Checkpoint at step 10000
Browse files- checkpoints/{checkpoint-6000 β checkpoint-10000}/config.json +0 -0
- checkpoints/{checkpoint-6000 β checkpoint-10000}/generation_config.json +0 -0
- checkpoints/{checkpoint-6000 β checkpoint-10000}/model-00001-of-00003.safetensors +0 -0
- checkpoints/{checkpoint-6000 β checkpoint-10000}/model-00002-of-00003.safetensors +1 -1
- checkpoints/{checkpoint-6000 β checkpoint-10000}/model-00003-of-00003.safetensors +1 -1
- checkpoints/{checkpoint-6000 β checkpoint-10000}/model.safetensors.index.json +0 -0
- checkpoints/{checkpoint-6000 β checkpoint-10000}/optimizer.pt +1 -1
- checkpoints/{checkpoint-6000 β checkpoint-10000}/rng_state.pth +0 -0
- checkpoints/{checkpoint-6000 β checkpoint-10000}/scheduler.pt +1 -1
- checkpoints/{checkpoint-6000 β checkpoint-10000}/trainer_state.json +2803 -3
- checkpoints/{checkpoint-6000 β checkpoint-10000}/training_args.bin +0 -0
- logs/events.out.tfevents.1752345702.5f638d684589.1618.0 +2 -2
checkpoints/{checkpoint-6000 β checkpoint-10000}/config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/generation_config.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/model-00001-of-00003.safetensors
RENAMED
|
File without changes
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/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 sha256:95ce03424f29b5a5326f65e763bb43ccf25b2d1c2196f5f1ccb6bc94e4e0f597
|
| 3 |
size 4999439616
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/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 sha256:3ee5f2e34146a859edbe91c7eb079daefda1e8d86d170772fa7498590ef7ad45
|
| 3 |
size 3892418912
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/model.safetensors.index.json
RENAMED
|
File without changes
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/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:cfa0f04dbfcbc97d18d5a64bae0e3edc47140b0acf3379561607a2e59c3aa5ef
|
| 3 |
size 101356346
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/rng_state.pth
RENAMED
|
File without changes
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/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 sha256:4628f65a9b6630de3e3b20d025ee2a85cbc417daf8b84231b5ed59b1aa0015ad
|
| 3 |
size 1064
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/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,
|
|
@@ -4208,6 +4208,2806 @@
|
|
| 4208 |
"learning_rate": 9.808698495748856e-05,
|
| 4209 |
"loss": 9.6542,
|
| 4210 |
"step": 6000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4211 |
}
|
| 4212 |
],
|
| 4213 |
"logging_steps": 10,
|
|
@@ -4227,7 +7027,7 @@
|
|
| 4227 |
"attributes": {}
|
| 4228 |
}
|
| 4229 |
},
|
| 4230 |
-
"total_flos":
|
| 4231 |
"train_batch_size": 2,
|
| 4232 |
"trial_name": null,
|
| 4233 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.4905658063368838,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 10000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 4208 |
"learning_rate": 9.808698495748856e-05,
|
| 4209 |
"loss": 9.6542,
|
| 4210 |
"step": 6000
|
| 4211 |
+
},
|
| 4212 |
+
{
|
| 4213 |
+
"epoch": 0.29483004960846715,
|
| 4214 |
+
"grad_norm": 1.75,
|
| 4215 |
+
"learning_rate": 9.825049051667757e-05,
|
| 4216 |
+
"loss": 9.5273,
|
| 4217 |
+
"step": 6010
|
| 4218 |
+
},
|
| 4219 |
+
{
|
| 4220 |
+
"epoch": 0.29532061541480403,
|
| 4221 |
+
"grad_norm": 2.796875,
|
| 4222 |
+
"learning_rate": 9.841399607586659e-05,
|
| 4223 |
+
"loss": 9.1164,
|
| 4224 |
+
"step": 6020
|
| 4225 |
+
},
|
| 4226 |
+
{
|
| 4227 |
+
"epoch": 0.2958111812211409,
|
| 4228 |
+
"grad_norm": 1.46875,
|
| 4229 |
+
"learning_rate": 9.857750163505559e-05,
|
| 4230 |
+
"loss": 9.6354,
|
| 4231 |
+
"step": 6030
|
| 4232 |
+
},
|
| 4233 |
+
{
|
| 4234 |
+
"epoch": 0.2963017470274778,
|
| 4235 |
+
"grad_norm": 1.4921875,
|
| 4236 |
+
"learning_rate": 9.874100719424461e-05,
|
| 4237 |
+
"loss": 9.8704,
|
| 4238 |
+
"step": 6040
|
| 4239 |
+
},
|
| 4240 |
+
{
|
| 4241 |
+
"epoch": 0.2967923128338147,
|
| 4242 |
+
"grad_norm": 1.25,
|
| 4243 |
+
"learning_rate": 9.890451275343363e-05,
|
| 4244 |
+
"loss": 9.1349,
|
| 4245 |
+
"step": 6050
|
| 4246 |
+
},
|
| 4247 |
+
{
|
| 4248 |
+
"epoch": 0.29728287864015157,
|
| 4249 |
+
"grad_norm": 1.8203125,
|
| 4250 |
+
"learning_rate": 9.906801831262264e-05,
|
| 4251 |
+
"loss": 9.3073,
|
| 4252 |
+
"step": 6060
|
| 4253 |
+
},
|
| 4254 |
+
{
|
| 4255 |
+
"epoch": 0.29777344444648846,
|
| 4256 |
+
"grad_norm": 1.53125,
|
| 4257 |
+
"learning_rate": 9.923152387181165e-05,
|
| 4258 |
+
"loss": 9.479,
|
| 4259 |
+
"step": 6070
|
| 4260 |
+
},
|
| 4261 |
+
{
|
| 4262 |
+
"epoch": 0.29826401025282534,
|
| 4263 |
+
"grad_norm": 3.1875,
|
| 4264 |
+
"learning_rate": 9.939502943100067e-05,
|
| 4265 |
+
"loss": 9.6812,
|
| 4266 |
+
"step": 6080
|
| 4267 |
+
},
|
| 4268 |
+
{
|
| 4269 |
+
"epoch": 0.2987545760591622,
|
| 4270 |
+
"grad_norm": 1.3125,
|
| 4271 |
+
"learning_rate": 9.955853499018967e-05,
|
| 4272 |
+
"loss": 8.847,
|
| 4273 |
+
"step": 6090
|
| 4274 |
+
},
|
| 4275 |
+
{
|
| 4276 |
+
"epoch": 0.2992451418654991,
|
| 4277 |
+
"grad_norm": 1.609375,
|
| 4278 |
+
"learning_rate": 9.972204054937868e-05,
|
| 4279 |
+
"loss": 9.6771,
|
| 4280 |
+
"step": 6100
|
| 4281 |
+
},
|
| 4282 |
+
{
|
| 4283 |
+
"epoch": 0.299735707671836,
|
| 4284 |
+
"grad_norm": 2.296875,
|
| 4285 |
+
"learning_rate": 9.988554610856769e-05,
|
| 4286 |
+
"loss": 9.7602,
|
| 4287 |
+
"step": 6110
|
| 4288 |
+
},
|
| 4289 |
+
{
|
| 4290 |
+
"epoch": 0.3002262734781729,
|
| 4291 |
+
"grad_norm": 0.984375,
|
| 4292 |
+
"learning_rate": 9.999999926693718e-05,
|
| 4293 |
+
"loss": 9.519,
|
| 4294 |
+
"step": 6120
|
| 4295 |
+
},
|
| 4296 |
+
{
|
| 4297 |
+
"epoch": 0.30071683928450976,
|
| 4298 |
+
"grad_norm": 2.015625,
|
| 4299 |
+
"learning_rate": 9.999998623470987e-05,
|
| 4300 |
+
"loss": 9.9142,
|
| 4301 |
+
"step": 6130
|
| 4302 |
+
},
|
| 4303 |
+
{
|
| 4304 |
+
"epoch": 0.30120740509084665,
|
| 4305 |
+
"grad_norm": 1.2109375,
|
| 4306 |
+
"learning_rate": 9.999995691220257e-05,
|
| 4307 |
+
"loss": 9.5174,
|
| 4308 |
+
"step": 6140
|
| 4309 |
+
},
|
| 4310 |
+
{
|
| 4311 |
+
"epoch": 0.30169797089718353,
|
| 4312 |
+
"grad_norm": 3.046875,
|
| 4313 |
+
"learning_rate": 9.999991129942481e-05,
|
| 4314 |
+
"loss": 9.6071,
|
| 4315 |
+
"step": 6150
|
| 4316 |
+
},
|
| 4317 |
+
{
|
| 4318 |
+
"epoch": 0.3021885367035204,
|
| 4319 |
+
"grad_norm": 1.921875,
|
| 4320 |
+
"learning_rate": 9.999984939639145e-05,
|
| 4321 |
+
"loss": 9.5535,
|
| 4322 |
+
"step": 6160
|
| 4323 |
+
},
|
| 4324 |
+
{
|
| 4325 |
+
"epoch": 0.3026791025098573,
|
| 4326 |
+
"grad_norm": 1.6484375,
|
| 4327 |
+
"learning_rate": 9.99997712031227e-05,
|
| 4328 |
+
"loss": 9.0336,
|
| 4329 |
+
"step": 6170
|
| 4330 |
+
},
|
| 4331 |
+
{
|
| 4332 |
+
"epoch": 0.3031696683161942,
|
| 4333 |
+
"grad_norm": 1.3828125,
|
| 4334 |
+
"learning_rate": 9.999967671964399e-05,
|
| 4335 |
+
"loss": 9.4898,
|
| 4336 |
+
"step": 6180
|
| 4337 |
+
},
|
| 4338 |
+
{
|
| 4339 |
+
"epoch": 0.30366023412253107,
|
| 4340 |
+
"grad_norm": 1.8203125,
|
| 4341 |
+
"learning_rate": 9.999956594598612e-05,
|
| 4342 |
+
"loss": 9.3706,
|
| 4343 |
+
"step": 6190
|
| 4344 |
+
},
|
| 4345 |
+
{
|
| 4346 |
+
"epoch": 0.30415079992886795,
|
| 4347 |
+
"grad_norm": 2.875,
|
| 4348 |
+
"learning_rate": 9.99994388821852e-05,
|
| 4349 |
+
"loss": 10.0704,
|
| 4350 |
+
"step": 6200
|
| 4351 |
+
},
|
| 4352 |
+
{
|
| 4353 |
+
"epoch": 0.30464136573520484,
|
| 4354 |
+
"grad_norm": 2.0625,
|
| 4355 |
+
"learning_rate": 9.99992955282826e-05,
|
| 4356 |
+
"loss": 9.6955,
|
| 4357 |
+
"step": 6210
|
| 4358 |
+
},
|
| 4359 |
+
{
|
| 4360 |
+
"epoch": 0.3051319315415417,
|
| 4361 |
+
"grad_norm": 1.21875,
|
| 4362 |
+
"learning_rate": 9.999913588432502e-05,
|
| 4363 |
+
"loss": 9.3493,
|
| 4364 |
+
"step": 6220
|
| 4365 |
+
},
|
| 4366 |
+
{
|
| 4367 |
+
"epoch": 0.3056224973478786,
|
| 4368 |
+
"grad_norm": 194.0,
|
| 4369 |
+
"learning_rate": 9.99989599503645e-05,
|
| 4370 |
+
"loss": 9.1016,
|
| 4371 |
+
"step": 6230
|
| 4372 |
+
},
|
| 4373 |
+
{
|
| 4374 |
+
"epoch": 0.3061130631542155,
|
| 4375 |
+
"grad_norm": 1.1328125,
|
| 4376 |
+
"learning_rate": 9.999876772645835e-05,
|
| 4377 |
+
"loss": 9.5685,
|
| 4378 |
+
"step": 6240
|
| 4379 |
+
},
|
| 4380 |
+
{
|
| 4381 |
+
"epoch": 0.3066036289605524,
|
| 4382 |
+
"grad_norm": 2.28125,
|
| 4383 |
+
"learning_rate": 9.999855921266921e-05,
|
| 4384 |
+
"loss": 9.7715,
|
| 4385 |
+
"step": 6250
|
| 4386 |
+
},
|
| 4387 |
+
{
|
| 4388 |
+
"epoch": 0.30709419476688926,
|
| 4389 |
+
"grad_norm": 2.046875,
|
| 4390 |
+
"learning_rate": 9.999833440906499e-05,
|
| 4391 |
+
"loss": 10.0416,
|
| 4392 |
+
"step": 6260
|
| 4393 |
+
},
|
| 4394 |
+
{
|
| 4395 |
+
"epoch": 0.30758476057322615,
|
| 4396 |
+
"grad_norm": 1.3515625,
|
| 4397 |
+
"learning_rate": 9.999809331571895e-05,
|
| 4398 |
+
"loss": 9.3091,
|
| 4399 |
+
"step": 6270
|
| 4400 |
+
},
|
| 4401 |
+
{
|
| 4402 |
+
"epoch": 0.30807532637956303,
|
| 4403 |
+
"grad_norm": 2.1875,
|
| 4404 |
+
"learning_rate": 9.999783593270962e-05,
|
| 4405 |
+
"loss": 9.6715,
|
| 4406 |
+
"step": 6280
|
| 4407 |
+
},
|
| 4408 |
+
{
|
| 4409 |
+
"epoch": 0.3085658921858999,
|
| 4410 |
+
"grad_norm": 1.5703125,
|
| 4411 |
+
"learning_rate": 9.999756226012087e-05,
|
| 4412 |
+
"loss": 9.9269,
|
| 4413 |
+
"step": 6290
|
| 4414 |
+
},
|
| 4415 |
+
{
|
| 4416 |
+
"epoch": 0.3090564579922368,
|
| 4417 |
+
"grad_norm": 1.7421875,
|
| 4418 |
+
"learning_rate": 9.999727229804187e-05,
|
| 4419 |
+
"loss": 9.7873,
|
| 4420 |
+
"step": 6300
|
| 4421 |
+
},
|
| 4422 |
+
{
|
| 4423 |
+
"epoch": 0.3095470237985737,
|
| 4424 |
+
"grad_norm": 1.2578125,
|
| 4425 |
+
"learning_rate": 9.999696604656709e-05,
|
| 4426 |
+
"loss": 9.9722,
|
| 4427 |
+
"step": 6310
|
| 4428 |
+
},
|
| 4429 |
+
{
|
| 4430 |
+
"epoch": 0.31003758960491057,
|
| 4431 |
+
"grad_norm": 1.3828125,
|
| 4432 |
+
"learning_rate": 9.999664350579631e-05,
|
| 4433 |
+
"loss": 9.2905,
|
| 4434 |
+
"step": 6320
|
| 4435 |
+
},
|
| 4436 |
+
{
|
| 4437 |
+
"epoch": 0.31052815541124745,
|
| 4438 |
+
"grad_norm": 1.1640625,
|
| 4439 |
+
"learning_rate": 9.99963046758346e-05,
|
| 4440 |
+
"loss": 9.6233,
|
| 4441 |
+
"step": 6330
|
| 4442 |
+
},
|
| 4443 |
+
{
|
| 4444 |
+
"epoch": 0.31101872121758434,
|
| 4445 |
+
"grad_norm": 2.4375,
|
| 4446 |
+
"learning_rate": 9.999594955679236e-05,
|
| 4447 |
+
"loss": 9.7319,
|
| 4448 |
+
"step": 6340
|
| 4449 |
+
},
|
| 4450 |
+
{
|
| 4451 |
+
"epoch": 0.3115092870239212,
|
| 4452 |
+
"grad_norm": 1.8828125,
|
| 4453 |
+
"learning_rate": 9.999557814878531e-05,
|
| 4454 |
+
"loss": 10.4584,
|
| 4455 |
+
"step": 6350
|
| 4456 |
+
},
|
| 4457 |
+
{
|
| 4458 |
+
"epoch": 0.3119998528302581,
|
| 4459 |
+
"grad_norm": 1.234375,
|
| 4460 |
+
"learning_rate": 9.999519045193442e-05,
|
| 4461 |
+
"loss": 9.4544,
|
| 4462 |
+
"step": 6360
|
| 4463 |
+
},
|
| 4464 |
+
{
|
| 4465 |
+
"epoch": 0.312490418636595,
|
| 4466 |
+
"grad_norm": 1.7578125,
|
| 4467 |
+
"learning_rate": 9.999478646636602e-05,
|
| 4468 |
+
"loss": 9.6401,
|
| 4469 |
+
"step": 6370
|
| 4470 |
+
},
|
| 4471 |
+
{
|
| 4472 |
+
"epoch": 0.3129809844429319,
|
| 4473 |
+
"grad_norm": 1.7109375,
|
| 4474 |
+
"learning_rate": 9.999436619221175e-05,
|
| 4475 |
+
"loss": 9.2365,
|
| 4476 |
+
"step": 6380
|
| 4477 |
+
},
|
| 4478 |
+
{
|
| 4479 |
+
"epoch": 0.31347155024926876,
|
| 4480 |
+
"grad_norm": 1.6328125,
|
| 4481 |
+
"learning_rate": 9.999392962960851e-05,
|
| 4482 |
+
"loss": 9.4004,
|
| 4483 |
+
"step": 6390
|
| 4484 |
+
},
|
| 4485 |
+
{
|
| 4486 |
+
"epoch": 0.31396211605560564,
|
| 4487 |
+
"grad_norm": 0.8515625,
|
| 4488 |
+
"learning_rate": 9.999347677869854e-05,
|
| 4489 |
+
"loss": 9.4101,
|
| 4490 |
+
"step": 6400
|
| 4491 |
+
},
|
| 4492 |
+
{
|
| 4493 |
+
"epoch": 0.31445268186194253,
|
| 4494 |
+
"grad_norm": 1.8125,
|
| 4495 |
+
"learning_rate": 9.99930076396294e-05,
|
| 4496 |
+
"loss": 9.4732,
|
| 4497 |
+
"step": 6410
|
| 4498 |
+
},
|
| 4499 |
+
{
|
| 4500 |
+
"epoch": 0.3149432476682794,
|
| 4501 |
+
"grad_norm": 1.765625,
|
| 4502 |
+
"learning_rate": 9.999252221255393e-05,
|
| 4503 |
+
"loss": 9.2319,
|
| 4504 |
+
"step": 6420
|
| 4505 |
+
},
|
| 4506 |
+
{
|
| 4507 |
+
"epoch": 0.3154338134746163,
|
| 4508 |
+
"grad_norm": 0.9765625,
|
| 4509 |
+
"learning_rate": 9.999202049763027e-05,
|
| 4510 |
+
"loss": 9.4702,
|
| 4511 |
+
"step": 6430
|
| 4512 |
+
},
|
| 4513 |
+
{
|
| 4514 |
+
"epoch": 0.3159243792809532,
|
| 4515 |
+
"grad_norm": 0.7578125,
|
| 4516 |
+
"learning_rate": 9.99915024950219e-05,
|
| 4517 |
+
"loss": 9.2764,
|
| 4518 |
+
"step": 6440
|
| 4519 |
+
},
|
| 4520 |
+
{
|
| 4521 |
+
"epoch": 0.31641494508729007,
|
| 4522 |
+
"grad_norm": 1.1953125,
|
| 4523 |
+
"learning_rate": 9.999096820489757e-05,
|
| 4524 |
+
"loss": 9.2799,
|
| 4525 |
+
"step": 6450
|
| 4526 |
+
},
|
| 4527 |
+
{
|
| 4528 |
+
"epoch": 0.31690551089362695,
|
| 4529 |
+
"grad_norm": 2.9375,
|
| 4530 |
+
"learning_rate": 9.999041762743137e-05,
|
| 4531 |
+
"loss": 9.2077,
|
| 4532 |
+
"step": 6460
|
| 4533 |
+
},
|
| 4534 |
+
{
|
| 4535 |
+
"epoch": 0.31739607669996384,
|
| 4536 |
+
"grad_norm": 1.234375,
|
| 4537 |
+
"learning_rate": 9.998985076280268e-05,
|
| 4538 |
+
"loss": 9.0989,
|
| 4539 |
+
"step": 6470
|
| 4540 |
+
},
|
| 4541 |
+
{
|
| 4542 |
+
"epoch": 0.3178866425063007,
|
| 4543 |
+
"grad_norm": 0.8515625,
|
| 4544 |
+
"learning_rate": 9.998926761119618e-05,
|
| 4545 |
+
"loss": 9.5567,
|
| 4546 |
+
"step": 6480
|
| 4547 |
+
},
|
| 4548 |
+
{
|
| 4549 |
+
"epoch": 0.3183772083126376,
|
| 4550 |
+
"grad_norm": 1.984375,
|
| 4551 |
+
"learning_rate": 9.998866817280187e-05,
|
| 4552 |
+
"loss": 9.6093,
|
| 4553 |
+
"step": 6490
|
| 4554 |
+
},
|
| 4555 |
+
{
|
| 4556 |
+
"epoch": 0.3188677741189745,
|
| 4557 |
+
"grad_norm": 1.0234375,
|
| 4558 |
+
"learning_rate": 9.998805244781504e-05,
|
| 4559 |
+
"loss": 9.0293,
|
| 4560 |
+
"step": 6500
|
| 4561 |
+
},
|
| 4562 |
+
{
|
| 4563 |
+
"epoch": 0.3193583399253114,
|
| 4564 |
+
"grad_norm": 1.3515625,
|
| 4565 |
+
"learning_rate": 9.998742043643632e-05,
|
| 4566 |
+
"loss": 9.4566,
|
| 4567 |
+
"step": 6510
|
| 4568 |
+
},
|
| 4569 |
+
{
|
| 4570 |
+
"epoch": 0.31984890573164826,
|
| 4571 |
+
"grad_norm": 1.359375,
|
| 4572 |
+
"learning_rate": 9.99867721388716e-05,
|
| 4573 |
+
"loss": 9.5419,
|
| 4574 |
+
"step": 6520
|
| 4575 |
+
},
|
| 4576 |
+
{
|
| 4577 |
+
"epoch": 0.32033947153798514,
|
| 4578 |
+
"grad_norm": 1.2578125,
|
| 4579 |
+
"learning_rate": 9.998610755533211e-05,
|
| 4580 |
+
"loss": 9.3771,
|
| 4581 |
+
"step": 6530
|
| 4582 |
+
},
|
| 4583 |
+
{
|
| 4584 |
+
"epoch": 0.320830037344322,
|
| 4585 |
+
"grad_norm": 1.8203125,
|
| 4586 |
+
"learning_rate": 9.998542668603436e-05,
|
| 4587 |
+
"loss": 10.0576,
|
| 4588 |
+
"step": 6540
|
| 4589 |
+
},
|
| 4590 |
+
{
|
| 4591 |
+
"epoch": 0.3213206031506589,
|
| 4592 |
+
"grad_norm": 0.99609375,
|
| 4593 |
+
"learning_rate": 9.998472953120022e-05,
|
| 4594 |
+
"loss": 8.9065,
|
| 4595 |
+
"step": 6550
|
| 4596 |
+
},
|
| 4597 |
+
{
|
| 4598 |
+
"epoch": 0.3218111689569958,
|
| 4599 |
+
"grad_norm": 1.15625,
|
| 4600 |
+
"learning_rate": 9.998401609105677e-05,
|
| 4601 |
+
"loss": 9.5357,
|
| 4602 |
+
"step": 6560
|
| 4603 |
+
},
|
| 4604 |
+
{
|
| 4605 |
+
"epoch": 0.3223017347633327,
|
| 4606 |
+
"grad_norm": 1.46875,
|
| 4607 |
+
"learning_rate": 9.99832863658365e-05,
|
| 4608 |
+
"loss": 9.4754,
|
| 4609 |
+
"step": 6570
|
| 4610 |
+
},
|
| 4611 |
+
{
|
| 4612 |
+
"epoch": 0.32279230056966957,
|
| 4613 |
+
"grad_norm": 1.7890625,
|
| 4614 |
+
"learning_rate": 9.998254035577714e-05,
|
| 4615 |
+
"loss": 10.0797,
|
| 4616 |
+
"step": 6580
|
| 4617 |
+
},
|
| 4618 |
+
{
|
| 4619 |
+
"epoch": 0.32328286637600645,
|
| 4620 |
+
"grad_norm": 1.0,
|
| 4621 |
+
"learning_rate": 9.998177806112175e-05,
|
| 4622 |
+
"loss": 9.659,
|
| 4623 |
+
"step": 6590
|
| 4624 |
+
},
|
| 4625 |
+
{
|
| 4626 |
+
"epoch": 0.32377343218234333,
|
| 4627 |
+
"grad_norm": 1.4140625,
|
| 4628 |
+
"learning_rate": 9.998099948211868e-05,
|
| 4629 |
+
"loss": 9.5543,
|
| 4630 |
+
"step": 6600
|
| 4631 |
+
},
|
| 4632 |
+
{
|
| 4633 |
+
"epoch": 0.3242639979886802,
|
| 4634 |
+
"grad_norm": 1.25,
|
| 4635 |
+
"learning_rate": 9.99802046190216e-05,
|
| 4636 |
+
"loss": 9.3885,
|
| 4637 |
+
"step": 6610
|
| 4638 |
+
},
|
| 4639 |
+
{
|
| 4640 |
+
"epoch": 0.3247545637950171,
|
| 4641 |
+
"grad_norm": 1.171875,
|
| 4642 |
+
"learning_rate": 9.997939347208949e-05,
|
| 4643 |
+
"loss": 8.6886,
|
| 4644 |
+
"step": 6620
|
| 4645 |
+
},
|
| 4646 |
+
{
|
| 4647 |
+
"epoch": 0.325245129601354,
|
| 4648 |
+
"grad_norm": 0.8515625,
|
| 4649 |
+
"learning_rate": 9.997856604158662e-05,
|
| 4650 |
+
"loss": 10.1025,
|
| 4651 |
+
"step": 6630
|
| 4652 |
+
},
|
| 4653 |
+
{
|
| 4654 |
+
"epoch": 0.3257356954076909,
|
| 4655 |
+
"grad_norm": 1.8671875,
|
| 4656 |
+
"learning_rate": 9.997772232778256e-05,
|
| 4657 |
+
"loss": 9.5884,
|
| 4658 |
+
"step": 6640
|
| 4659 |
+
},
|
| 4660 |
+
{
|
| 4661 |
+
"epoch": 0.3262262612140277,
|
| 4662 |
+
"grad_norm": 1.4375,
|
| 4663 |
+
"learning_rate": 9.997686233095222e-05,
|
| 4664 |
+
"loss": 9.5029,
|
| 4665 |
+
"step": 6650
|
| 4666 |
+
},
|
| 4667 |
+
{
|
| 4668 |
+
"epoch": 0.3267168270203646,
|
| 4669 |
+
"grad_norm": 1.765625,
|
| 4670 |
+
"learning_rate": 9.997598605137577e-05,
|
| 4671 |
+
"loss": 9.418,
|
| 4672 |
+
"step": 6660
|
| 4673 |
+
},
|
| 4674 |
+
{
|
| 4675 |
+
"epoch": 0.32720739282670147,
|
| 4676 |
+
"grad_norm": 1.1640625,
|
| 4677 |
+
"learning_rate": 9.997509348933871e-05,
|
| 4678 |
+
"loss": 9.5372,
|
| 4679 |
+
"step": 6670
|
| 4680 |
+
},
|
| 4681 |
+
{
|
| 4682 |
+
"epoch": 0.32769795863303836,
|
| 4683 |
+
"grad_norm": 2.25,
|
| 4684 |
+
"learning_rate": 9.997418464513187e-05,
|
| 4685 |
+
"loss": 9.3032,
|
| 4686 |
+
"step": 6680
|
| 4687 |
+
},
|
| 4688 |
+
{
|
| 4689 |
+
"epoch": 0.32818852443937524,
|
| 4690 |
+
"grad_norm": 1.71875,
|
| 4691 |
+
"learning_rate": 9.997325951905131e-05,
|
| 4692 |
+
"loss": 9.7677,
|
| 4693 |
+
"step": 6690
|
| 4694 |
+
},
|
| 4695 |
+
{
|
| 4696 |
+
"epoch": 0.3286790902457121,
|
| 4697 |
+
"grad_norm": 2.5625,
|
| 4698 |
+
"learning_rate": 9.997231811139848e-05,
|
| 4699 |
+
"loss": 9.8891,
|
| 4700 |
+
"step": 6700
|
| 4701 |
+
},
|
| 4702 |
+
{
|
| 4703 |
+
"epoch": 0.329169656052049,
|
| 4704 |
+
"grad_norm": 1.5078125,
|
| 4705 |
+
"learning_rate": 9.997136042248008e-05,
|
| 4706 |
+
"loss": 9.475,
|
| 4707 |
+
"step": 6710
|
| 4708 |
+
},
|
| 4709 |
+
{
|
| 4710 |
+
"epoch": 0.3296602218583859,
|
| 4711 |
+
"grad_norm": 1.9453125,
|
| 4712 |
+
"learning_rate": 9.997038645260812e-05,
|
| 4713 |
+
"loss": 9.2955,
|
| 4714 |
+
"step": 6720
|
| 4715 |
+
},
|
| 4716 |
+
{
|
| 4717 |
+
"epoch": 0.3301507876647228,
|
| 4718 |
+
"grad_norm": 1.578125,
|
| 4719 |
+
"learning_rate": 9.996939620209995e-05,
|
| 4720 |
+
"loss": 10.0806,
|
| 4721 |
+
"step": 6730
|
| 4722 |
+
},
|
| 4723 |
+
{
|
| 4724 |
+
"epoch": 0.33064135347105966,
|
| 4725 |
+
"grad_norm": 1.078125,
|
| 4726 |
+
"learning_rate": 9.996838967127818e-05,
|
| 4727 |
+
"loss": 8.6916,
|
| 4728 |
+
"step": 6740
|
| 4729 |
+
},
|
| 4730 |
+
{
|
| 4731 |
+
"epoch": 0.33113191927739655,
|
| 4732 |
+
"grad_norm": 1.734375,
|
| 4733 |
+
"learning_rate": 9.996736686047075e-05,
|
| 4734 |
+
"loss": 9.1876,
|
| 4735 |
+
"step": 6750
|
| 4736 |
+
},
|
| 4737 |
+
{
|
| 4738 |
+
"epoch": 0.33162248508373343,
|
| 4739 |
+
"grad_norm": 1.2578125,
|
| 4740 |
+
"learning_rate": 9.99663277700109e-05,
|
| 4741 |
+
"loss": 9.1853,
|
| 4742 |
+
"step": 6760
|
| 4743 |
+
},
|
| 4744 |
+
{
|
| 4745 |
+
"epoch": 0.3321130508900703,
|
| 4746 |
+
"grad_norm": 1.421875,
|
| 4747 |
+
"learning_rate": 9.996527240023717e-05,
|
| 4748 |
+
"loss": 9.2008,
|
| 4749 |
+
"step": 6770
|
| 4750 |
+
},
|
| 4751 |
+
{
|
| 4752 |
+
"epoch": 0.3326036166964072,
|
| 4753 |
+
"grad_norm": 1.75,
|
| 4754 |
+
"learning_rate": 9.996420075149341e-05,
|
| 4755 |
+
"loss": 9.2655,
|
| 4756 |
+
"step": 6780
|
| 4757 |
+
},
|
| 4758 |
+
{
|
| 4759 |
+
"epoch": 0.3330941825027441,
|
| 4760 |
+
"grad_norm": 1.7265625,
|
| 4761 |
+
"learning_rate": 9.996311282412876e-05,
|
| 4762 |
+
"loss": 9.4802,
|
| 4763 |
+
"step": 6790
|
| 4764 |
+
},
|
| 4765 |
+
{
|
| 4766 |
+
"epoch": 0.33358474830908097,
|
| 4767 |
+
"grad_norm": 1.5234375,
|
| 4768 |
+
"learning_rate": 9.996200861849767e-05,
|
| 4769 |
+
"loss": 9.3536,
|
| 4770 |
+
"step": 6800
|
| 4771 |
+
},
|
| 4772 |
+
{
|
| 4773 |
+
"epoch": 0.33407531411541785,
|
| 4774 |
+
"grad_norm": 1.390625,
|
| 4775 |
+
"learning_rate": 9.99608881349599e-05,
|
| 4776 |
+
"loss": 9.6765,
|
| 4777 |
+
"step": 6810
|
| 4778 |
+
},
|
| 4779 |
+
{
|
| 4780 |
+
"epoch": 0.33456587992175474,
|
| 4781 |
+
"grad_norm": 2.484375,
|
| 4782 |
+
"learning_rate": 9.995975137388052e-05,
|
| 4783 |
+
"loss": 9.1743,
|
| 4784 |
+
"step": 6820
|
| 4785 |
+
},
|
| 4786 |
+
{
|
| 4787 |
+
"epoch": 0.3350564457280916,
|
| 4788 |
+
"grad_norm": 1.171875,
|
| 4789 |
+
"learning_rate": 9.995859833562988e-05,
|
| 4790 |
+
"loss": 8.9971,
|
| 4791 |
+
"step": 6830
|
| 4792 |
+
},
|
| 4793 |
+
{
|
| 4794 |
+
"epoch": 0.3355470115344285,
|
| 4795 |
+
"grad_norm": 2.140625,
|
| 4796 |
+
"learning_rate": 9.995742902058367e-05,
|
| 4797 |
+
"loss": 9.143,
|
| 4798 |
+
"step": 6840
|
| 4799 |
+
},
|
| 4800 |
+
{
|
| 4801 |
+
"epoch": 0.3360375773407654,
|
| 4802 |
+
"grad_norm": 1.859375,
|
| 4803 |
+
"learning_rate": 9.995624342912282e-05,
|
| 4804 |
+
"loss": 9.8555,
|
| 4805 |
+
"step": 6850
|
| 4806 |
+
},
|
| 4807 |
+
{
|
| 4808 |
+
"epoch": 0.3365281431471023,
|
| 4809 |
+
"grad_norm": 2.203125,
|
| 4810 |
+
"learning_rate": 9.995504156163364e-05,
|
| 4811 |
+
"loss": 9.659,
|
| 4812 |
+
"step": 6860
|
| 4813 |
+
},
|
| 4814 |
+
{
|
| 4815 |
+
"epoch": 0.33701870895343916,
|
| 4816 |
+
"grad_norm": 1.9765625,
|
| 4817 |
+
"learning_rate": 9.995382341850769e-05,
|
| 4818 |
+
"loss": 9.178,
|
| 4819 |
+
"step": 6870
|
| 4820 |
+
},
|
| 4821 |
+
{
|
| 4822 |
+
"epoch": 0.33750927475977605,
|
| 4823 |
+
"grad_norm": 1.4765625,
|
| 4824 |
+
"learning_rate": 9.995258900014184e-05,
|
| 4825 |
+
"loss": 9.9946,
|
| 4826 |
+
"step": 6880
|
| 4827 |
+
},
|
| 4828 |
+
{
|
| 4829 |
+
"epoch": 0.33799984056611293,
|
| 4830 |
+
"grad_norm": 1.3046875,
|
| 4831 |
+
"learning_rate": 9.995133830693828e-05,
|
| 4832 |
+
"loss": 9.1998,
|
| 4833 |
+
"step": 6890
|
| 4834 |
+
},
|
| 4835 |
+
{
|
| 4836 |
+
"epoch": 0.3384904063724498,
|
| 4837 |
+
"grad_norm": 1.2265625,
|
| 4838 |
+
"learning_rate": 9.995007133930449e-05,
|
| 4839 |
+
"loss": 9.3834,
|
| 4840 |
+
"step": 6900
|
| 4841 |
+
},
|
| 4842 |
+
{
|
| 4843 |
+
"epoch": 0.3389809721787867,
|
| 4844 |
+
"grad_norm": 1.5078125,
|
| 4845 |
+
"learning_rate": 9.994878809765327e-05,
|
| 4846 |
+
"loss": 9.5709,
|
| 4847 |
+
"step": 6910
|
| 4848 |
+
},
|
| 4849 |
+
{
|
| 4850 |
+
"epoch": 0.3394715379851236,
|
| 4851 |
+
"grad_norm": 1.5859375,
|
| 4852 |
+
"learning_rate": 9.994748858240268e-05,
|
| 4853 |
+
"loss": 9.3729,
|
| 4854 |
+
"step": 6920
|
| 4855 |
+
},
|
| 4856 |
+
{
|
| 4857 |
+
"epoch": 0.33996210379146047,
|
| 4858 |
+
"grad_norm": 1.1484375,
|
| 4859 |
+
"learning_rate": 9.994617279397612e-05,
|
| 4860 |
+
"loss": 10.3316,
|
| 4861 |
+
"step": 6930
|
| 4862 |
+
},
|
| 4863 |
+
{
|
| 4864 |
+
"epoch": 0.34045266959779735,
|
| 4865 |
+
"grad_norm": 1.8125,
|
| 4866 |
+
"learning_rate": 9.99448407328023e-05,
|
| 4867 |
+
"loss": 10.2478,
|
| 4868 |
+
"step": 6940
|
| 4869 |
+
},
|
| 4870 |
+
{
|
| 4871 |
+
"epoch": 0.34094323540413424,
|
| 4872 |
+
"grad_norm": 1.421875,
|
| 4873 |
+
"learning_rate": 9.994349239931517e-05,
|
| 4874 |
+
"loss": 9.4121,
|
| 4875 |
+
"step": 6950
|
| 4876 |
+
},
|
| 4877 |
+
{
|
| 4878 |
+
"epoch": 0.3414338012104711,
|
| 4879 |
+
"grad_norm": 1.8359375,
|
| 4880 |
+
"learning_rate": 9.994212779395409e-05,
|
| 4881 |
+
"loss": 9.5804,
|
| 4882 |
+
"step": 6960
|
| 4883 |
+
},
|
| 4884 |
+
{
|
| 4885 |
+
"epoch": 0.341924367016808,
|
| 4886 |
+
"grad_norm": 1.359375,
|
| 4887 |
+
"learning_rate": 9.99407469171636e-05,
|
| 4888 |
+
"loss": 9.8077,
|
| 4889 |
+
"step": 6970
|
| 4890 |
+
},
|
| 4891 |
+
{
|
| 4892 |
+
"epoch": 0.3424149328231449,
|
| 4893 |
+
"grad_norm": 1.6484375,
|
| 4894 |
+
"learning_rate": 9.99393497693936e-05,
|
| 4895 |
+
"loss": 9.1917,
|
| 4896 |
+
"step": 6980
|
| 4897 |
+
},
|
| 4898 |
+
{
|
| 4899 |
+
"epoch": 0.3429054986294818,
|
| 4900 |
+
"grad_norm": 1.609375,
|
| 4901 |
+
"learning_rate": 9.993793635109933e-05,
|
| 4902 |
+
"loss": 9.6918,
|
| 4903 |
+
"step": 6990
|
| 4904 |
+
},
|
| 4905 |
+
{
|
| 4906 |
+
"epoch": 0.34339606443581866,
|
| 4907 |
+
"grad_norm": 1.1015625,
|
| 4908 |
+
"learning_rate": 9.993650666274126e-05,
|
| 4909 |
+
"loss": 10.0207,
|
| 4910 |
+
"step": 7000
|
| 4911 |
+
},
|
| 4912 |
+
{
|
| 4913 |
+
"epoch": 0.34388663024215554,
|
| 4914 |
+
"grad_norm": 1.5234375,
|
| 4915 |
+
"learning_rate": 9.993506070478519e-05,
|
| 4916 |
+
"loss": 9.6955,
|
| 4917 |
+
"step": 7010
|
| 4918 |
+
},
|
| 4919 |
+
{
|
| 4920 |
+
"epoch": 0.34437719604849243,
|
| 4921 |
+
"grad_norm": 5.125,
|
| 4922 |
+
"learning_rate": 9.993359847770223e-05,
|
| 4923 |
+
"loss": 9.4934,
|
| 4924 |
+
"step": 7020
|
| 4925 |
+
},
|
| 4926 |
+
{
|
| 4927 |
+
"epoch": 0.3448677618548293,
|
| 4928 |
+
"grad_norm": 1.2421875,
|
| 4929 |
+
"learning_rate": 9.993211998196877e-05,
|
| 4930 |
+
"loss": 9.6029,
|
| 4931 |
+
"step": 7030
|
| 4932 |
+
},
|
| 4933 |
+
{
|
| 4934 |
+
"epoch": 0.3453583276611662,
|
| 4935 |
+
"grad_norm": 1.5078125,
|
| 4936 |
+
"learning_rate": 9.993062521806653e-05,
|
| 4937 |
+
"loss": 10.157,
|
| 4938 |
+
"step": 7040
|
| 4939 |
+
},
|
| 4940 |
+
{
|
| 4941 |
+
"epoch": 0.3458488934675031,
|
| 4942 |
+
"grad_norm": 1.4453125,
|
| 4943 |
+
"learning_rate": 9.99291141864825e-05,
|
| 4944 |
+
"loss": 9.3385,
|
| 4945 |
+
"step": 7050
|
| 4946 |
+
},
|
| 4947 |
+
{
|
| 4948 |
+
"epoch": 0.34633945927383997,
|
| 4949 |
+
"grad_norm": 2.125,
|
| 4950 |
+
"learning_rate": 9.9927586887709e-05,
|
| 4951 |
+
"loss": 9.7968,
|
| 4952 |
+
"step": 7060
|
| 4953 |
+
},
|
| 4954 |
+
{
|
| 4955 |
+
"epoch": 0.34683002508017685,
|
| 4956 |
+
"grad_norm": 1.3515625,
|
| 4957 |
+
"learning_rate": 9.99260433222436e-05,
|
| 4958 |
+
"loss": 9.5476,
|
| 4959 |
+
"step": 7070
|
| 4960 |
+
},
|
| 4961 |
+
{
|
| 4962 |
+
"epoch": 0.34732059088651374,
|
| 4963 |
+
"grad_norm": 1.09375,
|
| 4964 |
+
"learning_rate": 9.992448349058923e-05,
|
| 4965 |
+
"loss": 9.7711,
|
| 4966 |
+
"step": 7080
|
| 4967 |
+
},
|
| 4968 |
+
{
|
| 4969 |
+
"epoch": 0.3478111566928506,
|
| 4970 |
+
"grad_norm": 1.296875,
|
| 4971 |
+
"learning_rate": 9.992290739325408e-05,
|
| 4972 |
+
"loss": 9.3846,
|
| 4973 |
+
"step": 7090
|
| 4974 |
+
},
|
| 4975 |
+
{
|
| 4976 |
+
"epoch": 0.3483017224991875,
|
| 4977 |
+
"grad_norm": 1.5,
|
| 4978 |
+
"learning_rate": 9.992131503075166e-05,
|
| 4979 |
+
"loss": 9.5349,
|
| 4980 |
+
"step": 7100
|
| 4981 |
+
},
|
| 4982 |
+
{
|
| 4983 |
+
"epoch": 0.3487922883055244,
|
| 4984 |
+
"grad_norm": 1.5546875,
|
| 4985 |
+
"learning_rate": 9.991970640360075e-05,
|
| 4986 |
+
"loss": 9.5095,
|
| 4987 |
+
"step": 7110
|
| 4988 |
+
},
|
| 4989 |
+
{
|
| 4990 |
+
"epoch": 0.3492828541118613,
|
| 4991 |
+
"grad_norm": 1.28125,
|
| 4992 |
+
"learning_rate": 9.99180815123255e-05,
|
| 4993 |
+
"loss": 9.7473,
|
| 4994 |
+
"step": 7120
|
| 4995 |
+
},
|
| 4996 |
+
{
|
| 4997 |
+
"epoch": 0.34977341991819816,
|
| 4998 |
+
"grad_norm": 2.359375,
|
| 4999 |
+
"learning_rate": 9.991644035745524e-05,
|
| 5000 |
+
"loss": 9.3305,
|
| 5001 |
+
"step": 7130
|
| 5002 |
+
},
|
| 5003 |
+
{
|
| 5004 |
+
"epoch": 0.35026398572453504,
|
| 5005 |
+
"grad_norm": 2.390625,
|
| 5006 |
+
"learning_rate": 9.991478293952472e-05,
|
| 5007 |
+
"loss": 9.5504,
|
| 5008 |
+
"step": 7140
|
| 5009 |
+
},
|
| 5010 |
+
{
|
| 5011 |
+
"epoch": 0.3507545515308719,
|
| 5012 |
+
"grad_norm": 1.9140625,
|
| 5013 |
+
"learning_rate": 9.99131092590739e-05,
|
| 5014 |
+
"loss": 9.4789,
|
| 5015 |
+
"step": 7150
|
| 5016 |
+
},
|
| 5017 |
+
{
|
| 5018 |
+
"epoch": 0.3512451173372088,
|
| 5019 |
+
"grad_norm": 1.2734375,
|
| 5020 |
+
"learning_rate": 9.991141931664811e-05,
|
| 5021 |
+
"loss": 9.5481,
|
| 5022 |
+
"step": 7160
|
| 5023 |
+
},
|
| 5024 |
+
{
|
| 5025 |
+
"epoch": 0.3517356831435457,
|
| 5026 |
+
"grad_norm": 2.328125,
|
| 5027 |
+
"learning_rate": 9.99097131127979e-05,
|
| 5028 |
+
"loss": 9.7359,
|
| 5029 |
+
"step": 7170
|
| 5030 |
+
},
|
| 5031 |
+
{
|
| 5032 |
+
"epoch": 0.3522262489498826,
|
| 5033 |
+
"grad_norm": 1.921875,
|
| 5034 |
+
"learning_rate": 9.990799064807921e-05,
|
| 5035 |
+
"loss": 9.6452,
|
| 5036 |
+
"step": 7180
|
| 5037 |
+
},
|
| 5038 |
+
{
|
| 5039 |
+
"epoch": 0.35271681475621947,
|
| 5040 |
+
"grad_norm": 1.28125,
|
| 5041 |
+
"learning_rate": 9.990625192305321e-05,
|
| 5042 |
+
"loss": 9.7301,
|
| 5043 |
+
"step": 7190
|
| 5044 |
+
},
|
| 5045 |
+
{
|
| 5046 |
+
"epoch": 0.35320738056255635,
|
| 5047 |
+
"grad_norm": 0.81640625,
|
| 5048 |
+
"learning_rate": 9.990449693828636e-05,
|
| 5049 |
+
"loss": 9.1542,
|
| 5050 |
+
"step": 7200
|
| 5051 |
+
},
|
| 5052 |
+
{
|
| 5053 |
+
"epoch": 0.35369794636889323,
|
| 5054 |
+
"grad_norm": 1.5390625,
|
| 5055 |
+
"learning_rate": 9.990272569435048e-05,
|
| 5056 |
+
"loss": 9.6957,
|
| 5057 |
+
"step": 7210
|
| 5058 |
+
},
|
| 5059 |
+
{
|
| 5060 |
+
"epoch": 0.3541885121752301,
|
| 5061 |
+
"grad_norm": 1.4765625,
|
| 5062 |
+
"learning_rate": 9.990093819182264e-05,
|
| 5063 |
+
"loss": 10.0123,
|
| 5064 |
+
"step": 7220
|
| 5065 |
+
},
|
| 5066 |
+
{
|
| 5067 |
+
"epoch": 0.354679077981567,
|
| 5068 |
+
"grad_norm": 1.0,
|
| 5069 |
+
"learning_rate": 9.989913443128521e-05,
|
| 5070 |
+
"loss": 9.1995,
|
| 5071 |
+
"step": 7230
|
| 5072 |
+
},
|
| 5073 |
+
{
|
| 5074 |
+
"epoch": 0.3551696437879039,
|
| 5075 |
+
"grad_norm": 1.046875,
|
| 5076 |
+
"learning_rate": 9.989731441332587e-05,
|
| 5077 |
+
"loss": 9.1959,
|
| 5078 |
+
"step": 7240
|
| 5079 |
+
},
|
| 5080 |
+
{
|
| 5081 |
+
"epoch": 0.3556602095942408,
|
| 5082 |
+
"grad_norm": 1.3828125,
|
| 5083 |
+
"learning_rate": 9.989547813853758e-05,
|
| 5084 |
+
"loss": 9.8696,
|
| 5085 |
+
"step": 7250
|
| 5086 |
+
},
|
| 5087 |
+
{
|
| 5088 |
+
"epoch": 0.35615077540057766,
|
| 5089 |
+
"grad_norm": 1.6328125,
|
| 5090 |
+
"learning_rate": 9.989362560751864e-05,
|
| 5091 |
+
"loss": 9.9184,
|
| 5092 |
+
"step": 7260
|
| 5093 |
+
},
|
| 5094 |
+
{
|
| 5095 |
+
"epoch": 0.35664134120691454,
|
| 5096 |
+
"grad_norm": 1.53125,
|
| 5097 |
+
"learning_rate": 9.989175682087261e-05,
|
| 5098 |
+
"loss": 8.766,
|
| 5099 |
+
"step": 7270
|
| 5100 |
+
},
|
| 5101 |
+
{
|
| 5102 |
+
"epoch": 0.3571319070132514,
|
| 5103 |
+
"grad_norm": 1.3046875,
|
| 5104 |
+
"learning_rate": 9.988987177920832e-05,
|
| 5105 |
+
"loss": 9.919,
|
| 5106 |
+
"step": 7280
|
| 5107 |
+
},
|
| 5108 |
+
{
|
| 5109 |
+
"epoch": 0.3576224728195883,
|
| 5110 |
+
"grad_norm": 2.03125,
|
| 5111 |
+
"learning_rate": 9.988797048313995e-05,
|
| 5112 |
+
"loss": 9.6369,
|
| 5113 |
+
"step": 7290
|
| 5114 |
+
},
|
| 5115 |
+
{
|
| 5116 |
+
"epoch": 0.3581130386259252,
|
| 5117 |
+
"grad_norm": 1.484375,
|
| 5118 |
+
"learning_rate": 9.988605293328694e-05,
|
| 5119 |
+
"loss": 9.3527,
|
| 5120 |
+
"step": 7300
|
| 5121 |
+
},
|
| 5122 |
+
{
|
| 5123 |
+
"epoch": 0.3586036044322621,
|
| 5124 |
+
"grad_norm": 1.0703125,
|
| 5125 |
+
"learning_rate": 9.988411913027405e-05,
|
| 5126 |
+
"loss": 9.7346,
|
| 5127 |
+
"step": 7310
|
| 5128 |
+
},
|
| 5129 |
+
{
|
| 5130 |
+
"epoch": 0.35909417023859896,
|
| 5131 |
+
"grad_norm": 0.9921875,
|
| 5132 |
+
"learning_rate": 9.988216907473134e-05,
|
| 5133 |
+
"loss": 9.8303,
|
| 5134 |
+
"step": 7320
|
| 5135 |
+
},
|
| 5136 |
+
{
|
| 5137 |
+
"epoch": 0.35958473604493585,
|
| 5138 |
+
"grad_norm": 1.03125,
|
| 5139 |
+
"learning_rate": 9.988020276729411e-05,
|
| 5140 |
+
"loss": 9.6438,
|
| 5141 |
+
"step": 7330
|
| 5142 |
+
},
|
| 5143 |
+
{
|
| 5144 |
+
"epoch": 0.36007530185127273,
|
| 5145 |
+
"grad_norm": 1.1015625,
|
| 5146 |
+
"learning_rate": 9.987822020860303e-05,
|
| 5147 |
+
"loss": 9.2781,
|
| 5148 |
+
"step": 7340
|
| 5149 |
+
},
|
| 5150 |
+
{
|
| 5151 |
+
"epoch": 0.3605658676576096,
|
| 5152 |
+
"grad_norm": 1.3515625,
|
| 5153 |
+
"learning_rate": 9.987622139930399e-05,
|
| 5154 |
+
"loss": 9.7961,
|
| 5155 |
+
"step": 7350
|
| 5156 |
+
},
|
| 5157 |
+
{
|
| 5158 |
+
"epoch": 0.3610564334639465,
|
| 5159 |
+
"grad_norm": 2.078125,
|
| 5160 |
+
"learning_rate": 9.987420634004826e-05,
|
| 5161 |
+
"loss": 9.5671,
|
| 5162 |
+
"step": 7360
|
| 5163 |
+
},
|
| 5164 |
+
{
|
| 5165 |
+
"epoch": 0.3615469992702834,
|
| 5166 |
+
"grad_norm": 1.53125,
|
| 5167 |
+
"learning_rate": 9.987217503149233e-05,
|
| 5168 |
+
"loss": 9.6844,
|
| 5169 |
+
"step": 7370
|
| 5170 |
+
},
|
| 5171 |
+
{
|
| 5172 |
+
"epoch": 0.36203756507662027,
|
| 5173 |
+
"grad_norm": 1.8203125,
|
| 5174 |
+
"learning_rate": 9.987012747429801e-05,
|
| 5175 |
+
"loss": 10.0758,
|
| 5176 |
+
"step": 7380
|
| 5177 |
+
},
|
| 5178 |
+
{
|
| 5179 |
+
"epoch": 0.36252813088295716,
|
| 5180 |
+
"grad_norm": 1.8828125,
|
| 5181 |
+
"learning_rate": 9.986806366913241e-05,
|
| 5182 |
+
"loss": 9.2339,
|
| 5183 |
+
"step": 7390
|
| 5184 |
+
},
|
| 5185 |
+
{
|
| 5186 |
+
"epoch": 0.36301869668929404,
|
| 5187 |
+
"grad_norm": 2.1875,
|
| 5188 |
+
"learning_rate": 9.986598361666794e-05,
|
| 5189 |
+
"loss": 9.7534,
|
| 5190 |
+
"step": 7400
|
| 5191 |
+
},
|
| 5192 |
+
{
|
| 5193 |
+
"epoch": 0.3635092624956309,
|
| 5194 |
+
"grad_norm": 1.4453125,
|
| 5195 |
+
"learning_rate": 9.98638873175823e-05,
|
| 5196 |
+
"loss": 9.531,
|
| 5197 |
+
"step": 7410
|
| 5198 |
+
},
|
| 5199 |
+
{
|
| 5200 |
+
"epoch": 0.3639998283019678,
|
| 5201 |
+
"grad_norm": 1.125,
|
| 5202 |
+
"learning_rate": 9.986177477255843e-05,
|
| 5203 |
+
"loss": 9.281,
|
| 5204 |
+
"step": 7420
|
| 5205 |
+
},
|
| 5206 |
+
{
|
| 5207 |
+
"epoch": 0.36449039410830464,
|
| 5208 |
+
"grad_norm": 4.5,
|
| 5209 |
+
"learning_rate": 9.985964598228465e-05,
|
| 5210 |
+
"loss": 9.3941,
|
| 5211 |
+
"step": 7430
|
| 5212 |
+
},
|
| 5213 |
+
{
|
| 5214 |
+
"epoch": 0.3649809599146415,
|
| 5215 |
+
"grad_norm": 2.640625,
|
| 5216 |
+
"learning_rate": 9.985750094745454e-05,
|
| 5217 |
+
"loss": 9.8157,
|
| 5218 |
+
"step": 7440
|
| 5219 |
+
},
|
| 5220 |
+
{
|
| 5221 |
+
"epoch": 0.3654715257209784,
|
| 5222 |
+
"grad_norm": 1.3515625,
|
| 5223 |
+
"learning_rate": 9.985533966876693e-05,
|
| 5224 |
+
"loss": 9.3986,
|
| 5225 |
+
"step": 7450
|
| 5226 |
+
},
|
| 5227 |
+
{
|
| 5228 |
+
"epoch": 0.3659620915273153,
|
| 5229 |
+
"grad_norm": 2.578125,
|
| 5230 |
+
"learning_rate": 9.9853162146926e-05,
|
| 5231 |
+
"loss": 10.3461,
|
| 5232 |
+
"step": 7460
|
| 5233 |
+
},
|
| 5234 |
+
{
|
| 5235 |
+
"epoch": 0.3664526573336522,
|
| 5236 |
+
"grad_norm": 1.5703125,
|
| 5237 |
+
"learning_rate": 9.985096838264118e-05,
|
| 5238 |
+
"loss": 9.5886,
|
| 5239 |
+
"step": 7470
|
| 5240 |
+
},
|
| 5241 |
+
{
|
| 5242 |
+
"epoch": 0.36694322313998906,
|
| 5243 |
+
"grad_norm": 1.1015625,
|
| 5244 |
+
"learning_rate": 9.984875837662725e-05,
|
| 5245 |
+
"loss": 9.5306,
|
| 5246 |
+
"step": 7480
|
| 5247 |
+
},
|
| 5248 |
+
{
|
| 5249 |
+
"epoch": 0.36743378894632595,
|
| 5250 |
+
"grad_norm": 1.8515625,
|
| 5251 |
+
"learning_rate": 9.98465321296042e-05,
|
| 5252 |
+
"loss": 9.589,
|
| 5253 |
+
"step": 7490
|
| 5254 |
+
},
|
| 5255 |
+
{
|
| 5256 |
+
"epoch": 0.36792435475266283,
|
| 5257 |
+
"grad_norm": 2.078125,
|
| 5258 |
+
"learning_rate": 9.984428964229737e-05,
|
| 5259 |
+
"loss": 9.4794,
|
| 5260 |
+
"step": 7500
|
| 5261 |
+
},
|
| 5262 |
+
{
|
| 5263 |
+
"epoch": 0.3684149205589997,
|
| 5264 |
+
"grad_norm": 1.640625,
|
| 5265 |
+
"learning_rate": 9.984203091543738e-05,
|
| 5266 |
+
"loss": 9.4859,
|
| 5267 |
+
"step": 7510
|
| 5268 |
+
},
|
| 5269 |
+
{
|
| 5270 |
+
"epoch": 0.3689054863653366,
|
| 5271 |
+
"grad_norm": 0.84375,
|
| 5272 |
+
"learning_rate": 9.983975594976012e-05,
|
| 5273 |
+
"loss": 9.3247,
|
| 5274 |
+
"step": 7520
|
| 5275 |
+
},
|
| 5276 |
+
{
|
| 5277 |
+
"epoch": 0.3693960521716735,
|
| 5278 |
+
"grad_norm": 1.40625,
|
| 5279 |
+
"learning_rate": 9.983746474600682e-05,
|
| 5280 |
+
"loss": 9.3625,
|
| 5281 |
+
"step": 7530
|
| 5282 |
+
},
|
| 5283 |
+
{
|
| 5284 |
+
"epoch": 0.36988661797801037,
|
| 5285 |
+
"grad_norm": 1.6640625,
|
| 5286 |
+
"learning_rate": 9.983515730492392e-05,
|
| 5287 |
+
"loss": 8.9516,
|
| 5288 |
+
"step": 7540
|
| 5289 |
+
},
|
| 5290 |
+
{
|
| 5291 |
+
"epoch": 0.37037718378434725,
|
| 5292 |
+
"grad_norm": 1.421875,
|
| 5293 |
+
"learning_rate": 9.983283362726324e-05,
|
| 5294 |
+
"loss": 9.6106,
|
| 5295 |
+
"step": 7550
|
| 5296 |
+
},
|
| 5297 |
+
{
|
| 5298 |
+
"epoch": 0.37086774959068414,
|
| 5299 |
+
"grad_norm": 1.4921875,
|
| 5300 |
+
"learning_rate": 9.983049371378182e-05,
|
| 5301 |
+
"loss": 9.892,
|
| 5302 |
+
"step": 7560
|
| 5303 |
+
},
|
| 5304 |
+
{
|
| 5305 |
+
"epoch": 0.371358315397021,
|
| 5306 |
+
"grad_norm": 0.9921875,
|
| 5307 |
+
"learning_rate": 9.982813756524203e-05,
|
| 5308 |
+
"loss": 9.1626,
|
| 5309 |
+
"step": 7570
|
| 5310 |
+
},
|
| 5311 |
+
{
|
| 5312 |
+
"epoch": 0.3718488812033579,
|
| 5313 |
+
"grad_norm": 1.3671875,
|
| 5314 |
+
"learning_rate": 9.98257651824115e-05,
|
| 5315 |
+
"loss": 9.6846,
|
| 5316 |
+
"step": 7580
|
| 5317 |
+
},
|
| 5318 |
+
{
|
| 5319 |
+
"epoch": 0.3723394470096948,
|
| 5320 |
+
"grad_norm": 1.0390625,
|
| 5321 |
+
"learning_rate": 9.982337656606323e-05,
|
| 5322 |
+
"loss": 9.525,
|
| 5323 |
+
"step": 7590
|
| 5324 |
+
},
|
| 5325 |
+
{
|
| 5326 |
+
"epoch": 0.3728300128160317,
|
| 5327 |
+
"grad_norm": 286.0,
|
| 5328 |
+
"learning_rate": 9.982097171697535e-05,
|
| 5329 |
+
"loss": 9.5155,
|
| 5330 |
+
"step": 7600
|
| 5331 |
+
},
|
| 5332 |
+
{
|
| 5333 |
+
"epoch": 0.37332057862236856,
|
| 5334 |
+
"grad_norm": 1.7265625,
|
| 5335 |
+
"learning_rate": 9.981855063593142e-05,
|
| 5336 |
+
"loss": 10.0883,
|
| 5337 |
+
"step": 7610
|
| 5338 |
+
},
|
| 5339 |
+
{
|
| 5340 |
+
"epoch": 0.37381114442870544,
|
| 5341 |
+
"grad_norm": 1.2265625,
|
| 5342 |
+
"learning_rate": 9.981611332372025e-05,
|
| 5343 |
+
"loss": 8.9297,
|
| 5344 |
+
"step": 7620
|
| 5345 |
+
},
|
| 5346 |
+
{
|
| 5347 |
+
"epoch": 0.37430171023504233,
|
| 5348 |
+
"grad_norm": 1.5234375,
|
| 5349 |
+
"learning_rate": 9.981365978113592e-05,
|
| 5350 |
+
"loss": 9.2481,
|
| 5351 |
+
"step": 7630
|
| 5352 |
+
},
|
| 5353 |
+
{
|
| 5354 |
+
"epoch": 0.3747922760413792,
|
| 5355 |
+
"grad_norm": 2.21875,
|
| 5356 |
+
"learning_rate": 9.98111900089778e-05,
|
| 5357 |
+
"loss": 9.2244,
|
| 5358 |
+
"step": 7640
|
| 5359 |
+
},
|
| 5360 |
+
{
|
| 5361 |
+
"epoch": 0.3752828418477161,
|
| 5362 |
+
"grad_norm": 1.5859375,
|
| 5363 |
+
"learning_rate": 9.980870400805058e-05,
|
| 5364 |
+
"loss": 9.7969,
|
| 5365 |
+
"step": 7650
|
| 5366 |
+
},
|
| 5367 |
+
{
|
| 5368 |
+
"epoch": 0.375773407654053,
|
| 5369 |
+
"grad_norm": 1.015625,
|
| 5370 |
+
"learning_rate": 9.980620177916419e-05,
|
| 5371 |
+
"loss": 9.0249,
|
| 5372 |
+
"step": 7660
|
| 5373 |
+
},
|
| 5374 |
+
{
|
| 5375 |
+
"epoch": 0.37626397346038987,
|
| 5376 |
+
"grad_norm": 1.59375,
|
| 5377 |
+
"learning_rate": 9.980368332313388e-05,
|
| 5378 |
+
"loss": 9.4258,
|
| 5379 |
+
"step": 7670
|
| 5380 |
+
},
|
| 5381 |
+
{
|
| 5382 |
+
"epoch": 0.37675453926672675,
|
| 5383 |
+
"grad_norm": 1.71875,
|
| 5384 |
+
"learning_rate": 9.980114864078016e-05,
|
| 5385 |
+
"loss": 9.5474,
|
| 5386 |
+
"step": 7680
|
| 5387 |
+
},
|
| 5388 |
+
{
|
| 5389 |
+
"epoch": 0.37724510507306364,
|
| 5390 |
+
"grad_norm": 2.28125,
|
| 5391 |
+
"learning_rate": 9.979859773292888e-05,
|
| 5392 |
+
"loss": 9.557,
|
| 5393 |
+
"step": 7690
|
| 5394 |
+
},
|
| 5395 |
+
{
|
| 5396 |
+
"epoch": 0.3777356708794005,
|
| 5397 |
+
"grad_norm": 0.98046875,
|
| 5398 |
+
"learning_rate": 9.979603060041111e-05,
|
| 5399 |
+
"loss": 9.2212,
|
| 5400 |
+
"step": 7700
|
| 5401 |
+
},
|
| 5402 |
+
{
|
| 5403 |
+
"epoch": 0.3782262366857374,
|
| 5404 |
+
"grad_norm": 1.265625,
|
| 5405 |
+
"learning_rate": 9.979344724406323e-05,
|
| 5406 |
+
"loss": 9.5588,
|
| 5407 |
+
"step": 7710
|
| 5408 |
+
},
|
| 5409 |
+
{
|
| 5410 |
+
"epoch": 0.3787168024920743,
|
| 5411 |
+
"grad_norm": 1.34375,
|
| 5412 |
+
"learning_rate": 9.979084766472695e-05,
|
| 5413 |
+
"loss": 9.7479,
|
| 5414 |
+
"step": 7720
|
| 5415 |
+
},
|
| 5416 |
+
{
|
| 5417 |
+
"epoch": 0.3792073682984112,
|
| 5418 |
+
"grad_norm": 1.53125,
|
| 5419 |
+
"learning_rate": 9.978823186324921e-05,
|
| 5420 |
+
"loss": 8.9423,
|
| 5421 |
+
"step": 7730
|
| 5422 |
+
},
|
| 5423 |
+
{
|
| 5424 |
+
"epoch": 0.37969793410474806,
|
| 5425 |
+
"grad_norm": 0.9921875,
|
| 5426 |
+
"learning_rate": 9.978559984048224e-05,
|
| 5427 |
+
"loss": 9.62,
|
| 5428 |
+
"step": 7740
|
| 5429 |
+
},
|
| 5430 |
+
{
|
| 5431 |
+
"epoch": 0.38018849991108494,
|
| 5432 |
+
"grad_norm": 1.421875,
|
| 5433 |
+
"learning_rate": 9.978295159728357e-05,
|
| 5434 |
+
"loss": 9.4276,
|
| 5435 |
+
"step": 7750
|
| 5436 |
+
},
|
| 5437 |
+
{
|
| 5438 |
+
"epoch": 0.3806790657174218,
|
| 5439 |
+
"grad_norm": 1.4140625,
|
| 5440 |
+
"learning_rate": 9.978028713451602e-05,
|
| 5441 |
+
"loss": 9.5612,
|
| 5442 |
+
"step": 7760
|
| 5443 |
+
},
|
| 5444 |
+
{
|
| 5445 |
+
"epoch": 0.3811696315237587,
|
| 5446 |
+
"grad_norm": 1.0859375,
|
| 5447 |
+
"learning_rate": 9.97776064530477e-05,
|
| 5448 |
+
"loss": 9.5241,
|
| 5449 |
+
"step": 7770
|
| 5450 |
+
},
|
| 5451 |
+
{
|
| 5452 |
+
"epoch": 0.3816601973300956,
|
| 5453 |
+
"grad_norm": 1.4140625,
|
| 5454 |
+
"learning_rate": 9.977490955375195e-05,
|
| 5455 |
+
"loss": 9.4522,
|
| 5456 |
+
"step": 7780
|
| 5457 |
+
},
|
| 5458 |
+
{
|
| 5459 |
+
"epoch": 0.3821507631364325,
|
| 5460 |
+
"grad_norm": 1.0390625,
|
| 5461 |
+
"learning_rate": 9.977219643750747e-05,
|
| 5462 |
+
"loss": 9.0719,
|
| 5463 |
+
"step": 7790
|
| 5464 |
+
},
|
| 5465 |
+
{
|
| 5466 |
+
"epoch": 0.38264132894276937,
|
| 5467 |
+
"grad_norm": 0.98046875,
|
| 5468 |
+
"learning_rate": 9.976946710519823e-05,
|
| 5469 |
+
"loss": 9.0643,
|
| 5470 |
+
"step": 7800
|
| 5471 |
+
},
|
| 5472 |
+
{
|
| 5473 |
+
"epoch": 0.38313189474910625,
|
| 5474 |
+
"grad_norm": 1.3984375,
|
| 5475 |
+
"learning_rate": 9.976672155771342e-05,
|
| 5476 |
+
"loss": 9.7797,
|
| 5477 |
+
"step": 7810
|
| 5478 |
+
},
|
| 5479 |
+
{
|
| 5480 |
+
"epoch": 0.38362246055544313,
|
| 5481 |
+
"grad_norm": 1.0546875,
|
| 5482 |
+
"learning_rate": 9.976395979594755e-05,
|
| 5483 |
+
"loss": 9.2137,
|
| 5484 |
+
"step": 7820
|
| 5485 |
+
},
|
| 5486 |
+
{
|
| 5487 |
+
"epoch": 0.38411302636178,
|
| 5488 |
+
"grad_norm": 1.375,
|
| 5489 |
+
"learning_rate": 9.976118182080045e-05,
|
| 5490 |
+
"loss": 9.1207,
|
| 5491 |
+
"step": 7830
|
| 5492 |
+
},
|
| 5493 |
+
{
|
| 5494 |
+
"epoch": 0.3846035921681169,
|
| 5495 |
+
"grad_norm": 1.578125,
|
| 5496 |
+
"learning_rate": 9.97583876331772e-05,
|
| 5497 |
+
"loss": 9.828,
|
| 5498 |
+
"step": 7840
|
| 5499 |
+
},
|
| 5500 |
+
{
|
| 5501 |
+
"epoch": 0.3850941579744538,
|
| 5502 |
+
"grad_norm": 1.265625,
|
| 5503 |
+
"learning_rate": 9.975557723398814e-05,
|
| 5504 |
+
"loss": 9.2545,
|
| 5505 |
+
"step": 7850
|
| 5506 |
+
},
|
| 5507 |
+
{
|
| 5508 |
+
"epoch": 0.3855847237807907,
|
| 5509 |
+
"grad_norm": 96.0,
|
| 5510 |
+
"learning_rate": 9.975275062414891e-05,
|
| 5511 |
+
"loss": 9.3358,
|
| 5512 |
+
"step": 7860
|
| 5513 |
+
},
|
| 5514 |
+
{
|
| 5515 |
+
"epoch": 0.38607528958712756,
|
| 5516 |
+
"grad_norm": 0.88671875,
|
| 5517 |
+
"learning_rate": 9.974990780458047e-05,
|
| 5518 |
+
"loss": 9.3523,
|
| 5519 |
+
"step": 7870
|
| 5520 |
+
},
|
| 5521 |
+
{
|
| 5522 |
+
"epoch": 0.38656585539346444,
|
| 5523 |
+
"grad_norm": 1.0625,
|
| 5524 |
+
"learning_rate": 9.974704877620899e-05,
|
| 5525 |
+
"loss": 8.9798,
|
| 5526 |
+
"step": 7880
|
| 5527 |
+
},
|
| 5528 |
+
{
|
| 5529 |
+
"epoch": 0.3870564211998013,
|
| 5530 |
+
"grad_norm": 1.8203125,
|
| 5531 |
+
"learning_rate": 9.974417353996597e-05,
|
| 5532 |
+
"loss": 9.2332,
|
| 5533 |
+
"step": 7890
|
| 5534 |
+
},
|
| 5535 |
+
{
|
| 5536 |
+
"epoch": 0.3875469870061382,
|
| 5537 |
+
"grad_norm": 1.828125,
|
| 5538 |
+
"learning_rate": 9.97412820967882e-05,
|
| 5539 |
+
"loss": 9.6917,
|
| 5540 |
+
"step": 7900
|
| 5541 |
+
},
|
| 5542 |
+
{
|
| 5543 |
+
"epoch": 0.3880375528124751,
|
| 5544 |
+
"grad_norm": 1.8125,
|
| 5545 |
+
"learning_rate": 9.973837444761767e-05,
|
| 5546 |
+
"loss": 10.2114,
|
| 5547 |
+
"step": 7910
|
| 5548 |
+
},
|
| 5549 |
+
{
|
| 5550 |
+
"epoch": 0.388528118618812,
|
| 5551 |
+
"grad_norm": 1.1328125,
|
| 5552 |
+
"learning_rate": 9.973545059340179e-05,
|
| 5553 |
+
"loss": 8.8373,
|
| 5554 |
+
"step": 7920
|
| 5555 |
+
},
|
| 5556 |
+
{
|
| 5557 |
+
"epoch": 0.38901868442514886,
|
| 5558 |
+
"grad_norm": 2.125,
|
| 5559 |
+
"learning_rate": 9.973251053509311e-05,
|
| 5560 |
+
"loss": 9.2673,
|
| 5561 |
+
"step": 7930
|
| 5562 |
+
},
|
| 5563 |
+
{
|
| 5564 |
+
"epoch": 0.38950925023148575,
|
| 5565 |
+
"grad_norm": 1.4609375,
|
| 5566 |
+
"learning_rate": 9.972955427364953e-05,
|
| 5567 |
+
"loss": 9.5442,
|
| 5568 |
+
"step": 7940
|
| 5569 |
+
},
|
| 5570 |
+
{
|
| 5571 |
+
"epoch": 0.38999981603782263,
|
| 5572 |
+
"grad_norm": 1.2421875,
|
| 5573 |
+
"learning_rate": 9.972658181003422e-05,
|
| 5574 |
+
"loss": 9.3631,
|
| 5575 |
+
"step": 7950
|
| 5576 |
+
},
|
| 5577 |
+
{
|
| 5578 |
+
"epoch": 0.3904903818441595,
|
| 5579 |
+
"grad_norm": 1.46875,
|
| 5580 |
+
"learning_rate": 9.972359314521565e-05,
|
| 5581 |
+
"loss": 9.2575,
|
| 5582 |
+
"step": 7960
|
| 5583 |
+
},
|
| 5584 |
+
{
|
| 5585 |
+
"epoch": 0.3909809476504964,
|
| 5586 |
+
"grad_norm": 1.4453125,
|
| 5587 |
+
"learning_rate": 9.972058828016749e-05,
|
| 5588 |
+
"loss": 9.5496,
|
| 5589 |
+
"step": 7970
|
| 5590 |
+
},
|
| 5591 |
+
{
|
| 5592 |
+
"epoch": 0.3914715134568333,
|
| 5593 |
+
"grad_norm": 2.21875,
|
| 5594 |
+
"learning_rate": 9.971756721586879e-05,
|
| 5595 |
+
"loss": 9.5364,
|
| 5596 |
+
"step": 7980
|
| 5597 |
+
},
|
| 5598 |
+
{
|
| 5599 |
+
"epoch": 0.39196207926317017,
|
| 5600 |
+
"grad_norm": 1.3359375,
|
| 5601 |
+
"learning_rate": 9.971452995330382e-05,
|
| 5602 |
+
"loss": 9.7381,
|
| 5603 |
+
"step": 7990
|
| 5604 |
+
},
|
| 5605 |
+
{
|
| 5606 |
+
"epoch": 0.39245264506950706,
|
| 5607 |
+
"grad_norm": 1.328125,
|
| 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 |
"logging_steps": 10,
|
|
|
|
| 7027 |
"attributes": {}
|
| 7028 |
}
|
| 7029 |
},
|
| 7030 |
+
"total_flos": 2.69016216109056e+19,
|
| 7031 |
"train_batch_size": 2,
|
| 7032 |
"trial_name": null,
|
| 7033 |
"trial_params": null
|
checkpoints/{checkpoint-6000 β checkpoint-10000}/training_args.bin
RENAMED
|
File without changes
|
logs/events.out.tfevents.1752345702.5f638d684589.1618.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f6ccf911ef8ac8ed9bc3e58f256b3fe099ed727d89e49446ed0edb60994b81b3
|
| 3 |
+
size 216247
|