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 +1447 -3
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 185097216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b11b2ff539c63d3c6bac569f7766703b625a1ff96963c99b63259f1624298ab
|
| 3 |
size 185097216
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 370329978
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa9d2a0ec897c5663a4be748f0fc34398833a23f3d592a7fbdfecc9b082648a1
|
| 3 |
size 370329978
|
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:c4c1665affd4ed287cc307c033aa9b83f5129c6f23f546aeee5169ca1b8994af
|
| 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 sha256:2abd31d46a5b8aba640a7054a1f1886e89e9f6b46ac0ecd782b859dc4214d256
|
| 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:b6356c5d3297eec0185fbb12716367c7cafc523337fc20be9926fc46325c429a
|
| 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 sha256:866ba9b91c55d95ef1da6feb07d2ffecceb23df8b5608e080221fc6dbac9ed2f
|
| 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 sha256:394d36bdefd8333a39a124a3477c46f23015ea9e02d59f4ba6dcce13ab46168f
|
| 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 sha256:577c7045aadc9972169583d0b06cfc68d51e9cd2d016cd56617f0749af75cb31
|
| 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 sha256:5fd7094860928ca612494a904481849cce54a143aaf53d511c3fdafa80f4fefa
|
| 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:b8e1cc2af93b92de277256d30ceb20997578b8263be191983b475c3ad46a0790
|
| 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:fe4f2fce602372e0140548196138ac3dca07af36635f6b15963598f4bc25bf6b
|
| 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:f4790382837b89dc610979dcc6f00fc79c7847cb5b0f57da5641fc8d16ea0a4a
|
| 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 sha256:75fc12f97c880c2c5755340b9a95c8af52d0a5dc3399cb79b9acc0bfca303aae
|
| 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": 1431,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3624,6 +3624,1450 @@
|
|
| 3624 |
"eval_samples_per_second": 109.582,
|
| 3625 |
"eval_steps_per_second": 1.425,
|
| 3626 |
"step": 7155
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3627 |
}
|
| 3628 |
],
|
| 3629 |
"logging_steps": 14,
|
|
@@ -3643,7 +5087,7 @@
|
|
| 3643 |
"attributes": {}
|
| 3644 |
}
|
| 3645 |
},
|
| 3646 |
-
"total_flos":
|
| 3647 |
"train_batch_size": 8,
|
| 3648 |
"trial_name": null,
|
| 3649 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.6997188411365105,
|
| 5 |
"eval_steps": 1431,
|
| 6 |
+
"global_step": 10017,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3624 |
"eval_samples_per_second": 109.582,
|
| 3625 |
"eval_steps_per_second": 1.425,
|
| 3626 |
"step": 7155
|
| 3627 |
+
},
|
| 3628 |
+
{
|
| 3629 |
+
"epoch": 0.5007072629795855,
|
| 3630 |
+
"grad_norm": 0.380859375,
|
| 3631 |
+
"learning_rate": 0.001,
|
| 3632 |
+
"loss": 1.2987,
|
| 3633 |
+
"step": 7168
|
| 3634 |
+
},
|
| 3635 |
+
{
|
| 3636 |
+
"epoch": 0.5016852068525924,
|
| 3637 |
+
"grad_norm": 0.267578125,
|
| 3638 |
+
"learning_rate": 0.001,
|
| 3639 |
+
"loss": 1.2759,
|
| 3640 |
+
"step": 7182
|
| 3641 |
+
},
|
| 3642 |
+
{
|
| 3643 |
+
"epoch": 0.5026631507255994,
|
| 3644 |
+
"grad_norm": 0.376953125,
|
| 3645 |
+
"learning_rate": 0.001,
|
| 3646 |
+
"loss": 1.2814,
|
| 3647 |
+
"step": 7196
|
| 3648 |
+
},
|
| 3649 |
+
{
|
| 3650 |
+
"epoch": 0.5036410945986064,
|
| 3651 |
+
"grad_norm": 0.359375,
|
| 3652 |
+
"learning_rate": 0.001,
|
| 3653 |
+
"loss": 1.2701,
|
| 3654 |
+
"step": 7210
|
| 3655 |
+
},
|
| 3656 |
+
{
|
| 3657 |
+
"epoch": 0.5046190384716134,
|
| 3658 |
+
"grad_norm": 0.314453125,
|
| 3659 |
+
"learning_rate": 0.001,
|
| 3660 |
+
"loss": 1.2857,
|
| 3661 |
+
"step": 7224
|
| 3662 |
+
},
|
| 3663 |
+
{
|
| 3664 |
+
"epoch": 0.5055969823446205,
|
| 3665 |
+
"grad_norm": 0.234375,
|
| 3666 |
+
"learning_rate": 0.001,
|
| 3667 |
+
"loss": 1.2707,
|
| 3668 |
+
"step": 7238
|
| 3669 |
+
},
|
| 3670 |
+
{
|
| 3671 |
+
"epoch": 0.5065749262176275,
|
| 3672 |
+
"grad_norm": 0.306640625,
|
| 3673 |
+
"learning_rate": 0.001,
|
| 3674 |
+
"loss": 1.2851,
|
| 3675 |
+
"step": 7252
|
| 3676 |
+
},
|
| 3677 |
+
{
|
| 3678 |
+
"epoch": 0.5075528700906344,
|
| 3679 |
+
"grad_norm": 0.337890625,
|
| 3680 |
+
"learning_rate": 0.001,
|
| 3681 |
+
"loss": 1.2722,
|
| 3682 |
+
"step": 7266
|
| 3683 |
+
},
|
| 3684 |
+
{
|
| 3685 |
+
"epoch": 0.5085308139636414,
|
| 3686 |
+
"grad_norm": 0.345703125,
|
| 3687 |
+
"learning_rate": 0.001,
|
| 3688 |
+
"loss": 1.277,
|
| 3689 |
+
"step": 7280
|
| 3690 |
+
},
|
| 3691 |
+
{
|
| 3692 |
+
"epoch": 0.5095087578366484,
|
| 3693 |
+
"grad_norm": 0.33203125,
|
| 3694 |
+
"learning_rate": 0.001,
|
| 3695 |
+
"loss": 1.3021,
|
| 3696 |
+
"step": 7294
|
| 3697 |
+
},
|
| 3698 |
+
{
|
| 3699 |
+
"epoch": 0.5104867017096555,
|
| 3700 |
+
"grad_norm": 0.2314453125,
|
| 3701 |
+
"learning_rate": 0.001,
|
| 3702 |
+
"loss": 1.2856,
|
| 3703 |
+
"step": 7308
|
| 3704 |
+
},
|
| 3705 |
+
{
|
| 3706 |
+
"epoch": 0.5114646455826625,
|
| 3707 |
+
"grad_norm": 0.25,
|
| 3708 |
+
"learning_rate": 0.001,
|
| 3709 |
+
"loss": 1.2704,
|
| 3710 |
+
"step": 7322
|
| 3711 |
+
},
|
| 3712 |
+
{
|
| 3713 |
+
"epoch": 0.5124425894556695,
|
| 3714 |
+
"grad_norm": 0.27734375,
|
| 3715 |
+
"learning_rate": 0.001,
|
| 3716 |
+
"loss": 1.2837,
|
| 3717 |
+
"step": 7336
|
| 3718 |
+
},
|
| 3719 |
+
{
|
| 3720 |
+
"epoch": 0.5134205333286764,
|
| 3721 |
+
"grad_norm": 0.283203125,
|
| 3722 |
+
"learning_rate": 0.001,
|
| 3723 |
+
"loss": 1.277,
|
| 3724 |
+
"step": 7350
|
| 3725 |
+
},
|
| 3726 |
+
{
|
| 3727 |
+
"epoch": 0.5143984772016834,
|
| 3728 |
+
"grad_norm": 0.2734375,
|
| 3729 |
+
"learning_rate": 0.001,
|
| 3730 |
+
"loss": 1.2838,
|
| 3731 |
+
"step": 7364
|
| 3732 |
+
},
|
| 3733 |
+
{
|
| 3734 |
+
"epoch": 0.5153764210746905,
|
| 3735 |
+
"grad_norm": 0.296875,
|
| 3736 |
+
"learning_rate": 0.001,
|
| 3737 |
+
"loss": 1.2762,
|
| 3738 |
+
"step": 7378
|
| 3739 |
+
},
|
| 3740 |
+
{
|
| 3741 |
+
"epoch": 0.5163543649476975,
|
| 3742 |
+
"grad_norm": 0.298828125,
|
| 3743 |
+
"learning_rate": 0.001,
|
| 3744 |
+
"loss": 1.2749,
|
| 3745 |
+
"step": 7392
|
| 3746 |
+
},
|
| 3747 |
+
{
|
| 3748 |
+
"epoch": 0.5173323088207045,
|
| 3749 |
+
"grad_norm": 0.28125,
|
| 3750 |
+
"learning_rate": 0.001,
|
| 3751 |
+
"loss": 1.2791,
|
| 3752 |
+
"step": 7406
|
| 3753 |
+
},
|
| 3754 |
+
{
|
| 3755 |
+
"epoch": 0.5183102526937114,
|
| 3756 |
+
"grad_norm": 0.400390625,
|
| 3757 |
+
"learning_rate": 0.001,
|
| 3758 |
+
"loss": 1.2708,
|
| 3759 |
+
"step": 7420
|
| 3760 |
+
},
|
| 3761 |
+
{
|
| 3762 |
+
"epoch": 0.5192881965667184,
|
| 3763 |
+
"grad_norm": 0.328125,
|
| 3764 |
+
"learning_rate": 0.001,
|
| 3765 |
+
"loss": 1.2773,
|
| 3766 |
+
"step": 7434
|
| 3767 |
+
},
|
| 3768 |
+
{
|
| 3769 |
+
"epoch": 0.5202661404397255,
|
| 3770 |
+
"grad_norm": 0.287109375,
|
| 3771 |
+
"learning_rate": 0.001,
|
| 3772 |
+
"loss": 1.2783,
|
| 3773 |
+
"step": 7448
|
| 3774 |
+
},
|
| 3775 |
+
{
|
| 3776 |
+
"epoch": 0.5212440843127325,
|
| 3777 |
+
"grad_norm": 0.341796875,
|
| 3778 |
+
"learning_rate": 0.001,
|
| 3779 |
+
"loss": 1.2944,
|
| 3780 |
+
"step": 7462
|
| 3781 |
+
},
|
| 3782 |
+
{
|
| 3783 |
+
"epoch": 0.5222220281857395,
|
| 3784 |
+
"grad_norm": 0.369140625,
|
| 3785 |
+
"learning_rate": 0.001,
|
| 3786 |
+
"loss": 1.2714,
|
| 3787 |
+
"step": 7476
|
| 3788 |
+
},
|
| 3789 |
+
{
|
| 3790 |
+
"epoch": 0.5231999720587465,
|
| 3791 |
+
"grad_norm": 0.2421875,
|
| 3792 |
+
"learning_rate": 0.001,
|
| 3793 |
+
"loss": 1.2711,
|
| 3794 |
+
"step": 7490
|
| 3795 |
+
},
|
| 3796 |
+
{
|
| 3797 |
+
"epoch": 0.5241779159317534,
|
| 3798 |
+
"grad_norm": 0.51953125,
|
| 3799 |
+
"learning_rate": 0.001,
|
| 3800 |
+
"loss": 1.2808,
|
| 3801 |
+
"step": 7504
|
| 3802 |
+
},
|
| 3803 |
+
{
|
| 3804 |
+
"epoch": 0.5251558598047605,
|
| 3805 |
+
"grad_norm": 0.271484375,
|
| 3806 |
+
"learning_rate": 0.001,
|
| 3807 |
+
"loss": 1.2765,
|
| 3808 |
+
"step": 7518
|
| 3809 |
+
},
|
| 3810 |
+
{
|
| 3811 |
+
"epoch": 0.5261338036777675,
|
| 3812 |
+
"grad_norm": 0.30078125,
|
| 3813 |
+
"learning_rate": 0.001,
|
| 3814 |
+
"loss": 1.2702,
|
| 3815 |
+
"step": 7532
|
| 3816 |
+
},
|
| 3817 |
+
{
|
| 3818 |
+
"epoch": 0.5271117475507745,
|
| 3819 |
+
"grad_norm": 0.26953125,
|
| 3820 |
+
"learning_rate": 0.001,
|
| 3821 |
+
"loss": 1.2802,
|
| 3822 |
+
"step": 7546
|
| 3823 |
+
},
|
| 3824 |
+
{
|
| 3825 |
+
"epoch": 0.5280896914237815,
|
| 3826 |
+
"grad_norm": 0.57421875,
|
| 3827 |
+
"learning_rate": 0.001,
|
| 3828 |
+
"loss": 1.2733,
|
| 3829 |
+
"step": 7560
|
| 3830 |
+
},
|
| 3831 |
+
{
|
| 3832 |
+
"epoch": 0.5290676352967885,
|
| 3833 |
+
"grad_norm": 0.494140625,
|
| 3834 |
+
"learning_rate": 0.001,
|
| 3835 |
+
"loss": 1.2575,
|
| 3836 |
+
"step": 7574
|
| 3837 |
+
},
|
| 3838 |
+
{
|
| 3839 |
+
"epoch": 0.5300455791697956,
|
| 3840 |
+
"grad_norm": 0.376953125,
|
| 3841 |
+
"learning_rate": 0.001,
|
| 3842 |
+
"loss": 1.2863,
|
| 3843 |
+
"step": 7588
|
| 3844 |
+
},
|
| 3845 |
+
{
|
| 3846 |
+
"epoch": 0.5310235230428025,
|
| 3847 |
+
"grad_norm": 0.369140625,
|
| 3848 |
+
"learning_rate": 0.001,
|
| 3849 |
+
"loss": 1.2815,
|
| 3850 |
+
"step": 7602
|
| 3851 |
+
},
|
| 3852 |
+
{
|
| 3853 |
+
"epoch": 0.5320014669158095,
|
| 3854 |
+
"grad_norm": 0.3046875,
|
| 3855 |
+
"learning_rate": 0.001,
|
| 3856 |
+
"loss": 1.2745,
|
| 3857 |
+
"step": 7616
|
| 3858 |
+
},
|
| 3859 |
+
{
|
| 3860 |
+
"epoch": 0.5329794107888165,
|
| 3861 |
+
"grad_norm": 0.271484375,
|
| 3862 |
+
"learning_rate": 0.001,
|
| 3863 |
+
"loss": 1.294,
|
| 3864 |
+
"step": 7630
|
| 3865 |
+
},
|
| 3866 |
+
{
|
| 3867 |
+
"epoch": 0.5339573546618235,
|
| 3868 |
+
"grad_norm": 0.2265625,
|
| 3869 |
+
"learning_rate": 0.001,
|
| 3870 |
+
"loss": 1.2797,
|
| 3871 |
+
"step": 7644
|
| 3872 |
+
},
|
| 3873 |
+
{
|
| 3874 |
+
"epoch": 0.5349352985348306,
|
| 3875 |
+
"grad_norm": 0.3203125,
|
| 3876 |
+
"learning_rate": 0.001,
|
| 3877 |
+
"loss": 1.2665,
|
| 3878 |
+
"step": 7658
|
| 3879 |
+
},
|
| 3880 |
+
{
|
| 3881 |
+
"epoch": 0.5359132424078376,
|
| 3882 |
+
"grad_norm": 0.458984375,
|
| 3883 |
+
"learning_rate": 0.001,
|
| 3884 |
+
"loss": 1.28,
|
| 3885 |
+
"step": 7672
|
| 3886 |
+
},
|
| 3887 |
+
{
|
| 3888 |
+
"epoch": 0.5368911862808445,
|
| 3889 |
+
"grad_norm": 0.359375,
|
| 3890 |
+
"learning_rate": 0.001,
|
| 3891 |
+
"loss": 1.3057,
|
| 3892 |
+
"step": 7686
|
| 3893 |
+
},
|
| 3894 |
+
{
|
| 3895 |
+
"epoch": 0.5378691301538515,
|
| 3896 |
+
"grad_norm": 0.37109375,
|
| 3897 |
+
"learning_rate": 0.001,
|
| 3898 |
+
"loss": 1.258,
|
| 3899 |
+
"step": 7700
|
| 3900 |
+
},
|
| 3901 |
+
{
|
| 3902 |
+
"epoch": 0.5388470740268585,
|
| 3903 |
+
"grad_norm": 0.33203125,
|
| 3904 |
+
"learning_rate": 0.001,
|
| 3905 |
+
"loss": 1.2742,
|
| 3906 |
+
"step": 7714
|
| 3907 |
+
},
|
| 3908 |
+
{
|
| 3909 |
+
"epoch": 0.5398250178998655,
|
| 3910 |
+
"grad_norm": 0.365234375,
|
| 3911 |
+
"learning_rate": 0.001,
|
| 3912 |
+
"loss": 1.277,
|
| 3913 |
+
"step": 7728
|
| 3914 |
+
},
|
| 3915 |
+
{
|
| 3916 |
+
"epoch": 0.5408029617728726,
|
| 3917 |
+
"grad_norm": 0.412109375,
|
| 3918 |
+
"learning_rate": 0.001,
|
| 3919 |
+
"loss": 1.2819,
|
| 3920 |
+
"step": 7742
|
| 3921 |
+
},
|
| 3922 |
+
{
|
| 3923 |
+
"epoch": 0.5417809056458796,
|
| 3924 |
+
"grad_norm": 0.263671875,
|
| 3925 |
+
"learning_rate": 0.001,
|
| 3926 |
+
"loss": 1.3018,
|
| 3927 |
+
"step": 7756
|
| 3928 |
+
},
|
| 3929 |
+
{
|
| 3930 |
+
"epoch": 0.5427588495188865,
|
| 3931 |
+
"grad_norm": 0.263671875,
|
| 3932 |
+
"learning_rate": 0.001,
|
| 3933 |
+
"loss": 1.2619,
|
| 3934 |
+
"step": 7770
|
| 3935 |
+
},
|
| 3936 |
+
{
|
| 3937 |
+
"epoch": 0.5437367933918935,
|
| 3938 |
+
"grad_norm": 0.296875,
|
| 3939 |
+
"learning_rate": 0.001,
|
| 3940 |
+
"loss": 1.2513,
|
| 3941 |
+
"step": 7784
|
| 3942 |
+
},
|
| 3943 |
+
{
|
| 3944 |
+
"epoch": 0.5447147372649005,
|
| 3945 |
+
"grad_norm": 0.271484375,
|
| 3946 |
+
"learning_rate": 0.001,
|
| 3947 |
+
"loss": 1.2456,
|
| 3948 |
+
"step": 7798
|
| 3949 |
+
},
|
| 3950 |
+
{
|
| 3951 |
+
"epoch": 0.5456926811379076,
|
| 3952 |
+
"grad_norm": 0.62109375,
|
| 3953 |
+
"learning_rate": 0.001,
|
| 3954 |
+
"loss": 1.2768,
|
| 3955 |
+
"step": 7812
|
| 3956 |
+
},
|
| 3957 |
+
{
|
| 3958 |
+
"epoch": 0.5466706250109146,
|
| 3959 |
+
"grad_norm": 0.44140625,
|
| 3960 |
+
"learning_rate": 0.001,
|
| 3961 |
+
"loss": 1.265,
|
| 3962 |
+
"step": 7826
|
| 3963 |
+
},
|
| 3964 |
+
{
|
| 3965 |
+
"epoch": 0.5476485688839215,
|
| 3966 |
+
"grad_norm": 0.306640625,
|
| 3967 |
+
"learning_rate": 0.001,
|
| 3968 |
+
"loss": 1.2683,
|
| 3969 |
+
"step": 7840
|
| 3970 |
+
},
|
| 3971 |
+
{
|
| 3972 |
+
"epoch": 0.5486265127569285,
|
| 3973 |
+
"grad_norm": 0.28515625,
|
| 3974 |
+
"learning_rate": 0.001,
|
| 3975 |
+
"loss": 1.2809,
|
| 3976 |
+
"step": 7854
|
| 3977 |
+
},
|
| 3978 |
+
{
|
| 3979 |
+
"epoch": 0.5496044566299355,
|
| 3980 |
+
"grad_norm": 0.263671875,
|
| 3981 |
+
"learning_rate": 0.001,
|
| 3982 |
+
"loss": 1.2498,
|
| 3983 |
+
"step": 7868
|
| 3984 |
+
},
|
| 3985 |
+
{
|
| 3986 |
+
"epoch": 0.5505824005029426,
|
| 3987 |
+
"grad_norm": 0.333984375,
|
| 3988 |
+
"learning_rate": 0.001,
|
| 3989 |
+
"loss": 1.2632,
|
| 3990 |
+
"step": 7882
|
| 3991 |
+
},
|
| 3992 |
+
{
|
| 3993 |
+
"epoch": 0.5515603443759496,
|
| 3994 |
+
"grad_norm": 0.29296875,
|
| 3995 |
+
"learning_rate": 0.001,
|
| 3996 |
+
"loss": 1.2711,
|
| 3997 |
+
"step": 7896
|
| 3998 |
+
},
|
| 3999 |
+
{
|
| 4000 |
+
"epoch": 0.5525382882489566,
|
| 4001 |
+
"grad_norm": 0.373046875,
|
| 4002 |
+
"learning_rate": 0.001,
|
| 4003 |
+
"loss": 1.2813,
|
| 4004 |
+
"step": 7910
|
| 4005 |
+
},
|
| 4006 |
+
{
|
| 4007 |
+
"epoch": 0.5535162321219635,
|
| 4008 |
+
"grad_norm": 0.427734375,
|
| 4009 |
+
"learning_rate": 0.001,
|
| 4010 |
+
"loss": 1.2993,
|
| 4011 |
+
"step": 7924
|
| 4012 |
+
},
|
| 4013 |
+
{
|
| 4014 |
+
"epoch": 0.5544941759949705,
|
| 4015 |
+
"grad_norm": 0.373046875,
|
| 4016 |
+
"learning_rate": 0.001,
|
| 4017 |
+
"loss": 1.3001,
|
| 4018 |
+
"step": 7938
|
| 4019 |
+
},
|
| 4020 |
+
{
|
| 4021 |
+
"epoch": 0.5554721198679776,
|
| 4022 |
+
"grad_norm": 0.416015625,
|
| 4023 |
+
"learning_rate": 0.001,
|
| 4024 |
+
"loss": 1.2786,
|
| 4025 |
+
"step": 7952
|
| 4026 |
+
},
|
| 4027 |
+
{
|
| 4028 |
+
"epoch": 0.5564500637409846,
|
| 4029 |
+
"grad_norm": 0.298828125,
|
| 4030 |
+
"learning_rate": 0.001,
|
| 4031 |
+
"loss": 1.2976,
|
| 4032 |
+
"step": 7966
|
| 4033 |
+
},
|
| 4034 |
+
{
|
| 4035 |
+
"epoch": 0.5574280076139916,
|
| 4036 |
+
"grad_norm": 0.30078125,
|
| 4037 |
+
"learning_rate": 0.001,
|
| 4038 |
+
"loss": 1.286,
|
| 4039 |
+
"step": 7980
|
| 4040 |
+
},
|
| 4041 |
+
{
|
| 4042 |
+
"epoch": 0.5584059514869986,
|
| 4043 |
+
"grad_norm": 0.59765625,
|
| 4044 |
+
"learning_rate": 0.001,
|
| 4045 |
+
"loss": 1.282,
|
| 4046 |
+
"step": 7994
|
| 4047 |
+
},
|
| 4048 |
+
{
|
| 4049 |
+
"epoch": 0.5593838953600055,
|
| 4050 |
+
"grad_norm": 0.244140625,
|
| 4051 |
+
"learning_rate": 0.001,
|
| 4052 |
+
"loss": 1.2853,
|
| 4053 |
+
"step": 8008
|
| 4054 |
+
},
|
| 4055 |
+
{
|
| 4056 |
+
"epoch": 0.5603618392330126,
|
| 4057 |
+
"grad_norm": 0.2265625,
|
| 4058 |
+
"learning_rate": 0.001,
|
| 4059 |
+
"loss": 1.2572,
|
| 4060 |
+
"step": 8022
|
| 4061 |
+
},
|
| 4062 |
+
{
|
| 4063 |
+
"epoch": 0.5613397831060196,
|
| 4064 |
+
"grad_norm": 0.306640625,
|
| 4065 |
+
"learning_rate": 0.001,
|
| 4066 |
+
"loss": 1.2572,
|
| 4067 |
+
"step": 8036
|
| 4068 |
+
},
|
| 4069 |
+
{
|
| 4070 |
+
"epoch": 0.5623177269790266,
|
| 4071 |
+
"grad_norm": 0.63671875,
|
| 4072 |
+
"learning_rate": 0.001,
|
| 4073 |
+
"loss": 1.315,
|
| 4074 |
+
"step": 8050
|
| 4075 |
+
},
|
| 4076 |
+
{
|
| 4077 |
+
"epoch": 0.5632956708520336,
|
| 4078 |
+
"grad_norm": 0.30859375,
|
| 4079 |
+
"learning_rate": 0.001,
|
| 4080 |
+
"loss": 1.3007,
|
| 4081 |
+
"step": 8064
|
| 4082 |
+
},
|
| 4083 |
+
{
|
| 4084 |
+
"epoch": 0.5642736147250406,
|
| 4085 |
+
"grad_norm": 0.271484375,
|
| 4086 |
+
"learning_rate": 0.001,
|
| 4087 |
+
"loss": 1.2737,
|
| 4088 |
+
"step": 8078
|
| 4089 |
+
},
|
| 4090 |
+
{
|
| 4091 |
+
"epoch": 0.5652515585980477,
|
| 4092 |
+
"grad_norm": 0.24609375,
|
| 4093 |
+
"learning_rate": 0.001,
|
| 4094 |
+
"loss": 1.2766,
|
| 4095 |
+
"step": 8092
|
| 4096 |
+
},
|
| 4097 |
+
{
|
| 4098 |
+
"epoch": 0.5662295024710546,
|
| 4099 |
+
"grad_norm": 0.314453125,
|
| 4100 |
+
"learning_rate": 0.001,
|
| 4101 |
+
"loss": 1.3102,
|
| 4102 |
+
"step": 8106
|
| 4103 |
+
},
|
| 4104 |
+
{
|
| 4105 |
+
"epoch": 0.5672074463440616,
|
| 4106 |
+
"grad_norm": 0.328125,
|
| 4107 |
+
"learning_rate": 0.001,
|
| 4108 |
+
"loss": 1.3044,
|
| 4109 |
+
"step": 8120
|
| 4110 |
+
},
|
| 4111 |
+
{
|
| 4112 |
+
"epoch": 0.5681853902170686,
|
| 4113 |
+
"grad_norm": 0.380859375,
|
| 4114 |
+
"learning_rate": 0.001,
|
| 4115 |
+
"loss": 1.2612,
|
| 4116 |
+
"step": 8134
|
| 4117 |
+
},
|
| 4118 |
+
{
|
| 4119 |
+
"epoch": 0.5691633340900756,
|
| 4120 |
+
"grad_norm": 0.255859375,
|
| 4121 |
+
"learning_rate": 0.001,
|
| 4122 |
+
"loss": 1.2701,
|
| 4123 |
+
"step": 8148
|
| 4124 |
+
},
|
| 4125 |
+
{
|
| 4126 |
+
"epoch": 0.5701412779630827,
|
| 4127 |
+
"grad_norm": 0.21875,
|
| 4128 |
+
"learning_rate": 0.001,
|
| 4129 |
+
"loss": 1.2649,
|
| 4130 |
+
"step": 8162
|
| 4131 |
+
},
|
| 4132 |
+
{
|
| 4133 |
+
"epoch": 0.5711192218360897,
|
| 4134 |
+
"grad_norm": 0.2236328125,
|
| 4135 |
+
"learning_rate": 0.001,
|
| 4136 |
+
"loss": 1.2761,
|
| 4137 |
+
"step": 8176
|
| 4138 |
+
},
|
| 4139 |
+
{
|
| 4140 |
+
"epoch": 0.5720971657090966,
|
| 4141 |
+
"grad_norm": 0.29296875,
|
| 4142 |
+
"learning_rate": 0.001,
|
| 4143 |
+
"loss": 1.2668,
|
| 4144 |
+
"step": 8190
|
| 4145 |
+
},
|
| 4146 |
+
{
|
| 4147 |
+
"epoch": 0.5730751095821036,
|
| 4148 |
+
"grad_norm": 0.31640625,
|
| 4149 |
+
"learning_rate": 0.001,
|
| 4150 |
+
"loss": 1.2847,
|
| 4151 |
+
"step": 8204
|
| 4152 |
+
},
|
| 4153 |
+
{
|
| 4154 |
+
"epoch": 0.5740530534551106,
|
| 4155 |
+
"grad_norm": 0.3203125,
|
| 4156 |
+
"learning_rate": 0.001,
|
| 4157 |
+
"loss": 1.2722,
|
| 4158 |
+
"step": 8218
|
| 4159 |
+
},
|
| 4160 |
+
{
|
| 4161 |
+
"epoch": 0.5750309973281177,
|
| 4162 |
+
"grad_norm": 0.224609375,
|
| 4163 |
+
"learning_rate": 0.001,
|
| 4164 |
+
"loss": 1.253,
|
| 4165 |
+
"step": 8232
|
| 4166 |
+
},
|
| 4167 |
+
{
|
| 4168 |
+
"epoch": 0.5760089412011247,
|
| 4169 |
+
"grad_norm": 0.2890625,
|
| 4170 |
+
"learning_rate": 0.001,
|
| 4171 |
+
"loss": 1.2454,
|
| 4172 |
+
"step": 8246
|
| 4173 |
+
},
|
| 4174 |
+
{
|
| 4175 |
+
"epoch": 0.5769868850741317,
|
| 4176 |
+
"grad_norm": 0.283203125,
|
| 4177 |
+
"learning_rate": 0.001,
|
| 4178 |
+
"loss": 1.2558,
|
| 4179 |
+
"step": 8260
|
| 4180 |
+
},
|
| 4181 |
+
{
|
| 4182 |
+
"epoch": 0.5779648289471386,
|
| 4183 |
+
"grad_norm": 0.265625,
|
| 4184 |
+
"learning_rate": 0.001,
|
| 4185 |
+
"loss": 1.2765,
|
| 4186 |
+
"step": 8274
|
| 4187 |
+
},
|
| 4188 |
+
{
|
| 4189 |
+
"epoch": 0.5789427728201456,
|
| 4190 |
+
"grad_norm": 0.314453125,
|
| 4191 |
+
"learning_rate": 0.001,
|
| 4192 |
+
"loss": 1.289,
|
| 4193 |
+
"step": 8288
|
| 4194 |
+
},
|
| 4195 |
+
{
|
| 4196 |
+
"epoch": 0.5799207166931526,
|
| 4197 |
+
"grad_norm": 0.333984375,
|
| 4198 |
+
"learning_rate": 0.001,
|
| 4199 |
+
"loss": 1.2724,
|
| 4200 |
+
"step": 8302
|
| 4201 |
+
},
|
| 4202 |
+
{
|
| 4203 |
+
"epoch": 0.5808986605661597,
|
| 4204 |
+
"grad_norm": 0.44140625,
|
| 4205 |
+
"learning_rate": 0.001,
|
| 4206 |
+
"loss": 1.2753,
|
| 4207 |
+
"step": 8316
|
| 4208 |
+
},
|
| 4209 |
+
{
|
| 4210 |
+
"epoch": 0.5818766044391667,
|
| 4211 |
+
"grad_norm": 0.326171875,
|
| 4212 |
+
"learning_rate": 0.001,
|
| 4213 |
+
"loss": 1.2558,
|
| 4214 |
+
"step": 8330
|
| 4215 |
+
},
|
| 4216 |
+
{
|
| 4217 |
+
"epoch": 0.5828545483121736,
|
| 4218 |
+
"grad_norm": 0.271484375,
|
| 4219 |
+
"learning_rate": 0.001,
|
| 4220 |
+
"loss": 1.2697,
|
| 4221 |
+
"step": 8344
|
| 4222 |
+
},
|
| 4223 |
+
{
|
| 4224 |
+
"epoch": 0.5838324921851806,
|
| 4225 |
+
"grad_norm": 0.34375,
|
| 4226 |
+
"learning_rate": 0.001,
|
| 4227 |
+
"loss": 1.2685,
|
| 4228 |
+
"step": 8358
|
| 4229 |
+
},
|
| 4230 |
+
{
|
| 4231 |
+
"epoch": 0.5848104360581876,
|
| 4232 |
+
"grad_norm": 0.275390625,
|
| 4233 |
+
"learning_rate": 0.001,
|
| 4234 |
+
"loss": 1.2724,
|
| 4235 |
+
"step": 8372
|
| 4236 |
+
},
|
| 4237 |
+
{
|
| 4238 |
+
"epoch": 0.5857883799311947,
|
| 4239 |
+
"grad_norm": 0.2255859375,
|
| 4240 |
+
"learning_rate": 0.001,
|
| 4241 |
+
"loss": 1.2287,
|
| 4242 |
+
"step": 8386
|
| 4243 |
+
},
|
| 4244 |
+
{
|
| 4245 |
+
"epoch": 0.5867663238042017,
|
| 4246 |
+
"grad_norm": 0.212890625,
|
| 4247 |
+
"learning_rate": 0.001,
|
| 4248 |
+
"loss": 1.2363,
|
| 4249 |
+
"step": 8400
|
| 4250 |
+
},
|
| 4251 |
+
{
|
| 4252 |
+
"epoch": 0.5877442676772087,
|
| 4253 |
+
"grad_norm": 0.279296875,
|
| 4254 |
+
"learning_rate": 0.001,
|
| 4255 |
+
"loss": 1.2648,
|
| 4256 |
+
"step": 8414
|
| 4257 |
+
},
|
| 4258 |
+
{
|
| 4259 |
+
"epoch": 0.5887222115502156,
|
| 4260 |
+
"grad_norm": 0.427734375,
|
| 4261 |
+
"learning_rate": 0.001,
|
| 4262 |
+
"loss": 1.2949,
|
| 4263 |
+
"step": 8428
|
| 4264 |
+
},
|
| 4265 |
+
{
|
| 4266 |
+
"epoch": 0.5897001554232226,
|
| 4267 |
+
"grad_norm": 0.23828125,
|
| 4268 |
+
"learning_rate": 0.001,
|
| 4269 |
+
"loss": 1.2571,
|
| 4270 |
+
"step": 8442
|
| 4271 |
+
},
|
| 4272 |
+
{
|
| 4273 |
+
"epoch": 0.5906780992962297,
|
| 4274 |
+
"grad_norm": 0.349609375,
|
| 4275 |
+
"learning_rate": 0.001,
|
| 4276 |
+
"loss": 1.2831,
|
| 4277 |
+
"step": 8456
|
| 4278 |
+
},
|
| 4279 |
+
{
|
| 4280 |
+
"epoch": 0.5916560431692367,
|
| 4281 |
+
"grad_norm": 0.34765625,
|
| 4282 |
+
"learning_rate": 0.001,
|
| 4283 |
+
"loss": 1.2965,
|
| 4284 |
+
"step": 8470
|
| 4285 |
+
},
|
| 4286 |
+
{
|
| 4287 |
+
"epoch": 0.5926339870422437,
|
| 4288 |
+
"grad_norm": 0.412109375,
|
| 4289 |
+
"learning_rate": 0.001,
|
| 4290 |
+
"loss": 1.2685,
|
| 4291 |
+
"step": 8484
|
| 4292 |
+
},
|
| 4293 |
+
{
|
| 4294 |
+
"epoch": 0.5936119309152507,
|
| 4295 |
+
"grad_norm": 0.439453125,
|
| 4296 |
+
"learning_rate": 0.001,
|
| 4297 |
+
"loss": 1.2637,
|
| 4298 |
+
"step": 8498
|
| 4299 |
+
},
|
| 4300 |
+
{
|
| 4301 |
+
"epoch": 0.5945898747882576,
|
| 4302 |
+
"grad_norm": 0.3671875,
|
| 4303 |
+
"learning_rate": 0.001,
|
| 4304 |
+
"loss": 1.28,
|
| 4305 |
+
"step": 8512
|
| 4306 |
+
},
|
| 4307 |
+
{
|
| 4308 |
+
"epoch": 0.5955678186612647,
|
| 4309 |
+
"grad_norm": 0.43359375,
|
| 4310 |
+
"learning_rate": 0.001,
|
| 4311 |
+
"loss": 1.2636,
|
| 4312 |
+
"step": 8526
|
| 4313 |
+
},
|
| 4314 |
+
{
|
| 4315 |
+
"epoch": 0.5965457625342717,
|
| 4316 |
+
"grad_norm": 0.333984375,
|
| 4317 |
+
"learning_rate": 0.001,
|
| 4318 |
+
"loss": 1.251,
|
| 4319 |
+
"step": 8540
|
| 4320 |
+
},
|
| 4321 |
+
{
|
| 4322 |
+
"epoch": 0.5975237064072787,
|
| 4323 |
+
"grad_norm": 0.328125,
|
| 4324 |
+
"learning_rate": 0.001,
|
| 4325 |
+
"loss": 1.262,
|
| 4326 |
+
"step": 8554
|
| 4327 |
+
},
|
| 4328 |
+
{
|
| 4329 |
+
"epoch": 0.5985016502802857,
|
| 4330 |
+
"grad_norm": 0.365234375,
|
| 4331 |
+
"learning_rate": 0.001,
|
| 4332 |
+
"loss": 1.2696,
|
| 4333 |
+
"step": 8568
|
| 4334 |
+
},
|
| 4335 |
+
{
|
| 4336 |
+
"epoch": 0.5994795941532927,
|
| 4337 |
+
"grad_norm": 0.28515625,
|
| 4338 |
+
"learning_rate": 0.001,
|
| 4339 |
+
"loss": 1.2872,
|
| 4340 |
+
"step": 8582
|
| 4341 |
+
},
|
| 4342 |
+
{
|
| 4343 |
+
"epoch": 0.5997590066884375,
|
| 4344 |
+
"eval_loss": 1.661841869354248,
|
| 4345 |
+
"eval_runtime": 9.1193,
|
| 4346 |
+
"eval_samples_per_second": 109.657,
|
| 4347 |
+
"eval_steps_per_second": 1.426,
|
| 4348 |
+
"step": 8586
|
| 4349 |
+
},
|
| 4350 |
+
{
|
| 4351 |
+
"epoch": 0.6004575380262998,
|
| 4352 |
+
"grad_norm": 0.2451171875,
|
| 4353 |
+
"learning_rate": 0.001,
|
| 4354 |
+
"loss": 1.2767,
|
| 4355 |
+
"step": 8596
|
| 4356 |
+
},
|
| 4357 |
+
{
|
| 4358 |
+
"epoch": 0.6014354818993067,
|
| 4359 |
+
"grad_norm": 0.33984375,
|
| 4360 |
+
"learning_rate": 0.001,
|
| 4361 |
+
"loss": 1.2623,
|
| 4362 |
+
"step": 8610
|
| 4363 |
+
},
|
| 4364 |
+
{
|
| 4365 |
+
"epoch": 0.6024134257723137,
|
| 4366 |
+
"grad_norm": 0.26953125,
|
| 4367 |
+
"learning_rate": 0.001,
|
| 4368 |
+
"loss": 1.2617,
|
| 4369 |
+
"step": 8624
|
| 4370 |
+
},
|
| 4371 |
+
{
|
| 4372 |
+
"epoch": 0.6033913696453207,
|
| 4373 |
+
"grad_norm": 0.25390625,
|
| 4374 |
+
"learning_rate": 0.001,
|
| 4375 |
+
"loss": 1.2514,
|
| 4376 |
+
"step": 8638
|
| 4377 |
+
},
|
| 4378 |
+
{
|
| 4379 |
+
"epoch": 0.6043693135183277,
|
| 4380 |
+
"grad_norm": 0.255859375,
|
| 4381 |
+
"learning_rate": 0.001,
|
| 4382 |
+
"loss": 1.2664,
|
| 4383 |
+
"step": 8652
|
| 4384 |
+
},
|
| 4385 |
+
{
|
| 4386 |
+
"epoch": 0.6053472573913348,
|
| 4387 |
+
"grad_norm": 0.357421875,
|
| 4388 |
+
"learning_rate": 0.001,
|
| 4389 |
+
"loss": 1.2421,
|
| 4390 |
+
"step": 8666
|
| 4391 |
+
},
|
| 4392 |
+
{
|
| 4393 |
+
"epoch": 0.6063252012643418,
|
| 4394 |
+
"grad_norm": 0.263671875,
|
| 4395 |
+
"learning_rate": 0.001,
|
| 4396 |
+
"loss": 1.2386,
|
| 4397 |
+
"step": 8680
|
| 4398 |
+
},
|
| 4399 |
+
{
|
| 4400 |
+
"epoch": 0.6073031451373487,
|
| 4401 |
+
"grad_norm": 0.259765625,
|
| 4402 |
+
"learning_rate": 0.001,
|
| 4403 |
+
"loss": 1.2601,
|
| 4404 |
+
"step": 8694
|
| 4405 |
+
},
|
| 4406 |
+
{
|
| 4407 |
+
"epoch": 0.6082810890103557,
|
| 4408 |
+
"grad_norm": 0.94921875,
|
| 4409 |
+
"learning_rate": 0.001,
|
| 4410 |
+
"loss": 1.2715,
|
| 4411 |
+
"step": 8708
|
| 4412 |
+
},
|
| 4413 |
+
{
|
| 4414 |
+
"epoch": 0.6092590328833627,
|
| 4415 |
+
"grad_norm": 0.43359375,
|
| 4416 |
+
"learning_rate": 0.001,
|
| 4417 |
+
"loss": 1.2848,
|
| 4418 |
+
"step": 8722
|
| 4419 |
+
},
|
| 4420 |
+
{
|
| 4421 |
+
"epoch": 0.6102369767563698,
|
| 4422 |
+
"grad_norm": 0.34375,
|
| 4423 |
+
"learning_rate": 0.001,
|
| 4424 |
+
"loss": 1.2632,
|
| 4425 |
+
"step": 8736
|
| 4426 |
+
},
|
| 4427 |
+
{
|
| 4428 |
+
"epoch": 0.6112149206293768,
|
| 4429 |
+
"grad_norm": 0.283203125,
|
| 4430 |
+
"learning_rate": 0.001,
|
| 4431 |
+
"loss": 1.2912,
|
| 4432 |
+
"step": 8750
|
| 4433 |
+
},
|
| 4434 |
+
{
|
| 4435 |
+
"epoch": 0.6121928645023837,
|
| 4436 |
+
"grad_norm": 0.388671875,
|
| 4437 |
+
"learning_rate": 0.001,
|
| 4438 |
+
"loss": 1.2613,
|
| 4439 |
+
"step": 8764
|
| 4440 |
+
},
|
| 4441 |
+
{
|
| 4442 |
+
"epoch": 0.6131708083753907,
|
| 4443 |
+
"grad_norm": 0.27734375,
|
| 4444 |
+
"learning_rate": 0.001,
|
| 4445 |
+
"loss": 1.2357,
|
| 4446 |
+
"step": 8778
|
| 4447 |
+
},
|
| 4448 |
+
{
|
| 4449 |
+
"epoch": 0.6141487522483977,
|
| 4450 |
+
"grad_norm": 0.30859375,
|
| 4451 |
+
"learning_rate": 0.001,
|
| 4452 |
+
"loss": 1.2541,
|
| 4453 |
+
"step": 8792
|
| 4454 |
+
},
|
| 4455 |
+
{
|
| 4456 |
+
"epoch": 0.6151266961214047,
|
| 4457 |
+
"grad_norm": 0.32421875,
|
| 4458 |
+
"learning_rate": 0.001,
|
| 4459 |
+
"loss": 1.2746,
|
| 4460 |
+
"step": 8806
|
| 4461 |
+
},
|
| 4462 |
+
{
|
| 4463 |
+
"epoch": 0.6161046399944118,
|
| 4464 |
+
"grad_norm": 0.271484375,
|
| 4465 |
+
"learning_rate": 0.001,
|
| 4466 |
+
"loss": 1.2445,
|
| 4467 |
+
"step": 8820
|
| 4468 |
+
},
|
| 4469 |
+
{
|
| 4470 |
+
"epoch": 0.6170825838674188,
|
| 4471 |
+
"grad_norm": 0.255859375,
|
| 4472 |
+
"learning_rate": 0.001,
|
| 4473 |
+
"loss": 1.2854,
|
| 4474 |
+
"step": 8834
|
| 4475 |
+
},
|
| 4476 |
+
{
|
| 4477 |
+
"epoch": 0.6180605277404257,
|
| 4478 |
+
"grad_norm": 0.54296875,
|
| 4479 |
+
"learning_rate": 0.001,
|
| 4480 |
+
"loss": 1.2746,
|
| 4481 |
+
"step": 8848
|
| 4482 |
+
},
|
| 4483 |
+
{
|
| 4484 |
+
"epoch": 0.6190384716134327,
|
| 4485 |
+
"grad_norm": 0.35546875,
|
| 4486 |
+
"learning_rate": 0.001,
|
| 4487 |
+
"loss": 1.2837,
|
| 4488 |
+
"step": 8862
|
| 4489 |
+
},
|
| 4490 |
+
{
|
| 4491 |
+
"epoch": 0.6200164154864397,
|
| 4492 |
+
"grad_norm": 0.341796875,
|
| 4493 |
+
"learning_rate": 0.001,
|
| 4494 |
+
"loss": 1.2682,
|
| 4495 |
+
"step": 8876
|
| 4496 |
+
},
|
| 4497 |
+
{
|
| 4498 |
+
"epoch": 0.6209943593594468,
|
| 4499 |
+
"grad_norm": 0.51953125,
|
| 4500 |
+
"learning_rate": 0.001,
|
| 4501 |
+
"loss": 1.2751,
|
| 4502 |
+
"step": 8890
|
| 4503 |
+
},
|
| 4504 |
+
{
|
| 4505 |
+
"epoch": 0.6219723032324538,
|
| 4506 |
+
"grad_norm": 0.41015625,
|
| 4507 |
+
"learning_rate": 0.001,
|
| 4508 |
+
"loss": 1.2666,
|
| 4509 |
+
"step": 8904
|
| 4510 |
+
},
|
| 4511 |
+
{
|
| 4512 |
+
"epoch": 0.6229502471054608,
|
| 4513 |
+
"grad_norm": 0.4140625,
|
| 4514 |
+
"learning_rate": 0.001,
|
| 4515 |
+
"loss": 1.2618,
|
| 4516 |
+
"step": 8918
|
| 4517 |
+
},
|
| 4518 |
+
{
|
| 4519 |
+
"epoch": 0.6239281909784677,
|
| 4520 |
+
"grad_norm": 0.267578125,
|
| 4521 |
+
"learning_rate": 0.001,
|
| 4522 |
+
"loss": 1.2721,
|
| 4523 |
+
"step": 8932
|
| 4524 |
+
},
|
| 4525 |
+
{
|
| 4526 |
+
"epoch": 0.6249061348514747,
|
| 4527 |
+
"grad_norm": 0.3515625,
|
| 4528 |
+
"learning_rate": 0.001,
|
| 4529 |
+
"loss": 1.2528,
|
| 4530 |
+
"step": 8946
|
| 4531 |
+
},
|
| 4532 |
+
{
|
| 4533 |
+
"epoch": 0.6258840787244818,
|
| 4534 |
+
"grad_norm": 0.34375,
|
| 4535 |
+
"learning_rate": 0.001,
|
| 4536 |
+
"loss": 1.2771,
|
| 4537 |
+
"step": 8960
|
| 4538 |
+
},
|
| 4539 |
+
{
|
| 4540 |
+
"epoch": 0.6268620225974888,
|
| 4541 |
+
"grad_norm": 0.275390625,
|
| 4542 |
+
"learning_rate": 0.001,
|
| 4543 |
+
"loss": 1.2751,
|
| 4544 |
+
"step": 8974
|
| 4545 |
+
},
|
| 4546 |
+
{
|
| 4547 |
+
"epoch": 0.6278399664704958,
|
| 4548 |
+
"grad_norm": 0.28515625,
|
| 4549 |
+
"learning_rate": 0.001,
|
| 4550 |
+
"loss": 1.2749,
|
| 4551 |
+
"step": 8988
|
| 4552 |
+
},
|
| 4553 |
+
{
|
| 4554 |
+
"epoch": 0.6288179103435028,
|
| 4555 |
+
"grad_norm": 0.27734375,
|
| 4556 |
+
"learning_rate": 0.001,
|
| 4557 |
+
"loss": 1.2851,
|
| 4558 |
+
"step": 9002
|
| 4559 |
+
},
|
| 4560 |
+
{
|
| 4561 |
+
"epoch": 0.6297958542165097,
|
| 4562 |
+
"grad_norm": 0.23828125,
|
| 4563 |
+
"learning_rate": 0.001,
|
| 4564 |
+
"loss": 1.2529,
|
| 4565 |
+
"step": 9016
|
| 4566 |
+
},
|
| 4567 |
+
{
|
| 4568 |
+
"epoch": 0.6307737980895168,
|
| 4569 |
+
"grad_norm": 0.2890625,
|
| 4570 |
+
"learning_rate": 0.001,
|
| 4571 |
+
"loss": 1.2673,
|
| 4572 |
+
"step": 9030
|
| 4573 |
+
},
|
| 4574 |
+
{
|
| 4575 |
+
"epoch": 0.6317517419625238,
|
| 4576 |
+
"grad_norm": 0.259765625,
|
| 4577 |
+
"learning_rate": 0.001,
|
| 4578 |
+
"loss": 1.2746,
|
| 4579 |
+
"step": 9044
|
| 4580 |
+
},
|
| 4581 |
+
{
|
| 4582 |
+
"epoch": 0.6327296858355308,
|
| 4583 |
+
"grad_norm": 0.380859375,
|
| 4584 |
+
"learning_rate": 0.001,
|
| 4585 |
+
"loss": 1.2649,
|
| 4586 |
+
"step": 9058
|
| 4587 |
+
},
|
| 4588 |
+
{
|
| 4589 |
+
"epoch": 0.6337076297085378,
|
| 4590 |
+
"grad_norm": 0.318359375,
|
| 4591 |
+
"learning_rate": 0.001,
|
| 4592 |
+
"loss": 1.2849,
|
| 4593 |
+
"step": 9072
|
| 4594 |
+
},
|
| 4595 |
+
{
|
| 4596 |
+
"epoch": 0.6346855735815448,
|
| 4597 |
+
"grad_norm": 0.30078125,
|
| 4598 |
+
"learning_rate": 0.001,
|
| 4599 |
+
"loss": 1.243,
|
| 4600 |
+
"step": 9086
|
| 4601 |
+
},
|
| 4602 |
+
{
|
| 4603 |
+
"epoch": 0.6356635174545519,
|
| 4604 |
+
"grad_norm": 0.375,
|
| 4605 |
+
"learning_rate": 0.001,
|
| 4606 |
+
"loss": 1.2641,
|
| 4607 |
+
"step": 9100
|
| 4608 |
+
},
|
| 4609 |
+
{
|
| 4610 |
+
"epoch": 0.6366414613275588,
|
| 4611 |
+
"grad_norm": 0.361328125,
|
| 4612 |
+
"learning_rate": 0.001,
|
| 4613 |
+
"loss": 1.2554,
|
| 4614 |
+
"step": 9114
|
| 4615 |
+
},
|
| 4616 |
+
{
|
| 4617 |
+
"epoch": 0.6376194052005658,
|
| 4618 |
+
"grad_norm": 0.396484375,
|
| 4619 |
+
"learning_rate": 0.001,
|
| 4620 |
+
"loss": 1.2396,
|
| 4621 |
+
"step": 9128
|
| 4622 |
+
},
|
| 4623 |
+
{
|
| 4624 |
+
"epoch": 0.6385973490735728,
|
| 4625 |
+
"grad_norm": 0.263671875,
|
| 4626 |
+
"learning_rate": 0.001,
|
| 4627 |
+
"loss": 1.2508,
|
| 4628 |
+
"step": 9142
|
| 4629 |
+
},
|
| 4630 |
+
{
|
| 4631 |
+
"epoch": 0.6395752929465798,
|
| 4632 |
+
"grad_norm": 0.33984375,
|
| 4633 |
+
"learning_rate": 0.001,
|
| 4634 |
+
"loss": 1.2772,
|
| 4635 |
+
"step": 9156
|
| 4636 |
+
},
|
| 4637 |
+
{
|
| 4638 |
+
"epoch": 0.6405532368195869,
|
| 4639 |
+
"grad_norm": 0.53515625,
|
| 4640 |
+
"learning_rate": 0.001,
|
| 4641 |
+
"loss": 1.2453,
|
| 4642 |
+
"step": 9170
|
| 4643 |
+
},
|
| 4644 |
+
{
|
| 4645 |
+
"epoch": 0.6415311806925938,
|
| 4646 |
+
"grad_norm": 0.2099609375,
|
| 4647 |
+
"learning_rate": 0.001,
|
| 4648 |
+
"loss": 1.2764,
|
| 4649 |
+
"step": 9184
|
| 4650 |
+
},
|
| 4651 |
+
{
|
| 4652 |
+
"epoch": 0.6425091245656008,
|
| 4653 |
+
"grad_norm": 0.2333984375,
|
| 4654 |
+
"learning_rate": 0.001,
|
| 4655 |
+
"loss": 1.251,
|
| 4656 |
+
"step": 9198
|
| 4657 |
+
},
|
| 4658 |
+
{
|
| 4659 |
+
"epoch": 0.6434870684386078,
|
| 4660 |
+
"grad_norm": 0.35546875,
|
| 4661 |
+
"learning_rate": 0.001,
|
| 4662 |
+
"loss": 1.2855,
|
| 4663 |
+
"step": 9212
|
| 4664 |
+
},
|
| 4665 |
+
{
|
| 4666 |
+
"epoch": 0.6444650123116148,
|
| 4667 |
+
"grad_norm": 1.1953125,
|
| 4668 |
+
"learning_rate": 0.001,
|
| 4669 |
+
"loss": 1.3198,
|
| 4670 |
+
"step": 9226
|
| 4671 |
+
},
|
| 4672 |
+
{
|
| 4673 |
+
"epoch": 0.6454429561846219,
|
| 4674 |
+
"grad_norm": 0.427734375,
|
| 4675 |
+
"learning_rate": 0.001,
|
| 4676 |
+
"loss": 1.2773,
|
| 4677 |
+
"step": 9240
|
| 4678 |
+
},
|
| 4679 |
+
{
|
| 4680 |
+
"epoch": 0.6464209000576289,
|
| 4681 |
+
"grad_norm": 0.5703125,
|
| 4682 |
+
"learning_rate": 0.001,
|
| 4683 |
+
"loss": 1.2786,
|
| 4684 |
+
"step": 9254
|
| 4685 |
+
},
|
| 4686 |
+
{
|
| 4687 |
+
"epoch": 0.6473988439306358,
|
| 4688 |
+
"grad_norm": 0.3125,
|
| 4689 |
+
"learning_rate": 0.001,
|
| 4690 |
+
"loss": 1.2389,
|
| 4691 |
+
"step": 9268
|
| 4692 |
+
},
|
| 4693 |
+
{
|
| 4694 |
+
"epoch": 0.6483767878036428,
|
| 4695 |
+
"grad_norm": 0.36328125,
|
| 4696 |
+
"learning_rate": 0.001,
|
| 4697 |
+
"loss": 1.2587,
|
| 4698 |
+
"step": 9282
|
| 4699 |
+
},
|
| 4700 |
+
{
|
| 4701 |
+
"epoch": 0.6493547316766498,
|
| 4702 |
+
"grad_norm": 0.380859375,
|
| 4703 |
+
"learning_rate": 0.001,
|
| 4704 |
+
"loss": 1.2806,
|
| 4705 |
+
"step": 9296
|
| 4706 |
+
},
|
| 4707 |
+
{
|
| 4708 |
+
"epoch": 0.6503326755496568,
|
| 4709 |
+
"grad_norm": 0.251953125,
|
| 4710 |
+
"learning_rate": 0.001,
|
| 4711 |
+
"loss": 1.2307,
|
| 4712 |
+
"step": 9310
|
| 4713 |
+
},
|
| 4714 |
+
{
|
| 4715 |
+
"epoch": 0.6513106194226639,
|
| 4716 |
+
"grad_norm": 0.263671875,
|
| 4717 |
+
"learning_rate": 0.001,
|
| 4718 |
+
"loss": 1.2657,
|
| 4719 |
+
"step": 9324
|
| 4720 |
+
},
|
| 4721 |
+
{
|
| 4722 |
+
"epoch": 0.6522885632956709,
|
| 4723 |
+
"grad_norm": 0.263671875,
|
| 4724 |
+
"learning_rate": 0.001,
|
| 4725 |
+
"loss": 1.2605,
|
| 4726 |
+
"step": 9338
|
| 4727 |
+
},
|
| 4728 |
+
{
|
| 4729 |
+
"epoch": 0.6532665071686778,
|
| 4730 |
+
"grad_norm": 0.2412109375,
|
| 4731 |
+
"learning_rate": 0.001,
|
| 4732 |
+
"loss": 1.2538,
|
| 4733 |
+
"step": 9352
|
| 4734 |
+
},
|
| 4735 |
+
{
|
| 4736 |
+
"epoch": 0.6542444510416848,
|
| 4737 |
+
"grad_norm": 0.3203125,
|
| 4738 |
+
"learning_rate": 0.001,
|
| 4739 |
+
"loss": 1.2633,
|
| 4740 |
+
"step": 9366
|
| 4741 |
+
},
|
| 4742 |
+
{
|
| 4743 |
+
"epoch": 0.6552223949146918,
|
| 4744 |
+
"grad_norm": 0.31640625,
|
| 4745 |
+
"learning_rate": 0.001,
|
| 4746 |
+
"loss": 1.2582,
|
| 4747 |
+
"step": 9380
|
| 4748 |
+
},
|
| 4749 |
+
{
|
| 4750 |
+
"epoch": 0.6562003387876989,
|
| 4751 |
+
"grad_norm": 0.328125,
|
| 4752 |
+
"learning_rate": 0.001,
|
| 4753 |
+
"loss": 1.2515,
|
| 4754 |
+
"step": 9394
|
| 4755 |
+
},
|
| 4756 |
+
{
|
| 4757 |
+
"epoch": 0.6571782826607059,
|
| 4758 |
+
"grad_norm": 0.33984375,
|
| 4759 |
+
"learning_rate": 0.001,
|
| 4760 |
+
"loss": 1.2679,
|
| 4761 |
+
"step": 9408
|
| 4762 |
+
},
|
| 4763 |
+
{
|
| 4764 |
+
"epoch": 0.6581562265337129,
|
| 4765 |
+
"grad_norm": 0.3828125,
|
| 4766 |
+
"learning_rate": 0.001,
|
| 4767 |
+
"loss": 1.2539,
|
| 4768 |
+
"step": 9422
|
| 4769 |
+
},
|
| 4770 |
+
{
|
| 4771 |
+
"epoch": 0.6591341704067198,
|
| 4772 |
+
"grad_norm": 0.26171875,
|
| 4773 |
+
"learning_rate": 0.001,
|
| 4774 |
+
"loss": 1.2632,
|
| 4775 |
+
"step": 9436
|
| 4776 |
+
},
|
| 4777 |
+
{
|
| 4778 |
+
"epoch": 0.6601121142797268,
|
| 4779 |
+
"grad_norm": 0.3203125,
|
| 4780 |
+
"learning_rate": 0.001,
|
| 4781 |
+
"loss": 1.2946,
|
| 4782 |
+
"step": 9450
|
| 4783 |
+
},
|
| 4784 |
+
{
|
| 4785 |
+
"epoch": 0.6610900581527339,
|
| 4786 |
+
"grad_norm": 0.38671875,
|
| 4787 |
+
"learning_rate": 0.001,
|
| 4788 |
+
"loss": 1.2691,
|
| 4789 |
+
"step": 9464
|
| 4790 |
+
},
|
| 4791 |
+
{
|
| 4792 |
+
"epoch": 0.6620680020257409,
|
| 4793 |
+
"grad_norm": 0.2890625,
|
| 4794 |
+
"learning_rate": 0.001,
|
| 4795 |
+
"loss": 1.246,
|
| 4796 |
+
"step": 9478
|
| 4797 |
+
},
|
| 4798 |
+
{
|
| 4799 |
+
"epoch": 0.6630459458987479,
|
| 4800 |
+
"grad_norm": 0.431640625,
|
| 4801 |
+
"learning_rate": 0.001,
|
| 4802 |
+
"loss": 1.2606,
|
| 4803 |
+
"step": 9492
|
| 4804 |
+
},
|
| 4805 |
+
{
|
| 4806 |
+
"epoch": 0.6640238897717549,
|
| 4807 |
+
"grad_norm": 0.8671875,
|
| 4808 |
+
"learning_rate": 0.001,
|
| 4809 |
+
"loss": 1.2782,
|
| 4810 |
+
"step": 9506
|
| 4811 |
+
},
|
| 4812 |
+
{
|
| 4813 |
+
"epoch": 0.6650018336447618,
|
| 4814 |
+
"grad_norm": 0.4375,
|
| 4815 |
+
"learning_rate": 0.001,
|
| 4816 |
+
"loss": 1.2687,
|
| 4817 |
+
"step": 9520
|
| 4818 |
+
},
|
| 4819 |
+
{
|
| 4820 |
+
"epoch": 0.6659797775177689,
|
| 4821 |
+
"grad_norm": 0.37109375,
|
| 4822 |
+
"learning_rate": 0.001,
|
| 4823 |
+
"loss": 1.2778,
|
| 4824 |
+
"step": 9534
|
| 4825 |
+
},
|
| 4826 |
+
{
|
| 4827 |
+
"epoch": 0.6669577213907759,
|
| 4828 |
+
"grad_norm": 0.2490234375,
|
| 4829 |
+
"learning_rate": 0.001,
|
| 4830 |
+
"loss": 1.2544,
|
| 4831 |
+
"step": 9548
|
| 4832 |
+
},
|
| 4833 |
+
{
|
| 4834 |
+
"epoch": 0.6679356652637829,
|
| 4835 |
+
"grad_norm": 0.291015625,
|
| 4836 |
+
"learning_rate": 0.001,
|
| 4837 |
+
"loss": 1.246,
|
| 4838 |
+
"step": 9562
|
| 4839 |
+
},
|
| 4840 |
+
{
|
| 4841 |
+
"epoch": 0.6689136091367899,
|
| 4842 |
+
"grad_norm": 0.287109375,
|
| 4843 |
+
"learning_rate": 0.001,
|
| 4844 |
+
"loss": 1.2577,
|
| 4845 |
+
"step": 9576
|
| 4846 |
+
},
|
| 4847 |
+
{
|
| 4848 |
+
"epoch": 0.6698915530097969,
|
| 4849 |
+
"grad_norm": 0.328125,
|
| 4850 |
+
"learning_rate": 0.001,
|
| 4851 |
+
"loss": 1.2639,
|
| 4852 |
+
"step": 9590
|
| 4853 |
+
},
|
| 4854 |
+
{
|
| 4855 |
+
"epoch": 0.670869496882804,
|
| 4856 |
+
"grad_norm": 0.271484375,
|
| 4857 |
+
"learning_rate": 0.001,
|
| 4858 |
+
"loss": 1.2493,
|
| 4859 |
+
"step": 9604
|
| 4860 |
+
},
|
| 4861 |
+
{
|
| 4862 |
+
"epoch": 0.6718474407558109,
|
| 4863 |
+
"grad_norm": 0.2119140625,
|
| 4864 |
+
"learning_rate": 0.001,
|
| 4865 |
+
"loss": 1.2586,
|
| 4866 |
+
"step": 9618
|
| 4867 |
+
},
|
| 4868 |
+
{
|
| 4869 |
+
"epoch": 0.6728253846288179,
|
| 4870 |
+
"grad_norm": 0.240234375,
|
| 4871 |
+
"learning_rate": 0.001,
|
| 4872 |
+
"loss": 1.2719,
|
| 4873 |
+
"step": 9632
|
| 4874 |
+
},
|
| 4875 |
+
{
|
| 4876 |
+
"epoch": 0.6738033285018249,
|
| 4877 |
+
"grad_norm": 0.294921875,
|
| 4878 |
+
"learning_rate": 0.001,
|
| 4879 |
+
"loss": 1.2272,
|
| 4880 |
+
"step": 9646
|
| 4881 |
+
},
|
| 4882 |
+
{
|
| 4883 |
+
"epoch": 0.6747812723748319,
|
| 4884 |
+
"grad_norm": 0.423828125,
|
| 4885 |
+
"learning_rate": 0.001,
|
| 4886 |
+
"loss": 1.2667,
|
| 4887 |
+
"step": 9660
|
| 4888 |
+
},
|
| 4889 |
+
{
|
| 4890 |
+
"epoch": 0.675759216247839,
|
| 4891 |
+
"grad_norm": 0.30859375,
|
| 4892 |
+
"learning_rate": 0.001,
|
| 4893 |
+
"loss": 1.2567,
|
| 4894 |
+
"step": 9674
|
| 4895 |
+
},
|
| 4896 |
+
{
|
| 4897 |
+
"epoch": 0.676737160120846,
|
| 4898 |
+
"grad_norm": 0.21875,
|
| 4899 |
+
"learning_rate": 0.001,
|
| 4900 |
+
"loss": 1.2403,
|
| 4901 |
+
"step": 9688
|
| 4902 |
+
},
|
| 4903 |
+
{
|
| 4904 |
+
"epoch": 0.6777151039938529,
|
| 4905 |
+
"grad_norm": 0.2490234375,
|
| 4906 |
+
"learning_rate": 0.001,
|
| 4907 |
+
"loss": 1.2642,
|
| 4908 |
+
"step": 9702
|
| 4909 |
+
},
|
| 4910 |
+
{
|
| 4911 |
+
"epoch": 0.6786930478668599,
|
| 4912 |
+
"grad_norm": 0.248046875,
|
| 4913 |
+
"learning_rate": 0.001,
|
| 4914 |
+
"loss": 1.2123,
|
| 4915 |
+
"step": 9716
|
| 4916 |
+
},
|
| 4917 |
+
{
|
| 4918 |
+
"epoch": 0.6796709917398669,
|
| 4919 |
+
"grad_norm": 0.271484375,
|
| 4920 |
+
"learning_rate": 0.001,
|
| 4921 |
+
"loss": 1.2413,
|
| 4922 |
+
"step": 9730
|
| 4923 |
+
},
|
| 4924 |
+
{
|
| 4925 |
+
"epoch": 0.680648935612874,
|
| 4926 |
+
"grad_norm": 0.259765625,
|
| 4927 |
+
"learning_rate": 0.001,
|
| 4928 |
+
"loss": 1.2442,
|
| 4929 |
+
"step": 9744
|
| 4930 |
+
},
|
| 4931 |
+
{
|
| 4932 |
+
"epoch": 0.681626879485881,
|
| 4933 |
+
"grad_norm": 0.271484375,
|
| 4934 |
+
"learning_rate": 0.001,
|
| 4935 |
+
"loss": 1.2298,
|
| 4936 |
+
"step": 9758
|
| 4937 |
+
},
|
| 4938 |
+
{
|
| 4939 |
+
"epoch": 0.682604823358888,
|
| 4940 |
+
"grad_norm": 0.232421875,
|
| 4941 |
+
"learning_rate": 0.001,
|
| 4942 |
+
"loss": 1.2421,
|
| 4943 |
+
"step": 9772
|
| 4944 |
+
},
|
| 4945 |
+
{
|
| 4946 |
+
"epoch": 0.6835827672318949,
|
| 4947 |
+
"grad_norm": 0.302734375,
|
| 4948 |
+
"learning_rate": 0.001,
|
| 4949 |
+
"loss": 1.2725,
|
| 4950 |
+
"step": 9786
|
| 4951 |
+
},
|
| 4952 |
+
{
|
| 4953 |
+
"epoch": 0.6845607111049019,
|
| 4954 |
+
"grad_norm": 0.267578125,
|
| 4955 |
+
"learning_rate": 0.001,
|
| 4956 |
+
"loss": 1.2417,
|
| 4957 |
+
"step": 9800
|
| 4958 |
+
},
|
| 4959 |
+
{
|
| 4960 |
+
"epoch": 0.685538654977909,
|
| 4961 |
+
"grad_norm": 0.232421875,
|
| 4962 |
+
"learning_rate": 0.001,
|
| 4963 |
+
"loss": 1.2526,
|
| 4964 |
+
"step": 9814
|
| 4965 |
+
},
|
| 4966 |
+
{
|
| 4967 |
+
"epoch": 0.686516598850916,
|
| 4968 |
+
"grad_norm": 0.265625,
|
| 4969 |
+
"learning_rate": 0.001,
|
| 4970 |
+
"loss": 1.2352,
|
| 4971 |
+
"step": 9828
|
| 4972 |
+
},
|
| 4973 |
+
{
|
| 4974 |
+
"epoch": 0.687494542723923,
|
| 4975 |
+
"grad_norm": 0.361328125,
|
| 4976 |
+
"learning_rate": 0.001,
|
| 4977 |
+
"loss": 1.2653,
|
| 4978 |
+
"step": 9842
|
| 4979 |
+
},
|
| 4980 |
+
{
|
| 4981 |
+
"epoch": 0.68847248659693,
|
| 4982 |
+
"grad_norm": 0.328125,
|
| 4983 |
+
"learning_rate": 0.001,
|
| 4984 |
+
"loss": 1.2569,
|
| 4985 |
+
"step": 9856
|
| 4986 |
+
},
|
| 4987 |
+
{
|
| 4988 |
+
"epoch": 0.6894504304699369,
|
| 4989 |
+
"grad_norm": 0.26171875,
|
| 4990 |
+
"learning_rate": 0.001,
|
| 4991 |
+
"loss": 1.248,
|
| 4992 |
+
"step": 9870
|
| 4993 |
+
},
|
| 4994 |
+
{
|
| 4995 |
+
"epoch": 0.6904283743429439,
|
| 4996 |
+
"grad_norm": 0.265625,
|
| 4997 |
+
"learning_rate": 0.001,
|
| 4998 |
+
"loss": 1.2864,
|
| 4999 |
+
"step": 9884
|
| 5000 |
+
},
|
| 5001 |
+
{
|
| 5002 |
+
"epoch": 0.691406318215951,
|
| 5003 |
+
"grad_norm": 0.345703125,
|
| 5004 |
+
"learning_rate": 0.001,
|
| 5005 |
+
"loss": 1.2663,
|
| 5006 |
+
"step": 9898
|
| 5007 |
+
},
|
| 5008 |
+
{
|
| 5009 |
+
"epoch": 0.692384262088958,
|
| 5010 |
+
"grad_norm": 0.36328125,
|
| 5011 |
+
"learning_rate": 0.001,
|
| 5012 |
+
"loss": 1.2534,
|
| 5013 |
+
"step": 9912
|
| 5014 |
+
},
|
| 5015 |
+
{
|
| 5016 |
+
"epoch": 0.693362205961965,
|
| 5017 |
+
"grad_norm": 0.333984375,
|
| 5018 |
+
"learning_rate": 0.001,
|
| 5019 |
+
"loss": 1.2439,
|
| 5020 |
+
"step": 9926
|
| 5021 |
+
},
|
| 5022 |
+
{
|
| 5023 |
+
"epoch": 0.6943401498349719,
|
| 5024 |
+
"grad_norm": 0.298828125,
|
| 5025 |
+
"learning_rate": 0.001,
|
| 5026 |
+
"loss": 1.2592,
|
| 5027 |
+
"step": 9940
|
| 5028 |
+
},
|
| 5029 |
+
{
|
| 5030 |
+
"epoch": 0.6953180937079789,
|
| 5031 |
+
"grad_norm": 0.3046875,
|
| 5032 |
+
"learning_rate": 0.001,
|
| 5033 |
+
"loss": 1.2655,
|
| 5034 |
+
"step": 9954
|
| 5035 |
+
},
|
| 5036 |
+
{
|
| 5037 |
+
"epoch": 0.696296037580986,
|
| 5038 |
+
"grad_norm": 0.265625,
|
| 5039 |
+
"learning_rate": 0.001,
|
| 5040 |
+
"loss": 1.2569,
|
| 5041 |
+
"step": 9968
|
| 5042 |
+
},
|
| 5043 |
+
{
|
| 5044 |
+
"epoch": 0.697273981453993,
|
| 5045 |
+
"grad_norm": 0.376953125,
|
| 5046 |
+
"learning_rate": 0.001,
|
| 5047 |
+
"loss": 1.2663,
|
| 5048 |
+
"step": 9982
|
| 5049 |
+
},
|
| 5050 |
+
{
|
| 5051 |
+
"epoch": 0.698251925327,
|
| 5052 |
+
"grad_norm": 0.291015625,
|
| 5053 |
+
"learning_rate": 0.001,
|
| 5054 |
+
"loss": 1.2419,
|
| 5055 |
+
"step": 9996
|
| 5056 |
+
},
|
| 5057 |
+
{
|
| 5058 |
+
"epoch": 0.699229869200007,
|
| 5059 |
+
"grad_norm": 0.251953125,
|
| 5060 |
+
"learning_rate": 0.001,
|
| 5061 |
+
"loss": 1.2406,
|
| 5062 |
+
"step": 10010
|
| 5063 |
+
},
|
| 5064 |
+
{
|
| 5065 |
+
"epoch": 0.6997188411365105,
|
| 5066 |
+
"eval_loss": 1.6404287815093994,
|
| 5067 |
+
"eval_runtime": 9.1224,
|
| 5068 |
+
"eval_samples_per_second": 109.621,
|
| 5069 |
+
"eval_steps_per_second": 1.425,
|
| 5070 |
+
"step": 10017
|
| 5071 |
}
|
| 5072 |
],
|
| 5073 |
"logging_steps": 14,
|
|
|
|
| 5087 |
"attributes": {}
|
| 5088 |
}
|
| 5089 |
},
|
| 5090 |
+
"total_flos": 2.66909141699448e+18,
|
| 5091 |
"train_batch_size": 8,
|
| 5092 |
"trial_name": null,
|
| 5093 |
"trial_params": null
|