Upload folder using huggingface_hub
Browse files- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/ema.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/model.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/optimizer.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/rng_state.pth +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/scaler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/scheduler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/trainer_state.json +1147 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/training_args.bin +3 -0
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2cb0dc805812a87f88738c1dcd844ccdec9e9d6f73164ec8d3aedcd87d998ada
|
| 3 |
+
size 550088
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e69e746c59f4e9bffd46a4b8fff59f5298117cc651aec90edc417c5a9eb3032
|
| 3 |
+
size 19331448
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d395fc7dd5e395ed1e754e2cf35239eadea0fa97a95d1668188b63fc998510f
|
| 3 |
+
size 1175115
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/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-c/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:68b7d31923f05c4b460794137c5c6eea5cd3094990a5ca753208ffd9f46866cb
|
| 3 |
+
size 1383
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fa0141c127e95bcc4d5966d55ec878adc1193cbedb02b1419969134a8445c716
|
| 3 |
+
size 1465
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-c/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": 0.509731113910675,
|
| 15 |
+
"learning_rate": 0.00049609375,
|
| 16 |
+
"loss": 2.1022446155548096,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.32036706805229187,
|
| 22 |
+
"learning_rate": 0.00099609375,
|
| 23 |
+
"loss": 1.3224166631698608,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.46608567237854004,
|
| 29 |
+
"learning_rate": 0.0009996438058859384,
|
| 30 |
+
"loss": 1.2243598699569702,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.3062893748283386,
|
| 36 |
+
"learning_rate": 0.0009985644996281977,
|
| 37 |
+
"loss": 1.1784849166870117,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.29423016282521425,
|
| 43 |
+
"eval_loss": 1.1209385933941358,
|
| 44 |
+
"eval_mse_loss": 0.9738235149753692,
|
| 45 |
+
"flow/cos_sim": 0.7057698449315546,
|
| 46 |
+
"flow/improvement_ratio": 0.47853941589457805,
|
| 47 |
+
"flow/mag_ratio_mean": 0.6953887133837835,
|
| 48 |
+
"flow/mag_ratio_std": 0.15273392692072207,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.29423016282521425,
|
| 54 |
+
"eval_loss": 1.1209385933941358,
|
| 55 |
+
"eval_mse_loss": 0.9738235149753692,
|
| 56 |
+
"eval_runtime": 37.7105,
|
| 57 |
+
"eval_samples_per_second": 742.314,
|
| 58 |
+
"eval_steps_per_second": 11.615,
|
| 59 |
+
"flow/cos_sim": 0.7057698449315546,
|
| 60 |
+
"flow/improvement_ratio": 0.47853941589457805,
|
| 61 |
+
"flow/mag_ratio_mean": 0.6953887133837835,
|
| 62 |
+
"flow/mag_ratio_std": 0.15273392692072207,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.42174944281578064,
|
| 68 |
+
"learning_rate": 0.0009967636212194833,
|
| 69 |
+
"loss": 1.1545671224594116,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.5216838717460632,
|
| 75 |
+
"learning_rate": 0.0009942437770702158,
|
| 76 |
+
"loss": 1.1316872835159302,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.45238178968429565,
|
| 82 |
+
"learning_rate": 0.000991008614149572,
|
| 83 |
+
"loss": 1.1131517887115479,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.633682906627655,
|
| 89 |
+
"learning_rate": 0.0009870628147072287,
|
| 90 |
+
"loss": 1.1000479459762573,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.26900983356857955,
|
| 96 |
+
"eval_loss": 1.0475259179666163,
|
| 97 |
+
"eval_mse_loss": 0.9130210040911148,
|
| 98 |
+
"flow/cos_sim": 0.7309901614983877,
|
| 99 |
+
"flow/improvement_ratio": 0.4792113368777924,
|
| 100 |
+
"flow/mag_ratio_mean": 0.7129153462305461,
|
| 101 |
+
"flow/mag_ratio_std": 0.15875548022250607,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.26900983356857955,
|
| 107 |
+
"eval_loss": 1.0475259179666163,
|
| 108 |
+
"eval_mse_loss": 0.9130210040911148,
|
| 109 |
+
"eval_runtime": 37.4057,
|
| 110 |
+
"eval_samples_per_second": 748.361,
|
| 111 |
+
"eval_steps_per_second": 11.709,
|
| 112 |
+
"flow/cos_sim": 0.7309901614983877,
|
| 113 |
+
"flow/improvement_ratio": 0.4792113368777924,
|
| 114 |
+
"flow/mag_ratio_mean": 0.7129153462305461,
|
| 115 |
+
"flow/mag_ratio_std": 0.15875548022250607,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.24040701985359192,
|
| 121 |
+
"learning_rate": 0.0009824120894967462,
|
| 122 |
+
"loss": 1.0860177278518677,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 0.3865581452846527,
|
| 128 |
+
"learning_rate": 0.0009770631695103949,
|
| 129 |
+
"loss": 1.0790908336639404,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.264023095369339,
|
| 135 |
+
"learning_rate": 0.0009710237962373965,
|
| 136 |
+
"loss": 1.0654981136322021,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 0.6001664996147156,
|
| 142 |
+
"learning_rate": 0.0009643027104596686,
|
| 143 |
+
"loss": 1.0555968284606934,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.2544627361101647,
|
| 149 |
+
"eval_loss": 1.0013415866518673,
|
| 150 |
+
"eval_mse_loss": 0.8741102180524504,
|
| 151 |
+
"flow/cos_sim": 0.7455372803559587,
|
| 152 |
+
"flow/improvement_ratio": 0.4856065721952752,
|
| 153 |
+
"flow/mag_ratio_mean": 0.7307753671794177,
|
| 154 |
+
"flow/mag_ratio_std": 0.16138841017876585,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.2544627361101647,
|
| 160 |
+
"eval_loss": 1.0013415866518673,
|
| 161 |
+
"eval_mse_loss": 0.8741102180524504,
|
| 162 |
+
"eval_runtime": 37.3598,
|
| 163 |
+
"eval_samples_per_second": 749.281,
|
| 164 |
+
"eval_steps_per_second": 11.724,
|
| 165 |
+
"flow/cos_sim": 0.7455372803559587,
|
| 166 |
+
"flow/improvement_ratio": 0.4856065721952752,
|
| 167 |
+
"flow/mag_ratio_mean": 0.7307753671794177,
|
| 168 |
+
"flow/mag_ratio_std": 0.16138841017876585,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.2462933361530304,
|
| 174 |
+
"learning_rate": 0.000956909639601296,
|
| 175 |
+
"loss": 1.0477466583251953,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.7574740648269653,
|
| 181 |
+
"learning_rate": 0.0009488552836500346,
|
| 182 |
+
"loss": 1.0434833765029907,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.545494794845581,
|
| 188 |
+
"learning_rate": 0.0009401512996712244,
|
| 189 |
+
"loss": 1.0306124687194824,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 1.4480839967727661,
|
| 195 |
+
"learning_rate": 0.0009308102849365255,
|
| 196 |
+
"loss": 1.0329992771148682,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.24506935215295722,
|
| 202 |
+
"eval_loss": 0.9690635023323912,
|
| 203 |
+
"eval_mse_loss": 0.8465288246058982,
|
| 204 |
+
"flow/cos_sim": 0.7549306665925675,
|
| 205 |
+
"flow/improvement_ratio": 0.47613726440630005,
|
| 206 |
+
"flow/mag_ratio_mean": 0.7428392282087509,
|
| 207 |
+
"flow/mag_ratio_std": 0.16521768938733017,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.24506935215295722,
|
| 213 |
+
"eval_loss": 0.9690635023323912,
|
| 214 |
+
"eval_mse_loss": 0.8465288246058982,
|
| 215 |
+
"eval_runtime": 37.7154,
|
| 216 |
+
"eval_samples_per_second": 742.217,
|
| 217 |
+
"eval_steps_per_second": 11.613,
|
| 218 |
+
"flow/cos_sim": 0.7549306665925675,
|
| 219 |
+
"flow/improvement_ratio": 0.47613726440630005,
|
| 220 |
+
"flow/mag_ratio_mean": 0.7428392282087509,
|
| 221 |
+
"flow/mag_ratio_std": 0.16521768938733017,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 1.0813051462173462,
|
| 227 |
+
"learning_rate": 0.0009208457586918912,
|
| 228 |
+
"loss": 1.0259068012237549,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.4623449146747589,
|
| 234 |
+
"learning_rate": 0.0009102721425911724,
|
| 235 |
+
"loss": 1.019983172416687,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.17263032495975494,
|
| 241 |
+
"learning_rate": 0.0008991047398236636,
|
| 242 |
+
"loss": 1.017630934715271,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.8617929816246033,
|
| 248 |
+
"learning_rate": 0.0008873597129658053,
|
| 249 |
+
"loss": 1.0077481269836426,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.23889310307562622,
|
| 255 |
+
"eval_loss": 0.9500942819466874,
|
| 256 |
+
"eval_mse_loss": 0.8306477298475292,
|
| 257 |
+
"flow/cos_sim": 0.7611069038306197,
|
| 258 |
+
"flow/improvement_ratio": 0.47608428826070814,
|
| 259 |
+
"flow/mag_ratio_mean": 0.7442350685596466,
|
| 260 |
+
"flow/mag_ratio_std": 0.16566321992030428,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.23889310307562622,
|
| 266 |
+
"eval_loss": 0.9500942819466874,
|
| 267 |
+
"eval_mse_loss": 0.8306477298475292,
|
| 268 |
+
"eval_runtime": 38.35,
|
| 269 |
+
"eval_samples_per_second": 729.934,
|
| 270 |
+
"eval_steps_per_second": 11.421,
|
| 271 |
+
"flow/cos_sim": 0.7611069038306197,
|
| 272 |
+
"flow/improvement_ratio": 0.47608428826070814,
|
| 273 |
+
"flow/mag_ratio_mean": 0.7442350685596466,
|
| 274 |
+
"flow/mag_ratio_std": 0.16566321992030428,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 1.1165090799331665,
|
| 280 |
+
"learning_rate": 0.0008750540605890948,
|
| 281 |
+
"loss": 1.0057861804962158,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.3366983234882355,
|
| 287 |
+
"learning_rate": 0.0008622055926580629,
|
| 288 |
+
"loss": 1.0057446956634521,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.18536005914211273,
|
| 294 |
+
"learning_rate": 0.0008488329047539208,
|
| 295 |
+
"loss": 1.0015759468078613,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.2887427508831024,
|
| 301 |
+
"learning_rate": 0.0008349553511611836,
|
| 302 |
+
"loss": 1.0002145767211914,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.23733438571838483,
|
| 308 |
+
"eval_loss": 0.9438223587051374,
|
| 309 |
+
"eval_mse_loss": 0.8251551663222378,
|
| 310 |
+
"flow/cos_sim": 0.7626656395931767,
|
| 311 |
+
"flow/improvement_ratio": 0.48116366265843447,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7489744677935561,
|
| 313 |
+
"flow/mag_ratio_std": 0.17211752354282223,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.23733438571838483,
|
| 319 |
+
"eval_loss": 0.9438223587051374,
|
| 320 |
+
"eval_mse_loss": 0.8251551663222378,
|
| 321 |
+
"eval_runtime": 37.8333,
|
| 322 |
+
"eval_samples_per_second": 739.903,
|
| 323 |
+
"eval_steps_per_second": 11.577,
|
| 324 |
+
"flow/cos_sim": 0.7626656395931767,
|
| 325 |
+
"flow/improvement_ratio": 0.48116366265843447,
|
| 326 |
+
"flow/mag_ratio_mean": 0.7489744677935561,
|
| 327 |
+
"flow/mag_ratio_std": 0.17211752354282223,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.3892139196395874,
|
| 333 |
+
"learning_rate": 0.0008205930168562264,
|
| 334 |
+
"loss": 0.9933722019195557,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.5060065388679504,
|
| 340 |
+
"learning_rate": 0.0008057666884383055,
|
| 341 |
+
"loss": 0.9938154220581055,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.5501173734664917,
|
| 347 |
+
"learning_rate": 0.0007904978240451256,
|
| 348 |
+
"loss": 0.9929619431495667,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.531117856502533,
|
| 354 |
+
"learning_rate": 0.000774870597388272,
|
| 355 |
+
"loss": 0.992885410785675,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.2344179714447287,
|
| 361 |
+
"eval_loss": 0.9344740850196037,
|
| 362 |
+
"eval_mse_loss": 0.8172650986338315,
|
| 363 |
+
"flow/cos_sim": 0.7655820315831328,
|
| 364 |
+
"flow/improvement_ratio": 0.4813905014703263,
|
| 365 |
+
"flow/mag_ratio_mean": 0.749662634988898,
|
| 366 |
+
"flow/mag_ratio_std": 0.17085277949157915,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.2344179714447287,
|
| 372 |
+
"eval_loss": 0.9344740850196037,
|
| 373 |
+
"eval_mse_loss": 0.8172650986338315,
|
| 374 |
+
"eval_runtime": 37.5903,
|
| 375 |
+
"eval_samples_per_second": 744.687,
|
| 376 |
+
"eval_steps_per_second": 11.652,
|
| 377 |
+
"flow/cos_sim": 0.7655820315831328,
|
| 378 |
+
"flow/improvement_ratio": 0.4813905014703263,
|
| 379 |
+
"flow/mag_ratio_mean": 0.749662634988898,
|
| 380 |
+
"flow/mag_ratio_std": 0.17085277949157915,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.1288619190454483,
|
| 386 |
+
"learning_rate": 0.0007587850741039065,
|
| 387 |
+
"loss": 0.9900345802307129,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.5061522722244263,
|
| 393 |
+
"learning_rate": 0.000742325011311212,
|
| 394 |
+
"loss": 0.9876418709754944,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.3961803615093231,
|
| 400 |
+
"learning_rate": 0.0007255142316508366,
|
| 401 |
+
"loss": 0.9861115217208862,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 1.0998958349227905,
|
| 407 |
+
"learning_rate": 0.0007083770653557752,
|
| 408 |
+
"loss": 0.9850316047668457,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.2321178341565067,
|
| 414 |
+
"eval_loss": 0.9271248780429091,
|
| 415 |
+
"eval_mse_loss": 0.8110659605564048,
|
| 416 |
+
"flow/cos_sim": 0.7678821885694652,
|
| 417 |
+
"flow/improvement_ratio": 0.47983810108274083,
|
| 418 |
+
"flow/mag_ratio_mean": 0.7483109420837333,
|
| 419 |
+
"flow/mag_ratio_std": 0.17305424319554683,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.2321178341565067,
|
| 425 |
+
"eval_loss": 0.9271248780429091,
|
| 426 |
+
"eval_mse_loss": 0.8110659605564048,
|
| 427 |
+
"eval_runtime": 37.6489,
|
| 428 |
+
"eval_samples_per_second": 743.528,
|
| 429 |
+
"eval_steps_per_second": 11.634,
|
| 430 |
+
"flow/cos_sim": 0.7678821885694652,
|
| 431 |
+
"flow/improvement_ratio": 0.47983810108274083,
|
| 432 |
+
"flow/mag_ratio_mean": 0.7483109420837333,
|
| 433 |
+
"flow/mag_ratio_std": 0.17305424319554683,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.9258986115455627,
|
| 439 |
+
"learning_rate": 0.0006909383150382365,
|
| 440 |
+
"loss": 0.9811054468154907,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 1.5331991910934448,
|
| 446 |
+
"learning_rate": 0.0006732232197928334,
|
| 447 |
+
"loss": 0.9832378625869751,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.13353095948696136,
|
| 453 |
+
"learning_rate": 0.0006552574186680553,
|
| 454 |
+
"loss": 0.9800243377685547,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 0.8382576704025269,
|
| 460 |
+
"learning_rate": 0.0006370669135588852,
|
| 461 |
+
"loss": 0.9766659140586853,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.23024527721753404,
|
| 467 |
+
"eval_loss": 0.9205626799091357,
|
| 468 |
+
"eval_mse_loss": 0.8054400410282013,
|
| 469 |
+
"flow/cos_sim": 0.7697547297227328,
|
| 470 |
+
"flow/improvement_ratio": 0.4783771674926967,
|
| 471 |
+
"flow/mag_ratio_mean": 0.7531407558482531,
|
| 472 |
+
"flow/mag_ratio_std": 0.17515080694330337,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.23024527721753404,
|
| 478 |
+
"eval_loss": 0.9205626799091357,
|
| 479 |
+
"eval_mse_loss": 0.8054400410282013,
|
| 480 |
+
"eval_runtime": 37.56,
|
| 481 |
+
"eval_samples_per_second": 745.287,
|
| 482 |
+
"eval_steps_per_second": 11.661,
|
| 483 |
+
"flow/cos_sim": 0.7697547297227328,
|
| 484 |
+
"flow/improvement_ratio": 0.4783771674926967,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7531407558482531,
|
| 486 |
+
"flow/mag_ratio_std": 0.17515080694330337,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.17838020622730255,
|
| 492 |
+
"learning_rate": 0.0006187502147205189,
|
| 493 |
+
"loss": 0.9791604280471802,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.24214127659797668,
|
| 499 |
+
"learning_rate": 0.000600190188915804,
|
| 500 |
+
"loss": 0.9759600162506104,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.5306867361068726,
|
| 506 |
+
"learning_rate": 0.0005814851579007778,
|
| 507 |
+
"loss": 0.9733525514602661,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.8362441062927246,
|
| 513 |
+
"learning_rate": 0.0005626621934574736,
|
| 514 |
+
"loss": 0.9725903272628784,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.22953055931688987,
|
| 520 |
+
"eval_loss": 0.9186442800580639,
|
| 521 |
+
"eval_mse_loss": 0.8038789996001274,
|
| 522 |
+
"flow/cos_sim": 0.7704694516038242,
|
| 523 |
+
"flow/improvement_ratio": 0.48191878424115375,
|
| 524 |
+
"flow/mag_ratio_mean": 0.7543752857539208,
|
| 525 |
+
"flow/mag_ratio_std": 0.1764680508132939,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.22953055931688987,
|
| 531 |
+
"eval_loss": 0.9186442800580639,
|
| 532 |
+
"eval_mse_loss": 0.8038789996001274,
|
| 533 |
+
"eval_runtime": 37.383,
|
| 534 |
+
"eval_samples_per_second": 748.816,
|
| 535 |
+
"eval_steps_per_second": 11.717,
|
| 536 |
+
"flow/cos_sim": 0.7704694516038242,
|
| 537 |
+
"flow/improvement_ratio": 0.48191878424115375,
|
| 538 |
+
"flow/mag_ratio_mean": 0.7543752857539208,
|
| 539 |
+
"flow/mag_ratio_std": 0.1764680508132939,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.5765674710273743,
|
| 545 |
+
"learning_rate": 0.0005437485380529165,
|
| 546 |
+
"loss": 0.9690040946006775,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 1.1299241781234741,
|
| 552 |
+
"learning_rate": 0.0005247715654111146,
|
| 553 |
+
"loss": 0.9676824808120728,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.9060980081558228,
|
| 559 |
+
"learning_rate": 0.0005057587408950816,
|
| 560 |
+
"loss": 0.9672520160675049,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 0.1904771476984024,
|
| 566 |
+
"learning_rate": 0.00048673758175623157,
|
| 567 |
+
"loss": 0.9666534662246704,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.2270588960889812,
|
| 573 |
+
"eval_loss": 0.9110365677369784,
|
| 574 |
+
"eval_mse_loss": 0.7975071200497075,
|
| 575 |
+
"flow/cos_sim": 0.7729411255823423,
|
| 576 |
+
"flow/improvement_ratio": 0.4843705101916779,
|
| 577 |
+
"flow/mag_ratio_mean": 0.7594509196880201,
|
| 578 |
+
"flow/mag_ratio_std": 0.17728658469572459,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.2270588960889812,
|
| 584 |
+
"eval_loss": 0.9110365677369784,
|
| 585 |
+
"eval_mse_loss": 0.7975071200497075,
|
| 586 |
+
"eval_runtime": 37.7887,
|
| 587 |
+
"eval_samples_per_second": 740.776,
|
| 588 |
+
"eval_steps_per_second": 11.591,
|
| 589 |
+
"flow/cos_sim": 0.7729411255823423,
|
| 590 |
+
"flow/improvement_ratio": 0.4843705101916779,
|
| 591 |
+
"flow/mag_ratio_mean": 0.7594509196880201,
|
| 592 |
+
"flow/mag_ratio_std": 0.17728658469572459,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 1.2817021608352661,
|
| 598 |
+
"learning_rate": 0.0004678097707725879,
|
| 599 |
+
"loss": 0.9641943573951721,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.21733422577381134,
|
| 605 |
+
"learning_rate": 0.00044885426671328947,
|
| 606 |
+
"loss": 0.9682671427726746,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 1.7028273344039917,
|
| 612 |
+
"learning_rate": 0.0004299727858481981,
|
| 613 |
+
"loss": 0.9672344326972961,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 0.37147364020347595,
|
| 619 |
+
"learning_rate": 0.00041119265533512737,
|
| 620 |
+
"loss": 0.9616128206253052,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.22645196407080787,
|
| 626 |
+
"eval_loss": 0.9077021104015716,
|
| 627 |
+
"eval_mse_loss": 0.7944761288764814,
|
| 628 |
+
"flow/cos_sim": 0.7735480565458672,
|
| 629 |
+
"flow/improvement_ratio": 0.48385851391374246,
|
| 630 |
+
"flow/mag_ratio_mean": 0.7533418638248967,
|
| 631 |
+
"flow/mag_ratio_std": 0.17509023045704245,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.22645196407080787,
|
| 637 |
+
"eval_loss": 0.9077021104015716,
|
| 638 |
+
"eval_mse_loss": 0.7944761288764814,
|
| 639 |
+
"eval_runtime": 37.7605,
|
| 640 |
+
"eval_samples_per_second": 741.331,
|
| 641 |
+
"eval_steps_per_second": 11.599,
|
| 642 |
+
"flow/cos_sim": 0.7735480565458672,
|
| 643 |
+
"flow/improvement_ratio": 0.48385851391374246,
|
| 644 |
+
"flow/mag_ratio_mean": 0.7533418638248967,
|
| 645 |
+
"flow/mag_ratio_std": 0.17509023045704245,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.16461700201034546,
|
| 651 |
+
"learning_rate": 0.0003925410556475763,
|
| 652 |
+
"loss": 0.9629362225532532,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.7557469010353088,
|
| 658 |
+
"learning_rate": 0.00037404498123644504,
|
| 659 |
+
"loss": 0.9638089537620544,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.22433336079120636,
|
| 665 |
+
"learning_rate": 0.0003557312014609784,
|
| 666 |
+
"loss": 0.9633023142814636,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 0.23267188668251038,
|
| 672 |
+
"learning_rate": 0.00033762622184548686,
|
| 673 |
+
"loss": 0.958705723285675,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.22426548933601814,
|
| 679 |
+
"eval_loss": 0.9016523400670318,
|
| 680 |
+
"eval_mse_loss": 0.7895195963175874,
|
| 681 |
+
"flow/cos_sim": 0.7757345382209238,
|
| 682 |
+
"flow/improvement_ratio": 0.47811366291078805,
|
| 683 |
+
"flow/mag_ratio_mean": 0.7582496140645519,
|
| 684 |
+
"flow/mag_ratio_std": 0.17954147260074746,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.22426548933601814,
|
| 690 |
+
"eval_loss": 0.9016523400670318,
|
| 691 |
+
"eval_mse_loss": 0.7895195963175874,
|
| 692 |
+
"eval_runtime": 37.7469,
|
| 693 |
+
"eval_samples_per_second": 741.598,
|
| 694 |
+
"eval_steps_per_second": 11.604,
|
| 695 |
+
"flow/cos_sim": 0.7757345382209238,
|
| 696 |
+
"flow/improvement_ratio": 0.47811366291078805,
|
| 697 |
+
"flow/mag_ratio_mean": 0.7582496140645519,
|
| 698 |
+
"flow/mag_ratio_std": 0.17954147260074746,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.8010876178741455,
|
| 704 |
+
"learning_rate": 0.00031975624571791317,
|
| 705 |
+
"loss": 0.9547749757766724,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.7472712397575378,
|
| 711 |
+
"learning_rate": 0.00030214713628576984,
|
| 712 |
+
"loss": 0.9569481015205383,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.37432172894477844,
|
| 718 |
+
"learning_rate": 0.000284891456531408,
|
| 719 |
+
"loss": 0.9571910500526428,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.5804730653762817,
|
| 725 |
+
"learning_rate": 0.00026787885831585745,
|
| 726 |
+
"loss": 0.9551773071289062,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.2237415857034731,
|
| 732 |
+
"eval_loss": 0.8986687686073181,
|
| 733 |
+
"eval_mse_loss": 0.7867979765210522,
|
| 734 |
+
"flow/cos_sim": 0.7762584320214241,
|
| 735 |
+
"flow/improvement_ratio": 0.4805531155437095,
|
| 736 |
+
"flow/mag_ratio_mean": 0.7577235157086969,
|
| 737 |
+
"flow/mag_ratio_std": 0.1783803814690407,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.2237415857034731,
|
| 743 |
+
"eval_loss": 0.8986687686073181,
|
| 744 |
+
"eval_mse_loss": 0.7867979765210522,
|
| 745 |
+
"eval_runtime": 37.8958,
|
| 746 |
+
"eval_samples_per_second": 738.683,
|
| 747 |
+
"eval_steps_per_second": 11.558,
|
| 748 |
+
"flow/cos_sim": 0.7762584320214241,
|
| 749 |
+
"flow/improvement_ratio": 0.4805531155437095,
|
| 750 |
+
"flow/mag_ratio_mean": 0.7577235157086969,
|
| 751 |
+
"flow/mag_ratio_std": 0.1783803814690407,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.4953038990497589,
|
| 757 |
+
"learning_rate": 0.0002512022089125798,
|
| 758 |
+
"loss": 0.9555047154426575,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 0.9414182901382446,
|
| 764 |
+
"learning_rate": 0.00023488564442791606,
|
| 765 |
+
"loss": 0.9574738144874573,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.4800240993499756,
|
| 771 |
+
"learning_rate": 0.0002189527798174848,
|
| 772 |
+
"loss": 0.9548642635345459,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 0.8845779299736023,
|
| 778 |
+
"learning_rate": 0.00020342667470826854,
|
| 779 |
+
"loss": 0.9534708261489868,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.22472695407529944,
|
| 785 |
+
"eval_loss": 0.9025066166980081,
|
| 786 |
+
"eval_mse_loss": 0.7901431399665467,
|
| 787 |
+
"flow/cos_sim": 0.7752730657248736,
|
| 788 |
+
"flow/improvement_ratio": 0.4882404270781774,
|
| 789 |
+
"flow/mag_ratio_mean": 0.7596338093280792,
|
| 790 |
+
"flow/mag_ratio_std": 0.17930837312381562,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.22472695407529944,
|
| 796 |
+
"eval_loss": 0.9025066166980081,
|
| 797 |
+
"eval_mse_loss": 0.7901431399665467,
|
| 798 |
+
"eval_runtime": 38.1115,
|
| 799 |
+
"eval_samples_per_second": 734.503,
|
| 800 |
+
"eval_steps_per_second": 11.493,
|
| 801 |
+
"flow/cos_sim": 0.7752730657248736,
|
| 802 |
+
"flow/improvement_ratio": 0.4882404270781774,
|
| 803 |
+
"flow/mag_ratio_mean": 0.7596338093280792,
|
| 804 |
+
"flow/mag_ratio_std": 0.17930837312381562,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.19289438426494598,
|
| 810 |
+
"learning_rate": 0.00018832980002442701,
|
| 811 |
+
"loss": 0.9571536779403687,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.377024382352829,
|
| 817 |
+
"learning_rate": 0.00017368400546514002,
|
| 818 |
+
"loss": 0.9562097191810608,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 0.2716529667377472,
|
| 824 |
+
"learning_rate": 0.00015951048788154866,
|
| 825 |
+
"loss": 0.9509574770927429,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.32783976197242737,
|
| 831 |
+
"learning_rate": 0.0001458297605985633,
|
| 832 |
+
"loss": 0.9516466856002808,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.22373602280741958,
|
| 838 |
+
"eval_loss": 0.8992317022253934,
|
| 839 |
+
"eval_mse_loss": 0.7873636899201293,
|
| 840 |
+
"flow/cos_sim": 0.7762640047835433,
|
| 841 |
+
"flow/improvement_ratio": 0.4824598703074129,
|
| 842 |
+
"flow/mag_ratio_mean": 0.7615949167236346,
|
| 843 |
+
"flow/mag_ratio_std": 0.179001236479032,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.22373602280741958,
|
| 849 |
+
"eval_loss": 0.8992317022253934,
|
| 850 |
+
"eval_mse_loss": 0.7873636899201293,
|
| 851 |
+
"eval_runtime": 37.8407,
|
| 852 |
+
"eval_samples_per_second": 739.758,
|
| 853 |
+
"eval_steps_per_second": 11.575,
|
| 854 |
+
"flow/cos_sim": 0.7762640047835433,
|
| 855 |
+
"flow/improvement_ratio": 0.4824598703074129,
|
| 856 |
+
"flow/mag_ratio_mean": 0.7615949167236346,
|
| 857 |
+
"flow/mag_ratio_std": 0.179001236479032,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.4769844710826874,
|
| 863 |
+
"learning_rate": 0.00013266162372593872,
|
| 864 |
+
"loss": 0.949449360370636,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.17671078443527222,
|
| 870 |
+
"learning_rate": 0.00012002513550158511,
|
| 871 |
+
"loss": 0.9504072666168213,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.5919731855392456,
|
| 877 |
+
"learning_rate": 0.00010798470498931034,
|
| 878 |
+
"loss": 0.9529365301132202,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.2042759358882904,
|
| 884 |
+
"learning_rate": 9.646333498481152e-05,
|
| 885 |
+
"loss": 0.9518898725509644,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.2236619240957308,
|
| 891 |
+
"eval_loss": 0.8992718941272666,
|
| 892 |
+
"eval_mse_loss": 0.7874409310077424,
|
| 893 |
+
"flow/cos_sim": 0.7763380925404971,
|
| 894 |
+
"flow/improvement_ratio": 0.4808275544480102,
|
| 895 |
+
"flow/mag_ratio_mean": 0.7600411922419996,
|
| 896 |
+
"flow/mag_ratio_std": 0.18042517123429197,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.2236619240957308,
|
| 902 |
+
"eval_loss": 0.8992718941272666,
|
| 903 |
+
"eval_mse_loss": 0.7874409310077424,
|
| 904 |
+
"eval_runtime": 37.719,
|
| 905 |
+
"eval_samples_per_second": 742.147,
|
| 906 |
+
"eval_steps_per_second": 11.612,
|
| 907 |
+
"flow/cos_sim": 0.7763380925404971,
|
| 908 |
+
"flow/improvement_ratio": 0.4808275544480102,
|
| 909 |
+
"flow/mag_ratio_mean": 0.7600411922419996,
|
| 910 |
+
"flow/mag_ratio_std": 0.18042517123429197,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.3804217576980591,
|
| 916 |
+
"learning_rate": 8.552600339377569e-05,
|
| 917 |
+
"loss": 0.9526547193527222,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.13327613472938538,
|
| 923 |
+
"learning_rate": 7.518853981074431e-05,
|
| 924 |
+
"loss": 0.9486812353134155,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.20542556047439575,
|
| 930 |
+
"learning_rate": 6.546590564159156e-05,
|
| 931 |
+
"loss": 0.949363112449646,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.45832833647727966,
|
| 937 |
+
"learning_rate": 5.637217244988779e-05,
|
| 938 |
+
"loss": 0.9490708708763123,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.22359807510354204,
|
| 944 |
+
"eval_loss": 0.8977929246207895,
|
| 945 |
+
"eval_mse_loss": 0.785993887681395,
|
| 946 |
+
"flow/cos_sim": 0.7764019544266131,
|
| 947 |
+
"flow/improvement_ratio": 0.47894510915970695,
|
| 948 |
+
"flow/mag_ratio_mean": 0.7616693674157199,
|
| 949 |
+
"flow/mag_ratio_std": 0.17972967324599828,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.22359807510354204,
|
| 955 |
+
"eval_loss": 0.8977929246207895,
|
| 956 |
+
"eval_mse_loss": 0.785993887681395,
|
| 957 |
+
"eval_runtime": 37.7351,
|
| 958 |
+
"eval_samples_per_second": 741.829,
|
| 959 |
+
"eval_steps_per_second": 11.607,
|
| 960 |
+
"flow/cos_sim": 0.7764019544266131,
|
| 961 |
+
"flow/improvement_ratio": 0.47894510915970695,
|
| 962 |
+
"flow/mag_ratio_mean": 0.7616693674157199,
|
| 963 |
+
"flow/mag_ratio_std": 0.17972967324599828,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.44674035906791687,
|
| 969 |
+
"learning_rate": 4.792050159113337e-05,
|
| 970 |
+
"loss": 0.9491938352584839,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.3422182500362396,
|
| 976 |
+
"learning_rate": 4.012312516433581e-05,
|
| 977 |
+
"loss": 0.9448590278625488,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.19175578653812408,
|
| 983 |
+
"learning_rate": 3.2991328308497545e-05,
|
| 984 |
+
"loss": 0.9500359892845154,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 0.5182244181632996,
|
| 990 |
+
"learning_rate": 2.653543286964183e-05,
|
| 991 |
+
"loss": 0.9520748853683472,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.2243065895623268,
|
| 997 |
+
"eval_loss": 0.9002899413239466,
|
| 998 |
+
"eval_mse_loss": 0.7881366464917519,
|
| 999 |
+
"flow/cos_sim": 0.7756934289790723,
|
| 1000 |
+
"flow/improvement_ratio": 0.48521961890943516,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.7599015614213465,
|
| 1002 |
+
"flow/mag_ratio_std": 0.17938740075178886,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.2243065895623268,
|
| 1008 |
+
"eval_loss": 0.9002899413239466,
|
| 1009 |
+
"eval_mse_loss": 0.7881366464917519,
|
| 1010 |
+
"eval_runtime": 38.0865,
|
| 1011 |
+
"eval_samples_per_second": 734.985,
|
| 1012 |
+
"eval_steps_per_second": 11.5,
|
| 1013 |
+
"flow/cos_sim": 0.7756934289790723,
|
| 1014 |
+
"flow/improvement_ratio": 0.48521961890943516,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.7599015614213465,
|
| 1016 |
+
"flow/mag_ratio_std": 0.17938740075178886,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.35820356011390686,
|
| 1022 |
+
"learning_rate": 2.076478246200819e-05,
|
| 1023 |
+
"loss": 0.9483771324157715,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.2084021270275116,
|
| 1029 |
+
"learning_rate": 1.5687728945045944e-05,
|
| 1030 |
+
"loss": 0.9474115371704102,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.39592018723487854,
|
| 1036 |
+
"learning_rate": 1.1311620335770879e-05,
|
| 1037 |
+
"loss": 0.9525643587112427,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.6320902109146118,
|
| 1043 |
+
"learning_rate": 7.642790173984836e-06,
|
| 1044 |
+
"loss": 0.9535797834396362,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.22213090517341275,
|
| 1050 |
+
"eval_loss": 0.8940546147899541,
|
| 1051 |
+
"eval_mse_loss": 0.7829891629687183,
|
| 1052 |
+
"flow/cos_sim": 0.7778691155453251,
|
| 1053 |
+
"flow/improvement_ratio": 0.4782056734986501,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.7631440506950361,
|
| 1055 |
+
"flow/mag_ratio_std": 0.18027915234026842,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.22213090517341275,
|
| 1061 |
+
"eval_loss": 0.8940546147899541,
|
| 1062 |
+
"eval_mse_loss": 0.7829891629687183,
|
| 1063 |
+
"eval_runtime": 37.8602,
|
| 1064 |
+
"eval_samples_per_second": 739.378,
|
| 1065 |
+
"eval_steps_per_second": 11.569,
|
| 1066 |
+
"flow/cos_sim": 0.7778691155453251,
|
| 1067 |
+
"flow/improvement_ratio": 0.4782056734986501,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.7631440506950361,
|
| 1069 |
+
"flow/mag_ratio_std": 0.18027915234026842,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.1165478378534317,
|
| 1075 |
+
"learning_rate": 4.686548355746001e-06,
|
| 1076 |
+
"loss": 0.9476454257965088,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.2622009813785553,
|
| 1082 |
+
"learning_rate": 2.447173448359541e-06,
|
| 1083 |
+
"loss": 0.9495624303817749,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.15396808087825775,
|
| 1089 |
+
"learning_rate": 9.324369820445933e-07,
|
| 1090 |
+
"loss": 0.9498383402824402,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.40939903259277344,
|
| 1096 |
+
"learning_rate": 1.3265238678672464e-07,
|
| 1097 |
+
"loss": 0.9499406218528748,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.2221550636639878,
|
| 1103 |
+
"eval_loss": 0.8939511321451021,
|
| 1104 |
+
"eval_mse_loss": 0.7828736011296102,
|
| 1105 |
+
"flow/cos_sim": 0.7778449566124781,
|
| 1106 |
+
"flow/improvement_ratio": 0.4865727505591362,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.7616105286498048,
|
| 1108 |
+
"flow/mag_ratio_std": 0.17894721432635774,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.2221550636639878,
|
| 1114 |
+
"eval_loss": 0.8939511321451021,
|
| 1115 |
+
"eval_mse_loss": 0.7828736011296102,
|
| 1116 |
+
"eval_runtime": 37.7989,
|
| 1117 |
+
"eval_samples_per_second": 740.578,
|
| 1118 |
+
"eval_steps_per_second": 11.588,
|
| 1119 |
+
"flow/cos_sim": 0.7778449566124781,
|
| 1120 |
+
"flow/improvement_ratio": 0.4865727505591362,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.7616105286498048,
|
| 1122 |
+
"flow/mag_ratio_std": 0.17894721432635774,
|
| 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-c/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03a8c5803c6e5663005c9dc7a4c98c213c8267649ccec42581475db390cc8017
|
| 3 |
+
size 5137
|