Upload folder using huggingface_hub
Browse files- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state_0.pth +1 -1
- rng_state_1.pth +1 -1
- rng_state_2.pth +1 -1
- rng_state_3.pth +1 -1
- rng_state_4.pth +1 -1
- rng_state_5.pth +1 -1
- rng_state_6.pth +1 -1
- rng_state_7.pth +1 -1
- rng_state_8.pth +1 -1
- rng_state_9.pth +1 -1
- scheduler.pt +1 -1
- trainer_state.json +1424 -3
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 213625344
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b0f1a637ff5efd3a742cf0adbf6c58e6934d9a4a6215f862074742126fcf08c2
|
| 3 |
size 213625344
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 427334458
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bf290f7d31c224d60ac5c8ed732a0ecc556301b4c61d776c8545f9109d532312
|
| 3 |
size 427334458
|
rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:435089f27ea78cb1b5ffb371da67f47c7a3dea92ab07479122e62b4e8dbeed97
|
| 3 |
size 16433
|
rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:35c3205fe632396691980ab13bb747592db3b39a8f9bc42c6b4bce2ebc4e86d2
|
| 3 |
size 16433
|
rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bdd18dda129c8617269378af8c4207e690d6ed4efdbda0ce1aa5947221052d4a
|
| 3 |
size 16433
|
rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7d14c531768fd1b817e6ad83f2878c07af9aa939a5372ffc3020b84164720063
|
| 3 |
size 16433
|
rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6343b5ecd40e08336f425dc8913f4b43aad1a7465797446cfa1892e3cb3133f3
|
| 3 |
size 16433
|
rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0492bd561ad6444b11560f52c5a570d34e2dcc461489aae4bd2703fa55d13f47
|
| 3 |
size 16433
|
rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a181670929d07bdff88356f8cc3404c6ef87b1e788fc5b1dc6f473ac9b2bc12
|
| 3 |
size 16433
|
rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:23b0154dd6c58d1151b932c0d3de52258209b2d646a7e6e34de818d4dfa12a13
|
| 3 |
size 16433
|
rng_state_8.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a0c53426dcd44d43fe238422e433fb361c04df5585664ed42f08da1365fc112f
|
| 3 |
size 16433
|
rng_state_9.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 16433
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:268b7b1ca55cb9d2d96542e3789b74164ec07561d7774458ecef281e2a3ad163
|
| 3 |
size 16433
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7fc0c0b23c1fbf47eda082a36b0831feacf86e3aa80c09efc720587a7e961c90
|
| 3 |
size 1064
|
trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 2318,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2856,6 +2856,1427 @@
|
|
| 2856 |
"eval_samples_per_second": 610.045,
|
| 2857 |
"eval_steps_per_second": 7.626,
|
| 2858 |
"step": 9272
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2859 |
}
|
| 2860 |
],
|
| 2861 |
"logging_steps": 23,
|
|
@@ -2875,7 +4296,7 @@
|
|
| 2875 |
"attributes": {}
|
| 2876 |
}
|
| 2877 |
},
|
| 2878 |
-
"total_flos":
|
| 2879 |
"train_batch_size": 8,
|
| 2880 |
"trial_name": null,
|
| 2881 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5998188640186312,
|
| 5 |
"eval_steps": 2318,
|
| 6 |
+
"global_step": 13908,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2856 |
"eval_samples_per_second": 610.045,
|
| 2857 |
"eval_steps_per_second": 7.626,
|
| 2858 |
"step": 9272
|
| 2859 |
+
},
|
| 2860 |
+
{
|
| 2861 |
+
"epoch": 0.4007417949713201,
|
| 2862 |
+
"grad_norm": 0.408203125,
|
| 2863 |
+
"learning_rate": 0.0005998532205145915,
|
| 2864 |
+
"loss": 1.0191,
|
| 2865 |
+
"step": 9292
|
| 2866 |
+
},
|
| 2867 |
+
{
|
| 2868 |
+
"epoch": 0.4017337301073878,
|
| 2869 |
+
"grad_norm": 0.46484375,
|
| 2870 |
+
"learning_rate": 0.0005988603004662408,
|
| 2871 |
+
"loss": 1.0074,
|
| 2872 |
+
"step": 9315
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 0.4027256652434554,
|
| 2876 |
+
"grad_norm": 0.353515625,
|
| 2877 |
+
"learning_rate": 0.0005978673804178898,
|
| 2878 |
+
"loss": 1.0052,
|
| 2879 |
+
"step": 9338
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 0.40371760037952303,
|
| 2883 |
+
"grad_norm": 0.361328125,
|
| 2884 |
+
"learning_rate": 0.000596874460369539,
|
| 2885 |
+
"loss": 1.0248,
|
| 2886 |
+
"step": 9361
|
| 2887 |
+
},
|
| 2888 |
+
{
|
| 2889 |
+
"epoch": 0.40470953551559063,
|
| 2890 |
+
"grad_norm": 0.376953125,
|
| 2891 |
+
"learning_rate": 0.000595881540321188,
|
| 2892 |
+
"loss": 1.0328,
|
| 2893 |
+
"step": 9384
|
| 2894 |
+
},
|
| 2895 |
+
{
|
| 2896 |
+
"epoch": 0.40570147065165824,
|
| 2897 |
+
"grad_norm": 0.4453125,
|
| 2898 |
+
"learning_rate": 0.0005948886202728372,
|
| 2899 |
+
"loss": 1.0254,
|
| 2900 |
+
"step": 9407
|
| 2901 |
+
},
|
| 2902 |
+
{
|
| 2903 |
+
"epoch": 0.4066934057877259,
|
| 2904 |
+
"grad_norm": 0.302734375,
|
| 2905 |
+
"learning_rate": 0.0005938957002244862,
|
| 2906 |
+
"loss": 1.0285,
|
| 2907 |
+
"step": 9430
|
| 2908 |
+
},
|
| 2909 |
+
{
|
| 2910 |
+
"epoch": 0.4076853409237935,
|
| 2911 |
+
"grad_norm": 0.439453125,
|
| 2912 |
+
"learning_rate": 0.0005929027801761354,
|
| 2913 |
+
"loss": 1.0077,
|
| 2914 |
+
"step": 9453
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 0.40867727605986115,
|
| 2918 |
+
"grad_norm": 0.3515625,
|
| 2919 |
+
"learning_rate": 0.0005919098601277845,
|
| 2920 |
+
"loss": 1.0181,
|
| 2921 |
+
"step": 9476
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 0.40966921119592875,
|
| 2925 |
+
"grad_norm": 0.33203125,
|
| 2926 |
+
"learning_rate": 0.0005909169400794336,
|
| 2927 |
+
"loss": 1.0427,
|
| 2928 |
+
"step": 9499
|
| 2929 |
+
},
|
| 2930 |
+
{
|
| 2931 |
+
"epoch": 0.41066114633199635,
|
| 2932 |
+
"grad_norm": 0.380859375,
|
| 2933 |
+
"learning_rate": 0.0005899240200310827,
|
| 2934 |
+
"loss": 1.0133,
|
| 2935 |
+
"step": 9522
|
| 2936 |
+
},
|
| 2937 |
+
{
|
| 2938 |
+
"epoch": 0.411653081468064,
|
| 2939 |
+
"grad_norm": 0.40234375,
|
| 2940 |
+
"learning_rate": 0.0005889310999827318,
|
| 2941 |
+
"loss": 1.0187,
|
| 2942 |
+
"step": 9545
|
| 2943 |
+
},
|
| 2944 |
+
{
|
| 2945 |
+
"epoch": 0.4126450166041316,
|
| 2946 |
+
"grad_norm": 0.44921875,
|
| 2947 |
+
"learning_rate": 0.000587938179934381,
|
| 2948 |
+
"loss": 1.0091,
|
| 2949 |
+
"step": 9568
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 0.41363695174019927,
|
| 2953 |
+
"grad_norm": 0.37109375,
|
| 2954 |
+
"learning_rate": 0.0005869452598860301,
|
| 2955 |
+
"loss": 1.002,
|
| 2956 |
+
"step": 9591
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 0.41462888687626687,
|
| 2960 |
+
"grad_norm": 0.47265625,
|
| 2961 |
+
"learning_rate": 0.0005859523398376792,
|
| 2962 |
+
"loss": 1.0321,
|
| 2963 |
+
"step": 9614
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 0.4156208220123345,
|
| 2967 |
+
"grad_norm": 0.4453125,
|
| 2968 |
+
"learning_rate": 0.0005849594197893283,
|
| 2969 |
+
"loss": 1.0244,
|
| 2970 |
+
"step": 9637
|
| 2971 |
+
},
|
| 2972 |
+
{
|
| 2973 |
+
"epoch": 0.4166127571484021,
|
| 2974 |
+
"grad_norm": 0.4375,
|
| 2975 |
+
"learning_rate": 0.0005839664997409773,
|
| 2976 |
+
"loss": 1.0345,
|
| 2977 |
+
"step": 9660
|
| 2978 |
+
},
|
| 2979 |
+
{
|
| 2980 |
+
"epoch": 0.4176046922844697,
|
| 2981 |
+
"grad_norm": 0.357421875,
|
| 2982 |
+
"learning_rate": 0.0005829735796926266,
|
| 2983 |
+
"loss": 1.0157,
|
| 2984 |
+
"step": 9683
|
| 2985 |
+
},
|
| 2986 |
+
{
|
| 2987 |
+
"epoch": 0.4185966274205374,
|
| 2988 |
+
"grad_norm": 0.3515625,
|
| 2989 |
+
"learning_rate": 0.0005819806596442756,
|
| 2990 |
+
"loss": 0.9774,
|
| 2991 |
+
"step": 9706
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 0.419588562556605,
|
| 2995 |
+
"grad_norm": 0.408203125,
|
| 2996 |
+
"learning_rate": 0.0005809877395959248,
|
| 2997 |
+
"loss": 1.0011,
|
| 2998 |
+
"step": 9729
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 0.42058049769267264,
|
| 3002 |
+
"grad_norm": 0.41015625,
|
| 3003 |
+
"learning_rate": 0.0005799948195475738,
|
| 3004 |
+
"loss": 1.0148,
|
| 3005 |
+
"step": 9752
|
| 3006 |
+
},
|
| 3007 |
+
{
|
| 3008 |
+
"epoch": 0.42157243282874024,
|
| 3009 |
+
"grad_norm": 0.53515625,
|
| 3010 |
+
"learning_rate": 0.000579001899499223,
|
| 3011 |
+
"loss": 1.0058,
|
| 3012 |
+
"step": 9775
|
| 3013 |
+
},
|
| 3014 |
+
{
|
| 3015 |
+
"epoch": 0.42256436796480784,
|
| 3016 |
+
"grad_norm": 0.37890625,
|
| 3017 |
+
"learning_rate": 0.000578008979450872,
|
| 3018 |
+
"loss": 1.0241,
|
| 3019 |
+
"step": 9798
|
| 3020 |
+
},
|
| 3021 |
+
{
|
| 3022 |
+
"epoch": 0.4235563031008755,
|
| 3023 |
+
"grad_norm": 0.38671875,
|
| 3024 |
+
"learning_rate": 0.0005770160594025212,
|
| 3025 |
+
"loss": 1.0067,
|
| 3026 |
+
"step": 9821
|
| 3027 |
+
},
|
| 3028 |
+
{
|
| 3029 |
+
"epoch": 0.4245482382369431,
|
| 3030 |
+
"grad_norm": 0.3359375,
|
| 3031 |
+
"learning_rate": 0.0005760231393541703,
|
| 3032 |
+
"loss": 1.0105,
|
| 3033 |
+
"step": 9844
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 0.42554017337301075,
|
| 3037 |
+
"grad_norm": 0.40625,
|
| 3038 |
+
"learning_rate": 0.0005750302193058193,
|
| 3039 |
+
"loss": 1.0226,
|
| 3040 |
+
"step": 9867
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 0.42653210850907836,
|
| 3044 |
+
"grad_norm": 0.392578125,
|
| 3045 |
+
"learning_rate": 0.0005740372992574685,
|
| 3046 |
+
"loss": 1.0243,
|
| 3047 |
+
"step": 9890
|
| 3048 |
+
},
|
| 3049 |
+
{
|
| 3050 |
+
"epoch": 0.427524043645146,
|
| 3051 |
+
"grad_norm": 0.419921875,
|
| 3052 |
+
"learning_rate": 0.0005730443792091175,
|
| 3053 |
+
"loss": 1.0128,
|
| 3054 |
+
"step": 9913
|
| 3055 |
+
},
|
| 3056 |
+
{
|
| 3057 |
+
"epoch": 0.4285159787812136,
|
| 3058 |
+
"grad_norm": 0.373046875,
|
| 3059 |
+
"learning_rate": 0.0005720514591607668,
|
| 3060 |
+
"loss": 0.9993,
|
| 3061 |
+
"step": 9936
|
| 3062 |
+
},
|
| 3063 |
+
{
|
| 3064 |
+
"epoch": 0.4295079139172812,
|
| 3065 |
+
"grad_norm": 0.365234375,
|
| 3066 |
+
"learning_rate": 0.0005710585391124158,
|
| 3067 |
+
"loss": 1.012,
|
| 3068 |
+
"step": 9959
|
| 3069 |
+
},
|
| 3070 |
+
{
|
| 3071 |
+
"epoch": 0.43049984905334887,
|
| 3072 |
+
"grad_norm": 0.302734375,
|
| 3073 |
+
"learning_rate": 0.000570065619064065,
|
| 3074 |
+
"loss": 1.0178,
|
| 3075 |
+
"step": 9982
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 0.43149178418941647,
|
| 3079 |
+
"grad_norm": 0.3671875,
|
| 3080 |
+
"learning_rate": 0.000569072699015714,
|
| 3081 |
+
"loss": 1.0192,
|
| 3082 |
+
"step": 10005
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.43248371932548413,
|
| 3086 |
+
"grad_norm": 0.345703125,
|
| 3087 |
+
"learning_rate": 0.0005680797789673631,
|
| 3088 |
+
"loss": 0.9992,
|
| 3089 |
+
"step": 10028
|
| 3090 |
+
},
|
| 3091 |
+
{
|
| 3092 |
+
"epoch": 0.43347565446155173,
|
| 3093 |
+
"grad_norm": 0.4140625,
|
| 3094 |
+
"learning_rate": 0.0005670868589190123,
|
| 3095 |
+
"loss": 1.0033,
|
| 3096 |
+
"step": 10051
|
| 3097 |
+
},
|
| 3098 |
+
{
|
| 3099 |
+
"epoch": 0.43446758959761933,
|
| 3100 |
+
"grad_norm": 0.359375,
|
| 3101 |
+
"learning_rate": 0.0005660939388706614,
|
| 3102 |
+
"loss": 1.0153,
|
| 3103 |
+
"step": 10074
|
| 3104 |
+
},
|
| 3105 |
+
{
|
| 3106 |
+
"epoch": 0.435459524733687,
|
| 3107 |
+
"grad_norm": 0.36328125,
|
| 3108 |
+
"learning_rate": 0.0005651010188223105,
|
| 3109 |
+
"loss": 0.997,
|
| 3110 |
+
"step": 10097
|
| 3111 |
+
},
|
| 3112 |
+
{
|
| 3113 |
+
"epoch": 0.4364514598697546,
|
| 3114 |
+
"grad_norm": 0.3671875,
|
| 3115 |
+
"learning_rate": 0.0005641080987739596,
|
| 3116 |
+
"loss": 1.0147,
|
| 3117 |
+
"step": 10120
|
| 3118 |
+
},
|
| 3119 |
+
{
|
| 3120 |
+
"epoch": 0.43744339500582224,
|
| 3121 |
+
"grad_norm": 0.3515625,
|
| 3122 |
+
"learning_rate": 0.0005631151787256087,
|
| 3123 |
+
"loss": 1.0046,
|
| 3124 |
+
"step": 10143
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.43843533014188985,
|
| 3128 |
+
"grad_norm": 0.37890625,
|
| 3129 |
+
"learning_rate": 0.0005621222586772579,
|
| 3130 |
+
"loss": 1.0008,
|
| 3131 |
+
"step": 10166
|
| 3132 |
+
},
|
| 3133 |
+
{
|
| 3134 |
+
"epoch": 0.4394272652779575,
|
| 3135 |
+
"grad_norm": 0.41796875,
|
| 3136 |
+
"learning_rate": 0.0005611293386289069,
|
| 3137 |
+
"loss": 1.0315,
|
| 3138 |
+
"step": 10189
|
| 3139 |
+
},
|
| 3140 |
+
{
|
| 3141 |
+
"epoch": 0.4404192004140251,
|
| 3142 |
+
"grad_norm": 0.365234375,
|
| 3143 |
+
"learning_rate": 0.0005601364185805561,
|
| 3144 |
+
"loss": 1.0164,
|
| 3145 |
+
"step": 10212
|
| 3146 |
+
},
|
| 3147 |
+
{
|
| 3148 |
+
"epoch": 0.4414111355500927,
|
| 3149 |
+
"grad_norm": 0.361328125,
|
| 3150 |
+
"learning_rate": 0.0005591434985322051,
|
| 3151 |
+
"loss": 1.0044,
|
| 3152 |
+
"step": 10235
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 0.44240307068616036,
|
| 3156 |
+
"grad_norm": 0.439453125,
|
| 3157 |
+
"learning_rate": 0.0005581505784838543,
|
| 3158 |
+
"loss": 1.0236,
|
| 3159 |
+
"step": 10258
|
| 3160 |
+
},
|
| 3161 |
+
{
|
| 3162 |
+
"epoch": 0.44339500582222796,
|
| 3163 |
+
"grad_norm": 0.341796875,
|
| 3164 |
+
"learning_rate": 0.0005571576584355034,
|
| 3165 |
+
"loss": 1.0156,
|
| 3166 |
+
"step": 10281
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.4443869409582956,
|
| 3170 |
+
"grad_norm": 0.373046875,
|
| 3171 |
+
"learning_rate": 0.0005561647383871526,
|
| 3172 |
+
"loss": 0.9916,
|
| 3173 |
+
"step": 10304
|
| 3174 |
+
},
|
| 3175 |
+
{
|
| 3176 |
+
"epoch": 0.4453788760943632,
|
| 3177 |
+
"grad_norm": 0.33203125,
|
| 3178 |
+
"learning_rate": 0.0005551718183388016,
|
| 3179 |
+
"loss": 0.9961,
|
| 3180 |
+
"step": 10327
|
| 3181 |
+
},
|
| 3182 |
+
{
|
| 3183 |
+
"epoch": 0.4463708112304308,
|
| 3184 |
+
"grad_norm": 0.392578125,
|
| 3185 |
+
"learning_rate": 0.0005541788982904507,
|
| 3186 |
+
"loss": 1.0021,
|
| 3187 |
+
"step": 10350
|
| 3188 |
+
},
|
| 3189 |
+
{
|
| 3190 |
+
"epoch": 0.4473627463664985,
|
| 3191 |
+
"grad_norm": 0.375,
|
| 3192 |
+
"learning_rate": 0.0005531859782420998,
|
| 3193 |
+
"loss": 1.0219,
|
| 3194 |
+
"step": 10373
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 0.4483546815025661,
|
| 3198 |
+
"grad_norm": 0.4140625,
|
| 3199 |
+
"learning_rate": 0.000552193058193749,
|
| 3200 |
+
"loss": 0.9982,
|
| 3201 |
+
"step": 10396
|
| 3202 |
+
},
|
| 3203 |
+
{
|
| 3204 |
+
"epoch": 0.44934661663863373,
|
| 3205 |
+
"grad_norm": 0.392578125,
|
| 3206 |
+
"learning_rate": 0.0005512001381453981,
|
| 3207 |
+
"loss": 0.994,
|
| 3208 |
+
"step": 10419
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 0.45033855177470133,
|
| 3212 |
+
"grad_norm": 0.34765625,
|
| 3213 |
+
"learning_rate": 0.0005502072180970471,
|
| 3214 |
+
"loss": 0.9899,
|
| 3215 |
+
"step": 10442
|
| 3216 |
+
},
|
| 3217 |
+
{
|
| 3218 |
+
"epoch": 0.451330486910769,
|
| 3219 |
+
"grad_norm": 0.3828125,
|
| 3220 |
+
"learning_rate": 0.0005492142980486963,
|
| 3221 |
+
"loss": 1.0096,
|
| 3222 |
+
"step": 10465
|
| 3223 |
+
},
|
| 3224 |
+
{
|
| 3225 |
+
"epoch": 0.4523224220468366,
|
| 3226 |
+
"grad_norm": 0.3984375,
|
| 3227 |
+
"learning_rate": 0.0005482213780003453,
|
| 3228 |
+
"loss": 0.9882,
|
| 3229 |
+
"step": 10488
|
| 3230 |
+
},
|
| 3231 |
+
{
|
| 3232 |
+
"epoch": 0.4533143571829042,
|
| 3233 |
+
"grad_norm": 0.390625,
|
| 3234 |
+
"learning_rate": 0.0005472284579519945,
|
| 3235 |
+
"loss": 0.999,
|
| 3236 |
+
"step": 10511
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 0.45430629231897185,
|
| 3240 |
+
"grad_norm": 0.3984375,
|
| 3241 |
+
"learning_rate": 0.0005462355379036436,
|
| 3242 |
+
"loss": 1.0087,
|
| 3243 |
+
"step": 10534
|
| 3244 |
+
},
|
| 3245 |
+
{
|
| 3246 |
+
"epoch": 0.45529822745503945,
|
| 3247 |
+
"grad_norm": 0.388671875,
|
| 3248 |
+
"learning_rate": 0.0005452426178552927,
|
| 3249 |
+
"loss": 0.9985,
|
| 3250 |
+
"step": 10557
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 0.4562901625911071,
|
| 3254 |
+
"grad_norm": 0.455078125,
|
| 3255 |
+
"learning_rate": 0.0005442496978069418,
|
| 3256 |
+
"loss": 1.0104,
|
| 3257 |
+
"step": 10580
|
| 3258 |
+
},
|
| 3259 |
+
{
|
| 3260 |
+
"epoch": 0.4572820977271747,
|
| 3261 |
+
"grad_norm": 0.61328125,
|
| 3262 |
+
"learning_rate": 0.0005432567777585909,
|
| 3263 |
+
"loss": 1.0056,
|
| 3264 |
+
"step": 10603
|
| 3265 |
+
},
|
| 3266 |
+
{
|
| 3267 |
+
"epoch": 0.4582740328632423,
|
| 3268 |
+
"grad_norm": 0.3359375,
|
| 3269 |
+
"learning_rate": 0.00054226385771024,
|
| 3270 |
+
"loss": 1.0115,
|
| 3271 |
+
"step": 10626
|
| 3272 |
+
},
|
| 3273 |
+
{
|
| 3274 |
+
"epoch": 0.45926596799930997,
|
| 3275 |
+
"grad_norm": 0.3515625,
|
| 3276 |
+
"learning_rate": 0.0005412709376618892,
|
| 3277 |
+
"loss": 1.0143,
|
| 3278 |
+
"step": 10649
|
| 3279 |
+
},
|
| 3280 |
+
{
|
| 3281 |
+
"epoch": 0.46025790313537757,
|
| 3282 |
+
"grad_norm": 0.388671875,
|
| 3283 |
+
"learning_rate": 0.0005402780176135383,
|
| 3284 |
+
"loss": 0.9916,
|
| 3285 |
+
"step": 10672
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.4612498382714452,
|
| 3289 |
+
"grad_norm": 0.396484375,
|
| 3290 |
+
"learning_rate": 0.0005392850975651874,
|
| 3291 |
+
"loss": 0.9967,
|
| 3292 |
+
"step": 10695
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 0.4622417734075128,
|
| 3296 |
+
"grad_norm": 0.41796875,
|
| 3297 |
+
"learning_rate": 0.0005382921775168364,
|
| 3298 |
+
"loss": 1.0009,
|
| 3299 |
+
"step": 10718
|
| 3300 |
+
},
|
| 3301 |
+
{
|
| 3302 |
+
"epoch": 0.4632337085435805,
|
| 3303 |
+
"grad_norm": 0.34765625,
|
| 3304 |
+
"learning_rate": 0.0005372992574684856,
|
| 3305 |
+
"loss": 0.9919,
|
| 3306 |
+
"step": 10741
|
| 3307 |
+
},
|
| 3308 |
+
{
|
| 3309 |
+
"epoch": 0.4642256436796481,
|
| 3310 |
+
"grad_norm": 0.3515625,
|
| 3311 |
+
"learning_rate": 0.0005363063374201347,
|
| 3312 |
+
"loss": 1.0128,
|
| 3313 |
+
"step": 10764
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 0.4652175788157157,
|
| 3317 |
+
"grad_norm": 0.400390625,
|
| 3318 |
+
"learning_rate": 0.0005353134173717839,
|
| 3319 |
+
"loss": 0.9982,
|
| 3320 |
+
"step": 10787
|
| 3321 |
+
},
|
| 3322 |
+
{
|
| 3323 |
+
"epoch": 0.46620951395178334,
|
| 3324 |
+
"grad_norm": 0.3515625,
|
| 3325 |
+
"learning_rate": 0.0005343204973234329,
|
| 3326 |
+
"loss": 0.9998,
|
| 3327 |
+
"step": 10810
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.46720144908785094,
|
| 3331 |
+
"grad_norm": 0.5390625,
|
| 3332 |
+
"learning_rate": 0.0005333275772750821,
|
| 3333 |
+
"loss": 1.0177,
|
| 3334 |
+
"step": 10833
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 0.4681933842239186,
|
| 3338 |
+
"grad_norm": 0.37890625,
|
| 3339 |
+
"learning_rate": 0.0005323346572267311,
|
| 3340 |
+
"loss": 0.9899,
|
| 3341 |
+
"step": 10856
|
| 3342 |
+
},
|
| 3343 |
+
{
|
| 3344 |
+
"epoch": 0.4691853193599862,
|
| 3345 |
+
"grad_norm": 0.38671875,
|
| 3346 |
+
"learning_rate": 0.0005313417371783802,
|
| 3347 |
+
"loss": 1.0052,
|
| 3348 |
+
"step": 10879
|
| 3349 |
+
},
|
| 3350 |
+
{
|
| 3351 |
+
"epoch": 0.4701772544960538,
|
| 3352 |
+
"grad_norm": 0.36328125,
|
| 3353 |
+
"learning_rate": 0.0005303488171300294,
|
| 3354 |
+
"loss": 0.9741,
|
| 3355 |
+
"step": 10902
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 0.47116918963212145,
|
| 3359 |
+
"grad_norm": 0.4453125,
|
| 3360 |
+
"learning_rate": 0.0005293558970816785,
|
| 3361 |
+
"loss": 1.0021,
|
| 3362 |
+
"step": 10925
|
| 3363 |
+
},
|
| 3364 |
+
{
|
| 3365 |
+
"epoch": 0.47216112476818906,
|
| 3366 |
+
"grad_norm": 0.322265625,
|
| 3367 |
+
"learning_rate": 0.0005283629770333276,
|
| 3368 |
+
"loss": 0.9896,
|
| 3369 |
+
"step": 10948
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.4731530599042567,
|
| 3373 |
+
"grad_norm": 0.36328125,
|
| 3374 |
+
"learning_rate": 0.0005273700569849767,
|
| 3375 |
+
"loss": 1.0046,
|
| 3376 |
+
"step": 10971
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 0.4741449950403243,
|
| 3380 |
+
"grad_norm": 0.345703125,
|
| 3381 |
+
"learning_rate": 0.0005263771369366258,
|
| 3382 |
+
"loss": 1.0004,
|
| 3383 |
+
"step": 10994
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.47513693017639197,
|
| 3387 |
+
"grad_norm": 0.357421875,
|
| 3388 |
+
"learning_rate": 0.0005253842168882749,
|
| 3389 |
+
"loss": 1.0031,
|
| 3390 |
+
"step": 11017
|
| 3391 |
+
},
|
| 3392 |
+
{
|
| 3393 |
+
"epoch": 0.47612886531245957,
|
| 3394 |
+
"grad_norm": 0.359375,
|
| 3395 |
+
"learning_rate": 0.0005243912968399241,
|
| 3396 |
+
"loss": 1.007,
|
| 3397 |
+
"step": 11040
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 0.47712080044852717,
|
| 3401 |
+
"grad_norm": 0.38671875,
|
| 3402 |
+
"learning_rate": 0.0005233983767915731,
|
| 3403 |
+
"loss": 1.0046,
|
| 3404 |
+
"step": 11063
|
| 3405 |
+
},
|
| 3406 |
+
{
|
| 3407 |
+
"epoch": 0.47811273558459483,
|
| 3408 |
+
"grad_norm": 0.341796875,
|
| 3409 |
+
"learning_rate": 0.0005224054567432222,
|
| 3410 |
+
"loss": 0.9956,
|
| 3411 |
+
"step": 11086
|
| 3412 |
+
},
|
| 3413 |
+
{
|
| 3414 |
+
"epoch": 0.47910467072066243,
|
| 3415 |
+
"grad_norm": 0.3515625,
|
| 3416 |
+
"learning_rate": 0.0005214125366948713,
|
| 3417 |
+
"loss": 1.01,
|
| 3418 |
+
"step": 11109
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 0.4800966058567301,
|
| 3422 |
+
"grad_norm": 0.431640625,
|
| 3423 |
+
"learning_rate": 0.0005204196166465205,
|
| 3424 |
+
"loss": 1.0211,
|
| 3425 |
+
"step": 11132
|
| 3426 |
+
},
|
| 3427 |
+
{
|
| 3428 |
+
"epoch": 0.4810885409927977,
|
| 3429 |
+
"grad_norm": 0.375,
|
| 3430 |
+
"learning_rate": 0.0005194266965981696,
|
| 3431 |
+
"loss": 1.0039,
|
| 3432 |
+
"step": 11155
|
| 3433 |
+
},
|
| 3434 |
+
{
|
| 3435 |
+
"epoch": 0.4820804761288653,
|
| 3436 |
+
"grad_norm": 0.392578125,
|
| 3437 |
+
"learning_rate": 0.0005184337765498187,
|
| 3438 |
+
"loss": 0.9886,
|
| 3439 |
+
"step": 11178
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 0.48307241126493294,
|
| 3443 |
+
"grad_norm": 0.515625,
|
| 3444 |
+
"learning_rate": 0.0005174408565014678,
|
| 3445 |
+
"loss": 0.9973,
|
| 3446 |
+
"step": 11201
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.48406434640100054,
|
| 3450 |
+
"grad_norm": 0.396484375,
|
| 3451 |
+
"learning_rate": 0.0005164479364531169,
|
| 3452 |
+
"loss": 1.013,
|
| 3453 |
+
"step": 11224
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"epoch": 0.4850562815370682,
|
| 3457 |
+
"grad_norm": 0.451171875,
|
| 3458 |
+
"learning_rate": 0.000515455016404766,
|
| 3459 |
+
"loss": 0.9876,
|
| 3460 |
+
"step": 11247
|
| 3461 |
+
},
|
| 3462 |
+
{
|
| 3463 |
+
"epoch": 0.4860482166731358,
|
| 3464 |
+
"grad_norm": 0.375,
|
| 3465 |
+
"learning_rate": 0.0005144620963564152,
|
| 3466 |
+
"loss": 0.9984,
|
| 3467 |
+
"step": 11270
|
| 3468 |
+
},
|
| 3469 |
+
{
|
| 3470 |
+
"epoch": 0.48704015180920346,
|
| 3471 |
+
"grad_norm": 0.416015625,
|
| 3472 |
+
"learning_rate": 0.0005134691763080642,
|
| 3473 |
+
"loss": 0.9875,
|
| 3474 |
+
"step": 11293
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 0.48803208694527106,
|
| 3478 |
+
"grad_norm": 0.369140625,
|
| 3479 |
+
"learning_rate": 0.0005124762562597134,
|
| 3480 |
+
"loss": 0.9954,
|
| 3481 |
+
"step": 11316
|
| 3482 |
+
},
|
| 3483 |
+
{
|
| 3484 |
+
"epoch": 0.48902402208133866,
|
| 3485 |
+
"grad_norm": 0.337890625,
|
| 3486 |
+
"learning_rate": 0.0005114833362113624,
|
| 3487 |
+
"loss": 0.9825,
|
| 3488 |
+
"step": 11339
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.4900159572174063,
|
| 3492 |
+
"grad_norm": 0.37890625,
|
| 3493 |
+
"learning_rate": 0.0005104904161630117,
|
| 3494 |
+
"loss": 0.9983,
|
| 3495 |
+
"step": 11362
|
| 3496 |
+
},
|
| 3497 |
+
{
|
| 3498 |
+
"epoch": 0.4910078923534739,
|
| 3499 |
+
"grad_norm": 0.328125,
|
| 3500 |
+
"learning_rate": 0.0005094974961146607,
|
| 3501 |
+
"loss": 0.9818,
|
| 3502 |
+
"step": 11385
|
| 3503 |
+
},
|
| 3504 |
+
{
|
| 3505 |
+
"epoch": 0.4919998274895416,
|
| 3506 |
+
"grad_norm": 0.357421875,
|
| 3507 |
+
"learning_rate": 0.0005085045760663098,
|
| 3508 |
+
"loss": 0.9928,
|
| 3509 |
+
"step": 11408
|
| 3510 |
+
},
|
| 3511 |
+
{
|
| 3512 |
+
"epoch": 0.4929917626256092,
|
| 3513 |
+
"grad_norm": 0.4921875,
|
| 3514 |
+
"learning_rate": 0.0005075116560179589,
|
| 3515 |
+
"loss": 0.9771,
|
| 3516 |
+
"step": 11431
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 0.4939836977616768,
|
| 3520 |
+
"grad_norm": 0.341796875,
|
| 3521 |
+
"learning_rate": 0.000506518735969608,
|
| 3522 |
+
"loss": 1.0059,
|
| 3523 |
+
"step": 11454
|
| 3524 |
+
},
|
| 3525 |
+
{
|
| 3526 |
+
"epoch": 0.49497563289774443,
|
| 3527 |
+
"grad_norm": 0.400390625,
|
| 3528 |
+
"learning_rate": 0.0005055258159212571,
|
| 3529 |
+
"loss": 1.0058,
|
| 3530 |
+
"step": 11477
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.49596756803381203,
|
| 3534 |
+
"grad_norm": 0.328125,
|
| 3535 |
+
"learning_rate": 0.0005045328958729063,
|
| 3536 |
+
"loss": 0.9962,
|
| 3537 |
+
"step": 11500
|
| 3538 |
+
},
|
| 3539 |
+
{
|
| 3540 |
+
"epoch": 0.4969595031698797,
|
| 3541 |
+
"grad_norm": 0.326171875,
|
| 3542 |
+
"learning_rate": 0.0005035399758245554,
|
| 3543 |
+
"loss": 0.9828,
|
| 3544 |
+
"step": 11523
|
| 3545 |
+
},
|
| 3546 |
+
{
|
| 3547 |
+
"epoch": 0.4979514383059473,
|
| 3548 |
+
"grad_norm": 0.5078125,
|
| 3549 |
+
"learning_rate": 0.0005025470557762045,
|
| 3550 |
+
"loss": 0.9881,
|
| 3551 |
+
"step": 11546
|
| 3552 |
+
},
|
| 3553 |
+
{
|
| 3554 |
+
"epoch": 0.49894337344201495,
|
| 3555 |
+
"grad_norm": 0.43359375,
|
| 3556 |
+
"learning_rate": 0.0005015541357278536,
|
| 3557 |
+
"loss": 0.9863,
|
| 3558 |
+
"step": 11569
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 0.49984905334885926,
|
| 3562 |
+
"eval_runtime": 163.9862,
|
| 3563 |
+
"eval_samples_per_second": 609.807,
|
| 3564 |
+
"eval_steps_per_second": 7.623,
|
| 3565 |
+
"step": 11590
|
| 3566 |
+
},
|
| 3567 |
+
{
|
| 3568 |
+
"epoch": 0.49993530857808255,
|
| 3569 |
+
"grad_norm": 0.353515625,
|
| 3570 |
+
"learning_rate": 0.0005005612156795026,
|
| 3571 |
+
"loss": 0.9764,
|
| 3572 |
+
"step": 11592
|
| 3573 |
+
},
|
| 3574 |
+
{
|
| 3575 |
+
"epoch": 0.5009272437141502,
|
| 3576 |
+
"grad_norm": 0.36328125,
|
| 3577 |
+
"learning_rate": 0.0004995682956311518,
|
| 3578 |
+
"loss": 0.9923,
|
| 3579 |
+
"step": 11615
|
| 3580 |
+
},
|
| 3581 |
+
{
|
| 3582 |
+
"epoch": 0.5019191788502178,
|
| 3583 |
+
"grad_norm": 0.39453125,
|
| 3584 |
+
"learning_rate": 0.0004985753755828009,
|
| 3585 |
+
"loss": 0.9738,
|
| 3586 |
+
"step": 11638
|
| 3587 |
+
},
|
| 3588 |
+
{
|
| 3589 |
+
"epoch": 0.5029111139862854,
|
| 3590 |
+
"grad_norm": 0.48828125,
|
| 3591 |
+
"learning_rate": 0.00049758245553445,
|
| 3592 |
+
"loss": 0.973,
|
| 3593 |
+
"step": 11661
|
| 3594 |
+
},
|
| 3595 |
+
{
|
| 3596 |
+
"epoch": 0.5039030491223531,
|
| 3597 |
+
"grad_norm": 0.384765625,
|
| 3598 |
+
"learning_rate": 0.0004965895354860991,
|
| 3599 |
+
"loss": 0.9741,
|
| 3600 |
+
"step": 11684
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 0.5048949842584207,
|
| 3604 |
+
"grad_norm": 0.359375,
|
| 3605 |
+
"learning_rate": 0.0004955966154377482,
|
| 3606 |
+
"loss": 0.9842,
|
| 3607 |
+
"step": 11707
|
| 3608 |
+
},
|
| 3609 |
+
{
|
| 3610 |
+
"epoch": 0.5058869193944883,
|
| 3611 |
+
"grad_norm": 0.443359375,
|
| 3612 |
+
"learning_rate": 0.0004946036953893974,
|
| 3613 |
+
"loss": 0.9927,
|
| 3614 |
+
"step": 11730
|
| 3615 |
+
},
|
| 3616 |
+
{
|
| 3617 |
+
"epoch": 0.5068788545305559,
|
| 3618 |
+
"grad_norm": 0.44921875,
|
| 3619 |
+
"learning_rate": 0.0004936107753410465,
|
| 3620 |
+
"loss": 0.9921,
|
| 3621 |
+
"step": 11753
|
| 3622 |
+
},
|
| 3623 |
+
{
|
| 3624 |
+
"epoch": 0.5078707896666236,
|
| 3625 |
+
"grad_norm": 0.40625,
|
| 3626 |
+
"learning_rate": 0.0004926178552926956,
|
| 3627 |
+
"loss": 0.9827,
|
| 3628 |
+
"step": 11776
|
| 3629 |
+
},
|
| 3630 |
+
{
|
| 3631 |
+
"epoch": 0.5088627248026911,
|
| 3632 |
+
"grad_norm": 0.416015625,
|
| 3633 |
+
"learning_rate": 0.0004916249352443447,
|
| 3634 |
+
"loss": 0.9836,
|
| 3635 |
+
"step": 11799
|
| 3636 |
+
},
|
| 3637 |
+
{
|
| 3638 |
+
"epoch": 0.5098546599387588,
|
| 3639 |
+
"grad_norm": 0.36328125,
|
| 3640 |
+
"learning_rate": 0.0004906320151959938,
|
| 3641 |
+
"loss": 0.9783,
|
| 3642 |
+
"step": 11822
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 0.5108465950748264,
|
| 3646 |
+
"grad_norm": 0.357421875,
|
| 3647 |
+
"learning_rate": 0.0004896390951476428,
|
| 3648 |
+
"loss": 1.0003,
|
| 3649 |
+
"step": 11845
|
| 3650 |
+
},
|
| 3651 |
+
{
|
| 3652 |
+
"epoch": 0.511838530210894,
|
| 3653 |
+
"grad_norm": 0.39453125,
|
| 3654 |
+
"learning_rate": 0.000488646175099292,
|
| 3655 |
+
"loss": 0.995,
|
| 3656 |
+
"step": 11868
|
| 3657 |
+
},
|
| 3658 |
+
{
|
| 3659 |
+
"epoch": 0.5128304653469616,
|
| 3660 |
+
"grad_norm": 0.376953125,
|
| 3661 |
+
"learning_rate": 0.0004876532550509411,
|
| 3662 |
+
"loss": 0.9952,
|
| 3663 |
+
"step": 11891
|
| 3664 |
+
},
|
| 3665 |
+
{
|
| 3666 |
+
"epoch": 0.5138224004830293,
|
| 3667 |
+
"grad_norm": 0.3828125,
|
| 3668 |
+
"learning_rate": 0.0004866603350025902,
|
| 3669 |
+
"loss": 0.9912,
|
| 3670 |
+
"step": 11914
|
| 3671 |
+
},
|
| 3672 |
+
{
|
| 3673 |
+
"epoch": 0.514814335619097,
|
| 3674 |
+
"grad_norm": 0.34375,
|
| 3675 |
+
"learning_rate": 0.00048566741495423933,
|
| 3676 |
+
"loss": 0.995,
|
| 3677 |
+
"step": 11937
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 0.5158062707551645,
|
| 3681 |
+
"grad_norm": 0.408203125,
|
| 3682 |
+
"learning_rate": 0.00048467449490588845,
|
| 3683 |
+
"loss": 0.9856,
|
| 3684 |
+
"step": 11960
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 0.5167982058912322,
|
| 3688 |
+
"grad_norm": 0.427734375,
|
| 3689 |
+
"learning_rate": 0.00048368157485753757,
|
| 3690 |
+
"loss": 0.9887,
|
| 3691 |
+
"step": 11983
|
| 3692 |
+
},
|
| 3693 |
+
{
|
| 3694 |
+
"epoch": 0.5177901410272998,
|
| 3695 |
+
"grad_norm": 0.376953125,
|
| 3696 |
+
"learning_rate": 0.0004826886548091867,
|
| 3697 |
+
"loss": 0.9815,
|
| 3698 |
+
"step": 12006
|
| 3699 |
+
},
|
| 3700 |
+
{
|
| 3701 |
+
"epoch": 0.5187820761633674,
|
| 3702 |
+
"grad_norm": 0.369140625,
|
| 3703 |
+
"learning_rate": 0.00048169573476083575,
|
| 3704 |
+
"loss": 0.9944,
|
| 3705 |
+
"step": 12029
|
| 3706 |
+
},
|
| 3707 |
+
{
|
| 3708 |
+
"epoch": 0.519774011299435,
|
| 3709 |
+
"grad_norm": 0.427734375,
|
| 3710 |
+
"learning_rate": 0.0004807028147124849,
|
| 3711 |
+
"loss": 0.9714,
|
| 3712 |
+
"step": 12052
|
| 3713 |
+
},
|
| 3714 |
+
{
|
| 3715 |
+
"epoch": 0.5207659464355027,
|
| 3716 |
+
"grad_norm": 0.326171875,
|
| 3717 |
+
"learning_rate": 0.000479709894664134,
|
| 3718 |
+
"loss": 0.9849,
|
| 3719 |
+
"step": 12075
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 0.5217578815715703,
|
| 3723 |
+
"grad_norm": 0.427734375,
|
| 3724 |
+
"learning_rate": 0.0004787169746157831,
|
| 3725 |
+
"loss": 0.9861,
|
| 3726 |
+
"step": 12098
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 0.5227498167076379,
|
| 3730 |
+
"grad_norm": 0.4765625,
|
| 3731 |
+
"learning_rate": 0.00047772405456743223,
|
| 3732 |
+
"loss": 1.0009,
|
| 3733 |
+
"step": 12121
|
| 3734 |
+
},
|
| 3735 |
+
{
|
| 3736 |
+
"epoch": 0.5237417518437055,
|
| 3737 |
+
"grad_norm": 0.345703125,
|
| 3738 |
+
"learning_rate": 0.00047673113451908135,
|
| 3739 |
+
"loss": 0.9892,
|
| 3740 |
+
"step": 12144
|
| 3741 |
+
},
|
| 3742 |
+
{
|
| 3743 |
+
"epoch": 0.5247336869797732,
|
| 3744 |
+
"grad_norm": 0.345703125,
|
| 3745 |
+
"learning_rate": 0.00047573821447073047,
|
| 3746 |
+
"loss": 0.9843,
|
| 3747 |
+
"step": 12167
|
| 3748 |
+
},
|
| 3749 |
+
{
|
| 3750 |
+
"epoch": 0.5257256221158407,
|
| 3751 |
+
"grad_norm": 0.40234375,
|
| 3752 |
+
"learning_rate": 0.0004747452944223796,
|
| 3753 |
+
"loss": 0.9767,
|
| 3754 |
+
"step": 12190
|
| 3755 |
+
},
|
| 3756 |
+
{
|
| 3757 |
+
"epoch": 0.5267175572519084,
|
| 3758 |
+
"grad_norm": 0.359375,
|
| 3759 |
+
"learning_rate": 0.00047375237437402866,
|
| 3760 |
+
"loss": 0.9599,
|
| 3761 |
+
"step": 12213
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 0.527709492387976,
|
| 3765 |
+
"grad_norm": 0.388671875,
|
| 3766 |
+
"learning_rate": 0.0004727594543256778,
|
| 3767 |
+
"loss": 0.9797,
|
| 3768 |
+
"step": 12236
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 0.5287014275240437,
|
| 3772 |
+
"grad_norm": 0.3359375,
|
| 3773 |
+
"learning_rate": 0.0004717665342773269,
|
| 3774 |
+
"loss": 0.9952,
|
| 3775 |
+
"step": 12259
|
| 3776 |
+
},
|
| 3777 |
+
{
|
| 3778 |
+
"epoch": 0.5296933626601112,
|
| 3779 |
+
"grad_norm": 0.359375,
|
| 3780 |
+
"learning_rate": 0.000470773614228976,
|
| 3781 |
+
"loss": 0.9851,
|
| 3782 |
+
"step": 12282
|
| 3783 |
+
},
|
| 3784 |
+
{
|
| 3785 |
+
"epoch": 0.5306852977961789,
|
| 3786 |
+
"grad_norm": 0.4140625,
|
| 3787 |
+
"learning_rate": 0.00046978069418062514,
|
| 3788 |
+
"loss": 0.9728,
|
| 3789 |
+
"step": 12305
|
| 3790 |
+
},
|
| 3791 |
+
{
|
| 3792 |
+
"epoch": 0.5316772329322466,
|
| 3793 |
+
"grad_norm": 0.376953125,
|
| 3794 |
+
"learning_rate": 0.00046878777413227426,
|
| 3795 |
+
"loss": 0.9813,
|
| 3796 |
+
"step": 12328
|
| 3797 |
+
},
|
| 3798 |
+
{
|
| 3799 |
+
"epoch": 0.5326691680683141,
|
| 3800 |
+
"grad_norm": 0.326171875,
|
| 3801 |
+
"learning_rate": 0.0004677948540839234,
|
| 3802 |
+
"loss": 0.9729,
|
| 3803 |
+
"step": 12351
|
| 3804 |
+
},
|
| 3805 |
+
{
|
| 3806 |
+
"epoch": 0.5336611032043818,
|
| 3807 |
+
"grad_norm": 0.33203125,
|
| 3808 |
+
"learning_rate": 0.0004668019340355725,
|
| 3809 |
+
"loss": 0.969,
|
| 3810 |
+
"step": 12374
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 0.5346530383404494,
|
| 3814 |
+
"grad_norm": 0.43359375,
|
| 3815 |
+
"learning_rate": 0.00046580901398722156,
|
| 3816 |
+
"loss": 0.9786,
|
| 3817 |
+
"step": 12397
|
| 3818 |
+
},
|
| 3819 |
+
{
|
| 3820 |
+
"epoch": 0.535644973476517,
|
| 3821 |
+
"grad_norm": 0.388671875,
|
| 3822 |
+
"learning_rate": 0.00046481609393887063,
|
| 3823 |
+
"loss": 0.9773,
|
| 3824 |
+
"step": 12420
|
| 3825 |
+
},
|
| 3826 |
+
{
|
| 3827 |
+
"epoch": 0.5366369086125846,
|
| 3828 |
+
"grad_norm": 0.451171875,
|
| 3829 |
+
"learning_rate": 0.00046382317389051975,
|
| 3830 |
+
"loss": 0.9972,
|
| 3831 |
+
"step": 12443
|
| 3832 |
+
},
|
| 3833 |
+
{
|
| 3834 |
+
"epoch": 0.5376288437486523,
|
| 3835 |
+
"grad_norm": 0.408203125,
|
| 3836 |
+
"learning_rate": 0.00046283025384216887,
|
| 3837 |
+
"loss": 0.9893,
|
| 3838 |
+
"step": 12466
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 0.5386207788847199,
|
| 3842 |
+
"grad_norm": 0.400390625,
|
| 3843 |
+
"learning_rate": 0.000461837333793818,
|
| 3844 |
+
"loss": 0.9747,
|
| 3845 |
+
"step": 12489
|
| 3846 |
+
},
|
| 3847 |
+
{
|
| 3848 |
+
"epoch": 0.5396127140207875,
|
| 3849 |
+
"grad_norm": 0.4921875,
|
| 3850 |
+
"learning_rate": 0.0004608444137454671,
|
| 3851 |
+
"loss": 0.9795,
|
| 3852 |
+
"step": 12512
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 0.5406046491568551,
|
| 3856 |
+
"grad_norm": 0.37109375,
|
| 3857 |
+
"learning_rate": 0.00045985149369711623,
|
| 3858 |
+
"loss": 0.9608,
|
| 3859 |
+
"step": 12535
|
| 3860 |
+
},
|
| 3861 |
+
{
|
| 3862 |
+
"epoch": 0.5415965842929228,
|
| 3863 |
+
"grad_norm": 0.3515625,
|
| 3864 |
+
"learning_rate": 0.00045885857364876535,
|
| 3865 |
+
"loss": 0.966,
|
| 3866 |
+
"step": 12558
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 0.5425885194289903,
|
| 3870 |
+
"grad_norm": 0.361328125,
|
| 3871 |
+
"learning_rate": 0.0004578656536004144,
|
| 3872 |
+
"loss": 0.9689,
|
| 3873 |
+
"step": 12581
|
| 3874 |
+
},
|
| 3875 |
+
{
|
| 3876 |
+
"epoch": 0.543580454565058,
|
| 3877 |
+
"grad_norm": 0.404296875,
|
| 3878 |
+
"learning_rate": 0.00045687273355206353,
|
| 3879 |
+
"loss": 0.9655,
|
| 3880 |
+
"step": 12604
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 0.5445723897011256,
|
| 3884 |
+
"grad_norm": 0.37890625,
|
| 3885 |
+
"learning_rate": 0.00045587981350371265,
|
| 3886 |
+
"loss": 0.9693,
|
| 3887 |
+
"step": 12627
|
| 3888 |
+
},
|
| 3889 |
+
{
|
| 3890 |
+
"epoch": 0.5455643248371933,
|
| 3891 |
+
"grad_norm": 0.41015625,
|
| 3892 |
+
"learning_rate": 0.00045488689345536177,
|
| 3893 |
+
"loss": 0.9986,
|
| 3894 |
+
"step": 12650
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 0.5465562599732608,
|
| 3898 |
+
"grad_norm": 0.345703125,
|
| 3899 |
+
"learning_rate": 0.0004538939734070109,
|
| 3900 |
+
"loss": 0.9715,
|
| 3901 |
+
"step": 12673
|
| 3902 |
+
},
|
| 3903 |
+
{
|
| 3904 |
+
"epoch": 0.5475481951093285,
|
| 3905 |
+
"grad_norm": 0.37890625,
|
| 3906 |
+
"learning_rate": 0.00045290105335866,
|
| 3907 |
+
"loss": 0.9781,
|
| 3908 |
+
"step": 12696
|
| 3909 |
+
},
|
| 3910 |
+
{
|
| 3911 |
+
"epoch": 0.5485401302453962,
|
| 3912 |
+
"grad_norm": 0.42578125,
|
| 3913 |
+
"learning_rate": 0.00045190813331030913,
|
| 3914 |
+
"loss": 1.0001,
|
| 3915 |
+
"step": 12719
|
| 3916 |
+
},
|
| 3917 |
+
{
|
| 3918 |
+
"epoch": 0.5495320653814637,
|
| 3919 |
+
"grad_norm": 0.43359375,
|
| 3920 |
+
"learning_rate": 0.0004509152132619582,
|
| 3921 |
+
"loss": 0.9811,
|
| 3922 |
+
"step": 12742
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 0.5505240005175314,
|
| 3926 |
+
"grad_norm": 0.341796875,
|
| 3927 |
+
"learning_rate": 0.0004499222932136073,
|
| 3928 |
+
"loss": 0.9584,
|
| 3929 |
+
"step": 12765
|
| 3930 |
+
},
|
| 3931 |
+
{
|
| 3932 |
+
"epoch": 0.551515935653599,
|
| 3933 |
+
"grad_norm": 0.419921875,
|
| 3934 |
+
"learning_rate": 0.00044892937316525644,
|
| 3935 |
+
"loss": 0.977,
|
| 3936 |
+
"step": 12788
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 0.5525078707896667,
|
| 3940 |
+
"grad_norm": 0.416015625,
|
| 3941 |
+
"learning_rate": 0.00044793645311690556,
|
| 3942 |
+
"loss": 0.9746,
|
| 3943 |
+
"step": 12811
|
| 3944 |
+
},
|
| 3945 |
+
{
|
| 3946 |
+
"epoch": 0.5534998059257342,
|
| 3947 |
+
"grad_norm": 0.390625,
|
| 3948 |
+
"learning_rate": 0.0004469435330685547,
|
| 3949 |
+
"loss": 0.9811,
|
| 3950 |
+
"step": 12834
|
| 3951 |
+
},
|
| 3952 |
+
{
|
| 3953 |
+
"epoch": 0.5544917410618019,
|
| 3954 |
+
"grad_norm": 0.35546875,
|
| 3955 |
+
"learning_rate": 0.0004459506130202038,
|
| 3956 |
+
"loss": 0.9523,
|
| 3957 |
+
"step": 12857
|
| 3958 |
+
},
|
| 3959 |
+
{
|
| 3960 |
+
"epoch": 0.5554836761978695,
|
| 3961 |
+
"grad_norm": 0.37890625,
|
| 3962 |
+
"learning_rate": 0.0004449576929718529,
|
| 3963 |
+
"loss": 0.9641,
|
| 3964 |
+
"step": 12880
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 0.5564756113339371,
|
| 3968 |
+
"grad_norm": 0.36328125,
|
| 3969 |
+
"learning_rate": 0.00044396477292350204,
|
| 3970 |
+
"loss": 0.9845,
|
| 3971 |
+
"step": 12903
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 0.5574675464700047,
|
| 3975 |
+
"grad_norm": 0.365234375,
|
| 3976 |
+
"learning_rate": 0.0004429718528751511,
|
| 3977 |
+
"loss": 0.9788,
|
| 3978 |
+
"step": 12926
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 0.5584594816060724,
|
| 3982 |
+
"grad_norm": 0.390625,
|
| 3983 |
+
"learning_rate": 0.0004419789328268002,
|
| 3984 |
+
"loss": 0.9795,
|
| 3985 |
+
"step": 12949
|
| 3986 |
+
},
|
| 3987 |
+
{
|
| 3988 |
+
"epoch": 0.5594514167421399,
|
| 3989 |
+
"grad_norm": 0.37109375,
|
| 3990 |
+
"learning_rate": 0.00044098601277844934,
|
| 3991 |
+
"loss": 0.9716,
|
| 3992 |
+
"step": 12972
|
| 3993 |
+
},
|
| 3994 |
+
{
|
| 3995 |
+
"epoch": 0.5604433518782076,
|
| 3996 |
+
"grad_norm": 0.38671875,
|
| 3997 |
+
"learning_rate": 0.00043999309273009846,
|
| 3998 |
+
"loss": 0.9814,
|
| 3999 |
+
"step": 12995
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 0.5614352870142753,
|
| 4003 |
+
"grad_norm": 0.34765625,
|
| 4004 |
+
"learning_rate": 0.00043900017268174753,
|
| 4005 |
+
"loss": 0.9724,
|
| 4006 |
+
"step": 13018
|
| 4007 |
+
},
|
| 4008 |
+
{
|
| 4009 |
+
"epoch": 0.5624272221503429,
|
| 4010 |
+
"grad_norm": 0.44921875,
|
| 4011 |
+
"learning_rate": 0.00043800725263339665,
|
| 4012 |
+
"loss": 0.9538,
|
| 4013 |
+
"step": 13041
|
| 4014 |
+
},
|
| 4015 |
+
{
|
| 4016 |
+
"epoch": 0.5634191572864105,
|
| 4017 |
+
"grad_norm": 0.3828125,
|
| 4018 |
+
"learning_rate": 0.00043701433258504577,
|
| 4019 |
+
"loss": 0.9744,
|
| 4020 |
+
"step": 13064
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 0.5644110924224781,
|
| 4024 |
+
"grad_norm": 0.423828125,
|
| 4025 |
+
"learning_rate": 0.0004360214125366949,
|
| 4026 |
+
"loss": 0.9777,
|
| 4027 |
+
"step": 13087
|
| 4028 |
+
},
|
| 4029 |
+
{
|
| 4030 |
+
"epoch": 0.5654030275585458,
|
| 4031 |
+
"grad_norm": 0.365234375,
|
| 4032 |
+
"learning_rate": 0.00043502849248834395,
|
| 4033 |
+
"loss": 0.9688,
|
| 4034 |
+
"step": 13110
|
| 4035 |
+
},
|
| 4036 |
+
{
|
| 4037 |
+
"epoch": 0.5663949626946133,
|
| 4038 |
+
"grad_norm": 0.470703125,
|
| 4039 |
+
"learning_rate": 0.00043403557243999307,
|
| 4040 |
+
"loss": 0.988,
|
| 4041 |
+
"step": 13133
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 0.567386897830681,
|
| 4045 |
+
"grad_norm": 0.341796875,
|
| 4046 |
+
"learning_rate": 0.0004330426523916422,
|
| 4047 |
+
"loss": 0.9678,
|
| 4048 |
+
"step": 13156
|
| 4049 |
+
},
|
| 4050 |
+
{
|
| 4051 |
+
"epoch": 0.5683788329667486,
|
| 4052 |
+
"grad_norm": 0.345703125,
|
| 4053 |
+
"learning_rate": 0.0004320497323432913,
|
| 4054 |
+
"loss": 0.9735,
|
| 4055 |
+
"step": 13179
|
| 4056 |
+
},
|
| 4057 |
+
{
|
| 4058 |
+
"epoch": 0.5693707681028163,
|
| 4059 |
+
"grad_norm": 0.416015625,
|
| 4060 |
+
"learning_rate": 0.00043105681229494043,
|
| 4061 |
+
"loss": 0.9612,
|
| 4062 |
+
"step": 13202
|
| 4063 |
+
},
|
| 4064 |
+
{
|
| 4065 |
+
"epoch": 0.5703627032388838,
|
| 4066 |
+
"grad_norm": 0.375,
|
| 4067 |
+
"learning_rate": 0.00043006389224658955,
|
| 4068 |
+
"loss": 0.9428,
|
| 4069 |
+
"step": 13225
|
| 4070 |
+
},
|
| 4071 |
+
{
|
| 4072 |
+
"epoch": 0.5713546383749515,
|
| 4073 |
+
"grad_norm": 0.4296875,
|
| 4074 |
+
"learning_rate": 0.00042907097219823867,
|
| 4075 |
+
"loss": 0.9654,
|
| 4076 |
+
"step": 13248
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 0.5723465735110191,
|
| 4080 |
+
"grad_norm": 0.353515625,
|
| 4081 |
+
"learning_rate": 0.0004280780521498878,
|
| 4082 |
+
"loss": 0.9739,
|
| 4083 |
+
"step": 13271
|
| 4084 |
+
},
|
| 4085 |
+
{
|
| 4086 |
+
"epoch": 0.5733385086470867,
|
| 4087 |
+
"grad_norm": 0.380859375,
|
| 4088 |
+
"learning_rate": 0.00042708513210153686,
|
| 4089 |
+
"loss": 0.9755,
|
| 4090 |
+
"step": 13294
|
| 4091 |
+
},
|
| 4092 |
+
{
|
| 4093 |
+
"epoch": 0.5743304437831543,
|
| 4094 |
+
"grad_norm": 0.357421875,
|
| 4095 |
+
"learning_rate": 0.000426092212053186,
|
| 4096 |
+
"loss": 0.9784,
|
| 4097 |
+
"step": 13317
|
| 4098 |
+
},
|
| 4099 |
+
{
|
| 4100 |
+
"epoch": 0.575322378919222,
|
| 4101 |
+
"grad_norm": 0.3125,
|
| 4102 |
+
"learning_rate": 0.0004250992920048351,
|
| 4103 |
+
"loss": 0.9625,
|
| 4104 |
+
"step": 13340
|
| 4105 |
+
},
|
| 4106 |
+
{
|
| 4107 |
+
"epoch": 0.5763143140552897,
|
| 4108 |
+
"grad_norm": 0.345703125,
|
| 4109 |
+
"learning_rate": 0.0004241063719564842,
|
| 4110 |
+
"loss": 0.9521,
|
| 4111 |
+
"step": 13363
|
| 4112 |
+
},
|
| 4113 |
+
{
|
| 4114 |
+
"epoch": 0.5773062491913572,
|
| 4115 |
+
"grad_norm": 0.333984375,
|
| 4116 |
+
"learning_rate": 0.00042311345190813334,
|
| 4117 |
+
"loss": 0.984,
|
| 4118 |
+
"step": 13386
|
| 4119 |
+
},
|
| 4120 |
+
{
|
| 4121 |
+
"epoch": 0.5782981843274249,
|
| 4122 |
+
"grad_norm": 0.45703125,
|
| 4123 |
+
"learning_rate": 0.00042212053185978246,
|
| 4124 |
+
"loss": 0.9794,
|
| 4125 |
+
"step": 13409
|
| 4126 |
+
},
|
| 4127 |
+
{
|
| 4128 |
+
"epoch": 0.5792901194634925,
|
| 4129 |
+
"grad_norm": 0.396484375,
|
| 4130 |
+
"learning_rate": 0.0004211276118114316,
|
| 4131 |
+
"loss": 0.9705,
|
| 4132 |
+
"step": 13432
|
| 4133 |
+
},
|
| 4134 |
+
{
|
| 4135 |
+
"epoch": 0.5802820545995601,
|
| 4136 |
+
"grad_norm": 0.400390625,
|
| 4137 |
+
"learning_rate": 0.00042013469176308064,
|
| 4138 |
+
"loss": 0.97,
|
| 4139 |
+
"step": 13455
|
| 4140 |
+
},
|
| 4141 |
+
{
|
| 4142 |
+
"epoch": 0.5812739897356277,
|
| 4143 |
+
"grad_norm": 0.37890625,
|
| 4144 |
+
"learning_rate": 0.00041914177171472976,
|
| 4145 |
+
"loss": 0.968,
|
| 4146 |
+
"step": 13478
|
| 4147 |
+
},
|
| 4148 |
+
{
|
| 4149 |
+
"epoch": 0.5822659248716954,
|
| 4150 |
+
"grad_norm": 0.365234375,
|
| 4151 |
+
"learning_rate": 0.0004181488516663789,
|
| 4152 |
+
"loss": 0.9664,
|
| 4153 |
+
"step": 13501
|
| 4154 |
+
},
|
| 4155 |
+
{
|
| 4156 |
+
"epoch": 0.5832578600077629,
|
| 4157 |
+
"grad_norm": 0.361328125,
|
| 4158 |
+
"learning_rate": 0.000417155931618028,
|
| 4159 |
+
"loss": 0.9722,
|
| 4160 |
+
"step": 13524
|
| 4161 |
+
},
|
| 4162 |
+
{
|
| 4163 |
+
"epoch": 0.5842497951438306,
|
| 4164 |
+
"grad_norm": 0.369140625,
|
| 4165 |
+
"learning_rate": 0.0004161630115696771,
|
| 4166 |
+
"loss": 0.9695,
|
| 4167 |
+
"step": 13547
|
| 4168 |
+
},
|
| 4169 |
+
{
|
| 4170 |
+
"epoch": 0.5852417302798982,
|
| 4171 |
+
"grad_norm": 0.337890625,
|
| 4172 |
+
"learning_rate": 0.00041517009152132624,
|
| 4173 |
+
"loss": 0.9628,
|
| 4174 |
+
"step": 13570
|
| 4175 |
+
},
|
| 4176 |
+
{
|
| 4177 |
+
"epoch": 0.5862336654159659,
|
| 4178 |
+
"grad_norm": 0.330078125,
|
| 4179 |
+
"learning_rate": 0.0004141771714729753,
|
| 4180 |
+
"loss": 0.9515,
|
| 4181 |
+
"step": 13593
|
| 4182 |
+
},
|
| 4183 |
+
{
|
| 4184 |
+
"epoch": 0.5872256005520334,
|
| 4185 |
+
"grad_norm": 0.359375,
|
| 4186 |
+
"learning_rate": 0.0004131842514246244,
|
| 4187 |
+
"loss": 0.965,
|
| 4188 |
+
"step": 13616
|
| 4189 |
+
},
|
| 4190 |
+
{
|
| 4191 |
+
"epoch": 0.5882175356881011,
|
| 4192 |
+
"grad_norm": 0.392578125,
|
| 4193 |
+
"learning_rate": 0.0004121913313762735,
|
| 4194 |
+
"loss": 0.9598,
|
| 4195 |
+
"step": 13639
|
| 4196 |
+
},
|
| 4197 |
+
{
|
| 4198 |
+
"epoch": 0.5892094708241687,
|
| 4199 |
+
"grad_norm": 0.41796875,
|
| 4200 |
+
"learning_rate": 0.0004111984113279226,
|
| 4201 |
+
"loss": 0.9575,
|
| 4202 |
+
"step": 13662
|
| 4203 |
+
},
|
| 4204 |
+
{
|
| 4205 |
+
"epoch": 0.5902014059602363,
|
| 4206 |
+
"grad_norm": 0.5234375,
|
| 4207 |
+
"learning_rate": 0.00041020549127957173,
|
| 4208 |
+
"loss": 0.9933,
|
| 4209 |
+
"step": 13685
|
| 4210 |
+
},
|
| 4211 |
+
{
|
| 4212 |
+
"epoch": 0.591193341096304,
|
| 4213 |
+
"grad_norm": 0.423828125,
|
| 4214 |
+
"learning_rate": 0.00040921257123122085,
|
| 4215 |
+
"loss": 0.9621,
|
| 4216 |
+
"step": 13708
|
| 4217 |
+
},
|
| 4218 |
+
{
|
| 4219 |
+
"epoch": 0.5921852762323716,
|
| 4220 |
+
"grad_norm": 0.33203125,
|
| 4221 |
+
"learning_rate": 0.00040821965118286997,
|
| 4222 |
+
"loss": 0.964,
|
| 4223 |
+
"step": 13731
|
| 4224 |
+
},
|
| 4225 |
+
{
|
| 4226 |
+
"epoch": 0.5931772113684393,
|
| 4227 |
+
"grad_norm": 0.423828125,
|
| 4228 |
+
"learning_rate": 0.0004072267311345191,
|
| 4229 |
+
"loss": 0.9854,
|
| 4230 |
+
"step": 13754
|
| 4231 |
+
},
|
| 4232 |
+
{
|
| 4233 |
+
"epoch": 0.5941691465045068,
|
| 4234 |
+
"grad_norm": 0.3515625,
|
| 4235 |
+
"learning_rate": 0.0004062338110861682,
|
| 4236 |
+
"loss": 0.9883,
|
| 4237 |
+
"step": 13777
|
| 4238 |
+
},
|
| 4239 |
+
{
|
| 4240 |
+
"epoch": 0.5951610816405745,
|
| 4241 |
+
"grad_norm": 0.408203125,
|
| 4242 |
+
"learning_rate": 0.00040524089103781733,
|
| 4243 |
+
"loss": 0.9853,
|
| 4244 |
+
"step": 13800
|
| 4245 |
+
},
|
| 4246 |
+
{
|
| 4247 |
+
"epoch": 0.5961530167766421,
|
| 4248 |
+
"grad_norm": 0.408203125,
|
| 4249 |
+
"learning_rate": 0.0004042479709894664,
|
| 4250 |
+
"loss": 0.9557,
|
| 4251 |
+
"step": 13823
|
| 4252 |
+
},
|
| 4253 |
+
{
|
| 4254 |
+
"epoch": 0.5971449519127097,
|
| 4255 |
+
"grad_norm": 0.42578125,
|
| 4256 |
+
"learning_rate": 0.0004032550509411155,
|
| 4257 |
+
"loss": 0.9587,
|
| 4258 |
+
"step": 13846
|
| 4259 |
+
},
|
| 4260 |
+
{
|
| 4261 |
+
"epoch": 0.5981368870487773,
|
| 4262 |
+
"grad_norm": 0.44921875,
|
| 4263 |
+
"learning_rate": 0.00040226213089276464,
|
| 4264 |
+
"loss": 0.9771,
|
| 4265 |
+
"step": 13869
|
| 4266 |
+
},
|
| 4267 |
+
{
|
| 4268 |
+
"epoch": 0.599128822184845,
|
| 4269 |
+
"grad_norm": 0.431640625,
|
| 4270 |
+
"learning_rate": 0.00040126921084441376,
|
| 4271 |
+
"loss": 0.9661,
|
| 4272 |
+
"step": 13892
|
| 4273 |
+
},
|
| 4274 |
+
{
|
| 4275 |
+
"epoch": 0.5998188640186312,
|
| 4276 |
+
"eval_runtime": 163.7921,
|
| 4277 |
+
"eval_samples_per_second": 610.53,
|
| 4278 |
+
"eval_steps_per_second": 7.632,
|
| 4279 |
+
"step": 13908
|
| 4280 |
}
|
| 4281 |
],
|
| 4282 |
"logging_steps": 23,
|
|
|
|
| 4296 |
"attributes": {}
|
| 4297 |
}
|
| 4298 |
},
|
| 4299 |
+
"total_flos": 1.0167159364234772e+18,
|
| 4300 |
"train_batch_size": 8,
|
| 4301 |
"trial_name": null,
|
| 4302 |
"trial_params": null
|