Checkpoint at step 12000
Browse files- checkpoints/{checkpoint-8000 β checkpoint-12000}/config.json +0 -0
- checkpoints/{checkpoint-8000 β checkpoint-12000}/generation_config.json +0 -0
- checkpoints/{checkpoint-8000 β checkpoint-12000}/model-00001-of-00003.safetensors +0 -0
- checkpoints/{checkpoint-8000 β checkpoint-12000}/model-00002-of-00003.safetensors +1 -1
- checkpoints/{checkpoint-8000 β checkpoint-12000}/model-00003-of-00003.safetensors +1 -1
- checkpoints/{checkpoint-8000 β checkpoint-12000}/model.safetensors.index.json +0 -0
- checkpoints/{checkpoint-8000 β checkpoint-12000}/optimizer.pt +1 -1
- checkpoints/{checkpoint-8000 β checkpoint-12000}/rng_state.pth +0 -0
- checkpoints/{checkpoint-8000 β checkpoint-12000}/scheduler.pt +1 -1
- checkpoints/{checkpoint-8000 β checkpoint-12000}/trainer_state.json +2803 -3
- checkpoints/{checkpoint-8000 β checkpoint-12000}/training_args.bin +0 -0
- logs/events.out.tfevents.1752541902.c28835c37b83.8174.0 +3 -0
- myolmoe/__pycache__/__init__.cpython-311.pyc +0 -0
- myolmoe/__pycache__/modeling_myolmoe.cpython-311.pyc +0 -0
- scripts/__pycache__/train.cpython-311.pyc +0 -0
checkpoints/{checkpoint-8000 β checkpoint-12000}/config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/generation_config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/model-00001-of-00003.safetensors
RENAMED
|
File without changes
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/model-00002-of-00003.safetensors
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4999439616
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a2e2d96888c8e1d6ff514f2505944786e1330bd3f8558ab8f5885fa9f858dcaa
|
| 3 |
size 4999439616
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/model-00003-of-00003.safetensors
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3892418912
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:beb7990e0ef0b1e2d7c8250221d310950dcaa366cb128d342e802233594cc78b
|
| 3 |
size 3892418912
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/model.safetensors.index.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 101356346
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fac14ccb2cde6e72781c03322ad642a3d47451f3c22ab7414b8a4a496f4f5fc3
|
| 3 |
size 101356346
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/rng_state.pth
RENAMED
|
File without changes
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/scheduler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:826efb50edce9db96af40792ceaf98c59834be42e3929c03f2e95d52a6df7f64
|
| 3 |
size 1064
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/trainer_state.json
RENAMED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -5608,6 +5608,2806 @@
|
|
| 5608 |
"learning_rate": 9.971147649346211e-05,
|
| 5609 |
"loss": 9.1089,
|
| 5610 |
"step": 8000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5611 |
}
|
| 5612 |
],
|
| 5613 |
"logging_steps": 10,
|
|
@@ -5627,7 +8427,7 @@
|
|
| 5627 |
"attributes": {}
|
| 5628 |
}
|
| 5629 |
},
|
| 5630 |
-
"total_flos":
|
| 5631 |
"train_batch_size": 2,
|
| 5632 |
"trial_name": null,
|
| 5633 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5886789676042605,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 12000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 5608 |
"learning_rate": 9.971147649346211e-05,
|
| 5609 |
"loss": 9.1089,
|
| 5610 |
"step": 8000
|
| 5611 |
+
},
|
| 5612 |
+
{
|
| 5613 |
+
"epoch": 0.39294321087584394,
|
| 5614 |
+
"grad_norm": 1.0078125,
|
| 5615 |
+
"learning_rate": 9.970840683733853e-05,
|
| 5616 |
+
"loss": 9.3147,
|
| 5617 |
+
"step": 8010
|
| 5618 |
+
},
|
| 5619 |
+
{
|
| 5620 |
+
"epoch": 0.3934337766821808,
|
| 5621 |
+
"grad_norm": 1.2421875,
|
| 5622 |
+
"learning_rate": 9.970532098593317e-05,
|
| 5623 |
+
"loss": 9.3678,
|
| 5624 |
+
"step": 8020
|
| 5625 |
+
},
|
| 5626 |
+
{
|
| 5627 |
+
"epoch": 0.3939243424885177,
|
| 5628 |
+
"grad_norm": 1.453125,
|
| 5629 |
+
"learning_rate": 9.970221894025143e-05,
|
| 5630 |
+
"loss": 8.9748,
|
| 5631 |
+
"step": 8030
|
| 5632 |
+
},
|
| 5633 |
+
{
|
| 5634 |
+
"epoch": 0.3944149082948546,
|
| 5635 |
+
"grad_norm": 0.92578125,
|
| 5636 |
+
"learning_rate": 9.969910070130395e-05,
|
| 5637 |
+
"loss": 9.4,
|
| 5638 |
+
"step": 8040
|
| 5639 |
+
},
|
| 5640 |
+
{
|
| 5641 |
+
"epoch": 0.3949054741011915,
|
| 5642 |
+
"grad_norm": 0.76171875,
|
| 5643 |
+
"learning_rate": 9.969596627010671e-05,
|
| 5644 |
+
"loss": 9.2384,
|
| 5645 |
+
"step": 8050
|
| 5646 |
+
},
|
| 5647 |
+
{
|
| 5648 |
+
"epoch": 0.39539603990752836,
|
| 5649 |
+
"grad_norm": 1.53125,
|
| 5650 |
+
"learning_rate": 9.969281564768089e-05,
|
| 5651 |
+
"loss": 9.7778,
|
| 5652 |
+
"step": 8060
|
| 5653 |
+
},
|
| 5654 |
+
{
|
| 5655 |
+
"epoch": 0.39588660571386525,
|
| 5656 |
+
"grad_norm": 1.0234375,
|
| 5657 |
+
"learning_rate": 9.9689648835053e-05,
|
| 5658 |
+
"loss": 9.2588,
|
| 5659 |
+
"step": 8070
|
| 5660 |
+
},
|
| 5661 |
+
{
|
| 5662 |
+
"epoch": 0.39637717152020213,
|
| 5663 |
+
"grad_norm": 1.5078125,
|
| 5664 |
+
"learning_rate": 9.96864658332548e-05,
|
| 5665 |
+
"loss": 9.7844,
|
| 5666 |
+
"step": 8080
|
| 5667 |
+
},
|
| 5668 |
+
{
|
| 5669 |
+
"epoch": 0.396867737326539,
|
| 5670 |
+
"grad_norm": 1.09375,
|
| 5671 |
+
"learning_rate": 9.968326664332333e-05,
|
| 5672 |
+
"loss": 8.7478,
|
| 5673 |
+
"step": 8090
|
| 5674 |
+
},
|
| 5675 |
+
{
|
| 5676 |
+
"epoch": 0.3973583031328759,
|
| 5677 |
+
"grad_norm": 2.40625,
|
| 5678 |
+
"learning_rate": 9.96800512663009e-05,
|
| 5679 |
+
"loss": 10.1442,
|
| 5680 |
+
"step": 8100
|
| 5681 |
+
},
|
| 5682 |
+
{
|
| 5683 |
+
"epoch": 0.3978488689392128,
|
| 5684 |
+
"grad_norm": 1.40625,
|
| 5685 |
+
"learning_rate": 9.967681970323512e-05,
|
| 5686 |
+
"loss": 9.3243,
|
| 5687 |
+
"step": 8110
|
| 5688 |
+
},
|
| 5689 |
+
{
|
| 5690 |
+
"epoch": 0.39833943474554967,
|
| 5691 |
+
"grad_norm": 1.3359375,
|
| 5692 |
+
"learning_rate": 9.967357195517881e-05,
|
| 5693 |
+
"loss": 9.4967,
|
| 5694 |
+
"step": 8120
|
| 5695 |
+
},
|
| 5696 |
+
{
|
| 5697 |
+
"epoch": 0.39883000055188655,
|
| 5698 |
+
"grad_norm": 1.453125,
|
| 5699 |
+
"learning_rate": 9.967030802319015e-05,
|
| 5700 |
+
"loss": 9.3689,
|
| 5701 |
+
"step": 8130
|
| 5702 |
+
},
|
| 5703 |
+
{
|
| 5704 |
+
"epoch": 0.39932056635822344,
|
| 5705 |
+
"grad_norm": 1.2578125,
|
| 5706 |
+
"learning_rate": 9.966702790833253e-05,
|
| 5707 |
+
"loss": 9.0353,
|
| 5708 |
+
"step": 8140
|
| 5709 |
+
},
|
| 5710 |
+
{
|
| 5711 |
+
"epoch": 0.3998111321645603,
|
| 5712 |
+
"grad_norm": 0.76171875,
|
| 5713 |
+
"learning_rate": 9.96637316116746e-05,
|
| 5714 |
+
"loss": 9.3215,
|
| 5715 |
+
"step": 8150
|
| 5716 |
+
},
|
| 5717 |
+
{
|
| 5718 |
+
"epoch": 0.4003016979708972,
|
| 5719 |
+
"grad_norm": 1.15625,
|
| 5720 |
+
"learning_rate": 9.966041913429036e-05,
|
| 5721 |
+
"loss": 10.0265,
|
| 5722 |
+
"step": 8160
|
| 5723 |
+
},
|
| 5724 |
+
{
|
| 5725 |
+
"epoch": 0.4007922637772341,
|
| 5726 |
+
"grad_norm": 1.1640625,
|
| 5727 |
+
"learning_rate": 9.965709047725901e-05,
|
| 5728 |
+
"loss": 8.743,
|
| 5729 |
+
"step": 8170
|
| 5730 |
+
},
|
| 5731 |
+
{
|
| 5732 |
+
"epoch": 0.401282829583571,
|
| 5733 |
+
"grad_norm": 0.82421875,
|
| 5734 |
+
"learning_rate": 9.965374564166505e-05,
|
| 5735 |
+
"loss": 8.9593,
|
| 5736 |
+
"step": 8180
|
| 5737 |
+
},
|
| 5738 |
+
{
|
| 5739 |
+
"epoch": 0.40177339538990786,
|
| 5740 |
+
"grad_norm": 1.0390625,
|
| 5741 |
+
"learning_rate": 9.965038462859824e-05,
|
| 5742 |
+
"loss": 8.961,
|
| 5743 |
+
"step": 8190
|
| 5744 |
+
},
|
| 5745 |
+
{
|
| 5746 |
+
"epoch": 0.40226396119624475,
|
| 5747 |
+
"grad_norm": 1.2890625,
|
| 5748 |
+
"learning_rate": 9.964700743915361e-05,
|
| 5749 |
+
"loss": 9.808,
|
| 5750 |
+
"step": 8200
|
| 5751 |
+
},
|
| 5752 |
+
{
|
| 5753 |
+
"epoch": 0.4027545270025816,
|
| 5754 |
+
"grad_norm": 8.375,
|
| 5755 |
+
"learning_rate": 9.96436140744315e-05,
|
| 5756 |
+
"loss": 9.4756,
|
| 5757 |
+
"step": 8210
|
| 5758 |
+
},
|
| 5759 |
+
{
|
| 5760 |
+
"epoch": 0.40324509280891846,
|
| 5761 |
+
"grad_norm": 0.99609375,
|
| 5762 |
+
"learning_rate": 9.964020453553746e-05,
|
| 5763 |
+
"loss": 9.3352,
|
| 5764 |
+
"step": 8220
|
| 5765 |
+
},
|
| 5766 |
+
{
|
| 5767 |
+
"epoch": 0.40373565861525534,
|
| 5768 |
+
"grad_norm": 1.7421875,
|
| 5769 |
+
"learning_rate": 9.963677882358233e-05,
|
| 5770 |
+
"loss": 9.3511,
|
| 5771 |
+
"step": 8230
|
| 5772 |
+
},
|
| 5773 |
+
{
|
| 5774 |
+
"epoch": 0.40422622442159223,
|
| 5775 |
+
"grad_norm": 1.203125,
|
| 5776 |
+
"learning_rate": 9.963333693968226e-05,
|
| 5777 |
+
"loss": 9.5166,
|
| 5778 |
+
"step": 8240
|
| 5779 |
+
},
|
| 5780 |
+
{
|
| 5781 |
+
"epoch": 0.4047167902279291,
|
| 5782 |
+
"grad_norm": 1.53125,
|
| 5783 |
+
"learning_rate": 9.962987888495862e-05,
|
| 5784 |
+
"loss": 9.3433,
|
| 5785 |
+
"step": 8250
|
| 5786 |
+
},
|
| 5787 |
+
{
|
| 5788 |
+
"epoch": 0.405207356034266,
|
| 5789 |
+
"grad_norm": 2.59375,
|
| 5790 |
+
"learning_rate": 9.962640466053804e-05,
|
| 5791 |
+
"loss": 9.9907,
|
| 5792 |
+
"step": 8260
|
| 5793 |
+
},
|
| 5794 |
+
{
|
| 5795 |
+
"epoch": 0.4056979218406029,
|
| 5796 |
+
"grad_norm": 1.625,
|
| 5797 |
+
"learning_rate": 9.962291426755248e-05,
|
| 5798 |
+
"loss": 9.5791,
|
| 5799 |
+
"step": 8270
|
| 5800 |
+
},
|
| 5801 |
+
{
|
| 5802 |
+
"epoch": 0.40618848764693977,
|
| 5803 |
+
"grad_norm": 1.125,
|
| 5804 |
+
"learning_rate": 9.96194077071391e-05,
|
| 5805 |
+
"loss": 9.2634,
|
| 5806 |
+
"step": 8280
|
| 5807 |
+
},
|
| 5808 |
+
{
|
| 5809 |
+
"epoch": 0.40667905345327665,
|
| 5810 |
+
"grad_norm": 1.1484375,
|
| 5811 |
+
"learning_rate": 9.961588498044037e-05,
|
| 5812 |
+
"loss": 9.2259,
|
| 5813 |
+
"step": 8290
|
| 5814 |
+
},
|
| 5815 |
+
{
|
| 5816 |
+
"epoch": 0.40716961925961354,
|
| 5817 |
+
"grad_norm": 1.6875,
|
| 5818 |
+
"learning_rate": 9.961234608860402e-05,
|
| 5819 |
+
"loss": 9.1666,
|
| 5820 |
+
"step": 8300
|
| 5821 |
+
},
|
| 5822 |
+
{
|
| 5823 |
+
"epoch": 0.4076601850659504,
|
| 5824 |
+
"grad_norm": 1.3359375,
|
| 5825 |
+
"learning_rate": 9.960879103278303e-05,
|
| 5826 |
+
"loss": 9.4266,
|
| 5827 |
+
"step": 8310
|
| 5828 |
+
},
|
| 5829 |
+
{
|
| 5830 |
+
"epoch": 0.4081507508722873,
|
| 5831 |
+
"grad_norm": 1.0703125,
|
| 5832 |
+
"learning_rate": 9.960521981413566e-05,
|
| 5833 |
+
"loss": 9.2596,
|
| 5834 |
+
"step": 8320
|
| 5835 |
+
},
|
| 5836 |
+
{
|
| 5837 |
+
"epoch": 0.4086413166786242,
|
| 5838 |
+
"grad_norm": 1.9296875,
|
| 5839 |
+
"learning_rate": 9.960163243382545e-05,
|
| 5840 |
+
"loss": 8.4992,
|
| 5841 |
+
"step": 8330
|
| 5842 |
+
},
|
| 5843 |
+
{
|
| 5844 |
+
"epoch": 0.4091318824849611,
|
| 5845 |
+
"grad_norm": 1.5234375,
|
| 5846 |
+
"learning_rate": 9.959802889302117e-05,
|
| 5847 |
+
"loss": 9.1112,
|
| 5848 |
+
"step": 8340
|
| 5849 |
+
},
|
| 5850 |
+
{
|
| 5851 |
+
"epoch": 0.40962244829129796,
|
| 5852 |
+
"grad_norm": 1.421875,
|
| 5853 |
+
"learning_rate": 9.959440919289686e-05,
|
| 5854 |
+
"loss": 9.6092,
|
| 5855 |
+
"step": 8350
|
| 5856 |
+
},
|
| 5857 |
+
{
|
| 5858 |
+
"epoch": 0.41011301409763484,
|
| 5859 |
+
"grad_norm": 2.09375,
|
| 5860 |
+
"learning_rate": 9.959077333463187e-05,
|
| 5861 |
+
"loss": 9.1144,
|
| 5862 |
+
"step": 8360
|
| 5863 |
+
},
|
| 5864 |
+
{
|
| 5865 |
+
"epoch": 0.4106035799039717,
|
| 5866 |
+
"grad_norm": 1.265625,
|
| 5867 |
+
"learning_rate": 9.958712131941077e-05,
|
| 5868 |
+
"loss": 9.007,
|
| 5869 |
+
"step": 8370
|
| 5870 |
+
},
|
| 5871 |
+
{
|
| 5872 |
+
"epoch": 0.4110941457103086,
|
| 5873 |
+
"grad_norm": 1.1328125,
|
| 5874 |
+
"learning_rate": 9.95834531484234e-05,
|
| 5875 |
+
"loss": 9.4561,
|
| 5876 |
+
"step": 8380
|
| 5877 |
+
},
|
| 5878 |
+
{
|
| 5879 |
+
"epoch": 0.4115847115166455,
|
| 5880 |
+
"grad_norm": 1.6484375,
|
| 5881 |
+
"learning_rate": 9.95797688228649e-05,
|
| 5882 |
+
"loss": 9.0578,
|
| 5883 |
+
"step": 8390
|
| 5884 |
+
},
|
| 5885 |
+
{
|
| 5886 |
+
"epoch": 0.4120752773229824,
|
| 5887 |
+
"grad_norm": 1.5703125,
|
| 5888 |
+
"learning_rate": 9.957606834393561e-05,
|
| 5889 |
+
"loss": 9.0047,
|
| 5890 |
+
"step": 8400
|
| 5891 |
+
},
|
| 5892 |
+
{
|
| 5893 |
+
"epoch": 0.41256584312931927,
|
| 5894 |
+
"grad_norm": 1.1328125,
|
| 5895 |
+
"learning_rate": 9.957235171284118e-05,
|
| 5896 |
+
"loss": 9.0116,
|
| 5897 |
+
"step": 8410
|
| 5898 |
+
},
|
| 5899 |
+
{
|
| 5900 |
+
"epoch": 0.41305640893565615,
|
| 5901 |
+
"grad_norm": 1.09375,
|
| 5902 |
+
"learning_rate": 9.956861893079253e-05,
|
| 5903 |
+
"loss": 9.5539,
|
| 5904 |
+
"step": 8420
|
| 5905 |
+
},
|
| 5906 |
+
{
|
| 5907 |
+
"epoch": 0.41354697474199303,
|
| 5908 |
+
"grad_norm": 2.375,
|
| 5909 |
+
"learning_rate": 9.956486999900578e-05,
|
| 5910 |
+
"loss": 9.4898,
|
| 5911 |
+
"step": 8430
|
| 5912 |
+
},
|
| 5913 |
+
{
|
| 5914 |
+
"epoch": 0.4140375405483299,
|
| 5915 |
+
"grad_norm": 1.0703125,
|
| 5916 |
+
"learning_rate": 9.956110491870237e-05,
|
| 5917 |
+
"loss": 9.4607,
|
| 5918 |
+
"step": 8440
|
| 5919 |
+
},
|
| 5920 |
+
{
|
| 5921 |
+
"epoch": 0.4145281063546668,
|
| 5922 |
+
"grad_norm": 0.9765625,
|
| 5923 |
+
"learning_rate": 9.9557323691109e-05,
|
| 5924 |
+
"loss": 9.0116,
|
| 5925 |
+
"step": 8450
|
| 5926 |
+
},
|
| 5927 |
+
{
|
| 5928 |
+
"epoch": 0.4150186721610037,
|
| 5929 |
+
"grad_norm": 1.0625,
|
| 5930 |
+
"learning_rate": 9.955352631745761e-05,
|
| 5931 |
+
"loss": 8.9376,
|
| 5932 |
+
"step": 8460
|
| 5933 |
+
},
|
| 5934 |
+
{
|
| 5935 |
+
"epoch": 0.4155092379673406,
|
| 5936 |
+
"grad_norm": 0.84375,
|
| 5937 |
+
"learning_rate": 9.954971279898538e-05,
|
| 5938 |
+
"loss": 9.4654,
|
| 5939 |
+
"step": 8470
|
| 5940 |
+
},
|
| 5941 |
+
{
|
| 5942 |
+
"epoch": 0.41599980377367746,
|
| 5943 |
+
"grad_norm": 1.65625,
|
| 5944 |
+
"learning_rate": 9.954588313693482e-05,
|
| 5945 |
+
"loss": 9.6625,
|
| 5946 |
+
"step": 8480
|
| 5947 |
+
},
|
| 5948 |
+
{
|
| 5949 |
+
"epoch": 0.41649036958001434,
|
| 5950 |
+
"grad_norm": 1.84375,
|
| 5951 |
+
"learning_rate": 9.954203733255362e-05,
|
| 5952 |
+
"loss": 9.5381,
|
| 5953 |
+
"step": 8490
|
| 5954 |
+
},
|
| 5955 |
+
{
|
| 5956 |
+
"epoch": 0.4169809353863512,
|
| 5957 |
+
"grad_norm": 5.375,
|
| 5958 |
+
"learning_rate": 9.953817538709478e-05,
|
| 5959 |
+
"loss": 9.4922,
|
| 5960 |
+
"step": 8500
|
| 5961 |
+
},
|
| 5962 |
+
{
|
| 5963 |
+
"epoch": 0.4174715011926881,
|
| 5964 |
+
"grad_norm": 0.9921875,
|
| 5965 |
+
"learning_rate": 9.953429730181653e-05,
|
| 5966 |
+
"loss": 9.4195,
|
| 5967 |
+
"step": 8510
|
| 5968 |
+
},
|
| 5969 |
+
{
|
| 5970 |
+
"epoch": 0.417962066999025,
|
| 5971 |
+
"grad_norm": 1.71875,
|
| 5972 |
+
"learning_rate": 9.95304030779824e-05,
|
| 5973 |
+
"loss": 9.0456,
|
| 5974 |
+
"step": 8520
|
| 5975 |
+
},
|
| 5976 |
+
{
|
| 5977 |
+
"epoch": 0.4184526328053619,
|
| 5978 |
+
"grad_norm": 1.3671875,
|
| 5979 |
+
"learning_rate": 9.952649271686114e-05,
|
| 5980 |
+
"loss": 9.5672,
|
| 5981 |
+
"step": 8530
|
| 5982 |
+
},
|
| 5983 |
+
{
|
| 5984 |
+
"epoch": 0.41894319861169876,
|
| 5985 |
+
"grad_norm": 0.83203125,
|
| 5986 |
+
"learning_rate": 9.952256621972676e-05,
|
| 5987 |
+
"loss": 8.7467,
|
| 5988 |
+
"step": 8540
|
| 5989 |
+
},
|
| 5990 |
+
{
|
| 5991 |
+
"epoch": 0.41943376441803565,
|
| 5992 |
+
"grad_norm": 1.515625,
|
| 5993 |
+
"learning_rate": 9.951862358785852e-05,
|
| 5994 |
+
"loss": 9.6638,
|
| 5995 |
+
"step": 8550
|
| 5996 |
+
},
|
| 5997 |
+
{
|
| 5998 |
+
"epoch": 0.41992433022437253,
|
| 5999 |
+
"grad_norm": 1.28125,
|
| 6000 |
+
"learning_rate": 9.951466482254097e-05,
|
| 6001 |
+
"loss": 9.5221,
|
| 6002 |
+
"step": 8560
|
| 6003 |
+
},
|
| 6004 |
+
{
|
| 6005 |
+
"epoch": 0.4204148960307094,
|
| 6006 |
+
"grad_norm": 1.6015625,
|
| 6007 |
+
"learning_rate": 9.951068992506391e-05,
|
| 6008 |
+
"loss": 9.3728,
|
| 6009 |
+
"step": 8570
|
| 6010 |
+
},
|
| 6011 |
+
{
|
| 6012 |
+
"epoch": 0.4209054618370463,
|
| 6013 |
+
"grad_norm": 1.6640625,
|
| 6014 |
+
"learning_rate": 9.950669889672238e-05,
|
| 6015 |
+
"loss": 9.0309,
|
| 6016 |
+
"step": 8580
|
| 6017 |
+
},
|
| 6018 |
+
{
|
| 6019 |
+
"epoch": 0.4213960276433832,
|
| 6020 |
+
"grad_norm": 0.77734375,
|
| 6021 |
+
"learning_rate": 9.950269173881663e-05,
|
| 6022 |
+
"loss": 9.546,
|
| 6023 |
+
"step": 8590
|
| 6024 |
+
},
|
| 6025 |
+
{
|
| 6026 |
+
"epoch": 0.42188659344972007,
|
| 6027 |
+
"grad_norm": 1.390625,
|
| 6028 |
+
"learning_rate": 9.94986684526523e-05,
|
| 6029 |
+
"loss": 9.3854,
|
| 6030 |
+
"step": 8600
|
| 6031 |
+
},
|
| 6032 |
+
{
|
| 6033 |
+
"epoch": 0.42237715925605696,
|
| 6034 |
+
"grad_norm": 1.734375,
|
| 6035 |
+
"learning_rate": 9.949462903954014e-05,
|
| 6036 |
+
"loss": 10.0409,
|
| 6037 |
+
"step": 8610
|
| 6038 |
+
},
|
| 6039 |
+
{
|
| 6040 |
+
"epoch": 0.42286772506239384,
|
| 6041 |
+
"grad_norm": 1.1875,
|
| 6042 |
+
"learning_rate": 9.949057350079622e-05,
|
| 6043 |
+
"loss": 10.0401,
|
| 6044 |
+
"step": 8620
|
| 6045 |
+
},
|
| 6046 |
+
{
|
| 6047 |
+
"epoch": 0.4233582908687307,
|
| 6048 |
+
"grad_norm": 1.171875,
|
| 6049 |
+
"learning_rate": 9.948650183774187e-05,
|
| 6050 |
+
"loss": 9.3531,
|
| 6051 |
+
"step": 8630
|
| 6052 |
+
},
|
| 6053 |
+
{
|
| 6054 |
+
"epoch": 0.4238488566750676,
|
| 6055 |
+
"grad_norm": 1.046875,
|
| 6056 |
+
"learning_rate": 9.948241405170367e-05,
|
| 6057 |
+
"loss": 9.6503,
|
| 6058 |
+
"step": 8640
|
| 6059 |
+
},
|
| 6060 |
+
{
|
| 6061 |
+
"epoch": 0.4243394224814045,
|
| 6062 |
+
"grad_norm": 1.03125,
|
| 6063 |
+
"learning_rate": 9.947831014401342e-05,
|
| 6064 |
+
"loss": 10.1025,
|
| 6065 |
+
"step": 8650
|
| 6066 |
+
},
|
| 6067 |
+
{
|
| 6068 |
+
"epoch": 0.4248299882877414,
|
| 6069 |
+
"grad_norm": 1.8515625,
|
| 6070 |
+
"learning_rate": 9.94741901160082e-05,
|
| 6071 |
+
"loss": 9.1217,
|
| 6072 |
+
"step": 8660
|
| 6073 |
+
},
|
| 6074 |
+
{
|
| 6075 |
+
"epoch": 0.42532055409407826,
|
| 6076 |
+
"grad_norm": 1.296875,
|
| 6077 |
+
"learning_rate": 9.947005396903036e-05,
|
| 6078 |
+
"loss": 9.4203,
|
| 6079 |
+
"step": 8670
|
| 6080 |
+
},
|
| 6081 |
+
{
|
| 6082 |
+
"epoch": 0.42581111990041515,
|
| 6083 |
+
"grad_norm": 1.3046875,
|
| 6084 |
+
"learning_rate": 9.946590170442747e-05,
|
| 6085 |
+
"loss": 8.9508,
|
| 6086 |
+
"step": 8680
|
| 6087 |
+
},
|
| 6088 |
+
{
|
| 6089 |
+
"epoch": 0.42630168570675203,
|
| 6090 |
+
"grad_norm": 1.78125,
|
| 6091 |
+
"learning_rate": 9.946173332355236e-05,
|
| 6092 |
+
"loss": 9.3826,
|
| 6093 |
+
"step": 8690
|
| 6094 |
+
},
|
| 6095 |
+
{
|
| 6096 |
+
"epoch": 0.4267922515130889,
|
| 6097 |
+
"grad_norm": 1.1796875,
|
| 6098 |
+
"learning_rate": 9.945754882776311e-05,
|
| 6099 |
+
"loss": 9.6903,
|
| 6100 |
+
"step": 8700
|
| 6101 |
+
},
|
| 6102 |
+
{
|
| 6103 |
+
"epoch": 0.4272828173194258,
|
| 6104 |
+
"grad_norm": 0.984375,
|
| 6105 |
+
"learning_rate": 9.945334821842303e-05,
|
| 6106 |
+
"loss": 9.1157,
|
| 6107 |
+
"step": 8710
|
| 6108 |
+
},
|
| 6109 |
+
{
|
| 6110 |
+
"epoch": 0.4277733831257627,
|
| 6111 |
+
"grad_norm": 3.171875,
|
| 6112 |
+
"learning_rate": 9.944913149690075e-05,
|
| 6113 |
+
"loss": 9.7155,
|
| 6114 |
+
"step": 8720
|
| 6115 |
+
},
|
| 6116 |
+
{
|
| 6117 |
+
"epoch": 0.42826394893209957,
|
| 6118 |
+
"grad_norm": 1.421875,
|
| 6119 |
+
"learning_rate": 9.944489866457008e-05,
|
| 6120 |
+
"loss": 9.4667,
|
| 6121 |
+
"step": 8730
|
| 6122 |
+
},
|
| 6123 |
+
{
|
| 6124 |
+
"epoch": 0.42875451473843645,
|
| 6125 |
+
"grad_norm": 1.53125,
|
| 6126 |
+
"learning_rate": 9.944064972281007e-05,
|
| 6127 |
+
"loss": 9.5808,
|
| 6128 |
+
"step": 8740
|
| 6129 |
+
},
|
| 6130 |
+
{
|
| 6131 |
+
"epoch": 0.42924508054477334,
|
| 6132 |
+
"grad_norm": 1.1328125,
|
| 6133 |
+
"learning_rate": 9.943638467300511e-05,
|
| 6134 |
+
"loss": 9.3587,
|
| 6135 |
+
"step": 8750
|
| 6136 |
+
},
|
| 6137 |
+
{
|
| 6138 |
+
"epoch": 0.4297356463511102,
|
| 6139 |
+
"grad_norm": 2.921875,
|
| 6140 |
+
"learning_rate": 9.943210351654473e-05,
|
| 6141 |
+
"loss": 9.6425,
|
| 6142 |
+
"step": 8760
|
| 6143 |
+
},
|
| 6144 |
+
{
|
| 6145 |
+
"epoch": 0.4302262121574471,
|
| 6146 |
+
"grad_norm": 1.40625,
|
| 6147 |
+
"learning_rate": 9.942780625482376e-05,
|
| 6148 |
+
"loss": 9.6145,
|
| 6149 |
+
"step": 8770
|
| 6150 |
+
},
|
| 6151 |
+
{
|
| 6152 |
+
"epoch": 0.430716777963784,
|
| 6153 |
+
"grad_norm": 1.703125,
|
| 6154 |
+
"learning_rate": 9.942349288924229e-05,
|
| 6155 |
+
"loss": 9.6646,
|
| 6156 |
+
"step": 8780
|
| 6157 |
+
},
|
| 6158 |
+
{
|
| 6159 |
+
"epoch": 0.4312073437701209,
|
| 6160 |
+
"grad_norm": 1.546875,
|
| 6161 |
+
"learning_rate": 9.941916342120564e-05,
|
| 6162 |
+
"loss": 9.7029,
|
| 6163 |
+
"step": 8790
|
| 6164 |
+
},
|
| 6165 |
+
{
|
| 6166 |
+
"epoch": 0.43169790957645776,
|
| 6167 |
+
"grad_norm": 1.71875,
|
| 6168 |
+
"learning_rate": 9.941481785212437e-05,
|
| 6169 |
+
"loss": 9.1658,
|
| 6170 |
+
"step": 8800
|
| 6171 |
+
},
|
| 6172 |
+
{
|
| 6173 |
+
"epoch": 0.43218847538279465,
|
| 6174 |
+
"grad_norm": 1.09375,
|
| 6175 |
+
"learning_rate": 9.941045618341427e-05,
|
| 6176 |
+
"loss": 9.9594,
|
| 6177 |
+
"step": 8810
|
| 6178 |
+
},
|
| 6179 |
+
{
|
| 6180 |
+
"epoch": 0.43267904118913153,
|
| 6181 |
+
"grad_norm": 1.0390625,
|
| 6182 |
+
"learning_rate": 9.940607841649643e-05,
|
| 6183 |
+
"loss": 9.3896,
|
| 6184 |
+
"step": 8820
|
| 6185 |
+
},
|
| 6186 |
+
{
|
| 6187 |
+
"epoch": 0.4331696069954684,
|
| 6188 |
+
"grad_norm": 1.4375,
|
| 6189 |
+
"learning_rate": 9.940168455279713e-05,
|
| 6190 |
+
"loss": 8.9524,
|
| 6191 |
+
"step": 8830
|
| 6192 |
+
},
|
| 6193 |
+
{
|
| 6194 |
+
"epoch": 0.4336601728018053,
|
| 6195 |
+
"grad_norm": 1.203125,
|
| 6196 |
+
"learning_rate": 9.939727459374792e-05,
|
| 6197 |
+
"loss": 9.8094,
|
| 6198 |
+
"step": 8840
|
| 6199 |
+
},
|
| 6200 |
+
{
|
| 6201 |
+
"epoch": 0.4341507386081422,
|
| 6202 |
+
"grad_norm": 2.734375,
|
| 6203 |
+
"learning_rate": 9.93928485407856e-05,
|
| 6204 |
+
"loss": 9.6122,
|
| 6205 |
+
"step": 8850
|
| 6206 |
+
},
|
| 6207 |
+
{
|
| 6208 |
+
"epoch": 0.43464130441447907,
|
| 6209 |
+
"grad_norm": 0.88671875,
|
| 6210 |
+
"learning_rate": 9.938840639535219e-05,
|
| 6211 |
+
"loss": 9.5027,
|
| 6212 |
+
"step": 8860
|
| 6213 |
+
},
|
| 6214 |
+
{
|
| 6215 |
+
"epoch": 0.43513187022081595,
|
| 6216 |
+
"grad_norm": 1.1015625,
|
| 6217 |
+
"learning_rate": 9.938394815889497e-05,
|
| 6218 |
+
"loss": 8.9665,
|
| 6219 |
+
"step": 8870
|
| 6220 |
+
},
|
| 6221 |
+
{
|
| 6222 |
+
"epoch": 0.43562243602715284,
|
| 6223 |
+
"grad_norm": 0.96484375,
|
| 6224 |
+
"learning_rate": 9.937947383286646e-05,
|
| 6225 |
+
"loss": 9.6251,
|
| 6226 |
+
"step": 8880
|
| 6227 |
+
},
|
| 6228 |
+
{
|
| 6229 |
+
"epoch": 0.4361130018334897,
|
| 6230 |
+
"grad_norm": 1.7265625,
|
| 6231 |
+
"learning_rate": 9.937498341872443e-05,
|
| 6232 |
+
"loss": 10.0964,
|
| 6233 |
+
"step": 8890
|
| 6234 |
+
},
|
| 6235 |
+
{
|
| 6236 |
+
"epoch": 0.4366035676398266,
|
| 6237 |
+
"grad_norm": 0.91796875,
|
| 6238 |
+
"learning_rate": 9.937047691793186e-05,
|
| 6239 |
+
"loss": 9.6408,
|
| 6240 |
+
"step": 8900
|
| 6241 |
+
},
|
| 6242 |
+
{
|
| 6243 |
+
"epoch": 0.4370941334461635,
|
| 6244 |
+
"grad_norm": 121.5,
|
| 6245 |
+
"learning_rate": 9.936595433195701e-05,
|
| 6246 |
+
"loss": 9.911,
|
| 6247 |
+
"step": 8910
|
| 6248 |
+
},
|
| 6249 |
+
{
|
| 6250 |
+
"epoch": 0.4375846992525004,
|
| 6251 |
+
"grad_norm": 1.328125,
|
| 6252 |
+
"learning_rate": 9.936141566227335e-05,
|
| 6253 |
+
"loss": 9.3583,
|
| 6254 |
+
"step": 8920
|
| 6255 |
+
},
|
| 6256 |
+
{
|
| 6257 |
+
"epoch": 0.43807526505883726,
|
| 6258 |
+
"grad_norm": 1.1171875,
|
| 6259 |
+
"learning_rate": 9.935686091035963e-05,
|
| 6260 |
+
"loss": 9.074,
|
| 6261 |
+
"step": 8930
|
| 6262 |
+
},
|
| 6263 |
+
{
|
| 6264 |
+
"epoch": 0.43856583086517414,
|
| 6265 |
+
"grad_norm": 1.4765625,
|
| 6266 |
+
"learning_rate": 9.93522900776998e-05,
|
| 6267 |
+
"loss": 9.409,
|
| 6268 |
+
"step": 8940
|
| 6269 |
+
},
|
| 6270 |
+
{
|
| 6271 |
+
"epoch": 0.43905639667151103,
|
| 6272 |
+
"grad_norm": 1.3515625,
|
| 6273 |
+
"learning_rate": 9.934770316578306e-05,
|
| 6274 |
+
"loss": 9.5203,
|
| 6275 |
+
"step": 8950
|
| 6276 |
+
},
|
| 6277 |
+
{
|
| 6278 |
+
"epoch": 0.4395469624778479,
|
| 6279 |
+
"grad_norm": 1.2421875,
|
| 6280 |
+
"learning_rate": 9.934310017610385e-05,
|
| 6281 |
+
"loss": 9.6801,
|
| 6282 |
+
"step": 8960
|
| 6283 |
+
},
|
| 6284 |
+
{
|
| 6285 |
+
"epoch": 0.4400375282841848,
|
| 6286 |
+
"grad_norm": 1.1015625,
|
| 6287 |
+
"learning_rate": 9.933848111016186e-05,
|
| 6288 |
+
"loss": 9.5165,
|
| 6289 |
+
"step": 8970
|
| 6290 |
+
},
|
| 6291 |
+
{
|
| 6292 |
+
"epoch": 0.4405280940905217,
|
| 6293 |
+
"grad_norm": 1.5234375,
|
| 6294 |
+
"learning_rate": 9.933384596946201e-05,
|
| 6295 |
+
"loss": 9.6044,
|
| 6296 |
+
"step": 8980
|
| 6297 |
+
},
|
| 6298 |
+
{
|
| 6299 |
+
"epoch": 0.44101865989685857,
|
| 6300 |
+
"grad_norm": 1.1875,
|
| 6301 |
+
"learning_rate": 9.932919475551443e-05,
|
| 6302 |
+
"loss": 9.2709,
|
| 6303 |
+
"step": 8990
|
| 6304 |
+
},
|
| 6305 |
+
{
|
| 6306 |
+
"epoch": 0.4415092257031954,
|
| 6307 |
+
"grad_norm": 1.234375,
|
| 6308 |
+
"learning_rate": 9.932452746983455e-05,
|
| 6309 |
+
"loss": 9.0887,
|
| 6310 |
+
"step": 9000
|
| 6311 |
+
},
|
| 6312 |
+
{
|
| 6313 |
+
"epoch": 0.4419997915095323,
|
| 6314 |
+
"grad_norm": 0.94140625,
|
| 6315 |
+
"learning_rate": 9.931984411394297e-05,
|
| 6316 |
+
"loss": 8.9249,
|
| 6317 |
+
"step": 9010
|
| 6318 |
+
},
|
| 6319 |
+
{
|
| 6320 |
+
"epoch": 0.44249035731586916,
|
| 6321 |
+
"grad_norm": 1.234375,
|
| 6322 |
+
"learning_rate": 9.931514468936556e-05,
|
| 6323 |
+
"loss": 9.3311,
|
| 6324 |
+
"step": 9020
|
| 6325 |
+
},
|
| 6326 |
+
{
|
| 6327 |
+
"epoch": 0.44298092312220605,
|
| 6328 |
+
"grad_norm": 1.5625,
|
| 6329 |
+
"learning_rate": 9.931042919763343e-05,
|
| 6330 |
+
"loss": 9.1204,
|
| 6331 |
+
"step": 9030
|
| 6332 |
+
},
|
| 6333 |
+
{
|
| 6334 |
+
"epoch": 0.44347148892854293,
|
| 6335 |
+
"grad_norm": 1.0078125,
|
| 6336 |
+
"learning_rate": 9.930569764028289e-05,
|
| 6337 |
+
"loss": 8.936,
|
| 6338 |
+
"step": 9040
|
| 6339 |
+
},
|
| 6340 |
+
{
|
| 6341 |
+
"epoch": 0.4439620547348798,
|
| 6342 |
+
"grad_norm": 1.5078125,
|
| 6343 |
+
"learning_rate": 9.930095001885554e-05,
|
| 6344 |
+
"loss": 9.8463,
|
| 6345 |
+
"step": 9050
|
| 6346 |
+
},
|
| 6347 |
+
{
|
| 6348 |
+
"epoch": 0.4444526205412167,
|
| 6349 |
+
"grad_norm": 1.3125,
|
| 6350 |
+
"learning_rate": 9.929618633489815e-05,
|
| 6351 |
+
"loss": 9.4598,
|
| 6352 |
+
"step": 9060
|
| 6353 |
+
},
|
| 6354 |
+
{
|
| 6355 |
+
"epoch": 0.4449431863475536,
|
| 6356 |
+
"grad_norm": 1.390625,
|
| 6357 |
+
"learning_rate": 9.929140658996278e-05,
|
| 6358 |
+
"loss": 9.1565,
|
| 6359 |
+
"step": 9070
|
| 6360 |
+
},
|
| 6361 |
+
{
|
| 6362 |
+
"epoch": 0.44543375215389047,
|
| 6363 |
+
"grad_norm": 1.0625,
|
| 6364 |
+
"learning_rate": 9.928661078560669e-05,
|
| 6365 |
+
"loss": 9.5713,
|
| 6366 |
+
"step": 9080
|
| 6367 |
+
},
|
| 6368 |
+
{
|
| 6369 |
+
"epoch": 0.44592431796022736,
|
| 6370 |
+
"grad_norm": 1.265625,
|
| 6371 |
+
"learning_rate": 9.928179892339238e-05,
|
| 6372 |
+
"loss": 9.5154,
|
| 6373 |
+
"step": 9090
|
| 6374 |
+
},
|
| 6375 |
+
{
|
| 6376 |
+
"epoch": 0.44641488376656424,
|
| 6377 |
+
"grad_norm": 2.671875,
|
| 6378 |
+
"learning_rate": 9.927697100488757e-05,
|
| 6379 |
+
"loss": 9.8792,
|
| 6380 |
+
"step": 9100
|
| 6381 |
+
},
|
| 6382 |
+
{
|
| 6383 |
+
"epoch": 0.4469054495729011,
|
| 6384 |
+
"grad_norm": 1.1640625,
|
| 6385 |
+
"learning_rate": 9.927212703166526e-05,
|
| 6386 |
+
"loss": 8.8977,
|
| 6387 |
+
"step": 9110
|
| 6388 |
+
},
|
| 6389 |
+
{
|
| 6390 |
+
"epoch": 0.447396015379238,
|
| 6391 |
+
"grad_norm": 1.2109375,
|
| 6392 |
+
"learning_rate": 9.92672670053036e-05,
|
| 6393 |
+
"loss": 9.3305,
|
| 6394 |
+
"step": 9120
|
| 6395 |
+
},
|
| 6396 |
+
{
|
| 6397 |
+
"epoch": 0.4478865811855749,
|
| 6398 |
+
"grad_norm": 1.5390625,
|
| 6399 |
+
"learning_rate": 9.926239092738606e-05,
|
| 6400 |
+
"loss": 9.4327,
|
| 6401 |
+
"step": 9130
|
| 6402 |
+
},
|
| 6403 |
+
{
|
| 6404 |
+
"epoch": 0.4483771469919118,
|
| 6405 |
+
"grad_norm": 1.0859375,
|
| 6406 |
+
"learning_rate": 9.925749879950123e-05,
|
| 6407 |
+
"loss": 9.3707,
|
| 6408 |
+
"step": 9140
|
| 6409 |
+
},
|
| 6410 |
+
{
|
| 6411 |
+
"epoch": 0.44886771279824866,
|
| 6412 |
+
"grad_norm": 3.578125,
|
| 6413 |
+
"learning_rate": 9.925259062324305e-05,
|
| 6414 |
+
"loss": 9.2381,
|
| 6415 |
+
"step": 9150
|
| 6416 |
+
},
|
| 6417 |
+
{
|
| 6418 |
+
"epoch": 0.44935827860458555,
|
| 6419 |
+
"grad_norm": 6.03125,
|
| 6420 |
+
"learning_rate": 9.924766640021061e-05,
|
| 6421 |
+
"loss": 8.6538,
|
| 6422 |
+
"step": 9160
|
| 6423 |
+
},
|
| 6424 |
+
{
|
| 6425 |
+
"epoch": 0.44984884441092243,
|
| 6426 |
+
"grad_norm": 2.25,
|
| 6427 |
+
"learning_rate": 9.924272613200825e-05,
|
| 6428 |
+
"loss": 9.581,
|
| 6429 |
+
"step": 9170
|
| 6430 |
+
},
|
| 6431 |
+
{
|
| 6432 |
+
"epoch": 0.4503394102172593,
|
| 6433 |
+
"grad_norm": 1.9921875,
|
| 6434 |
+
"learning_rate": 9.923776982024554e-05,
|
| 6435 |
+
"loss": 9.4617,
|
| 6436 |
+
"step": 9180
|
| 6437 |
+
},
|
| 6438 |
+
{
|
| 6439 |
+
"epoch": 0.4508299760235962,
|
| 6440 |
+
"grad_norm": 1.59375,
|
| 6441 |
+
"learning_rate": 9.923279746653728e-05,
|
| 6442 |
+
"loss": 9.5088,
|
| 6443 |
+
"step": 9190
|
| 6444 |
+
},
|
| 6445 |
+
{
|
| 6446 |
+
"epoch": 0.4513205418299331,
|
| 6447 |
+
"grad_norm": 1.3125,
|
| 6448 |
+
"learning_rate": 9.922780907250348e-05,
|
| 6449 |
+
"loss": 8.9247,
|
| 6450 |
+
"step": 9200
|
| 6451 |
+
},
|
| 6452 |
+
{
|
| 6453 |
+
"epoch": 0.45181110763626997,
|
| 6454 |
+
"grad_norm": 1.4921875,
|
| 6455 |
+
"learning_rate": 9.922280463976938e-05,
|
| 6456 |
+
"loss": 9.1817,
|
| 6457 |
+
"step": 9210
|
| 6458 |
+
},
|
| 6459 |
+
{
|
| 6460 |
+
"epoch": 0.45230167344260686,
|
| 6461 |
+
"grad_norm": 1.6953125,
|
| 6462 |
+
"learning_rate": 9.921778416996549e-05,
|
| 6463 |
+
"loss": 9.3479,
|
| 6464 |
+
"step": 9220
|
| 6465 |
+
},
|
| 6466 |
+
{
|
| 6467 |
+
"epoch": 0.45279223924894374,
|
| 6468 |
+
"grad_norm": 1.3125,
|
| 6469 |
+
"learning_rate": 9.921274766472748e-05,
|
| 6470 |
+
"loss": 10.0808,
|
| 6471 |
+
"step": 9230
|
| 6472 |
+
},
|
| 6473 |
+
{
|
| 6474 |
+
"epoch": 0.4532828050552806,
|
| 6475 |
+
"grad_norm": 1.2421875,
|
| 6476 |
+
"learning_rate": 9.920769512569625e-05,
|
| 6477 |
+
"loss": 9.1088,
|
| 6478 |
+
"step": 9240
|
| 6479 |
+
},
|
| 6480 |
+
{
|
| 6481 |
+
"epoch": 0.4537733708616175,
|
| 6482 |
+
"grad_norm": 1.171875,
|
| 6483 |
+
"learning_rate": 9.9202626554518e-05,
|
| 6484 |
+
"loss": 10.0015,
|
| 6485 |
+
"step": 9250
|
| 6486 |
+
},
|
| 6487 |
+
{
|
| 6488 |
+
"epoch": 0.4542639366679544,
|
| 6489 |
+
"grad_norm": 1.8046875,
|
| 6490 |
+
"learning_rate": 9.919754195284406e-05,
|
| 6491 |
+
"loss": 10.0798,
|
| 6492 |
+
"step": 9260
|
| 6493 |
+
},
|
| 6494 |
+
{
|
| 6495 |
+
"epoch": 0.4547545024742913,
|
| 6496 |
+
"grad_norm": 1.09375,
|
| 6497 |
+
"learning_rate": 9.919244132233104e-05,
|
| 6498 |
+
"loss": 9.0094,
|
| 6499 |
+
"step": 9270
|
| 6500 |
+
},
|
| 6501 |
+
{
|
| 6502 |
+
"epoch": 0.45524506828062816,
|
| 6503 |
+
"grad_norm": 2.640625,
|
| 6504 |
+
"learning_rate": 9.918732466464072e-05,
|
| 6505 |
+
"loss": 9.6669,
|
| 6506 |
+
"step": 9280
|
| 6507 |
+
},
|
| 6508 |
+
{
|
| 6509 |
+
"epoch": 0.45573563408696505,
|
| 6510 |
+
"grad_norm": 1.484375,
|
| 6511 |
+
"learning_rate": 9.918219198144019e-05,
|
| 6512 |
+
"loss": 9.5382,
|
| 6513 |
+
"step": 9290
|
| 6514 |
+
},
|
| 6515 |
+
{
|
| 6516 |
+
"epoch": 0.45622619989330193,
|
| 6517 |
+
"grad_norm": 1.421875,
|
| 6518 |
+
"learning_rate": 9.917704327440166e-05,
|
| 6519 |
+
"loss": 9.4784,
|
| 6520 |
+
"step": 9300
|
| 6521 |
+
},
|
| 6522 |
+
{
|
| 6523 |
+
"epoch": 0.4567167656996388,
|
| 6524 |
+
"grad_norm": 1.1484375,
|
| 6525 |
+
"learning_rate": 9.917187854520264e-05,
|
| 6526 |
+
"loss": 10.2777,
|
| 6527 |
+
"step": 9310
|
| 6528 |
+
},
|
| 6529 |
+
{
|
| 6530 |
+
"epoch": 0.4572073315059757,
|
| 6531 |
+
"grad_norm": 1.1953125,
|
| 6532 |
+
"learning_rate": 9.916669779552581e-05,
|
| 6533 |
+
"loss": 9.0473,
|
| 6534 |
+
"step": 9320
|
| 6535 |
+
},
|
| 6536 |
+
{
|
| 6537 |
+
"epoch": 0.4576978973123126,
|
| 6538 |
+
"grad_norm": 1.46875,
|
| 6539 |
+
"learning_rate": 9.91615010270591e-05,
|
| 6540 |
+
"loss": 9.1827,
|
| 6541 |
+
"step": 9330
|
| 6542 |
+
},
|
| 6543 |
+
{
|
| 6544 |
+
"epoch": 0.45818846311864947,
|
| 6545 |
+
"grad_norm": 1.8828125,
|
| 6546 |
+
"learning_rate": 9.915628824149564e-05,
|
| 6547 |
+
"loss": 9.3418,
|
| 6548 |
+
"step": 9340
|
| 6549 |
+
},
|
| 6550 |
+
{
|
| 6551 |
+
"epoch": 0.45867902892498635,
|
| 6552 |
+
"grad_norm": 1.0546875,
|
| 6553 |
+
"learning_rate": 9.91510594405338e-05,
|
| 6554 |
+
"loss": 9.5231,
|
| 6555 |
+
"step": 9350
|
| 6556 |
+
},
|
| 6557 |
+
{
|
| 6558 |
+
"epoch": 0.45916959473132324,
|
| 6559 |
+
"grad_norm": 0.93359375,
|
| 6560 |
+
"learning_rate": 9.914581462587712e-05,
|
| 6561 |
+
"loss": 9.2647,
|
| 6562 |
+
"step": 9360
|
| 6563 |
+
},
|
| 6564 |
+
{
|
| 6565 |
+
"epoch": 0.4596601605376601,
|
| 6566 |
+
"grad_norm": 1.1484375,
|
| 6567 |
+
"learning_rate": 9.914055379923442e-05,
|
| 6568 |
+
"loss": 9.6544,
|
| 6569 |
+
"step": 9370
|
| 6570 |
+
},
|
| 6571 |
+
{
|
| 6572 |
+
"epoch": 0.460150726343997,
|
| 6573 |
+
"grad_norm": 2.015625,
|
| 6574 |
+
"learning_rate": 9.913527696231969e-05,
|
| 6575 |
+
"loss": 9.1867,
|
| 6576 |
+
"step": 9380
|
| 6577 |
+
},
|
| 6578 |
+
{
|
| 6579 |
+
"epoch": 0.4606412921503339,
|
| 6580 |
+
"grad_norm": 1.125,
|
| 6581 |
+
"learning_rate": 9.912998411685216e-05,
|
| 6582 |
+
"loss": 9.5706,
|
| 6583 |
+
"step": 9390
|
| 6584 |
+
},
|
| 6585 |
+
{
|
| 6586 |
+
"epoch": 0.4611318579566708,
|
| 6587 |
+
"grad_norm": 0.9375,
|
| 6588 |
+
"learning_rate": 9.912467526455626e-05,
|
| 6589 |
+
"loss": 9.2712,
|
| 6590 |
+
"step": 9400
|
| 6591 |
+
},
|
| 6592 |
+
{
|
| 6593 |
+
"epoch": 0.46162242376300766,
|
| 6594 |
+
"grad_norm": 1.15625,
|
| 6595 |
+
"learning_rate": 9.911935040716169e-05,
|
| 6596 |
+
"loss": 9.6751,
|
| 6597 |
+
"step": 9410
|
| 6598 |
+
},
|
| 6599 |
+
{
|
| 6600 |
+
"epoch": 0.46211298956934455,
|
| 6601 |
+
"grad_norm": 1.265625,
|
| 6602 |
+
"learning_rate": 9.911400954640325e-05,
|
| 6603 |
+
"loss": 9.2709,
|
| 6604 |
+
"step": 9420
|
| 6605 |
+
},
|
| 6606 |
+
{
|
| 6607 |
+
"epoch": 0.46260355537568143,
|
| 6608 |
+
"grad_norm": 1.6953125,
|
| 6609 |
+
"learning_rate": 9.910865268402106e-05,
|
| 6610 |
+
"loss": 9.6717,
|
| 6611 |
+
"step": 9430
|
| 6612 |
+
},
|
| 6613 |
+
{
|
| 6614 |
+
"epoch": 0.4630941211820183,
|
| 6615 |
+
"grad_norm": 1.0625,
|
| 6616 |
+
"learning_rate": 9.910327982176042e-05,
|
| 6617 |
+
"loss": 8.833,
|
| 6618 |
+
"step": 9440
|
| 6619 |
+
},
|
| 6620 |
+
{
|
| 6621 |
+
"epoch": 0.4635846869883552,
|
| 6622 |
+
"grad_norm": 1.3125,
|
| 6623 |
+
"learning_rate": 9.909789096137184e-05,
|
| 6624 |
+
"loss": 9.7037,
|
| 6625 |
+
"step": 9450
|
| 6626 |
+
},
|
| 6627 |
+
{
|
| 6628 |
+
"epoch": 0.4640752527946921,
|
| 6629 |
+
"grad_norm": 1.21875,
|
| 6630 |
+
"learning_rate": 9.909248610461101e-05,
|
| 6631 |
+
"loss": 9.3282,
|
| 6632 |
+
"step": 9460
|
| 6633 |
+
},
|
| 6634 |
+
{
|
| 6635 |
+
"epoch": 0.46456581860102897,
|
| 6636 |
+
"grad_norm": 1.4921875,
|
| 6637 |
+
"learning_rate": 9.90870652532389e-05,
|
| 6638 |
+
"loss": 9.7322,
|
| 6639 |
+
"step": 9470
|
| 6640 |
+
},
|
| 6641 |
+
{
|
| 6642 |
+
"epoch": 0.46505638440736585,
|
| 6643 |
+
"grad_norm": 2.609375,
|
| 6644 |
+
"learning_rate": 9.908162840902163e-05,
|
| 6645 |
+
"loss": 9.5577,
|
| 6646 |
+
"step": 9480
|
| 6647 |
+
},
|
| 6648 |
+
{
|
| 6649 |
+
"epoch": 0.46554695021370274,
|
| 6650 |
+
"grad_norm": 1.3671875,
|
| 6651 |
+
"learning_rate": 9.907617557373057e-05,
|
| 6652 |
+
"loss": 9.5126,
|
| 6653 |
+
"step": 9490
|
| 6654 |
+
},
|
| 6655 |
+
{
|
| 6656 |
+
"epoch": 0.4660375160200396,
|
| 6657 |
+
"grad_norm": 1.5546875,
|
| 6658 |
+
"learning_rate": 9.907070674914228e-05,
|
| 6659 |
+
"loss": 9.4535,
|
| 6660 |
+
"step": 9500
|
| 6661 |
+
},
|
| 6662 |
+
{
|
| 6663 |
+
"epoch": 0.4665280818263765,
|
| 6664 |
+
"grad_norm": 1.0390625,
|
| 6665 |
+
"learning_rate": 9.906522193703853e-05,
|
| 6666 |
+
"loss": 9.1627,
|
| 6667 |
+
"step": 9510
|
| 6668 |
+
},
|
| 6669 |
+
{
|
| 6670 |
+
"epoch": 0.4670186476327134,
|
| 6671 |
+
"grad_norm": 1.15625,
|
| 6672 |
+
"learning_rate": 9.905972113920632e-05,
|
| 6673 |
+
"loss": 8.9282,
|
| 6674 |
+
"step": 9520
|
| 6675 |
+
},
|
| 6676 |
+
{
|
| 6677 |
+
"epoch": 0.4675092134390503,
|
| 6678 |
+
"grad_norm": 1.8828125,
|
| 6679 |
+
"learning_rate": 9.905420435743782e-05,
|
| 6680 |
+
"loss": 9.3116,
|
| 6681 |
+
"step": 9530
|
| 6682 |
+
},
|
| 6683 |
+
{
|
| 6684 |
+
"epoch": 0.46799977924538716,
|
| 6685 |
+
"grad_norm": 1.1875,
|
| 6686 |
+
"learning_rate": 9.904867159353041e-05,
|
| 6687 |
+
"loss": 9.3445,
|
| 6688 |
+
"step": 9540
|
| 6689 |
+
},
|
| 6690 |
+
{
|
| 6691 |
+
"epoch": 0.46849034505172404,
|
| 6692 |
+
"grad_norm": 1.3984375,
|
| 6693 |
+
"learning_rate": 9.904312284928677e-05,
|
| 6694 |
+
"loss": 8.8062,
|
| 6695 |
+
"step": 9550
|
| 6696 |
+
},
|
| 6697 |
+
{
|
| 6698 |
+
"epoch": 0.46898091085806093,
|
| 6699 |
+
"grad_norm": 1.7109375,
|
| 6700 |
+
"learning_rate": 9.903755812651463e-05,
|
| 6701 |
+
"loss": 9.8019,
|
| 6702 |
+
"step": 9560
|
| 6703 |
+
},
|
| 6704 |
+
{
|
| 6705 |
+
"epoch": 0.4694714766643978,
|
| 6706 |
+
"grad_norm": 1.4140625,
|
| 6707 |
+
"learning_rate": 9.903197742702706e-05,
|
| 6708 |
+
"loss": 9.7276,
|
| 6709 |
+
"step": 9570
|
| 6710 |
+
},
|
| 6711 |
+
{
|
| 6712 |
+
"epoch": 0.4699620424707347,
|
| 6713 |
+
"grad_norm": 1.4140625,
|
| 6714 |
+
"learning_rate": 9.902638075264226e-05,
|
| 6715 |
+
"loss": 8.9828,
|
| 6716 |
+
"step": 9580
|
| 6717 |
+
},
|
| 6718 |
+
{
|
| 6719 |
+
"epoch": 0.4704526082770716,
|
| 6720 |
+
"grad_norm": 1.0234375,
|
| 6721 |
+
"learning_rate": 9.902076810518366e-05,
|
| 6722 |
+
"loss": 9.4312,
|
| 6723 |
+
"step": 9590
|
| 6724 |
+
},
|
| 6725 |
+
{
|
| 6726 |
+
"epoch": 0.47094317408340847,
|
| 6727 |
+
"grad_norm": 1.6875,
|
| 6728 |
+
"learning_rate": 9.90151394864799e-05,
|
| 6729 |
+
"loss": 9.4076,
|
| 6730 |
+
"step": 9600
|
| 6731 |
+
},
|
| 6732 |
+
{
|
| 6733 |
+
"epoch": 0.47143373988974535,
|
| 6734 |
+
"grad_norm": 1.265625,
|
| 6735 |
+
"learning_rate": 9.900949489836482e-05,
|
| 6736 |
+
"loss": 9.443,
|
| 6737 |
+
"step": 9610
|
| 6738 |
+
},
|
| 6739 |
+
{
|
| 6740 |
+
"epoch": 0.47192430569608224,
|
| 6741 |
+
"grad_norm": 0.9765625,
|
| 6742 |
+
"learning_rate": 9.900383434267745e-05,
|
| 6743 |
+
"loss": 9.1485,
|
| 6744 |
+
"step": 9620
|
| 6745 |
+
},
|
| 6746 |
+
{
|
| 6747 |
+
"epoch": 0.4724148715024191,
|
| 6748 |
+
"grad_norm": 1.390625,
|
| 6749 |
+
"learning_rate": 9.899815782126203e-05,
|
| 6750 |
+
"loss": 9.0047,
|
| 6751 |
+
"step": 9630
|
| 6752 |
+
},
|
| 6753 |
+
{
|
| 6754 |
+
"epoch": 0.472905437308756,
|
| 6755 |
+
"grad_norm": 1.2421875,
|
| 6756 |
+
"learning_rate": 9.8992465335968e-05,
|
| 6757 |
+
"loss": 9.5106,
|
| 6758 |
+
"step": 9640
|
| 6759 |
+
},
|
| 6760 |
+
{
|
| 6761 |
+
"epoch": 0.4733960031150929,
|
| 6762 |
+
"grad_norm": 1.2734375,
|
| 6763 |
+
"learning_rate": 9.898675688865004e-05,
|
| 6764 |
+
"loss": 8.8625,
|
| 6765 |
+
"step": 9650
|
| 6766 |
+
},
|
| 6767 |
+
{
|
| 6768 |
+
"epoch": 0.4738865689214298,
|
| 6769 |
+
"grad_norm": 1.578125,
|
| 6770 |
+
"learning_rate": 9.898103248116795e-05,
|
| 6771 |
+
"loss": 9.7804,
|
| 6772 |
+
"step": 9660
|
| 6773 |
+
},
|
| 6774 |
+
{
|
| 6775 |
+
"epoch": 0.47437713472776666,
|
| 6776 |
+
"grad_norm": 1.3203125,
|
| 6777 |
+
"learning_rate": 9.897529211538678e-05,
|
| 6778 |
+
"loss": 9.1752,
|
| 6779 |
+
"step": 9670
|
| 6780 |
+
},
|
| 6781 |
+
{
|
| 6782 |
+
"epoch": 0.47486770053410354,
|
| 6783 |
+
"grad_norm": 1.6015625,
|
| 6784 |
+
"learning_rate": 9.896953579317679e-05,
|
| 6785 |
+
"loss": 9.4863,
|
| 6786 |
+
"step": 9680
|
| 6787 |
+
},
|
| 6788 |
+
{
|
| 6789 |
+
"epoch": 0.4753582663404404,
|
| 6790 |
+
"grad_norm": 0.9296875,
|
| 6791 |
+
"learning_rate": 9.89637635164134e-05,
|
| 6792 |
+
"loss": 9.5613,
|
| 6793 |
+
"step": 9690
|
| 6794 |
+
},
|
| 6795 |
+
{
|
| 6796 |
+
"epoch": 0.4758488321467773,
|
| 6797 |
+
"grad_norm": 1.6796875,
|
| 6798 |
+
"learning_rate": 9.89579752869773e-05,
|
| 6799 |
+
"loss": 9.3063,
|
| 6800 |
+
"step": 9700
|
| 6801 |
+
},
|
| 6802 |
+
{
|
| 6803 |
+
"epoch": 0.4763393979531142,
|
| 6804 |
+
"grad_norm": 1.28125,
|
| 6805 |
+
"learning_rate": 9.895217110675428e-05,
|
| 6806 |
+
"loss": 9.356,
|
| 6807 |
+
"step": 9710
|
| 6808 |
+
},
|
| 6809 |
+
{
|
| 6810 |
+
"epoch": 0.4768299637594511,
|
| 6811 |
+
"grad_norm": 1.4609375,
|
| 6812 |
+
"learning_rate": 9.894635097763538e-05,
|
| 6813 |
+
"loss": 9.4519,
|
| 6814 |
+
"step": 9720
|
| 6815 |
+
},
|
| 6816 |
+
{
|
| 6817 |
+
"epoch": 0.47732052956578797,
|
| 6818 |
+
"grad_norm": 1.1484375,
|
| 6819 |
+
"learning_rate": 9.894051490151686e-05,
|
| 6820 |
+
"loss": 9.8951,
|
| 6821 |
+
"step": 9730
|
| 6822 |
+
},
|
| 6823 |
+
{
|
| 6824 |
+
"epoch": 0.47781109537212485,
|
| 6825 |
+
"grad_norm": 1.703125,
|
| 6826 |
+
"learning_rate": 9.893466288030011e-05,
|
| 6827 |
+
"loss": 9.0165,
|
| 6828 |
+
"step": 9740
|
| 6829 |
+
},
|
| 6830 |
+
{
|
| 6831 |
+
"epoch": 0.47830166117846173,
|
| 6832 |
+
"grad_norm": 1.625,
|
| 6833 |
+
"learning_rate": 9.892879491589179e-05,
|
| 6834 |
+
"loss": 9.2397,
|
| 6835 |
+
"step": 9750
|
| 6836 |
+
},
|
| 6837 |
+
{
|
| 6838 |
+
"epoch": 0.4787922269847986,
|
| 6839 |
+
"grad_norm": 1.875,
|
| 6840 |
+
"learning_rate": 9.892291101020368e-05,
|
| 6841 |
+
"loss": 9.1272,
|
| 6842 |
+
"step": 9760
|
| 6843 |
+
},
|
| 6844 |
+
{
|
| 6845 |
+
"epoch": 0.4792827927911355,
|
| 6846 |
+
"grad_norm": 1.1796875,
|
| 6847 |
+
"learning_rate": 9.891701116515282e-05,
|
| 6848 |
+
"loss": 9.2885,
|
| 6849 |
+
"step": 9770
|
| 6850 |
+
},
|
| 6851 |
+
{
|
| 6852 |
+
"epoch": 0.47977335859747233,
|
| 6853 |
+
"grad_norm": 1.5390625,
|
| 6854 |
+
"learning_rate": 9.89110953826614e-05,
|
| 6855 |
+
"loss": 10.1661,
|
| 6856 |
+
"step": 9780
|
| 6857 |
+
},
|
| 6858 |
+
{
|
| 6859 |
+
"epoch": 0.4802639244038092,
|
| 6860 |
+
"grad_norm": 0.8515625,
|
| 6861 |
+
"learning_rate": 9.890516366465678e-05,
|
| 6862 |
+
"loss": 9.3322,
|
| 6863 |
+
"step": 9790
|
| 6864 |
+
},
|
| 6865 |
+
{
|
| 6866 |
+
"epoch": 0.4807544902101461,
|
| 6867 |
+
"grad_norm": 1.28125,
|
| 6868 |
+
"learning_rate": 9.889921601307161e-05,
|
| 6869 |
+
"loss": 9.2185,
|
| 6870 |
+
"step": 9800
|
| 6871 |
+
},
|
| 6872 |
+
{
|
| 6873 |
+
"epoch": 0.481245056016483,
|
| 6874 |
+
"grad_norm": 1.046875,
|
| 6875 |
+
"learning_rate": 9.889325242984365e-05,
|
| 6876 |
+
"loss": 9.7809,
|
| 6877 |
+
"step": 9810
|
| 6878 |
+
},
|
| 6879 |
+
{
|
| 6880 |
+
"epoch": 0.48173562182281987,
|
| 6881 |
+
"grad_norm": 0.9765625,
|
| 6882 |
+
"learning_rate": 9.888727291691584e-05,
|
| 6883 |
+
"loss": 9.2243,
|
| 6884 |
+
"step": 9820
|
| 6885 |
+
},
|
| 6886 |
+
{
|
| 6887 |
+
"epoch": 0.48222618762915676,
|
| 6888 |
+
"grad_norm": 1.6875,
|
| 6889 |
+
"learning_rate": 9.888127747623637e-05,
|
| 6890 |
+
"loss": 9.6819,
|
| 6891 |
+
"step": 9830
|
| 6892 |
+
},
|
| 6893 |
+
{
|
| 6894 |
+
"epoch": 0.48271675343549364,
|
| 6895 |
+
"grad_norm": 1.7578125,
|
| 6896 |
+
"learning_rate": 9.887526610975857e-05,
|
| 6897 |
+
"loss": 9.172,
|
| 6898 |
+
"step": 9840
|
| 6899 |
+
},
|
| 6900 |
+
{
|
| 6901 |
+
"epoch": 0.4832073192418305,
|
| 6902 |
+
"grad_norm": 1.1171875,
|
| 6903 |
+
"learning_rate": 9.8869238819441e-05,
|
| 6904 |
+
"loss": 9.2786,
|
| 6905 |
+
"step": 9850
|
| 6906 |
+
},
|
| 6907 |
+
{
|
| 6908 |
+
"epoch": 0.4836978850481674,
|
| 6909 |
+
"grad_norm": 1.9296875,
|
| 6910 |
+
"learning_rate": 9.886319560724735e-05,
|
| 6911 |
+
"loss": 9.4504,
|
| 6912 |
+
"step": 9860
|
| 6913 |
+
},
|
| 6914 |
+
{
|
| 6915 |
+
"epoch": 0.4841884508545043,
|
| 6916 |
+
"grad_norm": 1.6640625,
|
| 6917 |
+
"learning_rate": 9.885713647514658e-05,
|
| 6918 |
+
"loss": 10.0407,
|
| 6919 |
+
"step": 9870
|
| 6920 |
+
},
|
| 6921 |
+
{
|
| 6922 |
+
"epoch": 0.4846790166608412,
|
| 6923 |
+
"grad_norm": 0.9765625,
|
| 6924 |
+
"learning_rate": 9.885106142511275e-05,
|
| 6925 |
+
"loss": 8.9902,
|
| 6926 |
+
"step": 9880
|
| 6927 |
+
},
|
| 6928 |
+
{
|
| 6929 |
+
"epoch": 0.48516958246717806,
|
| 6930 |
+
"grad_norm": 1.0234375,
|
| 6931 |
+
"learning_rate": 9.884497045912515e-05,
|
| 6932 |
+
"loss": 8.8224,
|
| 6933 |
+
"step": 9890
|
| 6934 |
+
},
|
| 6935 |
+
{
|
| 6936 |
+
"epoch": 0.48566014827351495,
|
| 6937 |
+
"grad_norm": 1.4453125,
|
| 6938 |
+
"learning_rate": 9.883886357916828e-05,
|
| 6939 |
+
"loss": 9.893,
|
| 6940 |
+
"step": 9900
|
| 6941 |
+
},
|
| 6942 |
+
{
|
| 6943 |
+
"epoch": 0.48615071407985183,
|
| 6944 |
+
"grad_norm": 1.2890625,
|
| 6945 |
+
"learning_rate": 9.883274078723177e-05,
|
| 6946 |
+
"loss": 9.0176,
|
| 6947 |
+
"step": 9910
|
| 6948 |
+
},
|
| 6949 |
+
{
|
| 6950 |
+
"epoch": 0.4866412798861887,
|
| 6951 |
+
"grad_norm": 3.359375,
|
| 6952 |
+
"learning_rate": 9.882660208531046e-05,
|
| 6953 |
+
"loss": 9.8512,
|
| 6954 |
+
"step": 9920
|
| 6955 |
+
},
|
| 6956 |
+
{
|
| 6957 |
+
"epoch": 0.4871318456925256,
|
| 6958 |
+
"grad_norm": 1.1640625,
|
| 6959 |
+
"learning_rate": 9.882044747540439e-05,
|
| 6960 |
+
"loss": 9.3034,
|
| 6961 |
+
"step": 9930
|
| 6962 |
+
},
|
| 6963 |
+
{
|
| 6964 |
+
"epoch": 0.4876224114988625,
|
| 6965 |
+
"grad_norm": 1.265625,
|
| 6966 |
+
"learning_rate": 9.881427695951875e-05,
|
| 6967 |
+
"loss": 9.5191,
|
| 6968 |
+
"step": 9940
|
| 6969 |
+
},
|
| 6970 |
+
{
|
| 6971 |
+
"epoch": 0.48811297730519937,
|
| 6972 |
+
"grad_norm": 1.1796875,
|
| 6973 |
+
"learning_rate": 9.880809053966395e-05,
|
| 6974 |
+
"loss": 9.4207,
|
| 6975 |
+
"step": 9950
|
| 6976 |
+
},
|
| 6977 |
+
{
|
| 6978 |
+
"epoch": 0.48860354311153625,
|
| 6979 |
+
"grad_norm": 1.3828125,
|
| 6980 |
+
"learning_rate": 9.880188821785554e-05,
|
| 6981 |
+
"loss": 9.2694,
|
| 6982 |
+
"step": 9960
|
| 6983 |
+
},
|
| 6984 |
+
{
|
| 6985 |
+
"epoch": 0.48909410891787314,
|
| 6986 |
+
"grad_norm": 1.609375,
|
| 6987 |
+
"learning_rate": 9.879566999611429e-05,
|
| 6988 |
+
"loss": 9.7303,
|
| 6989 |
+
"step": 9970
|
| 6990 |
+
},
|
| 6991 |
+
{
|
| 6992 |
+
"epoch": 0.48958467472421,
|
| 6993 |
+
"grad_norm": 1.3515625,
|
| 6994 |
+
"learning_rate": 9.878943587646611e-05,
|
| 6995 |
+
"loss": 9.449,
|
| 6996 |
+
"step": 9980
|
| 6997 |
+
},
|
| 6998 |
+
{
|
| 6999 |
+
"epoch": 0.4900752405305469,
|
| 7000 |
+
"grad_norm": 1.6484375,
|
| 7001 |
+
"learning_rate": 9.878318586094213e-05,
|
| 7002 |
+
"loss": 9.3069,
|
| 7003 |
+
"step": 9990
|
| 7004 |
+
},
|
| 7005 |
+
{
|
| 7006 |
+
"epoch": 0.4905658063368838,
|
| 7007 |
+
"grad_norm": 1.2734375,
|
| 7008 |
+
"learning_rate": 9.877691995157862e-05,
|
| 7009 |
+
"loss": 9.0935,
|
| 7010 |
+
"step": 10000
|
| 7011 |
+
},
|
| 7012 |
+
{
|
| 7013 |
+
"epoch": 0.4910563721432207,
|
| 7014 |
+
"grad_norm": 1.640625,
|
| 7015 |
+
"learning_rate": 9.877063815041709e-05,
|
| 7016 |
+
"loss": 9.6356,
|
| 7017 |
+
"step": 10010
|
| 7018 |
+
},
|
| 7019 |
+
{
|
| 7020 |
+
"epoch": 0.49154693794955756,
|
| 7021 |
+
"grad_norm": 1.25,
|
| 7022 |
+
"learning_rate": 9.876434045950414e-05,
|
| 7023 |
+
"loss": 9.878,
|
| 7024 |
+
"step": 10020
|
| 7025 |
+
},
|
| 7026 |
+
{
|
| 7027 |
+
"epoch": 0.49203750375589445,
|
| 7028 |
+
"grad_norm": 1.5390625,
|
| 7029 |
+
"learning_rate": 9.875802688089163e-05,
|
| 7030 |
+
"loss": 9.4215,
|
| 7031 |
+
"step": 10030
|
| 7032 |
+
},
|
| 7033 |
+
{
|
| 7034 |
+
"epoch": 0.49252806956223133,
|
| 7035 |
+
"grad_norm": 1.25,
|
| 7036 |
+
"learning_rate": 9.875169741663653e-05,
|
| 7037 |
+
"loss": 9.108,
|
| 7038 |
+
"step": 10040
|
| 7039 |
+
},
|
| 7040 |
+
{
|
| 7041 |
+
"epoch": 0.4930186353685682,
|
| 7042 |
+
"grad_norm": 1.2734375,
|
| 7043 |
+
"learning_rate": 9.874535206880105e-05,
|
| 7044 |
+
"loss": 9.5489,
|
| 7045 |
+
"step": 10050
|
| 7046 |
+
},
|
| 7047 |
+
{
|
| 7048 |
+
"epoch": 0.4935092011749051,
|
| 7049 |
+
"grad_norm": 1.203125,
|
| 7050 |
+
"learning_rate": 9.873899083945248e-05,
|
| 7051 |
+
"loss": 9.9763,
|
| 7052 |
+
"step": 10060
|
| 7053 |
+
},
|
| 7054 |
+
{
|
| 7055 |
+
"epoch": 0.493999766981242,
|
| 7056 |
+
"grad_norm": 1.234375,
|
| 7057 |
+
"learning_rate": 9.873261373066342e-05,
|
| 7058 |
+
"loss": 9.0065,
|
| 7059 |
+
"step": 10070
|
| 7060 |
+
},
|
| 7061 |
+
{
|
| 7062 |
+
"epoch": 0.49449033278757887,
|
| 7063 |
+
"grad_norm": 1.4609375,
|
| 7064 |
+
"learning_rate": 9.872622074451153e-05,
|
| 7065 |
+
"loss": 9.4474,
|
| 7066 |
+
"step": 10080
|
| 7067 |
+
},
|
| 7068 |
+
{
|
| 7069 |
+
"epoch": 0.49498089859391575,
|
| 7070 |
+
"grad_norm": 1.359375,
|
| 7071 |
+
"learning_rate": 9.871981188307966e-05,
|
| 7072 |
+
"loss": 9.4922,
|
| 7073 |
+
"step": 10090
|
| 7074 |
+
},
|
| 7075 |
+
{
|
| 7076 |
+
"epoch": 0.49547146440025264,
|
| 7077 |
+
"grad_norm": 1.09375,
|
| 7078 |
+
"learning_rate": 9.871338714845589e-05,
|
| 7079 |
+
"loss": 9.3891,
|
| 7080 |
+
"step": 10100
|
| 7081 |
+
},
|
| 7082 |
+
{
|
| 7083 |
+
"epoch": 0.4959620302065895,
|
| 7084 |
+
"grad_norm": 0.98828125,
|
| 7085 |
+
"learning_rate": 9.870694654273342e-05,
|
| 7086 |
+
"loss": 9.0705,
|
| 7087 |
+
"step": 10110
|
| 7088 |
+
},
|
| 7089 |
+
{
|
| 7090 |
+
"epoch": 0.4964525960129264,
|
| 7091 |
+
"grad_norm": 1.1015625,
|
| 7092 |
+
"learning_rate": 9.870049006801065e-05,
|
| 7093 |
+
"loss": 9.4599,
|
| 7094 |
+
"step": 10120
|
| 7095 |
+
},
|
| 7096 |
+
{
|
| 7097 |
+
"epoch": 0.4969431618192633,
|
| 7098 |
+
"grad_norm": 1.234375,
|
| 7099 |
+
"learning_rate": 9.86940177263911e-05,
|
| 7100 |
+
"loss": 9.3144,
|
| 7101 |
+
"step": 10130
|
| 7102 |
+
},
|
| 7103 |
+
{
|
| 7104 |
+
"epoch": 0.4974337276256002,
|
| 7105 |
+
"grad_norm": 1.421875,
|
| 7106 |
+
"learning_rate": 9.868752951998355e-05,
|
| 7107 |
+
"loss": 9.1926,
|
| 7108 |
+
"step": 10140
|
| 7109 |
+
},
|
| 7110 |
+
{
|
| 7111 |
+
"epoch": 0.49792429343193706,
|
| 7112 |
+
"grad_norm": 1.0546875,
|
| 7113 |
+
"learning_rate": 9.868102545090186e-05,
|
| 7114 |
+
"loss": 9.8184,
|
| 7115 |
+
"step": 10150
|
| 7116 |
+
},
|
| 7117 |
+
{
|
| 7118 |
+
"epoch": 0.49841485923827394,
|
| 7119 |
+
"grad_norm": 1.3984375,
|
| 7120 |
+
"learning_rate": 9.867450552126506e-05,
|
| 7121 |
+
"loss": 9.8685,
|
| 7122 |
+
"step": 10160
|
| 7123 |
+
},
|
| 7124 |
+
{
|
| 7125 |
+
"epoch": 0.49890542504461083,
|
| 7126 |
+
"grad_norm": 0.84375,
|
| 7127 |
+
"learning_rate": 9.866796973319747e-05,
|
| 7128 |
+
"loss": 9.5477,
|
| 7129 |
+
"step": 10170
|
| 7130 |
+
},
|
| 7131 |
+
{
|
| 7132 |
+
"epoch": 0.4993959908509477,
|
| 7133 |
+
"grad_norm": 1.4140625,
|
| 7134 |
+
"learning_rate": 9.86614180888284e-05,
|
| 7135 |
+
"loss": 9.5079,
|
| 7136 |
+
"step": 10180
|
| 7137 |
+
},
|
| 7138 |
+
{
|
| 7139 |
+
"epoch": 0.4998865566572846,
|
| 7140 |
+
"grad_norm": 1.21875,
|
| 7141 |
+
"learning_rate": 9.865485059029246e-05,
|
| 7142 |
+
"loss": 9.8399,
|
| 7143 |
+
"step": 10190
|
| 7144 |
+
},
|
| 7145 |
+
{
|
| 7146 |
+
"epoch": 0.5003771224636214,
|
| 7147 |
+
"grad_norm": 1.015625,
|
| 7148 |
+
"learning_rate": 9.864826723972938e-05,
|
| 7149 |
+
"loss": 9.1246,
|
| 7150 |
+
"step": 10200
|
| 7151 |
+
},
|
| 7152 |
+
{
|
| 7153 |
+
"epoch": 0.5008676882699583,
|
| 7154 |
+
"grad_norm": 1.21875,
|
| 7155 |
+
"learning_rate": 9.864166803928401e-05,
|
| 7156 |
+
"loss": 8.9986,
|
| 7157 |
+
"step": 10210
|
| 7158 |
+
},
|
| 7159 |
+
{
|
| 7160 |
+
"epoch": 0.5013582540762952,
|
| 7161 |
+
"grad_norm": 1.390625,
|
| 7162 |
+
"learning_rate": 9.863505299110645e-05,
|
| 7163 |
+
"loss": 8.951,
|
| 7164 |
+
"step": 10220
|
| 7165 |
+
},
|
| 7166 |
+
{
|
| 7167 |
+
"epoch": 0.5018488198826321,
|
| 7168 |
+
"grad_norm": 1.296875,
|
| 7169 |
+
"learning_rate": 9.862842209735191e-05,
|
| 7170 |
+
"loss": 9.4867,
|
| 7171 |
+
"step": 10230
|
| 7172 |
+
},
|
| 7173 |
+
{
|
| 7174 |
+
"epoch": 0.502339385688969,
|
| 7175 |
+
"grad_norm": 1.0234375,
|
| 7176 |
+
"learning_rate": 9.862177536018075e-05,
|
| 7177 |
+
"loss": 9.658,
|
| 7178 |
+
"step": 10240
|
| 7179 |
+
},
|
| 7180 |
+
{
|
| 7181 |
+
"epoch": 0.5028299514953058,
|
| 7182 |
+
"grad_norm": 1.0625,
|
| 7183 |
+
"learning_rate": 9.861511278175857e-05,
|
| 7184 |
+
"loss": 9.2603,
|
| 7185 |
+
"step": 10250
|
| 7186 |
+
},
|
| 7187 |
+
{
|
| 7188 |
+
"epoch": 0.5033205173016427,
|
| 7189 |
+
"grad_norm": 0.8359375,
|
| 7190 |
+
"learning_rate": 9.8608434364256e-05,
|
| 7191 |
+
"loss": 9.1361,
|
| 7192 |
+
"step": 10260
|
| 7193 |
+
},
|
| 7194 |
+
{
|
| 7195 |
+
"epoch": 0.5038110831079796,
|
| 7196 |
+
"grad_norm": 1.1640625,
|
| 7197 |
+
"learning_rate": 9.860174010984898e-05,
|
| 7198 |
+
"loss": 9.3492,
|
| 7199 |
+
"step": 10270
|
| 7200 |
+
},
|
| 7201 |
+
{
|
| 7202 |
+
"epoch": 0.5043016489143165,
|
| 7203 |
+
"grad_norm": 1.140625,
|
| 7204 |
+
"learning_rate": 9.859503002071848e-05,
|
| 7205 |
+
"loss": 9.3052,
|
| 7206 |
+
"step": 10280
|
| 7207 |
+
},
|
| 7208 |
+
{
|
| 7209 |
+
"epoch": 0.5047922147206534,
|
| 7210 |
+
"grad_norm": 1.984375,
|
| 7211 |
+
"learning_rate": 9.858830409905072e-05,
|
| 7212 |
+
"loss": 9.3797,
|
| 7213 |
+
"step": 10290
|
| 7214 |
+
},
|
| 7215 |
+
{
|
| 7216 |
+
"epoch": 0.5052827805269903,
|
| 7217 |
+
"grad_norm": 1.28125,
|
| 7218 |
+
"learning_rate": 9.858156234703704e-05,
|
| 7219 |
+
"loss": 9.5779,
|
| 7220 |
+
"step": 10300
|
| 7221 |
+
},
|
| 7222 |
+
{
|
| 7223 |
+
"epoch": 0.5057733463333272,
|
| 7224 |
+
"grad_norm": 0.70703125,
|
| 7225 |
+
"learning_rate": 9.85748047668739e-05,
|
| 7226 |
+
"loss": 8.8646,
|
| 7227 |
+
"step": 10310
|
| 7228 |
+
},
|
| 7229 |
+
{
|
| 7230 |
+
"epoch": 0.506263912139664,
|
| 7231 |
+
"grad_norm": 1.546875,
|
| 7232 |
+
"learning_rate": 9.856803136076302e-05,
|
| 7233 |
+
"loss": 9.3479,
|
| 7234 |
+
"step": 10320
|
| 7235 |
+
},
|
| 7236 |
+
{
|
| 7237 |
+
"epoch": 0.5067544779460009,
|
| 7238 |
+
"grad_norm": 0.9609375,
|
| 7239 |
+
"learning_rate": 9.856124213091118e-05,
|
| 7240 |
+
"loss": 9.5916,
|
| 7241 |
+
"step": 10330
|
| 7242 |
+
},
|
| 7243 |
+
{
|
| 7244 |
+
"epoch": 0.5072450437523378,
|
| 7245 |
+
"grad_norm": 1.5859375,
|
| 7246 |
+
"learning_rate": 9.855443707953034e-05,
|
| 7247 |
+
"loss": 10.0713,
|
| 7248 |
+
"step": 10340
|
| 7249 |
+
},
|
| 7250 |
+
{
|
| 7251 |
+
"epoch": 0.5077356095586747,
|
| 7252 |
+
"grad_norm": 1.1875,
|
| 7253 |
+
"learning_rate": 9.854761620883766e-05,
|
| 7254 |
+
"loss": 9.7502,
|
| 7255 |
+
"step": 10350
|
| 7256 |
+
},
|
| 7257 |
+
{
|
| 7258 |
+
"epoch": 0.5082261753650116,
|
| 7259 |
+
"grad_norm": 1.078125,
|
| 7260 |
+
"learning_rate": 9.854077952105539e-05,
|
| 7261 |
+
"loss": 9.5492,
|
| 7262 |
+
"step": 10360
|
| 7263 |
+
},
|
| 7264 |
+
{
|
| 7265 |
+
"epoch": 0.5087167411713485,
|
| 7266 |
+
"grad_norm": 1.3671875,
|
| 7267 |
+
"learning_rate": 9.853392701841097e-05,
|
| 7268 |
+
"loss": 9.0267,
|
| 7269 |
+
"step": 10370
|
| 7270 |
+
},
|
| 7271 |
+
{
|
| 7272 |
+
"epoch": 0.5092073069776853,
|
| 7273 |
+
"grad_norm": 1.4375,
|
| 7274 |
+
"learning_rate": 9.852705870313697e-05,
|
| 7275 |
+
"loss": 9.5468,
|
| 7276 |
+
"step": 10380
|
| 7277 |
+
},
|
| 7278 |
+
{
|
| 7279 |
+
"epoch": 0.5096978727840222,
|
| 7280 |
+
"grad_norm": 1.0390625,
|
| 7281 |
+
"learning_rate": 9.852017457747116e-05,
|
| 7282 |
+
"loss": 9.4679,
|
| 7283 |
+
"step": 10390
|
| 7284 |
+
},
|
| 7285 |
+
{
|
| 7286 |
+
"epoch": 0.5101884385903591,
|
| 7287 |
+
"grad_norm": 1.375,
|
| 7288 |
+
"learning_rate": 9.851327464365639e-05,
|
| 7289 |
+
"loss": 9.2376,
|
| 7290 |
+
"step": 10400
|
| 7291 |
+
},
|
| 7292 |
+
{
|
| 7293 |
+
"epoch": 0.510679004396696,
|
| 7294 |
+
"grad_norm": 1.5546875,
|
| 7295 |
+
"learning_rate": 9.850635890394073e-05,
|
| 7296 |
+
"loss": 8.5749,
|
| 7297 |
+
"step": 10410
|
| 7298 |
+
},
|
| 7299 |
+
{
|
| 7300 |
+
"epoch": 0.5111695702030329,
|
| 7301 |
+
"grad_norm": 1.0625,
|
| 7302 |
+
"learning_rate": 9.849942736057734e-05,
|
| 7303 |
+
"loss": 8.9378,
|
| 7304 |
+
"step": 10420
|
| 7305 |
+
},
|
| 7306 |
+
{
|
| 7307 |
+
"epoch": 0.5116601360093698,
|
| 7308 |
+
"grad_norm": 1.484375,
|
| 7309 |
+
"learning_rate": 9.849248001582458e-05,
|
| 7310 |
+
"loss": 9.6109,
|
| 7311 |
+
"step": 10430
|
| 7312 |
+
},
|
| 7313 |
+
{
|
| 7314 |
+
"epoch": 0.5121507018157067,
|
| 7315 |
+
"grad_norm": 1.2890625,
|
| 7316 |
+
"learning_rate": 9.848551687194591e-05,
|
| 7317 |
+
"loss": 9.7468,
|
| 7318 |
+
"step": 10440
|
| 7319 |
+
},
|
| 7320 |
+
{
|
| 7321 |
+
"epoch": 0.5126412676220435,
|
| 7322 |
+
"grad_norm": 1.0390625,
|
| 7323 |
+
"learning_rate": 9.847853793121e-05,
|
| 7324 |
+
"loss": 9.2302,
|
| 7325 |
+
"step": 10450
|
| 7326 |
+
},
|
| 7327 |
+
{
|
| 7328 |
+
"epoch": 0.5131318334283804,
|
| 7329 |
+
"grad_norm": 0.87109375,
|
| 7330 |
+
"learning_rate": 9.847154319589058e-05,
|
| 7331 |
+
"loss": 9.3679,
|
| 7332 |
+
"step": 10460
|
| 7333 |
+
},
|
| 7334 |
+
{
|
| 7335 |
+
"epoch": 0.5136223992347173,
|
| 7336 |
+
"grad_norm": 1.59375,
|
| 7337 |
+
"learning_rate": 9.846453266826659e-05,
|
| 7338 |
+
"loss": 9.499,
|
| 7339 |
+
"step": 10470
|
| 7340 |
+
},
|
| 7341 |
+
{
|
| 7342 |
+
"epoch": 0.5141129650410542,
|
| 7343 |
+
"grad_norm": 1.5234375,
|
| 7344 |
+
"learning_rate": 9.845750635062213e-05,
|
| 7345 |
+
"loss": 9.6534,
|
| 7346 |
+
"step": 10480
|
| 7347 |
+
},
|
| 7348 |
+
{
|
| 7349 |
+
"epoch": 0.5146035308473911,
|
| 7350 |
+
"grad_norm": 1.53125,
|
| 7351 |
+
"learning_rate": 9.845046424524638e-05,
|
| 7352 |
+
"loss": 9.8588,
|
| 7353 |
+
"step": 10490
|
| 7354 |
+
},
|
| 7355 |
+
{
|
| 7356 |
+
"epoch": 0.515094096653728,
|
| 7357 |
+
"grad_norm": 1.4609375,
|
| 7358 |
+
"learning_rate": 9.844340635443372e-05,
|
| 7359 |
+
"loss": 10.2558,
|
| 7360 |
+
"step": 10500
|
| 7361 |
+
},
|
| 7362 |
+
{
|
| 7363 |
+
"epoch": 0.5155846624600648,
|
| 7364 |
+
"grad_norm": 2.40625,
|
| 7365 |
+
"learning_rate": 9.843633268048363e-05,
|
| 7366 |
+
"loss": 9.9006,
|
| 7367 |
+
"step": 10510
|
| 7368 |
+
},
|
| 7369 |
+
{
|
| 7370 |
+
"epoch": 0.5160752282664017,
|
| 7371 |
+
"grad_norm": 1.390625,
|
| 7372 |
+
"learning_rate": 9.842924322570077e-05,
|
| 7373 |
+
"loss": 9.0815,
|
| 7374 |
+
"step": 10520
|
| 7375 |
+
},
|
| 7376 |
+
{
|
| 7377 |
+
"epoch": 0.5165657940727386,
|
| 7378 |
+
"grad_norm": 1.28125,
|
| 7379 |
+
"learning_rate": 9.842213799239492e-05,
|
| 7380 |
+
"loss": 10.4063,
|
| 7381 |
+
"step": 10530
|
| 7382 |
+
},
|
| 7383 |
+
{
|
| 7384 |
+
"epoch": 0.5170563598790755,
|
| 7385 |
+
"grad_norm": 1.328125,
|
| 7386 |
+
"learning_rate": 9.8415016982881e-05,
|
| 7387 |
+
"loss": 8.8923,
|
| 7388 |
+
"step": 10540
|
| 7389 |
+
},
|
| 7390 |
+
{
|
| 7391 |
+
"epoch": 0.5175469256854124,
|
| 7392 |
+
"grad_norm": 1.8359375,
|
| 7393 |
+
"learning_rate": 9.840788019947908e-05,
|
| 7394 |
+
"loss": 9.8298,
|
| 7395 |
+
"step": 10550
|
| 7396 |
+
},
|
| 7397 |
+
{
|
| 7398 |
+
"epoch": 0.5180374914917493,
|
| 7399 |
+
"grad_norm": 1.2421875,
|
| 7400 |
+
"learning_rate": 9.840072764451438e-05,
|
| 7401 |
+
"loss": 8.764,
|
| 7402 |
+
"step": 10560
|
| 7403 |
+
},
|
| 7404 |
+
{
|
| 7405 |
+
"epoch": 0.5185280572980862,
|
| 7406 |
+
"grad_norm": 1.765625,
|
| 7407 |
+
"learning_rate": 9.839355932031721e-05,
|
| 7408 |
+
"loss": 9.384,
|
| 7409 |
+
"step": 10570
|
| 7410 |
+
},
|
| 7411 |
+
{
|
| 7412 |
+
"epoch": 0.519018623104423,
|
| 7413 |
+
"grad_norm": 1.6015625,
|
| 7414 |
+
"learning_rate": 9.838637522922308e-05,
|
| 7415 |
+
"loss": 9.458,
|
| 7416 |
+
"step": 10580
|
| 7417 |
+
},
|
| 7418 |
+
{
|
| 7419 |
+
"epoch": 0.5195091889107599,
|
| 7420 |
+
"grad_norm": 1.265625,
|
| 7421 |
+
"learning_rate": 9.837917537357257e-05,
|
| 7422 |
+
"loss": 9.5669,
|
| 7423 |
+
"step": 10590
|
| 7424 |
+
},
|
| 7425 |
+
{
|
| 7426 |
+
"epoch": 0.5199997547170968,
|
| 7427 |
+
"grad_norm": 1.3203125,
|
| 7428 |
+
"learning_rate": 9.83719597557115e-05,
|
| 7429 |
+
"loss": 9.0019,
|
| 7430 |
+
"step": 10600
|
| 7431 |
+
},
|
| 7432 |
+
{
|
| 7433 |
+
"epoch": 0.5204903205234337,
|
| 7434 |
+
"grad_norm": 1.484375,
|
| 7435 |
+
"learning_rate": 9.836472837799069e-05,
|
| 7436 |
+
"loss": 9.708,
|
| 7437 |
+
"step": 10610
|
| 7438 |
+
},
|
| 7439 |
+
{
|
| 7440 |
+
"epoch": 0.5209808863297706,
|
| 7441 |
+
"grad_norm": 1.46875,
|
| 7442 |
+
"learning_rate": 9.83574812427662e-05,
|
| 7443 |
+
"loss": 9.3858,
|
| 7444 |
+
"step": 10620
|
| 7445 |
+
},
|
| 7446 |
+
{
|
| 7447 |
+
"epoch": 0.5214714521361075,
|
| 7448 |
+
"grad_norm": 1.8984375,
|
| 7449 |
+
"learning_rate": 9.835021835239918e-05,
|
| 7450 |
+
"loss": 8.7211,
|
| 7451 |
+
"step": 10630
|
| 7452 |
+
},
|
| 7453 |
+
{
|
| 7454 |
+
"epoch": 0.5219620179424443,
|
| 7455 |
+
"grad_norm": 1.09375,
|
| 7456 |
+
"learning_rate": 9.834293970925593e-05,
|
| 7457 |
+
"loss": 9.141,
|
| 7458 |
+
"step": 10640
|
| 7459 |
+
},
|
| 7460 |
+
{
|
| 7461 |
+
"epoch": 0.5224525837487812,
|
| 7462 |
+
"grad_norm": 1.640625,
|
| 7463 |
+
"learning_rate": 9.833564531570785e-05,
|
| 7464 |
+
"loss": 9.6482,
|
| 7465 |
+
"step": 10650
|
| 7466 |
+
},
|
| 7467 |
+
{
|
| 7468 |
+
"epoch": 0.5229431495551181,
|
| 7469 |
+
"grad_norm": 1.5078125,
|
| 7470 |
+
"learning_rate": 9.832833517413153e-05,
|
| 7471 |
+
"loss": 8.9195,
|
| 7472 |
+
"step": 10660
|
| 7473 |
+
},
|
| 7474 |
+
{
|
| 7475 |
+
"epoch": 0.523433715361455,
|
| 7476 |
+
"grad_norm": 1.125,
|
| 7477 |
+
"learning_rate": 9.832100928690864e-05,
|
| 7478 |
+
"loss": 9.2983,
|
| 7479 |
+
"step": 10670
|
| 7480 |
+
},
|
| 7481 |
+
{
|
| 7482 |
+
"epoch": 0.5239242811677919,
|
| 7483 |
+
"grad_norm": 1.140625,
|
| 7484 |
+
"learning_rate": 9.831366765642597e-05,
|
| 7485 |
+
"loss": 9.4973,
|
| 7486 |
+
"step": 10680
|
| 7487 |
+
},
|
| 7488 |
+
{
|
| 7489 |
+
"epoch": 0.5244148469741288,
|
| 7490 |
+
"grad_norm": 1.4375,
|
| 7491 |
+
"learning_rate": 9.830631028507551e-05,
|
| 7492 |
+
"loss": 10.2198,
|
| 7493 |
+
"step": 10690
|
| 7494 |
+
},
|
| 7495 |
+
{
|
| 7496 |
+
"epoch": 0.5249054127804657,
|
| 7497 |
+
"grad_norm": 1.4609375,
|
| 7498 |
+
"learning_rate": 9.829893717525428e-05,
|
| 7499 |
+
"loss": 8.9836,
|
| 7500 |
+
"step": 10700
|
| 7501 |
+
},
|
| 7502 |
+
{
|
| 7503 |
+
"epoch": 0.5253959785868025,
|
| 7504 |
+
"grad_norm": 1.0078125,
|
| 7505 |
+
"learning_rate": 9.829154832936454e-05,
|
| 7506 |
+
"loss": 9.1102,
|
| 7507 |
+
"step": 10710
|
| 7508 |
+
},
|
| 7509 |
+
{
|
| 7510 |
+
"epoch": 0.5258865443931394,
|
| 7511 |
+
"grad_norm": 1.65625,
|
| 7512 |
+
"learning_rate": 9.828414374981359e-05,
|
| 7513 |
+
"loss": 9.444,
|
| 7514 |
+
"step": 10720
|
| 7515 |
+
},
|
| 7516 |
+
{
|
| 7517 |
+
"epoch": 0.5263771101994763,
|
| 7518 |
+
"grad_norm": 1.3515625,
|
| 7519 |
+
"learning_rate": 9.827672343901386e-05,
|
| 7520 |
+
"loss": 8.9337,
|
| 7521 |
+
"step": 10730
|
| 7522 |
+
},
|
| 7523 |
+
{
|
| 7524 |
+
"epoch": 0.5268676760058132,
|
| 7525 |
+
"grad_norm": 1.2578125,
|
| 7526 |
+
"learning_rate": 9.826928739938297e-05,
|
| 7527 |
+
"loss": 9.6987,
|
| 7528 |
+
"step": 10740
|
| 7529 |
+
},
|
| 7530 |
+
{
|
| 7531 |
+
"epoch": 0.5273582418121501,
|
| 7532 |
+
"grad_norm": 1.2890625,
|
| 7533 |
+
"learning_rate": 9.826183563334361e-05,
|
| 7534 |
+
"loss": 9.2521,
|
| 7535 |
+
"step": 10750
|
| 7536 |
+
},
|
| 7537 |
+
{
|
| 7538 |
+
"epoch": 0.527848807618487,
|
| 7539 |
+
"grad_norm": 0.984375,
|
| 7540 |
+
"learning_rate": 9.825436814332359e-05,
|
| 7541 |
+
"loss": 9.3628,
|
| 7542 |
+
"step": 10760
|
| 7543 |
+
},
|
| 7544 |
+
{
|
| 7545 |
+
"epoch": 0.5283393734248238,
|
| 7546 |
+
"grad_norm": 0.7421875,
|
| 7547 |
+
"learning_rate": 9.824688493175589e-05,
|
| 7548 |
+
"loss": 9.5574,
|
| 7549 |
+
"step": 10770
|
| 7550 |
+
},
|
| 7551 |
+
{
|
| 7552 |
+
"epoch": 0.5288299392311607,
|
| 7553 |
+
"grad_norm": 1.265625,
|
| 7554 |
+
"learning_rate": 9.823938600107856e-05,
|
| 7555 |
+
"loss": 8.8906,
|
| 7556 |
+
"step": 10780
|
| 7557 |
+
},
|
| 7558 |
+
{
|
| 7559 |
+
"epoch": 0.5293205050374976,
|
| 7560 |
+
"grad_norm": 0.9296875,
|
| 7561 |
+
"learning_rate": 9.823187135373481e-05,
|
| 7562 |
+
"loss": 9.2323,
|
| 7563 |
+
"step": 10790
|
| 7564 |
+
},
|
| 7565 |
+
{
|
| 7566 |
+
"epoch": 0.5298110708438345,
|
| 7567 |
+
"grad_norm": 1.109375,
|
| 7568 |
+
"learning_rate": 9.822434099217295e-05,
|
| 7569 |
+
"loss": 9.0068,
|
| 7570 |
+
"step": 10800
|
| 7571 |
+
},
|
| 7572 |
+
{
|
| 7573 |
+
"epoch": 0.5303016366501714,
|
| 7574 |
+
"grad_norm": 1.0234375,
|
| 7575 |
+
"learning_rate": 9.821679491884641e-05,
|
| 7576 |
+
"loss": 9.5788,
|
| 7577 |
+
"step": 10810
|
| 7578 |
+
},
|
| 7579 |
+
{
|
| 7580 |
+
"epoch": 0.5307922024565083,
|
| 7581 |
+
"grad_norm": 1.1171875,
|
| 7582 |
+
"learning_rate": 9.820923313621374e-05,
|
| 7583 |
+
"loss": 9.619,
|
| 7584 |
+
"step": 10820
|
| 7585 |
+
},
|
| 7586 |
+
{
|
| 7587 |
+
"epoch": 0.5312827682628452,
|
| 7588 |
+
"grad_norm": 1.046875,
|
| 7589 |
+
"learning_rate": 9.820165564673862e-05,
|
| 7590 |
+
"loss": 9.9449,
|
| 7591 |
+
"step": 10830
|
| 7592 |
+
},
|
| 7593 |
+
{
|
| 7594 |
+
"epoch": 0.531773334069182,
|
| 7595 |
+
"grad_norm": 1.2734375,
|
| 7596 |
+
"learning_rate": 9.819406245288984e-05,
|
| 7597 |
+
"loss": 9.6994,
|
| 7598 |
+
"step": 10840
|
| 7599 |
+
},
|
| 7600 |
+
{
|
| 7601 |
+
"epoch": 0.5322638998755189,
|
| 7602 |
+
"grad_norm": 1.1953125,
|
| 7603 |
+
"learning_rate": 9.81864535571413e-05,
|
| 7604 |
+
"loss": 9.9071,
|
| 7605 |
+
"step": 10850
|
| 7606 |
+
},
|
| 7607 |
+
{
|
| 7608 |
+
"epoch": 0.5327544656818558,
|
| 7609 |
+
"grad_norm": 1.3984375,
|
| 7610 |
+
"learning_rate": 9.817882896197203e-05,
|
| 7611 |
+
"loss": 9.1392,
|
| 7612 |
+
"step": 10860
|
| 7613 |
+
},
|
| 7614 |
+
{
|
| 7615 |
+
"epoch": 0.5332450314881927,
|
| 7616 |
+
"grad_norm": 1.140625,
|
| 7617 |
+
"learning_rate": 9.817118866986615e-05,
|
| 7618 |
+
"loss": 9.3059,
|
| 7619 |
+
"step": 10870
|
| 7620 |
+
},
|
| 7621 |
+
{
|
| 7622 |
+
"epoch": 0.5337355972945296,
|
| 7623 |
+
"grad_norm": 1.2109375,
|
| 7624 |
+
"learning_rate": 9.816353268331293e-05,
|
| 7625 |
+
"loss": 9.3637,
|
| 7626 |
+
"step": 10880
|
| 7627 |
+
},
|
| 7628 |
+
{
|
| 7629 |
+
"epoch": 0.5342261631008665,
|
| 7630 |
+
"grad_norm": 1.140625,
|
| 7631 |
+
"learning_rate": 9.815586100480674e-05,
|
| 7632 |
+
"loss": 8.9703,
|
| 7633 |
+
"step": 10890
|
| 7634 |
+
},
|
| 7635 |
+
{
|
| 7636 |
+
"epoch": 0.5347167289072033,
|
| 7637 |
+
"grad_norm": 1.5703125,
|
| 7638 |
+
"learning_rate": 9.8148173636847e-05,
|
| 7639 |
+
"loss": 10.3108,
|
| 7640 |
+
"step": 10900
|
| 7641 |
+
},
|
| 7642 |
+
{
|
| 7643 |
+
"epoch": 0.5352072947135402,
|
| 7644 |
+
"grad_norm": 1.2578125,
|
| 7645 |
+
"learning_rate": 9.814047058193839e-05,
|
| 7646 |
+
"loss": 9.5247,
|
| 7647 |
+
"step": 10910
|
| 7648 |
+
},
|
| 7649 |
+
{
|
| 7650 |
+
"epoch": 0.5356978605198771,
|
| 7651 |
+
"grad_norm": 1.15625,
|
| 7652 |
+
"learning_rate": 9.813275184259052e-05,
|
| 7653 |
+
"loss": 9.4193,
|
| 7654 |
+
"step": 10920
|
| 7655 |
+
},
|
| 7656 |
+
{
|
| 7657 |
+
"epoch": 0.536188426326214,
|
| 7658 |
+
"grad_norm": 0.94921875,
|
| 7659 |
+
"learning_rate": 9.812501742131825e-05,
|
| 7660 |
+
"loss": 9.0993,
|
| 7661 |
+
"step": 10930
|
| 7662 |
+
},
|
| 7663 |
+
{
|
| 7664 |
+
"epoch": 0.5366789921325509,
|
| 7665 |
+
"grad_norm": 1.234375,
|
| 7666 |
+
"learning_rate": 9.81172673206415e-05,
|
| 7667 |
+
"loss": 9.294,
|
| 7668 |
+
"step": 10940
|
| 7669 |
+
},
|
| 7670 |
+
{
|
| 7671 |
+
"epoch": 0.5371695579388878,
|
| 7672 |
+
"grad_norm": 14.8125,
|
| 7673 |
+
"learning_rate": 9.810950154308528e-05,
|
| 7674 |
+
"loss": 9.8395,
|
| 7675 |
+
"step": 10950
|
| 7676 |
+
},
|
| 7677 |
+
{
|
| 7678 |
+
"epoch": 0.5376601237452246,
|
| 7679 |
+
"grad_norm": 1.5234375,
|
| 7680 |
+
"learning_rate": 9.810172009117973e-05,
|
| 7681 |
+
"loss": 10.1739,
|
| 7682 |
+
"step": 10960
|
| 7683 |
+
},
|
| 7684 |
+
{
|
| 7685 |
+
"epoch": 0.5381506895515615,
|
| 7686 |
+
"grad_norm": 1.1953125,
|
| 7687 |
+
"learning_rate": 9.809392296746007e-05,
|
| 7688 |
+
"loss": 8.8649,
|
| 7689 |
+
"step": 10970
|
| 7690 |
+
},
|
| 7691 |
+
{
|
| 7692 |
+
"epoch": 0.5386412553578984,
|
| 7693 |
+
"grad_norm": 1.0703125,
|
| 7694 |
+
"learning_rate": 9.80861101744667e-05,
|
| 7695 |
+
"loss": 9.1894,
|
| 7696 |
+
"step": 10980
|
| 7697 |
+
},
|
| 7698 |
+
{
|
| 7699 |
+
"epoch": 0.5391318211642353,
|
| 7700 |
+
"grad_norm": 1.03125,
|
| 7701 |
+
"learning_rate": 9.807828171474502e-05,
|
| 7702 |
+
"loss": 9.866,
|
| 7703 |
+
"step": 10990
|
| 7704 |
+
},
|
| 7705 |
+
{
|
| 7706 |
+
"epoch": 0.5396223869705722,
|
| 7707 |
+
"grad_norm": 1.8984375,
|
| 7708 |
+
"learning_rate": 9.80704375908456e-05,
|
| 7709 |
+
"loss": 9.6843,
|
| 7710 |
+
"step": 11000
|
| 7711 |
+
},
|
| 7712 |
+
{
|
| 7713 |
+
"epoch": 0.5401129527769091,
|
| 7714 |
+
"grad_norm": 1.53125,
|
| 7715 |
+
"learning_rate": 9.80625778053241e-05,
|
| 7716 |
+
"loss": 9.3401,
|
| 7717 |
+
"step": 11010
|
| 7718 |
+
},
|
| 7719 |
+
{
|
| 7720 |
+
"epoch": 0.540603518583246,
|
| 7721 |
+
"grad_norm": 1.0546875,
|
| 7722 |
+
"learning_rate": 9.805470236074131e-05,
|
| 7723 |
+
"loss": 9.4714,
|
| 7724 |
+
"step": 11020
|
| 7725 |
+
},
|
| 7726 |
+
{
|
| 7727 |
+
"epoch": 0.5410940843895828,
|
| 7728 |
+
"grad_norm": 1.6171875,
|
| 7729 |
+
"learning_rate": 9.804681125966306e-05,
|
| 7730 |
+
"loss": 9.2955,
|
| 7731 |
+
"step": 11030
|
| 7732 |
+
},
|
| 7733 |
+
{
|
| 7734 |
+
"epoch": 0.5415846501959197,
|
| 7735 |
+
"grad_norm": 1.3671875,
|
| 7736 |
+
"learning_rate": 9.803890450466035e-05,
|
| 7737 |
+
"loss": 9.4521,
|
| 7738 |
+
"step": 11040
|
| 7739 |
+
},
|
| 7740 |
+
{
|
| 7741 |
+
"epoch": 0.5420752160022566,
|
| 7742 |
+
"grad_norm": 1.546875,
|
| 7743 |
+
"learning_rate": 9.80309820983092e-05,
|
| 7744 |
+
"loss": 10.1109,
|
| 7745 |
+
"step": 11050
|
| 7746 |
+
},
|
| 7747 |
+
{
|
| 7748 |
+
"epoch": 0.5425657818085935,
|
| 7749 |
+
"grad_norm": 1.53125,
|
| 7750 |
+
"learning_rate": 9.802304404319081e-05,
|
| 7751 |
+
"loss": 9.9055,
|
| 7752 |
+
"step": 11060
|
| 7753 |
+
},
|
| 7754 |
+
{
|
| 7755 |
+
"epoch": 0.5430563476149304,
|
| 7756 |
+
"grad_norm": 1.234375,
|
| 7757 |
+
"learning_rate": 9.801509034189144e-05,
|
| 7758 |
+
"loss": 9.3527,
|
| 7759 |
+
"step": 11070
|
| 7760 |
+
},
|
| 7761 |
+
{
|
| 7762 |
+
"epoch": 0.5435469134212673,
|
| 7763 |
+
"grad_norm": 0.87890625,
|
| 7764 |
+
"learning_rate": 9.800712099700244e-05,
|
| 7765 |
+
"loss": 9.4049,
|
| 7766 |
+
"step": 11080
|
| 7767 |
+
},
|
| 7768 |
+
{
|
| 7769 |
+
"epoch": 0.5440374792276041,
|
| 7770 |
+
"grad_norm": 1.25,
|
| 7771 |
+
"learning_rate": 9.799913601112028e-05,
|
| 7772 |
+
"loss": 9.8058,
|
| 7773 |
+
"step": 11090
|
| 7774 |
+
},
|
| 7775 |
+
{
|
| 7776 |
+
"epoch": 0.544528045033941,
|
| 7777 |
+
"grad_norm": 1.25,
|
| 7778 |
+
"learning_rate": 9.79911353868465e-05,
|
| 7779 |
+
"loss": 9.1666,
|
| 7780 |
+
"step": 11100
|
| 7781 |
+
},
|
| 7782 |
+
{
|
| 7783 |
+
"epoch": 0.5450186108402779,
|
| 7784 |
+
"grad_norm": 1.21875,
|
| 7785 |
+
"learning_rate": 9.798311912678775e-05,
|
| 7786 |
+
"loss": 9.0768,
|
| 7787 |
+
"step": 11110
|
| 7788 |
+
},
|
| 7789 |
+
{
|
| 7790 |
+
"epoch": 0.5455091766466148,
|
| 7791 |
+
"grad_norm": 0.91796875,
|
| 7792 |
+
"learning_rate": 9.797508723355578e-05,
|
| 7793 |
+
"loss": 8.8543,
|
| 7794 |
+
"step": 11120
|
| 7795 |
+
},
|
| 7796 |
+
{
|
| 7797 |
+
"epoch": 0.5459997424529517,
|
| 7798 |
+
"grad_norm": 1.9609375,
|
| 7799 |
+
"learning_rate": 9.796703970976744e-05,
|
| 7800 |
+
"loss": 9.5889,
|
| 7801 |
+
"step": 11130
|
| 7802 |
+
},
|
| 7803 |
+
{
|
| 7804 |
+
"epoch": 0.5464903082592886,
|
| 7805 |
+
"grad_norm": 1.546875,
|
| 7806 |
+
"learning_rate": 9.795897655804462e-05,
|
| 7807 |
+
"loss": 9.3185,
|
| 7808 |
+
"step": 11140
|
| 7809 |
+
},
|
| 7810 |
+
{
|
| 7811 |
+
"epoch": 0.5469808740656255,
|
| 7812 |
+
"grad_norm": 1.125,
|
| 7813 |
+
"learning_rate": 9.795089778101439e-05,
|
| 7814 |
+
"loss": 9.3281,
|
| 7815 |
+
"step": 11150
|
| 7816 |
+
},
|
| 7817 |
+
{
|
| 7818 |
+
"epoch": 0.5474714398719623,
|
| 7819 |
+
"grad_norm": 1.046875,
|
| 7820 |
+
"learning_rate": 9.794280338130882e-05,
|
| 7821 |
+
"loss": 9.5044,
|
| 7822 |
+
"step": 11160
|
| 7823 |
+
},
|
| 7824 |
+
{
|
| 7825 |
+
"epoch": 0.5479620056782992,
|
| 7826 |
+
"grad_norm": 0.828125,
|
| 7827 |
+
"learning_rate": 9.793469336156513e-05,
|
| 7828 |
+
"loss": 9.1314,
|
| 7829 |
+
"step": 11170
|
| 7830 |
+
},
|
| 7831 |
+
{
|
| 7832 |
+
"epoch": 0.5484525714846361,
|
| 7833 |
+
"grad_norm": 0.984375,
|
| 7834 |
+
"learning_rate": 9.792656772442561e-05,
|
| 7835 |
+
"loss": 9.2528,
|
| 7836 |
+
"step": 11180
|
| 7837 |
+
},
|
| 7838 |
+
{
|
| 7839 |
+
"epoch": 0.548943137290973,
|
| 7840 |
+
"grad_norm": 1.4765625,
|
| 7841 |
+
"learning_rate": 9.791842647253764e-05,
|
| 7842 |
+
"loss": 9.0135,
|
| 7843 |
+
"step": 11190
|
| 7844 |
+
},
|
| 7845 |
+
{
|
| 7846 |
+
"epoch": 0.5494337030973099,
|
| 7847 |
+
"grad_norm": 1.4609375,
|
| 7848 |
+
"learning_rate": 9.791026960855369e-05,
|
| 7849 |
+
"loss": 8.8728,
|
| 7850 |
+
"step": 11200
|
| 7851 |
+
},
|
| 7852 |
+
{
|
| 7853 |
+
"epoch": 0.5499242689036468,
|
| 7854 |
+
"grad_norm": 0.90625,
|
| 7855 |
+
"learning_rate": 9.79020971351313e-05,
|
| 7856 |
+
"loss": 9.2327,
|
| 7857 |
+
"step": 11210
|
| 7858 |
+
},
|
| 7859 |
+
{
|
| 7860 |
+
"epoch": 0.5504148347099836,
|
| 7861 |
+
"grad_norm": 1.3671875,
|
| 7862 |
+
"learning_rate": 9.789390905493311e-05,
|
| 7863 |
+
"loss": 9.3236,
|
| 7864 |
+
"step": 11220
|
| 7865 |
+
},
|
| 7866 |
+
{
|
| 7867 |
+
"epoch": 0.5509054005163205,
|
| 7868 |
+
"grad_norm": 1.3046875,
|
| 7869 |
+
"learning_rate": 9.788570537062685e-05,
|
| 7870 |
+
"loss": 9.4911,
|
| 7871 |
+
"step": 11230
|
| 7872 |
+
},
|
| 7873 |
+
{
|
| 7874 |
+
"epoch": 0.5513959663226574,
|
| 7875 |
+
"grad_norm": 1.203125,
|
| 7876 |
+
"learning_rate": 9.787748608488533e-05,
|
| 7877 |
+
"loss": 8.5407,
|
| 7878 |
+
"step": 11240
|
| 7879 |
+
},
|
| 7880 |
+
{
|
| 7881 |
+
"epoch": 0.5518865321289943,
|
| 7882 |
+
"grad_norm": 1.125,
|
| 7883 |
+
"learning_rate": 9.786925120038644e-05,
|
| 7884 |
+
"loss": 9.4687,
|
| 7885 |
+
"step": 11250
|
| 7886 |
+
},
|
| 7887 |
+
{
|
| 7888 |
+
"epoch": 0.5523770979353312,
|
| 7889 |
+
"grad_norm": 0.9921875,
|
| 7890 |
+
"learning_rate": 9.786100071981313e-05,
|
| 7891 |
+
"loss": 9.7501,
|
| 7892 |
+
"step": 11260
|
| 7893 |
+
},
|
| 7894 |
+
{
|
| 7895 |
+
"epoch": 0.5528676637416681,
|
| 7896 |
+
"grad_norm": 1.3046875,
|
| 7897 |
+
"learning_rate": 9.785273464585347e-05,
|
| 7898 |
+
"loss": 9.2585,
|
| 7899 |
+
"step": 11270
|
| 7900 |
+
},
|
| 7901 |
+
{
|
| 7902 |
+
"epoch": 0.553358229548005,
|
| 7903 |
+
"grad_norm": 1.5078125,
|
| 7904 |
+
"learning_rate": 9.784445298120063e-05,
|
| 7905 |
+
"loss": 9.5025,
|
| 7906 |
+
"step": 11280
|
| 7907 |
+
},
|
| 7908 |
+
{
|
| 7909 |
+
"epoch": 0.5538487953543418,
|
| 7910 |
+
"grad_norm": 1.765625,
|
| 7911 |
+
"learning_rate": 9.783615572855274e-05,
|
| 7912 |
+
"loss": 10.4089,
|
| 7913 |
+
"step": 11290
|
| 7914 |
+
},
|
| 7915 |
+
{
|
| 7916 |
+
"epoch": 0.5543393611606787,
|
| 7917 |
+
"grad_norm": 1.09375,
|
| 7918 |
+
"learning_rate": 9.782784289061317e-05,
|
| 7919 |
+
"loss": 9.3678,
|
| 7920 |
+
"step": 11300
|
| 7921 |
+
},
|
| 7922 |
+
{
|
| 7923 |
+
"epoch": 0.5548299269670156,
|
| 7924 |
+
"grad_norm": 1.4453125,
|
| 7925 |
+
"learning_rate": 9.781951447009025e-05,
|
| 7926 |
+
"loss": 9.685,
|
| 7927 |
+
"step": 11310
|
| 7928 |
+
},
|
| 7929 |
+
{
|
| 7930 |
+
"epoch": 0.5553204927733525,
|
| 7931 |
+
"grad_norm": 1.5234375,
|
| 7932 |
+
"learning_rate": 9.781117046969743e-05,
|
| 7933 |
+
"loss": 9.4025,
|
| 7934 |
+
"step": 11320
|
| 7935 |
+
},
|
| 7936 |
+
{
|
| 7937 |
+
"epoch": 0.5558110585796894,
|
| 7938 |
+
"grad_norm": 1.5546875,
|
| 7939 |
+
"learning_rate": 9.780281089215325e-05,
|
| 7940 |
+
"loss": 9.1134,
|
| 7941 |
+
"step": 11330
|
| 7942 |
+
},
|
| 7943 |
+
{
|
| 7944 |
+
"epoch": 0.5563016243860263,
|
| 7945 |
+
"grad_norm": 1.3125,
|
| 7946 |
+
"learning_rate": 9.77944357401813e-05,
|
| 7947 |
+
"loss": 9.4538,
|
| 7948 |
+
"step": 11340
|
| 7949 |
+
},
|
| 7950 |
+
{
|
| 7951 |
+
"epoch": 0.5567921901923631,
|
| 7952 |
+
"grad_norm": 1.359375,
|
| 7953 |
+
"learning_rate": 9.778604501651024e-05,
|
| 7954 |
+
"loss": 9.0207,
|
| 7955 |
+
"step": 11350
|
| 7956 |
+
},
|
| 7957 |
+
{
|
| 7958 |
+
"epoch": 0.5572827559987,
|
| 7959 |
+
"grad_norm": 1.2109375,
|
| 7960 |
+
"learning_rate": 9.777763872387384e-05,
|
| 7961 |
+
"loss": 9.7341,
|
| 7962 |
+
"step": 11360
|
| 7963 |
+
},
|
| 7964 |
+
{
|
| 7965 |
+
"epoch": 0.5577733218050369,
|
| 7966 |
+
"grad_norm": 2.46875,
|
| 7967 |
+
"learning_rate": 9.776921686501089e-05,
|
| 7968 |
+
"loss": 9.9912,
|
| 7969 |
+
"step": 11370
|
| 7970 |
+
},
|
| 7971 |
+
{
|
| 7972 |
+
"epoch": 0.5582638876113738,
|
| 7973 |
+
"grad_norm": 1.21875,
|
| 7974 |
+
"learning_rate": 9.776077944266529e-05,
|
| 7975 |
+
"loss": 8.8371,
|
| 7976 |
+
"step": 11380
|
| 7977 |
+
},
|
| 7978 |
+
{
|
| 7979 |
+
"epoch": 0.5587544534177107,
|
| 7980 |
+
"grad_norm": 0.98828125,
|
| 7981 |
+
"learning_rate": 9.7752326459586e-05,
|
| 7982 |
+
"loss": 9.3957,
|
| 7983 |
+
"step": 11390
|
| 7984 |
+
},
|
| 7985 |
+
{
|
| 7986 |
+
"epoch": 0.5592450192240476,
|
| 7987 |
+
"grad_norm": 1.0859375,
|
| 7988 |
+
"learning_rate": 9.774385791852708e-05,
|
| 7989 |
+
"loss": 9.0049,
|
| 7990 |
+
"step": 11400
|
| 7991 |
+
},
|
| 7992 |
+
{
|
| 7993 |
+
"epoch": 0.5597355850303845,
|
| 7994 |
+
"grad_norm": 1.84375,
|
| 7995 |
+
"learning_rate": 9.773537382224758e-05,
|
| 7996 |
+
"loss": 9.448,
|
| 7997 |
+
"step": 11410
|
| 7998 |
+
},
|
| 7999 |
+
{
|
| 8000 |
+
"epoch": 0.5602261508367213,
|
| 8001 |
+
"grad_norm": 1.9140625,
|
| 8002 |
+
"learning_rate": 9.772687417351169e-05,
|
| 8003 |
+
"loss": 9.2995,
|
| 8004 |
+
"step": 11420
|
| 8005 |
+
},
|
| 8006 |
+
{
|
| 8007 |
+
"epoch": 0.5607167166430582,
|
| 8008 |
+
"grad_norm": 1.015625,
|
| 8009 |
+
"learning_rate": 9.771835897508864e-05,
|
| 8010 |
+
"loss": 8.856,
|
| 8011 |
+
"step": 11430
|
| 8012 |
+
},
|
| 8013 |
+
{
|
| 8014 |
+
"epoch": 0.5612072824493951,
|
| 8015 |
+
"grad_norm": 1.4375,
|
| 8016 |
+
"learning_rate": 9.770982822975275e-05,
|
| 8017 |
+
"loss": 9.4798,
|
| 8018 |
+
"step": 11440
|
| 8019 |
+
},
|
| 8020 |
+
{
|
| 8021 |
+
"epoch": 0.561697848255732,
|
| 8022 |
+
"grad_norm": 1.4296875,
|
| 8023 |
+
"learning_rate": 9.770128194028334e-05,
|
| 8024 |
+
"loss": 10.1526,
|
| 8025 |
+
"step": 11450
|
| 8026 |
+
},
|
| 8027 |
+
{
|
| 8028 |
+
"epoch": 0.5621884140620689,
|
| 8029 |
+
"grad_norm": 1.8203125,
|
| 8030 |
+
"learning_rate": 9.769272010946488e-05,
|
| 8031 |
+
"loss": 9.4594,
|
| 8032 |
+
"step": 11460
|
| 8033 |
+
},
|
| 8034 |
+
{
|
| 8035 |
+
"epoch": 0.5626789798684058,
|
| 8036 |
+
"grad_norm": 1.109375,
|
| 8037 |
+
"learning_rate": 9.768414274008686e-05,
|
| 8038 |
+
"loss": 8.7307,
|
| 8039 |
+
"step": 11470
|
| 8040 |
+
},
|
| 8041 |
+
{
|
| 8042 |
+
"epoch": 0.5631695456747426,
|
| 8043 |
+
"grad_norm": 0.921875,
|
| 8044 |
+
"learning_rate": 9.767554983494383e-05,
|
| 8045 |
+
"loss": 9.1459,
|
| 8046 |
+
"step": 11480
|
| 8047 |
+
},
|
| 8048 |
+
{
|
| 8049 |
+
"epoch": 0.5636601114810795,
|
| 8050 |
+
"grad_norm": 2.390625,
|
| 8051 |
+
"learning_rate": 9.766694139683539e-05,
|
| 8052 |
+
"loss": 9.6059,
|
| 8053 |
+
"step": 11490
|
| 8054 |
+
},
|
| 8055 |
+
{
|
| 8056 |
+
"epoch": 0.5641506772874164,
|
| 8057 |
+
"grad_norm": 1.5703125,
|
| 8058 |
+
"learning_rate": 9.765831742856624e-05,
|
| 8059 |
+
"loss": 9.5779,
|
| 8060 |
+
"step": 11500
|
| 8061 |
+
},
|
| 8062 |
+
{
|
| 8063 |
+
"epoch": 0.5646412430937533,
|
| 8064 |
+
"grad_norm": 1.40625,
|
| 8065 |
+
"learning_rate": 9.764967793294611e-05,
|
| 8066 |
+
"loss": 10.1103,
|
| 8067 |
+
"step": 11510
|
| 8068 |
+
},
|
| 8069 |
+
{
|
| 8070 |
+
"epoch": 0.5651318089000902,
|
| 8071 |
+
"grad_norm": 1.3359375,
|
| 8072 |
+
"learning_rate": 9.764102291278978e-05,
|
| 8073 |
+
"loss": 9.0898,
|
| 8074 |
+
"step": 11520
|
| 8075 |
+
},
|
| 8076 |
+
{
|
| 8077 |
+
"epoch": 0.5656223747064271,
|
| 8078 |
+
"grad_norm": 1.515625,
|
| 8079 |
+
"learning_rate": 9.763235237091714e-05,
|
| 8080 |
+
"loss": 9.6021,
|
| 8081 |
+
"step": 11530
|
| 8082 |
+
},
|
| 8083 |
+
{
|
| 8084 |
+
"epoch": 0.566112940512764,
|
| 8085 |
+
"grad_norm": 1.3125,
|
| 8086 |
+
"learning_rate": 9.762366631015307e-05,
|
| 8087 |
+
"loss": 9.3611,
|
| 8088 |
+
"step": 11540
|
| 8089 |
+
},
|
| 8090 |
+
{
|
| 8091 |
+
"epoch": 0.5666035063191008,
|
| 8092 |
+
"grad_norm": 1.1640625,
|
| 8093 |
+
"learning_rate": 9.761496473332756e-05,
|
| 8094 |
+
"loss": 8.6628,
|
| 8095 |
+
"step": 11550
|
| 8096 |
+
},
|
| 8097 |
+
{
|
| 8098 |
+
"epoch": 0.5670940721254377,
|
| 8099 |
+
"grad_norm": 2.765625,
|
| 8100 |
+
"learning_rate": 9.760624764327564e-05,
|
| 8101 |
+
"loss": 9.3294,
|
| 8102 |
+
"step": 11560
|
| 8103 |
+
},
|
| 8104 |
+
{
|
| 8105 |
+
"epoch": 0.5675846379317746,
|
| 8106 |
+
"grad_norm": 1.2421875,
|
| 8107 |
+
"learning_rate": 9.759751504283735e-05,
|
| 8108 |
+
"loss": 9.7098,
|
| 8109 |
+
"step": 11570
|
| 8110 |
+
},
|
| 8111 |
+
{
|
| 8112 |
+
"epoch": 0.5680752037381115,
|
| 8113 |
+
"grad_norm": 1.1484375,
|
| 8114 |
+
"learning_rate": 9.758876693485785e-05,
|
| 8115 |
+
"loss": 9.3732,
|
| 8116 |
+
"step": 11580
|
| 8117 |
+
},
|
| 8118 |
+
{
|
| 8119 |
+
"epoch": 0.5685657695444484,
|
| 8120 |
+
"grad_norm": 1.0234375,
|
| 8121 |
+
"learning_rate": 9.758000332218732e-05,
|
| 8122 |
+
"loss": 9.2857,
|
| 8123 |
+
"step": 11590
|
| 8124 |
+
},
|
| 8125 |
+
{
|
| 8126 |
+
"epoch": 0.5690563353507853,
|
| 8127 |
+
"grad_norm": 1.171875,
|
| 8128 |
+
"learning_rate": 9.757122420768099e-05,
|
| 8129 |
+
"loss": 8.9375,
|
| 8130 |
+
"step": 11600
|
| 8131 |
+
},
|
| 8132 |
+
{
|
| 8133 |
+
"epoch": 0.5695469011571221,
|
| 8134 |
+
"grad_norm": 1.765625,
|
| 8135 |
+
"learning_rate": 9.756242959419915e-05,
|
| 8136 |
+
"loss": 9.5477,
|
| 8137 |
+
"step": 11610
|
| 8138 |
+
},
|
| 8139 |
+
{
|
| 8140 |
+
"epoch": 0.570037466963459,
|
| 8141 |
+
"grad_norm": 1.3125,
|
| 8142 |
+
"learning_rate": 9.755361948460713e-05,
|
| 8143 |
+
"loss": 9.0618,
|
| 8144 |
+
"step": 11620
|
| 8145 |
+
},
|
| 8146 |
+
{
|
| 8147 |
+
"epoch": 0.5705280327697959,
|
| 8148 |
+
"grad_norm": 2.09375,
|
| 8149 |
+
"learning_rate": 9.754479388177531e-05,
|
| 8150 |
+
"loss": 9.4139,
|
| 8151 |
+
"step": 11630
|
| 8152 |
+
},
|
| 8153 |
+
{
|
| 8154 |
+
"epoch": 0.5710185985761328,
|
| 8155 |
+
"grad_norm": 1.2265625,
|
| 8156 |
+
"learning_rate": 9.753595278857912e-05,
|
| 8157 |
+
"loss": 9.6154,
|
| 8158 |
+
"step": 11640
|
| 8159 |
+
},
|
| 8160 |
+
{
|
| 8161 |
+
"epoch": 0.5715091643824697,
|
| 8162 |
+
"grad_norm": 1.0078125,
|
| 8163 |
+
"learning_rate": 9.752709620789906e-05,
|
| 8164 |
+
"loss": 8.8043,
|
| 8165 |
+
"step": 11650
|
| 8166 |
+
},
|
| 8167 |
+
{
|
| 8168 |
+
"epoch": 0.5719997301888066,
|
| 8169 |
+
"grad_norm": 1.265625,
|
| 8170 |
+
"learning_rate": 9.751822414262064e-05,
|
| 8171 |
+
"loss": 9.5509,
|
| 8172 |
+
"step": 11660
|
| 8173 |
+
},
|
| 8174 |
+
{
|
| 8175 |
+
"epoch": 0.5724902959951435,
|
| 8176 |
+
"grad_norm": 1.6875,
|
| 8177 |
+
"learning_rate": 9.750933659563444e-05,
|
| 8178 |
+
"loss": 9.223,
|
| 8179 |
+
"step": 11670
|
| 8180 |
+
},
|
| 8181 |
+
{
|
| 8182 |
+
"epoch": 0.5729808618014803,
|
| 8183 |
+
"grad_norm": 1.7890625,
|
| 8184 |
+
"learning_rate": 9.750043356983605e-05,
|
| 8185 |
+
"loss": 9.4112,
|
| 8186 |
+
"step": 11680
|
| 8187 |
+
},
|
| 8188 |
+
{
|
| 8189 |
+
"epoch": 0.5734714276078172,
|
| 8190 |
+
"grad_norm": 1.3515625,
|
| 8191 |
+
"learning_rate": 9.749151506812612e-05,
|
| 8192 |
+
"loss": 8.8999,
|
| 8193 |
+
"step": 11690
|
| 8194 |
+
},
|
| 8195 |
+
{
|
| 8196 |
+
"epoch": 0.5739619934141541,
|
| 8197 |
+
"grad_norm": 1.2265625,
|
| 8198 |
+
"learning_rate": 9.74825810934104e-05,
|
| 8199 |
+
"loss": 9.2815,
|
| 8200 |
+
"step": 11700
|
| 8201 |
+
},
|
| 8202 |
+
{
|
| 8203 |
+
"epoch": 0.574452559220491,
|
| 8204 |
+
"grad_norm": 1.0,
|
| 8205 |
+
"learning_rate": 9.747363164859959e-05,
|
| 8206 |
+
"loss": 9.3178,
|
| 8207 |
+
"step": 11710
|
| 8208 |
+
},
|
| 8209 |
+
{
|
| 8210 |
+
"epoch": 0.5749431250268279,
|
| 8211 |
+
"grad_norm": 1.1015625,
|
| 8212 |
+
"learning_rate": 9.746466673660947e-05,
|
| 8213 |
+
"loss": 9.0209,
|
| 8214 |
+
"step": 11720
|
| 8215 |
+
},
|
| 8216 |
+
{
|
| 8217 |
+
"epoch": 0.5754336908331646,
|
| 8218 |
+
"grad_norm": 2.0625,
|
| 8219 |
+
"learning_rate": 9.745568636036087e-05,
|
| 8220 |
+
"loss": 9.3736,
|
| 8221 |
+
"step": 11730
|
| 8222 |
+
},
|
| 8223 |
+
{
|
| 8224 |
+
"epoch": 0.5759242566395015,
|
| 8225 |
+
"grad_norm": 1.140625,
|
| 8226 |
+
"learning_rate": 9.744669052277966e-05,
|
| 8227 |
+
"loss": 9.5949,
|
| 8228 |
+
"step": 11740
|
| 8229 |
+
},
|
| 8230 |
+
{
|
| 8231 |
+
"epoch": 0.5764148224458384,
|
| 8232 |
+
"grad_norm": 0.91796875,
|
| 8233 |
+
"learning_rate": 9.74376792267967e-05,
|
| 8234 |
+
"loss": 9.2923,
|
| 8235 |
+
"step": 11750
|
| 8236 |
+
},
|
| 8237 |
+
{
|
| 8238 |
+
"epoch": 0.5769053882521753,
|
| 8239 |
+
"grad_norm": 1.09375,
|
| 8240 |
+
"learning_rate": 9.742865247534795e-05,
|
| 8241 |
+
"loss": 9.1704,
|
| 8242 |
+
"step": 11760
|
| 8243 |
+
},
|
| 8244 |
+
{
|
| 8245 |
+
"epoch": 0.5773959540585122,
|
| 8246 |
+
"grad_norm": 1.7734375,
|
| 8247 |
+
"learning_rate": 9.741961027137437e-05,
|
| 8248 |
+
"loss": 9.6,
|
| 8249 |
+
"step": 11770
|
| 8250 |
+
},
|
| 8251 |
+
{
|
| 8252 |
+
"epoch": 0.5778865198648491,
|
| 8253 |
+
"grad_norm": 0.9375,
|
| 8254 |
+
"learning_rate": 9.741055261782195e-05,
|
| 8255 |
+
"loss": 9.1026,
|
| 8256 |
+
"step": 11780
|
| 8257 |
+
},
|
| 8258 |
+
{
|
| 8259 |
+
"epoch": 0.578377085671186,
|
| 8260 |
+
"grad_norm": 1.2109375,
|
| 8261 |
+
"learning_rate": 9.740147951764175e-05,
|
| 8262 |
+
"loss": 9.5726,
|
| 8263 |
+
"step": 11790
|
| 8264 |
+
},
|
| 8265 |
+
{
|
| 8266 |
+
"epoch": 0.5788676514775228,
|
| 8267 |
+
"grad_norm": 1.1953125,
|
| 8268 |
+
"learning_rate": 9.739239097378981e-05,
|
| 8269 |
+
"loss": 8.7005,
|
| 8270 |
+
"step": 11800
|
| 8271 |
+
},
|
| 8272 |
+
{
|
| 8273 |
+
"epoch": 0.5793582172838597,
|
| 8274 |
+
"grad_norm": 1.1953125,
|
| 8275 |
+
"learning_rate": 9.738328698922725e-05,
|
| 8276 |
+
"loss": 9.0097,
|
| 8277 |
+
"step": 11810
|
| 8278 |
+
},
|
| 8279 |
+
{
|
| 8280 |
+
"epoch": 0.5798487830901966,
|
| 8281 |
+
"grad_norm": 1.515625,
|
| 8282 |
+
"learning_rate": 9.73741675669202e-05,
|
| 8283 |
+
"loss": 9.0148,
|
| 8284 |
+
"step": 11820
|
| 8285 |
+
},
|
| 8286 |
+
{
|
| 8287 |
+
"epoch": 0.5803393488965335,
|
| 8288 |
+
"grad_norm": 1.3359375,
|
| 8289 |
+
"learning_rate": 9.736503270983979e-05,
|
| 8290 |
+
"loss": 9.1574,
|
| 8291 |
+
"step": 11830
|
| 8292 |
+
},
|
| 8293 |
+
{
|
| 8294 |
+
"epoch": 0.5808299147028704,
|
| 8295 |
+
"grad_norm": 1.109375,
|
| 8296 |
+
"learning_rate": 9.735588242096225e-05,
|
| 8297 |
+
"loss": 9.2741,
|
| 8298 |
+
"step": 11840
|
| 8299 |
+
},
|
| 8300 |
+
{
|
| 8301 |
+
"epoch": 0.5813204805092073,
|
| 8302 |
+
"grad_norm": 1.625,
|
| 8303 |
+
"learning_rate": 9.734671670326875e-05,
|
| 8304 |
+
"loss": 9.3417,
|
| 8305 |
+
"step": 11850
|
| 8306 |
+
},
|
| 8307 |
+
{
|
| 8308 |
+
"epoch": 0.5818110463155441,
|
| 8309 |
+
"grad_norm": 1.3203125,
|
| 8310 |
+
"learning_rate": 9.733753555974558e-05,
|
| 8311 |
+
"loss": 9.5609,
|
| 8312 |
+
"step": 11860
|
| 8313 |
+
},
|
| 8314 |
+
{
|
| 8315 |
+
"epoch": 0.582301612121881,
|
| 8316 |
+
"grad_norm": 1.3515625,
|
| 8317 |
+
"learning_rate": 9.732833899338398e-05,
|
| 8318 |
+
"loss": 9.3026,
|
| 8319 |
+
"step": 11870
|
| 8320 |
+
},
|
| 8321 |
+
{
|
| 8322 |
+
"epoch": 0.5827921779282179,
|
| 8323 |
+
"grad_norm": 1.1953125,
|
| 8324 |
+
"learning_rate": 9.731912700718027e-05,
|
| 8325 |
+
"loss": 9.1469,
|
| 8326 |
+
"step": 11880
|
| 8327 |
+
},
|
| 8328 |
+
{
|
| 8329 |
+
"epoch": 0.5832827437345548,
|
| 8330 |
+
"grad_norm": 1.140625,
|
| 8331 |
+
"learning_rate": 9.730989960413572e-05,
|
| 8332 |
+
"loss": 9.0359,
|
| 8333 |
+
"step": 11890
|
| 8334 |
+
},
|
| 8335 |
+
{
|
| 8336 |
+
"epoch": 0.5837733095408917,
|
| 8337 |
+
"grad_norm": 1.515625,
|
| 8338 |
+
"learning_rate": 9.730065678725672e-05,
|
| 8339 |
+
"loss": 9.7309,
|
| 8340 |
+
"step": 11900
|
| 8341 |
+
},
|
| 8342 |
+
{
|
| 8343 |
+
"epoch": 0.5842638753472286,
|
| 8344 |
+
"grad_norm": 1.9140625,
|
| 8345 |
+
"learning_rate": 9.729139855955461e-05,
|
| 8346 |
+
"loss": 9.5244,
|
| 8347 |
+
"step": 11910
|
| 8348 |
+
},
|
| 8349 |
+
{
|
| 8350 |
+
"epoch": 0.5847544411535655,
|
| 8351 |
+
"grad_norm": 1.171875,
|
| 8352 |
+
"learning_rate": 9.728212492404578e-05,
|
| 8353 |
+
"loss": 8.8631,
|
| 8354 |
+
"step": 11920
|
| 8355 |
+
},
|
| 8356 |
+
{
|
| 8357 |
+
"epoch": 0.5852450069599023,
|
| 8358 |
+
"grad_norm": 0.78125,
|
| 8359 |
+
"learning_rate": 9.727283588375162e-05,
|
| 8360 |
+
"loss": 9.0954,
|
| 8361 |
+
"step": 11930
|
| 8362 |
+
},
|
| 8363 |
+
{
|
| 8364 |
+
"epoch": 0.5857355727662392,
|
| 8365 |
+
"grad_norm": 1.09375,
|
| 8366 |
+
"learning_rate": 9.726353144169856e-05,
|
| 8367 |
+
"loss": 9.5136,
|
| 8368 |
+
"step": 11940
|
| 8369 |
+
},
|
| 8370 |
+
{
|
| 8371 |
+
"epoch": 0.5862261385725761,
|
| 8372 |
+
"grad_norm": 1.7265625,
|
| 8373 |
+
"learning_rate": 9.725421160091805e-05,
|
| 8374 |
+
"loss": 9.4926,
|
| 8375 |
+
"step": 11950
|
| 8376 |
+
},
|
| 8377 |
+
{
|
| 8378 |
+
"epoch": 0.586716704378913,
|
| 8379 |
+
"grad_norm": 0.9375,
|
| 8380 |
+
"learning_rate": 9.724487636444653e-05,
|
| 8381 |
+
"loss": 8.8427,
|
| 8382 |
+
"step": 11960
|
| 8383 |
+
},
|
| 8384 |
+
{
|
| 8385 |
+
"epoch": 0.5872072701852499,
|
| 8386 |
+
"grad_norm": 1.53125,
|
| 8387 |
+
"learning_rate": 9.723552573532549e-05,
|
| 8388 |
+
"loss": 9.2992,
|
| 8389 |
+
"step": 11970
|
| 8390 |
+
},
|
| 8391 |
+
{
|
| 8392 |
+
"epoch": 0.5876978359915868,
|
| 8393 |
+
"grad_norm": 1.4609375,
|
| 8394 |
+
"learning_rate": 9.722615971660139e-05,
|
| 8395 |
+
"loss": 8.8732,
|
| 8396 |
+
"step": 11980
|
| 8397 |
+
},
|
| 8398 |
+
{
|
| 8399 |
+
"epoch": 0.5881884017979236,
|
| 8400 |
+
"grad_norm": 1.390625,
|
| 8401 |
+
"learning_rate": 9.721677831132576e-05,
|
| 8402 |
+
"loss": 9.6405,
|
| 8403 |
+
"step": 11990
|
| 8404 |
+
},
|
| 8405 |
+
{
|
| 8406 |
+
"epoch": 0.5886789676042605,
|
| 8407 |
+
"grad_norm": 1.328125,
|
| 8408 |
+
"learning_rate": 9.72073815225551e-05,
|
| 8409 |
+
"loss": 9.0488,
|
| 8410 |
+
"step": 12000
|
| 8411 |
}
|
| 8412 |
],
|
| 8413 |
"logging_steps": 10,
|
|
|
|
| 8427 |
"attributes": {}
|
| 8428 |
}
|
| 8429 |
},
|
| 8430 |
+
"total_flos": 3.228194593308672e+19,
|
| 8431 |
"train_batch_size": 2,
|
| 8432 |
"trial_name": null,
|
| 8433 |
"trial_params": null
|
checkpoints/{checkpoint-8000 β checkpoint-12000}/training_args.bin
RENAMED
|
File without changes
|
logs/events.out.tfevents.1752541902.c28835c37b83.8174.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:71ae173ee4aec44b839460a33e432e3a735e5f987d28c32455535ba3b815df66
|
| 3 |
+
size 47495
|
myolmoe/__pycache__/__init__.cpython-311.pyc
CHANGED
|
Binary files a/myolmoe/__pycache__/__init__.cpython-311.pyc and b/myolmoe/__pycache__/__init__.cpython-311.pyc differ
|
|
|
myolmoe/__pycache__/modeling_myolmoe.cpython-311.pyc
CHANGED
|
Binary files a/myolmoe/__pycache__/modeling_myolmoe.cpython-311.pyc and b/myolmoe/__pycache__/modeling_myolmoe.cpython-311.pyc differ
|
|
|
scripts/__pycache__/train.cpython-311.pyc
CHANGED
|
Binary files a/scripts/__pycache__/train.cpython-311.pyc and b/scripts/__pycache__/train.cpython-311.pyc differ
|
|
|