Upload 10 files
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +2733 -3
- training_args.bin +1 -1
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 598635032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:369fb121dccbfa708387d7542d49616a29ba6ab3a23f6d40be115e7f2da11de3
|
| 3 |
size 598635032
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1197359627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:97b91fd1b382931771f0f6e73df67cd95ac4ed5c3d36a8a1a7e104521c538d24
|
| 3 |
size 1197359627
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f0ea071ae1c5cc927b5c63cfe47ea1c2ca149560b4f3efa336fcab339b6f065e
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c5e5def1dbadcea7f355dc3c47dab1ec2eea5e39c560e66d4be8f163dff9ae8
|
| 3 |
size 1465
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 1000,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -1341,6 +1341,2736 @@
|
|
| 1341 |
"eval_samples_per_second": 202.013,
|
| 1342 |
"eval_steps_per_second": 3.166,
|
| 1343 |
"step": 17000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1344 |
}
|
| 1345 |
],
|
| 1346 |
"logging_steps": 100,
|
|
@@ -1360,7 +4090,7 @@
|
|
| 1360 |
"attributes": {}
|
| 1361 |
}
|
| 1362 |
},
|
| 1363 |
-
"total_flos":
|
| 1364 |
"train_batch_size": 64,
|
| 1365 |
"trial_name": null,
|
| 1366 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.044,
|
| 6 |
"eval_steps": 1000,
|
| 7 |
+
"global_step": 52000,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 1341 |
"eval_samples_per_second": 202.013,
|
| 1342 |
"eval_steps_per_second": 3.166,
|
| 1343 |
"step": 17000
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 0.0342,
|
| 1347 |
+
"grad_norm": 0.9049177765846252,
|
| 1348 |
+
"learning_rate": 4.992633097789499e-05,
|
| 1349 |
+
"loss": 1.1995,
|
| 1350 |
+
"step": 17100
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.0344,
|
| 1354 |
+
"grad_norm": 1.2447205781936646,
|
| 1355 |
+
"learning_rate": 4.992510878581858e-05,
|
| 1356 |
+
"loss": 1.2174,
|
| 1357 |
+
"step": 17200
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.0346,
|
| 1361 |
+
"grad_norm": 1.0060733556747437,
|
| 1362 |
+
"learning_rate": 4.9923876553911334e-05,
|
| 1363 |
+
"loss": 1.2098,
|
| 1364 |
+
"step": 17300
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.0348,
|
| 1368 |
+
"grad_norm": 1.3275829553604126,
|
| 1369 |
+
"learning_rate": 4.992263428266958e-05,
|
| 1370 |
+
"loss": 1.2256,
|
| 1371 |
+
"step": 17400
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 0.035,
|
| 1375 |
+
"grad_norm": 1.3165931701660156,
|
| 1376 |
+
"learning_rate": 4.992138197259373e-05,
|
| 1377 |
+
"loss": 1.2276,
|
| 1378 |
+
"step": 17500
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 0.0352,
|
| 1382 |
+
"grad_norm": 1.2749327421188354,
|
| 1383 |
+
"learning_rate": 4.9920119624188196e-05,
|
| 1384 |
+
"loss": 1.2758,
|
| 1385 |
+
"step": 17600
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.0354,
|
| 1389 |
+
"grad_norm": 1.0836033821105957,
|
| 1390 |
+
"learning_rate": 4.991884723796146e-05,
|
| 1391 |
+
"loss": 1.2407,
|
| 1392 |
+
"step": 17700
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.0356,
|
| 1396 |
+
"grad_norm": 1.343475103378296,
|
| 1397 |
+
"learning_rate": 4.9917564814426034e-05,
|
| 1398 |
+
"loss": 1.2466,
|
| 1399 |
+
"step": 17800
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.0358,
|
| 1403 |
+
"grad_norm": 1.3868790864944458,
|
| 1404 |
+
"learning_rate": 4.991627235409848e-05,
|
| 1405 |
+
"loss": 1.2402,
|
| 1406 |
+
"step": 17900
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.036,
|
| 1410 |
+
"grad_norm": 1.5200074911117554,
|
| 1411 |
+
"learning_rate": 4.99149698574994e-05,
|
| 1412 |
+
"loss": 1.2183,
|
| 1413 |
+
"step": 18000
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 0.036,
|
| 1417 |
+
"eval_loss": 1.0960842370986938,
|
| 1418 |
+
"eval_runtime": 76.481,
|
| 1419 |
+
"eval_samples_per_second": 201.92,
|
| 1420 |
+
"eval_steps_per_second": 3.164,
|
| 1421 |
+
"step": 18000
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 0.0362,
|
| 1425 |
+
"grad_norm": 1.4647791385650635,
|
| 1426 |
+
"learning_rate": 4.991365732515345e-05,
|
| 1427 |
+
"loss": 1.2386,
|
| 1428 |
+
"step": 18100
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.0364,
|
| 1432 |
+
"grad_norm": 0.9076351523399353,
|
| 1433 |
+
"learning_rate": 4.991233475758931e-05,
|
| 1434 |
+
"loss": 1.2011,
|
| 1435 |
+
"step": 18200
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 0.0366,
|
| 1439 |
+
"grad_norm": 0.9813222289085388,
|
| 1440 |
+
"learning_rate": 4.99110021553397e-05,
|
| 1441 |
+
"loss": 1.214,
|
| 1442 |
+
"step": 18300
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.0368,
|
| 1446 |
+
"grad_norm": 1.5431565046310425,
|
| 1447 |
+
"learning_rate": 4.99096595189414e-05,
|
| 1448 |
+
"loss": 1.2206,
|
| 1449 |
+
"step": 18400
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.037,
|
| 1453 |
+
"grad_norm": 0.9991932511329651,
|
| 1454 |
+
"learning_rate": 4.990830684893523e-05,
|
| 1455 |
+
"loss": 1.2334,
|
| 1456 |
+
"step": 18500
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.0372,
|
| 1460 |
+
"grad_norm": 0.6322658658027649,
|
| 1461 |
+
"learning_rate": 4.9906944145866035e-05,
|
| 1462 |
+
"loss": 1.2354,
|
| 1463 |
+
"step": 18600
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 0.0374,
|
| 1467 |
+
"grad_norm": 0.9555477499961853,
|
| 1468 |
+
"learning_rate": 4.990557141028272e-05,
|
| 1469 |
+
"loss": 1.2017,
|
| 1470 |
+
"step": 18700
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.0376,
|
| 1474 |
+
"grad_norm": 1.171019196510315,
|
| 1475 |
+
"learning_rate": 4.990418864273822e-05,
|
| 1476 |
+
"loss": 1.286,
|
| 1477 |
+
"step": 18800
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.0378,
|
| 1481 |
+
"grad_norm": 1.2275811433792114,
|
| 1482 |
+
"learning_rate": 4.990279584378951e-05,
|
| 1483 |
+
"loss": 1.2345,
|
| 1484 |
+
"step": 18900
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.038,
|
| 1488 |
+
"grad_norm": 1.6589407920837402,
|
| 1489 |
+
"learning_rate": 4.9901393013997616e-05,
|
| 1490 |
+
"loss": 1.2376,
|
| 1491 |
+
"step": 19000
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 0.038,
|
| 1495 |
+
"eval_loss": 1.107132077217102,
|
| 1496 |
+
"eval_runtime": 76.3932,
|
| 1497 |
+
"eval_samples_per_second": 202.152,
|
| 1498 |
+
"eval_steps_per_second": 3.168,
|
| 1499 |
+
"step": 19000
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 0.0382,
|
| 1503 |
+
"grad_norm": 0.7907335758209229,
|
| 1504 |
+
"learning_rate": 4.9899980153927596e-05,
|
| 1505 |
+
"loss": 1.2554,
|
| 1506 |
+
"step": 19100
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.0384,
|
| 1510 |
+
"grad_norm": 1.4444235563278198,
|
| 1511 |
+
"learning_rate": 4.989855726414854e-05,
|
| 1512 |
+
"loss": 1.2618,
|
| 1513 |
+
"step": 19200
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 0.0386,
|
| 1517 |
+
"grad_norm": 1.1591296195983887,
|
| 1518 |
+
"learning_rate": 4.98971243452336e-05,
|
| 1519 |
+
"loss": 1.2028,
|
| 1520 |
+
"step": 19300
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.0388,
|
| 1524 |
+
"grad_norm": 0.9183579087257385,
|
| 1525 |
+
"learning_rate": 4.989568139775995e-05,
|
| 1526 |
+
"loss": 1.2259,
|
| 1527 |
+
"step": 19400
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 0.039,
|
| 1531 |
+
"grad_norm": 1.0866785049438477,
|
| 1532 |
+
"learning_rate": 4.9894228422308805e-05,
|
| 1533 |
+
"loss": 1.2307,
|
| 1534 |
+
"step": 19500
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 0.0392,
|
| 1538 |
+
"grad_norm": 1.5889687538146973,
|
| 1539 |
+
"learning_rate": 4.9892765419465436e-05,
|
| 1540 |
+
"loss": 1.2346,
|
| 1541 |
+
"step": 19600
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 0.0394,
|
| 1545 |
+
"grad_norm": 1.300850510597229,
|
| 1546 |
+
"learning_rate": 4.989129238981913e-05,
|
| 1547 |
+
"loss": 1.2748,
|
| 1548 |
+
"step": 19700
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 0.0396,
|
| 1552 |
+
"grad_norm": 1.2363704442977905,
|
| 1553 |
+
"learning_rate": 4.988980933396323e-05,
|
| 1554 |
+
"loss": 1.2536,
|
| 1555 |
+
"step": 19800
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 0.0398,
|
| 1559 |
+
"grad_norm": 0.8141745328903198,
|
| 1560 |
+
"learning_rate": 4.9888316252495106e-05,
|
| 1561 |
+
"loss": 1.2198,
|
| 1562 |
+
"step": 19900
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 0.04,
|
| 1566 |
+
"grad_norm": 1.0759721994400024,
|
| 1567 |
+
"learning_rate": 4.988681314601617e-05,
|
| 1568 |
+
"loss": 1.2225,
|
| 1569 |
+
"step": 20000
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 0.04,
|
| 1573 |
+
"eval_loss": 1.0962127447128296,
|
| 1574 |
+
"eval_runtime": 76.7102,
|
| 1575 |
+
"eval_samples_per_second": 201.316,
|
| 1576 |
+
"eval_steps_per_second": 3.155,
|
| 1577 |
+
"step": 20000
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.0402,
|
| 1581 |
+
"grad_norm": 1.2748645544052124,
|
| 1582 |
+
"learning_rate": 4.988530001513187e-05,
|
| 1583 |
+
"loss": 1.2245,
|
| 1584 |
+
"step": 20100
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.0404,
|
| 1588 |
+
"grad_norm": 1.3192243576049805,
|
| 1589 |
+
"learning_rate": 4.9883776860451704e-05,
|
| 1590 |
+
"loss": 1.2292,
|
| 1591 |
+
"step": 20200
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.0406,
|
| 1595 |
+
"grad_norm": 1.329868197441101,
|
| 1596 |
+
"learning_rate": 4.98822436825892e-05,
|
| 1597 |
+
"loss": 1.2243,
|
| 1598 |
+
"step": 20300
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.0408,
|
| 1602 |
+
"grad_norm": 1.3394356966018677,
|
| 1603 |
+
"learning_rate": 4.988070048216191e-05,
|
| 1604 |
+
"loss": 1.216,
|
| 1605 |
+
"step": 20400
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.041,
|
| 1609 |
+
"grad_norm": 1.3035671710968018,
|
| 1610 |
+
"learning_rate": 4.987914725979144e-05,
|
| 1611 |
+
"loss": 1.2335,
|
| 1612 |
+
"step": 20500
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.0412,
|
| 1616 |
+
"grad_norm": 1.2765480279922485,
|
| 1617 |
+
"learning_rate": 4.987758401610343e-05,
|
| 1618 |
+
"loss": 1.261,
|
| 1619 |
+
"step": 20600
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.0414,
|
| 1623 |
+
"grad_norm": 1.0472270250320435,
|
| 1624 |
+
"learning_rate": 4.9876010751727553e-05,
|
| 1625 |
+
"loss": 1.2173,
|
| 1626 |
+
"step": 20700
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.0416,
|
| 1630 |
+
"grad_norm": 1.163237452507019,
|
| 1631 |
+
"learning_rate": 4.9874427467297525e-05,
|
| 1632 |
+
"loss": 1.2316,
|
| 1633 |
+
"step": 20800
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.0418,
|
| 1637 |
+
"grad_norm": 1.3546457290649414,
|
| 1638 |
+
"learning_rate": 4.987283416345109e-05,
|
| 1639 |
+
"loss": 1.2268,
|
| 1640 |
+
"step": 20900
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.042,
|
| 1644 |
+
"grad_norm": 1.0812748670578003,
|
| 1645 |
+
"learning_rate": 4.9871230840830016e-05,
|
| 1646 |
+
"loss": 1.2267,
|
| 1647 |
+
"step": 21000
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.042,
|
| 1651 |
+
"eval_loss": 1.1046785116195679,
|
| 1652 |
+
"eval_runtime": 76.3631,
|
| 1653 |
+
"eval_samples_per_second": 202.231,
|
| 1654 |
+
"eval_steps_per_second": 3.169,
|
| 1655 |
+
"step": 21000
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.0422,
|
| 1659 |
+
"grad_norm": 0.7458230257034302,
|
| 1660 |
+
"learning_rate": 4.986961750008014e-05,
|
| 1661 |
+
"loss": 1.1918,
|
| 1662 |
+
"step": 21100
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 0.0424,
|
| 1666 |
+
"grad_norm": 1.2837951183319092,
|
| 1667 |
+
"learning_rate": 4.986799414185131e-05,
|
| 1668 |
+
"loss": 1.2206,
|
| 1669 |
+
"step": 21200
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 0.0426,
|
| 1673 |
+
"grad_norm": 1.4213489294052124,
|
| 1674 |
+
"learning_rate": 4.986636076679742e-05,
|
| 1675 |
+
"loss": 1.2552,
|
| 1676 |
+
"step": 21300
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 0.0428,
|
| 1680 |
+
"grad_norm": 1.297608733177185,
|
| 1681 |
+
"learning_rate": 4.986471737557638e-05,
|
| 1682 |
+
"loss": 1.2234,
|
| 1683 |
+
"step": 21400
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 0.043,
|
| 1687 |
+
"grad_norm": 1.3617885112762451,
|
| 1688 |
+
"learning_rate": 4.986306396885015e-05,
|
| 1689 |
+
"loss": 1.2381,
|
| 1690 |
+
"step": 21500
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 0.0432,
|
| 1694 |
+
"grad_norm": 1.500025749206543,
|
| 1695 |
+
"learning_rate": 4.986140054728473e-05,
|
| 1696 |
+
"loss": 1.1957,
|
| 1697 |
+
"step": 21600
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.0434,
|
| 1701 |
+
"grad_norm": 0.6222732663154602,
|
| 1702 |
+
"learning_rate": 4.9859727111550147e-05,
|
| 1703 |
+
"loss": 1.2579,
|
| 1704 |
+
"step": 21700
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 0.0436,
|
| 1708 |
+
"grad_norm": 1.4154349565505981,
|
| 1709 |
+
"learning_rate": 4.985804366232045e-05,
|
| 1710 |
+
"loss": 1.2073,
|
| 1711 |
+
"step": 21800
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 0.0438,
|
| 1715 |
+
"grad_norm": 1.334390640258789,
|
| 1716 |
+
"learning_rate": 4.9856350200273746e-05,
|
| 1717 |
+
"loss": 1.2317,
|
| 1718 |
+
"step": 21900
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 0.044,
|
| 1722 |
+
"grad_norm": 0.8164774179458618,
|
| 1723 |
+
"learning_rate": 4.985464672609215e-05,
|
| 1724 |
+
"loss": 1.2248,
|
| 1725 |
+
"step": 22000
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 0.044,
|
| 1729 |
+
"eval_loss": 1.1025385856628418,
|
| 1730 |
+
"eval_runtime": 76.8498,
|
| 1731 |
+
"eval_samples_per_second": 200.951,
|
| 1732 |
+
"eval_steps_per_second": 3.149,
|
| 1733 |
+
"step": 22000
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.0442,
|
| 1737 |
+
"grad_norm": 1.1641725301742554,
|
| 1738 |
+
"learning_rate": 4.985293324046182e-05,
|
| 1739 |
+
"loss": 1.1928,
|
| 1740 |
+
"step": 22100
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.0444,
|
| 1744 |
+
"grad_norm": 1.2185006141662598,
|
| 1745 |
+
"learning_rate": 4.9851209744072954e-05,
|
| 1746 |
+
"loss": 1.2435,
|
| 1747 |
+
"step": 22200
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.0446,
|
| 1751 |
+
"grad_norm": 1.0973742008209229,
|
| 1752 |
+
"learning_rate": 4.9849476237619784e-05,
|
| 1753 |
+
"loss": 1.2515,
|
| 1754 |
+
"step": 22300
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.0448,
|
| 1758 |
+
"grad_norm": 1.0242998600006104,
|
| 1759 |
+
"learning_rate": 4.984773272180056e-05,
|
| 1760 |
+
"loss": 1.2511,
|
| 1761 |
+
"step": 22400
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.045,
|
| 1765 |
+
"grad_norm": 0.598416805267334,
|
| 1766 |
+
"learning_rate": 4.984597919731755e-05,
|
| 1767 |
+
"loss": 1.215,
|
| 1768 |
+
"step": 22500
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.0452,
|
| 1772 |
+
"grad_norm": 0.9391146302223206,
|
| 1773 |
+
"learning_rate": 4.98442156648771e-05,
|
| 1774 |
+
"loss": 1.2303,
|
| 1775 |
+
"step": 22600
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.0454,
|
| 1779 |
+
"grad_norm": 0.9301611185073853,
|
| 1780 |
+
"learning_rate": 4.9842442125189556e-05,
|
| 1781 |
+
"loss": 1.2621,
|
| 1782 |
+
"step": 22700
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 0.0456,
|
| 1786 |
+
"grad_norm": 1.3423951864242554,
|
| 1787 |
+
"learning_rate": 4.984065857896928e-05,
|
| 1788 |
+
"loss": 1.2251,
|
| 1789 |
+
"step": 22800
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 0.0458,
|
| 1793 |
+
"grad_norm": 1.3373651504516602,
|
| 1794 |
+
"learning_rate": 4.983886502693471e-05,
|
| 1795 |
+
"loss": 1.2738,
|
| 1796 |
+
"step": 22900
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.046,
|
| 1800 |
+
"grad_norm": 1.007158637046814,
|
| 1801 |
+
"learning_rate": 4.983706146980828e-05,
|
| 1802 |
+
"loss": 1.1923,
|
| 1803 |
+
"step": 23000
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 0.046,
|
| 1807 |
+
"eval_loss": 1.1094993352890015,
|
| 1808 |
+
"eval_runtime": 76.6473,
|
| 1809 |
+
"eval_samples_per_second": 201.481,
|
| 1810 |
+
"eval_steps_per_second": 3.157,
|
| 1811 |
+
"step": 23000
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.0462,
|
| 1815 |
+
"grad_norm": 0.7804542779922485,
|
| 1816 |
+
"learning_rate": 4.9835247908316454e-05,
|
| 1817 |
+
"loss": 1.2098,
|
| 1818 |
+
"step": 23100
|
| 1819 |
+
},
|
| 1820 |
+
{
|
| 1821 |
+
"epoch": 0.0464,
|
| 1822 |
+
"grad_norm": 1.377008318901062,
|
| 1823 |
+
"learning_rate": 4.983342434318975e-05,
|
| 1824 |
+
"loss": 1.2202,
|
| 1825 |
+
"step": 23200
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 0.0466,
|
| 1829 |
+
"grad_norm": 1.1037031412124634,
|
| 1830 |
+
"learning_rate": 4.983159077516268e-05,
|
| 1831 |
+
"loss": 1.1977,
|
| 1832 |
+
"step": 23300
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 0.0468,
|
| 1836 |
+
"grad_norm": 0.7141278386116028,
|
| 1837 |
+
"learning_rate": 4.982974720497382e-05,
|
| 1838 |
+
"loss": 1.2054,
|
| 1839 |
+
"step": 23400
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 0.047,
|
| 1843 |
+
"grad_norm": 0.570811927318573,
|
| 1844 |
+
"learning_rate": 4.9827893633365754e-05,
|
| 1845 |
+
"loss": 1.2163,
|
| 1846 |
+
"step": 23500
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.0472,
|
| 1850 |
+
"grad_norm": 0.7255613803863525,
|
| 1851 |
+
"learning_rate": 4.98260300610851e-05,
|
| 1852 |
+
"loss": 1.2212,
|
| 1853 |
+
"step": 23600
|
| 1854 |
+
},
|
| 1855 |
+
{
|
| 1856 |
+
"epoch": 0.0474,
|
| 1857 |
+
"grad_norm": 0.8988520503044128,
|
| 1858 |
+
"learning_rate": 4.982415648888251e-05,
|
| 1859 |
+
"loss": 1.2332,
|
| 1860 |
+
"step": 23700
|
| 1861 |
+
},
|
| 1862 |
+
{
|
| 1863 |
+
"epoch": 0.0476,
|
| 1864 |
+
"grad_norm": 1.2191438674926758,
|
| 1865 |
+
"learning_rate": 4.9822272917512644e-05,
|
| 1866 |
+
"loss": 1.1974,
|
| 1867 |
+
"step": 23800
|
| 1868 |
+
},
|
| 1869 |
+
{
|
| 1870 |
+
"epoch": 0.0478,
|
| 1871 |
+
"grad_norm": 1.2043516635894775,
|
| 1872 |
+
"learning_rate": 4.982037934773423e-05,
|
| 1873 |
+
"loss": 1.2229,
|
| 1874 |
+
"step": 23900
|
| 1875 |
+
},
|
| 1876 |
+
{
|
| 1877 |
+
"epoch": 0.048,
|
| 1878 |
+
"grad_norm": 1.3503689765930176,
|
| 1879 |
+
"learning_rate": 4.981847578030998e-05,
|
| 1880 |
+
"loss": 1.2307,
|
| 1881 |
+
"step": 24000
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.048,
|
| 1885 |
+
"eval_loss": 1.0969973802566528,
|
| 1886 |
+
"eval_runtime": 76.7433,
|
| 1887 |
+
"eval_samples_per_second": 201.229,
|
| 1888 |
+
"eval_steps_per_second": 3.153,
|
| 1889 |
+
"step": 24000
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.0482,
|
| 1893 |
+
"grad_norm": 1.3795185089111328,
|
| 1894 |
+
"learning_rate": 4.9816562216006645e-05,
|
| 1895 |
+
"loss": 1.1894,
|
| 1896 |
+
"step": 24100
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 0.0484,
|
| 1900 |
+
"grad_norm": 1.1966140270233154,
|
| 1901 |
+
"learning_rate": 4.9814638655595024e-05,
|
| 1902 |
+
"loss": 1.2011,
|
| 1903 |
+
"step": 24200
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 0.0486,
|
| 1907 |
+
"grad_norm": 1.179077386856079,
|
| 1908 |
+
"learning_rate": 4.981270509984992e-05,
|
| 1909 |
+
"loss": 1.2596,
|
| 1910 |
+
"step": 24300
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 0.0488,
|
| 1914 |
+
"grad_norm": 1.24593186378479,
|
| 1915 |
+
"learning_rate": 4.9810761549550166e-05,
|
| 1916 |
+
"loss": 1.2219,
|
| 1917 |
+
"step": 24400
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 0.049,
|
| 1921 |
+
"grad_norm": 1.2809820175170898,
|
| 1922 |
+
"learning_rate": 4.9808808005478635e-05,
|
| 1923 |
+
"loss": 1.2033,
|
| 1924 |
+
"step": 24500
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 0.0492,
|
| 1928 |
+
"grad_norm": 0.9016757011413574,
|
| 1929 |
+
"learning_rate": 4.9806844468422196e-05,
|
| 1930 |
+
"loss": 1.2394,
|
| 1931 |
+
"step": 24600
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.0494,
|
| 1935 |
+
"grad_norm": 0.7064381837844849,
|
| 1936 |
+
"learning_rate": 4.9804870939171774e-05,
|
| 1937 |
+
"loss": 1.2154,
|
| 1938 |
+
"step": 24700
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 0.0496,
|
| 1942 |
+
"grad_norm": 0.626646101474762,
|
| 1943 |
+
"learning_rate": 4.980288741852231e-05,
|
| 1944 |
+
"loss": 1.2021,
|
| 1945 |
+
"step": 24800
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.0498,
|
| 1949 |
+
"grad_norm": 1.049187421798706,
|
| 1950 |
+
"learning_rate": 4.980089390727275e-05,
|
| 1951 |
+
"loss": 1.1839,
|
| 1952 |
+
"step": 24900
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 0.05,
|
| 1956 |
+
"grad_norm": 1.2987581491470337,
|
| 1957 |
+
"learning_rate": 4.97988904062261e-05,
|
| 1958 |
+
"loss": 1.1969,
|
| 1959 |
+
"step": 25000
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 0.05,
|
| 1963 |
+
"eval_loss": 1.090114951133728,
|
| 1964 |
+
"eval_runtime": 77.5992,
|
| 1965 |
+
"eval_samples_per_second": 199.01,
|
| 1966 |
+
"eval_steps_per_second": 3.119,
|
| 1967 |
+
"step": 25000
|
| 1968 |
+
},
|
| 1969 |
+
{
|
| 1970 |
+
"epoch": 0.0502,
|
| 1971 |
+
"grad_norm": 1.105361819267273,
|
| 1972 |
+
"learning_rate": 4.979687691618936e-05,
|
| 1973 |
+
"loss": 1.1784,
|
| 1974 |
+
"step": 25100
|
| 1975 |
+
},
|
| 1976 |
+
{
|
| 1977 |
+
"epoch": 0.0504,
|
| 1978 |
+
"grad_norm": 0.7138956189155579,
|
| 1979 |
+
"learning_rate": 4.9794853437973555e-05,
|
| 1980 |
+
"loss": 1.2016,
|
| 1981 |
+
"step": 25200
|
| 1982 |
+
},
|
| 1983 |
+
{
|
| 1984 |
+
"epoch": 0.0506,
|
| 1985 |
+
"grad_norm": 1.250241756439209,
|
| 1986 |
+
"learning_rate": 4.9792819972393756e-05,
|
| 1987 |
+
"loss": 1.2032,
|
| 1988 |
+
"step": 25300
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 0.0508,
|
| 1992 |
+
"grad_norm": 0.5875529050827026,
|
| 1993 |
+
"learning_rate": 4.9790776520269034e-05,
|
| 1994 |
+
"loss": 1.2034,
|
| 1995 |
+
"step": 25400
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.051,
|
| 1999 |
+
"grad_norm": 1.2880475521087646,
|
| 2000 |
+
"learning_rate": 4.9788723082422495e-05,
|
| 2001 |
+
"loss": 1.2172,
|
| 2002 |
+
"step": 25500
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.0512,
|
| 2006 |
+
"grad_norm": 0.8775302767753601,
|
| 2007 |
+
"learning_rate": 4.978665965968127e-05,
|
| 2008 |
+
"loss": 1.2264,
|
| 2009 |
+
"step": 25600
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.0514,
|
| 2013 |
+
"grad_norm": 0.7336851954460144,
|
| 2014 |
+
"learning_rate": 4.978458625287649e-05,
|
| 2015 |
+
"loss": 1.2248,
|
| 2016 |
+
"step": 25700
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.0516,
|
| 2020 |
+
"grad_norm": 1.431084156036377,
|
| 2021 |
+
"learning_rate": 4.978250286284333e-05,
|
| 2022 |
+
"loss": 1.2353,
|
| 2023 |
+
"step": 25800
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 0.0518,
|
| 2027 |
+
"grad_norm": 1.6342276334762573,
|
| 2028 |
+
"learning_rate": 4.978040949042099e-05,
|
| 2029 |
+
"loss": 1.1984,
|
| 2030 |
+
"step": 25900
|
| 2031 |
+
},
|
| 2032 |
+
{
|
| 2033 |
+
"epoch": 0.052,
|
| 2034 |
+
"grad_norm": 1.5883526802062988,
|
| 2035 |
+
"learning_rate": 4.977830613645266e-05,
|
| 2036 |
+
"loss": 1.2251,
|
| 2037 |
+
"step": 26000
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.052,
|
| 2041 |
+
"eval_loss": 1.0901614427566528,
|
| 2042 |
+
"eval_runtime": 76.7254,
|
| 2043 |
+
"eval_samples_per_second": 201.276,
|
| 2044 |
+
"eval_steps_per_second": 3.154,
|
| 2045 |
+
"step": 26000
|
| 2046 |
+
},
|
| 2047 |
+
{
|
| 2048 |
+
"epoch": 0.0522,
|
| 2049 |
+
"grad_norm": 1.1527795791625977,
|
| 2050 |
+
"learning_rate": 4.977619280178558e-05,
|
| 2051 |
+
"loss": 1.2043,
|
| 2052 |
+
"step": 26100
|
| 2053 |
+
},
|
| 2054 |
+
{
|
| 2055 |
+
"epoch": 0.0524,
|
| 2056 |
+
"grad_norm": 1.5160431861877441,
|
| 2057 |
+
"learning_rate": 4.9774069487271014e-05,
|
| 2058 |
+
"loss": 1.1931,
|
| 2059 |
+
"step": 26200
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 0.0526,
|
| 2063 |
+
"grad_norm": 1.2551748752593994,
|
| 2064 |
+
"learning_rate": 4.977193619376421e-05,
|
| 2065 |
+
"loss": 1.2397,
|
| 2066 |
+
"step": 26300
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 0.0528,
|
| 2070 |
+
"grad_norm": 1.2745076417922974,
|
| 2071 |
+
"learning_rate": 4.976979292212448e-05,
|
| 2072 |
+
"loss": 1.2336,
|
| 2073 |
+
"step": 26400
|
| 2074 |
+
},
|
| 2075 |
+
{
|
| 2076 |
+
"epoch": 0.053,
|
| 2077 |
+
"grad_norm": 1.4893673658370972,
|
| 2078 |
+
"learning_rate": 4.976763967321511e-05,
|
| 2079 |
+
"loss": 1.1827,
|
| 2080 |
+
"step": 26500
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 0.0532,
|
| 2084 |
+
"grad_norm": 0.857379138469696,
|
| 2085 |
+
"learning_rate": 4.976547644790346e-05,
|
| 2086 |
+
"loss": 1.2441,
|
| 2087 |
+
"step": 26600
|
| 2088 |
+
},
|
| 2089 |
+
{
|
| 2090 |
+
"epoch": 0.0534,
|
| 2091 |
+
"grad_norm": 1.167006492614746,
|
| 2092 |
+
"learning_rate": 4.976330324706084e-05,
|
| 2093 |
+
"loss": 1.2779,
|
| 2094 |
+
"step": 26700
|
| 2095 |
+
},
|
| 2096 |
+
{
|
| 2097 |
+
"epoch": 0.0536,
|
| 2098 |
+
"grad_norm": 0.634842574596405,
|
| 2099 |
+
"learning_rate": 4.976112007156265e-05,
|
| 2100 |
+
"loss": 1.2828,
|
| 2101 |
+
"step": 26800
|
| 2102 |
+
},
|
| 2103 |
+
{
|
| 2104 |
+
"epoch": 0.0538,
|
| 2105 |
+
"grad_norm": 0.9239290952682495,
|
| 2106 |
+
"learning_rate": 4.975892692228825e-05,
|
| 2107 |
+
"loss": 1.2094,
|
| 2108 |
+
"step": 26900
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 0.054,
|
| 2112 |
+
"grad_norm": 1.2031028270721436,
|
| 2113 |
+
"learning_rate": 4.9756723800121044e-05,
|
| 2114 |
+
"loss": 1.222,
|
| 2115 |
+
"step": 27000
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 0.054,
|
| 2119 |
+
"eval_loss": 1.0867078304290771,
|
| 2120 |
+
"eval_runtime": 76.6606,
|
| 2121 |
+
"eval_samples_per_second": 201.446,
|
| 2122 |
+
"eval_steps_per_second": 3.157,
|
| 2123 |
+
"step": 27000
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.0542,
|
| 2127 |
+
"grad_norm": 1.3575947284698486,
|
| 2128 |
+
"learning_rate": 4.9754510705948456e-05,
|
| 2129 |
+
"loss": 1.1622,
|
| 2130 |
+
"step": 27100
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.0544,
|
| 2134 |
+
"grad_norm": 1.142074465751648,
|
| 2135 |
+
"learning_rate": 4.975228764066191e-05,
|
| 2136 |
+
"loss": 1.2703,
|
| 2137 |
+
"step": 27200
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.0546,
|
| 2141 |
+
"grad_norm": 0.8273721933364868,
|
| 2142 |
+
"learning_rate": 4.975005460515686e-05,
|
| 2143 |
+
"loss": 1.1921,
|
| 2144 |
+
"step": 27300
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.0548,
|
| 2148 |
+
"grad_norm": 1.3859556913375854,
|
| 2149 |
+
"learning_rate": 4.974781160033278e-05,
|
| 2150 |
+
"loss": 1.2195,
|
| 2151 |
+
"step": 27400
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.055,
|
| 2155 |
+
"grad_norm": 1.2232416868209839,
|
| 2156 |
+
"learning_rate": 4.974555862709315e-05,
|
| 2157 |
+
"loss": 1.1851,
|
| 2158 |
+
"step": 27500
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.0552,
|
| 2162 |
+
"grad_norm": 0.7069573998451233,
|
| 2163 |
+
"learning_rate": 4.974329568634546e-05,
|
| 2164 |
+
"loss": 1.2098,
|
| 2165 |
+
"step": 27600
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.0554,
|
| 2169 |
+
"grad_norm": 1.2497153282165527,
|
| 2170 |
+
"learning_rate": 4.974102277900122e-05,
|
| 2171 |
+
"loss": 1.206,
|
| 2172 |
+
"step": 27700
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.0556,
|
| 2176 |
+
"grad_norm": 1.206449031829834,
|
| 2177 |
+
"learning_rate": 4.9738739905975976e-05,
|
| 2178 |
+
"loss": 1.2352,
|
| 2179 |
+
"step": 27800
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.0558,
|
| 2183 |
+
"grad_norm": 1.3927749395370483,
|
| 2184 |
+
"learning_rate": 4.973644706818925e-05,
|
| 2185 |
+
"loss": 1.1952,
|
| 2186 |
+
"step": 27900
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.056,
|
| 2190 |
+
"grad_norm": 1.3856321573257446,
|
| 2191 |
+
"learning_rate": 4.973414426656461e-05,
|
| 2192 |
+
"loss": 1.2499,
|
| 2193 |
+
"step": 28000
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.056,
|
| 2197 |
+
"eval_loss": 1.0941141843795776,
|
| 2198 |
+
"eval_runtime": 76.7063,
|
| 2199 |
+
"eval_samples_per_second": 201.326,
|
| 2200 |
+
"eval_steps_per_second": 3.155,
|
| 2201 |
+
"step": 28000
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.0562,
|
| 2205 |
+
"grad_norm": 0.6676329970359802,
|
| 2206 |
+
"learning_rate": 4.9731831502029606e-05,
|
| 2207 |
+
"loss": 1.2333,
|
| 2208 |
+
"step": 28100
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.0564,
|
| 2212 |
+
"grad_norm": 1.2670732736587524,
|
| 2213 |
+
"learning_rate": 4.972950877551584e-05,
|
| 2214 |
+
"loss": 1.183,
|
| 2215 |
+
"step": 28200
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.0566,
|
| 2219 |
+
"grad_norm": 1.2089595794677734,
|
| 2220 |
+
"learning_rate": 4.972717608795889e-05,
|
| 2221 |
+
"loss": 1.2445,
|
| 2222 |
+
"step": 28300
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.0568,
|
| 2226 |
+
"grad_norm": 1.1897366046905518,
|
| 2227 |
+
"learning_rate": 4.972483344029838e-05,
|
| 2228 |
+
"loss": 1.2217,
|
| 2229 |
+
"step": 28400
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.057,
|
| 2233 |
+
"grad_norm": 1.4963501691818237,
|
| 2234 |
+
"learning_rate": 4.97224808334779e-05,
|
| 2235 |
+
"loss": 1.2079,
|
| 2236 |
+
"step": 28500
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.0572,
|
| 2240 |
+
"grad_norm": 1.594019889831543,
|
| 2241 |
+
"learning_rate": 4.972011826844511e-05,
|
| 2242 |
+
"loss": 1.1822,
|
| 2243 |
+
"step": 28600
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.0574,
|
| 2247 |
+
"grad_norm": 1.3324779272079468,
|
| 2248 |
+
"learning_rate": 4.971774574615163e-05,
|
| 2249 |
+
"loss": 1.2562,
|
| 2250 |
+
"step": 28700
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.0576,
|
| 2254 |
+
"grad_norm": 1.3334344625473022,
|
| 2255 |
+
"learning_rate": 4.971536326755313e-05,
|
| 2256 |
+
"loss": 1.2509,
|
| 2257 |
+
"step": 28800
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.0578,
|
| 2261 |
+
"grad_norm": 0.9475389719009399,
|
| 2262 |
+
"learning_rate": 4.971297083360925e-05,
|
| 2263 |
+
"loss": 1.1826,
|
| 2264 |
+
"step": 28900
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.058,
|
| 2268 |
+
"grad_norm": 0.8067657947540283,
|
| 2269 |
+
"learning_rate": 4.971056844528368e-05,
|
| 2270 |
+
"loss": 1.1895,
|
| 2271 |
+
"step": 29000
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.058,
|
| 2275 |
+
"eval_loss": 1.0870901346206665,
|
| 2276 |
+
"eval_runtime": 76.6141,
|
| 2277 |
+
"eval_samples_per_second": 201.569,
|
| 2278 |
+
"eval_steps_per_second": 3.159,
|
| 2279 |
+
"step": 29000
|
| 2280 |
+
},
|
| 2281 |
+
{
|
| 2282 |
+
"epoch": 0.0582,
|
| 2283 |
+
"grad_norm": 0.7364763617515564,
|
| 2284 |
+
"learning_rate": 4.970815610354409e-05,
|
| 2285 |
+
"loss": 1.1821,
|
| 2286 |
+
"step": 29100
|
| 2287 |
+
},
|
| 2288 |
+
{
|
| 2289 |
+
"epoch": 0.0584,
|
| 2290 |
+
"grad_norm": 1.494878888130188,
|
| 2291 |
+
"learning_rate": 4.970573380936218e-05,
|
| 2292 |
+
"loss": 1.1592,
|
| 2293 |
+
"step": 29200
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.0586,
|
| 2297 |
+
"grad_norm": 0.7247675061225891,
|
| 2298 |
+
"learning_rate": 4.9703301563713645e-05,
|
| 2299 |
+
"loss": 1.2347,
|
| 2300 |
+
"step": 29300
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.0588,
|
| 2304 |
+
"grad_norm": 1.0013625621795654,
|
| 2305 |
+
"learning_rate": 4.970085936757819e-05,
|
| 2306 |
+
"loss": 1.2536,
|
| 2307 |
+
"step": 29400
|
| 2308 |
+
},
|
| 2309 |
+
{
|
| 2310 |
+
"epoch": 0.059,
|
| 2311 |
+
"grad_norm": 1.012537956237793,
|
| 2312 |
+
"learning_rate": 4.969840722193955e-05,
|
| 2313 |
+
"loss": 1.2461,
|
| 2314 |
+
"step": 29500
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.0592,
|
| 2318 |
+
"grad_norm": 0.8702846169471741,
|
| 2319 |
+
"learning_rate": 4.969594512778541e-05,
|
| 2320 |
+
"loss": 1.2005,
|
| 2321 |
+
"step": 29600
|
| 2322 |
+
},
|
| 2323 |
+
{
|
| 2324 |
+
"epoch": 0.0594,
|
| 2325 |
+
"grad_norm": 1.1068499088287354,
|
| 2326 |
+
"learning_rate": 4.969347308610755e-05,
|
| 2327 |
+
"loss": 1.1942,
|
| 2328 |
+
"step": 29700
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 0.0596,
|
| 2332 |
+
"grad_norm": 1.6333682537078857,
|
| 2333 |
+
"learning_rate": 4.969099109790167e-05,
|
| 2334 |
+
"loss": 1.2372,
|
| 2335 |
+
"step": 29800
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.0598,
|
| 2339 |
+
"grad_norm": 1.0337685346603394,
|
| 2340 |
+
"learning_rate": 4.9688499164167536e-05,
|
| 2341 |
+
"loss": 1.2435,
|
| 2342 |
+
"step": 29900
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.06,
|
| 2346 |
+
"grad_norm": 0.8429011702537537,
|
| 2347 |
+
"learning_rate": 4.9685997285908894e-05,
|
| 2348 |
+
"loss": 1.2023,
|
| 2349 |
+
"step": 30000
|
| 2350 |
+
},
|
| 2351 |
+
{
|
| 2352 |
+
"epoch": 0.06,
|
| 2353 |
+
"eval_loss": 1.086748480796814,
|
| 2354 |
+
"eval_runtime": 76.8684,
|
| 2355 |
+
"eval_samples_per_second": 200.902,
|
| 2356 |
+
"eval_steps_per_second": 3.148,
|
| 2357 |
+
"step": 30000
|
| 2358 |
+
},
|
| 2359 |
+
{
|
| 2360 |
+
"epoch": 0.0002,
|
| 2361 |
+
"grad_norm": 0.8381020426750183,
|
| 2362 |
+
"learning_rate": 4.9683485464133484e-05,
|
| 2363 |
+
"loss": 1.2362,
|
| 2364 |
+
"step": 30100
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 0.0004,
|
| 2368 |
+
"grad_norm": 0.6860467791557312,
|
| 2369 |
+
"learning_rate": 4.968096369985309e-05,
|
| 2370 |
+
"loss": 1.2125,
|
| 2371 |
+
"step": 30200
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.0006,
|
| 2375 |
+
"grad_norm": 0.9316505193710327,
|
| 2376 |
+
"learning_rate": 4.967843199408347e-05,
|
| 2377 |
+
"loss": 1.1904,
|
| 2378 |
+
"step": 30300
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.0008,
|
| 2382 |
+
"grad_norm": 1.3389461040496826,
|
| 2383 |
+
"learning_rate": 4.967589034784439e-05,
|
| 2384 |
+
"loss": 1.2689,
|
| 2385 |
+
"step": 30400
|
| 2386 |
+
},
|
| 2387 |
+
{
|
| 2388 |
+
"epoch": 0.001,
|
| 2389 |
+
"grad_norm": 0.9387079477310181,
|
| 2390 |
+
"learning_rate": 4.967333876215963e-05,
|
| 2391 |
+
"loss": 1.2205,
|
| 2392 |
+
"step": 30500
|
| 2393 |
+
},
|
| 2394 |
+
{
|
| 2395 |
+
"epoch": 0.0012,
|
| 2396 |
+
"grad_norm": 0.7549923062324524,
|
| 2397 |
+
"learning_rate": 4.967077723805697e-05,
|
| 2398 |
+
"loss": 1.21,
|
| 2399 |
+
"step": 30600
|
| 2400 |
+
},
|
| 2401 |
+
{
|
| 2402 |
+
"epoch": 0.0014,
|
| 2403 |
+
"grad_norm": 1.1242858171463013,
|
| 2404 |
+
"learning_rate": 4.966820577656819e-05,
|
| 2405 |
+
"loss": 1.203,
|
| 2406 |
+
"step": 30700
|
| 2407 |
+
},
|
| 2408 |
+
{
|
| 2409 |
+
"epoch": 0.0016,
|
| 2410 |
+
"grad_norm": 1.5065937042236328,
|
| 2411 |
+
"learning_rate": 4.966562437872907e-05,
|
| 2412 |
+
"loss": 1.2233,
|
| 2413 |
+
"step": 30800
|
| 2414 |
+
},
|
| 2415 |
+
{
|
| 2416 |
+
"epoch": 0.0018,
|
| 2417 |
+
"grad_norm": 1.1448508501052856,
|
| 2418 |
+
"learning_rate": 4.96630330455794e-05,
|
| 2419 |
+
"loss": 1.2242,
|
| 2420 |
+
"step": 30900
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.002,
|
| 2424 |
+
"grad_norm": 0.7356053590774536,
|
| 2425 |
+
"learning_rate": 4.966043177816296e-05,
|
| 2426 |
+
"loss": 1.2541,
|
| 2427 |
+
"step": 31000
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 0.002,
|
| 2431 |
+
"eval_loss": 1.0892270803451538,
|
| 2432 |
+
"eval_runtime": 78.1396,
|
| 2433 |
+
"eval_samples_per_second": 197.633,
|
| 2434 |
+
"eval_steps_per_second": 3.097,
|
| 2435 |
+
"step": 31000
|
| 2436 |
+
},
|
| 2437 |
+
{
|
| 2438 |
+
"epoch": 0.0022,
|
| 2439 |
+
"grad_norm": 1.290472149848938,
|
| 2440 |
+
"learning_rate": 4.965782057752757e-05,
|
| 2441 |
+
"loss": 1.2005,
|
| 2442 |
+
"step": 31100
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 0.0024,
|
| 2446 |
+
"grad_norm": 0.7970076203346252,
|
| 2447 |
+
"learning_rate": 4.965519944472498e-05,
|
| 2448 |
+
"loss": 1.2718,
|
| 2449 |
+
"step": 31200
|
| 2450 |
+
},
|
| 2451 |
+
{
|
| 2452 |
+
"epoch": 0.0026,
|
| 2453 |
+
"grad_norm": 1.3415039777755737,
|
| 2454 |
+
"learning_rate": 4.9652568380811016e-05,
|
| 2455 |
+
"loss": 1.2673,
|
| 2456 |
+
"step": 31300
|
| 2457 |
+
},
|
| 2458 |
+
{
|
| 2459 |
+
"epoch": 0.0028,
|
| 2460 |
+
"grad_norm": 1.3146836757659912,
|
| 2461 |
+
"learning_rate": 4.9649927386845444e-05,
|
| 2462 |
+
"loss": 1.2717,
|
| 2463 |
+
"step": 31400
|
| 2464 |
+
},
|
| 2465 |
+
{
|
| 2466 |
+
"epoch": 0.003,
|
| 2467 |
+
"grad_norm": 0.9725894927978516,
|
| 2468 |
+
"learning_rate": 4.964727646389208e-05,
|
| 2469 |
+
"loss": 1.2418,
|
| 2470 |
+
"step": 31500
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 0.0032,
|
| 2474 |
+
"grad_norm": 0.9590099453926086,
|
| 2475 |
+
"learning_rate": 4.96446156130187e-05,
|
| 2476 |
+
"loss": 1.2389,
|
| 2477 |
+
"step": 31600
|
| 2478 |
+
},
|
| 2479 |
+
{
|
| 2480 |
+
"epoch": 0.0034,
|
| 2481 |
+
"grad_norm": 1.5478194952011108,
|
| 2482 |
+
"learning_rate": 4.964194483529709e-05,
|
| 2483 |
+
"loss": 1.2693,
|
| 2484 |
+
"step": 31700
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 0.0036,
|
| 2488 |
+
"grad_norm": 0.7029865384101868,
|
| 2489 |
+
"learning_rate": 4.9639264131803056e-05,
|
| 2490 |
+
"loss": 1.25,
|
| 2491 |
+
"step": 31800
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 0.0038,
|
| 2495 |
+
"grad_norm": 0.7784998416900635,
|
| 2496 |
+
"learning_rate": 4.963657350361637e-05,
|
| 2497 |
+
"loss": 1.2339,
|
| 2498 |
+
"step": 31900
|
| 2499 |
+
},
|
| 2500 |
+
{
|
| 2501 |
+
"epoch": 0.004,
|
| 2502 |
+
"grad_norm": 0.6479517817497253,
|
| 2503 |
+
"learning_rate": 4.963387295182083e-05,
|
| 2504 |
+
"loss": 1.2538,
|
| 2505 |
+
"step": 32000
|
| 2506 |
+
},
|
| 2507 |
+
{
|
| 2508 |
+
"epoch": 0.004,
|
| 2509 |
+
"eval_loss": 1.0948545932769775,
|
| 2510 |
+
"eval_runtime": 77.4713,
|
| 2511 |
+
"eval_samples_per_second": 199.338,
|
| 2512 |
+
"eval_steps_per_second": 3.124,
|
| 2513 |
+
"step": 32000
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 0.0042,
|
| 2517 |
+
"grad_norm": 1.4759093523025513,
|
| 2518 |
+
"learning_rate": 4.963116247750421e-05,
|
| 2519 |
+
"loss": 1.2646,
|
| 2520 |
+
"step": 32100
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 0.0044,
|
| 2524 |
+
"grad_norm": 0.7561829686164856,
|
| 2525 |
+
"learning_rate": 4.9628442081758285e-05,
|
| 2526 |
+
"loss": 1.2083,
|
| 2527 |
+
"step": 32200
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 0.0046,
|
| 2531 |
+
"grad_norm": 0.6289774775505066,
|
| 2532 |
+
"learning_rate": 4.962571176567884e-05,
|
| 2533 |
+
"loss": 1.2492,
|
| 2534 |
+
"step": 32300
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 0.0048,
|
| 2538 |
+
"grad_norm": 0.8146848678588867,
|
| 2539 |
+
"learning_rate": 4.962297153036564e-05,
|
| 2540 |
+
"loss": 1.2693,
|
| 2541 |
+
"step": 32400
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 0.005,
|
| 2545 |
+
"grad_norm": 1.1135525703430176,
|
| 2546 |
+
"learning_rate": 4.962022137692245e-05,
|
| 2547 |
+
"loss": 1.2218,
|
| 2548 |
+
"step": 32500
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 0.0052,
|
| 2552 |
+
"grad_norm": 1.1507619619369507,
|
| 2553 |
+
"learning_rate": 4.961746130645703e-05,
|
| 2554 |
+
"loss": 1.2118,
|
| 2555 |
+
"step": 32600
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 0.0054,
|
| 2559 |
+
"grad_norm": 0.8586376905441284,
|
| 2560 |
+
"learning_rate": 4.961469132008114e-05,
|
| 2561 |
+
"loss": 1.2115,
|
| 2562 |
+
"step": 32700
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 0.0056,
|
| 2566 |
+
"grad_norm": 1.5335224866867065,
|
| 2567 |
+
"learning_rate": 4.961191141891054e-05,
|
| 2568 |
+
"loss": 1.2239,
|
| 2569 |
+
"step": 32800
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.0058,
|
| 2573 |
+
"grad_norm": 1.2822892665863037,
|
| 2574 |
+
"learning_rate": 4.960912160406496e-05,
|
| 2575 |
+
"loss": 1.2443,
|
| 2576 |
+
"step": 32900
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 0.006,
|
| 2580 |
+
"grad_norm": 0.9584761261940002,
|
| 2581 |
+
"learning_rate": 4.960632187666814e-05,
|
| 2582 |
+
"loss": 1.243,
|
| 2583 |
+
"step": 33000
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 0.006,
|
| 2587 |
+
"eval_loss": 1.0964241027832031,
|
| 2588 |
+
"eval_runtime": 76.2571,
|
| 2589 |
+
"eval_samples_per_second": 202.512,
|
| 2590 |
+
"eval_steps_per_second": 3.173,
|
| 2591 |
+
"step": 33000
|
| 2592 |
+
},
|
| 2593 |
+
{
|
| 2594 |
+
"epoch": 0.0062,
|
| 2595 |
+
"grad_norm": 0.7512497305870056,
|
| 2596 |
+
"learning_rate": 4.960351223784781e-05,
|
| 2597 |
+
"loss": 1.1821,
|
| 2598 |
+
"step": 33100
|
| 2599 |
+
},
|
| 2600 |
+
{
|
| 2601 |
+
"epoch": 0.0064,
|
| 2602 |
+
"grad_norm": 1.3305505514144897,
|
| 2603 |
+
"learning_rate": 4.960069268873568e-05,
|
| 2604 |
+
"loss": 1.2393,
|
| 2605 |
+
"step": 33200
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 0.0066,
|
| 2609 |
+
"grad_norm": 1.5360506772994995,
|
| 2610 |
+
"learning_rate": 4.959786323046749e-05,
|
| 2611 |
+
"loss": 1.2475,
|
| 2612 |
+
"step": 33300
|
| 2613 |
+
},
|
| 2614 |
+
{
|
| 2615 |
+
"epoch": 0.0068,
|
| 2616 |
+
"grad_norm": 0.7005806565284729,
|
| 2617 |
+
"learning_rate": 4.959502386418293e-05,
|
| 2618 |
+
"loss": 1.2122,
|
| 2619 |
+
"step": 33400
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 0.007,
|
| 2623 |
+
"grad_norm": 1.381052017211914,
|
| 2624 |
+
"learning_rate": 4.95921745910257e-05,
|
| 2625 |
+
"loss": 1.2336,
|
| 2626 |
+
"step": 33500
|
| 2627 |
+
},
|
| 2628 |
+
{
|
| 2629 |
+
"epoch": 0.0072,
|
| 2630 |
+
"grad_norm": 1.074300765991211,
|
| 2631 |
+
"learning_rate": 4.958931541214349e-05,
|
| 2632 |
+
"loss": 1.2661,
|
| 2633 |
+
"step": 33600
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 0.0074,
|
| 2637 |
+
"grad_norm": 1.1441256999969482,
|
| 2638 |
+
"learning_rate": 4.9586446328687967e-05,
|
| 2639 |
+
"loss": 1.2296,
|
| 2640 |
+
"step": 33700
|
| 2641 |
+
},
|
| 2642 |
+
{
|
| 2643 |
+
"epoch": 0.0076,
|
| 2644 |
+
"grad_norm": 0.8737586140632629,
|
| 2645 |
+
"learning_rate": 4.958356734181481e-05,
|
| 2646 |
+
"loss": 1.2067,
|
| 2647 |
+
"step": 33800
|
| 2648 |
+
},
|
| 2649 |
+
{
|
| 2650 |
+
"epoch": 0.0078,
|
| 2651 |
+
"grad_norm": 1.1493791341781616,
|
| 2652 |
+
"learning_rate": 4.958067845268366e-05,
|
| 2653 |
+
"loss": 1.2643,
|
| 2654 |
+
"step": 33900
|
| 2655 |
+
},
|
| 2656 |
+
{
|
| 2657 |
+
"epoch": 0.008,
|
| 2658 |
+
"grad_norm": 1.3028621673583984,
|
| 2659 |
+
"learning_rate": 4.957777966245817e-05,
|
| 2660 |
+
"loss": 1.2427,
|
| 2661 |
+
"step": 34000
|
| 2662 |
+
},
|
| 2663 |
+
{
|
| 2664 |
+
"epoch": 0.008,
|
| 2665 |
+
"eval_loss": 1.096444845199585,
|
| 2666 |
+
"eval_runtime": 76.4253,
|
| 2667 |
+
"eval_samples_per_second": 202.067,
|
| 2668 |
+
"eval_steps_per_second": 3.166,
|
| 2669 |
+
"step": 34000
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 0.0082,
|
| 2673 |
+
"grad_norm": 1.31423819065094,
|
| 2674 |
+
"learning_rate": 4.957487097230597e-05,
|
| 2675 |
+
"loss": 1.2137,
|
| 2676 |
+
"step": 34100
|
| 2677 |
+
},
|
| 2678 |
+
{
|
| 2679 |
+
"epoch": 0.0084,
|
| 2680 |
+
"grad_norm": 1.1846545934677124,
|
| 2681 |
+
"learning_rate": 4.957195238339868e-05,
|
| 2682 |
+
"loss": 1.2141,
|
| 2683 |
+
"step": 34200
|
| 2684 |
+
},
|
| 2685 |
+
{
|
| 2686 |
+
"epoch": 0.0086,
|
| 2687 |
+
"grad_norm": 0.9421952366828918,
|
| 2688 |
+
"learning_rate": 4.9569023896911914e-05,
|
| 2689 |
+
"loss": 1.219,
|
| 2690 |
+
"step": 34300
|
| 2691 |
+
},
|
| 2692 |
+
{
|
| 2693 |
+
"epoch": 0.0088,
|
| 2694 |
+
"grad_norm": 1.4107282161712646,
|
| 2695 |
+
"learning_rate": 4.9566085514025256e-05,
|
| 2696 |
+
"loss": 1.2141,
|
| 2697 |
+
"step": 34400
|
| 2698 |
+
},
|
| 2699 |
+
{
|
| 2700 |
+
"epoch": 0.009,
|
| 2701 |
+
"grad_norm": 0.7364057302474976,
|
| 2702 |
+
"learning_rate": 4.95631372359223e-05,
|
| 2703 |
+
"loss": 1.246,
|
| 2704 |
+
"step": 34500
|
| 2705 |
+
},
|
| 2706 |
+
{
|
| 2707 |
+
"epoch": 0.0092,
|
| 2708 |
+
"grad_norm": 0.8100732564926147,
|
| 2709 |
+
"learning_rate": 4.956017906379059e-05,
|
| 2710 |
+
"loss": 1.1891,
|
| 2711 |
+
"step": 34600
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 0.0094,
|
| 2715 |
+
"grad_norm": 1.2455086708068848,
|
| 2716 |
+
"learning_rate": 4.955721099882169e-05,
|
| 2717 |
+
"loss": 1.2458,
|
| 2718 |
+
"step": 34700
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.0096,
|
| 2722 |
+
"grad_norm": 0.676437497138977,
|
| 2723 |
+
"learning_rate": 4.9554233042211146e-05,
|
| 2724 |
+
"loss": 1.2058,
|
| 2725 |
+
"step": 34800
|
| 2726 |
+
},
|
| 2727 |
+
{
|
| 2728 |
+
"epoch": 0.0098,
|
| 2729 |
+
"grad_norm": 1.3339647054672241,
|
| 2730 |
+
"learning_rate": 4.955124519515847e-05,
|
| 2731 |
+
"loss": 1.2407,
|
| 2732 |
+
"step": 34900
|
| 2733 |
+
},
|
| 2734 |
+
{
|
| 2735 |
+
"epoch": 0.01,
|
| 2736 |
+
"grad_norm": 0.9411395192146301,
|
| 2737 |
+
"learning_rate": 4.954824745886716e-05,
|
| 2738 |
+
"loss": 1.1974,
|
| 2739 |
+
"step": 35000
|
| 2740 |
+
},
|
| 2741 |
+
{
|
| 2742 |
+
"epoch": 0.01,
|
| 2743 |
+
"eval_loss": 1.0945005416870117,
|
| 2744 |
+
"eval_runtime": 76.9422,
|
| 2745 |
+
"eval_samples_per_second": 200.709,
|
| 2746 |
+
"eval_steps_per_second": 3.145,
|
| 2747 |
+
"step": 35000
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.0102,
|
| 2751 |
+
"grad_norm": 0.6638602018356323,
|
| 2752 |
+
"learning_rate": 4.95452398345447e-05,
|
| 2753 |
+
"loss": 1.2259,
|
| 2754 |
+
"step": 35100
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.0104,
|
| 2758 |
+
"grad_norm": 0.6337453722953796,
|
| 2759 |
+
"learning_rate": 4.954222232340259e-05,
|
| 2760 |
+
"loss": 1.1686,
|
| 2761 |
+
"step": 35200
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.0106,
|
| 2765 |
+
"grad_norm": 0.809762179851532,
|
| 2766 |
+
"learning_rate": 4.953919492665625e-05,
|
| 2767 |
+
"loss": 1.2174,
|
| 2768 |
+
"step": 35300
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.0108,
|
| 2772 |
+
"grad_norm": 0.9431924819946289,
|
| 2773 |
+
"learning_rate": 4.953615764552513e-05,
|
| 2774 |
+
"loss": 1.2128,
|
| 2775 |
+
"step": 35400
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.011,
|
| 2779 |
+
"grad_norm": 0.7606577277183533,
|
| 2780 |
+
"learning_rate": 4.953311048123265e-05,
|
| 2781 |
+
"loss": 1.2473,
|
| 2782 |
+
"step": 35500
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.0112,
|
| 2786 |
+
"grad_norm": 1.1843669414520264,
|
| 2787 |
+
"learning_rate": 4.953005343500619e-05,
|
| 2788 |
+
"loss": 1.2194,
|
| 2789 |
+
"step": 35600
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.0114,
|
| 2793 |
+
"grad_norm": 0.9086577296257019,
|
| 2794 |
+
"learning_rate": 4.952698650807715e-05,
|
| 2795 |
+
"loss": 1.2572,
|
| 2796 |
+
"step": 35700
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.0116,
|
| 2800 |
+
"grad_norm": 1.36215078830719,
|
| 2801 |
+
"learning_rate": 4.9523909701680874e-05,
|
| 2802 |
+
"loss": 1.2263,
|
| 2803 |
+
"step": 35800
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.0118,
|
| 2807 |
+
"grad_norm": 0.8537183403968811,
|
| 2808 |
+
"learning_rate": 4.952082301705671e-05,
|
| 2809 |
+
"loss": 1.2297,
|
| 2810 |
+
"step": 35900
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.012,
|
| 2814 |
+
"grad_norm": 0.6182298064231873,
|
| 2815 |
+
"learning_rate": 4.9517726455447955e-05,
|
| 2816 |
+
"loss": 1.2101,
|
| 2817 |
+
"step": 36000
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.012,
|
| 2821 |
+
"eval_loss": 1.0894391536712646,
|
| 2822 |
+
"eval_runtime": 76.3033,
|
| 2823 |
+
"eval_samples_per_second": 202.39,
|
| 2824 |
+
"eval_steps_per_second": 3.172,
|
| 2825 |
+
"step": 36000
|
| 2826 |
+
},
|
| 2827 |
+
{
|
| 2828 |
+
"epoch": 0.0122,
|
| 2829 |
+
"grad_norm": 1.1102640628814697,
|
| 2830 |
+
"learning_rate": 4.951462001810192e-05,
|
| 2831 |
+
"loss": 1.2086,
|
| 2832 |
+
"step": 36100
|
| 2833 |
+
},
|
| 2834 |
+
{
|
| 2835 |
+
"epoch": 0.0124,
|
| 2836 |
+
"grad_norm": 0.9391844868659973,
|
| 2837 |
+
"learning_rate": 4.951150370626988e-05,
|
| 2838 |
+
"loss": 1.2595,
|
| 2839 |
+
"step": 36200
|
| 2840 |
+
},
|
| 2841 |
+
{
|
| 2842 |
+
"epoch": 0.0126,
|
| 2843 |
+
"grad_norm": 1.3386393785476685,
|
| 2844 |
+
"learning_rate": 4.950837752120707e-05,
|
| 2845 |
+
"loss": 1.1953,
|
| 2846 |
+
"step": 36300
|
| 2847 |
+
},
|
| 2848 |
+
{
|
| 2849 |
+
"epoch": 0.0128,
|
| 2850 |
+
"grad_norm": 1.0943065881729126,
|
| 2851 |
+
"learning_rate": 4.950524146417273e-05,
|
| 2852 |
+
"loss": 1.2759,
|
| 2853 |
+
"step": 36400
|
| 2854 |
+
},
|
| 2855 |
+
{
|
| 2856 |
+
"epoch": 0.013,
|
| 2857 |
+
"grad_norm": 0.9743318557739258,
|
| 2858 |
+
"learning_rate": 4.950209553643006e-05,
|
| 2859 |
+
"loss": 1.2421,
|
| 2860 |
+
"step": 36500
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 0.0132,
|
| 2864 |
+
"grad_norm": 1.2555447816848755,
|
| 2865 |
+
"learning_rate": 4.949893973924623e-05,
|
| 2866 |
+
"loss": 1.242,
|
| 2867 |
+
"step": 36600
|
| 2868 |
+
},
|
| 2869 |
+
{
|
| 2870 |
+
"epoch": 0.0134,
|
| 2871 |
+
"grad_norm": 1.3289902210235596,
|
| 2872 |
+
"learning_rate": 4.949577407389241e-05,
|
| 2873 |
+
"loss": 1.2337,
|
| 2874 |
+
"step": 36700
|
| 2875 |
+
},
|
| 2876 |
+
{
|
| 2877 |
+
"epoch": 0.0136,
|
| 2878 |
+
"grad_norm": 0.8806101679801941,
|
| 2879 |
+
"learning_rate": 4.949259854164372e-05,
|
| 2880 |
+
"loss": 1.244,
|
| 2881 |
+
"step": 36800
|
| 2882 |
+
},
|
| 2883 |
+
{
|
| 2884 |
+
"epoch": 0.0138,
|
| 2885 |
+
"grad_norm": 1.211584448814392,
|
| 2886 |
+
"learning_rate": 4.948941314377927e-05,
|
| 2887 |
+
"loss": 1.2344,
|
| 2888 |
+
"step": 36900
|
| 2889 |
+
},
|
| 2890 |
+
{
|
| 2891 |
+
"epoch": 0.014,
|
| 2892 |
+
"grad_norm": 1.6472032070159912,
|
| 2893 |
+
"learning_rate": 4.9486217881582134e-05,
|
| 2894 |
+
"loss": 1.1866,
|
| 2895 |
+
"step": 37000
|
| 2896 |
+
},
|
| 2897 |
+
{
|
| 2898 |
+
"epoch": 0.014,
|
| 2899 |
+
"eval_loss": 1.0940065383911133,
|
| 2900 |
+
"eval_runtime": 76.3383,
|
| 2901 |
+
"eval_samples_per_second": 202.297,
|
| 2902 |
+
"eval_steps_per_second": 3.17,
|
| 2903 |
+
"step": 37000
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 0.0142,
|
| 2907 |
+
"grad_norm": 1.5010918378829956,
|
| 2908 |
+
"learning_rate": 4.948301275633936e-05,
|
| 2909 |
+
"loss": 1.2057,
|
| 2910 |
+
"step": 37100
|
| 2911 |
+
},
|
| 2912 |
+
{
|
| 2913 |
+
"epoch": 0.0144,
|
| 2914 |
+
"grad_norm": 0.6793572306632996,
|
| 2915 |
+
"learning_rate": 4.947979776934197e-05,
|
| 2916 |
+
"loss": 1.2104,
|
| 2917 |
+
"step": 37200
|
| 2918 |
+
},
|
| 2919 |
+
{
|
| 2920 |
+
"epoch": 0.0146,
|
| 2921 |
+
"grad_norm": 0.7654362916946411,
|
| 2922 |
+
"learning_rate": 4.947657292188498e-05,
|
| 2923 |
+
"loss": 1.2266,
|
| 2924 |
+
"step": 37300
|
| 2925 |
+
},
|
| 2926 |
+
{
|
| 2927 |
+
"epoch": 0.0148,
|
| 2928 |
+
"grad_norm": 1.0618220567703247,
|
| 2929 |
+
"learning_rate": 4.947333821526734e-05,
|
| 2930 |
+
"loss": 1.2509,
|
| 2931 |
+
"step": 37400
|
| 2932 |
+
},
|
| 2933 |
+
{
|
| 2934 |
+
"epoch": 0.015,
|
| 2935 |
+
"grad_norm": 1.2712790966033936,
|
| 2936 |
+
"learning_rate": 4.947009365079199e-05,
|
| 2937 |
+
"loss": 1.2179,
|
| 2938 |
+
"step": 37500
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 0.0152,
|
| 2942 |
+
"grad_norm": 1.3342602252960205,
|
| 2943 |
+
"learning_rate": 4.946683922976584e-05,
|
| 2944 |
+
"loss": 1.2224,
|
| 2945 |
+
"step": 37600
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 0.0154,
|
| 2949 |
+
"grad_norm": 0.8218332529067993,
|
| 2950 |
+
"learning_rate": 4.946357495349978e-05,
|
| 2951 |
+
"loss": 1.2402,
|
| 2952 |
+
"step": 37700
|
| 2953 |
+
},
|
| 2954 |
+
{
|
| 2955 |
+
"epoch": 0.0156,
|
| 2956 |
+
"grad_norm": 2.0291969776153564,
|
| 2957 |
+
"learning_rate": 4.946030082330865e-05,
|
| 2958 |
+
"loss": 1.1599,
|
| 2959 |
+
"step": 37800
|
| 2960 |
+
},
|
| 2961 |
+
{
|
| 2962 |
+
"epoch": 0.0158,
|
| 2963 |
+
"grad_norm": 1.5702838897705078,
|
| 2964 |
+
"learning_rate": 4.945701684051128e-05,
|
| 2965 |
+
"loss": 1.1784,
|
| 2966 |
+
"step": 37900
|
| 2967 |
+
},
|
| 2968 |
+
{
|
| 2969 |
+
"epoch": 0.016,
|
| 2970 |
+
"grad_norm": 1.268508791923523,
|
| 2971 |
+
"learning_rate": 4.9453723006430444e-05,
|
| 2972 |
+
"loss": 1.2172,
|
| 2973 |
+
"step": 38000
|
| 2974 |
+
},
|
| 2975 |
+
{
|
| 2976 |
+
"epoch": 0.016,
|
| 2977 |
+
"eval_loss": 1.088572382926941,
|
| 2978 |
+
"eval_runtime": 76.533,
|
| 2979 |
+
"eval_samples_per_second": 201.782,
|
| 2980 |
+
"eval_steps_per_second": 3.162,
|
| 2981 |
+
"step": 38000
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 0.0162,
|
| 2985 |
+
"grad_norm": 1.3127037286758423,
|
| 2986 |
+
"learning_rate": 4.945041932239292e-05,
|
| 2987 |
+
"loss": 1.2299,
|
| 2988 |
+
"step": 38100
|
| 2989 |
+
},
|
| 2990 |
+
{
|
| 2991 |
+
"epoch": 0.0164,
|
| 2992 |
+
"grad_norm": 0.7277888655662537,
|
| 2993 |
+
"learning_rate": 4.9447105789729396e-05,
|
| 2994 |
+
"loss": 1.2655,
|
| 2995 |
+
"step": 38200
|
| 2996 |
+
},
|
| 2997 |
+
{
|
| 2998 |
+
"epoch": 0.0166,
|
| 2999 |
+
"grad_norm": 1.031909704208374,
|
| 3000 |
+
"learning_rate": 4.94437824097746e-05,
|
| 3001 |
+
"loss": 1.2179,
|
| 3002 |
+
"step": 38300
|
| 3003 |
+
},
|
| 3004 |
+
{
|
| 3005 |
+
"epoch": 0.0168,
|
| 3006 |
+
"grad_norm": 1.2462060451507568,
|
| 3007 |
+
"learning_rate": 4.9440449183867166e-05,
|
| 3008 |
+
"loss": 1.2311,
|
| 3009 |
+
"step": 38400
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 0.017,
|
| 3013 |
+
"grad_norm": 0.5426816344261169,
|
| 3014 |
+
"learning_rate": 4.9437106113349716e-05,
|
| 3015 |
+
"loss": 1.1637,
|
| 3016 |
+
"step": 38500
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 0.0172,
|
| 3020 |
+
"grad_norm": 1.2320595979690552,
|
| 3021 |
+
"learning_rate": 4.9433753199568856e-05,
|
| 3022 |
+
"loss": 1.2282,
|
| 3023 |
+
"step": 38600
|
| 3024 |
+
},
|
| 3025 |
+
{
|
| 3026 |
+
"epoch": 0.0174,
|
| 3027 |
+
"grad_norm": 0.928945779800415,
|
| 3028 |
+
"learning_rate": 4.943039044387513e-05,
|
| 3029 |
+
"loss": 1.1936,
|
| 3030 |
+
"step": 38700
|
| 3031 |
+
},
|
| 3032 |
+
{
|
| 3033 |
+
"epoch": 0.0176,
|
| 3034 |
+
"grad_norm": 1.4080160856246948,
|
| 3035 |
+
"learning_rate": 4.9427017847623044e-05,
|
| 3036 |
+
"loss": 1.251,
|
| 3037 |
+
"step": 38800
|
| 3038 |
+
},
|
| 3039 |
+
{
|
| 3040 |
+
"epoch": 0.0178,
|
| 3041 |
+
"grad_norm": 1.3436859846115112,
|
| 3042 |
+
"learning_rate": 4.9423635412171106e-05,
|
| 3043 |
+
"loss": 1.287,
|
| 3044 |
+
"step": 38900
|
| 3045 |
+
},
|
| 3046 |
+
{
|
| 3047 |
+
"epoch": 0.018,
|
| 3048 |
+
"grad_norm": 0.9334709048271179,
|
| 3049 |
+
"learning_rate": 4.9420243138881734e-05,
|
| 3050 |
+
"loss": 1.1766,
|
| 3051 |
+
"step": 39000
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 0.018,
|
| 3055 |
+
"eval_loss": 1.092005968093872,
|
| 3056 |
+
"eval_runtime": 76.2687,
|
| 3057 |
+
"eval_samples_per_second": 202.482,
|
| 3058 |
+
"eval_steps_per_second": 3.173,
|
| 3059 |
+
"step": 39000
|
| 3060 |
+
},
|
| 3061 |
+
{
|
| 3062 |
+
"epoch": 0.0182,
|
| 3063 |
+
"grad_norm": 0.9674895405769348,
|
| 3064 |
+
"learning_rate": 4.9416841029121355e-05,
|
| 3065 |
+
"loss": 1.2388,
|
| 3066 |
+
"step": 39100
|
| 3067 |
+
},
|
| 3068 |
+
{
|
| 3069 |
+
"epoch": 0.0184,
|
| 3070 |
+
"grad_norm": 1.3673955202102661,
|
| 3071 |
+
"learning_rate": 4.941342908426032e-05,
|
| 3072 |
+
"loss": 1.183,
|
| 3073 |
+
"step": 39200
|
| 3074 |
+
},
|
| 3075 |
+
{
|
| 3076 |
+
"epoch": 0.0186,
|
| 3077 |
+
"grad_norm": 0.8423133492469788,
|
| 3078 |
+
"learning_rate": 4.941000730567297e-05,
|
| 3079 |
+
"loss": 1.1847,
|
| 3080 |
+
"step": 39300
|
| 3081 |
+
},
|
| 3082 |
+
{
|
| 3083 |
+
"epoch": 0.0188,
|
| 3084 |
+
"grad_norm": 0.9814749360084534,
|
| 3085 |
+
"learning_rate": 4.94065756947376e-05,
|
| 3086 |
+
"loss": 1.2022,
|
| 3087 |
+
"step": 39400
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 0.019,
|
| 3091 |
+
"grad_norm": 1.125647783279419,
|
| 3092 |
+
"learning_rate": 4.9403134252836456e-05,
|
| 3093 |
+
"loss": 1.1966,
|
| 3094 |
+
"step": 39500
|
| 3095 |
+
},
|
| 3096 |
+
{
|
| 3097 |
+
"epoch": 0.0192,
|
| 3098 |
+
"grad_norm": 0.8501796722412109,
|
| 3099 |
+
"learning_rate": 4.9399682981355755e-05,
|
| 3100 |
+
"loss": 1.2347,
|
| 3101 |
+
"step": 39600
|
| 3102 |
+
},
|
| 3103 |
+
{
|
| 3104 |
+
"epoch": 0.0194,
|
| 3105 |
+
"grad_norm": 0.8226144909858704,
|
| 3106 |
+
"learning_rate": 4.9396221881685665e-05,
|
| 3107 |
+
"loss": 1.2129,
|
| 3108 |
+
"step": 39700
|
| 3109 |
+
},
|
| 3110 |
+
{
|
| 3111 |
+
"epoch": 0.0196,
|
| 3112 |
+
"grad_norm": 0.9265516400337219,
|
| 3113 |
+
"learning_rate": 4.939275095522032e-05,
|
| 3114 |
+
"loss": 1.1917,
|
| 3115 |
+
"step": 39800
|
| 3116 |
+
},
|
| 3117 |
+
{
|
| 3118 |
+
"epoch": 0.0198,
|
| 3119 |
+
"grad_norm": 0.8538194298744202,
|
| 3120 |
+
"learning_rate": 4.938927020335781e-05,
|
| 3121 |
+
"loss": 1.2548,
|
| 3122 |
+
"step": 39900
|
| 3123 |
+
},
|
| 3124 |
+
{
|
| 3125 |
+
"epoch": 0.02,
|
| 3126 |
+
"grad_norm": 1.2129065990447998,
|
| 3127 |
+
"learning_rate": 4.9385779627500174e-05,
|
| 3128 |
+
"loss": 1.2219,
|
| 3129 |
+
"step": 40000
|
| 3130 |
+
},
|
| 3131 |
+
{
|
| 3132 |
+
"epoch": 0.02,
|
| 3133 |
+
"eval_loss": 1.087021827697754,
|
| 3134 |
+
"eval_runtime": 76.3535,
|
| 3135 |
+
"eval_samples_per_second": 202.257,
|
| 3136 |
+
"eval_steps_per_second": 3.169,
|
| 3137 |
+
"step": 40000
|
| 3138 |
+
},
|
| 3139 |
+
{
|
| 3140 |
+
"epoch": 0.0202,
|
| 3141 |
+
"grad_norm": 1.2157970666885376,
|
| 3142 |
+
"learning_rate": 4.938227922905342e-05,
|
| 3143 |
+
"loss": 1.1623,
|
| 3144 |
+
"step": 40100
|
| 3145 |
+
},
|
| 3146 |
+
{
|
| 3147 |
+
"epoch": 0.0204,
|
| 3148 |
+
"grad_norm": 0.6873258948326111,
|
| 3149 |
+
"learning_rate": 4.9378769009427515e-05,
|
| 3150 |
+
"loss": 1.2088,
|
| 3151 |
+
"step": 40200
|
| 3152 |
+
},
|
| 3153 |
+
{
|
| 3154 |
+
"epoch": 0.0206,
|
| 3155 |
+
"grad_norm": 1.139224886894226,
|
| 3156 |
+
"learning_rate": 4.937524897003637e-05,
|
| 3157 |
+
"loss": 1.2158,
|
| 3158 |
+
"step": 40300
|
| 3159 |
+
},
|
| 3160 |
+
{
|
| 3161 |
+
"epoch": 0.0208,
|
| 3162 |
+
"grad_norm": 1.2190488576889038,
|
| 3163 |
+
"learning_rate": 4.9371719112297845e-05,
|
| 3164 |
+
"loss": 1.19,
|
| 3165 |
+
"step": 40400
|
| 3166 |
+
},
|
| 3167 |
+
{
|
| 3168 |
+
"epoch": 0.021,
|
| 3169 |
+
"grad_norm": 1.2439500093460083,
|
| 3170 |
+
"learning_rate": 4.936817943763378e-05,
|
| 3171 |
+
"loss": 1.173,
|
| 3172 |
+
"step": 40500
|
| 3173 |
+
},
|
| 3174 |
+
{
|
| 3175 |
+
"epoch": 0.0212,
|
| 3176 |
+
"grad_norm": 1.030110478401184,
|
| 3177 |
+
"learning_rate": 4.936462994746995e-05,
|
| 3178 |
+
"loss": 1.1995,
|
| 3179 |
+
"step": 40600
|
| 3180 |
+
},
|
| 3181 |
+
{
|
| 3182 |
+
"epoch": 0.0214,
|
| 3183 |
+
"grad_norm": 0.666333794593811,
|
| 3184 |
+
"learning_rate": 4.93610706432361e-05,
|
| 3185 |
+
"loss": 1.2476,
|
| 3186 |
+
"step": 40700
|
| 3187 |
+
},
|
| 3188 |
+
{
|
| 3189 |
+
"epoch": 0.0216,
|
| 3190 |
+
"grad_norm": 0.8477672934532166,
|
| 3191 |
+
"learning_rate": 4.93575015263659e-05,
|
| 3192 |
+
"loss": 1.2225,
|
| 3193 |
+
"step": 40800
|
| 3194 |
+
},
|
| 3195 |
+
{
|
| 3196 |
+
"epoch": 0.0218,
|
| 3197 |
+
"grad_norm": 1.087173342704773,
|
| 3198 |
+
"learning_rate": 4.9353922598296995e-05,
|
| 3199 |
+
"loss": 1.1758,
|
| 3200 |
+
"step": 40900
|
| 3201 |
+
},
|
| 3202 |
+
{
|
| 3203 |
+
"epoch": 0.022,
|
| 3204 |
+
"grad_norm": 1.2760623693466187,
|
| 3205 |
+
"learning_rate": 4.935033386047099e-05,
|
| 3206 |
+
"loss": 1.2811,
|
| 3207 |
+
"step": 41000
|
| 3208 |
+
},
|
| 3209 |
+
{
|
| 3210 |
+
"epoch": 0.022,
|
| 3211 |
+
"eval_loss": 1.082631230354309,
|
| 3212 |
+
"eval_runtime": 75.9811,
|
| 3213 |
+
"eval_samples_per_second": 203.248,
|
| 3214 |
+
"eval_steps_per_second": 3.185,
|
| 3215 |
+
"step": 41000
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.0222,
|
| 3219 |
+
"grad_norm": 1.0236754417419434,
|
| 3220 |
+
"learning_rate": 4.934673531433341e-05,
|
| 3221 |
+
"loss": 1.2283,
|
| 3222 |
+
"step": 41100
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.0224,
|
| 3226 |
+
"grad_norm": 1.509448766708374,
|
| 3227 |
+
"learning_rate": 4.934312696133376e-05,
|
| 3228 |
+
"loss": 1.1989,
|
| 3229 |
+
"step": 41200
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.0226,
|
| 3233 |
+
"grad_norm": 1.2022035121917725,
|
| 3234 |
+
"learning_rate": 4.9339508802925475e-05,
|
| 3235 |
+
"loss": 1.2247,
|
| 3236 |
+
"step": 41300
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.0228,
|
| 3240 |
+
"grad_norm": 1.4019054174423218,
|
| 3241 |
+
"learning_rate": 4.933588084056596e-05,
|
| 3242 |
+
"loss": 1.2201,
|
| 3243 |
+
"step": 41400
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.023,
|
| 3247 |
+
"grad_norm": 1.06856107711792,
|
| 3248 |
+
"learning_rate": 4.933224307571655e-05,
|
| 3249 |
+
"loss": 1.1789,
|
| 3250 |
+
"step": 41500
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.0232,
|
| 3254 |
+
"grad_norm": 1.0807596445083618,
|
| 3255 |
+
"learning_rate": 4.932859550984255e-05,
|
| 3256 |
+
"loss": 1.2361,
|
| 3257 |
+
"step": 41600
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.0234,
|
| 3261 |
+
"grad_norm": 1.20824134349823,
|
| 3262 |
+
"learning_rate": 4.932493814441318e-05,
|
| 3263 |
+
"loss": 1.2167,
|
| 3264 |
+
"step": 41700
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.0236,
|
| 3268 |
+
"grad_norm": 0.7066964507102966,
|
| 3269 |
+
"learning_rate": 4.9321270980901635e-05,
|
| 3270 |
+
"loss": 1.1941,
|
| 3271 |
+
"step": 41800
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.0238,
|
| 3275 |
+
"grad_norm": 0.7342857122421265,
|
| 3276 |
+
"learning_rate": 4.9317594020785044e-05,
|
| 3277 |
+
"loss": 1.1709,
|
| 3278 |
+
"step": 41900
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.024,
|
| 3282 |
+
"grad_norm": 1.239176630973816,
|
| 3283 |
+
"learning_rate": 4.931390726554449e-05,
|
| 3284 |
+
"loss": 1.2238,
|
| 3285 |
+
"step": 42000
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.024,
|
| 3289 |
+
"eval_loss": 1.0859261751174927,
|
| 3290 |
+
"eval_runtime": 76.6051,
|
| 3291 |
+
"eval_samples_per_second": 201.592,
|
| 3292 |
+
"eval_steps_per_second": 3.159,
|
| 3293 |
+
"step": 42000
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 0.0242,
|
| 3297 |
+
"grad_norm": 0.9031541347503662,
|
| 3298 |
+
"learning_rate": 4.9310210716665003e-05,
|
| 3299 |
+
"loss": 1.1621,
|
| 3300 |
+
"step": 42100
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 0.0244,
|
| 3304 |
+
"grad_norm": 0.744767963886261,
|
| 3305 |
+
"learning_rate": 4.930650437563554e-05,
|
| 3306 |
+
"loss": 1.21,
|
| 3307 |
+
"step": 42200
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 0.0246,
|
| 3311 |
+
"grad_norm": 1.2594637870788574,
|
| 3312 |
+
"learning_rate": 4.9302788243949025e-05,
|
| 3313 |
+
"loss": 1.21,
|
| 3314 |
+
"step": 42300
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 0.0248,
|
| 3318 |
+
"grad_norm": 0.67472243309021,
|
| 3319 |
+
"learning_rate": 4.929906232310231e-05,
|
| 3320 |
+
"loss": 1.1785,
|
| 3321 |
+
"step": 42400
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.025,
|
| 3325 |
+
"grad_norm": 1.3947267532348633,
|
| 3326 |
+
"learning_rate": 4.92953266145962e-05,
|
| 3327 |
+
"loss": 1.1598,
|
| 3328 |
+
"step": 42500
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 0.0252,
|
| 3332 |
+
"grad_norm": 0.7739892601966858,
|
| 3333 |
+
"learning_rate": 4.929158111993543e-05,
|
| 3334 |
+
"loss": 1.1492,
|
| 3335 |
+
"step": 42600
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 0.0254,
|
| 3339 |
+
"grad_norm": 0.8620167970657349,
|
| 3340 |
+
"learning_rate": 4.9287825840628695e-05,
|
| 3341 |
+
"loss": 1.1863,
|
| 3342 |
+
"step": 42700
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 0.0256,
|
| 3346 |
+
"grad_norm": 0.7649038434028625,
|
| 3347 |
+
"learning_rate": 4.928406077818861e-05,
|
| 3348 |
+
"loss": 1.1782,
|
| 3349 |
+
"step": 42800
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 0.0258,
|
| 3353 |
+
"grad_norm": 1.2743923664093018,
|
| 3354 |
+
"learning_rate": 4.9280285934131755e-05,
|
| 3355 |
+
"loss": 1.2254,
|
| 3356 |
+
"step": 42900
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 0.026,
|
| 3360 |
+
"grad_norm": 0.6955134272575378,
|
| 3361 |
+
"learning_rate": 4.927650130997862e-05,
|
| 3362 |
+
"loss": 1.2254,
|
| 3363 |
+
"step": 43000
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 0.026,
|
| 3367 |
+
"eval_loss": 1.0833112001419067,
|
| 3368 |
+
"eval_runtime": 77.475,
|
| 3369 |
+
"eval_samples_per_second": 199.329,
|
| 3370 |
+
"eval_steps_per_second": 3.124,
|
| 3371 |
+
"step": 43000
|
| 3372 |
+
},
|
| 3373 |
+
{
|
| 3374 |
+
"epoch": 0.0262,
|
| 3375 |
+
"grad_norm": 0.8997926115989685,
|
| 3376 |
+
"learning_rate": 4.927270690725367e-05,
|
| 3377 |
+
"loss": 1.1989,
|
| 3378 |
+
"step": 43100
|
| 3379 |
+
},
|
| 3380 |
+
{
|
| 3381 |
+
"epoch": 0.0264,
|
| 3382 |
+
"grad_norm": 1.3762701749801636,
|
| 3383 |
+
"learning_rate": 4.9268902727485276e-05,
|
| 3384 |
+
"loss": 1.1928,
|
| 3385 |
+
"step": 43200
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 0.0266,
|
| 3389 |
+
"grad_norm": 0.7553657293319702,
|
| 3390 |
+
"learning_rate": 4.926508877220577e-05,
|
| 3391 |
+
"loss": 1.2266,
|
| 3392 |
+
"step": 43300
|
| 3393 |
+
},
|
| 3394 |
+
{
|
| 3395 |
+
"epoch": 0.0268,
|
| 3396 |
+
"grad_norm": 0.6331331133842468,
|
| 3397 |
+
"learning_rate": 4.92612650429514e-05,
|
| 3398 |
+
"loss": 1.2034,
|
| 3399 |
+
"step": 43400
|
| 3400 |
+
},
|
| 3401 |
+
{
|
| 3402 |
+
"epoch": 0.027,
|
| 3403 |
+
"grad_norm": 0.6229783892631531,
|
| 3404 |
+
"learning_rate": 4.925743154126238e-05,
|
| 3405 |
+
"loss": 1.2123,
|
| 3406 |
+
"step": 43500
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 0.0272,
|
| 3410 |
+
"grad_norm": 1.2101593017578125,
|
| 3411 |
+
"learning_rate": 4.9253588268682835e-05,
|
| 3412 |
+
"loss": 1.2473,
|
| 3413 |
+
"step": 43600
|
| 3414 |
+
},
|
| 3415 |
+
{
|
| 3416 |
+
"epoch": 0.0274,
|
| 3417 |
+
"grad_norm": 1.2178127765655518,
|
| 3418 |
+
"learning_rate": 4.924973522676083e-05,
|
| 3419 |
+
"loss": 1.2391,
|
| 3420 |
+
"step": 43700
|
| 3421 |
+
},
|
| 3422 |
+
{
|
| 3423 |
+
"epoch": 0.0276,
|
| 3424 |
+
"grad_norm": 1.4870595932006836,
|
| 3425 |
+
"learning_rate": 4.924587241704838e-05,
|
| 3426 |
+
"loss": 1.2358,
|
| 3427 |
+
"step": 43800
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 0.0278,
|
| 3431 |
+
"grad_norm": 1.2042150497436523,
|
| 3432 |
+
"learning_rate": 4.924199984110142e-05,
|
| 3433 |
+
"loss": 1.1996,
|
| 3434 |
+
"step": 43900
|
| 3435 |
+
},
|
| 3436 |
+
{
|
| 3437 |
+
"epoch": 0.028,
|
| 3438 |
+
"grad_norm": 1.3220444917678833,
|
| 3439 |
+
"learning_rate": 4.923811750047982e-05,
|
| 3440 |
+
"loss": 1.2052,
|
| 3441 |
+
"step": 44000
|
| 3442 |
+
},
|
| 3443 |
+
{
|
| 3444 |
+
"epoch": 0.028,
|
| 3445 |
+
"eval_loss": 1.0859400033950806,
|
| 3446 |
+
"eval_runtime": 76.6882,
|
| 3447 |
+
"eval_samples_per_second": 201.374,
|
| 3448 |
+
"eval_steps_per_second": 3.156,
|
| 3449 |
+
"step": 44000
|
| 3450 |
+
},
|
| 3451 |
+
{
|
| 3452 |
+
"epoch": 0.0282,
|
| 3453 |
+
"grad_norm": 1.464141607284546,
|
| 3454 |
+
"learning_rate": 4.923422539674739e-05,
|
| 3455 |
+
"loss": 1.2326,
|
| 3456 |
+
"step": 44100
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 0.0284,
|
| 3460 |
+
"grad_norm": 1.2406100034713745,
|
| 3461 |
+
"learning_rate": 4.923032353147187e-05,
|
| 3462 |
+
"loss": 1.2092,
|
| 3463 |
+
"step": 44200
|
| 3464 |
+
},
|
| 3465 |
+
{
|
| 3466 |
+
"epoch": 0.0286,
|
| 3467 |
+
"grad_norm": 0.9459540247917175,
|
| 3468 |
+
"learning_rate": 4.9226411906224935e-05,
|
| 3469 |
+
"loss": 1.2023,
|
| 3470 |
+
"step": 44300
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 0.0288,
|
| 3474 |
+
"grad_norm": 1.2143398523330688,
|
| 3475 |
+
"learning_rate": 4.922249052258217e-05,
|
| 3476 |
+
"loss": 1.2348,
|
| 3477 |
+
"step": 44400
|
| 3478 |
+
},
|
| 3479 |
+
{
|
| 3480 |
+
"epoch": 0.029,
|
| 3481 |
+
"grad_norm": 1.1002607345581055,
|
| 3482 |
+
"learning_rate": 4.921855938212312e-05,
|
| 3483 |
+
"loss": 1.1912,
|
| 3484 |
+
"step": 44500
|
| 3485 |
+
},
|
| 3486 |
+
{
|
| 3487 |
+
"epoch": 0.0292,
|
| 3488 |
+
"grad_norm": 1.169640302658081,
|
| 3489 |
+
"learning_rate": 4.921461848643126e-05,
|
| 3490 |
+
"loss": 1.1797,
|
| 3491 |
+
"step": 44600
|
| 3492 |
+
},
|
| 3493 |
+
{
|
| 3494 |
+
"epoch": 0.0294,
|
| 3495 |
+
"grad_norm": 1.2756543159484863,
|
| 3496 |
+
"learning_rate": 4.921066783709396e-05,
|
| 3497 |
+
"loss": 1.1691,
|
| 3498 |
+
"step": 44700
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 0.0296,
|
| 3502 |
+
"grad_norm": 0.5525041222572327,
|
| 3503 |
+
"learning_rate": 4.920670743570255e-05,
|
| 3504 |
+
"loss": 1.2011,
|
| 3505 |
+
"step": 44800
|
| 3506 |
+
},
|
| 3507 |
+
{
|
| 3508 |
+
"epoch": 0.0298,
|
| 3509 |
+
"grad_norm": 0.7082927823066711,
|
| 3510 |
+
"learning_rate": 4.9202737283852284e-05,
|
| 3511 |
+
"loss": 1.1831,
|
| 3512 |
+
"step": 44900
|
| 3513 |
+
},
|
| 3514 |
+
{
|
| 3515 |
+
"epoch": 0.03,
|
| 3516 |
+
"grad_norm": 0.7773894667625427,
|
| 3517 |
+
"learning_rate": 4.919875738314233e-05,
|
| 3518 |
+
"loss": 1.1947,
|
| 3519 |
+
"step": 45000
|
| 3520 |
+
},
|
| 3521 |
+
{
|
| 3522 |
+
"epoch": 0.03,
|
| 3523 |
+
"eval_loss": 1.0890144109725952,
|
| 3524 |
+
"eval_runtime": 76.6594,
|
| 3525 |
+
"eval_samples_per_second": 201.45,
|
| 3526 |
+
"eval_steps_per_second": 3.157,
|
| 3527 |
+
"step": 45000
|
| 3528 |
+
},
|
| 3529 |
+
{
|
| 3530 |
+
"epoch": 0.0302,
|
| 3531 |
+
"grad_norm": 0.7057791352272034,
|
| 3532 |
+
"learning_rate": 4.91947677351758e-05,
|
| 3533 |
+
"loss": 1.2717,
|
| 3534 |
+
"step": 45100
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 0.0304,
|
| 3538 |
+
"grad_norm": 0.9837706685066223,
|
| 3539 |
+
"learning_rate": 4.919076834155971e-05,
|
| 3540 |
+
"loss": 1.206,
|
| 3541 |
+
"step": 45200
|
| 3542 |
+
},
|
| 3543 |
+
{
|
| 3544 |
+
"epoch": 0.0306,
|
| 3545 |
+
"grad_norm": 0.5716899633407593,
|
| 3546 |
+
"learning_rate": 4.918675920390504e-05,
|
| 3547 |
+
"loss": 1.2071,
|
| 3548 |
+
"step": 45300
|
| 3549 |
+
},
|
| 3550 |
+
{
|
| 3551 |
+
"epoch": 0.0308,
|
| 3552 |
+
"grad_norm": 0.6972540020942688,
|
| 3553 |
+
"learning_rate": 4.918274032382665e-05,
|
| 3554 |
+
"loss": 1.1761,
|
| 3555 |
+
"step": 45400
|
| 3556 |
+
},
|
| 3557 |
+
{
|
| 3558 |
+
"epoch": 0.031,
|
| 3559 |
+
"grad_norm": 1.4802424907684326,
|
| 3560 |
+
"learning_rate": 4.917871170294334e-05,
|
| 3561 |
+
"loss": 1.2109,
|
| 3562 |
+
"step": 45500
|
| 3563 |
+
},
|
| 3564 |
+
{
|
| 3565 |
+
"epoch": 0.0312,
|
| 3566 |
+
"grad_norm": 0.7575565576553345,
|
| 3567 |
+
"learning_rate": 4.9174673342877854e-05,
|
| 3568 |
+
"loss": 1.2169,
|
| 3569 |
+
"step": 45600
|
| 3570 |
+
},
|
| 3571 |
+
{
|
| 3572 |
+
"epoch": 0.0314,
|
| 3573 |
+
"grad_norm": 2.227360963821411,
|
| 3574 |
+
"learning_rate": 4.917062524525684e-05,
|
| 3575 |
+
"loss": 1.1657,
|
| 3576 |
+
"step": 45700
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 0.0316,
|
| 3580 |
+
"grad_norm": 0.8020743727684021,
|
| 3581 |
+
"learning_rate": 4.916656741171086e-05,
|
| 3582 |
+
"loss": 1.2073,
|
| 3583 |
+
"step": 45800
|
| 3584 |
+
},
|
| 3585 |
+
{
|
| 3586 |
+
"epoch": 0.0318,
|
| 3587 |
+
"grad_norm": 1.1863917112350464,
|
| 3588 |
+
"learning_rate": 4.916249984387443e-05,
|
| 3589 |
+
"loss": 1.211,
|
| 3590 |
+
"step": 45900
|
| 3591 |
+
},
|
| 3592 |
+
{
|
| 3593 |
+
"epoch": 0.032,
|
| 3594 |
+
"grad_norm": 0.5976528525352478,
|
| 3595 |
+
"learning_rate": 4.915842254338594e-05,
|
| 3596 |
+
"loss": 1.2468,
|
| 3597 |
+
"step": 46000
|
| 3598 |
+
},
|
| 3599 |
+
{
|
| 3600 |
+
"epoch": 0.032,
|
| 3601 |
+
"eval_loss": 1.0842978954315186,
|
| 3602 |
+
"eval_runtime": 76.5369,
|
| 3603 |
+
"eval_samples_per_second": 201.772,
|
| 3604 |
+
"eval_steps_per_second": 3.162,
|
| 3605 |
+
"step": 46000
|
| 3606 |
+
},
|
| 3607 |
+
{
|
| 3608 |
+
"epoch": 0.0322,
|
| 3609 |
+
"grad_norm": 1.4908519983291626,
|
| 3610 |
+
"learning_rate": 4.915433551188774e-05,
|
| 3611 |
+
"loss": 1.1695,
|
| 3612 |
+
"step": 46100
|
| 3613 |
+
},
|
| 3614 |
+
{
|
| 3615 |
+
"epoch": 0.0324,
|
| 3616 |
+
"grad_norm": 1.1190279722213745,
|
| 3617 |
+
"learning_rate": 4.915023875102609e-05,
|
| 3618 |
+
"loss": 1.2017,
|
| 3619 |
+
"step": 46200
|
| 3620 |
+
},
|
| 3621 |
+
{
|
| 3622 |
+
"epoch": 0.0326,
|
| 3623 |
+
"grad_norm": 1.1334049701690674,
|
| 3624 |
+
"learning_rate": 4.914613226245115e-05,
|
| 3625 |
+
"loss": 1.2083,
|
| 3626 |
+
"step": 46300
|
| 3627 |
+
},
|
| 3628 |
+
{
|
| 3629 |
+
"epoch": 0.0328,
|
| 3630 |
+
"grad_norm": 0.6902172565460205,
|
| 3631 |
+
"learning_rate": 4.914201604781703e-05,
|
| 3632 |
+
"loss": 1.233,
|
| 3633 |
+
"step": 46400
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 0.033,
|
| 3637 |
+
"grad_norm": 0.7509928941726685,
|
| 3638 |
+
"learning_rate": 4.913789010878174e-05,
|
| 3639 |
+
"loss": 1.2437,
|
| 3640 |
+
"step": 46500
|
| 3641 |
+
},
|
| 3642 |
+
{
|
| 3643 |
+
"epoch": 0.0332,
|
| 3644 |
+
"grad_norm": 1.4217336177825928,
|
| 3645 |
+
"learning_rate": 4.9133754447007185e-05,
|
| 3646 |
+
"loss": 1.1909,
|
| 3647 |
+
"step": 46600
|
| 3648 |
+
},
|
| 3649 |
+
{
|
| 3650 |
+
"epoch": 0.0334,
|
| 3651 |
+
"grad_norm": 1.212930679321289,
|
| 3652 |
+
"learning_rate": 4.912960906415923e-05,
|
| 3653 |
+
"loss": 1.1828,
|
| 3654 |
+
"step": 46700
|
| 3655 |
+
},
|
| 3656 |
+
{
|
| 3657 |
+
"epoch": 0.0336,
|
| 3658 |
+
"grad_norm": 1.1408753395080566,
|
| 3659 |
+
"learning_rate": 4.912545396190763e-05,
|
| 3660 |
+
"loss": 1.2118,
|
| 3661 |
+
"step": 46800
|
| 3662 |
+
},
|
| 3663 |
+
{
|
| 3664 |
+
"epoch": 0.0338,
|
| 3665 |
+
"grad_norm": 0.649695634841919,
|
| 3666 |
+
"learning_rate": 4.9121289141926066e-05,
|
| 3667 |
+
"loss": 1.1877,
|
| 3668 |
+
"step": 46900
|
| 3669 |
+
},
|
| 3670 |
+
{
|
| 3671 |
+
"epoch": 0.034,
|
| 3672 |
+
"grad_norm": 1.4613287448883057,
|
| 3673 |
+
"learning_rate": 4.911711460589211e-05,
|
| 3674 |
+
"loss": 1.1977,
|
| 3675 |
+
"step": 47000
|
| 3676 |
+
},
|
| 3677 |
+
{
|
| 3678 |
+
"epoch": 0.034,
|
| 3679 |
+
"eval_loss": 1.0870256423950195,
|
| 3680 |
+
"eval_runtime": 76.7051,
|
| 3681 |
+
"eval_samples_per_second": 201.33,
|
| 3682 |
+
"eval_steps_per_second": 3.155,
|
| 3683 |
+
"step": 47000
|
| 3684 |
+
},
|
| 3685 |
+
{
|
| 3686 |
+
"epoch": 0.0342,
|
| 3687 |
+
"grad_norm": 1.1586204767227173,
|
| 3688 |
+
"learning_rate": 4.9112930355487284e-05,
|
| 3689 |
+
"loss": 1.2222,
|
| 3690 |
+
"step": 47100
|
| 3691 |
+
},
|
| 3692 |
+
{
|
| 3693 |
+
"epoch": 0.0344,
|
| 3694 |
+
"grad_norm": 1.220306158065796,
|
| 3695 |
+
"learning_rate": 4.910873639239699e-05,
|
| 3696 |
+
"loss": 1.1909,
|
| 3697 |
+
"step": 47200
|
| 3698 |
+
},
|
| 3699 |
+
{
|
| 3700 |
+
"epoch": 0.0346,
|
| 3701 |
+
"grad_norm": 0.589338481426239,
|
| 3702 |
+
"learning_rate": 4.910453271831056e-05,
|
| 3703 |
+
"loss": 1.2034,
|
| 3704 |
+
"step": 47300
|
| 3705 |
+
},
|
| 3706 |
+
{
|
| 3707 |
+
"epoch": 0.0348,
|
| 3708 |
+
"grad_norm": 1.4743396043777466,
|
| 3709 |
+
"learning_rate": 4.910031933492123e-05,
|
| 3710 |
+
"loss": 1.2019,
|
| 3711 |
+
"step": 47400
|
| 3712 |
+
},
|
| 3713 |
+
{
|
| 3714 |
+
"epoch": 0.035,
|
| 3715 |
+
"grad_norm": 0.6481319069862366,
|
| 3716 |
+
"learning_rate": 4.909609624392616e-05,
|
| 3717 |
+
"loss": 1.2107,
|
| 3718 |
+
"step": 47500
|
| 3719 |
+
},
|
| 3720 |
+
{
|
| 3721 |
+
"epoch": 0.0352,
|
| 3722 |
+
"grad_norm": 1.1668992042541504,
|
| 3723 |
+
"learning_rate": 4.9091863447026404e-05,
|
| 3724 |
+
"loss": 1.2498,
|
| 3725 |
+
"step": 47600
|
| 3726 |
+
},
|
| 3727 |
+
{
|
| 3728 |
+
"epoch": 0.0354,
|
| 3729 |
+
"grad_norm": 1.115519404411316,
|
| 3730 |
+
"learning_rate": 4.908762094592693e-05,
|
| 3731 |
+
"loss": 1.206,
|
| 3732 |
+
"step": 47700
|
| 3733 |
+
},
|
| 3734 |
+
{
|
| 3735 |
+
"epoch": 0.0356,
|
| 3736 |
+
"grad_norm": 1.3867928981781006,
|
| 3737 |
+
"learning_rate": 4.908336874233662e-05,
|
| 3738 |
+
"loss": 1.2082,
|
| 3739 |
+
"step": 47800
|
| 3740 |
+
},
|
| 3741 |
+
{
|
| 3742 |
+
"epoch": 0.0358,
|
| 3743 |
+
"grad_norm": 0.6380243301391602,
|
| 3744 |
+
"learning_rate": 4.9079106837968264e-05,
|
| 3745 |
+
"loss": 1.1693,
|
| 3746 |
+
"step": 47900
|
| 3747 |
+
},
|
| 3748 |
+
{
|
| 3749 |
+
"epoch": 0.036,
|
| 3750 |
+
"grad_norm": 1.8375539779663086,
|
| 3751 |
+
"learning_rate": 4.907483523453855e-05,
|
| 3752 |
+
"loss": 1.1531,
|
| 3753 |
+
"step": 48000
|
| 3754 |
+
},
|
| 3755 |
+
{
|
| 3756 |
+
"epoch": 0.036,
|
| 3757 |
+
"eval_loss": 1.0780328512191772,
|
| 3758 |
+
"eval_runtime": 76.5805,
|
| 3759 |
+
"eval_samples_per_second": 201.657,
|
| 3760 |
+
"eval_steps_per_second": 3.16,
|
| 3761 |
+
"step": 48000
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.0362,
|
| 3765 |
+
"grad_norm": 1.231332778930664,
|
| 3766 |
+
"learning_rate": 4.907055393376808e-05,
|
| 3767 |
+
"loss": 1.1618,
|
| 3768 |
+
"step": 48100
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.0364,
|
| 3772 |
+
"grad_norm": 1.2306678295135498,
|
| 3773 |
+
"learning_rate": 4.906626293738137e-05,
|
| 3774 |
+
"loss": 1.2365,
|
| 3775 |
+
"step": 48200
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.0366,
|
| 3779 |
+
"grad_norm": 1.057521104812622,
|
| 3780 |
+
"learning_rate": 4.906196224710683e-05,
|
| 3781 |
+
"loss": 1.1775,
|
| 3782 |
+
"step": 48300
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.0368,
|
| 3786 |
+
"grad_norm": 0.9679245352745056,
|
| 3787 |
+
"learning_rate": 4.905765186467677e-05,
|
| 3788 |
+
"loss": 1.2175,
|
| 3789 |
+
"step": 48400
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.037,
|
| 3793 |
+
"grad_norm": 1.325900912284851,
|
| 3794 |
+
"learning_rate": 4.9053331791827404e-05,
|
| 3795 |
+
"loss": 1.1848,
|
| 3796 |
+
"step": 48500
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.0372,
|
| 3800 |
+
"grad_norm": 1.3124104738235474,
|
| 3801 |
+
"learning_rate": 4.9049002030298887e-05,
|
| 3802 |
+
"loss": 1.1779,
|
| 3803 |
+
"step": 48600
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.0374,
|
| 3807 |
+
"grad_norm": 1.7284040451049805,
|
| 3808 |
+
"learning_rate": 4.904466258183522e-05,
|
| 3809 |
+
"loss": 1.2144,
|
| 3810 |
+
"step": 48700
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.0376,
|
| 3814 |
+
"grad_norm": 0.9314505457878113,
|
| 3815 |
+
"learning_rate": 4.904031344818434e-05,
|
| 3816 |
+
"loss": 1.219,
|
| 3817 |
+
"step": 48800
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.0378,
|
| 3821 |
+
"grad_norm": 1.1688934564590454,
|
| 3822 |
+
"learning_rate": 4.903595463109808e-05,
|
| 3823 |
+
"loss": 1.2268,
|
| 3824 |
+
"step": 48900
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.038,
|
| 3828 |
+
"grad_norm": 1.0910236835479736,
|
| 3829 |
+
"learning_rate": 4.903158613233216e-05,
|
| 3830 |
+
"loss": 1.2213,
|
| 3831 |
+
"step": 49000
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.038,
|
| 3835 |
+
"eval_loss": 1.0866200923919678,
|
| 3836 |
+
"eval_runtime": 76.736,
|
| 3837 |
+
"eval_samples_per_second": 201.248,
|
| 3838 |
+
"eval_steps_per_second": 3.154,
|
| 3839 |
+
"step": 49000
|
| 3840 |
+
},
|
| 3841 |
+
{
|
| 3842 |
+
"epoch": 0.0382,
|
| 3843 |
+
"grad_norm": 1.0715341567993164,
|
| 3844 |
+
"learning_rate": 4.902720795364623e-05,
|
| 3845 |
+
"loss": 1.2007,
|
| 3846 |
+
"step": 49100
|
| 3847 |
+
},
|
| 3848 |
+
{
|
| 3849 |
+
"epoch": 0.0384,
|
| 3850 |
+
"grad_norm": 0.6578232645988464,
|
| 3851 |
+
"learning_rate": 4.902282009680381e-05,
|
| 3852 |
+
"loss": 1.2078,
|
| 3853 |
+
"step": 49200
|
| 3854 |
+
},
|
| 3855 |
+
{
|
| 3856 |
+
"epoch": 0.0386,
|
| 3857 |
+
"grad_norm": 1.34630286693573,
|
| 3858 |
+
"learning_rate": 4.9018422563572326e-05,
|
| 3859 |
+
"loss": 1.1894,
|
| 3860 |
+
"step": 49300
|
| 3861 |
+
},
|
| 3862 |
+
{
|
| 3863 |
+
"epoch": 0.0388,
|
| 3864 |
+
"grad_norm": 1.1832722425460815,
|
| 3865 |
+
"learning_rate": 4.9014015355723104e-05,
|
| 3866 |
+
"loss": 1.1846,
|
| 3867 |
+
"step": 49400
|
| 3868 |
+
},
|
| 3869 |
+
{
|
| 3870 |
+
"epoch": 0.039,
|
| 3871 |
+
"grad_norm": 0.9175591468811035,
|
| 3872 |
+
"learning_rate": 4.900959847503137e-05,
|
| 3873 |
+
"loss": 1.1984,
|
| 3874 |
+
"step": 49500
|
| 3875 |
+
},
|
| 3876 |
+
{
|
| 3877 |
+
"epoch": 0.0392,
|
| 3878 |
+
"grad_norm": 1.077879548072815,
|
| 3879 |
+
"learning_rate": 4.9005171923276236e-05,
|
| 3880 |
+
"loss": 1.1868,
|
| 3881 |
+
"step": 49600
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 0.0394,
|
| 3885 |
+
"grad_norm": 0.5999984741210938,
|
| 3886 |
+
"learning_rate": 4.900073570224073e-05,
|
| 3887 |
+
"loss": 1.1816,
|
| 3888 |
+
"step": 49700
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 0.0396,
|
| 3892 |
+
"grad_norm": 1.24228835105896,
|
| 3893 |
+
"learning_rate": 4.899628981371175e-05,
|
| 3894 |
+
"loss": 1.191,
|
| 3895 |
+
"step": 49800
|
| 3896 |
+
},
|
| 3897 |
+
{
|
| 3898 |
+
"epoch": 0.0398,
|
| 3899 |
+
"grad_norm": 0.7666544318199158,
|
| 3900 |
+
"learning_rate": 4.899183425948011e-05,
|
| 3901 |
+
"loss": 1.1813,
|
| 3902 |
+
"step": 49900
|
| 3903 |
+
},
|
| 3904 |
+
{
|
| 3905 |
+
"epoch": 0.04,
|
| 3906 |
+
"grad_norm": 1.2996748685836792,
|
| 3907 |
+
"learning_rate": 4.8987369041340486e-05,
|
| 3908 |
+
"loss": 1.184,
|
| 3909 |
+
"step": 50000
|
| 3910 |
+
},
|
| 3911 |
+
{
|
| 3912 |
+
"epoch": 0.04,
|
| 3913 |
+
"eval_loss": 1.0817583799362183,
|
| 3914 |
+
"eval_runtime": 77.0972,
|
| 3915 |
+
"eval_samples_per_second": 200.306,
|
| 3916 |
+
"eval_steps_per_second": 3.139,
|
| 3917 |
+
"step": 50000
|
| 3918 |
+
},
|
| 3919 |
+
{
|
| 3920 |
+
"epoch": 0.0402,
|
| 3921 |
+
"grad_norm": 1.1717365980148315,
|
| 3922 |
+
"learning_rate": 4.898289416109149e-05,
|
| 3923 |
+
"loss": 1.1936,
|
| 3924 |
+
"step": 50100
|
| 3925 |
+
},
|
| 3926 |
+
{
|
| 3927 |
+
"epoch": 0.0404,
|
| 3928 |
+
"grad_norm": 1.3680170774459839,
|
| 3929 |
+
"learning_rate": 4.8978409620535595e-05,
|
| 3930 |
+
"loss": 1.2138,
|
| 3931 |
+
"step": 50200
|
| 3932 |
+
},
|
| 3933 |
+
{
|
| 3934 |
+
"epoch": 0.0406,
|
| 3935 |
+
"grad_norm": 1.6390254497528076,
|
| 3936 |
+
"learning_rate": 4.897391542147916e-05,
|
| 3937 |
+
"loss": 1.1883,
|
| 3938 |
+
"step": 50300
|
| 3939 |
+
},
|
| 3940 |
+
{
|
| 3941 |
+
"epoch": 0.0408,
|
| 3942 |
+
"grad_norm": 1.2523001432418823,
|
| 3943 |
+
"learning_rate": 4.896941156573247e-05,
|
| 3944 |
+
"loss": 1.2157,
|
| 3945 |
+
"step": 50400
|
| 3946 |
+
},
|
| 3947 |
+
{
|
| 3948 |
+
"epoch": 0.041,
|
| 3949 |
+
"grad_norm": 1.4317930936813354,
|
| 3950 |
+
"learning_rate": 4.896489805510966e-05,
|
| 3951 |
+
"loss": 1.1721,
|
| 3952 |
+
"step": 50500
|
| 3953 |
+
},
|
| 3954 |
+
{
|
| 3955 |
+
"epoch": 0.0412,
|
| 3956 |
+
"grad_norm": 0.9794881939888,
|
| 3957 |
+
"learning_rate": 4.896037489142879e-05,
|
| 3958 |
+
"loss": 1.2073,
|
| 3959 |
+
"step": 50600
|
| 3960 |
+
},
|
| 3961 |
+
{
|
| 3962 |
+
"epoch": 0.0414,
|
| 3963 |
+
"grad_norm": 0.8774665594100952,
|
| 3964 |
+
"learning_rate": 4.895584207651178e-05,
|
| 3965 |
+
"loss": 1.1934,
|
| 3966 |
+
"step": 50700
|
| 3967 |
+
},
|
| 3968 |
+
{
|
| 3969 |
+
"epoch": 0.0416,
|
| 3970 |
+
"grad_norm": 1.421742558479309,
|
| 3971 |
+
"learning_rate": 4.895129961218444e-05,
|
| 3972 |
+
"loss": 1.2078,
|
| 3973 |
+
"step": 50800
|
| 3974 |
+
},
|
| 3975 |
+
{
|
| 3976 |
+
"epoch": 0.0418,
|
| 3977 |
+
"grad_norm": 1.0715827941894531,
|
| 3978 |
+
"learning_rate": 4.894674750027648e-05,
|
| 3979 |
+
"loss": 1.1713,
|
| 3980 |
+
"step": 50900
|
| 3981 |
+
},
|
| 3982 |
+
{
|
| 3983 |
+
"epoch": 0.042,
|
| 3984 |
+
"grad_norm": 0.7623746991157532,
|
| 3985 |
+
"learning_rate": 4.894218574262149e-05,
|
| 3986 |
+
"loss": 1.1779,
|
| 3987 |
+
"step": 51000
|
| 3988 |
+
},
|
| 3989 |
+
{
|
| 3990 |
+
"epoch": 0.042,
|
| 3991 |
+
"eval_loss": 1.0817545652389526,
|
| 3992 |
+
"eval_runtime": 76.5318,
|
| 3993 |
+
"eval_samples_per_second": 201.785,
|
| 3994 |
+
"eval_steps_per_second": 3.162,
|
| 3995 |
+
"step": 51000
|
| 3996 |
+
},
|
| 3997 |
+
{
|
| 3998 |
+
"epoch": 0.0422,
|
| 3999 |
+
"grad_norm": 0.710477888584137,
|
| 4000 |
+
"learning_rate": 4.893761434105695e-05,
|
| 4001 |
+
"loss": 1.1876,
|
| 4002 |
+
"step": 51100
|
| 4003 |
+
},
|
| 4004 |
+
{
|
| 4005 |
+
"epoch": 0.0424,
|
| 4006 |
+
"grad_norm": 1.244310736656189,
|
| 4007 |
+
"learning_rate": 4.893303329742421e-05,
|
| 4008 |
+
"loss": 1.2077,
|
| 4009 |
+
"step": 51200
|
| 4010 |
+
},
|
| 4011 |
+
{
|
| 4012 |
+
"epoch": 0.0426,
|
| 4013 |
+
"grad_norm": 1.6161651611328125,
|
| 4014 |
+
"learning_rate": 4.8928442613568535e-05,
|
| 4015 |
+
"loss": 1.1896,
|
| 4016 |
+
"step": 51300
|
| 4017 |
+
},
|
| 4018 |
+
{
|
| 4019 |
+
"epoch": 0.0428,
|
| 4020 |
+
"grad_norm": 1.0831233263015747,
|
| 4021 |
+
"learning_rate": 4.892384229133902e-05,
|
| 4022 |
+
"loss": 1.1904,
|
| 4023 |
+
"step": 51400
|
| 4024 |
+
},
|
| 4025 |
+
{
|
| 4026 |
+
"epoch": 0.043,
|
| 4027 |
+
"grad_norm": 0.8258353471755981,
|
| 4028 |
+
"learning_rate": 4.89192323325887e-05,
|
| 4029 |
+
"loss": 1.1906,
|
| 4030 |
+
"step": 51500
|
| 4031 |
+
},
|
| 4032 |
+
{
|
| 4033 |
+
"epoch": 0.0432,
|
| 4034 |
+
"grad_norm": 0.7877621054649353,
|
| 4035 |
+
"learning_rate": 4.8914612739174456e-05,
|
| 4036 |
+
"loss": 1.1416,
|
| 4037 |
+
"step": 51600
|
| 4038 |
+
},
|
| 4039 |
+
{
|
| 4040 |
+
"epoch": 0.0434,
|
| 4041 |
+
"grad_norm": 1.2102254629135132,
|
| 4042 |
+
"learning_rate": 4.890998351295706e-05,
|
| 4043 |
+
"loss": 1.1782,
|
| 4044 |
+
"step": 51700
|
| 4045 |
+
},
|
| 4046 |
+
{
|
| 4047 |
+
"epoch": 0.0436,
|
| 4048 |
+
"grad_norm": 1.139289140701294,
|
| 4049 |
+
"learning_rate": 4.890534465580115e-05,
|
| 4050 |
+
"loss": 1.1471,
|
| 4051 |
+
"step": 51800
|
| 4052 |
+
},
|
| 4053 |
+
{
|
| 4054 |
+
"epoch": 0.0438,
|
| 4055 |
+
"grad_norm": 1.2521135807037354,
|
| 4056 |
+
"learning_rate": 4.890069616957529e-05,
|
| 4057 |
+
"loss": 1.206,
|
| 4058 |
+
"step": 51900
|
| 4059 |
+
},
|
| 4060 |
+
{
|
| 4061 |
+
"epoch": 0.044,
|
| 4062 |
+
"grad_norm": 1.3690674304962158,
|
| 4063 |
+
"learning_rate": 4.889603805615187e-05,
|
| 4064 |
+
"loss": 1.2328,
|
| 4065 |
+
"step": 52000
|
| 4066 |
+
},
|
| 4067 |
+
{
|
| 4068 |
+
"epoch": 0.044,
|
| 4069 |
+
"eval_loss": 1.0797057151794434,
|
| 4070 |
+
"eval_runtime": 76.4385,
|
| 4071 |
+
"eval_samples_per_second": 202.032,
|
| 4072 |
+
"eval_steps_per_second": 3.166,
|
| 4073 |
+
"step": 52000
|
| 4074 |
}
|
| 4075 |
],
|
| 4076 |
"logging_steps": 100,
|
|
|
|
| 4090 |
"attributes": {}
|
| 4091 |
}
|
| 4092 |
},
|
| 4093 |
+
"total_flos": 2.269082661421056e+18,
|
| 4094 |
"train_batch_size": 64,
|
| 4095 |
"trial_name": null,
|
| 4096 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5777
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6ae00380f8558778cd5ee84adf2e34d3acc852a57545540e601bcec753dc98c
|
| 3 |
size 5777
|