Upload folder using huggingface_hub
Browse files- checkpoint-1000/optimizer.pt +3 -0
- checkpoint-1000/pytorch_model.bin +3 -0
- checkpoint-1000/rng_state_0.pth +3 -0
- checkpoint-1000/rng_state_1.pth +3 -0
- checkpoint-1000/rng_state_2.pth +3 -0
- checkpoint-1000/rng_state_3.pth +3 -0
- checkpoint-1000/rng_state_4.pth +3 -0
- checkpoint-1000/rng_state_5.pth +3 -0
- checkpoint-1000/rng_state_6.pth +3 -0
- checkpoint-1000/rng_state_7.pth +3 -0
- checkpoint-1000/scheduler.pt +3 -0
- checkpoint-1000/trainer_state.json +2012 -0
- checkpoint-1000/training_args.bin +3 -0
checkpoint-1000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73e0bebd95a2e6a37ad397ea656dc2d8e18c70e39c9590d7d105ea57d54f2f98
|
| 3 |
+
size 558664971
|
checkpoint-1000/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac6fa33968d4ed26b5e442c2cc888a6a86a98b5325a529d3bfea288be0093c5b
|
| 3 |
+
size 279336283
|
checkpoint-1000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b093dfe59b41efeb45cc3d628d3360abaa2303bbaa489081411faf431e52941d
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:450a0ac1645503c0b14fe9c37d77060cc76b1c9942dcfdd0e779cd526b2e98d9
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:938b37918eac9a4cbef3805f7d2abdcef094a334f848e73ac19fcdc39d38663a
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8b27a54988f134299ab296b95e8c1e63d476dffdba7c6f120f2076e8688f355
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_4.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d95f73d920296d5d9558e47894c5a2c0d649d7cb10a3b07a013d6bfbd3b8cf90
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_5.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70b945bb634c9daf4a00433296ecc5245b34a2b5f09017993b5f5f03b84dabea
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_6.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bfdd1fca0dace16a59c8592c531a70661218184bb0249c5862bbfb5ab0844fc9
|
| 3 |
+
size 16389
|
checkpoint-1000/rng_state_7.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d106363f9f1b0ff898c86d083a097bf22fd84de35e5670aa299504abcc99752a
|
| 3 |
+
size 16389
|
checkpoint-1000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1f10ac0b70778fb49c17b30c87ede08deb3a2bbd486b8970ed9515f2cf51c5eb
|
| 3 |
+
size 1529
|
checkpoint-1000/trainer_state.json
ADDED
|
@@ -0,0 +1,2012 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1000,
|
| 3 |
+
"best_metric": 3.2542863563067255,
|
| 4 |
+
"best_model_checkpoint": "/gpfs/scratch/guoh/DNAFM/output/gencode_human_12.8k_12800/GeneZip-4R_12.8K-100B/checkpoint-1000",
|
| 5 |
+
"epoch": 0.28368794326241137,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"comp/rl_weight": 0.03,
|
| 14 |
+
"comp/strictness": 1.0,
|
| 15 |
+
"epoch": 0.0028368794326241137,
|
| 16 |
+
"grad_norm": 1631.21875,
|
| 17 |
+
"loss": 88.5514,
|
| 18 |
+
"loss_ce": 61.01824188232422,
|
| 19 |
+
"loss_region": 0.08141374588012695,
|
| 20 |
+
"loss_total": 61.09965515136719,
|
| 21 |
+
"lr": 2.20454076850486e-05,
|
| 22 |
+
"router/avg_deficit_s0": 0.0,
|
| 23 |
+
"router/avg_deficit_s1": 0.0,
|
| 24 |
+
"router/selected_tokens_s0": 829.25,
|
| 25 |
+
"router/selected_tokens_s1": 174.53125,
|
| 26 |
+
"router/trigger_rate_s0": 0.0,
|
| 27 |
+
"router/trigger_rate_s1": 0.0,
|
| 28 |
+
"step": 10,
|
| 29 |
+
"tokens_trained": 0.03276544
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"comp/rl_weight": 0.03,
|
| 33 |
+
"comp/strictness": 1.0,
|
| 34 |
+
"epoch": 0.005673758865248227,
|
| 35 |
+
"grad_norm": 434.8985595703125,
|
| 36 |
+
"loss": 27.284,
|
| 37 |
+
"loss_ce": 15.031281471252441,
|
| 38 |
+
"loss_region": 0.13326193392276764,
|
| 39 |
+
"loss_total": 15.164543151855469,
|
| 40 |
+
"lr": 4.654030511288038e-05,
|
| 41 |
+
"router/avg_deficit_s0": 7.0,
|
| 42 |
+
"router/avg_deficit_s1": 7.0,
|
| 43 |
+
"router/selected_tokens_s0": 1.0,
|
| 44 |
+
"router/selected_tokens_s1": 1.0,
|
| 45 |
+
"router/trigger_rate_s0": 1.0,
|
| 46 |
+
"router/trigger_rate_s1": 1.0,
|
| 47 |
+
"step": 20,
|
| 48 |
+
"tokens_trained": 0.06553088
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"comp/rl_weight": 0.03,
|
| 52 |
+
"comp/strictness": 1.0,
|
| 53 |
+
"epoch": 0.00851063829787234,
|
| 54 |
+
"grad_norm": 583.2597045898438,
|
| 55 |
+
"loss": 10.2036,
|
| 56 |
+
"loss_ce": 7.8456501960754395,
|
| 57 |
+
"loss_region": 0.06420987844467163,
|
| 58 |
+
"loss_total": 7.909860134124756,
|
| 59 |
+
"lr": 7.103520254071216e-05,
|
| 60 |
+
"router/avg_deficit_s0": 0.0,
|
| 61 |
+
"router/avg_deficit_s1": 7.0,
|
| 62 |
+
"router/selected_tokens_s0": 99.875,
|
| 63 |
+
"router/selected_tokens_s1": 1.0,
|
| 64 |
+
"router/trigger_rate_s0": 0.0,
|
| 65 |
+
"router/trigger_rate_s1": 1.0,
|
| 66 |
+
"step": 30,
|
| 67 |
+
"tokens_trained": 0.09829632
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"comp/rl_weight": 0.03,
|
| 71 |
+
"comp/strictness": 1.0,
|
| 72 |
+
"epoch": 0.011347517730496455,
|
| 73 |
+
"grad_norm": 476.65606689453125,
|
| 74 |
+
"loss": 14.0147,
|
| 75 |
+
"loss_ce": 18.72601318359375,
|
| 76 |
+
"loss_region": 0.13252411782741547,
|
| 77 |
+
"loss_total": 18.858537673950195,
|
| 78 |
+
"lr": 9.553009996854394e-05,
|
| 79 |
+
"router/avg_deficit_s0": 7.0,
|
| 80 |
+
"router/avg_deficit_s1": 7.0,
|
| 81 |
+
"router/selected_tokens_s0": 1.0,
|
| 82 |
+
"router/selected_tokens_s1": 1.0,
|
| 83 |
+
"router/trigger_rate_s0": 1.0,
|
| 84 |
+
"router/trigger_rate_s1": 1.0,
|
| 85 |
+
"step": 40,
|
| 86 |
+
"tokens_trained": 0.13106176
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"comp/rl_weight": 0.03,
|
| 90 |
+
"comp/strictness": 1.0,
|
| 91 |
+
"epoch": 0.014184397163120567,
|
| 92 |
+
"grad_norm": 574.13232421875,
|
| 93 |
+
"loss": 11.397,
|
| 94 |
+
"loss_ce": 12.435722351074219,
|
| 95 |
+
"loss_region": 0.13453753292560577,
|
| 96 |
+
"loss_total": 12.570260047912598,
|
| 97 |
+
"lr": 0.00012002499739637572,
|
| 98 |
+
"router/avg_deficit_s0": 7.0,
|
| 99 |
+
"router/avg_deficit_s1": 7.0,
|
| 100 |
+
"router/selected_tokens_s0": 1.0,
|
| 101 |
+
"router/selected_tokens_s1": 1.0,
|
| 102 |
+
"router/trigger_rate_s0": 1.0,
|
| 103 |
+
"router/trigger_rate_s1": 1.0,
|
| 104 |
+
"step": 50,
|
| 105 |
+
"tokens_trained": 0.1638272
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"comp/rl_weight": 0.03,
|
| 109 |
+
"comp/strictness": 1.0,
|
| 110 |
+
"epoch": 0.01702127659574468,
|
| 111 |
+
"grad_norm": 572.3484497070312,
|
| 112 |
+
"loss": 11.4654,
|
| 113 |
+
"loss_ce": 15.20344066619873,
|
| 114 |
+
"loss_region": 0.1364777684211731,
|
| 115 |
+
"loss_total": 15.33991813659668,
|
| 116 |
+
"lr": 0.00014451989482420748,
|
| 117 |
+
"router/avg_deficit_s0": 7.0,
|
| 118 |
+
"router/avg_deficit_s1": 7.0,
|
| 119 |
+
"router/selected_tokens_s0": 1.0,
|
| 120 |
+
"router/selected_tokens_s1": 1.0,
|
| 121 |
+
"router/trigger_rate_s0": 1.0,
|
| 122 |
+
"router/trigger_rate_s1": 1.0,
|
| 123 |
+
"step": 60,
|
| 124 |
+
"tokens_trained": 0.19659264
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"comp/rl_weight": 0.03,
|
| 128 |
+
"comp/strictness": 1.0,
|
| 129 |
+
"epoch": 0.019858156028368795,
|
| 130 |
+
"grad_norm": 506.12725830078125,
|
| 131 |
+
"loss": 10.0282,
|
| 132 |
+
"loss_ce": 7.821432590484619,
|
| 133 |
+
"loss_region": 0.12491649389266968,
|
| 134 |
+
"loss_total": 7.946349143981934,
|
| 135 |
+
"lr": 0.00016901479225203927,
|
| 136 |
+
"router/avg_deficit_s0": 7.0,
|
| 137 |
+
"router/avg_deficit_s1": 7.0,
|
| 138 |
+
"router/selected_tokens_s0": 1.0,
|
| 139 |
+
"router/selected_tokens_s1": 1.0,
|
| 140 |
+
"router/trigger_rate_s0": 1.0,
|
| 141 |
+
"router/trigger_rate_s1": 1.0,
|
| 142 |
+
"step": 70,
|
| 143 |
+
"tokens_trained": 0.22935808
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"comp/rl_weight": 0.03,
|
| 147 |
+
"comp/strictness": 1.0,
|
| 148 |
+
"epoch": 0.02269503546099291,
|
| 149 |
+
"grad_norm": 327.83026123046875,
|
| 150 |
+
"loss": 9.0494,
|
| 151 |
+
"loss_ce": 13.648162841796875,
|
| 152 |
+
"loss_region": 0.1402742862701416,
|
| 153 |
+
"loss_total": 13.788436889648438,
|
| 154 |
+
"lr": 0.00019350968967987104,
|
| 155 |
+
"router/avg_deficit_s0": 7.0,
|
| 156 |
+
"router/avg_deficit_s1": 7.0,
|
| 157 |
+
"router/selected_tokens_s0": 1.0,
|
| 158 |
+
"router/selected_tokens_s1": 1.0,
|
| 159 |
+
"router/trigger_rate_s0": 1.0,
|
| 160 |
+
"router/trigger_rate_s1": 1.0,
|
| 161 |
+
"step": 80,
|
| 162 |
+
"tokens_trained": 0.26212352
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"comp/rl_weight": 0.03,
|
| 166 |
+
"comp/strictness": 1.0,
|
| 167 |
+
"epoch": 0.02553191489361702,
|
| 168 |
+
"grad_norm": 533.38037109375,
|
| 169 |
+
"loss": 10.6353,
|
| 170 |
+
"loss_ce": 11.381054878234863,
|
| 171 |
+
"loss_region": 0.12512199580669403,
|
| 172 |
+
"loss_total": 11.506176948547363,
|
| 173 |
+
"lr": 0.0002180045871077028,
|
| 174 |
+
"router/avg_deficit_s0": 7.0,
|
| 175 |
+
"router/avg_deficit_s1": 7.0,
|
| 176 |
+
"router/selected_tokens_s0": 1.0,
|
| 177 |
+
"router/selected_tokens_s1": 1.0,
|
| 178 |
+
"router/trigger_rate_s0": 1.0,
|
| 179 |
+
"router/trigger_rate_s1": 1.0,
|
| 180 |
+
"step": 90,
|
| 181 |
+
"tokens_trained": 0.29488896
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"comp/rl_weight": 0.03,
|
| 185 |
+
"comp/strictness": 1.0,
|
| 186 |
+
"epoch": 0.028368794326241134,
|
| 187 |
+
"grad_norm": 491.7203063964844,
|
| 188 |
+
"loss": 10.7556,
|
| 189 |
+
"loss_ce": 14.666193962097168,
|
| 190 |
+
"loss_region": 0.13100658357143402,
|
| 191 |
+
"loss_total": 14.797200202941895,
|
| 192 |
+
"lr": 0.00024249948453553463,
|
| 193 |
+
"router/avg_deficit_s0": 7.0,
|
| 194 |
+
"router/avg_deficit_s1": 7.0,
|
| 195 |
+
"router/selected_tokens_s0": 1.0,
|
| 196 |
+
"router/selected_tokens_s1": 1.0,
|
| 197 |
+
"router/trigger_rate_s0": 1.0,
|
| 198 |
+
"router/trigger_rate_s1": 1.0,
|
| 199 |
+
"step": 100,
|
| 200 |
+
"tokens_trained": 0.3276544
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"comp/rl_weight": 0.03,
|
| 204 |
+
"comp/strictness": 1.0,
|
| 205 |
+
"epoch": 0.031205673758865248,
|
| 206 |
+
"grad_norm": 412.38726806640625,
|
| 207 |
+
"loss": 8.1796,
|
| 208 |
+
"loss_ce": 7.042880535125732,
|
| 209 |
+
"loss_region": 0.12635233998298645,
|
| 210 |
+
"loss_total": 7.1692328453063965,
|
| 211 |
+
"lr": 0.00026699438196336637,
|
| 212 |
+
"router/avg_deficit_s0": 7.0,
|
| 213 |
+
"router/avg_deficit_s1": 7.0,
|
| 214 |
+
"router/selected_tokens_s0": 1.0,
|
| 215 |
+
"router/selected_tokens_s1": 1.0,
|
| 216 |
+
"router/trigger_rate_s0": 1.0,
|
| 217 |
+
"router/trigger_rate_s1": 1.0,
|
| 218 |
+
"step": 110,
|
| 219 |
+
"tokens_trained": 0.36041984
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"comp/rl_weight": 0.03,
|
| 223 |
+
"comp/strictness": 1.0,
|
| 224 |
+
"epoch": 0.03404255319148936,
|
| 225 |
+
"grad_norm": 380.94964599609375,
|
| 226 |
+
"loss": 10.906,
|
| 227 |
+
"loss_ce": 11.391271591186523,
|
| 228 |
+
"loss_region": 0.13119234144687653,
|
| 229 |
+
"loss_total": 11.52246379852295,
|
| 230 |
+
"lr": 0.00029148927939119814,
|
| 231 |
+
"router/avg_deficit_s0": 6.96875,
|
| 232 |
+
"router/avg_deficit_s1": 7.0,
|
| 233 |
+
"router/selected_tokens_s0": 1.03125,
|
| 234 |
+
"router/selected_tokens_s1": 1.0,
|
| 235 |
+
"router/trigger_rate_s0": 1.0,
|
| 236 |
+
"router/trigger_rate_s1": 1.0,
|
| 237 |
+
"step": 120,
|
| 238 |
+
"tokens_trained": 0.39318448
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"comp/rl_weight": 0.03,
|
| 242 |
+
"comp/strictness": 1.0,
|
| 243 |
+
"epoch": 0.03687943262411347,
|
| 244 |
+
"grad_norm": 199.84938049316406,
|
| 245 |
+
"loss": 9.1563,
|
| 246 |
+
"loss_ce": 3.2591464519500732,
|
| 247 |
+
"loss_region": 0.06539206206798553,
|
| 248 |
+
"loss_total": 3.324538469314575,
|
| 249 |
+
"lr": 0.00031598417681902996,
|
| 250 |
+
"router/avg_deficit_s0": 5.0,
|
| 251 |
+
"router/avg_deficit_s1": 7.0,
|
| 252 |
+
"router/selected_tokens_s0": 47.1875,
|
| 253 |
+
"router/selected_tokens_s1": 1.0,
|
| 254 |
+
"router/trigger_rate_s0": 0.03125,
|
| 255 |
+
"router/trigger_rate_s1": 1.0,
|
| 256 |
+
"step": 130,
|
| 257 |
+
"tokens_trained": 0.42594992
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"comp/rl_weight": 0.03,
|
| 261 |
+
"comp/strictness": 1.0,
|
| 262 |
+
"epoch": 0.03971631205673759,
|
| 263 |
+
"grad_norm": 296.4170227050781,
|
| 264 |
+
"loss": 12.2809,
|
| 265 |
+
"loss_ce": 11.650270462036133,
|
| 266 |
+
"loss_region": 0.09621138125658035,
|
| 267 |
+
"loss_total": 11.746481895446777,
|
| 268 |
+
"lr": 0.00034047907424686173,
|
| 269 |
+
"router/avg_deficit_s0": 0.0,
|
| 270 |
+
"router/avg_deficit_s1": 7.0,
|
| 271 |
+
"router/selected_tokens_s0": 2378.96875,
|
| 272 |
+
"router/selected_tokens_s1": 1.0,
|
| 273 |
+
"router/trigger_rate_s0": 0.0,
|
| 274 |
+
"router/trigger_rate_s1": 1.0,
|
| 275 |
+
"step": 140,
|
| 276 |
+
"tokens_trained": 0.45871536
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"comp/rl_weight": 0.03,
|
| 280 |
+
"comp/strictness": 1.0,
|
| 281 |
+
"epoch": 0.0425531914893617,
|
| 282 |
+
"grad_norm": 286.05914306640625,
|
| 283 |
+
"loss": 9.714,
|
| 284 |
+
"loss_ce": 13.918790817260742,
|
| 285 |
+
"loss_region": 0.1489456295967102,
|
| 286 |
+
"loss_total": 14.067736625671387,
|
| 287 |
+
"lr": 0.0003649739716746935,
|
| 288 |
+
"router/avg_deficit_s0": 0.0,
|
| 289 |
+
"router/avg_deficit_s1": 7.0,
|
| 290 |
+
"router/selected_tokens_s0": 4390.6875,
|
| 291 |
+
"router/selected_tokens_s1": 1.0,
|
| 292 |
+
"router/trigger_rate_s0": 0.0,
|
| 293 |
+
"router/trigger_rate_s1": 1.0,
|
| 294 |
+
"step": 150,
|
| 295 |
+
"tokens_trained": 0.4914808
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"comp/rl_weight": 0.03,
|
| 299 |
+
"comp/strictness": 1.0,
|
| 300 |
+
"epoch": 0.04539007092198582,
|
| 301 |
+
"grad_norm": 256.81048583984375,
|
| 302 |
+
"loss": 7.6435,
|
| 303 |
+
"loss_ce": 7.968850135803223,
|
| 304 |
+
"loss_region": 0.1997712403535843,
|
| 305 |
+
"loss_total": 8.168621063232422,
|
| 306 |
+
"lr": 0.00038946886910252526,
|
| 307 |
+
"router/avg_deficit_s0": 0.0,
|
| 308 |
+
"router/avg_deficit_s1": 7.0,
|
| 309 |
+
"router/selected_tokens_s0": 6185.8125,
|
| 310 |
+
"router/selected_tokens_s1": 1.0,
|
| 311 |
+
"router/trigger_rate_s0": 0.0,
|
| 312 |
+
"router/trigger_rate_s1": 1.0,
|
| 313 |
+
"step": 160,
|
| 314 |
+
"tokens_trained": 0.52424544
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"comp/rl_weight": 0.03,
|
| 318 |
+
"comp/strictness": 1.0,
|
| 319 |
+
"epoch": 0.04822695035460993,
|
| 320 |
+
"grad_norm": 223.51954650878906,
|
| 321 |
+
"loss": 5.6721,
|
| 322 |
+
"loss_ce": 7.768090724945068,
|
| 323 |
+
"loss_region": 0.10872498154640198,
|
| 324 |
+
"loss_total": 7.8768157958984375,
|
| 325 |
+
"lr": 0.0004139637665303571,
|
| 326 |
+
"router/avg_deficit_s0": 0.0,
|
| 327 |
+
"router/avg_deficit_s1": 7.0,
|
| 328 |
+
"router/selected_tokens_s0": 2846.78125,
|
| 329 |
+
"router/selected_tokens_s1": 1.0,
|
| 330 |
+
"router/trigger_rate_s0": 0.0,
|
| 331 |
+
"router/trigger_rate_s1": 1.0,
|
| 332 |
+
"step": 170,
|
| 333 |
+
"tokens_trained": 0.55701008
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"comp/rl_weight": 0.03,
|
| 337 |
+
"comp/strictness": 1.0,
|
| 338 |
+
"epoch": 0.05106382978723404,
|
| 339 |
+
"grad_norm": 218.51199340820312,
|
| 340 |
+
"loss": 5.0097,
|
| 341 |
+
"loss_ce": 4.9551897048950195,
|
| 342 |
+
"loss_region": 0.11287109553813934,
|
| 343 |
+
"loss_total": 5.068060874938965,
|
| 344 |
+
"lr": 0.0004384586639581888,
|
| 345 |
+
"router/avg_deficit_s0": 0.0,
|
| 346 |
+
"router/avg_deficit_s1": 7.0,
|
| 347 |
+
"router/selected_tokens_s0": 3095.65625,
|
| 348 |
+
"router/selected_tokens_s1": 1.0,
|
| 349 |
+
"router/trigger_rate_s0": 0.0,
|
| 350 |
+
"router/trigger_rate_s1": 1.0,
|
| 351 |
+
"step": 180,
|
| 352 |
+
"tokens_trained": 0.58977552
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"comp/rl_weight": 0.03,
|
| 356 |
+
"comp/strictness": 1.0,
|
| 357 |
+
"epoch": 0.05390070921985816,
|
| 358 |
+
"grad_norm": 111.65990447998047,
|
| 359 |
+
"loss": 4.0707,
|
| 360 |
+
"loss_ce": 1.8088655471801758,
|
| 361 |
+
"loss_region": 0.1163250282406807,
|
| 362 |
+
"loss_total": 1.9251905679702759,
|
| 363 |
+
"lr": 0.0004629535613860206,
|
| 364 |
+
"router/avg_deficit_s0": 0.0,
|
| 365 |
+
"router/avg_deficit_s1": 7.0,
|
| 366 |
+
"router/selected_tokens_s0": 3243.25,
|
| 367 |
+
"router/selected_tokens_s1": 1.0,
|
| 368 |
+
"router/trigger_rate_s0": 0.0,
|
| 369 |
+
"router/trigger_rate_s1": 1.0,
|
| 370 |
+
"step": 190,
|
| 371 |
+
"tokens_trained": 0.62254016
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"comp/rl_weight": 0.03,
|
| 375 |
+
"comp/strictness": 1.0,
|
| 376 |
+
"epoch": 0.05673758865248227,
|
| 377 |
+
"grad_norm": 202.0291748046875,
|
| 378 |
+
"loss": 3.8831,
|
| 379 |
+
"loss_ce": 4.9208855628967285,
|
| 380 |
+
"loss_region": 0.15217137336730957,
|
| 381 |
+
"loss_total": 5.073057174682617,
|
| 382 |
+
"lr": 0.00048744845881385244,
|
| 383 |
+
"router/avg_deficit_s0": 0.0,
|
| 384 |
+
"router/avg_deficit_s1": 7.0,
|
| 385 |
+
"router/selected_tokens_s0": 4667.375,
|
| 386 |
+
"router/selected_tokens_s1": 1.0,
|
| 387 |
+
"router/trigger_rate_s0": 0.0,
|
| 388 |
+
"router/trigger_rate_s1": 1.0,
|
| 389 |
+
"step": 200,
|
| 390 |
+
"tokens_trained": 0.6553056
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"comp/rl_weight": 0.03,
|
| 394 |
+
"comp/strictness": 1.0,
|
| 395 |
+
"epoch": 0.059574468085106386,
|
| 396 |
+
"grad_norm": 148.11932373046875,
|
| 397 |
+
"loss": 5.9535,
|
| 398 |
+
"loss_ce": 6.731753349304199,
|
| 399 |
+
"loss_region": 0.1624327301979065,
|
| 400 |
+
"loss_total": 6.894186019897461,
|
| 401 |
+
"lr": 0.0005119433562416841,
|
| 402 |
+
"router/avg_deficit_s0": 0.0,
|
| 403 |
+
"router/avg_deficit_s1": 7.0,
|
| 404 |
+
"router/selected_tokens_s0": 5666.625,
|
| 405 |
+
"router/selected_tokens_s1": 1.0,
|
| 406 |
+
"router/trigger_rate_s0": 0.0,
|
| 407 |
+
"router/trigger_rate_s1": 1.0,
|
| 408 |
+
"step": 210,
|
| 409 |
+
"tokens_trained": 0.68807104
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"comp/rl_weight": 0.03,
|
| 413 |
+
"comp/strictness": 1.0,
|
| 414 |
+
"epoch": 0.062411347517730496,
|
| 415 |
+
"grad_norm": 170.54896545410156,
|
| 416 |
+
"loss": 6.0952,
|
| 417 |
+
"loss_ce": 6.673608303070068,
|
| 418 |
+
"loss_region": 0.05599266663193703,
|
| 419 |
+
"loss_total": 6.72960090637207,
|
| 420 |
+
"lr": 0.0005364382536695159,
|
| 421 |
+
"router/avg_deficit_s0": 0.0,
|
| 422 |
+
"router/avg_deficit_s1": 7.0,
|
| 423 |
+
"router/selected_tokens_s0": 173.8125,
|
| 424 |
+
"router/selected_tokens_s1": 1.0,
|
| 425 |
+
"router/trigger_rate_s0": 0.0,
|
| 426 |
+
"router/trigger_rate_s1": 1.0,
|
| 427 |
+
"step": 220,
|
| 428 |
+
"tokens_trained": 0.72083648
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"comp/rl_weight": 0.03,
|
| 432 |
+
"comp/strictness": 1.0,
|
| 433 |
+
"epoch": 0.06524822695035461,
|
| 434 |
+
"grad_norm": 97.6951904296875,
|
| 435 |
+
"loss": 5.0635,
|
| 436 |
+
"loss_ce": 4.121210098266602,
|
| 437 |
+
"loss_region": 0.20329658687114716,
|
| 438 |
+
"loss_total": 4.324506759643555,
|
| 439 |
+
"lr": 0.0005609331510973477,
|
| 440 |
+
"router/avg_deficit_s0": 0.0,
|
| 441 |
+
"router/avg_deficit_s1": 7.0,
|
| 442 |
+
"router/selected_tokens_s0": 6635.03125,
|
| 443 |
+
"router/selected_tokens_s1": 1.0,
|
| 444 |
+
"router/trigger_rate_s0": 0.0,
|
| 445 |
+
"router/trigger_rate_s1": 1.0,
|
| 446 |
+
"step": 230,
|
| 447 |
+
"tokens_trained": 0.75360192
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"comp/rl_weight": 0.03,
|
| 451 |
+
"comp/strictness": 1.0,
|
| 452 |
+
"epoch": 0.06808510638297872,
|
| 453 |
+
"grad_norm": 138.48391723632812,
|
| 454 |
+
"loss": 3.7867,
|
| 455 |
+
"loss_ce": 4.167834281921387,
|
| 456 |
+
"loss_region": 0.06501111388206482,
|
| 457 |
+
"loss_total": 4.232845306396484,
|
| 458 |
+
"lr": 0.0005854280485251795,
|
| 459 |
+
"router/avg_deficit_s0": 0.0,
|
| 460 |
+
"router/avg_deficit_s1": 7.0,
|
| 461 |
+
"router/selected_tokens_s0": 1004.15625,
|
| 462 |
+
"router/selected_tokens_s1": 1.0,
|
| 463 |
+
"router/trigger_rate_s0": 0.0,
|
| 464 |
+
"router/trigger_rate_s1": 1.0,
|
| 465 |
+
"step": 240,
|
| 466 |
+
"tokens_trained": 0.78636736
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"comp/rl_weight": 0.03,
|
| 470 |
+
"comp/strictness": 1.0,
|
| 471 |
+
"epoch": 0.07092198581560284,
|
| 472 |
+
"grad_norm": 132.49085998535156,
|
| 473 |
+
"loss": 3.4818,
|
| 474 |
+
"loss_ce": 4.135341167449951,
|
| 475 |
+
"loss_region": 0.059110887348651886,
|
| 476 |
+
"loss_total": 4.194452285766602,
|
| 477 |
+
"lr": 0.0006099229459530113,
|
| 478 |
+
"router/avg_deficit_s0": 0.0,
|
| 479 |
+
"router/avg_deficit_s1": 7.0,
|
| 480 |
+
"router/selected_tokens_s0": 102.625,
|
| 481 |
+
"router/selected_tokens_s1": 1.0,
|
| 482 |
+
"router/trigger_rate_s0": 0.0,
|
| 483 |
+
"router/trigger_rate_s1": 1.0,
|
| 484 |
+
"step": 250,
|
| 485 |
+
"tokens_trained": 0.8191328
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"comp/rl_weight": 0.03,
|
| 489 |
+
"comp/strictness": 1.0,
|
| 490 |
+
"epoch": 0.07375886524822695,
|
| 491 |
+
"grad_norm": 114.51081848144531,
|
| 492 |
+
"loss": 3.9975,
|
| 493 |
+
"loss_ce": 4.087859630584717,
|
| 494 |
+
"loss_region": 0.06149622052907944,
|
| 495 |
+
"loss_total": 4.149355888366699,
|
| 496 |
+
"lr": 0.0006344178433808431,
|
| 497 |
+
"router/avg_deficit_s0": 0.0,
|
| 498 |
+
"router/avg_deficit_s1": 7.0,
|
| 499 |
+
"router/selected_tokens_s0": 696.5625,
|
| 500 |
+
"router/selected_tokens_s1": 1.0,
|
| 501 |
+
"router/trigger_rate_s0": 0.0,
|
| 502 |
+
"router/trigger_rate_s1": 1.0,
|
| 503 |
+
"step": 260,
|
| 504 |
+
"tokens_trained": 0.85189824
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"comp/rl_weight": 0.03,
|
| 508 |
+
"comp/strictness": 1.0,
|
| 509 |
+
"epoch": 0.07659574468085106,
|
| 510 |
+
"grad_norm": 97.07891845703125,
|
| 511 |
+
"loss": 4.1583,
|
| 512 |
+
"loss_ce": 5.67254638671875,
|
| 513 |
+
"loss_region": 0.08243336528539658,
|
| 514 |
+
"loss_total": 5.754979610443115,
|
| 515 |
+
"lr": 0.0006589127408086749,
|
| 516 |
+
"router/avg_deficit_s0": 0.0,
|
| 517 |
+
"router/avg_deficit_s1": 7.0,
|
| 518 |
+
"router/selected_tokens_s0": 1881.78125,
|
| 519 |
+
"router/selected_tokens_s1": 1.0,
|
| 520 |
+
"router/trigger_rate_s0": 0.0,
|
| 521 |
+
"router/trigger_rate_s1": 1.0,
|
| 522 |
+
"step": 270,
|
| 523 |
+
"tokens_trained": 0.88466288
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"comp/rl_weight": 0.03,
|
| 527 |
+
"comp/strictness": 1.0,
|
| 528 |
+
"epoch": 0.07943262411347518,
|
| 529 |
+
"grad_norm": 117.09831237792969,
|
| 530 |
+
"loss": 4.3864,
|
| 531 |
+
"loss_ce": 3.5589284896850586,
|
| 532 |
+
"loss_region": 0.08375530689954758,
|
| 533 |
+
"loss_total": 3.642683744430542,
|
| 534 |
+
"lr": 0.0006834076382365066,
|
| 535 |
+
"router/avg_deficit_s0": 0.0,
|
| 536 |
+
"router/avg_deficit_s1": 7.0,
|
| 537 |
+
"router/selected_tokens_s0": 2048.4375,
|
| 538 |
+
"router/selected_tokens_s1": 1.0,
|
| 539 |
+
"router/trigger_rate_s0": 0.0,
|
| 540 |
+
"router/trigger_rate_s1": 1.0,
|
| 541 |
+
"step": 280,
|
| 542 |
+
"tokens_trained": 0.91742832
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"comp/rl_weight": 0.03,
|
| 546 |
+
"comp/strictness": 1.0,
|
| 547 |
+
"epoch": 0.08226950354609928,
|
| 548 |
+
"grad_norm": 50.94511413574219,
|
| 549 |
+
"loss": 2.6175,
|
| 550 |
+
"loss_ce": 1.673315167427063,
|
| 551 |
+
"loss_region": 0.05678440257906914,
|
| 552 |
+
"loss_total": 1.7300995588302612,
|
| 553 |
+
"lr": 0.0007079025356643384,
|
| 554 |
+
"router/avg_deficit_s0": 0.0,
|
| 555 |
+
"router/avg_deficit_s1": 7.0,
|
| 556 |
+
"router/selected_tokens_s0": 916.25,
|
| 557 |
+
"router/selected_tokens_s1": 1.0,
|
| 558 |
+
"router/trigger_rate_s0": 0.0,
|
| 559 |
+
"router/trigger_rate_s1": 1.0,
|
| 560 |
+
"step": 290,
|
| 561 |
+
"tokens_trained": 0.95019376
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"comp/rl_weight": 0.03,
|
| 565 |
+
"comp/strictness": 1.0,
|
| 566 |
+
"epoch": 0.0851063829787234,
|
| 567 |
+
"grad_norm": 101.19383239746094,
|
| 568 |
+
"loss": 2.602,
|
| 569 |
+
"loss_ce": 2.7257542610168457,
|
| 570 |
+
"loss_region": 0.06887104362249374,
|
| 571 |
+
"loss_total": 2.7946252822875977,
|
| 572 |
+
"lr": 0.0007323974330921702,
|
| 573 |
+
"router/avg_deficit_s0": 0.0,
|
| 574 |
+
"router/avg_deficit_s1": 7.0,
|
| 575 |
+
"router/selected_tokens_s0": 1558.4375,
|
| 576 |
+
"router/selected_tokens_s1": 1.0,
|
| 577 |
+
"router/trigger_rate_s0": 0.0,
|
| 578 |
+
"router/trigger_rate_s1": 1.0,
|
| 579 |
+
"step": 300,
|
| 580 |
+
"tokens_trained": 0.9829592
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"comp/rl_weight": 0.03,
|
| 584 |
+
"comp/strictness": 1.0,
|
| 585 |
+
"epoch": 0.08794326241134752,
|
| 586 |
+
"grad_norm": 103.9866943359375,
|
| 587 |
+
"loss": 2.7451,
|
| 588 |
+
"loss_ce": 4.545335292816162,
|
| 589 |
+
"loss_region": 0.06630731374025345,
|
| 590 |
+
"loss_total": 4.611642837524414,
|
| 591 |
+
"lr": 0.000756892330520002,
|
| 592 |
+
"router/avg_deficit_s0": 0.0,
|
| 593 |
+
"router/avg_deficit_s1": 7.0,
|
| 594 |
+
"router/selected_tokens_s0": 1316.53125,
|
| 595 |
+
"router/selected_tokens_s1": 1.0,
|
| 596 |
+
"router/trigger_rate_s0": 0.0,
|
| 597 |
+
"router/trigger_rate_s1": 1.0,
|
| 598 |
+
"step": 310,
|
| 599 |
+
"tokens_trained": 1.01572464
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"comp/rl_weight": 0.03,
|
| 603 |
+
"comp/strictness": 1.0,
|
| 604 |
+
"epoch": 0.09078014184397164,
|
| 605 |
+
"grad_norm": 68.10435485839844,
|
| 606 |
+
"loss": 3.1077,
|
| 607 |
+
"loss_ce": 3.9576547145843506,
|
| 608 |
+
"loss_region": 0.06686295568943024,
|
| 609 |
+
"loss_total": 4.02451753616333,
|
| 610 |
+
"lr": 0.0007813872279478337,
|
| 611 |
+
"router/avg_deficit_s0": 0.0,
|
| 612 |
+
"router/avg_deficit_s1": 7.0,
|
| 613 |
+
"router/selected_tokens_s0": 1619.4375,
|
| 614 |
+
"router/selected_tokens_s1": 1.0,
|
| 615 |
+
"router/trigger_rate_s0": 0.0,
|
| 616 |
+
"router/trigger_rate_s1": 1.0,
|
| 617 |
+
"step": 320,
|
| 618 |
+
"tokens_trained": 1.04848992
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"comp/rl_weight": 0.03,
|
| 622 |
+
"comp/strictness": 1.0,
|
| 623 |
+
"epoch": 0.09361702127659574,
|
| 624 |
+
"grad_norm": 66.0996322631836,
|
| 625 |
+
"loss": 3.6268,
|
| 626 |
+
"loss_ce": 2.3814711570739746,
|
| 627 |
+
"loss_region": 0.06678226590156555,
|
| 628 |
+
"loss_total": 2.4482533931732178,
|
| 629 |
+
"lr": 0.0008058821253756655,
|
| 630 |
+
"router/avg_deficit_s0": 0.0,
|
| 631 |
+
"router/avg_deficit_s1": 7.0,
|
| 632 |
+
"router/selected_tokens_s0": 1653.875,
|
| 633 |
+
"router/selected_tokens_s1": 1.0,
|
| 634 |
+
"router/trigger_rate_s0": 0.0,
|
| 635 |
+
"router/trigger_rate_s1": 1.0,
|
| 636 |
+
"step": 330,
|
| 637 |
+
"tokens_trained": 1.08125456
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"comp/rl_weight": 0.03,
|
| 641 |
+
"comp/strictness": 1.0,
|
| 642 |
+
"epoch": 0.09645390070921986,
|
| 643 |
+
"grad_norm": 85.41226959228516,
|
| 644 |
+
"loss": 2.2121,
|
| 645 |
+
"loss_ce": 2.704575300216675,
|
| 646 |
+
"loss_region": 0.06229915842413902,
|
| 647 |
+
"loss_total": 2.7668745517730713,
|
| 648 |
+
"lr": 0.0008303770228034974,
|
| 649 |
+
"router/avg_deficit_s0": 0.0,
|
| 650 |
+
"router/avg_deficit_s1": 7.0,
|
| 651 |
+
"router/selected_tokens_s0": 1233.90625,
|
| 652 |
+
"router/selected_tokens_s1": 1.0,
|
| 653 |
+
"router/trigger_rate_s0": 0.0,
|
| 654 |
+
"router/trigger_rate_s1": 1.0,
|
| 655 |
+
"step": 340,
|
| 656 |
+
"tokens_trained": 1.11402
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"comp/rl_weight": 0.03,
|
| 660 |
+
"comp/strictness": 1.0,
|
| 661 |
+
"epoch": 0.09929078014184398,
|
| 662 |
+
"grad_norm": 65.33210754394531,
|
| 663 |
+
"loss": 3.0251,
|
| 664 |
+
"loss_ce": 2.7039618492126465,
|
| 665 |
+
"loss_region": 0.06328069418668747,
|
| 666 |
+
"loss_total": 2.767242431640625,
|
| 667 |
+
"lr": 0.0008548719202313291,
|
| 668 |
+
"router/avg_deficit_s0": 0.0,
|
| 669 |
+
"router/avg_deficit_s1": 7.0,
|
| 670 |
+
"router/selected_tokens_s0": 1259.8125,
|
| 671 |
+
"router/selected_tokens_s1": 1.0,
|
| 672 |
+
"router/trigger_rate_s0": 0.0,
|
| 673 |
+
"router/trigger_rate_s1": 1.0,
|
| 674 |
+
"step": 350,
|
| 675 |
+
"tokens_trained": 1.14678544
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"comp/rl_weight": 0.03,
|
| 679 |
+
"comp/strictness": 1.0,
|
| 680 |
+
"epoch": 0.10212765957446808,
|
| 681 |
+
"grad_norm": 90.26396179199219,
|
| 682 |
+
"loss": 3.2803,
|
| 683 |
+
"loss_ce": 3.2153475284576416,
|
| 684 |
+
"loss_region": 0.09488862007856369,
|
| 685 |
+
"loss_total": 3.3102362155914307,
|
| 686 |
+
"lr": 0.0008793668176591608,
|
| 687 |
+
"router/avg_deficit_s0": 0.0,
|
| 688 |
+
"router/avg_deficit_s1": 7.0,
|
| 689 |
+
"router/selected_tokens_s0": 2469.34375,
|
| 690 |
+
"router/selected_tokens_s1": 1.0,
|
| 691 |
+
"router/trigger_rate_s0": 0.0,
|
| 692 |
+
"router/trigger_rate_s1": 1.0,
|
| 693 |
+
"step": 360,
|
| 694 |
+
"tokens_trained": 1.17955088
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"comp/rl_weight": 0.03,
|
| 698 |
+
"comp/strictness": 1.0,
|
| 699 |
+
"epoch": 0.1049645390070922,
|
| 700 |
+
"grad_norm": 54.07102584838867,
|
| 701 |
+
"loss": 2.0596,
|
| 702 |
+
"loss_ce": 2.0937747955322266,
|
| 703 |
+
"loss_region": 0.06421175599098206,
|
| 704 |
+
"loss_total": 2.157986640930176,
|
| 705 |
+
"lr": 0.0009038617150869926,
|
| 706 |
+
"router/avg_deficit_s0": 0.0,
|
| 707 |
+
"router/avg_deficit_s1": 7.0,
|
| 708 |
+
"router/selected_tokens_s0": 1472.28125,
|
| 709 |
+
"router/selected_tokens_s1": 1.0,
|
| 710 |
+
"router/trigger_rate_s0": 0.0,
|
| 711 |
+
"router/trigger_rate_s1": 1.0,
|
| 712 |
+
"step": 370,
|
| 713 |
+
"tokens_trained": 1.21231632
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"comp/rl_weight": 0.03,
|
| 717 |
+
"comp/strictness": 1.0,
|
| 718 |
+
"epoch": 0.10780141843971631,
|
| 719 |
+
"grad_norm": 48.10196304321289,
|
| 720 |
+
"loss": 2.1087,
|
| 721 |
+
"loss_ce": 2.1116416454315186,
|
| 722 |
+
"loss_region": 0.05196397379040718,
|
| 723 |
+
"loss_total": 2.1636056900024414,
|
| 724 |
+
"lr": 0.0009283566125148244,
|
| 725 |
+
"router/avg_deficit_s0": 0.0,
|
| 726 |
+
"router/avg_deficit_s1": 7.0,
|
| 727 |
+
"router/selected_tokens_s0": 449.125,
|
| 728 |
+
"router/selected_tokens_s1": 1.0,
|
| 729 |
+
"router/trigger_rate_s0": 0.0,
|
| 730 |
+
"router/trigger_rate_s1": 1.0,
|
| 731 |
+
"step": 380,
|
| 732 |
+
"tokens_trained": 1.24508176
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"comp/rl_weight": 0.03,
|
| 736 |
+
"comp/strictness": 1.0,
|
| 737 |
+
"epoch": 0.11063829787234042,
|
| 738 |
+
"grad_norm": 54.023807525634766,
|
| 739 |
+
"loss": 2.2794,
|
| 740 |
+
"loss_ce": 2.8213136196136475,
|
| 741 |
+
"loss_region": 0.059752676635980606,
|
| 742 |
+
"loss_total": 2.88106632232666,
|
| 743 |
+
"lr": 0.0009528515099426562,
|
| 744 |
+
"router/avg_deficit_s0": 0.0,
|
| 745 |
+
"router/avg_deficit_s1": 7.0,
|
| 746 |
+
"router/selected_tokens_s0": 1112.1875,
|
| 747 |
+
"router/selected_tokens_s1": 1.0,
|
| 748 |
+
"router/trigger_rate_s0": 0.0,
|
| 749 |
+
"router/trigger_rate_s1": 1.0,
|
| 750 |
+
"step": 390,
|
| 751 |
+
"tokens_trained": 1.2778448
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"comp/rl_weight": 0.03,
|
| 755 |
+
"comp/strictness": 1.0,
|
| 756 |
+
"epoch": 0.11347517730496454,
|
| 757 |
+
"grad_norm": 89.19528198242188,
|
| 758 |
+
"loss": 2.4597,
|
| 759 |
+
"loss_ce": 3.5597033500671387,
|
| 760 |
+
"loss_region": 0.2661536931991577,
|
| 761 |
+
"loss_total": 3.825857162475586,
|
| 762 |
+
"lr": 0.000977346407370488,
|
| 763 |
+
"router/avg_deficit_s0": 7.0,
|
| 764 |
+
"router/avg_deficit_s1": 7.0,
|
| 765 |
+
"router/selected_tokens_s0": 1.0,
|
| 766 |
+
"router/selected_tokens_s1": 1.0,
|
| 767 |
+
"router/trigger_rate_s0": 1.0,
|
| 768 |
+
"router/trigger_rate_s1": 1.0,
|
| 769 |
+
"step": 400,
|
| 770 |
+
"tokens_trained": 1.31061024
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"comp/rl_weight": 0.03,
|
| 774 |
+
"comp/strictness": 1.0,
|
| 775 |
+
"epoch": 0.11631205673758865,
|
| 776 |
+
"grad_norm": 37.59690856933594,
|
| 777 |
+
"loss": 2.0656,
|
| 778 |
+
"loss_ce": 1.5812817811965942,
|
| 779 |
+
"loss_region": 0.07297738641500473,
|
| 780 |
+
"loss_total": 1.654259204864502,
|
| 781 |
+
"lr": 0.0010018413047983197,
|
| 782 |
+
"router/avg_deficit_s0": 0.0,
|
| 783 |
+
"router/avg_deficit_s1": 7.0,
|
| 784 |
+
"router/selected_tokens_s0": 1352.25,
|
| 785 |
+
"router/selected_tokens_s1": 1.0,
|
| 786 |
+
"router/trigger_rate_s0": 0.0,
|
| 787 |
+
"router/trigger_rate_s1": 1.0,
|
| 788 |
+
"step": 410,
|
| 789 |
+
"tokens_trained": 1.34337568
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"comp/rl_weight": 0.03,
|
| 793 |
+
"comp/strictness": 1.0,
|
| 794 |
+
"epoch": 0.11914893617021277,
|
| 795 |
+
"grad_norm": 31.402502059936523,
|
| 796 |
+
"loss": 1.7999,
|
| 797 |
+
"loss_ce": 1.4559017419815063,
|
| 798 |
+
"loss_region": 0.07083795219659805,
|
| 799 |
+
"loss_total": 1.5267397165298462,
|
| 800 |
+
"lr": 0.0010263362022261515,
|
| 801 |
+
"router/avg_deficit_s0": 0.0,
|
| 802 |
+
"router/avg_deficit_s1": 7.0,
|
| 803 |
+
"router/selected_tokens_s0": 1392.46875,
|
| 804 |
+
"router/selected_tokens_s1": 1.0,
|
| 805 |
+
"router/trigger_rate_s0": 0.0,
|
| 806 |
+
"router/trigger_rate_s1": 1.0,
|
| 807 |
+
"step": 420,
|
| 808 |
+
"tokens_trained": 1.376141104
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"comp/rl_weight": 0.03,
|
| 812 |
+
"comp/strictness": 1.0,
|
| 813 |
+
"epoch": 0.12198581560283688,
|
| 814 |
+
"grad_norm": 32.30119705200195,
|
| 815 |
+
"loss": 1.7606,
|
| 816 |
+
"loss_ce": 1.619798183441162,
|
| 817 |
+
"loss_region": 0.05854247137904167,
|
| 818 |
+
"loss_total": 1.6783406734466553,
|
| 819 |
+
"lr": 0.0010508310996539833,
|
| 820 |
+
"router/avg_deficit_s0": 0.0,
|
| 821 |
+
"router/avg_deficit_s1": 7.0,
|
| 822 |
+
"router/selected_tokens_s0": 107.46875,
|
| 823 |
+
"router/selected_tokens_s1": 1.0,
|
| 824 |
+
"router/trigger_rate_s0": 0.0,
|
| 825 |
+
"router/trigger_rate_s1": 1.0,
|
| 826 |
+
"step": 430,
|
| 827 |
+
"tokens_trained": 1.408906544
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"comp/rl_weight": 0.03,
|
| 831 |
+
"comp/strictness": 1.0,
|
| 832 |
+
"epoch": 0.12482269503546099,
|
| 833 |
+
"grad_norm": 46.65761947631836,
|
| 834 |
+
"loss": 1.9924,
|
| 835 |
+
"loss_ce": 2.4070842266082764,
|
| 836 |
+
"loss_region": 0.06459856033325195,
|
| 837 |
+
"loss_total": 2.4716827869415283,
|
| 838 |
+
"lr": 0.0010753259970818151,
|
| 839 |
+
"router/avg_deficit_s0": 0.0,
|
| 840 |
+
"router/avg_deficit_s1": 7.0,
|
| 841 |
+
"router/selected_tokens_s0": 1131.90625,
|
| 842 |
+
"router/selected_tokens_s1": 1.0,
|
| 843 |
+
"router/trigger_rate_s0": 0.0,
|
| 844 |
+
"router/trigger_rate_s1": 1.0,
|
| 845 |
+
"step": 440,
|
| 846 |
+
"tokens_trained": 1.441671984
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"comp/rl_weight": 0.03,
|
| 850 |
+
"comp/strictness": 1.0,
|
| 851 |
+
"epoch": 0.1276595744680851,
|
| 852 |
+
"grad_norm": 30.38531494140625,
|
| 853 |
+
"loss": 1.8704,
|
| 854 |
+
"loss_ce": 1.4392523765563965,
|
| 855 |
+
"loss_region": 0.055338773876428604,
|
| 856 |
+
"loss_total": 1.4945911169052124,
|
| 857 |
+
"lr": 0.001099820894509647,
|
| 858 |
+
"router/avg_deficit_s0": 0.0,
|
| 859 |
+
"router/avg_deficit_s1": 7.0,
|
| 860 |
+
"router/selected_tokens_s0": 631.5625,
|
| 861 |
+
"router/selected_tokens_s1": 1.0,
|
| 862 |
+
"router/trigger_rate_s0": 0.0,
|
| 863 |
+
"router/trigger_rate_s1": 1.0,
|
| 864 |
+
"step": 450,
|
| 865 |
+
"tokens_trained": 1.474437424
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"comp/rl_weight": 0.03,
|
| 869 |
+
"comp/strictness": 1.0,
|
| 870 |
+
"epoch": 0.13049645390070921,
|
| 871 |
+
"grad_norm": 37.19720458984375,
|
| 872 |
+
"loss": 1.7504,
|
| 873 |
+
"loss_ce": 1.5982601642608643,
|
| 874 |
+
"loss_region": 0.07102015614509583,
|
| 875 |
+
"loss_total": 1.6692802906036377,
|
| 876 |
+
"lr": 0.0011243157919374788,
|
| 877 |
+
"router/avg_deficit_s0": 0.0,
|
| 878 |
+
"router/avg_deficit_s1": 0.0,
|
| 879 |
+
"router/selected_tokens_s0": 929.3125,
|
| 880 |
+
"router/selected_tokens_s1": 48.28125,
|
| 881 |
+
"router/trigger_rate_s0": 0.0,
|
| 882 |
+
"router/trigger_rate_s1": 0.0,
|
| 883 |
+
"step": 460,
|
| 884 |
+
"tokens_trained": 1.507202864
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"comp/rl_weight": 0.03,
|
| 888 |
+
"comp/strictness": 1.0,
|
| 889 |
+
"epoch": 0.13333333333333333,
|
| 890 |
+
"grad_norm": 30.36027717590332,
|
| 891 |
+
"loss": 1.7281,
|
| 892 |
+
"loss_ce": 1.4914231300354004,
|
| 893 |
+
"loss_region": 0.05361942574381828,
|
| 894 |
+
"loss_total": 1.5450425148010254,
|
| 895 |
+
"lr": 0.0011488106893653104,
|
| 896 |
+
"router/avg_deficit_s0": 0.0,
|
| 897 |
+
"router/avg_deficit_s1": 7.0,
|
| 898 |
+
"router/selected_tokens_s0": 377.4375,
|
| 899 |
+
"router/selected_tokens_s1": 1.0,
|
| 900 |
+
"router/trigger_rate_s0": 0.0,
|
| 901 |
+
"router/trigger_rate_s1": 1.0,
|
| 902 |
+
"step": 470,
|
| 903 |
+
"tokens_trained": 1.539968304
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"comp/rl_weight": 0.03,
|
| 907 |
+
"comp/strictness": 1.0,
|
| 908 |
+
"epoch": 0.13617021276595745,
|
| 909 |
+
"grad_norm": 12.751656532287598,
|
| 910 |
+
"loss": 1.502,
|
| 911 |
+
"loss_ce": 1.3392599821090698,
|
| 912 |
+
"loss_region": 0.05824059993028641,
|
| 913 |
+
"loss_total": 1.3975006341934204,
|
| 914 |
+
"lr": 0.0011733055867931422,
|
| 915 |
+
"router/avg_deficit_s0": 0.0,
|
| 916 |
+
"router/avg_deficit_s1": 7.0,
|
| 917 |
+
"router/selected_tokens_s0": 779.8125,
|
| 918 |
+
"router/selected_tokens_s1": 1.0,
|
| 919 |
+
"router/trigger_rate_s0": 0.0,
|
| 920 |
+
"router/trigger_rate_s1": 1.0,
|
| 921 |
+
"step": 480,
|
| 922 |
+
"tokens_trained": 1.572733744
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"comp/rl_weight": 0.03,
|
| 926 |
+
"comp/strictness": 1.0,
|
| 927 |
+
"epoch": 0.13900709219858157,
|
| 928 |
+
"grad_norm": 30.712886810302734,
|
| 929 |
+
"loss": 1.5265,
|
| 930 |
+
"loss_ce": 1.441983699798584,
|
| 931 |
+
"loss_region": 0.05336770787835121,
|
| 932 |
+
"loss_total": 1.4953514337539673,
|
| 933 |
+
"lr": 0.001197800484220974,
|
| 934 |
+
"router/avg_deficit_s0": 0.0,
|
| 935 |
+
"router/avg_deficit_s1": 7.0,
|
| 936 |
+
"router/selected_tokens_s0": 630.71875,
|
| 937 |
+
"router/selected_tokens_s1": 1.0,
|
| 938 |
+
"router/trigger_rate_s0": 0.0,
|
| 939 |
+
"router/trigger_rate_s1": 1.0,
|
| 940 |
+
"step": 490,
|
| 941 |
+
"tokens_trained": 1.605499184
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"comp/rl_weight": 0.03,
|
| 945 |
+
"comp/strictness": 1.0,
|
| 946 |
+
"epoch": 0.14184397163120568,
|
| 947 |
+
"grad_norm": 23.531269073486328,
|
| 948 |
+
"loss": 1.6006,
|
| 949 |
+
"loss_ce": 1.3859856128692627,
|
| 950 |
+
"loss_region": 0.05309389531612396,
|
| 951 |
+
"loss_total": 1.4390795230865479,
|
| 952 |
+
"lr": 0.0012222953816488059,
|
| 953 |
+
"router/avg_deficit_s0": 0.0,
|
| 954 |
+
"router/avg_deficit_s1": 7.0,
|
| 955 |
+
"router/selected_tokens_s0": 355.1875,
|
| 956 |
+
"router/selected_tokens_s1": 1.0,
|
| 957 |
+
"router/trigger_rate_s0": 0.0,
|
| 958 |
+
"router/trigger_rate_s1": 1.0,
|
| 959 |
+
"step": 500,
|
| 960 |
+
"tokens_trained": 1.638264624
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.14184397163120568,
|
| 964 |
+
"eval_ppl": 5.639791369232795,
|
| 965 |
+
"eval_runtime": 0.6154,
|
| 966 |
+
"step": 500,
|
| 967 |
+
"tokens_trained": 1.638264624
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.14184397163120568,
|
| 971 |
+
"eval_F": 0.0006250488319399953,
|
| 972 |
+
"eval_F_cds": 0.000417908784442649,
|
| 973 |
+
"eval_F_exon": 0.0010869565217391304,
|
| 974 |
+
"eval_F_intron": 0.0006231804525964301,
|
| 975 |
+
"eval_F_nig": 0.0007117661631871394,
|
| 976 |
+
"eval_F_promoter": 0.00046730843692003687,
|
| 977 |
+
"eval_F_utr": 0.0007665184730952016,
|
| 978 |
+
"eval_G": 0.018260951935942262,
|
| 979 |
+
"eval_G_cds": 0.017367320490764216,
|
| 980 |
+
"eval_G_exon": 0.01836432857789855,
|
| 981 |
+
"eval_G_intron": 0.018287128753366702,
|
| 982 |
+
"eval_G_nig": 0.019185528203734217,
|
| 983 |
+
"eval_G_promoter": 0.016835490531997278,
|
| 984 |
+
"eval_G_utr": 0.017702484637692268,
|
| 985 |
+
"eval_avg_bp_per_token": 1599.875,
|
| 986 |
+
"eval_bp_per_token/cds": 2392.866666666667,
|
| 987 |
+
"eval_bp_per_token/exon": 920.0,
|
| 988 |
+
"eval_bp_per_token/intron": 1604.6716417910447,
|
| 989 |
+
"eval_bp_per_token/nig": 1404.9558011049724,
|
| 990 |
+
"eval_bp_per_token/promoter": 2139.9142857142856,
|
| 991 |
+
"eval_bp_per_token/utr": 1304.6,
|
| 992 |
+
"eval_ppl_cds": 6.105901405038029,
|
| 993 |
+
"eval_ppl_exon": 5.801578519769911,
|
| 994 |
+
"eval_ppl_intron": 5.612093447668663,
|
| 995 |
+
"eval_ppl_nig": 5.620627240075904,
|
| 996 |
+
"eval_ppl_promoter": 5.688336002196773,
|
| 997 |
+
"eval_ppl_utr": 5.695504465687492,
|
| 998 |
+
"step": 500,
|
| 999 |
+
"tokens_trained": 1.638264624
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"comp/rl_weight": 0.03,
|
| 1003 |
+
"comp/strictness": 1.0,
|
| 1004 |
+
"epoch": 0.14468085106382977,
|
| 1005 |
+
"grad_norm": 9.532892227172852,
|
| 1006 |
+
"loss": 1.5975,
|
| 1007 |
+
"loss_ce": 1.3092715740203857,
|
| 1008 |
+
"loss_region": 0.05799846723675728,
|
| 1009 |
+
"loss_total": 1.3672699928283691,
|
| 1010 |
+
"lr": 0.0012243786686061229,
|
| 1011 |
+
"router/avg_deficit_s0": 0.0,
|
| 1012 |
+
"router/avg_deficit_s1": 7.0,
|
| 1013 |
+
"router/selected_tokens_s0": 946.25,
|
| 1014 |
+
"router/selected_tokens_s1": 1.0,
|
| 1015 |
+
"router/trigger_rate_s0": 0.0,
|
| 1016 |
+
"router/trigger_rate_s1": 1.0,
|
| 1017 |
+
"step": 510,
|
| 1018 |
+
"tokens_trained": 1.671030064
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"comp/rl_weight": 0.03,
|
| 1022 |
+
"comp/strictness": 1.0,
|
| 1023 |
+
"epoch": 0.1475177304964539,
|
| 1024 |
+
"grad_norm": 18.65001678466797,
|
| 1025 |
+
"loss": 1.6168,
|
| 1026 |
+
"loss_ce": 1.343700885772705,
|
| 1027 |
+
"loss_region": 0.05204938352108002,
|
| 1028 |
+
"loss_total": 1.3957502841949463,
|
| 1029 |
+
"lr": 0.0012239717766222718,
|
| 1030 |
+
"router/avg_deficit_s0": 0.0,
|
| 1031 |
+
"router/avg_deficit_s1": 7.0,
|
| 1032 |
+
"router/selected_tokens_s0": 541.875,
|
| 1033 |
+
"router/selected_tokens_s1": 1.0,
|
| 1034 |
+
"router/trigger_rate_s0": 0.0,
|
| 1035 |
+
"router/trigger_rate_s1": 1.0,
|
| 1036 |
+
"step": 520,
|
| 1037 |
+
"tokens_trained": 1.703795504
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"comp/rl_weight": 0.03,
|
| 1041 |
+
"comp/strictness": 1.0,
|
| 1042 |
+
"epoch": 0.150354609929078,
|
| 1043 |
+
"grad_norm": 34.98784255981445,
|
| 1044 |
+
"loss": 1.6703,
|
| 1045 |
+
"loss_ce": 1.456465721130371,
|
| 1046 |
+
"loss_region": 0.05966515839099884,
|
| 1047 |
+
"loss_total": 1.5161309242248535,
|
| 1048 |
+
"lr": 0.001223564884638421,
|
| 1049 |
+
"router/avg_deficit_s0": 0.0,
|
| 1050 |
+
"router/avg_deficit_s1": 7.0,
|
| 1051 |
+
"router/selected_tokens_s0": 980.71875,
|
| 1052 |
+
"router/selected_tokens_s1": 1.0,
|
| 1053 |
+
"router/trigger_rate_s0": 0.0,
|
| 1054 |
+
"router/trigger_rate_s1": 1.0,
|
| 1055 |
+
"step": 530,
|
| 1056 |
+
"tokens_trained": 1.736560944
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"comp/rl_weight": 0.03,
|
| 1060 |
+
"comp/strictness": 1.0,
|
| 1061 |
+
"epoch": 0.15319148936170213,
|
| 1062 |
+
"grad_norm": 27.90123176574707,
|
| 1063 |
+
"loss": 1.5678,
|
| 1064 |
+
"loss_ce": 1.4636605978012085,
|
| 1065 |
+
"loss_region": 0.055262189358472824,
|
| 1066 |
+
"loss_total": 1.5189228057861328,
|
| 1067 |
+
"lr": 0.00122315799265457,
|
| 1068 |
+
"router/avg_deficit_s0": 0.0,
|
| 1069 |
+
"router/avg_deficit_s1": 7.0,
|
| 1070 |
+
"router/selected_tokens_s0": 515.9375,
|
| 1071 |
+
"router/selected_tokens_s1": 1.0,
|
| 1072 |
+
"router/trigger_rate_s0": 0.0,
|
| 1073 |
+
"router/trigger_rate_s1": 1.0,
|
| 1074 |
+
"step": 540,
|
| 1075 |
+
"tokens_trained": 1.769326384
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"comp/rl_weight": 0.03,
|
| 1079 |
+
"comp/strictness": 1.0,
|
| 1080 |
+
"epoch": 0.15602836879432624,
|
| 1081 |
+
"grad_norm": 36.412662506103516,
|
| 1082 |
+
"loss": 1.617,
|
| 1083 |
+
"loss_ce": 1.481337070465088,
|
| 1084 |
+
"loss_region": 0.057017069309949875,
|
| 1085 |
+
"loss_total": 1.5383541584014893,
|
| 1086 |
+
"lr": 0.001222751100670719,
|
| 1087 |
+
"router/avg_deficit_s0": 0.0,
|
| 1088 |
+
"router/avg_deficit_s1": 7.0,
|
| 1089 |
+
"router/selected_tokens_s0": 653.09375,
|
| 1090 |
+
"router/selected_tokens_s1": 1.0,
|
| 1091 |
+
"router/trigger_rate_s0": 0.0,
|
| 1092 |
+
"router/trigger_rate_s1": 1.0,
|
| 1093 |
+
"step": 550,
|
| 1094 |
+
"tokens_trained": 1.802091024
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"comp/rl_weight": 0.03,
|
| 1098 |
+
"comp/strictness": 1.0,
|
| 1099 |
+
"epoch": 0.15886524822695036,
|
| 1100 |
+
"grad_norm": 33.760154724121094,
|
| 1101 |
+
"loss": 1.6967,
|
| 1102 |
+
"loss_ce": 1.6242401599884033,
|
| 1103 |
+
"loss_region": 0.05576888844370842,
|
| 1104 |
+
"loss_total": 1.6800090074539185,
|
| 1105 |
+
"lr": 0.001222344208686868,
|
| 1106 |
+
"router/avg_deficit_s0": 0.0,
|
| 1107 |
+
"router/avg_deficit_s1": 6.46875,
|
| 1108 |
+
"router/selected_tokens_s0": 594.71875,
|
| 1109 |
+
"router/selected_tokens_s1": 1.53125,
|
| 1110 |
+
"router/trigger_rate_s0": 0.0,
|
| 1111 |
+
"router/trigger_rate_s1": 1.0,
|
| 1112 |
+
"step": 560,
|
| 1113 |
+
"tokens_trained": 1.834856464
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"comp/rl_weight": 0.03,
|
| 1117 |
+
"comp/strictness": 1.0,
|
| 1118 |
+
"epoch": 0.16170212765957448,
|
| 1119 |
+
"grad_norm": 30.55908966064453,
|
| 1120 |
+
"loss": 1.7022,
|
| 1121 |
+
"loss_ce": 1.4764841794967651,
|
| 1122 |
+
"loss_region": 0.053381841629743576,
|
| 1123 |
+
"loss_total": 1.5298659801483154,
|
| 1124 |
+
"lr": 0.0012219373167030169,
|
| 1125 |
+
"router/avg_deficit_s0": 0.0,
|
| 1126 |
+
"router/avg_deficit_s1": 7.0,
|
| 1127 |
+
"router/selected_tokens_s0": 584.40625,
|
| 1128 |
+
"router/selected_tokens_s1": 1.0,
|
| 1129 |
+
"router/trigger_rate_s0": 0.0,
|
| 1130 |
+
"router/trigger_rate_s1": 1.0,
|
| 1131 |
+
"step": 570,
|
| 1132 |
+
"tokens_trained": 1.867621904
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"comp/rl_weight": 0.03,
|
| 1136 |
+
"comp/strictness": 1.0,
|
| 1137 |
+
"epoch": 0.16453900709219857,
|
| 1138 |
+
"grad_norm": 24.999780654907227,
|
| 1139 |
+
"loss": 1.4763,
|
| 1140 |
+
"loss_ce": 1.4031097888946533,
|
| 1141 |
+
"loss_region": 0.05390815809369087,
|
| 1142 |
+
"loss_total": 1.4570178985595703,
|
| 1143 |
+
"lr": 0.0012215304247191658,
|
| 1144 |
+
"router/avg_deficit_s0": 0.0,
|
| 1145 |
+
"router/avg_deficit_s1": 7.0,
|
| 1146 |
+
"router/selected_tokens_s0": 671.75,
|
| 1147 |
+
"router/selected_tokens_s1": 1.0,
|
| 1148 |
+
"router/trigger_rate_s0": 0.0,
|
| 1149 |
+
"router/trigger_rate_s1": 1.0,
|
| 1150 |
+
"step": 580,
|
| 1151 |
+
"tokens_trained": 1.900385744
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"comp/rl_weight": 0.03,
|
| 1155 |
+
"comp/strictness": 1.0,
|
| 1156 |
+
"epoch": 0.1673758865248227,
|
| 1157 |
+
"grad_norm": 34.729034423828125,
|
| 1158 |
+
"loss": 1.4617,
|
| 1159 |
+
"loss_ce": 1.487898349761963,
|
| 1160 |
+
"loss_region": 0.10076552629470825,
|
| 1161 |
+
"loss_total": 1.5886638164520264,
|
| 1162 |
+
"lr": 0.0012211235327353148,
|
| 1163 |
+
"router/avg_deficit_s0": 0.0,
|
| 1164 |
+
"router/avg_deficit_s1": 5.5,
|
| 1165 |
+
"router/selected_tokens_s0": 540.78125,
|
| 1166 |
+
"router/selected_tokens_s1": 105.71875,
|
| 1167 |
+
"router/trigger_rate_s0": 0.0,
|
| 1168 |
+
"router/trigger_rate_s1": 0.125,
|
| 1169 |
+
"step": 590,
|
| 1170 |
+
"tokens_trained": 1.933151184
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"comp/rl_weight": 0.03,
|
| 1174 |
+
"comp/strictness": 1.0,
|
| 1175 |
+
"epoch": 0.1702127659574468,
|
| 1176 |
+
"grad_norm": 23.622325897216797,
|
| 1177 |
+
"loss": 1.4675,
|
| 1178 |
+
"loss_ce": 1.3250795602798462,
|
| 1179 |
+
"loss_region": 0.0658915713429451,
|
| 1180 |
+
"loss_total": 1.3909711837768555,
|
| 1181 |
+
"lr": 0.0012207166407514638,
|
| 1182 |
+
"router/avg_deficit_s0": 0.0,
|
| 1183 |
+
"router/avg_deficit_s1": 5.0,
|
| 1184 |
+
"router/selected_tokens_s0": 539.9375,
|
| 1185 |
+
"router/selected_tokens_s1": 32.5,
|
| 1186 |
+
"router/trigger_rate_s0": 0.0,
|
| 1187 |
+
"router/trigger_rate_s1": 0.3125,
|
| 1188 |
+
"step": 600,
|
| 1189 |
+
"tokens_trained": 1.965916624
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"comp/rl_weight": 0.03,
|
| 1193 |
+
"comp/strictness": 1.0,
|
| 1194 |
+
"epoch": 0.17304964539007092,
|
| 1195 |
+
"grad_norm": 20.5329532623291,
|
| 1196 |
+
"loss": 1.3919,
|
| 1197 |
+
"loss_ce": 1.31636381149292,
|
| 1198 |
+
"loss_region": 0.06174392253160477,
|
| 1199 |
+
"loss_total": 1.3781077861785889,
|
| 1200 |
+
"lr": 0.0012203097487676127,
|
| 1201 |
+
"router/avg_deficit_s0": 0.0,
|
| 1202 |
+
"router/avg_deficit_s1": 6.84375,
|
| 1203 |
+
"router/selected_tokens_s0": 1102.0,
|
| 1204 |
+
"router/selected_tokens_s1": 1.15625,
|
| 1205 |
+
"router/trigger_rate_s0": 0.0,
|
| 1206 |
+
"router/trigger_rate_s1": 1.0,
|
| 1207 |
+
"step": 610,
|
| 1208 |
+
"tokens_trained": 1.998682064
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"comp/rl_weight": 0.03,
|
| 1212 |
+
"comp/strictness": 1.0,
|
| 1213 |
+
"epoch": 0.17588652482269504,
|
| 1214 |
+
"grad_norm": 25.521442413330078,
|
| 1215 |
+
"loss": 1.371,
|
| 1216 |
+
"loss_ce": 1.3530433177947998,
|
| 1217 |
+
"loss_region": 0.055061180144548416,
|
| 1218 |
+
"loss_total": 1.4081045389175415,
|
| 1219 |
+
"lr": 0.0012199028567837617,
|
| 1220 |
+
"router/avg_deficit_s0": 0.0,
|
| 1221 |
+
"router/avg_deficit_s1": 6.96875,
|
| 1222 |
+
"router/selected_tokens_s0": 267.96875,
|
| 1223 |
+
"router/selected_tokens_s1": 1.03125,
|
| 1224 |
+
"router/trigger_rate_s0": 0.0,
|
| 1225 |
+
"router/trigger_rate_s1": 1.0,
|
| 1226 |
+
"step": 620,
|
| 1227 |
+
"tokens_trained": 2.031447504
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"comp/rl_weight": 0.03,
|
| 1231 |
+
"comp/strictness": 1.0,
|
| 1232 |
+
"epoch": 0.17872340425531916,
|
| 1233 |
+
"grad_norm": 19.09522819519043,
|
| 1234 |
+
"loss": 1.3462,
|
| 1235 |
+
"loss_ce": 1.2844027280807495,
|
| 1236 |
+
"loss_region": 0.05663185566663742,
|
| 1237 |
+
"loss_total": 1.3410345315933228,
|
| 1238 |
+
"lr": 0.0012194959647999107,
|
| 1239 |
+
"router/avg_deficit_s0": 0.0,
|
| 1240 |
+
"router/avg_deficit_s1": 5.741935483870968,
|
| 1241 |
+
"router/selected_tokens_s0": 761.0625,
|
| 1242 |
+
"router/selected_tokens_s1": 2.4375,
|
| 1243 |
+
"router/trigger_rate_s0": 0.0,
|
| 1244 |
+
"router/trigger_rate_s1": 0.96875,
|
| 1245 |
+
"step": 630,
|
| 1246 |
+
"tokens_trained": 2.064212944
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"comp/rl_weight": 0.03,
|
| 1250 |
+
"comp/strictness": 1.0,
|
| 1251 |
+
"epoch": 0.18156028368794327,
|
| 1252 |
+
"grad_norm": 13.425186157226562,
|
| 1253 |
+
"loss": 1.3235,
|
| 1254 |
+
"loss_ce": 1.2834771871566772,
|
| 1255 |
+
"loss_region": 0.057796087116003036,
|
| 1256 |
+
"loss_total": 1.341273307800293,
|
| 1257 |
+
"lr": 0.0012190890728160596,
|
| 1258 |
+
"router/avg_deficit_s0": 0.0,
|
| 1259 |
+
"router/avg_deficit_s1": 4.384615384615385,
|
| 1260 |
+
"router/selected_tokens_s0": 976.3125,
|
| 1261 |
+
"router/selected_tokens_s1": 5.75,
|
| 1262 |
+
"router/trigger_rate_s0": 0.0,
|
| 1263 |
+
"router/trigger_rate_s1": 0.8125,
|
| 1264 |
+
"step": 640,
|
| 1265 |
+
"tokens_trained": 2.096975976
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"comp/rl_weight": 0.03,
|
| 1269 |
+
"comp/strictness": 1.0,
|
| 1270 |
+
"epoch": 0.18439716312056736,
|
| 1271 |
+
"grad_norm": 30.06955909729004,
|
| 1272 |
+
"loss": 1.3965,
|
| 1273 |
+
"loss_ce": 1.4326735734939575,
|
| 1274 |
+
"loss_region": 0.051951903849840164,
|
| 1275 |
+
"loss_total": 1.4846254587173462,
|
| 1276 |
+
"lr": 0.0012186821808322086,
|
| 1277 |
+
"router/avg_deficit_s0": 0.0,
|
| 1278 |
+
"router/avg_deficit_s1": 6.96875,
|
| 1279 |
+
"router/selected_tokens_s0": 366.9375,
|
| 1280 |
+
"router/selected_tokens_s1": 1.03125,
|
| 1281 |
+
"router/trigger_rate_s0": 0.0,
|
| 1282 |
+
"router/trigger_rate_s1": 1.0,
|
| 1283 |
+
"step": 650,
|
| 1284 |
+
"tokens_trained": 2.129741416
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"comp/rl_weight": 0.03,
|
| 1288 |
+
"comp/strictness": 1.0,
|
| 1289 |
+
"epoch": 0.18723404255319148,
|
| 1290 |
+
"grad_norm": 26.722410202026367,
|
| 1291 |
+
"loss": 1.3767,
|
| 1292 |
+
"loss_ce": 1.3280962705612183,
|
| 1293 |
+
"loss_region": 0.05729561671614647,
|
| 1294 |
+
"loss_total": 1.3853918313980103,
|
| 1295 |
+
"lr": 0.0012182752888483576,
|
| 1296 |
+
"router/avg_deficit_s0": 0.0,
|
| 1297 |
+
"router/avg_deficit_s1": 7.0,
|
| 1298 |
+
"router/selected_tokens_s0": 938.03125,
|
| 1299 |
+
"router/selected_tokens_s1": 1.0,
|
| 1300 |
+
"router/trigger_rate_s0": 0.0,
|
| 1301 |
+
"router/trigger_rate_s1": 1.0,
|
| 1302 |
+
"step": 660,
|
| 1303 |
+
"tokens_trained": 2.162506856
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"comp/rl_weight": 0.03,
|
| 1307 |
+
"comp/strictness": 1.0,
|
| 1308 |
+
"epoch": 0.1900709219858156,
|
| 1309 |
+
"grad_norm": 13.13896656036377,
|
| 1310 |
+
"loss": 1.3329,
|
| 1311 |
+
"loss_ce": 1.2802703380584717,
|
| 1312 |
+
"loss_region": 0.05178743973374367,
|
| 1313 |
+
"loss_total": 1.3320578336715698,
|
| 1314 |
+
"lr": 0.0012178683968645065,
|
| 1315 |
+
"router/avg_deficit_s0": 0.0,
|
| 1316 |
+
"router/avg_deficit_s1": 7.0,
|
| 1317 |
+
"router/selected_tokens_s0": 558.8125,
|
| 1318 |
+
"router/selected_tokens_s1": 1.0,
|
| 1319 |
+
"router/trigger_rate_s0": 0.0,
|
| 1320 |
+
"router/trigger_rate_s1": 1.0,
|
| 1321 |
+
"step": 670,
|
| 1322 |
+
"tokens_trained": 2.195272296
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"comp/rl_weight": 0.03,
|
| 1326 |
+
"comp/strictness": 1.0,
|
| 1327 |
+
"epoch": 0.19290780141843972,
|
| 1328 |
+
"grad_norm": 13.104703903198242,
|
| 1329 |
+
"loss": 1.3014,
|
| 1330 |
+
"loss_ce": 1.2344958782196045,
|
| 1331 |
+
"loss_region": 0.05391368269920349,
|
| 1332 |
+
"loss_total": 1.2884095907211304,
|
| 1333 |
+
"lr": 0.0012174615048806555,
|
| 1334 |
+
"router/avg_deficit_s0": 0.0,
|
| 1335 |
+
"router/avg_deficit_s1": 7.0,
|
| 1336 |
+
"router/selected_tokens_s0": 328.4375,
|
| 1337 |
+
"router/selected_tokens_s1": 1.0,
|
| 1338 |
+
"router/trigger_rate_s0": 0.0,
|
| 1339 |
+
"router/trigger_rate_s1": 1.0,
|
| 1340 |
+
"step": 680,
|
| 1341 |
+
"tokens_trained": 2.228037736
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"comp/rl_weight": 0.03,
|
| 1345 |
+
"comp/strictness": 1.0,
|
| 1346 |
+
"epoch": 0.19574468085106383,
|
| 1347 |
+
"grad_norm": 15.946929931640625,
|
| 1348 |
+
"loss": 1.2954,
|
| 1349 |
+
"loss_ce": 1.2488086223602295,
|
| 1350 |
+
"loss_region": 0.05314250662922859,
|
| 1351 |
+
"loss_total": 1.3019511699676514,
|
| 1352 |
+
"lr": 0.0012170546128968045,
|
| 1353 |
+
"router/avg_deficit_s0": 0.0,
|
| 1354 |
+
"router/avg_deficit_s1": 7.0,
|
| 1355 |
+
"router/selected_tokens_s0": 583.71875,
|
| 1356 |
+
"router/selected_tokens_s1": 1.0,
|
| 1357 |
+
"router/trigger_rate_s0": 0.0,
|
| 1358 |
+
"router/trigger_rate_s1": 1.0,
|
| 1359 |
+
"step": 690,
|
| 1360 |
+
"tokens_trained": 2.260803176
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"comp/rl_weight": 0.03,
|
| 1364 |
+
"comp/strictness": 1.0,
|
| 1365 |
+
"epoch": 0.19858156028368795,
|
| 1366 |
+
"grad_norm": 36.70180130004883,
|
| 1367 |
+
"loss": 1.3394,
|
| 1368 |
+
"loss_ce": 1.3971507549285889,
|
| 1369 |
+
"loss_region": 0.06877385824918747,
|
| 1370 |
+
"loss_total": 1.465924620628357,
|
| 1371 |
+
"lr": 0.0012166477209129534,
|
| 1372 |
+
"router/avg_deficit_s0": 0.0,
|
| 1373 |
+
"router/avg_deficit_s1": 6.96875,
|
| 1374 |
+
"router/selected_tokens_s0": 1593.3125,
|
| 1375 |
+
"router/selected_tokens_s1": 1.03125,
|
| 1376 |
+
"router/trigger_rate_s0": 0.0,
|
| 1377 |
+
"router/trigger_rate_s1": 1.0,
|
| 1378 |
+
"step": 700,
|
| 1379 |
+
"tokens_trained": 2.293568616
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"comp/rl_weight": 0.03,
|
| 1383 |
+
"comp/strictness": 1.0,
|
| 1384 |
+
"epoch": 0.20141843971631207,
|
| 1385 |
+
"grad_norm": 9.902652740478516,
|
| 1386 |
+
"loss": 1.4144,
|
| 1387 |
+
"loss_ce": 1.231865644454956,
|
| 1388 |
+
"loss_region": 0.23015622794628143,
|
| 1389 |
+
"loss_total": 1.462021827697754,
|
| 1390 |
+
"lr": 0.0012162408289291026,
|
| 1391 |
+
"router/avg_deficit_s0": 6.333333333333333,
|
| 1392 |
+
"router/avg_deficit_s1": 7.0,
|
| 1393 |
+
"router/selected_tokens_s0": 36.34375,
|
| 1394 |
+
"router/selected_tokens_s1": 1.0,
|
| 1395 |
+
"router/trigger_rate_s0": 0.1875,
|
| 1396 |
+
"router/trigger_rate_s1": 1.0,
|
| 1397 |
+
"step": 710,
|
| 1398 |
+
"tokens_trained": 2.326334056
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"comp/rl_weight": 0.03,
|
| 1402 |
+
"comp/strictness": 1.0,
|
| 1403 |
+
"epoch": 0.20425531914893616,
|
| 1404 |
+
"grad_norm": 16.028406143188477,
|
| 1405 |
+
"loss": 1.3595,
|
| 1406 |
+
"loss_ce": 1.2959296703338623,
|
| 1407 |
+
"loss_region": 0.05357291176915169,
|
| 1408 |
+
"loss_total": 1.3495025634765625,
|
| 1409 |
+
"lr": 0.0012158339369452516,
|
| 1410 |
+
"router/avg_deficit_s0": 0.0,
|
| 1411 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1412 |
+
"router/selected_tokens_s0": 248.96875,
|
| 1413 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1414 |
+
"router/trigger_rate_s0": 0.0,
|
| 1415 |
+
"router/trigger_rate_s1": 1.0,
|
| 1416 |
+
"step": 720,
|
| 1417 |
+
"tokens_trained": 2.359099496
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"comp/rl_weight": 0.03,
|
| 1421 |
+
"comp/strictness": 1.0,
|
| 1422 |
+
"epoch": 0.20709219858156028,
|
| 1423 |
+
"grad_norm": 12.643508911132812,
|
| 1424 |
+
"loss": 1.347,
|
| 1425 |
+
"loss_ce": 1.2560234069824219,
|
| 1426 |
+
"loss_region": 0.05382990837097168,
|
| 1427 |
+
"loss_total": 1.3098533153533936,
|
| 1428 |
+
"lr": 0.0012154270449614005,
|
| 1429 |
+
"router/avg_deficit_s0": 0.0,
|
| 1430 |
+
"router/avg_deficit_s1": 7.0,
|
| 1431 |
+
"router/selected_tokens_s0": 212.34375,
|
| 1432 |
+
"router/selected_tokens_s1": 1.0,
|
| 1433 |
+
"router/trigger_rate_s0": 0.0,
|
| 1434 |
+
"router/trigger_rate_s1": 1.0,
|
| 1435 |
+
"step": 730,
|
| 1436 |
+
"tokens_trained": 2.391864136
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"comp/rl_weight": 0.03,
|
| 1440 |
+
"comp/strictness": 1.0,
|
| 1441 |
+
"epoch": 0.2099290780141844,
|
| 1442 |
+
"grad_norm": 6.794911861419678,
|
| 1443 |
+
"loss": 1.2899,
|
| 1444 |
+
"loss_ce": 1.2041139602661133,
|
| 1445 |
+
"loss_region": 0.05598007142543793,
|
| 1446 |
+
"loss_total": 1.2600940465927124,
|
| 1447 |
+
"lr": 0.0012150201529775495,
|
| 1448 |
+
"router/avg_deficit_s0": 0.0,
|
| 1449 |
+
"router/avg_deficit_s1": 7.0,
|
| 1450 |
+
"router/selected_tokens_s0": 436.8125,
|
| 1451 |
+
"router/selected_tokens_s1": 1.0,
|
| 1452 |
+
"router/trigger_rate_s0": 0.0,
|
| 1453 |
+
"router/trigger_rate_s1": 1.0,
|
| 1454 |
+
"step": 740,
|
| 1455 |
+
"tokens_trained": 2.424628776
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"comp/rl_weight": 0.03,
|
| 1459 |
+
"comp/strictness": 1.0,
|
| 1460 |
+
"epoch": 0.2127659574468085,
|
| 1461 |
+
"grad_norm": 8.187490463256836,
|
| 1462 |
+
"loss": 1.2837,
|
| 1463 |
+
"loss_ce": 1.21809720993042,
|
| 1464 |
+
"loss_region": 0.05316738411784172,
|
| 1465 |
+
"loss_total": 1.2712645530700684,
|
| 1466 |
+
"lr": 0.0012146132609936982,
|
| 1467 |
+
"router/avg_deficit_s0": 0.0,
|
| 1468 |
+
"router/avg_deficit_s1": 7.0,
|
| 1469 |
+
"router/selected_tokens_s0": 324.8125,
|
| 1470 |
+
"router/selected_tokens_s1": 1.0,
|
| 1471 |
+
"router/trigger_rate_s0": 0.0,
|
| 1472 |
+
"router/trigger_rate_s1": 1.0,
|
| 1473 |
+
"step": 750,
|
| 1474 |
+
"tokens_trained": 2.457394216
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"comp/rl_weight": 0.03,
|
| 1478 |
+
"comp/strictness": 1.0,
|
| 1479 |
+
"epoch": 0.21560283687943263,
|
| 1480 |
+
"grad_norm": 6.42191743850708,
|
| 1481 |
+
"loss": 1.2935,
|
| 1482 |
+
"loss_ce": 1.232576847076416,
|
| 1483 |
+
"loss_region": 0.053723808377981186,
|
| 1484 |
+
"loss_total": 1.2863006591796875,
|
| 1485 |
+
"lr": 0.0012142063690098472,
|
| 1486 |
+
"router/avg_deficit_s0": 0.0,
|
| 1487 |
+
"router/avg_deficit_s1": 6.5625,
|
| 1488 |
+
"router/selected_tokens_s0": 596.46875,
|
| 1489 |
+
"router/selected_tokens_s1": 1.4375,
|
| 1490 |
+
"router/trigger_rate_s0": 0.0,
|
| 1491 |
+
"router/trigger_rate_s1": 1.0,
|
| 1492 |
+
"step": 760,
|
| 1493 |
+
"tokens_trained": 2.490159656
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"comp/rl_weight": 0.03,
|
| 1497 |
+
"comp/strictness": 1.0,
|
| 1498 |
+
"epoch": 0.21843971631205675,
|
| 1499 |
+
"grad_norm": 18.290868759155273,
|
| 1500 |
+
"loss": 1.303,
|
| 1501 |
+
"loss_ce": 1.2971560955047607,
|
| 1502 |
+
"loss_region": 0.0527898408472538,
|
| 1503 |
+
"loss_total": 1.3499459028244019,
|
| 1504 |
+
"lr": 0.0012137994770259962,
|
| 1505 |
+
"router/avg_deficit_s0": 0.0,
|
| 1506 |
+
"router/avg_deficit_s1": 7.0,
|
| 1507 |
+
"router/selected_tokens_s0": 454.40625,
|
| 1508 |
+
"router/selected_tokens_s1": 1.0,
|
| 1509 |
+
"router/trigger_rate_s0": 0.0,
|
| 1510 |
+
"router/trigger_rate_s1": 1.0,
|
| 1511 |
+
"step": 770,
|
| 1512 |
+
"tokens_trained": 2.522925096
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"comp/rl_weight": 0.03,
|
| 1516 |
+
"comp/strictness": 1.0,
|
| 1517 |
+
"epoch": 0.22127659574468084,
|
| 1518 |
+
"grad_norm": 15.430254936218262,
|
| 1519 |
+
"loss": 1.313,
|
| 1520 |
+
"loss_ce": 1.2602447271347046,
|
| 1521 |
+
"loss_region": 0.052465811371803284,
|
| 1522 |
+
"loss_total": 1.3127105236053467,
|
| 1523 |
+
"lr": 0.0012133925850421454,
|
| 1524 |
+
"router/avg_deficit_s0": 0.0,
|
| 1525 |
+
"router/avg_deficit_s1": 7.0,
|
| 1526 |
+
"router/selected_tokens_s0": 236.28125,
|
| 1527 |
+
"router/selected_tokens_s1": 1.0,
|
| 1528 |
+
"router/trigger_rate_s0": 0.0,
|
| 1529 |
+
"router/trigger_rate_s1": 1.0,
|
| 1530 |
+
"step": 780,
|
| 1531 |
+
"tokens_trained": 2.555690536
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"comp/rl_weight": 0.03,
|
| 1535 |
+
"comp/strictness": 1.0,
|
| 1536 |
+
"epoch": 0.22411347517730495,
|
| 1537 |
+
"grad_norm": 7.031615734100342,
|
| 1538 |
+
"loss": 1.301,
|
| 1539 |
+
"loss_ce": 1.185632586479187,
|
| 1540 |
+
"loss_region": 0.06820279359817505,
|
| 1541 |
+
"loss_total": 1.2538354396820068,
|
| 1542 |
+
"lr": 0.0012129856930582943,
|
| 1543 |
+
"router/avg_deficit_s0": 0.0,
|
| 1544 |
+
"router/avg_deficit_s1": 7.0,
|
| 1545 |
+
"router/selected_tokens_s0": 1199.375,
|
| 1546 |
+
"router/selected_tokens_s1": 1.0,
|
| 1547 |
+
"router/trigger_rate_s0": 0.0,
|
| 1548 |
+
"router/trigger_rate_s1": 1.0,
|
| 1549 |
+
"step": 790,
|
| 1550 |
+
"tokens_trained": 2.588455976
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"comp/rl_weight": 0.03,
|
| 1554 |
+
"comp/strictness": 1.0,
|
| 1555 |
+
"epoch": 0.22695035460992907,
|
| 1556 |
+
"grad_norm": 5.123201847076416,
|
| 1557 |
+
"loss": 1.2839,
|
| 1558 |
+
"loss_ce": 1.2030867338180542,
|
| 1559 |
+
"loss_region": 0.06425300985574722,
|
| 1560 |
+
"loss_total": 1.2673397064208984,
|
| 1561 |
+
"lr": 0.0012125788010744433,
|
| 1562 |
+
"router/avg_deficit_s0": 0.0,
|
| 1563 |
+
"router/avg_deficit_s1": 7.0,
|
| 1564 |
+
"router/selected_tokens_s0": 1046.25,
|
| 1565 |
+
"router/selected_tokens_s1": 1.0,
|
| 1566 |
+
"router/trigger_rate_s0": 0.0,
|
| 1567 |
+
"router/trigger_rate_s1": 1.0,
|
| 1568 |
+
"step": 800,
|
| 1569 |
+
"tokens_trained": 2.621221416
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"comp/rl_weight": 0.03,
|
| 1573 |
+
"comp/strictness": 1.0,
|
| 1574 |
+
"epoch": 0.2297872340425532,
|
| 1575 |
+
"grad_norm": 8.874433517456055,
|
| 1576 |
+
"loss": 1.282,
|
| 1577 |
+
"loss_ce": 1.2380104064941406,
|
| 1578 |
+
"loss_region": 0.05459459125995636,
|
| 1579 |
+
"loss_total": 1.2926050424575806,
|
| 1580 |
+
"lr": 0.0012121719090905923,
|
| 1581 |
+
"router/avg_deficit_s0": 0.0,
|
| 1582 |
+
"router/avg_deficit_s1": 5.448275862068965,
|
| 1583 |
+
"router/selected_tokens_s0": 606.28125,
|
| 1584 |
+
"router/selected_tokens_s1": 3.125,
|
| 1585 |
+
"router/trigger_rate_s0": 0.0,
|
| 1586 |
+
"router/trigger_rate_s1": 0.90625,
|
| 1587 |
+
"step": 810,
|
| 1588 |
+
"tokens_trained": 2.653986856
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"comp/rl_weight": 0.03,
|
| 1592 |
+
"comp/strictness": 1.0,
|
| 1593 |
+
"epoch": 0.2326241134751773,
|
| 1594 |
+
"grad_norm": 3.3697080612182617,
|
| 1595 |
+
"loss": 1.279,
|
| 1596 |
+
"loss_ce": 1.2127931118011475,
|
| 1597 |
+
"loss_region": 0.05367187783122063,
|
| 1598 |
+
"loss_total": 1.2664649486541748,
|
| 1599 |
+
"lr": 0.0012117650171067412,
|
| 1600 |
+
"router/avg_deficit_s0": 0.0,
|
| 1601 |
+
"router/avg_deficit_s1": 5.387096774193548,
|
| 1602 |
+
"router/selected_tokens_s0": 550.875,
|
| 1603 |
+
"router/selected_tokens_s1": 2.78125,
|
| 1604 |
+
"router/trigger_rate_s0": 0.0,
|
| 1605 |
+
"router/trigger_rate_s1": 0.96875,
|
| 1606 |
+
"step": 820,
|
| 1607 |
+
"tokens_trained": 2.686752296
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"comp/rl_weight": 0.03,
|
| 1611 |
+
"comp/strictness": 1.0,
|
| 1612 |
+
"epoch": 0.23546099290780143,
|
| 1613 |
+
"grad_norm": 18.332046508789062,
|
| 1614 |
+
"loss": 1.3407,
|
| 1615 |
+
"loss_ce": 1.366118311882019,
|
| 1616 |
+
"loss_region": 0.0635012835264206,
|
| 1617 |
+
"loss_total": 1.429619550704956,
|
| 1618 |
+
"lr": 0.0012113581251228902,
|
| 1619 |
+
"router/avg_deficit_s0": 0.0,
|
| 1620 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1621 |
+
"router/selected_tokens_s0": 1186.375,
|
| 1622 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1623 |
+
"router/trigger_rate_s0": 0.0,
|
| 1624 |
+
"router/trigger_rate_s1": 1.0,
|
| 1625 |
+
"step": 830,
|
| 1626 |
+
"tokens_trained": 2.719517736
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"comp/rl_weight": 0.03,
|
| 1630 |
+
"comp/strictness": 1.0,
|
| 1631 |
+
"epoch": 0.23829787234042554,
|
| 1632 |
+
"grad_norm": 9.937188148498535,
|
| 1633 |
+
"loss": 1.3486,
|
| 1634 |
+
"loss_ce": 1.243201494216919,
|
| 1635 |
+
"loss_region": 0.09887048602104187,
|
| 1636 |
+
"loss_total": 1.3420720100402832,
|
| 1637 |
+
"lr": 0.0012109512331390391,
|
| 1638 |
+
"router/avg_deficit_s0": 0.0,
|
| 1639 |
+
"router/avg_deficit_s1": 6.96875,
|
| 1640 |
+
"router/selected_tokens_s0": 67.71875,
|
| 1641 |
+
"router/selected_tokens_s1": 1.03125,
|
| 1642 |
+
"router/trigger_rate_s0": 0.0,
|
| 1643 |
+
"router/trigger_rate_s1": 1.0,
|
| 1644 |
+
"step": 840,
|
| 1645 |
+
"tokens_trained": 2.752283176
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"comp/rl_weight": 0.03,
|
| 1649 |
+
"comp/strictness": 1.0,
|
| 1650 |
+
"epoch": 0.24113475177304963,
|
| 1651 |
+
"grad_norm": 6.917110443115234,
|
| 1652 |
+
"loss": 1.311,
|
| 1653 |
+
"loss_ce": 1.2267051935195923,
|
| 1654 |
+
"loss_region": 0.0637051984667778,
|
| 1655 |
+
"loss_total": 1.2904103994369507,
|
| 1656 |
+
"lr": 0.0012105443411551881,
|
| 1657 |
+
"router/avg_deficit_s0": 0.0,
|
| 1658 |
+
"router/avg_deficit_s1": 6.28125,
|
| 1659 |
+
"router/selected_tokens_s0": 1052.40625,
|
| 1660 |
+
"router/selected_tokens_s1": 1.71875,
|
| 1661 |
+
"router/trigger_rate_s0": 0.0,
|
| 1662 |
+
"router/trigger_rate_s1": 1.0,
|
| 1663 |
+
"step": 850,
|
| 1664 |
+
"tokens_trained": 2.7850478
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"comp/rl_weight": 0.03,
|
| 1668 |
+
"comp/strictness": 1.0,
|
| 1669 |
+
"epoch": 0.24397163120567375,
|
| 1670 |
+
"grad_norm": 6.469630718231201,
|
| 1671 |
+
"loss": 1.269,
|
| 1672 |
+
"loss_ce": 1.1876814365386963,
|
| 1673 |
+
"loss_region": 0.0656404048204422,
|
| 1674 |
+
"loss_total": 1.253321886062622,
|
| 1675 |
+
"lr": 0.001210137449171337,
|
| 1676 |
+
"router/avg_deficit_s0": 0.0,
|
| 1677 |
+
"router/avg_deficit_s1": 6.516129032258065,
|
| 1678 |
+
"router/selected_tokens_s0": 1256.46875,
|
| 1679 |
+
"router/selected_tokens_s1": 3.09375,
|
| 1680 |
+
"router/trigger_rate_s0": 0.0,
|
| 1681 |
+
"router/trigger_rate_s1": 0.96875,
|
| 1682 |
+
"step": 860,
|
| 1683 |
+
"tokens_trained": 2.817809728
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"comp/rl_weight": 0.03,
|
| 1687 |
+
"comp/strictness": 1.0,
|
| 1688 |
+
"epoch": 0.24680851063829787,
|
| 1689 |
+
"grad_norm": 9.715760231018066,
|
| 1690 |
+
"loss": 1.2708,
|
| 1691 |
+
"loss_ce": 1.2158164978027344,
|
| 1692 |
+
"loss_region": 0.12369807064533234,
|
| 1693 |
+
"loss_total": 1.3395146131515503,
|
| 1694 |
+
"lr": 0.001209730557187486,
|
| 1695 |
+
"router/avg_deficit_s0": 0.0,
|
| 1696 |
+
"router/avg_deficit_s1": 3.6666666666666665,
|
| 1697 |
+
"router/selected_tokens_s0": 780.3125,
|
| 1698 |
+
"router/selected_tokens_s1": 226.90625,
|
| 1699 |
+
"router/trigger_rate_s0": 0.0,
|
| 1700 |
+
"router/trigger_rate_s1": 0.09375,
|
| 1701 |
+
"step": 870,
|
| 1702 |
+
"tokens_trained": 2.850575168
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"comp/rl_weight": 0.03,
|
| 1706 |
+
"comp/strictness": 1.0,
|
| 1707 |
+
"epoch": 0.24964539007092199,
|
| 1708 |
+
"grad_norm": 5.817880630493164,
|
| 1709 |
+
"loss": 1.2945,
|
| 1710 |
+
"loss_ce": 1.2212210893630981,
|
| 1711 |
+
"loss_region": 0.0879848301410675,
|
| 1712 |
+
"loss_total": 1.3092058897018433,
|
| 1713 |
+
"lr": 0.001209323665203635,
|
| 1714 |
+
"router/avg_deficit_s0": 0.0,
|
| 1715 |
+
"router/avg_deficit_s1": 0.0,
|
| 1716 |
+
"router/selected_tokens_s0": 245.15625,
|
| 1717 |
+
"router/selected_tokens_s1": 37.03125,
|
| 1718 |
+
"router/trigger_rate_s0": 0.0,
|
| 1719 |
+
"router/trigger_rate_s1": 0.0,
|
| 1720 |
+
"step": 880,
|
| 1721 |
+
"tokens_trained": 2.883340608
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"comp/rl_weight": 0.03,
|
| 1725 |
+
"comp/strictness": 1.0,
|
| 1726 |
+
"epoch": 0.2524822695035461,
|
| 1727 |
+
"grad_norm": 13.84002685546875,
|
| 1728 |
+
"loss": 1.3021,
|
| 1729 |
+
"loss_ce": 1.2078100442886353,
|
| 1730 |
+
"loss_region": 0.055929187685251236,
|
| 1731 |
+
"loss_total": 1.2637392282485962,
|
| 1732 |
+
"lr": 0.001208916773219784,
|
| 1733 |
+
"router/avg_deficit_s0": 0.0,
|
| 1734 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1735 |
+
"router/selected_tokens_s0": 808.375,
|
| 1736 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1737 |
+
"router/trigger_rate_s0": 0.0,
|
| 1738 |
+
"router/trigger_rate_s1": 1.0,
|
| 1739 |
+
"step": 890,
|
| 1740 |
+
"tokens_trained": 2.916106048
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"comp/rl_weight": 0.03,
|
| 1744 |
+
"comp/strictness": 1.0,
|
| 1745 |
+
"epoch": 0.2553191489361702,
|
| 1746 |
+
"grad_norm": 12.068603515625,
|
| 1747 |
+
"loss": 1.2995,
|
| 1748 |
+
"loss_ce": 1.2349709272384644,
|
| 1749 |
+
"loss_region": 0.07071184366941452,
|
| 1750 |
+
"loss_total": 1.3056827783584595,
|
| 1751 |
+
"lr": 0.001208509881235933,
|
| 1752 |
+
"router/avg_deficit_s0": 0.0,
|
| 1753 |
+
"router/avg_deficit_s1": 6.71875,
|
| 1754 |
+
"router/selected_tokens_s0": 1462.46875,
|
| 1755 |
+
"router/selected_tokens_s1": 1.28125,
|
| 1756 |
+
"router/trigger_rate_s0": 0.0,
|
| 1757 |
+
"router/trigger_rate_s1": 1.0,
|
| 1758 |
+
"step": 900,
|
| 1759 |
+
"tokens_trained": 2.948871488
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"comp/rl_weight": 0.03,
|
| 1763 |
+
"comp/strictness": 1.0,
|
| 1764 |
+
"epoch": 0.2581560283687943,
|
| 1765 |
+
"grad_norm": 3.185922384262085,
|
| 1766 |
+
"loss": 1.2773,
|
| 1767 |
+
"loss_ce": 1.2066916227340698,
|
| 1768 |
+
"loss_region": 0.05171738564968109,
|
| 1769 |
+
"loss_total": 1.258409023284912,
|
| 1770 |
+
"lr": 0.001208102989252082,
|
| 1771 |
+
"router/avg_deficit_s0": 0.0,
|
| 1772 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1773 |
+
"router/selected_tokens_s0": 612.25,
|
| 1774 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1775 |
+
"router/trigger_rate_s0": 0.0,
|
| 1776 |
+
"router/trigger_rate_s1": 1.0,
|
| 1777 |
+
"step": 910,
|
| 1778 |
+
"tokens_trained": 2.981630496
|
| 1779 |
+
},
|
| 1780 |
+
{
|
| 1781 |
+
"comp/rl_weight": 0.03,
|
| 1782 |
+
"comp/strictness": 1.0,
|
| 1783 |
+
"epoch": 0.26099290780141843,
|
| 1784 |
+
"grad_norm": 13.117000579833984,
|
| 1785 |
+
"loss": 1.3082,
|
| 1786 |
+
"loss_ce": 1.2591320276260376,
|
| 1787 |
+
"loss_region": 0.057609450072050095,
|
| 1788 |
+
"loss_total": 1.3167414665222168,
|
| 1789 |
+
"lr": 0.0012076960972682309,
|
| 1790 |
+
"router/avg_deficit_s0": 0.0,
|
| 1791 |
+
"router/avg_deficit_s1": 7.0,
|
| 1792 |
+
"router/selected_tokens_s0": 873.21875,
|
| 1793 |
+
"router/selected_tokens_s1": 1.0,
|
| 1794 |
+
"router/trigger_rate_s0": 0.0,
|
| 1795 |
+
"router/trigger_rate_s1": 1.0,
|
| 1796 |
+
"step": 920,
|
| 1797 |
+
"tokens_trained": 3.014395936
|
| 1798 |
+
},
|
| 1799 |
+
{
|
| 1800 |
+
"comp/rl_weight": 0.03,
|
| 1801 |
+
"comp/strictness": 1.0,
|
| 1802 |
+
"epoch": 0.26382978723404255,
|
| 1803 |
+
"grad_norm": 3.454926013946533,
|
| 1804 |
+
"loss": 1.2552,
|
| 1805 |
+
"loss_ce": 1.192643165588379,
|
| 1806 |
+
"loss_region": 0.05184994637966156,
|
| 1807 |
+
"loss_total": 1.2444931268692017,
|
| 1808 |
+
"lr": 0.0012072892052843798,
|
| 1809 |
+
"router/avg_deficit_s0": 0.0,
|
| 1810 |
+
"router/avg_deficit_s1": 7.0,
|
| 1811 |
+
"router/selected_tokens_s0": 660.6875,
|
| 1812 |
+
"router/selected_tokens_s1": 1.0,
|
| 1813 |
+
"router/trigger_rate_s0": 0.0,
|
| 1814 |
+
"router/trigger_rate_s1": 1.0,
|
| 1815 |
+
"step": 930,
|
| 1816 |
+
"tokens_trained": 3.047160576
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"comp/rl_weight": 0.03,
|
| 1820 |
+
"comp/strictness": 1.0,
|
| 1821 |
+
"epoch": 0.26666666666666666,
|
| 1822 |
+
"grad_norm": 3.770582437515259,
|
| 1823 |
+
"loss": 1.256,
|
| 1824 |
+
"loss_ce": 1.1592071056365967,
|
| 1825 |
+
"loss_region": 0.049208104610443115,
|
| 1826 |
+
"loss_total": 1.2084152698516846,
|
| 1827 |
+
"lr": 0.0012068823133005288,
|
| 1828 |
+
"router/avg_deficit_s0": 0.0,
|
| 1829 |
+
"router/avg_deficit_s1": 6.8125,
|
| 1830 |
+
"router/selected_tokens_s0": 356.875,
|
| 1831 |
+
"router/selected_tokens_s1": 1.1875,
|
| 1832 |
+
"router/trigger_rate_s0": 0.0,
|
| 1833 |
+
"router/trigger_rate_s1": 1.0,
|
| 1834 |
+
"step": 940,
|
| 1835 |
+
"tokens_trained": 3.079925216
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"comp/rl_weight": 0.03,
|
| 1839 |
+
"comp/strictness": 1.0,
|
| 1840 |
+
"epoch": 0.2695035460992908,
|
| 1841 |
+
"grad_norm": 7.1153244972229,
|
| 1842 |
+
"loss": 1.2572,
|
| 1843 |
+
"loss_ce": 1.1901915073394775,
|
| 1844 |
+
"loss_region": 0.050450198352336884,
|
| 1845 |
+
"loss_total": 1.240641713142395,
|
| 1846 |
+
"lr": 0.0012064754213166778,
|
| 1847 |
+
"router/avg_deficit_s0": 0.0,
|
| 1848 |
+
"router/avg_deficit_s1": 6.935483870967742,
|
| 1849 |
+
"router/selected_tokens_s0": 375.375,
|
| 1850 |
+
"router/selected_tokens_s1": 1.3125,
|
| 1851 |
+
"router/trigger_rate_s0": 0.0,
|
| 1852 |
+
"router/trigger_rate_s1": 0.96875,
|
| 1853 |
+
"step": 950,
|
| 1854 |
+
"tokens_trained": 3.112689856
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"comp/rl_weight": 0.03,
|
| 1858 |
+
"comp/strictness": 1.0,
|
| 1859 |
+
"epoch": 0.2723404255319149,
|
| 1860 |
+
"grad_norm": 9.772027015686035,
|
| 1861 |
+
"loss": 1.2898,
|
| 1862 |
+
"loss_ce": 1.2428486347198486,
|
| 1863 |
+
"loss_region": 0.04990971088409424,
|
| 1864 |
+
"loss_total": 1.2927583456039429,
|
| 1865 |
+
"lr": 0.001206068529332827,
|
| 1866 |
+
"router/avg_deficit_s0": 0.0,
|
| 1867 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1868 |
+
"router/selected_tokens_s0": 411.8125,
|
| 1869 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1870 |
+
"router/trigger_rate_s0": 0.0,
|
| 1871 |
+
"router/trigger_rate_s1": 1.0,
|
| 1872 |
+
"step": 960,
|
| 1873 |
+
"tokens_trained": 3.145454496
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"comp/rl_weight": 0.03,
|
| 1877 |
+
"comp/strictness": 1.0,
|
| 1878 |
+
"epoch": 0.275177304964539,
|
| 1879 |
+
"grad_norm": 11.92085075378418,
|
| 1880 |
+
"loss": 1.2779,
|
| 1881 |
+
"loss_ce": 1.224082350730896,
|
| 1882 |
+
"loss_region": 0.061600904911756516,
|
| 1883 |
+
"loss_total": 1.285683274269104,
|
| 1884 |
+
"lr": 0.001205661637348976,
|
| 1885 |
+
"router/avg_deficit_s0": 0.0,
|
| 1886 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1887 |
+
"router/selected_tokens_s0": 1274.90625,
|
| 1888 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1889 |
+
"router/trigger_rate_s0": 0.0,
|
| 1890 |
+
"router/trigger_rate_s1": 1.0,
|
| 1891 |
+
"step": 970,
|
| 1892 |
+
"tokens_trained": 3.178219936
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"comp/rl_weight": 0.03,
|
| 1896 |
+
"comp/strictness": 1.0,
|
| 1897 |
+
"epoch": 0.27801418439716313,
|
| 1898 |
+
"grad_norm": 6.2715935707092285,
|
| 1899 |
+
"loss": 1.256,
|
| 1900 |
+
"loss_ce": 1.2057243585586548,
|
| 1901 |
+
"loss_region": 0.05960024148225784,
|
| 1902 |
+
"loss_total": 1.265324592590332,
|
| 1903 |
+
"lr": 0.0012052547453651249,
|
| 1904 |
+
"router/avg_deficit_s0": 0.0,
|
| 1905 |
+
"router/avg_deficit_s1": 6.9375,
|
| 1906 |
+
"router/selected_tokens_s0": 1017.5,
|
| 1907 |
+
"router/selected_tokens_s1": 1.0625,
|
| 1908 |
+
"router/trigger_rate_s0": 0.0,
|
| 1909 |
+
"router/trigger_rate_s1": 1.0,
|
| 1910 |
+
"step": 980,
|
| 1911 |
+
"tokens_trained": 3.210985376
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"comp/rl_weight": 0.03,
|
| 1915 |
+
"comp/strictness": 1.0,
|
| 1916 |
+
"epoch": 0.28085106382978725,
|
| 1917 |
+
"grad_norm": 4.448913097381592,
|
| 1918 |
+
"loss": 1.2509,
|
| 1919 |
+
"loss_ce": 1.2143372297286987,
|
| 1920 |
+
"loss_region": 0.056969594210386276,
|
| 1921 |
+
"loss_total": 1.2713068723678589,
|
| 1922 |
+
"lr": 0.0012048478533812738,
|
| 1923 |
+
"router/avg_deficit_s0": 0.0,
|
| 1924 |
+
"router/avg_deficit_s1": 7.0,
|
| 1925 |
+
"router/selected_tokens_s0": 965.90625,
|
| 1926 |
+
"router/selected_tokens_s1": 1.0,
|
| 1927 |
+
"router/trigger_rate_s0": 0.0,
|
| 1928 |
+
"router/trigger_rate_s1": 1.0,
|
| 1929 |
+
"step": 990,
|
| 1930 |
+
"tokens_trained": 3.243750816
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"comp/rl_weight": 0.03,
|
| 1934 |
+
"comp/strictness": 1.0,
|
| 1935 |
+
"epoch": 0.28368794326241137,
|
| 1936 |
+
"grad_norm": 5.426684856414795,
|
| 1937 |
+
"loss": 1.2402,
|
| 1938 |
+
"loss_ce": 1.16555655002594,
|
| 1939 |
+
"loss_region": 0.05230846256017685,
|
| 1940 |
+
"loss_total": 1.217864990234375,
|
| 1941 |
+
"lr": 0.0012044409613974226,
|
| 1942 |
+
"router/avg_deficit_s0": 0.0,
|
| 1943 |
+
"router/avg_deficit_s1": 7.0,
|
| 1944 |
+
"router/selected_tokens_s0": 619.40625,
|
| 1945 |
+
"router/selected_tokens_s1": 1.0,
|
| 1946 |
+
"router/trigger_rate_s0": 0.0,
|
| 1947 |
+
"router/trigger_rate_s1": 1.0,
|
| 1948 |
+
"step": 1000,
|
| 1949 |
+
"tokens_trained": 3.276516256
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 0.28368794326241137,
|
| 1953 |
+
"eval_ppl": 3.2542863563067255,
|
| 1954 |
+
"eval_runtime": 0.578,
|
| 1955 |
+
"step": 1000,
|
| 1956 |
+
"tokens_trained": 3.276516256
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 0.28368794326241137,
|
| 1960 |
+
"eval_F": 0.0006250488319399953,
|
| 1961 |
+
"eval_F_cds": 0.0011144234251803972,
|
| 1962 |
+
"eval_F_exon": 0.00042270531400966186,
|
| 1963 |
+
"eval_F_intron": 0.000556743304131992,
|
| 1964 |
+
"eval_F_nig": 0.0006567124268080237,
|
| 1965 |
+
"eval_F_promoter": 0.000801100177577206,
|
| 1966 |
+
"eval_F_utr": 0.0007665184730952016,
|
| 1967 |
+
"eval_G": 0.02674509616687583,
|
| 1968 |
+
"eval_G_cds": 0.03768764539743125,
|
| 1969 |
+
"eval_G_exon": 0.030816443991545893,
|
| 1970 |
+
"eval_G_intron": 0.02521758477546237,
|
| 1971 |
+
"eval_G_nig": 0.02268251573209672,
|
| 1972 |
+
"eval_G_promoter": 0.037520551661782184,
|
| 1973 |
+
"eval_G_utr": 0.032285099359957076,
|
| 1974 |
+
"eval_avg_bp_per_token": 1599.875,
|
| 1975 |
+
"eval_bp_per_token/cds": 897.325,
|
| 1976 |
+
"eval_bp_per_token/exon": 2365.714285714286,
|
| 1977 |
+
"eval_bp_per_token/intron": 1796.1599045346063,
|
| 1978 |
+
"eval_bp_per_token/nig": 1522.7365269461077,
|
| 1979 |
+
"eval_bp_per_token/promoter": 1248.2833333333333,
|
| 1980 |
+
"eval_bp_per_token/utr": 1304.6,
|
| 1981 |
+
"eval_ppl_cds": 3.858434678372507,
|
| 1982 |
+
"eval_ppl_exon": 3.4414248334927393,
|
| 1983 |
+
"eval_ppl_intron": 3.301873903034357,
|
| 1984 |
+
"eval_ppl_nig": 3.1578400796505584,
|
| 1985 |
+
"eval_ppl_promoter": 3.4346062040854677,
|
| 1986 |
+
"eval_ppl_utr": 3.5520807511116885,
|
| 1987 |
+
"step": 1000,
|
| 1988 |
+
"tokens_trained": 3.276516256
|
| 1989 |
+
}
|
| 1990 |
+
],
|
| 1991 |
+
"logging_steps": 10,
|
| 1992 |
+
"max_steps": 30600,
|
| 1993 |
+
"num_input_tokens_seen": 0,
|
| 1994 |
+
"num_train_epochs": 9,
|
| 1995 |
+
"save_steps": 1000,
|
| 1996 |
+
"stateful_callbacks": {
|
| 1997 |
+
"TrainerControl": {
|
| 1998 |
+
"args": {
|
| 1999 |
+
"should_epoch_stop": false,
|
| 2000 |
+
"should_evaluate": false,
|
| 2001 |
+
"should_log": false,
|
| 2002 |
+
"should_save": true,
|
| 2003 |
+
"should_training_stop": false
|
| 2004 |
+
},
|
| 2005 |
+
"attributes": {}
|
| 2006 |
+
}
|
| 2007 |
+
},
|
| 2008 |
+
"total_flos": 0.0,
|
| 2009 |
+
"train_batch_size": 32,
|
| 2010 |
+
"trial_name": null,
|
| 2011 |
+
"trial_params": null
|
| 2012 |
+
}
|
checkpoint-1000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f7cb4518b673fe0a40ad5d569c2df0284b1ddc1094eaecb8b6f4b2a826f0f109
|
| 3 |
+
size 5969
|