Training in progress, epoch 1, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/optimizer.pt +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/rng_state_4.pth +1 -1
- last-checkpoint/rng_state_5.pth +1 -1
- last-checkpoint/rng_state_6.pth +1 -1
- last-checkpoint/rng_state_7.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/tokenizer.model +3 -0
- last-checkpoint/trainer_state.json +956 -4
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 242273720
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6ff2191251bdc5ec70dd110ae634b13ebf4810e1bca6c85177e8a00270e99df
|
| 3 |
size 242273720
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 123464052
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:247b185c94b121cfad45276c33bc430abee09ddf71df2950d9535045e5736c7b
|
| 3 |
size 123464052
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:effd1c28ebf0d99d4f55c0db7dbc3f8310000e3e983a8abbdfc87f12f555cd10
|
| 3 |
size 15984
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f67c5ae92a13c1dfff98fe2ed3c24293bba80d0fbfd82b323e01b99d5047705
|
| 3 |
size 15984
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2e824cc8f44e0c3bdb90370120243f177219f1b2337606f172a19235eb54fa3a
|
| 3 |
size 15984
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7f383805aa6de04f9b0524a89b48769947b5028ac7f403c5f175e44d3966d979
|
| 3 |
size 15984
|
last-checkpoint/rng_state_4.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:13d25717bd4cc4e246680a6efed9c3ff28ae6dea06ff25b07f0ca41bc7afd64c
|
| 3 |
size 15984
|
last-checkpoint/rng_state_5.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:636a4df001af52276924ce01200b1487c01c67a6fd7d60811f3d21bbd7d5e6a1
|
| 3 |
size 15984
|
last-checkpoint/rng_state_6.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f716607386a91dc4a5317c0077dfdf351c6712bb99b4353fd837f97b6a19c612
|
| 3 |
size 15984
|
last-checkpoint/rng_state_7.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 15984
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:25797d2b598ccc5302513242855c7194b2e66ccfafbd9173713f63204e6f5e01
|
| 3 |
size 15984
|
last-checkpoint/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 sha256:5f4fe8e328be6e10e053e14666b2e571c45c73d9a8291556b08910e3da67b3e6
|
| 3 |
size 1064
|
last-checkpoint/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:583f2ebd2a1936009b7da991ea255504db68c7a9713a78673d1335a87098966c
|
| 3 |
+
size 4241023
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2563,6 +2563,958 @@
|
|
| 2563 |
"eval_samples_per_second": 38.39,
|
| 2564 |
"eval_steps_per_second": 2.405,
|
| 2565 |
"step": 364
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2566 |
}
|
| 2567 |
],
|
| 2568 |
"logging_steps": 1,
|
|
@@ -2577,12 +3529,12 @@
|
|
| 2577 |
"should_evaluate": false,
|
| 2578 |
"should_log": false,
|
| 2579 |
"should_save": true,
|
| 2580 |
-
"should_training_stop":
|
| 2581 |
},
|
| 2582 |
"attributes": {}
|
| 2583 |
}
|
| 2584 |
},
|
| 2585 |
-
"total_flos": 1.
|
| 2586 |
"train_batch_size": 2,
|
| 2587 |
"trial_name": null,
|
| 2588 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.3714873200822482,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2563 |
"eval_samples_per_second": 38.39,
|
| 2564 |
"eval_steps_per_second": 2.405,
|
| 2565 |
"step": 364
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 1.0013708019191228,
|
| 2569 |
+
"grad_norm": 0.27995359897613525,
|
| 2570 |
+
"learning_rate": 5.113787585030454e-05,
|
| 2571 |
+
"loss": 0.9121,
|
| 2572 |
+
"step": 365
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 1.004112405757368,
|
| 2576 |
+
"grad_norm": 0.17540809512138367,
|
| 2577 |
+
"learning_rate": 5.045413315675924e-05,
|
| 2578 |
+
"loss": 0.6274,
|
| 2579 |
+
"step": 366
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 1.0068540095956133,
|
| 2583 |
+
"grad_norm": 0.17872898280620575,
|
| 2584 |
+
"learning_rate": 4.977344668566275e-05,
|
| 2585 |
+
"loss": 0.6557,
|
| 2586 |
+
"step": 367
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 1.0095956134338588,
|
| 2590 |
+
"grad_norm": 0.19064383208751678,
|
| 2591 |
+
"learning_rate": 4.909585842496287e-05,
|
| 2592 |
+
"loss": 0.6214,
|
| 2593 |
+
"step": 368
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 1.0123372172721041,
|
| 2597 |
+
"grad_norm": 0.18582047522068024,
|
| 2598 |
+
"learning_rate": 4.842141017149526e-05,
|
| 2599 |
+
"loss": 0.6319,
|
| 2600 |
+
"step": 369
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 1.0150788211103496,
|
| 2604 |
+
"grad_norm": 0.192233607172966,
|
| 2605 |
+
"learning_rate": 4.7750143528405126e-05,
|
| 2606 |
+
"loss": 0.6574,
|
| 2607 |
+
"step": 370
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 1.017820424948595,
|
| 2611 |
+
"grad_norm": 0.1876421719789505,
|
| 2612 |
+
"learning_rate": 4.708209990258095e-05,
|
| 2613 |
+
"loss": 0.6348,
|
| 2614 |
+
"step": 371
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 1.0205620287868402,
|
| 2618 |
+
"grad_norm": 0.18002265691757202,
|
| 2619 |
+
"learning_rate": 4.6417320502100316e-05,
|
| 2620 |
+
"loss": 0.6705,
|
| 2621 |
+
"step": 372
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 1.0233036326250857,
|
| 2625 |
+
"grad_norm": 0.1792839914560318,
|
| 2626 |
+
"learning_rate": 4.575584633368815e-05,
|
| 2627 |
+
"loss": 0.6675,
|
| 2628 |
+
"step": 373
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 1.026045236463331,
|
| 2632 |
+
"grad_norm": 0.18165361881256104,
|
| 2633 |
+
"learning_rate": 4.5097718200186814e-05,
|
| 2634 |
+
"loss": 0.626,
|
| 2635 |
+
"step": 374
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 1.0287868403015765,
|
| 2639 |
+
"grad_norm": 0.18540449440479279,
|
| 2640 |
+
"learning_rate": 4.444297669803981e-05,
|
| 2641 |
+
"loss": 0.6514,
|
| 2642 |
+
"step": 375
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 1.0315284441398218,
|
| 2646 |
+
"grad_norm": 0.1831628531217575,
|
| 2647 |
+
"learning_rate": 4.379166221478697e-05,
|
| 2648 |
+
"loss": 0.5291,
|
| 2649 |
+
"step": 376
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 1.034270047978067,
|
| 2653 |
+
"grad_norm": 0.20461496710777283,
|
| 2654 |
+
"learning_rate": 4.31438149265736e-05,
|
| 2655 |
+
"loss": 0.6833,
|
| 2656 |
+
"step": 377
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 1.0370116518163126,
|
| 2660 |
+
"grad_norm": 0.18456245958805084,
|
| 2661 |
+
"learning_rate": 4.249947479567218e-05,
|
| 2662 |
+
"loss": 0.6176,
|
| 2663 |
+
"step": 378
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 1.039753255654558,
|
| 2667 |
+
"grad_norm": 0.1796509176492691,
|
| 2668 |
+
"learning_rate": 4.185868156801694e-05,
|
| 2669 |
+
"loss": 0.5667,
|
| 2670 |
+
"step": 379
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 1.0424948594928032,
|
| 2674 |
+
"grad_norm": 0.19982261955738068,
|
| 2675 |
+
"learning_rate": 4.12214747707527e-05,
|
| 2676 |
+
"loss": 0.7026,
|
| 2677 |
+
"step": 380
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 1.0452364633310487,
|
| 2681 |
+
"grad_norm": 0.1934226006269455,
|
| 2682 |
+
"learning_rate": 4.058789370979615e-05,
|
| 2683 |
+
"loss": 0.7126,
|
| 2684 |
+
"step": 381
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 1.047978067169294,
|
| 2688 |
+
"grad_norm": 0.22090677917003632,
|
| 2689 |
+
"learning_rate": 3.9957977467411615e-05,
|
| 2690 |
+
"loss": 0.7282,
|
| 2691 |
+
"step": 382
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 1.0507196710075395,
|
| 2695 |
+
"grad_norm": 0.1960362195968628,
|
| 2696 |
+
"learning_rate": 3.933176489980005e-05,
|
| 2697 |
+
"loss": 0.5993,
|
| 2698 |
+
"step": 383
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 1.0534612748457848,
|
| 2702 |
+
"grad_norm": 0.1959688514471054,
|
| 2703 |
+
"learning_rate": 3.8709294634702376e-05,
|
| 2704 |
+
"loss": 0.6483,
|
| 2705 |
+
"step": 384
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 1.05620287868403,
|
| 2709 |
+
"grad_norm": 0.19136562943458557,
|
| 2710 |
+
"learning_rate": 3.8090605069016595e-05,
|
| 2711 |
+
"loss": 0.6883,
|
| 2712 |
+
"step": 385
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 1.0589444825222756,
|
| 2716 |
+
"grad_norm": 0.202112078666687,
|
| 2717 |
+
"learning_rate": 3.747573436642951e-05,
|
| 2718 |
+
"loss": 0.6789,
|
| 2719 |
+
"step": 386
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 1.0616860863605209,
|
| 2723 |
+
"grad_norm": 0.20548656582832336,
|
| 2724 |
+
"learning_rate": 3.686472045506223e-05,
|
| 2725 |
+
"loss": 0.585,
|
| 2726 |
+
"step": 387
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 1.0644276901987664,
|
| 2730 |
+
"grad_norm": 0.21226055920124054,
|
| 2731 |
+
"learning_rate": 3.6257601025131026e-05,
|
| 2732 |
+
"loss": 0.6373,
|
| 2733 |
+
"step": 388
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 1.0671692940370117,
|
| 2737 |
+
"grad_norm": 0.20208019018173218,
|
| 2738 |
+
"learning_rate": 3.565441352662211e-05,
|
| 2739 |
+
"loss": 0.5311,
|
| 2740 |
+
"step": 389
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 1.069910897875257,
|
| 2744 |
+
"grad_norm": 0.24004848301410675,
|
| 2745 |
+
"learning_rate": 3.5055195166981645e-05,
|
| 2746 |
+
"loss": 0.6887,
|
| 2747 |
+
"step": 390
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 1.0726525017135025,
|
| 2751 |
+
"grad_norm": 0.22098727524280548,
|
| 2752 |
+
"learning_rate": 3.445998290882062e-05,
|
| 2753 |
+
"loss": 0.6883,
|
| 2754 |
+
"step": 391
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 1.0753941055517477,
|
| 2758 |
+
"grad_norm": 0.22314122319221497,
|
| 2759 |
+
"learning_rate": 3.386881346763483e-05,
|
| 2760 |
+
"loss": 0.6937,
|
| 2761 |
+
"step": 392
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 1.078135709389993,
|
| 2765 |
+
"grad_norm": 0.19696767628192902,
|
| 2766 |
+
"learning_rate": 3.328172330954001e-05,
|
| 2767 |
+
"loss": 0.4911,
|
| 2768 |
+
"step": 393
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 1.0808773132282385,
|
| 2772 |
+
"grad_norm": 0.23884212970733643,
|
| 2773 |
+
"learning_rate": 3.269874864902269e-05,
|
| 2774 |
+
"loss": 0.6735,
|
| 2775 |
+
"step": 394
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 1.0836189170664838,
|
| 2779 |
+
"grad_norm": 0.2059262990951538,
|
| 2780 |
+
"learning_rate": 3.211992544670582e-05,
|
| 2781 |
+
"loss": 0.6691,
|
| 2782 |
+
"step": 395
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 1.0863605209047293,
|
| 2786 |
+
"grad_norm": 0.20793655514717102,
|
| 2787 |
+
"learning_rate": 3.154528940713113e-05,
|
| 2788 |
+
"loss": 0.5767,
|
| 2789 |
+
"step": 396
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 1.0891021247429746,
|
| 2793 |
+
"grad_norm": 0.21672658622264862,
|
| 2794 |
+
"learning_rate": 3.0974875976556284e-05,
|
| 2795 |
+
"loss": 0.6456,
|
| 2796 |
+
"step": 397
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 1.09184372858122,
|
| 2800 |
+
"grad_norm": 0.2148715704679489,
|
| 2801 |
+
"learning_rate": 3.0408720340768572e-05,
|
| 2802 |
+
"loss": 0.6234,
|
| 2803 |
+
"step": 398
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 1.0945853324194654,
|
| 2807 |
+
"grad_norm": 0.2160511016845703,
|
| 2808 |
+
"learning_rate": 2.9846857422914433e-05,
|
| 2809 |
+
"loss": 0.6743,
|
| 2810 |
+
"step": 399
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 1.0973269362577107,
|
| 2814 |
+
"grad_norm": 0.20896896719932556,
|
| 2815 |
+
"learning_rate": 2.9289321881345254e-05,
|
| 2816 |
+
"loss": 0.5162,
|
| 2817 |
+
"step": 400
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 1.1000685400959562,
|
| 2821 |
+
"grad_norm": 0.22997024655342102,
|
| 2822 |
+
"learning_rate": 2.8736148107479467e-05,
|
| 2823 |
+
"loss": 0.5936,
|
| 2824 |
+
"step": 401
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 1.1028101439342015,
|
| 2828 |
+
"grad_norm": 0.21282993257045746,
|
| 2829 |
+
"learning_rate": 2.8187370223681132e-05,
|
| 2830 |
+
"loss": 0.5867,
|
| 2831 |
+
"step": 402
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 1.1055517477724468,
|
| 2835 |
+
"grad_norm": 0.2123803347349167,
|
| 2836 |
+
"learning_rate": 2.764302208115508e-05,
|
| 2837 |
+
"loss": 0.6076,
|
| 2838 |
+
"step": 403
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 1.1082933516106923,
|
| 2842 |
+
"grad_norm": 0.1950678676366806,
|
| 2843 |
+
"learning_rate": 2.7103137257858868e-05,
|
| 2844 |
+
"loss": 0.614,
|
| 2845 |
+
"step": 404
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 1.1110349554489376,
|
| 2849 |
+
"grad_norm": 0.19433154165744781,
|
| 2850 |
+
"learning_rate": 2.6567749056431467e-05,
|
| 2851 |
+
"loss": 0.6088,
|
| 2852 |
+
"step": 405
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 1.1137765592871829,
|
| 2856 |
+
"grad_norm": 0.2161804586648941,
|
| 2857 |
+
"learning_rate": 2.603689050213902e-05,
|
| 2858 |
+
"loss": 0.598,
|
| 2859 |
+
"step": 406
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 1.1165181631254284,
|
| 2863 |
+
"grad_norm": 0.1916419416666031,
|
| 2864 |
+
"learning_rate": 2.5510594340837824e-05,
|
| 2865 |
+
"loss": 0.4941,
|
| 2866 |
+
"step": 407
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 1.1192597669636737,
|
| 2870 |
+
"grad_norm": 0.20069898664951324,
|
| 2871 |
+
"learning_rate": 2.4988893036954043e-05,
|
| 2872 |
+
"loss": 0.5563,
|
| 2873 |
+
"step": 408
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 1.1220013708019192,
|
| 2877 |
+
"grad_norm": 0.21479913592338562,
|
| 2878 |
+
"learning_rate": 2.4471818771481648e-05,
|
| 2879 |
+
"loss": 0.6706,
|
| 2880 |
+
"step": 409
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 1.1247429746401645,
|
| 2884 |
+
"grad_norm": 0.21888920664787292,
|
| 2885 |
+
"learning_rate": 2.3959403439996907e-05,
|
| 2886 |
+
"loss": 0.802,
|
| 2887 |
+
"step": 410
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 1.12748457847841,
|
| 2891 |
+
"grad_norm": 0.20712052285671234,
|
| 2892 |
+
"learning_rate": 2.34516786506912e-05,
|
| 2893 |
+
"loss": 0.6035,
|
| 2894 |
+
"step": 411
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 1.1302261823166553,
|
| 2898 |
+
"grad_norm": 0.21758322417736053,
|
| 2899 |
+
"learning_rate": 2.2948675722421086e-05,
|
| 2900 |
+
"loss": 0.6168,
|
| 2901 |
+
"step": 412
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 1.1329677861549006,
|
| 2905 |
+
"grad_norm": 0.20871827006340027,
|
| 2906 |
+
"learning_rate": 2.2450425682776565e-05,
|
| 2907 |
+
"loss": 0.5873,
|
| 2908 |
+
"step": 413
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 1.135709389993146,
|
| 2912 |
+
"grad_norm": 0.20848055183887482,
|
| 2913 |
+
"learning_rate": 2.195695926616702e-05,
|
| 2914 |
+
"loss": 0.5853,
|
| 2915 |
+
"step": 414
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 1.1384509938313914,
|
| 2919 |
+
"grad_norm": 0.21239161491394043,
|
| 2920 |
+
"learning_rate": 2.146830691192553e-05,
|
| 2921 |
+
"loss": 0.6978,
|
| 2922 |
+
"step": 415
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 1.1411925976696367,
|
| 2926 |
+
"grad_norm": 0.22280415892601013,
|
| 2927 |
+
"learning_rate": 2.098449876243096e-05,
|
| 2928 |
+
"loss": 0.6442,
|
| 2929 |
+
"step": 416
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 1.1439342015078822,
|
| 2933 |
+
"grad_norm": 0.19891239702701569,
|
| 2934 |
+
"learning_rate": 2.0505564661249e-05,
|
| 2935 |
+
"loss": 0.5775,
|
| 2936 |
+
"step": 417
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 1.1466758053461275,
|
| 2940 |
+
"grad_norm": 0.19570530951023102,
|
| 2941 |
+
"learning_rate": 2.0031534151290943e-05,
|
| 2942 |
+
"loss": 0.6288,
|
| 2943 |
+
"step": 418
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 1.1494174091843727,
|
| 2947 |
+
"grad_norm": 0.20307469367980957,
|
| 2948 |
+
"learning_rate": 1.9562436472991552e-05,
|
| 2949 |
+
"loss": 0.6071,
|
| 2950 |
+
"step": 419
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 1.1521590130226183,
|
| 2954 |
+
"grad_norm": 0.20724999904632568,
|
| 2955 |
+
"learning_rate": 1.9098300562505266e-05,
|
| 2956 |
+
"loss": 0.5735,
|
| 2957 |
+
"step": 420
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 1.1549006168608635,
|
| 2961 |
+
"grad_norm": 0.21279802918434143,
|
| 2962 |
+
"learning_rate": 1.863915504992131e-05,
|
| 2963 |
+
"loss": 0.6961,
|
| 2964 |
+
"step": 421
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 1.157642220699109,
|
| 2968 |
+
"grad_norm": 0.21184802055358887,
|
| 2969 |
+
"learning_rate": 1.818502825749764e-05,
|
| 2970 |
+
"loss": 0.6069,
|
| 2971 |
+
"step": 422
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 1.1603838245373543,
|
| 2975 |
+
"grad_norm": 0.2107073813676834,
|
| 2976 |
+
"learning_rate": 1.773594819791403e-05,
|
| 2977 |
+
"loss": 0.5622,
|
| 2978 |
+
"step": 423
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 1.1631254283755998,
|
| 2982 |
+
"grad_norm": 0.22249971330165863,
|
| 2983 |
+
"learning_rate": 1.7291942572543807e-05,
|
| 2984 |
+
"loss": 0.6211,
|
| 2985 |
+
"step": 424
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 1.1658670322138451,
|
| 2989 |
+
"grad_norm": 0.20617300271987915,
|
| 2990 |
+
"learning_rate": 1.6853038769745467e-05,
|
| 2991 |
+
"loss": 0.5564,
|
| 2992 |
+
"step": 425
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 1.1686086360520904,
|
| 2996 |
+
"grad_norm": 0.1992446780204773,
|
| 2997 |
+
"learning_rate": 1.6419263863172997e-05,
|
| 2998 |
+
"loss": 0.5611,
|
| 2999 |
+
"step": 426
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 1.171350239890336,
|
| 3003 |
+
"grad_norm": 0.2055218517780304,
|
| 3004 |
+
"learning_rate": 1.599064461010581e-05,
|
| 3005 |
+
"loss": 0.6667,
|
| 3006 |
+
"step": 427
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 1.1740918437285812,
|
| 3010 |
+
"grad_norm": 0.20699098706245422,
|
| 3011 |
+
"learning_rate": 1.5567207449798515e-05,
|
| 3012 |
+
"loss": 0.5557,
|
| 3013 |
+
"step": 428
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 1.1768334475668265,
|
| 3017 |
+
"grad_norm": 0.20959362387657166,
|
| 3018 |
+
"learning_rate": 1.5148978501849642e-05,
|
| 3019 |
+
"loss": 0.5549,
|
| 3020 |
+
"step": 429
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 1.179575051405072,
|
| 3024 |
+
"grad_norm": 0.2295706868171692,
|
| 3025 |
+
"learning_rate": 1.4735983564590783e-05,
|
| 3026 |
+
"loss": 0.7443,
|
| 3027 |
+
"step": 430
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 1.1823166552433173,
|
| 3031 |
+
"grad_norm": 0.20821189880371094,
|
| 3032 |
+
"learning_rate": 1.4328248113495047e-05,
|
| 3033 |
+
"loss": 0.5313,
|
| 3034 |
+
"step": 431
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 1.1850582590815626,
|
| 3038 |
+
"grad_norm": 0.2135133445262909,
|
| 3039 |
+
"learning_rate": 1.3925797299605647e-05,
|
| 3040 |
+
"loss": 0.6516,
|
| 3041 |
+
"step": 432
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 1.187799862919808,
|
| 3045 |
+
"grad_norm": 0.1996772736310959,
|
| 3046 |
+
"learning_rate": 1.3528655947984504e-05,
|
| 3047 |
+
"loss": 0.5504,
|
| 3048 |
+
"step": 433
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 1.1905414667580534,
|
| 3052 |
+
"grad_norm": 0.20553068816661835,
|
| 3053 |
+
"learning_rate": 1.3136848556180892e-05,
|
| 3054 |
+
"loss": 0.6391,
|
| 3055 |
+
"step": 434
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 1.193283070596299,
|
| 3059 |
+
"grad_norm": 0.23739643394947052,
|
| 3060 |
+
"learning_rate": 1.2750399292720283e-05,
|
| 3061 |
+
"loss": 0.6973,
|
| 3062 |
+
"step": 435
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 1.1960246744345442,
|
| 3066 |
+
"grad_norm": 0.21224342286586761,
|
| 3067 |
+
"learning_rate": 1.2369331995613665e-05,
|
| 3068 |
+
"loss": 0.5342,
|
| 3069 |
+
"step": 436
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 1.1987662782727897,
|
| 3073 |
+
"grad_norm": 0.2122921198606491,
|
| 3074 |
+
"learning_rate": 1.1993670170886806e-05,
|
| 3075 |
+
"loss": 0.5486,
|
| 3076 |
+
"step": 437
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 1.201507882111035,
|
| 3080 |
+
"grad_norm": 0.21199892461299896,
|
| 3081 |
+
"learning_rate": 1.1623436991130654e-05,
|
| 3082 |
+
"loss": 0.8089,
|
| 3083 |
+
"step": 438
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 1.2042494859492803,
|
| 3087 |
+
"grad_norm": 0.19737985730171204,
|
| 3088 |
+
"learning_rate": 1.1258655294071685e-05,
|
| 3089 |
+
"loss": 0.5941,
|
| 3090 |
+
"step": 439
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 1.2069910897875258,
|
| 3094 |
+
"grad_norm": 0.20743800699710846,
|
| 3095 |
+
"learning_rate": 1.0899347581163221e-05,
|
| 3096 |
+
"loss": 0.6042,
|
| 3097 |
+
"step": 440
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 1.209732693625771,
|
| 3101 |
+
"grad_norm": 0.20547975599765778,
|
| 3102 |
+
"learning_rate": 1.0545536016197499e-05,
|
| 3103 |
+
"loss": 0.6674,
|
| 3104 |
+
"step": 441
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 1.2124742974640164,
|
| 3108 |
+
"grad_norm": 0.2322980910539627,
|
| 3109 |
+
"learning_rate": 1.0197242423938446e-05,
|
| 3110 |
+
"loss": 0.7908,
|
| 3111 |
+
"step": 442
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 1.2152159013022619,
|
| 3115 |
+
"grad_norm": 0.2214648574590683,
|
| 3116 |
+
"learning_rate": 9.854488288775422e-06,
|
| 3117 |
+
"loss": 0.5208,
|
| 3118 |
+
"step": 443
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 1.2179575051405072,
|
| 3122 |
+
"grad_norm": 0.21311421692371368,
|
| 3123 |
+
"learning_rate": 9.517294753398064e-06,
|
| 3124 |
+
"loss": 0.609,
|
| 3125 |
+
"step": 444
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 1.2206991089787527,
|
| 3129 |
+
"grad_norm": 0.20191775262355804,
|
| 3130 |
+
"learning_rate": 9.185682617491863e-06,
|
| 3131 |
+
"loss": 0.6295,
|
| 3132 |
+
"step": 445
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 1.223440712816998,
|
| 3136 |
+
"grad_norm": 0.1986418217420578,
|
| 3137 |
+
"learning_rate": 8.85967233645547e-06,
|
| 3138 |
+
"loss": 0.6387,
|
| 3139 |
+
"step": 446
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 1.2261823166552432,
|
| 3143 |
+
"grad_norm": 0.20926733314990997,
|
| 3144 |
+
"learning_rate": 8.539284020138638e-06,
|
| 3145 |
+
"loss": 0.6503,
|
| 3146 |
+
"step": 447
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 1.2289239204934888,
|
| 3150 |
+
"grad_norm": 0.21828588843345642,
|
| 3151 |
+
"learning_rate": 8.224537431601886e-06,
|
| 3152 |
+
"loss": 0.7032,
|
| 3153 |
+
"step": 448
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 1.231665524331734,
|
| 3157 |
+
"grad_norm": 0.21580448746681213,
|
| 3158 |
+
"learning_rate": 7.915451985897382e-06,
|
| 3159 |
+
"loss": 0.6478,
|
| 3160 |
+
"step": 449
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 1.2344071281699796,
|
| 3164 |
+
"grad_norm": 0.20177645981311798,
|
| 3165 |
+
"learning_rate": 7.612046748871327e-06,
|
| 3166 |
+
"loss": 0.621,
|
| 3167 |
+
"step": 450
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 1.2371487320082248,
|
| 3171 |
+
"grad_norm": 0.21839949488639832,
|
| 3172 |
+
"learning_rate": 7.314340435987921e-06,
|
| 3173 |
+
"loss": 0.6387,
|
| 3174 |
+
"step": 451
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 1.2398903358464701,
|
| 3178 |
+
"grad_norm": 0.21275487542152405,
|
| 3179 |
+
"learning_rate": 7.022351411174866e-06,
|
| 3180 |
+
"loss": 0.6035,
|
| 3181 |
+
"step": 452
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 1.2426319396847156,
|
| 3185 |
+
"grad_norm": 0.21411579847335815,
|
| 3186 |
+
"learning_rate": 6.736097685690601e-06,
|
| 3187 |
+
"loss": 0.6683,
|
| 3188 |
+
"step": 453
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 1.245373543522961,
|
| 3192 |
+
"grad_norm": 0.193324014544487,
|
| 3193 |
+
"learning_rate": 6.455596917013273e-06,
|
| 3194 |
+
"loss": 0.4833,
|
| 3195 |
+
"step": 454
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 1.2481151473612062,
|
| 3199 |
+
"grad_norm": 0.23513925075531006,
|
| 3200 |
+
"learning_rate": 6.180866407751595e-06,
|
| 3201 |
+
"loss": 0.6459,
|
| 3202 |
+
"step": 455
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 1.2508567511994517,
|
| 3206 |
+
"grad_norm": 0.22250422835350037,
|
| 3207 |
+
"learning_rate": 5.911923104577455e-06,
|
| 3208 |
+
"loss": 0.5959,
|
| 3209 |
+
"step": 456
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 1.253598355037697,
|
| 3213 |
+
"grad_norm": 0.21887513995170593,
|
| 3214 |
+
"learning_rate": 5.6487835971806515e-06,
|
| 3215 |
+
"loss": 0.5718,
|
| 3216 |
+
"step": 457
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 1.2563399588759423,
|
| 3220 |
+
"grad_norm": 0.21817608177661896,
|
| 3221 |
+
"learning_rate": 5.39146411724547e-06,
|
| 3222 |
+
"loss": 0.6227,
|
| 3223 |
+
"step": 458
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 1.2590815627141878,
|
| 3227 |
+
"grad_norm": 0.21858827769756317,
|
| 3228 |
+
"learning_rate": 5.13998053744954e-06,
|
| 3229 |
+
"loss": 0.5857,
|
| 3230 |
+
"step": 459
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 1.261823166552433,
|
| 3234 |
+
"grad_norm": 0.20830455422401428,
|
| 3235 |
+
"learning_rate": 4.8943483704846475e-06,
|
| 3236 |
+
"loss": 0.5913,
|
| 3237 |
+
"step": 460
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 1.2645647703906786,
|
| 3241 |
+
"grad_norm": 0.21001742780208588,
|
| 3242 |
+
"learning_rate": 4.65458276809988e-06,
|
| 3243 |
+
"loss": 0.6005,
|
| 3244 |
+
"step": 461
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 1.267306374228924,
|
| 3248 |
+
"grad_norm": 0.20365364849567413,
|
| 3249 |
+
"learning_rate": 4.420698520166988e-06,
|
| 3250 |
+
"loss": 0.5436,
|
| 3251 |
+
"step": 462
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 1.2700479780671694,
|
| 3255 |
+
"grad_norm": 0.22307349741458893,
|
| 3256 |
+
"learning_rate": 4.19271005376809e-06,
|
| 3257 |
+
"loss": 0.7224,
|
| 3258 |
+
"step": 463
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 1.2727895819054147,
|
| 3262 |
+
"grad_norm": 0.20303815603256226,
|
| 3263 |
+
"learning_rate": 3.970631432305694e-06,
|
| 3264 |
+
"loss": 0.6244,
|
| 3265 |
+
"step": 464
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 1.27553118574366,
|
| 3269 |
+
"grad_norm": 0.20350705087184906,
|
| 3270 |
+
"learning_rate": 3.7544763546352834e-06,
|
| 3271 |
+
"loss": 0.566,
|
| 3272 |
+
"step": 465
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 1.2782727895819055,
|
| 3276 |
+
"grad_norm": 0.21413709223270416,
|
| 3277 |
+
"learning_rate": 3.5442581542201923e-06,
|
| 3278 |
+
"loss": 0.6124,
|
| 3279 |
+
"step": 466
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 1.2810143934201508,
|
| 3283 |
+
"grad_norm": 0.21414998173713684,
|
| 3284 |
+
"learning_rate": 3.339989798309273e-06,
|
| 3285 |
+
"loss": 0.5667,
|
| 3286 |
+
"step": 467
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 1.283755997258396,
|
| 3290 |
+
"grad_norm": 0.20336252450942993,
|
| 3291 |
+
"learning_rate": 3.1416838871368924e-06,
|
| 3292 |
+
"loss": 0.6196,
|
| 3293 |
+
"step": 468
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 1.2864976010966416,
|
| 3297 |
+
"grad_norm": 0.215332493185997,
|
| 3298 |
+
"learning_rate": 2.949352653145754e-06,
|
| 3299 |
+
"loss": 0.5635,
|
| 3300 |
+
"step": 469
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 1.2892392049348869,
|
| 3304 |
+
"grad_norm": 0.22713740170001984,
|
| 3305 |
+
"learning_rate": 2.7630079602323442e-06,
|
| 3306 |
+
"loss": 0.6363,
|
| 3307 |
+
"step": 470
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 1.2919808087731321,
|
| 3311 |
+
"grad_norm": 0.2187124490737915,
|
| 3312 |
+
"learning_rate": 2.582661303015066e-06,
|
| 3313 |
+
"loss": 0.669,
|
| 3314 |
+
"step": 471
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 1.2947224126113777,
|
| 3318 |
+
"grad_norm": 0.19379432499408722,
|
| 3319 |
+
"learning_rate": 2.4083238061252567e-06,
|
| 3320 |
+
"loss": 0.4984,
|
| 3321 |
+
"step": 472
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 1.297464016449623,
|
| 3325 |
+
"grad_norm": 0.21041575074195862,
|
| 3326 |
+
"learning_rate": 2.240006223520941e-06,
|
| 3327 |
+
"loss": 0.6761,
|
| 3328 |
+
"step": 473
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 1.3002056202878685,
|
| 3332 |
+
"grad_norm": 0.21407803893089294,
|
| 3333 |
+
"learning_rate": 2.0777189378234143e-06,
|
| 3334 |
+
"loss": 0.5696,
|
| 3335 |
+
"step": 474
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 1.3029472241261137,
|
| 3339 |
+
"grad_norm": 0.23364566266536713,
|
| 3340 |
+
"learning_rate": 1.921471959676957e-06,
|
| 3341 |
+
"loss": 0.6658,
|
| 3342 |
+
"step": 475
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 1.3056888279643593,
|
| 3346 |
+
"grad_norm": 0.20902545750141144,
|
| 3347 |
+
"learning_rate": 1.771274927131139e-06,
|
| 3348 |
+
"loss": 0.5335,
|
| 3349 |
+
"step": 476
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 1.3084304318026045,
|
| 3353 |
+
"grad_norm": 0.22559121251106262,
|
| 3354 |
+
"learning_rate": 1.6271371050464056e-06,
|
| 3355 |
+
"loss": 0.6188,
|
| 3356 |
+
"step": 477
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 1.3111720356408498,
|
| 3360 |
+
"grad_norm": 0.21116699278354645,
|
| 3361 |
+
"learning_rate": 1.4890673845226133e-06,
|
| 3362 |
+
"loss": 0.6316,
|
| 3363 |
+
"step": 478
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 1.3139136394790953,
|
| 3367 |
+
"grad_norm": 0.19573360681533813,
|
| 3368 |
+
"learning_rate": 1.3570742823504567e-06,
|
| 3369 |
+
"loss": 0.5112,
|
| 3370 |
+
"step": 479
|
| 3371 |
+
},
|
| 3372 |
+
{
|
| 3373 |
+
"epoch": 1.3166552433173406,
|
| 3374 |
+
"grad_norm": 0.26849284768104553,
|
| 3375 |
+
"learning_rate": 1.231165940486234e-06,
|
| 3376 |
+
"loss": 0.6787,
|
| 3377 |
+
"step": 480
|
| 3378 |
+
},
|
| 3379 |
+
{
|
| 3380 |
+
"epoch": 1.319396847155586,
|
| 3381 |
+
"grad_norm": 0.19678130745887756,
|
| 3382 |
+
"learning_rate": 1.1113501255495485e-06,
|
| 3383 |
+
"loss": 0.5658,
|
| 3384 |
+
"step": 481
|
| 3385 |
+
},
|
| 3386 |
+
{
|
| 3387 |
+
"epoch": 1.3221384509938314,
|
| 3388 |
+
"grad_norm": 0.21285156905651093,
|
| 3389 |
+
"learning_rate": 9.976342283442463e-07,
|
| 3390 |
+
"loss": 0.6645,
|
| 3391 |
+
"step": 482
|
| 3392 |
+
},
|
| 3393 |
+
{
|
| 3394 |
+
"epoch": 1.3248800548320767,
|
| 3395 |
+
"grad_norm": 0.20968644320964813,
|
| 3396 |
+
"learning_rate": 8.900252634025274e-07,
|
| 3397 |
+
"loss": 0.6224,
|
| 3398 |
+
"step": 483
|
| 3399 |
+
},
|
| 3400 |
+
{
|
| 3401 |
+
"epoch": 1.327621658670322,
|
| 3402 |
+
"grad_norm": 0.20871570706367493,
|
| 3403 |
+
"learning_rate": 7.885298685522235e-07,
|
| 3404 |
+
"loss": 0.56,
|
| 3405 |
+
"step": 484
|
| 3406 |
+
},
|
| 3407 |
+
{
|
| 3408 |
+
"epoch": 1.3303632625085675,
|
| 3409 |
+
"grad_norm": 0.21702443063259125,
|
| 3410 |
+
"learning_rate": 6.931543045073708e-07,
|
| 3411 |
+
"loss": 0.6191,
|
| 3412 |
+
"step": 485
|
| 3413 |
+
},
|
| 3414 |
+
{
|
| 3415 |
+
"epoch": 1.333104866346813,
|
| 3416 |
+
"grad_norm": 0.2229696661233902,
|
| 3417 |
+
"learning_rate": 6.039044544820404e-07,
|
| 3418 |
+
"loss": 0.7263,
|
| 3419 |
+
"step": 486
|
| 3420 |
+
},
|
| 3421 |
+
{
|
| 3422 |
+
"epoch": 1.3358464701850583,
|
| 3423 |
+
"grad_norm": 0.20075109601020813,
|
| 3424 |
+
"learning_rate": 5.207858238273522e-07,
|
| 3425 |
+
"loss": 0.5889,
|
| 3426 |
+
"step": 487
|
| 3427 |
+
},
|
| 3428 |
+
{
|
| 3429 |
+
"epoch": 1.3385880740233036,
|
| 3430 |
+
"grad_norm": 0.23196394741535187,
|
| 3431 |
+
"learning_rate": 4.438035396920004e-07,
|
| 3432 |
+
"loss": 0.6561,
|
| 3433 |
+
"step": 488
|
| 3434 |
+
},
|
| 3435 |
+
{
|
| 3436 |
+
"epoch": 1.341329677861549,
|
| 3437 |
+
"grad_norm": 0.19844023883342743,
|
| 3438 |
+
"learning_rate": 3.7296235070587435e-07,
|
| 3439 |
+
"loss": 0.5336,
|
| 3440 |
+
"step": 489
|
| 3441 |
+
},
|
| 3442 |
+
{
|
| 3443 |
+
"epoch": 1.3440712816997944,
|
| 3444 |
+
"grad_norm": 0.215968519449234,
|
| 3445 |
+
"learning_rate": 3.0826662668720364e-07,
|
| 3446 |
+
"loss": 0.6015,
|
| 3447 |
+
"step": 490
|
| 3448 |
+
},
|
| 3449 |
+
{
|
| 3450 |
+
"epoch": 1.3468128855380397,
|
| 3451 |
+
"grad_norm": 0.20940376818180084,
|
| 3452 |
+
"learning_rate": 2.497203583729957e-07,
|
| 3453 |
+
"loss": 0.5726,
|
| 3454 |
+
"step": 491
|
| 3455 |
+
},
|
| 3456 |
+
{
|
| 3457 |
+
"epoch": 1.3495544893762852,
|
| 3458 |
+
"grad_norm": 0.20483621954917908,
|
| 3459 |
+
"learning_rate": 1.973271571728441e-07,
|
| 3460 |
+
"loss": 0.6296,
|
| 3461 |
+
"step": 492
|
| 3462 |
+
},
|
| 3463 |
+
{
|
| 3464 |
+
"epoch": 1.3522960932145305,
|
| 3465 |
+
"grad_norm": 0.20878347754478455,
|
| 3466 |
+
"learning_rate": 1.5109025494620677e-07,
|
| 3467 |
+
"loss": 0.5886,
|
| 3468 |
+
"step": 493
|
| 3469 |
+
},
|
| 3470 |
+
{
|
| 3471 |
+
"epoch": 1.3550376970527758,
|
| 3472 |
+
"grad_norm": 0.21599401533603668,
|
| 3473 |
+
"learning_rate": 1.1101250380300965e-07,
|
| 3474 |
+
"loss": 0.6098,
|
| 3475 |
+
"step": 494
|
| 3476 |
+
},
|
| 3477 |
+
{
|
| 3478 |
+
"epoch": 1.3577793008910213,
|
| 3479 |
+
"grad_norm": 0.2182384878396988,
|
| 3480 |
+
"learning_rate": 7.709637592770991e-08,
|
| 3481 |
+
"loss": 0.7436,
|
| 3482 |
+
"step": 495
|
| 3483 |
+
},
|
| 3484 |
+
{
|
| 3485 |
+
"epoch": 1.3605209047292666,
|
| 3486 |
+
"grad_norm": 0.20137697458267212,
|
| 3487 |
+
"learning_rate": 4.934396342684e-08,
|
| 3488 |
+
"loss": 0.6014,
|
| 3489 |
+
"step": 496
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 1.363262508567512,
|
| 3493 |
+
"grad_norm": 0.20790421962738037,
|
| 3494 |
+
"learning_rate": 2.7756978199944273e-08,
|
| 3495 |
+
"loss": 0.6034,
|
| 3496 |
+
"step": 497
|
| 3497 |
+
},
|
| 3498 |
+
{
|
| 3499 |
+
"epoch": 1.3660041124057574,
|
| 3500 |
+
"grad_norm": 0.23739448189735413,
|
| 3501 |
+
"learning_rate": 1.2336751833941229e-08,
|
| 3502 |
+
"loss": 0.6228,
|
| 3503 |
+
"step": 498
|
| 3504 |
+
},
|
| 3505 |
+
{
|
| 3506 |
+
"epoch": 1.3687457162440029,
|
| 3507 |
+
"grad_norm": 0.2084830105304718,
|
| 3508 |
+
"learning_rate": 3.0842355210336515e-09,
|
| 3509 |
+
"loss": 0.5611,
|
| 3510 |
+
"step": 499
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 1.3714873200822482,
|
| 3514 |
+
"grad_norm": 0.19652077555656433,
|
| 3515 |
+
"learning_rate": 0.0,
|
| 3516 |
+
"loss": 0.5083,
|
| 3517 |
+
"step": 500
|
| 3518 |
}
|
| 3519 |
],
|
| 3520 |
"logging_steps": 1,
|
|
|
|
| 3529 |
"should_evaluate": false,
|
| 3530 |
"should_log": false,
|
| 3531 |
"should_save": true,
|
| 3532 |
+
"should_training_stop": true
|
| 3533 |
},
|
| 3534 |
"attributes": {}
|
| 3535 |
}
|
| 3536 |
},
|
| 3537 |
+
"total_flos": 1.5051838837405778e+18,
|
| 3538 |
"train_batch_size": 2,
|
| 3539 |
"trial_name": null,
|
| 3540 |
"trial_params": null
|