Upload folder using huggingface_hub
Browse files- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/ema.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/model.safetensors +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/optimizer.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/rng_state.pth +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/scaler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/scheduler.pt +3 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/trainer_state.json +1147 -0
- checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/training_args.bin +3 -0
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e95344803241bd9f9a2479032833623864645b5cb9f62543b600aeb722a3c0c8
|
| 3 |
+
size 550088
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:01c09c0140d9223aedb48dbae1713669e280c6ca5dd0c2a5f692719d9eab01d1
|
| 3 |
+
size 19331448
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71717e3164209d80a1ea7fee841888c0663af047a2979c421226c9e95dc61082
|
| 3 |
+
size 1175115
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/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-b/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c681c3ef28514ea1cfb29fe1861a5692253b9f3548b5de1939978218e27b227d
|
| 3 |
+
size 1383
|
checkpoints-v2.6-b-whitened/checkpoint-21651-cos-b/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-cos-b/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": 1.0396935939788818,
|
| 15 |
+
"learning_rate": 0.000494140625,
|
| 16 |
+
"loss": 2.385955572128296,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 0.3018300235271454,
|
| 22 |
+
"learning_rate": 0.000994140625,
|
| 23 |
+
"loss": 1.499626636505127,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.3906877040863037,
|
| 29 |
+
"learning_rate": 0.0009996466047131989,
|
| 30 |
+
"loss": 1.3862546682357788,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.6580488681793213,
|
| 36 |
+
"learning_rate": 0.000998570120840354,
|
| 37 |
+
"loss": 1.3300455808639526,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_cos_loss": 0.2912509761172343,
|
| 43 |
+
"eval_loss": 1.261210564064653,
|
| 44 |
+
"eval_mse_loss": 0.969959588219586,
|
| 45 |
+
"flow/cos_sim": 0.708749028645694,
|
| 46 |
+
"flow/improvement_ratio": 0.4786223828384321,
|
| 47 |
+
"flow/mag_ratio_mean": 0.6964796499846733,
|
| 48 |
+
"flow/mag_ratio_std": 0.14973047248100582,
|
| 49 |
+
"step": 1024
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.047295736917463395,
|
| 53 |
+
"eval_cos_loss": 0.2912509761172343,
|
| 54 |
+
"eval_loss": 1.261210564064653,
|
| 55 |
+
"eval_mse_loss": 0.969959588219586,
|
| 56 |
+
"eval_runtime": 39.6995,
|
| 57 |
+
"eval_samples_per_second": 705.121,
|
| 58 |
+
"eval_steps_per_second": 11.033,
|
| 59 |
+
"flow/cos_sim": 0.708749028645694,
|
| 60 |
+
"flow/improvement_ratio": 0.4786223828384321,
|
| 61 |
+
"flow/mag_ratio_mean": 0.6964796499846733,
|
| 62 |
+
"flow/mag_ratio_std": 0.14973047248100582,
|
| 63 |
+
"step": 1024
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.05911967114682925,
|
| 67 |
+
"grad_norm": 0.9962846040725708,
|
| 68 |
+
"learning_rate": 0.0009967720566809583,
|
| 69 |
+
"loss": 1.2950860261917114,
|
| 70 |
+
"step": 1280
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.0709436053761951,
|
| 74 |
+
"grad_norm": 0.29216381907463074,
|
| 75 |
+
"learning_rate": 0.0009942550145723697,
|
| 76 |
+
"loss": 1.2622082233428955,
|
| 77 |
+
"step": 1536
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.08276753960556095,
|
| 81 |
+
"grad_norm": 0.6702899932861328,
|
| 82 |
+
"learning_rate": 0.0009910226374283738,
|
| 83 |
+
"loss": 1.2361210584640503,
|
| 84 |
+
"step": 1792
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 0.09459147383492679,
|
| 88 |
+
"grad_norm": 0.4554673135280609,
|
| 89 |
+
"learning_rate": 0.0009870796034667942,
|
| 90 |
+
"loss": 1.2209129333496094,
|
| 91 |
+
"step": 2048
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.09459147383492679,
|
| 95 |
+
"eval_cos_loss": 0.26044403312549197,
|
| 96 |
+
"eval_loss": 1.152398099365844,
|
| 97 |
+
"eval_mse_loss": 0.8919540645052854,
|
| 98 |
+
"flow/cos_sim": 0.7395559614651823,
|
| 99 |
+
"flow/improvement_ratio": 0.479380850718446,
|
| 100 |
+
"flow/mag_ratio_mean": 0.7208647506149937,
|
| 101 |
+
"flow/mag_ratio_std": 0.15676253396150183,
|
| 102 |
+
"step": 2048
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.09459147383492679,
|
| 106 |
+
"eval_cos_loss": 0.26044403312549197,
|
| 107 |
+
"eval_loss": 1.152398099365844,
|
| 108 |
+
"eval_mse_loss": 0.8919540645052854,
|
| 109 |
+
"eval_runtime": 38.3879,
|
| 110 |
+
"eval_samples_per_second": 729.213,
|
| 111 |
+
"eval_steps_per_second": 11.41,
|
| 112 |
+
"flow/cos_sim": 0.7395559614651823,
|
| 113 |
+
"flow/improvement_ratio": 0.479380850718446,
|
| 114 |
+
"flow/mag_ratio_mean": 0.7208647506149937,
|
| 115 |
+
"flow/mag_ratio_std": 0.15676253396150183,
|
| 116 |
+
"step": 2048
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.10641540806429264,
|
| 120 |
+
"grad_norm": 0.28728047013282776,
|
| 121 |
+
"learning_rate": 0.0009824316194387115,
|
| 122 |
+
"loss": 1.2018136978149414,
|
| 123 |
+
"step": 2304
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.1182393422936585,
|
| 127 |
+
"grad_norm": 1.280680775642395,
|
| 128 |
+
"learning_rate": 0.0009770854123690854,
|
| 129 |
+
"loss": 1.1904348134994507,
|
| 130 |
+
"step": 2560
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.13006327652302435,
|
| 134 |
+
"grad_norm": 0.5957157611846924,
|
| 135 |
+
"learning_rate": 0.0009710487198207337,
|
| 136 |
+
"loss": 1.1770356893539429,
|
| 137 |
+
"step": 2816
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.1418872107523902,
|
| 141 |
+
"grad_norm": 1.3396246433258057,
|
| 142 |
+
"learning_rate": 0.0009643302786957629,
|
| 143 |
+
"loss": 1.1677006483078003,
|
| 144 |
+
"step": 3072
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.1418872107523902,
|
| 148 |
+
"eval_cos_loss": 0.24673845867315927,
|
| 149 |
+
"eval_loss": 1.1009203881433565,
|
| 150 |
+
"eval_mse_loss": 0.8541819294021554,
|
| 151 |
+
"flow/cos_sim": 0.7532615565001692,
|
| 152 |
+
"flow/improvement_ratio": 0.486232013190718,
|
| 153 |
+
"flow/mag_ratio_mean": 0.7330454062653459,
|
| 154 |
+
"flow/mag_ratio_std": 0.16009155762930438,
|
| 155 |
+
"step": 3072
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 0.1418872107523902,
|
| 159 |
+
"eval_cos_loss": 0.24673845867315927,
|
| 160 |
+
"eval_loss": 1.1009203881433565,
|
| 161 |
+
"eval_mse_loss": 0.8541819294021554,
|
| 162 |
+
"eval_runtime": 38.2149,
|
| 163 |
+
"eval_samples_per_second": 732.516,
|
| 164 |
+
"eval_steps_per_second": 11.462,
|
| 165 |
+
"flow/cos_sim": 0.7532615565001692,
|
| 166 |
+
"flow/improvement_ratio": 0.486232013190718,
|
| 167 |
+
"flow/mag_ratio_mean": 0.7330454062653459,
|
| 168 |
+
"flow/mag_ratio_std": 0.16009155762930438,
|
| 169 |
+
"step": 3072
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.15371114498175603,
|
| 173 |
+
"grad_norm": 0.3938586115837097,
|
| 174 |
+
"learning_rate": 0.0009569398125906527,
|
| 175 |
+
"loss": 1.161599040031433,
|
| 176 |
+
"step": 3328
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.1655350792111219,
|
| 180 |
+
"grad_norm": 0.2642618715763092,
|
| 181 |
+
"learning_rate": 0.0009488880177233017,
|
| 182 |
+
"loss": 1.1554925441741943,
|
| 183 |
+
"step": 3584
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.17735901344048774,
|
| 187 |
+
"grad_norm": 0.15925352275371552,
|
| 188 |
+
"learning_rate": 0.000940186547452394,
|
| 189 |
+
"loss": 1.142865538597107,
|
| 190 |
+
"step": 3840
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18918294766985358,
|
| 194 |
+
"grad_norm": 0.28947100043296814,
|
| 195 |
+
"learning_rate": 0.0009308479954115012,
|
| 196 |
+
"loss": 1.1450462341308594,
|
| 197 |
+
"step": 4096
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.18918294766985358,
|
| 201 |
+
"eval_cos_loss": 0.23922484527984167,
|
| 202 |
+
"eval_loss": 1.070648630596187,
|
| 203 |
+
"eval_mse_loss": 0.8314237849080943,
|
| 204 |
+
"flow/cos_sim": 0.7607751724110347,
|
| 205 |
+
"flow/improvement_ratio": 0.4767084107415317,
|
| 206 |
+
"flow/mag_ratio_mean": 0.7456348471717748,
|
| 207 |
+
"flow/mag_ratio_std": 0.16709686444910693,
|
| 208 |
+
"step": 4096
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.18918294766985358,
|
| 212 |
+
"eval_cos_loss": 0.23922484527984167,
|
| 213 |
+
"eval_loss": 1.070648630596187,
|
| 214 |
+
"eval_mse_loss": 0.8314237849080943,
|
| 215 |
+
"eval_runtime": 37.9421,
|
| 216 |
+
"eval_samples_per_second": 737.782,
|
| 217 |
+
"eval_steps_per_second": 11.544,
|
| 218 |
+
"flow/cos_sim": 0.7607751724110347,
|
| 219 |
+
"flow/improvement_ratio": 0.4767084107415317,
|
| 220 |
+
"flow/mag_ratio_mean": 0.7456348471717748,
|
| 221 |
+
"flow/mag_ratio_std": 0.16709686444910693,
|
| 222 |
+
"step": 4096
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.20100688189921945,
|
| 226 |
+
"grad_norm": 0.2802768349647522,
|
| 227 |
+
"learning_rate": 0.0009208858772823219,
|
| 228 |
+
"loss": 1.1381992101669312,
|
| 229 |
+
"step": 4352
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.2128308161285853,
|
| 233 |
+
"grad_norm": 0.21698614954948425,
|
| 234 |
+
"learning_rate": 0.000910314611233442,
|
| 235 |
+
"loss": 1.132680058479309,
|
| 236 |
+
"step": 4608
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.22465475035795113,
|
| 240 |
+
"grad_norm": 0.32471123337745667,
|
| 241 |
+
"learning_rate": 0.0008991494970529273,
|
| 242 |
+
"loss": 1.1287866830825806,
|
| 243 |
+
"step": 4864
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 0.236478684587317,
|
| 247 |
+
"grad_norm": 0.2330758422613144,
|
| 248 |
+
"learning_rate": 0.0008874066940049477,
|
| 249 |
+
"loss": 1.1182652711868286,
|
| 250 |
+
"step": 5120
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.236478684587317,
|
| 254 |
+
"eval_cos_loss": 0.23335642752037744,
|
| 255 |
+
"eval_loss": 1.0504317106721608,
|
| 256 |
+
"eval_mse_loss": 0.8170752830156996,
|
| 257 |
+
"flow/cos_sim": 0.7666435818693954,
|
| 258 |
+
"flow/improvement_ratio": 0.4764664819523624,
|
| 259 |
+
"flow/mag_ratio_mean": 0.7472941716239877,
|
| 260 |
+
"flow/mag_ratio_std": 0.16736687576934084,
|
| 261 |
+
"step": 5120
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.236478684587317,
|
| 265 |
+
"eval_cos_loss": 0.23335642752037744,
|
| 266 |
+
"eval_loss": 1.0504317106721608,
|
| 267 |
+
"eval_mse_loss": 0.8170752830156996,
|
| 268 |
+
"eval_runtime": 37.1297,
|
| 269 |
+
"eval_samples_per_second": 753.925,
|
| 270 |
+
"eval_steps_per_second": 11.796,
|
| 271 |
+
"flow/cos_sim": 0.7666435818693954,
|
| 272 |
+
"flow/improvement_ratio": 0.4764664819523624,
|
| 273 |
+
"flow/mag_ratio_mean": 0.7472941716239877,
|
| 274 |
+
"flow/mag_ratio_std": 0.16736687576934084,
|
| 275 |
+
"step": 5120
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.24830261881668284,
|
| 279 |
+
"grad_norm": 0.22401200234889984,
|
| 280 |
+
"learning_rate": 0.0008751031974424814,
|
| 281 |
+
"loss": 1.1156164407730103,
|
| 282 |
+
"step": 5376
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2601265530460487,
|
| 286 |
+
"grad_norm": 0.30990755558013916,
|
| 287 |
+
"learning_rate": 0.0008622568142099504,
|
| 288 |
+
"loss": 1.1163040399551392,
|
| 289 |
+
"step": 5632
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.27195048727541454,
|
| 293 |
+
"grad_norm": 0.36347290873527527,
|
| 294 |
+
"learning_rate": 0.0008488861368713825,
|
| 295 |
+
"loss": 1.1100337505340576,
|
| 296 |
+
"step": 5888
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2837744215047804,
|
| 300 |
+
"grad_norm": 0.272342324256897,
|
| 301 |
+
"learning_rate": 0.0008350105168014028,
|
| 302 |
+
"loss": 1.111609697341919,
|
| 303 |
+
"step": 6144
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2837744215047804,
|
| 307 |
+
"eval_cos_loss": 0.2336256698318268,
|
| 308 |
+
"eval_loss": 1.049514850797174,
|
| 309 |
+
"eval_mse_loss": 0.8158891825643304,
|
| 310 |
+
"flow/cos_sim": 0.7663743587117217,
|
| 311 |
+
"flow/improvement_ratio": 0.4811141695742193,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7543826833979724,
|
| 313 |
+
"flow/mag_ratio_std": 0.17351540572703156,
|
| 314 |
+
"step": 6144
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.2837744215047804,
|
| 318 |
+
"eval_cos_loss": 0.2336256698318268,
|
| 319 |
+
"eval_loss": 1.049514850797174,
|
| 320 |
+
"eval_mse_loss": 0.8158891825643304,
|
| 321 |
+
"eval_runtime": 37.1083,
|
| 322 |
+
"eval_samples_per_second": 754.36,
|
| 323 |
+
"eval_steps_per_second": 11.803,
|
| 324 |
+
"flow/cos_sim": 0.7663743587117217,
|
| 325 |
+
"flow/improvement_ratio": 0.4811141695742193,
|
| 326 |
+
"flow/mag_ratio_mean": 0.7543826833979724,
|
| 327 |
+
"flow/mag_ratio_std": 0.17351540572703156,
|
| 328 |
+
"step": 6144
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.2955983557341462,
|
| 332 |
+
"grad_norm": 0.1812877058982849,
|
| 333 |
+
"learning_rate": 0.0008206500361779995,
|
| 334 |
+
"loss": 1.102935552597046,
|
| 335 |
+
"step": 6400
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.30742228996351206,
|
| 339 |
+
"grad_norm": 0.7799150347709656,
|
| 340 |
+
"learning_rate": 0.0008058254789175967,
|
| 341 |
+
"loss": 1.1035031080245972,
|
| 342 |
+
"step": 6656
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 0.3192462241928779,
|
| 346 |
+
"grad_norm": 0.2194916158914566,
|
| 347 |
+
"learning_rate": 0.0007905583005945037,
|
| 348 |
+
"loss": 1.102286696434021,
|
| 349 |
+
"step": 6912
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 0.3310701584222438,
|
| 353 |
+
"grad_norm": 0.4366890788078308,
|
| 354 |
+
"learning_rate": 0.000774870597388272,
|
| 355 |
+
"loss": 1.101851463317871,
|
| 356 |
+
"step": 7168
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.3310701584222438,
|
| 360 |
+
"eval_cos_loss": 0.229756694372114,
|
| 361 |
+
"eval_loss": 1.035625888876719,
|
| 362 |
+
"eval_mse_loss": 0.8058691928375802,
|
| 363 |
+
"flow/cos_sim": 0.7702433085877057,
|
| 364 |
+
"flow/improvement_ratio": 0.4816136150599615,
|
| 365 |
+
"flow/mag_ratio_mean": 0.7532055862418049,
|
| 366 |
+
"flow/mag_ratio_std": 0.1725723232444563,
|
| 367 |
+
"step": 7168
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.3310701584222438,
|
| 371 |
+
"eval_cos_loss": 0.229756694372114,
|
| 372 |
+
"eval_loss": 1.035625888876719,
|
| 373 |
+
"eval_mse_loss": 0.8058691928375802,
|
| 374 |
+
"eval_runtime": 37.4619,
|
| 375 |
+
"eval_samples_per_second": 747.238,
|
| 376 |
+
"eval_steps_per_second": 11.692,
|
| 377 |
+
"flow/cos_sim": 0.7702433085877057,
|
| 378 |
+
"flow/improvement_ratio": 0.4816136150599615,
|
| 379 |
+
"flow/mag_ratio_mean": 0.7532055862418049,
|
| 380 |
+
"flow/mag_ratio_std": 0.1725723232444563,
|
| 381 |
+
"step": 7168
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.34289409265160964,
|
| 385 |
+
"grad_norm": 0.6802489161491394,
|
| 386 |
+
"learning_rate": 0.0007587850741039065,
|
| 387 |
+
"loss": 1.1002918481826782,
|
| 388 |
+
"step": 7424
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.3547180268809755,
|
| 392 |
+
"grad_norm": 0.19382832944393158,
|
| 393 |
+
"learning_rate": 0.000742325011311212,
|
| 394 |
+
"loss": 1.0985705852508545,
|
| 395 |
+
"step": 7680
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.3665419611103413,
|
| 399 |
+
"grad_norm": 0.8463470935821533,
|
| 400 |
+
"learning_rate": 0.0007255142316508366,
|
| 401 |
+
"loss": 1.0968726873397827,
|
| 402 |
+
"step": 7936
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.37836589533970716,
|
| 406 |
+
"grad_norm": 0.20921163260936737,
|
| 407 |
+
"learning_rate": 0.0007083770653557752,
|
| 408 |
+
"loss": 1.0953953266143799,
|
| 409 |
+
"step": 8192
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.37836589533970716,
|
| 413 |
+
"eval_cos_loss": 0.2280988702422952,
|
| 414 |
+
"eval_loss": 1.0291743364236126,
|
| 415 |
+
"eval_mse_loss": 0.8010754635617069,
|
| 416 |
+
"flow/cos_sim": 0.7719011534022414,
|
| 417 |
+
"flow/improvement_ratio": 0.4799054505890363,
|
| 418 |
+
"flow/mag_ratio_mean": 0.7548060637630828,
|
| 419 |
+
"flow/mag_ratio_std": 0.17471813757533897,
|
| 420 |
+
"step": 8192
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.37836589533970716,
|
| 424 |
+
"eval_cos_loss": 0.2280988702422952,
|
| 425 |
+
"eval_loss": 1.0291743364236126,
|
| 426 |
+
"eval_mse_loss": 0.8010754635617069,
|
| 427 |
+
"eval_runtime": 37.2649,
|
| 428 |
+
"eval_samples_per_second": 751.19,
|
| 429 |
+
"eval_steps_per_second": 11.754,
|
| 430 |
+
"flow/cos_sim": 0.7719011534022414,
|
| 431 |
+
"flow/improvement_ratio": 0.4799054505890363,
|
| 432 |
+
"flow/mag_ratio_mean": 0.7548060637630828,
|
| 433 |
+
"flow/mag_ratio_std": 0.17471813757533897,
|
| 434 |
+
"step": 8192
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.390189829569073,
|
| 438 |
+
"grad_norm": 0.3799210488796234,
|
| 439 |
+
"learning_rate": 0.000691006989350691,
|
| 440 |
+
"loss": 1.0918513536453247,
|
| 441 |
+
"step": 8448
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.4020137637984389,
|
| 445 |
+
"grad_norm": 0.9599165916442871,
|
| 446 |
+
"learning_rate": 0.0006732929239454054,
|
| 447 |
+
"loss": 1.0943893194198608,
|
| 448 |
+
"step": 8704
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.41383769802780473,
|
| 452 |
+
"grad_norm": 0.19293536245822906,
|
| 453 |
+
"learning_rate": 0.0006553280517779596,
|
| 454 |
+
"loss": 1.0910601615905762,
|
| 455 |
+
"step": 8960
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.4256616322571706,
|
| 459 |
+
"grad_norm": 1.4826339483261108,
|
| 460 |
+
"learning_rate": 0.000637138373398857,
|
| 461 |
+
"loss": 1.0878775119781494,
|
| 462 |
+
"step": 9216
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.4256616322571706,
|
| 466 |
+
"eval_cos_loss": 0.2281181326467697,
|
| 467 |
+
"eval_loss": 1.0284691565929482,
|
| 468 |
+
"eval_mse_loss": 0.8003510253070152,
|
| 469 |
+
"flow/cos_sim": 0.7718818755182502,
|
| 470 |
+
"flow/improvement_ratio": 0.47831420069687985,
|
| 471 |
+
"flow/mag_ratio_mean": 0.7562852190237611,
|
| 472 |
+
"flow/mag_ratio_std": 0.17577999796225055,
|
| 473 |
+
"step": 9216
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.4256616322571706,
|
| 477 |
+
"eval_cos_loss": 0.2281181326467697,
|
| 478 |
+
"eval_loss": 1.0284691565929482,
|
| 479 |
+
"eval_mse_loss": 0.8003510253070152,
|
| 480 |
+
"eval_runtime": 37.4975,
|
| 481 |
+
"eval_samples_per_second": 746.53,
|
| 482 |
+
"eval_steps_per_second": 11.681,
|
| 483 |
+
"flow/cos_sim": 0.7718818755182502,
|
| 484 |
+
"flow/improvement_ratio": 0.47831420069687985,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7562852190237611,
|
| 486 |
+
"flow/mag_ratio_std": 0.17577999796225055,
|
| 487 |
+
"step": 9216
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.4374855664865364,
|
| 491 |
+
"grad_norm": 0.531550407409668,
|
| 492 |
+
"learning_rate": 0.0006187502147205189,
|
| 493 |
+
"loss": 1.0887291431427002,
|
| 494 |
+
"step": 9472
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.44930950071590225,
|
| 498 |
+
"grad_norm": 0.25300586223602295,
|
| 499 |
+
"learning_rate": 0.000600190188915804,
|
| 500 |
+
"loss": 1.0868346691131592,
|
| 501 |
+
"step": 9728
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.4611334349452681,
|
| 505 |
+
"grad_norm": 0.668536901473999,
|
| 506 |
+
"learning_rate": 0.0005814851579007778,
|
| 507 |
+
"loss": 1.0852059125900269,
|
| 508 |
+
"step": 9984
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 0.472957369174634,
|
| 512 |
+
"grad_norm": 0.7231985926628113,
|
| 513 |
+
"learning_rate": 0.0005626621934574736,
|
| 514 |
+
"loss": 1.0829912424087524,
|
| 515 |
+
"step": 10240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 0.472957369174634,
|
| 519 |
+
"eval_cos_loss": 0.22645327139254576,
|
| 520 |
+
"eval_loss": 1.0233922791263284,
|
| 521 |
+
"eval_mse_loss": 0.7969390074956363,
|
| 522 |
+
"flow/cos_sim": 0.7735467391199173,
|
| 523 |
+
"flow/improvement_ratio": 0.4819793237126581,
|
| 524 |
+
"flow/mag_ratio_mean": 0.7563864052567852,
|
| 525 |
+
"flow/mag_ratio_std": 0.17542334914751792,
|
| 526 |
+
"step": 10240
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.472957369174634,
|
| 530 |
+
"eval_cos_loss": 0.22645327139254576,
|
| 531 |
+
"eval_loss": 1.0233922791263284,
|
| 532 |
+
"eval_mse_loss": 0.7969390074956363,
|
| 533 |
+
"eval_runtime": 37.4007,
|
| 534 |
+
"eval_samples_per_second": 748.463,
|
| 535 |
+
"eval_steps_per_second": 11.711,
|
| 536 |
+
"flow/cos_sim": 0.7735467391199173,
|
| 537 |
+
"flow/improvement_ratio": 0.4819793237126581,
|
| 538 |
+
"flow/mag_ratio_mean": 0.7563864052567852,
|
| 539 |
+
"flow/mag_ratio_std": 0.17542334914751792,
|
| 540 |
+
"step": 10240
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.48478130340399983,
|
| 544 |
+
"grad_norm": 0.9182093143463135,
|
| 545 |
+
"learning_rate": 0.0005437485380529165,
|
| 546 |
+
"loss": 1.0810939073562622,
|
| 547 |
+
"step": 10496
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"epoch": 0.49660523763336567,
|
| 551 |
+
"grad_norm": 0.7298511862754822,
|
| 552 |
+
"learning_rate": 0.0005247715654111146,
|
| 553 |
+
"loss": 1.0796959400177002,
|
| 554 |
+
"step": 10752
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.5084291718627315,
|
| 558 |
+
"grad_norm": 0.23945634067058563,
|
| 559 |
+
"learning_rate": 0.0005057587408950816,
|
| 560 |
+
"loss": 1.0790259838104248,
|
| 561 |
+
"step": 11008
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.5202531060920974,
|
| 565 |
+
"grad_norm": 1.0996612310409546,
|
| 566 |
+
"learning_rate": 0.00048673758175623157,
|
| 567 |
+
"loss": 1.0788182020187378,
|
| 568 |
+
"step": 11264
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.5202531060920974,
|
| 572 |
+
"eval_cos_loss": 0.2244913134400703,
|
| 573 |
+
"eval_loss": 1.01628939324318,
|
| 574 |
+
"eval_mse_loss": 0.7917980787144404,
|
| 575 |
+
"flow/cos_sim": 0.7755086994878778,
|
| 576 |
+
"flow/improvement_ratio": 0.4844034041988251,
|
| 577 |
+
"flow/mag_ratio_mean": 0.7616438624793536,
|
| 578 |
+
"flow/mag_ratio_std": 0.17625076667358885,
|
| 579 |
+
"step": 11264
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.5202531060920974,
|
| 583 |
+
"eval_cos_loss": 0.2244913134400703,
|
| 584 |
+
"eval_loss": 1.01628939324318,
|
| 585 |
+
"eval_mse_loss": 0.7917980787144404,
|
| 586 |
+
"eval_runtime": 37.3175,
|
| 587 |
+
"eval_samples_per_second": 750.131,
|
| 588 |
+
"eval_steps_per_second": 11.737,
|
| 589 |
+
"flow/cos_sim": 0.7755086994878778,
|
| 590 |
+
"flow/improvement_ratio": 0.4844034041988251,
|
| 591 |
+
"flow/mag_ratio_mean": 0.7616438624793536,
|
| 592 |
+
"flow/mag_ratio_std": 0.17625076667358885,
|
| 593 |
+
"step": 11264
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5320770403214632,
|
| 597 |
+
"grad_norm": 0.23548179864883423,
|
| 598 |
+
"learning_rate": 0.00046773561730867265,
|
| 599 |
+
"loss": 1.0772168636322021,
|
| 600 |
+
"step": 11520
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.5439009745508291,
|
| 604 |
+
"grad_norm": 0.4916202425956726,
|
| 605 |
+
"learning_rate": 0.0004487803490860472,
|
| 606 |
+
"loss": 1.0812937021255493,
|
| 607 |
+
"step": 11776
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.5557249087801949,
|
| 611 |
+
"grad_norm": 1.28579843044281,
|
| 612 |
+
"learning_rate": 0.00042989921103857385,
|
| 613 |
+
"loss": 1.0812631845474243,
|
| 614 |
+
"step": 12032
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.5675488430095608,
|
| 618 |
+
"grad_norm": 1.1461721658706665,
|
| 619 |
+
"learning_rate": 0.0004111195298279063,
|
| 620 |
+
"loss": 1.0747902393341064,
|
| 621 |
+
"step": 12288
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.5675488430095608,
|
| 625 |
+
"eval_cos_loss": 0.2239211694630858,
|
| 626 |
+
"eval_loss": 1.0131171265421393,
|
| 627 |
+
"eval_mse_loss": 0.789195956432656,
|
| 628 |
+
"flow/cos_sim": 0.7760788515278193,
|
| 629 |
+
"flow/improvement_ratio": 0.48404773266892454,
|
| 630 |
+
"flow/mag_ratio_mean": 0.753236744637903,
|
| 631 |
+
"flow/mag_ratio_std": 0.17332864685417854,
|
| 632 |
+
"step": 12288
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5675488430095608,
|
| 636 |
+
"eval_cos_loss": 0.2239211694630858,
|
| 637 |
+
"eval_loss": 1.0131171265421393,
|
| 638 |
+
"eval_mse_loss": 0.789195956432656,
|
| 639 |
+
"eval_runtime": 37.4967,
|
| 640 |
+
"eval_samples_per_second": 746.546,
|
| 641 |
+
"eval_steps_per_second": 11.681,
|
| 642 |
+
"flow/cos_sim": 0.7760788515278193,
|
| 643 |
+
"flow/improvement_ratio": 0.48404773266892454,
|
| 644 |
+
"flow/mag_ratio_mean": 0.753236744637903,
|
| 645 |
+
"flow/mag_ratio_std": 0.17332864685417854,
|
| 646 |
+
"step": 12288
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.5793727772389267,
|
| 650 |
+
"grad_norm": 0.2644420266151428,
|
| 651 |
+
"learning_rate": 0.0003925410556475763,
|
| 652 |
+
"loss": 1.0753477811813354,
|
| 653 |
+
"step": 12544
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5911967114682924,
|
| 657 |
+
"grad_norm": 0.403468519449234,
|
| 658 |
+
"learning_rate": 0.00037404498123644504,
|
| 659 |
+
"loss": 1.077595829963684,
|
| 660 |
+
"step": 12800
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6030206456976583,
|
| 664 |
+
"grad_norm": 0.1891508400440216,
|
| 665 |
+
"learning_rate": 0.0003557312014609784,
|
| 666 |
+
"loss": 1.077918291091919,
|
| 667 |
+
"step": 13056
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.6148445799270241,
|
| 671 |
+
"grad_norm": 1.224328637123108,
|
| 672 |
+
"learning_rate": 0.00033762622184548686,
|
| 673 |
+
"loss": 1.072798490524292,
|
| 674 |
+
"step": 13312
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6148445799270241,
|
| 678 |
+
"eval_cos_loss": 0.22266034091308237,
|
| 679 |
+
"eval_loss": 1.009104470956271,
|
| 680 |
+
"eval_mse_loss": 0.7864441286483311,
|
| 681 |
+
"flow/cos_sim": 0.7773396887191354,
|
| 682 |
+
"flow/improvement_ratio": 0.47814115698206916,
|
| 683 |
+
"flow/mag_ratio_mean": 0.7620742424437988,
|
| 684 |
+
"flow/mag_ratio_std": 0.1770942181212717,
|
| 685 |
+
"step": 13312
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.6148445799270241,
|
| 689 |
+
"eval_cos_loss": 0.22266034091308237,
|
| 690 |
+
"eval_loss": 1.009104470956271,
|
| 691 |
+
"eval_mse_loss": 0.7864441286483311,
|
| 692 |
+
"eval_runtime": 37.4951,
|
| 693 |
+
"eval_samples_per_second": 746.577,
|
| 694 |
+
"eval_steps_per_second": 11.682,
|
| 695 |
+
"flow/cos_sim": 0.7773396887191354,
|
| 696 |
+
"flow/improvement_ratio": 0.47814115698206916,
|
| 697 |
+
"flow/mag_ratio_mean": 0.7620742424437988,
|
| 698 |
+
"flow/mag_ratio_std": 0.1770942181212717,
|
| 699 |
+
"step": 13312
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.62666851415639,
|
| 703 |
+
"grad_norm": 0.3658396601676941,
|
| 704 |
+
"learning_rate": 0.00031975624571791317,
|
| 705 |
+
"loss": 1.069403052330017,
|
| 706 |
+
"step": 13568
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 0.6384924483857558,
|
| 710 |
+
"grad_norm": 0.5004097819328308,
|
| 711 |
+
"learning_rate": 0.00030214713628576984,
|
| 712 |
+
"loss": 1.0717116594314575,
|
| 713 |
+
"step": 13824
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6503163826151217,
|
| 717 |
+
"grad_norm": 0.8231380581855774,
|
| 718 |
+
"learning_rate": 0.00028482437920433144,
|
| 719 |
+
"loss": 1.0720727443695068,
|
| 720 |
+
"step": 14080
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.6621403168444876,
|
| 724 |
+
"grad_norm": 0.4187723994255066,
|
| 725 |
+
"learning_rate": 0.00026781304569125866,
|
| 726 |
+
"loss": 1.0699665546417236,
|
| 727 |
+
"step": 14336
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.6621403168444876,
|
| 731 |
+
"eval_cos_loss": 0.22183906633968223,
|
| 732 |
+
"eval_loss": 1.0048139836962364,
|
| 733 |
+
"eval_mse_loss": 0.782974916642115,
|
| 734 |
+
"flow/cos_sim": 0.7781609424717351,
|
| 735 |
+
"flow/improvement_ratio": 0.4805064608925554,
|
| 736 |
+
"flow/mag_ratio_mean": 0.7602974135581761,
|
| 737 |
+
"flow/mag_ratio_std": 0.17573183972269432,
|
| 738 |
+
"step": 14336
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.6621403168444876,
|
| 742 |
+
"eval_cos_loss": 0.22183906633968223,
|
| 743 |
+
"eval_loss": 1.0048139836962364,
|
| 744 |
+
"eval_mse_loss": 0.782974916642115,
|
| 745 |
+
"eval_runtime": 37.5509,
|
| 746 |
+
"eval_samples_per_second": 745.468,
|
| 747 |
+
"eval_steps_per_second": 11.664,
|
| 748 |
+
"flow/cos_sim": 0.7781609424717351,
|
| 749 |
+
"flow/improvement_ratio": 0.4805064608925554,
|
| 750 |
+
"flow/mag_ratio_mean": 0.7602974135581761,
|
| 751 |
+
"flow/mag_ratio_std": 0.17573183972269432,
|
| 752 |
+
"step": 14336
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.6739642510738534,
|
| 756 |
+
"grad_norm": 0.4959767460823059,
|
| 757 |
+
"learning_rate": 0.0002511377562410373,
|
| 758 |
+
"loss": 1.0708597898483276,
|
| 759 |
+
"step": 14592
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.6857881853032193,
|
| 763 |
+
"grad_norm": 1.066572904586792,
|
| 764 |
+
"learning_rate": 0.00023482264499174915,
|
| 765 |
+
"loss": 1.073146104812622,
|
| 766 |
+
"step": 14848
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 0.6976121195325851,
|
| 770 |
+
"grad_norm": 0.3535749018192291,
|
| 771 |
+
"learning_rate": 0.00021889132479574586,
|
| 772 |
+
"loss": 1.0699703693389893,
|
| 773 |
+
"step": 15104
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.709436053761951,
|
| 777 |
+
"grad_norm": 0.37873294949531555,
|
| 778 |
+
"learning_rate": 0.0002033668530447801,
|
| 779 |
+
"loss": 1.0678328275680542,
|
| 780 |
+
"step": 15360
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.709436053761951,
|
| 784 |
+
"eval_cos_loss": 0.22249556676437865,
|
| 785 |
+
"eval_loss": 1.0087477605636805,
|
| 786 |
+
"eval_mse_loss": 0.7862521942075529,
|
| 787 |
+
"flow/cos_sim": 0.777504450790414,
|
| 788 |
+
"flow/improvement_ratio": 0.4885656609791055,
|
| 789 |
+
"flow/mag_ratio_mean": 0.7583253439430777,
|
| 790 |
+
"flow/mag_ratio_std": 0.17545615629926664,
|
| 791 |
+
"step": 15360
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.709436053761951,
|
| 795 |
+
"eval_cos_loss": 0.22249556676437865,
|
| 796 |
+
"eval_loss": 1.0087477605636805,
|
| 797 |
+
"eval_mse_loss": 0.7862521942075529,
|
| 798 |
+
"eval_runtime": 37.6925,
|
| 799 |
+
"eval_samples_per_second": 742.668,
|
| 800 |
+
"eval_steps_per_second": 11.62,
|
| 801 |
+
"flow/cos_sim": 0.777504450790414,
|
| 802 |
+
"flow/improvement_ratio": 0.4885656609791055,
|
| 803 |
+
"flow/mag_ratio_mean": 0.7583253439430777,
|
| 804 |
+
"flow/mag_ratio_std": 0.17545615629926664,
|
| 805 |
+
"step": 15360
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.7212599879913169,
|
| 809 |
+
"grad_norm": 0.4826560914516449,
|
| 810 |
+
"learning_rate": 0.0001882716982990524,
|
| 811 |
+
"loss": 1.0732594728469849,
|
| 812 |
+
"step": 15616
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.7330839222206826,
|
| 816 |
+
"grad_norm": 0.22264961898326874,
|
| 817 |
+
"learning_rate": 0.00017362770776847765,
|
| 818 |
+
"loss": 1.0719412565231323,
|
| 819 |
+
"step": 15872
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.7449078564500485,
|
| 823 |
+
"grad_norm": 0.38622406125068665,
|
| 824 |
+
"learning_rate": 0.00015945607569322633,
|
| 825 |
+
"loss": 1.0658717155456543,
|
| 826 |
+
"step": 16128
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.7567317906794143,
|
| 830 |
+
"grad_norm": 0.44819656014442444,
|
| 831 |
+
"learning_rate": 0.0001458297605985633,
|
| 832 |
+
"loss": 1.0668818950653076,
|
| 833 |
+
"step": 16384
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.7567317906794143,
|
| 837 |
+
"eval_cos_loss": 0.22197851242676173,
|
| 838 |
+
"eval_loss": 1.0059498877830158,
|
| 839 |
+
"eval_mse_loss": 0.7839713738933546,
|
| 840 |
+
"flow/cos_sim": 0.7780215108775657,
|
| 841 |
+
"flow/improvement_ratio": 0.48246467950409405,
|
| 842 |
+
"flow/mag_ratio_mean": 0.7628998027000253,
|
| 843 |
+
"flow/mag_ratio_std": 0.1759639097607299,
|
| 844 |
+
"step": 16384
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.7567317906794143,
|
| 848 |
+
"eval_cos_loss": 0.22197851242676173,
|
| 849 |
+
"eval_loss": 1.0059498877830158,
|
| 850 |
+
"eval_mse_loss": 0.7839713738933546,
|
| 851 |
+
"eval_runtime": 37.5688,
|
| 852 |
+
"eval_samples_per_second": 745.113,
|
| 853 |
+
"eval_steps_per_second": 11.659,
|
| 854 |
+
"flow/cos_sim": 0.7780215108775657,
|
| 855 |
+
"flow/improvement_ratio": 0.48246467950409405,
|
| 856 |
+
"flow/mag_ratio_mean": 0.7628998027000253,
|
| 857 |
+
"flow/mag_ratio_std": 0.1759639097607299,
|
| 858 |
+
"step": 16384
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.7685557249087802,
|
| 862 |
+
"grad_norm": 0.40678733587265015,
|
| 863 |
+
"learning_rate": 0.00013266162372593872,
|
| 864 |
+
"loss": 1.0644735097885132,
|
| 865 |
+
"step": 16640
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.780379659138146,
|
| 869 |
+
"grad_norm": 0.2160959541797638,
|
| 870 |
+
"learning_rate": 0.00012002513550158511,
|
| 871 |
+
"loss": 1.066280484199524,
|
| 872 |
+
"step": 16896
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.7922035933675119,
|
| 876 |
+
"grad_norm": 0.32006150484085083,
|
| 877 |
+
"learning_rate": 0.00010793858470858986,
|
| 878 |
+
"loss": 1.0685601234436035,
|
| 879 |
+
"step": 17152
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.8040275275968778,
|
| 883 |
+
"grad_norm": 0.23137597739696503,
|
| 884 |
+
"learning_rate": 9.641946420587128e-05,
|
| 885 |
+
"loss": 1.0673739910125732,
|
| 886 |
+
"step": 17408
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.8040275275968778,
|
| 890 |
+
"eval_cos_loss": 0.22172908348852097,
|
| 891 |
+
"eval_loss": 1.0058240679573252,
|
| 892 |
+
"eval_mse_loss": 0.784094982495591,
|
| 893 |
+
"flow/cos_sim": 0.7782709329095605,
|
| 894 |
+
"flow/improvement_ratio": 0.4807120756879789,
|
| 895 |
+
"flow/mag_ratio_mean": 0.7601865844367302,
|
| 896 |
+
"flow/mag_ratio_std": 0.17613858596919335,
|
| 897 |
+
"step": 17408
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.8040275275968778,
|
| 901 |
+
"eval_cos_loss": 0.22172908348852097,
|
| 902 |
+
"eval_loss": 1.0058240679573252,
|
| 903 |
+
"eval_mse_loss": 0.784094982495591,
|
| 904 |
+
"eval_runtime": 37.7929,
|
| 905 |
+
"eval_samples_per_second": 740.695,
|
| 906 |
+
"eval_steps_per_second": 11.589,
|
| 907 |
+
"flow/cos_sim": 0.7782709329095605,
|
| 908 |
+
"flow/improvement_ratio": 0.4807120756879789,
|
| 909 |
+
"flow/mag_ratio_mean": 0.7601865844367302,
|
| 910 |
+
"flow/mag_ratio_std": 0.17613858596919335,
|
| 911 |
+
"step": 17408
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 0.8158514618262436,
|
| 915 |
+
"grad_norm": 0.1639220118522644,
|
| 916 |
+
"learning_rate": 8.548444561077174e-05,
|
| 917 |
+
"loss": 1.068544864654541,
|
| 918 |
+
"step": 17664
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.8276753960556095,
|
| 922 |
+
"grad_norm": 0.217204749584198,
|
| 923 |
+
"learning_rate": 7.514935517023558e-05,
|
| 924 |
+
"loss": 1.064216136932373,
|
| 925 |
+
"step": 17920
|
| 926 |
+
},
|
| 927 |
+
{
|
| 928 |
+
"epoch": 0.8394993302849753,
|
| 929 |
+
"grad_norm": 0.3630645275115967,
|
| 930 |
+
"learning_rate": 6.542915085548828e-05,
|
| 931 |
+
"loss": 1.0645893812179565,
|
| 932 |
+
"step": 18176
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.8513232645143411,
|
| 936 |
+
"grad_norm": 0.42965033650398254,
|
| 937 |
+
"learning_rate": 5.6337900713373745e-05,
|
| 938 |
+
"loss": 1.0648459196090698,
|
| 939 |
+
"step": 18432
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.8513232645143411,
|
| 943 |
+
"eval_cos_loss": 0.2218880422039119,
|
| 944 |
+
"eval_loss": 1.005011047130306,
|
| 945 |
+
"eval_mse_loss": 0.7831230051985615,
|
| 946 |
+
"flow/cos_sim": 0.7781119878705778,
|
| 947 |
+
"flow/improvement_ratio": 0.4789293587684087,
|
| 948 |
+
"flow/mag_ratio_mean": 0.7607750161869885,
|
| 949 |
+
"flow/mag_ratio_std": 0.17580983801383407,
|
| 950 |
+
"step": 18432
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.8513232645143411,
|
| 954 |
+
"eval_cos_loss": 0.2218880422039119,
|
| 955 |
+
"eval_loss": 1.005011047130306,
|
| 956 |
+
"eval_mse_loss": 0.7831230051985615,
|
| 957 |
+
"eval_runtime": 37.6426,
|
| 958 |
+
"eval_samples_per_second": 743.652,
|
| 959 |
+
"eval_steps_per_second": 11.636,
|
| 960 |
+
"flow/cos_sim": 0.7781119878705778,
|
| 961 |
+
"flow/improvement_ratio": 0.4789293587684087,
|
| 962 |
+
"flow/mag_ratio_mean": 0.7607750161869885,
|
| 963 |
+
"flow/mag_ratio_std": 0.17580983801383407,
|
| 964 |
+
"step": 18432
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 0.8631471987437069,
|
| 968 |
+
"grad_norm": 0.1396653950214386,
|
| 969 |
+
"learning_rate": 4.78887625056757e-05,
|
| 970 |
+
"loss": 1.0648523569107056,
|
| 971 |
+
"step": 18688
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 0.8749711329730728,
|
| 975 |
+
"grad_norm": 0.4940822124481201,
|
| 976 |
+
"learning_rate": 4.009396466589682e-05,
|
| 977 |
+
"loss": 1.0595393180847168,
|
| 978 |
+
"step": 18944
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 0.8867950672024387,
|
| 982 |
+
"grad_norm": 0.4651182293891907,
|
| 983 |
+
"learning_rate": 3.296478860105229e-05,
|
| 984 |
+
"loss": 1.0656297206878662,
|
| 985 |
+
"step": 19200
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8986190014318045,
|
| 989 |
+
"grad_norm": 0.3197803795337677,
|
| 990 |
+
"learning_rate": 2.6511552364095358e-05,
|
| 991 |
+
"loss": 1.0682142972946167,
|
| 992 |
+
"step": 19456
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.8986190014318045,
|
| 996 |
+
"eval_cos_loss": 0.22248767602117095,
|
| 997 |
+
"eval_loss": 1.0076421013433638,
|
| 998 |
+
"eval_mse_loss": 0.7851544253902348,
|
| 999 |
+
"flow/cos_sim": 0.7775123350696477,
|
| 1000 |
+
"flow/improvement_ratio": 0.48520407645397534,
|
| 1001 |
+
"flow/mag_ratio_mean": 0.7587072534920418,
|
| 1002 |
+
"flow/mag_ratio_std": 0.17524336564214263,
|
| 1003 |
+
"step": 19456
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8986190014318045,
|
| 1007 |
+
"eval_cos_loss": 0.22248767602117095,
|
| 1008 |
+
"eval_loss": 1.0076421013433638,
|
| 1009 |
+
"eval_mse_loss": 0.7851544253902348,
|
| 1010 |
+
"eval_runtime": 37.6327,
|
| 1011 |
+
"eval_samples_per_second": 743.849,
|
| 1012 |
+
"eval_steps_per_second": 11.639,
|
| 1013 |
+
"flow/cos_sim": 0.7775123350696477,
|
| 1014 |
+
"flow/improvement_ratio": 0.48520407645397534,
|
| 1015 |
+
"flow/mag_ratio_mean": 0.7587072534920418,
|
| 1016 |
+
"flow/mag_ratio_std": 0.17524336564214263,
|
| 1017 |
+
"step": 19456
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9104429356611704,
|
| 1021 |
+
"grad_norm": 0.13613785803318024,
|
| 1022 |
+
"learning_rate": 2.074359572060447e-05,
|
| 1023 |
+
"loss": 1.0635831356048584,
|
| 1024 |
+
"step": 19712
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9222668698905362,
|
| 1028 |
+
"grad_norm": 0.1770780384540558,
|
| 1029 |
+
"learning_rate": 1.5669266631345104e-05,
|
| 1030 |
+
"loss": 1.062790036201477,
|
| 1031 |
+
"step": 19968
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9340908041199021,
|
| 1035 |
+
"grad_norm": 0.1503819078207016,
|
| 1036 |
+
"learning_rate": 1.1295909170270401e-05,
|
| 1037 |
+
"loss": 1.0686960220336914,
|
| 1038 |
+
"step": 20224
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.945914738349268,
|
| 1042 |
+
"grad_norm": 0.33569076657295227,
|
| 1043 |
+
"learning_rate": 7.642790173984836e-06,
|
| 1044 |
+
"loss": 1.0699816942214966,
|
| 1045 |
+
"step": 20480
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.945914738349268,
|
| 1049 |
+
"eval_cos_loss": 0.22027743961577,
|
| 1050 |
+
"eval_loss": 0.9999760730898,
|
| 1051 |
+
"eval_mse_loss": 0.779698631807005,
|
| 1052 |
+
"flow/cos_sim": 0.779722579401922,
|
| 1053 |
+
"flow/improvement_ratio": 0.4781283199378889,
|
| 1054 |
+
"flow/mag_ratio_mean": 0.7626172565434077,
|
| 1055 |
+
"flow/mag_ratio_std": 0.17652299278947317,
|
| 1056 |
+
"step": 20480
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.945914738349268,
|
| 1060 |
+
"eval_cos_loss": 0.22027743961577,
|
| 1061 |
+
"eval_loss": 0.9999760730898,
|
| 1062 |
+
"eval_mse_loss": 0.779698631807005,
|
| 1063 |
+
"eval_runtime": 37.3018,
|
| 1064 |
+
"eval_samples_per_second": 750.447,
|
| 1065 |
+
"eval_steps_per_second": 11.742,
|
| 1066 |
+
"flow/cos_sim": 0.779722579401922,
|
| 1067 |
+
"flow/improvement_ratio": 0.4781283199378889,
|
| 1068 |
+
"flow/mag_ratio_mean": 0.7626172565434077,
|
| 1069 |
+
"flow/mag_ratio_std": 0.17652299278947317,
|
| 1070 |
+
"step": 20480
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.9577386725786338,
|
| 1074 |
+
"grad_norm": 0.13765716552734375,
|
| 1075 |
+
"learning_rate": 4.686548355746001e-06,
|
| 1076 |
+
"loss": 1.0629913806915283,
|
| 1077 |
+
"step": 20736
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.9695626068079997,
|
| 1081 |
+
"grad_norm": 0.2119743674993515,
|
| 1082 |
+
"learning_rate": 2.447173448359541e-06,
|
| 1083 |
+
"loss": 1.0651545524597168,
|
| 1084 |
+
"step": 20992
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.9813865410373654,
|
| 1088 |
+
"grad_norm": 0.35947737097740173,
|
| 1089 |
+
"learning_rate": 9.279064980089457e-07,
|
| 1090 |
+
"loss": 1.0654728412628174,
|
| 1091 |
+
"step": 21248
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.9932104752667313,
|
| 1095 |
+
"grad_norm": 0.1823749542236328,
|
| 1096 |
+
"learning_rate": 1.3094633899163322e-07,
|
| 1097 |
+
"loss": 1.065650463104248,
|
| 1098 |
+
"step": 21504
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 0.9932104752667313,
|
| 1102 |
+
"eval_cos_loss": 0.2202744213821681,
|
| 1103 |
+
"eval_loss": 0.9997453873288141,
|
| 1104 |
+
"eval_mse_loss": 0.7794709677837756,
|
| 1105 |
+
"flow/cos_sim": 0.7797256029087659,
|
| 1106 |
+
"flow/improvement_ratio": 0.48657261542805796,
|
| 1107 |
+
"flow/mag_ratio_mean": 0.7617599999523599,
|
| 1108 |
+
"flow/mag_ratio_std": 0.17516998038308262,
|
| 1109 |
+
"step": 21504
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.9932104752667313,
|
| 1113 |
+
"eval_cos_loss": 0.2202744213821681,
|
| 1114 |
+
"eval_loss": 0.9997453873288141,
|
| 1115 |
+
"eval_mse_loss": 0.7794709677837756,
|
| 1116 |
+
"eval_runtime": 37.3277,
|
| 1117 |
+
"eval_samples_per_second": 749.926,
|
| 1118 |
+
"eval_steps_per_second": 11.734,
|
| 1119 |
+
"flow/cos_sim": 0.7797256029087659,
|
| 1120 |
+
"flow/improvement_ratio": 0.48657261542805796,
|
| 1121 |
+
"flow/mag_ratio_mean": 0.7617599999523599,
|
| 1122 |
+
"flow/mag_ratio_std": 0.17516998038308262,
|
| 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-b/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03a8c5803c6e5663005c9dc7a4c98c213c8267649ccec42581475db390cc8017
|
| 3 |
+
size 5137
|