390cc8017
|
| 3 |
+
size 5137
|