Upload folder using huggingface_hub
Browse files- optimizer.pt +1 -1
- pytorch_model-00001-of-00003.bin +1 -1
- pytorch_model-00002-of-00003.bin +1 -1
- pytorch_model-00003-of-00003.bin +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
- scheduler.pt +1 -1
- trainer_state.json +1451 -3
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13476859646
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d51eb87875752e699b0b9ce682165fb110c71a672b28fc084bbf55fc1977239
|
| 3 |
size 13476859646
|
pytorch_model-00001-of-00003.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9877989650
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0109702a17586a6669f6211be7906e986e5292f56b1e53682a9decc684dff422
|
| 3 |
size 9877989650
|
pytorch_model-00002-of-00003.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 9894801206
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cec7a4785a52bef4efdb94137e6c54a479ed55762251e2ecc1658ee1bf38ff38
|
| 3 |
size 9894801206
|
pytorch_model-00003-of-00003.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7180990841
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:519cc27e43a806ed30c155517838a9b34b0e96de21ac23267a2e2fdfd4c4c2e8
|
| 3 |
size 7180990841
|
rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14583
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2f6bc3b332b1d7b34dd8e7d7ed0389c868155059ddb1d908e9ac3feb6672b23c
|
| 3 |
size 14583
|
rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14583
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8de5e0c7dadcd828a8d62fffc136e170202022509240a895985c7bc45cabbced
|
| 3 |
size 14583
|
rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14583
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c420d12d8aa09a561480241f19154d4aedd8a866de54ed145d69f860bae6f94
|
| 3 |
size 14583
|
rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14583
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec873fd7c31f869e7956f098c0d1e17d2296924b3c55e4971a059dd097690b6f
|
| 3 |
size 14583
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 627
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fab38bc378f5a3363deb1f9a4cf2ada87fd73ebe45f0144d5d4b18eed6403727
|
| 3 |
size 627
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -2894,11 +2894,1459 @@
|
|
| 2894 |
"learning_rate": 8.570056464180998e-06,
|
| 2895 |
"loss": 0.2311,
|
| 2896 |
"step": 480
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2897 |
}
|
| 2898 |
],
|
| 2899 |
"max_steps": 858,
|
| 2900 |
"num_train_epochs": 3,
|
| 2901 |
-
"total_flos": 1.
|
| 2902 |
"trial_name": null,
|
| 2903 |
"trial_params": null
|
| 2904 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.5174825174825175,
|
| 5 |
+
"global_step": 720,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 2894 |
"learning_rate": 8.570056464180998e-06,
|
| 2895 |
"loss": 0.2311,
|
| 2896 |
"step": 480
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 1.68,
|
| 2900 |
+
"learning_rate": 8.532695255446384e-06,
|
| 2901 |
+
"loss": 0.2439,
|
| 2902 |
+
"step": 481
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 1.69,
|
| 2906 |
+
"learning_rate": 8.49535496725217e-06,
|
| 2907 |
+
"loss": 0.238,
|
| 2908 |
+
"step": 482
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 1.69,
|
| 2912 |
+
"learning_rate": 8.458036131988792e-06,
|
| 2913 |
+
"loss": 0.2322,
|
| 2914 |
+
"step": 483
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 1.69,
|
| 2918 |
+
"learning_rate": 8.420739281740806e-06,
|
| 2919 |
+
"loss": 0.2276,
|
| 2920 |
+
"step": 484
|
| 2921 |
+
},
|
| 2922 |
+
{
|
| 2923 |
+
"epoch": 1.7,
|
| 2924 |
+
"learning_rate": 8.383464948279319e-06,
|
| 2925 |
+
"loss": 0.2401,
|
| 2926 |
+
"step": 485
|
| 2927 |
+
},
|
| 2928 |
+
{
|
| 2929 |
+
"epoch": 1.7,
|
| 2930 |
+
"learning_rate": 8.346213663054388e-06,
|
| 2931 |
+
"loss": 0.2328,
|
| 2932 |
+
"step": 486
|
| 2933 |
+
},
|
| 2934 |
+
{
|
| 2935 |
+
"epoch": 1.7,
|
| 2936 |
+
"learning_rate": 8.308985957187466e-06,
|
| 2937 |
+
"loss": 0.2292,
|
| 2938 |
+
"step": 487
|
| 2939 |
+
},
|
| 2940 |
+
{
|
| 2941 |
+
"epoch": 1.71,
|
| 2942 |
+
"learning_rate": 8.271782361463806e-06,
|
| 2943 |
+
"loss": 0.2211,
|
| 2944 |
+
"step": 488
|
| 2945 |
+
},
|
| 2946 |
+
{
|
| 2947 |
+
"epoch": 1.71,
|
| 2948 |
+
"learning_rate": 8.234603406324909e-06,
|
| 2949 |
+
"loss": 0.2379,
|
| 2950 |
+
"step": 489
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 1.71,
|
| 2954 |
+
"learning_rate": 8.197449621860944e-06,
|
| 2955 |
+
"loss": 0.2438,
|
| 2956 |
+
"step": 490
|
| 2957 |
+
},
|
| 2958 |
+
{
|
| 2959 |
+
"epoch": 1.72,
|
| 2960 |
+
"learning_rate": 8.16032153780322e-06,
|
| 2961 |
+
"loss": 0.2556,
|
| 2962 |
+
"step": 491
|
| 2963 |
+
},
|
| 2964 |
+
{
|
| 2965 |
+
"epoch": 1.72,
|
| 2966 |
+
"learning_rate": 8.123219683516604e-06,
|
| 2967 |
+
"loss": 0.217,
|
| 2968 |
+
"step": 492
|
| 2969 |
+
},
|
| 2970 |
+
{
|
| 2971 |
+
"epoch": 1.72,
|
| 2972 |
+
"learning_rate": 8.08614458799198e-06,
|
| 2973 |
+
"loss": 0.2489,
|
| 2974 |
+
"step": 493
|
| 2975 |
+
},
|
| 2976 |
+
{
|
| 2977 |
+
"epoch": 1.73,
|
| 2978 |
+
"learning_rate": 8.04909677983872e-06,
|
| 2979 |
+
"loss": 0.2447,
|
| 2980 |
+
"step": 494
|
| 2981 |
+
},
|
| 2982 |
+
{
|
| 2983 |
+
"epoch": 1.73,
|
| 2984 |
+
"learning_rate": 8.01207678727713e-06,
|
| 2985 |
+
"loss": 0.2486,
|
| 2986 |
+
"step": 495
|
| 2987 |
+
},
|
| 2988 |
+
{
|
| 2989 |
+
"epoch": 1.73,
|
| 2990 |
+
"learning_rate": 7.975085138130938e-06,
|
| 2991 |
+
"loss": 0.2513,
|
| 2992 |
+
"step": 496
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 1.74,
|
| 2996 |
+
"learning_rate": 7.938122359819745e-06,
|
| 2997 |
+
"loss": 0.2451,
|
| 2998 |
+
"step": 497
|
| 2999 |
+
},
|
| 3000 |
+
{
|
| 3001 |
+
"epoch": 1.74,
|
| 3002 |
+
"learning_rate": 7.901188979351527e-06,
|
| 3003 |
+
"loss": 0.2301,
|
| 3004 |
+
"step": 498
|
| 3005 |
+
},
|
| 3006 |
+
{
|
| 3007 |
+
"epoch": 1.74,
|
| 3008 |
+
"learning_rate": 7.864285523315097e-06,
|
| 3009 |
+
"loss": 0.2475,
|
| 3010 |
+
"step": 499
|
| 3011 |
+
},
|
| 3012 |
+
{
|
| 3013 |
+
"epoch": 1.75,
|
| 3014 |
+
"learning_rate": 7.827412517872634e-06,
|
| 3015 |
+
"loss": 0.2394,
|
| 3016 |
+
"step": 500
|
| 3017 |
+
},
|
| 3018 |
+
{
|
| 3019 |
+
"epoch": 1.75,
|
| 3020 |
+
"learning_rate": 7.790570488752137e-06,
|
| 3021 |
+
"loss": 0.2304,
|
| 3022 |
+
"step": 501
|
| 3023 |
+
},
|
| 3024 |
+
{
|
| 3025 |
+
"epoch": 1.76,
|
| 3026 |
+
"learning_rate": 7.753759961239965e-06,
|
| 3027 |
+
"loss": 0.2185,
|
| 3028 |
+
"step": 502
|
| 3029 |
+
},
|
| 3030 |
+
{
|
| 3031 |
+
"epoch": 1.76,
|
| 3032 |
+
"learning_rate": 7.716981460173319e-06,
|
| 3033 |
+
"loss": 0.2383,
|
| 3034 |
+
"step": 503
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 1.76,
|
| 3038 |
+
"learning_rate": 7.680235509932791e-06,
|
| 3039 |
+
"loss": 0.2208,
|
| 3040 |
+
"step": 504
|
| 3041 |
+
},
|
| 3042 |
+
{
|
| 3043 |
+
"epoch": 1.77,
|
| 3044 |
+
"learning_rate": 7.643522634434856e-06,
|
| 3045 |
+
"loss": 0.232,
|
| 3046 |
+
"step": 505
|
| 3047 |
+
},
|
| 3048 |
+
{
|
| 3049 |
+
"epoch": 1.77,
|
| 3050 |
+
"learning_rate": 7.606843357124426e-06,
|
| 3051 |
+
"loss": 0.2213,
|
| 3052 |
+
"step": 506
|
| 3053 |
+
},
|
| 3054 |
+
{
|
| 3055 |
+
"epoch": 1.77,
|
| 3056 |
+
"learning_rate": 7.570198200967363e-06,
|
| 3057 |
+
"loss": 0.2409,
|
| 3058 |
+
"step": 507
|
| 3059 |
+
},
|
| 3060 |
+
{
|
| 3061 |
+
"epoch": 1.78,
|
| 3062 |
+
"learning_rate": 7.5335876884430495e-06,
|
| 3063 |
+
"loss": 0.2161,
|
| 3064 |
+
"step": 508
|
| 3065 |
+
},
|
| 3066 |
+
{
|
| 3067 |
+
"epoch": 1.78,
|
| 3068 |
+
"learning_rate": 7.497012341536924e-06,
|
| 3069 |
+
"loss": 0.2244,
|
| 3070 |
+
"step": 509
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 1.78,
|
| 3074 |
+
"learning_rate": 7.460472681733031e-06,
|
| 3075 |
+
"loss": 0.2406,
|
| 3076 |
+
"step": 510
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 1.79,
|
| 3080 |
+
"learning_rate": 7.423969230006609e-06,
|
| 3081 |
+
"loss": 0.2504,
|
| 3082 |
+
"step": 511
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 1.79,
|
| 3086 |
+
"learning_rate": 7.387502506816638e-06,
|
| 3087 |
+
"loss": 0.235,
|
| 3088 |
+
"step": 512
|
| 3089 |
+
},
|
| 3090 |
+
{
|
| 3091 |
+
"epoch": 1.79,
|
| 3092 |
+
"learning_rate": 7.351073032098437e-06,
|
| 3093 |
+
"loss": 0.2376,
|
| 3094 |
+
"step": 513
|
| 3095 |
+
},
|
| 3096 |
+
{
|
| 3097 |
+
"epoch": 1.8,
|
| 3098 |
+
"learning_rate": 7.314681325256232e-06,
|
| 3099 |
+
"loss": 0.2281,
|
| 3100 |
+
"step": 514
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 1.8,
|
| 3104 |
+
"learning_rate": 7.278327905155783e-06,
|
| 3105 |
+
"loss": 0.2215,
|
| 3106 |
+
"step": 515
|
| 3107 |
+
},
|
| 3108 |
+
{
|
| 3109 |
+
"epoch": 1.8,
|
| 3110 |
+
"learning_rate": 7.242013290116944e-06,
|
| 3111 |
+
"loss": 0.2446,
|
| 3112 |
+
"step": 516
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 1.81,
|
| 3116 |
+
"learning_rate": 7.205737997906307e-06,
|
| 3117 |
+
"loss": 0.2385,
|
| 3118 |
+
"step": 517
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 1.81,
|
| 3122 |
+
"learning_rate": 7.169502545729798e-06,
|
| 3123 |
+
"loss": 0.2401,
|
| 3124 |
+
"step": 518
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 1.81,
|
| 3128 |
+
"learning_rate": 7.133307450225322e-06,
|
| 3129 |
+
"loss": 0.2354,
|
| 3130 |
+
"step": 519
|
| 3131 |
+
},
|
| 3132 |
+
{
|
| 3133 |
+
"epoch": 1.82,
|
| 3134 |
+
"learning_rate": 7.097153227455379e-06,
|
| 3135 |
+
"loss": 0.2367,
|
| 3136 |
+
"step": 520
|
| 3137 |
+
},
|
| 3138 |
+
{
|
| 3139 |
+
"epoch": 1.82,
|
| 3140 |
+
"learning_rate": 7.0610403928997114e-06,
|
| 3141 |
+
"loss": 0.246,
|
| 3142 |
+
"step": 521
|
| 3143 |
+
},
|
| 3144 |
+
{
|
| 3145 |
+
"epoch": 1.83,
|
| 3146 |
+
"learning_rate": 7.024969461447973e-06,
|
| 3147 |
+
"loss": 0.2227,
|
| 3148 |
+
"step": 522
|
| 3149 |
+
},
|
| 3150 |
+
{
|
| 3151 |
+
"epoch": 1.83,
|
| 3152 |
+
"learning_rate": 6.9889409473923445e-06,
|
| 3153 |
+
"loss": 0.2342,
|
| 3154 |
+
"step": 523
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 1.83,
|
| 3158 |
+
"learning_rate": 6.952955364420255e-06,
|
| 3159 |
+
"loss": 0.2398,
|
| 3160 |
+
"step": 524
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 1.84,
|
| 3164 |
+
"learning_rate": 6.91701322560701e-06,
|
| 3165 |
+
"loss": 0.22,
|
| 3166 |
+
"step": 525
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 1.84,
|
| 3170 |
+
"learning_rate": 6.881115043408512e-06,
|
| 3171 |
+
"loss": 0.2343,
|
| 3172 |
+
"step": 526
|
| 3173 |
+
},
|
| 3174 |
+
{
|
| 3175 |
+
"epoch": 1.84,
|
| 3176 |
+
"learning_rate": 6.845261329653923e-06,
|
| 3177 |
+
"loss": 0.244,
|
| 3178 |
+
"step": 527
|
| 3179 |
+
},
|
| 3180 |
+
{
|
| 3181 |
+
"epoch": 1.85,
|
| 3182 |
+
"learning_rate": 6.809452595538403e-06,
|
| 3183 |
+
"loss": 0.2396,
|
| 3184 |
+
"step": 528
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 1.85,
|
| 3188 |
+
"learning_rate": 6.7736893516157795e-06,
|
| 3189 |
+
"loss": 0.2296,
|
| 3190 |
+
"step": 529
|
| 3191 |
+
},
|
| 3192 |
+
{
|
| 3193 |
+
"epoch": 1.85,
|
| 3194 |
+
"learning_rate": 6.7379721077913095e-06,
|
| 3195 |
+
"loss": 0.2295,
|
| 3196 |
+
"step": 530
|
| 3197 |
+
},
|
| 3198 |
+
{
|
| 3199 |
+
"epoch": 1.86,
|
| 3200 |
+
"learning_rate": 6.70230137331437e-06,
|
| 3201 |
+
"loss": 0.245,
|
| 3202 |
+
"step": 531
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 1.86,
|
| 3206 |
+
"learning_rate": 6.666677656771239e-06,
|
| 3207 |
+
"loss": 0.2305,
|
| 3208 |
+
"step": 532
|
| 3209 |
+
},
|
| 3210 |
+
{
|
| 3211 |
+
"epoch": 1.86,
|
| 3212 |
+
"learning_rate": 6.631101466077801e-06,
|
| 3213 |
+
"loss": 0.2527,
|
| 3214 |
+
"step": 533
|
| 3215 |
+
},
|
| 3216 |
+
{
|
| 3217 |
+
"epoch": 1.87,
|
| 3218 |
+
"learning_rate": 6.595573308472338e-06,
|
| 3219 |
+
"loss": 0.2522,
|
| 3220 |
+
"step": 534
|
| 3221 |
+
},
|
| 3222 |
+
{
|
| 3223 |
+
"epoch": 1.87,
|
| 3224 |
+
"learning_rate": 6.56009369050829e-06,
|
| 3225 |
+
"loss": 0.2481,
|
| 3226 |
+
"step": 535
|
| 3227 |
+
},
|
| 3228 |
+
{
|
| 3229 |
+
"epoch": 1.87,
|
| 3230 |
+
"learning_rate": 6.524663118047015e-06,
|
| 3231 |
+
"loss": 0.2197,
|
| 3232 |
+
"step": 536
|
| 3233 |
+
},
|
| 3234 |
+
{
|
| 3235 |
+
"epoch": 1.88,
|
| 3236 |
+
"learning_rate": 6.489282096250602e-06,
|
| 3237 |
+
"loss": 0.2209,
|
| 3238 |
+
"step": 537
|
| 3239 |
+
},
|
| 3240 |
+
{
|
| 3241 |
+
"epoch": 1.88,
|
| 3242 |
+
"learning_rate": 6.453951129574644e-06,
|
| 3243 |
+
"loss": 0.2253,
|
| 3244 |
+
"step": 538
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 1.88,
|
| 3248 |
+
"learning_rate": 6.4186707217610735e-06,
|
| 3249 |
+
"loss": 0.2205,
|
| 3250 |
+
"step": 539
|
| 3251 |
+
},
|
| 3252 |
+
{
|
| 3253 |
+
"epoch": 1.89,
|
| 3254 |
+
"learning_rate": 6.3834413758309525e-06,
|
| 3255 |
+
"loss": 0.2536,
|
| 3256 |
+
"step": 540
|
| 3257 |
+
},
|
| 3258 |
+
{
|
| 3259 |
+
"epoch": 1.89,
|
| 3260 |
+
"learning_rate": 6.348263594077319e-06,
|
| 3261 |
+
"loss": 0.229,
|
| 3262 |
+
"step": 541
|
| 3263 |
+
},
|
| 3264 |
+
{
|
| 3265 |
+
"epoch": 1.9,
|
| 3266 |
+
"learning_rate": 6.3131378780580134e-06,
|
| 3267 |
+
"loss": 0.2434,
|
| 3268 |
+
"step": 542
|
| 3269 |
+
},
|
| 3270 |
+
{
|
| 3271 |
+
"epoch": 1.9,
|
| 3272 |
+
"learning_rate": 6.278064728588542e-06,
|
| 3273 |
+
"loss": 0.2428,
|
| 3274 |
+
"step": 543
|
| 3275 |
+
},
|
| 3276 |
+
{
|
| 3277 |
+
"epoch": 1.9,
|
| 3278 |
+
"learning_rate": 6.243044645734917e-06,
|
| 3279 |
+
"loss": 0.2441,
|
| 3280 |
+
"step": 544
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"epoch": 1.91,
|
| 3284 |
+
"learning_rate": 6.208078128806549e-06,
|
| 3285 |
+
"loss": 0.2493,
|
| 3286 |
+
"step": 545
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 1.91,
|
| 3290 |
+
"learning_rate": 6.173165676349103e-06,
|
| 3291 |
+
"loss": 0.234,
|
| 3292 |
+
"step": 546
|
| 3293 |
+
},
|
| 3294 |
+
{
|
| 3295 |
+
"epoch": 1.91,
|
| 3296 |
+
"learning_rate": 6.138307786137415e-06,
|
| 3297 |
+
"loss": 0.2335,
|
| 3298 |
+
"step": 547
|
| 3299 |
+
},
|
| 3300 |
+
{
|
| 3301 |
+
"epoch": 1.92,
|
| 3302 |
+
"learning_rate": 6.103504955168382e-06,
|
| 3303 |
+
"loss": 0.2273,
|
| 3304 |
+
"step": 548
|
| 3305 |
+
},
|
| 3306 |
+
{
|
| 3307 |
+
"epoch": 1.92,
|
| 3308 |
+
"learning_rate": 6.0687576796538685e-06,
|
| 3309 |
+
"loss": 0.2285,
|
| 3310 |
+
"step": 549
|
| 3311 |
+
},
|
| 3312 |
+
{
|
| 3313 |
+
"epoch": 1.92,
|
| 3314 |
+
"learning_rate": 6.0340664550136494e-06,
|
| 3315 |
+
"loss": 0.23,
|
| 3316 |
+
"step": 550
|
| 3317 |
+
},
|
| 3318 |
+
{
|
| 3319 |
+
"epoch": 1.93,
|
| 3320 |
+
"learning_rate": 5.999431775868329e-06,
|
| 3321 |
+
"loss": 0.2425,
|
| 3322 |
+
"step": 551
|
| 3323 |
+
},
|
| 3324 |
+
{
|
| 3325 |
+
"epoch": 1.93,
|
| 3326 |
+
"learning_rate": 5.96485413603231e-06,
|
| 3327 |
+
"loss": 0.2407,
|
| 3328 |
+
"step": 552
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 1.93,
|
| 3332 |
+
"learning_rate": 5.930334028506726e-06,
|
| 3333 |
+
"loss": 0.2289,
|
| 3334 |
+
"step": 553
|
| 3335 |
+
},
|
| 3336 |
+
{
|
| 3337 |
+
"epoch": 1.94,
|
| 3338 |
+
"learning_rate": 5.895871945472434e-06,
|
| 3339 |
+
"loss": 0.2091,
|
| 3340 |
+
"step": 554
|
| 3341 |
+
},
|
| 3342 |
+
{
|
| 3343 |
+
"epoch": 1.94,
|
| 3344 |
+
"learning_rate": 5.861468378282984e-06,
|
| 3345 |
+
"loss": 0.2212,
|
| 3346 |
+
"step": 555
|
| 3347 |
+
},
|
| 3348 |
+
{
|
| 3349 |
+
"epoch": 1.94,
|
| 3350 |
+
"learning_rate": 5.827123817457631e-06,
|
| 3351 |
+
"loss": 0.2393,
|
| 3352 |
+
"step": 556
|
| 3353 |
+
},
|
| 3354 |
+
{
|
| 3355 |
+
"epoch": 1.95,
|
| 3356 |
+
"learning_rate": 5.792838752674309e-06,
|
| 3357 |
+
"loss": 0.2308,
|
| 3358 |
+
"step": 557
|
| 3359 |
+
},
|
| 3360 |
+
{
|
| 3361 |
+
"epoch": 1.95,
|
| 3362 |
+
"learning_rate": 5.75861367276269e-06,
|
| 3363 |
+
"loss": 0.2339,
|
| 3364 |
+
"step": 558
|
| 3365 |
+
},
|
| 3366 |
+
{
|
| 3367 |
+
"epoch": 1.95,
|
| 3368 |
+
"learning_rate": 5.724449065697182e-06,
|
| 3369 |
+
"loss": 0.2286,
|
| 3370 |
+
"step": 559
|
| 3371 |
+
},
|
| 3372 |
+
{
|
| 3373 |
+
"epoch": 1.96,
|
| 3374 |
+
"learning_rate": 5.690345418589978e-06,
|
| 3375 |
+
"loss": 0.2509,
|
| 3376 |
+
"step": 560
|
| 3377 |
+
},
|
| 3378 |
+
{
|
| 3379 |
+
"epoch": 1.96,
|
| 3380 |
+
"learning_rate": 5.656303217684133e-06,
|
| 3381 |
+
"loss": 0.2434,
|
| 3382 |
+
"step": 561
|
| 3383 |
+
},
|
| 3384 |
+
{
|
| 3385 |
+
"epoch": 1.97,
|
| 3386 |
+
"learning_rate": 5.622322948346595e-06,
|
| 3387 |
+
"loss": 0.2346,
|
| 3388 |
+
"step": 562
|
| 3389 |
+
},
|
| 3390 |
+
{
|
| 3391 |
+
"epoch": 1.97,
|
| 3392 |
+
"learning_rate": 5.588405095061322e-06,
|
| 3393 |
+
"loss": 0.245,
|
| 3394 |
+
"step": 563
|
| 3395 |
+
},
|
| 3396 |
+
{
|
| 3397 |
+
"epoch": 1.97,
|
| 3398 |
+
"learning_rate": 5.55455014142234e-06,
|
| 3399 |
+
"loss": 0.2255,
|
| 3400 |
+
"step": 564
|
| 3401 |
+
},
|
| 3402 |
+
{
|
| 3403 |
+
"epoch": 1.98,
|
| 3404 |
+
"learning_rate": 5.5207585701268805e-06,
|
| 3405 |
+
"loss": 0.2267,
|
| 3406 |
+
"step": 565
|
| 3407 |
+
},
|
| 3408 |
+
{
|
| 3409 |
+
"epoch": 1.98,
|
| 3410 |
+
"learning_rate": 5.4870308629684675e-06,
|
| 3411 |
+
"loss": 0.2529,
|
| 3412 |
+
"step": 566
|
| 3413 |
+
},
|
| 3414 |
+
{
|
| 3415 |
+
"epoch": 1.98,
|
| 3416 |
+
"learning_rate": 5.453367500830069e-06,
|
| 3417 |
+
"loss": 0.2457,
|
| 3418 |
+
"step": 567
|
| 3419 |
+
},
|
| 3420 |
+
{
|
| 3421 |
+
"epoch": 1.99,
|
| 3422 |
+
"learning_rate": 5.419768963677233e-06,
|
| 3423 |
+
"loss": 0.2378,
|
| 3424 |
+
"step": 568
|
| 3425 |
+
},
|
| 3426 |
+
{
|
| 3427 |
+
"epoch": 1.99,
|
| 3428 |
+
"learning_rate": 5.3862357305512524e-06,
|
| 3429 |
+
"loss": 0.2198,
|
| 3430 |
+
"step": 569
|
| 3431 |
+
},
|
| 3432 |
+
{
|
| 3433 |
+
"epoch": 1.99,
|
| 3434 |
+
"learning_rate": 5.352768279562315e-06,
|
| 3435 |
+
"loss": 0.2448,
|
| 3436 |
+
"step": 570
|
| 3437 |
+
},
|
| 3438 |
+
{
|
| 3439 |
+
"epoch": 2.0,
|
| 3440 |
+
"learning_rate": 5.319367087882717e-06,
|
| 3441 |
+
"loss": 0.2713,
|
| 3442 |
+
"step": 571
|
| 3443 |
+
},
|
| 3444 |
+
{
|
| 3445 |
+
"epoch": 2.0,
|
| 3446 |
+
"learning_rate": 5.286032631740023e-06,
|
| 3447 |
+
"loss": 0.2335,
|
| 3448 |
+
"step": 572
|
| 3449 |
+
},
|
| 3450 |
+
{
|
| 3451 |
+
"epoch": 2.0,
|
| 3452 |
+
"eval_loss": 0.3002837300300598,
|
| 3453 |
+
"eval_runtime": 165.4835,
|
| 3454 |
+
"eval_samples_per_second": 4.514,
|
| 3455 |
+
"eval_steps_per_second": 0.145,
|
| 3456 |
+
"step": 572
|
| 3457 |
+
},
|
| 3458 |
+
{
|
| 3459 |
+
"epoch": 2.0,
|
| 3460 |
+
"learning_rate": 5.2527653864103124e-06,
|
| 3461 |
+
"loss": 0.1816,
|
| 3462 |
+
"step": 573
|
| 3463 |
+
},
|
| 3464 |
+
{
|
| 3465 |
+
"epoch": 2.01,
|
| 3466 |
+
"learning_rate": 5.219565826211382e-06,
|
| 3467 |
+
"loss": 0.1793,
|
| 3468 |
+
"step": 574
|
| 3469 |
+
},
|
| 3470 |
+
{
|
| 3471 |
+
"epoch": 2.01,
|
| 3472 |
+
"learning_rate": 5.18643442449598e-06,
|
| 3473 |
+
"loss": 0.1813,
|
| 3474 |
+
"step": 575
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 2.01,
|
| 3478 |
+
"learning_rate": 5.15337165364507e-06,
|
| 3479 |
+
"loss": 0.1827,
|
| 3480 |
+
"step": 576
|
| 3481 |
+
},
|
| 3482 |
+
{
|
| 3483 |
+
"epoch": 2.02,
|
| 3484 |
+
"learning_rate": 5.1203779850610865e-06,
|
| 3485 |
+
"loss": 0.1763,
|
| 3486 |
+
"step": 577
|
| 3487 |
+
},
|
| 3488 |
+
{
|
| 3489 |
+
"epoch": 2.02,
|
| 3490 |
+
"learning_rate": 5.087453889161229e-06,
|
| 3491 |
+
"loss": 0.1937,
|
| 3492 |
+
"step": 578
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 2.02,
|
| 3496 |
+
"learning_rate": 5.054599835370724e-06,
|
| 3497 |
+
"loss": 0.1733,
|
| 3498 |
+
"step": 579
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 2.03,
|
| 3502 |
+
"learning_rate": 5.021816292116175e-06,
|
| 3503 |
+
"loss": 0.1892,
|
| 3504 |
+
"step": 580
|
| 3505 |
+
},
|
| 3506 |
+
{
|
| 3507 |
+
"epoch": 2.03,
|
| 3508 |
+
"learning_rate": 4.989103726818836e-06,
|
| 3509 |
+
"loss": 0.1766,
|
| 3510 |
+
"step": 581
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 2.03,
|
| 3514 |
+
"learning_rate": 4.956462605887994e-06,
|
| 3515 |
+
"loss": 0.1728,
|
| 3516 |
+
"step": 582
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 2.04,
|
| 3520 |
+
"learning_rate": 4.92389339471428e-06,
|
| 3521 |
+
"loss": 0.178,
|
| 3522 |
+
"step": 583
|
| 3523 |
+
},
|
| 3524 |
+
{
|
| 3525 |
+
"epoch": 2.04,
|
| 3526 |
+
"learning_rate": 4.891396557663056e-06,
|
| 3527 |
+
"loss": 0.1998,
|
| 3528 |
+
"step": 584
|
| 3529 |
+
},
|
| 3530 |
+
{
|
| 3531 |
+
"epoch": 2.05,
|
| 3532 |
+
"learning_rate": 4.858972558067784e-06,
|
| 3533 |
+
"loss": 0.1779,
|
| 3534 |
+
"step": 585
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 2.05,
|
| 3538 |
+
"learning_rate": 4.826621858223431e-06,
|
| 3539 |
+
"loss": 0.1827,
|
| 3540 |
+
"step": 586
|
| 3541 |
+
},
|
| 3542 |
+
{
|
| 3543 |
+
"epoch": 2.05,
|
| 3544 |
+
"learning_rate": 4.794344919379872e-06,
|
| 3545 |
+
"loss": 0.1717,
|
| 3546 |
+
"step": 587
|
| 3547 |
+
},
|
| 3548 |
+
{
|
| 3549 |
+
"epoch": 2.06,
|
| 3550 |
+
"learning_rate": 4.762142201735299e-06,
|
| 3551 |
+
"loss": 0.1767,
|
| 3552 |
+
"step": 588
|
| 3553 |
+
},
|
| 3554 |
+
{
|
| 3555 |
+
"epoch": 2.06,
|
| 3556 |
+
"learning_rate": 4.730014164429689e-06,
|
| 3557 |
+
"loss": 0.1754,
|
| 3558 |
+
"step": 589
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 2.06,
|
| 3562 |
+
"learning_rate": 4.697961265538231e-06,
|
| 3563 |
+
"loss": 0.1693,
|
| 3564 |
+
"step": 590
|
| 3565 |
+
},
|
| 3566 |
+
{
|
| 3567 |
+
"epoch": 2.07,
|
| 3568 |
+
"learning_rate": 4.665983962064807e-06,
|
| 3569 |
+
"loss": 0.1814,
|
| 3570 |
+
"step": 591
|
| 3571 |
+
},
|
| 3572 |
+
{
|
| 3573 |
+
"epoch": 2.07,
|
| 3574 |
+
"learning_rate": 4.6340827099354734e-06,
|
| 3575 |
+
"loss": 0.1857,
|
| 3576 |
+
"step": 592
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 2.07,
|
| 3580 |
+
"learning_rate": 4.60225796399197e-06,
|
| 3581 |
+
"loss": 0.1727,
|
| 3582 |
+
"step": 593
|
| 3583 |
+
},
|
| 3584 |
+
{
|
| 3585 |
+
"epoch": 2.08,
|
| 3586 |
+
"learning_rate": 4.570510177985213e-06,
|
| 3587 |
+
"loss": 0.1726,
|
| 3588 |
+
"step": 594
|
| 3589 |
+
},
|
| 3590 |
+
{
|
| 3591 |
+
"epoch": 2.08,
|
| 3592 |
+
"learning_rate": 4.538839804568857e-06,
|
| 3593 |
+
"loss": 0.1771,
|
| 3594 |
+
"step": 595
|
| 3595 |
+
},
|
| 3596 |
+
{
|
| 3597 |
+
"epoch": 2.08,
|
| 3598 |
+
"learning_rate": 4.5072472952928015e-06,
|
| 3599 |
+
"loss": 0.1735,
|
| 3600 |
+
"step": 596
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 2.09,
|
| 3604 |
+
"learning_rate": 4.475733100596795e-06,
|
| 3605 |
+
"loss": 0.1716,
|
| 3606 |
+
"step": 597
|
| 3607 |
+
},
|
| 3608 |
+
{
|
| 3609 |
+
"epoch": 2.09,
|
| 3610 |
+
"learning_rate": 4.444297669803981e-06,
|
| 3611 |
+
"loss": 0.1667,
|
| 3612 |
+
"step": 598
|
| 3613 |
+
},
|
| 3614 |
+
{
|
| 3615 |
+
"epoch": 2.09,
|
| 3616 |
+
"learning_rate": 4.412941451114499e-06,
|
| 3617 |
+
"loss": 0.1808,
|
| 3618 |
+
"step": 599
|
| 3619 |
+
},
|
| 3620 |
+
{
|
| 3621 |
+
"epoch": 2.1,
|
| 3622 |
+
"learning_rate": 4.381664891599111e-06,
|
| 3623 |
+
"loss": 0.1858,
|
| 3624 |
+
"step": 600
|
| 3625 |
+
},
|
| 3626 |
+
{
|
| 3627 |
+
"epoch": 2.1,
|
| 3628 |
+
"learning_rate": 4.350468437192801e-06,
|
| 3629 |
+
"loss": 0.1798,
|
| 3630 |
+
"step": 601
|
| 3631 |
+
},
|
| 3632 |
+
{
|
| 3633 |
+
"epoch": 2.1,
|
| 3634 |
+
"learning_rate": 4.319352532688444e-06,
|
| 3635 |
+
"loss": 0.172,
|
| 3636 |
+
"step": 602
|
| 3637 |
+
},
|
| 3638 |
+
{
|
| 3639 |
+
"epoch": 2.11,
|
| 3640 |
+
"learning_rate": 4.288317621730434e-06,
|
| 3641 |
+
"loss": 0.1827,
|
| 3642 |
+
"step": 603
|
| 3643 |
+
},
|
| 3644 |
+
{
|
| 3645 |
+
"epoch": 2.11,
|
| 3646 |
+
"learning_rate": 4.257364146808394e-06,
|
| 3647 |
+
"loss": 0.1782,
|
| 3648 |
+
"step": 604
|
| 3649 |
+
},
|
| 3650 |
+
{
|
| 3651 |
+
"epoch": 2.12,
|
| 3652 |
+
"learning_rate": 4.226492549250829e-06,
|
| 3653 |
+
"loss": 0.1728,
|
| 3654 |
+
"step": 605
|
| 3655 |
+
},
|
| 3656 |
+
{
|
| 3657 |
+
"epoch": 2.12,
|
| 3658 |
+
"learning_rate": 4.1957032692188685e-06,
|
| 3659 |
+
"loss": 0.1727,
|
| 3660 |
+
"step": 606
|
| 3661 |
+
},
|
| 3662 |
+
{
|
| 3663 |
+
"epoch": 2.12,
|
| 3664 |
+
"learning_rate": 4.164996745699966e-06,
|
| 3665 |
+
"loss": 0.1807,
|
| 3666 |
+
"step": 607
|
| 3667 |
+
},
|
| 3668 |
+
{
|
| 3669 |
+
"epoch": 2.13,
|
| 3670 |
+
"learning_rate": 4.134373416501652e-06,
|
| 3671 |
+
"loss": 0.1742,
|
| 3672 |
+
"step": 608
|
| 3673 |
+
},
|
| 3674 |
+
{
|
| 3675 |
+
"epoch": 2.13,
|
| 3676 |
+
"learning_rate": 4.103833718245282e-06,
|
| 3677 |
+
"loss": 0.1783,
|
| 3678 |
+
"step": 609
|
| 3679 |
+
},
|
| 3680 |
+
{
|
| 3681 |
+
"epoch": 2.13,
|
| 3682 |
+
"learning_rate": 4.073378086359834e-06,
|
| 3683 |
+
"loss": 0.1821,
|
| 3684 |
+
"step": 610
|
| 3685 |
+
},
|
| 3686 |
+
{
|
| 3687 |
+
"epoch": 2.14,
|
| 3688 |
+
"learning_rate": 4.043006955075667e-06,
|
| 3689 |
+
"loss": 0.171,
|
| 3690 |
+
"step": 611
|
| 3691 |
+
},
|
| 3692 |
+
{
|
| 3693 |
+
"epoch": 2.14,
|
| 3694 |
+
"learning_rate": 4.0127207574183576e-06,
|
| 3695 |
+
"loss": 0.1798,
|
| 3696 |
+
"step": 612
|
| 3697 |
+
},
|
| 3698 |
+
{
|
| 3699 |
+
"epoch": 2.14,
|
| 3700 |
+
"learning_rate": 3.9825199252025185e-06,
|
| 3701 |
+
"loss": 0.1747,
|
| 3702 |
+
"step": 613
|
| 3703 |
+
},
|
| 3704 |
+
{
|
| 3705 |
+
"epoch": 2.15,
|
| 3706 |
+
"learning_rate": 3.952404889025626e-06,
|
| 3707 |
+
"loss": 0.1792,
|
| 3708 |
+
"step": 614
|
| 3709 |
+
},
|
| 3710 |
+
{
|
| 3711 |
+
"epoch": 2.15,
|
| 3712 |
+
"learning_rate": 3.922376078261905e-06,
|
| 3713 |
+
"loss": 0.1712,
|
| 3714 |
+
"step": 615
|
| 3715 |
+
},
|
| 3716 |
+
{
|
| 3717 |
+
"epoch": 2.15,
|
| 3718 |
+
"learning_rate": 3.892433921056184e-06,
|
| 3719 |
+
"loss": 0.1724,
|
| 3720 |
+
"step": 616
|
| 3721 |
+
},
|
| 3722 |
+
{
|
| 3723 |
+
"epoch": 2.16,
|
| 3724 |
+
"learning_rate": 3.862578844317817e-06,
|
| 3725 |
+
"loss": 0.1755,
|
| 3726 |
+
"step": 617
|
| 3727 |
+
},
|
| 3728 |
+
{
|
| 3729 |
+
"epoch": 2.16,
|
| 3730 |
+
"learning_rate": 3.832811273714569e-06,
|
| 3731 |
+
"loss": 0.1834,
|
| 3732 |
+
"step": 618
|
| 3733 |
+
},
|
| 3734 |
+
{
|
| 3735 |
+
"epoch": 2.16,
|
| 3736 |
+
"learning_rate": 3.8031316336665725e-06,
|
| 3737 |
+
"loss": 0.1735,
|
| 3738 |
+
"step": 619
|
| 3739 |
+
},
|
| 3740 |
+
{
|
| 3741 |
+
"epoch": 2.17,
|
| 3742 |
+
"learning_rate": 3.7735403473402475e-06,
|
| 3743 |
+
"loss": 0.1738,
|
| 3744 |
+
"step": 620
|
| 3745 |
+
},
|
| 3746 |
+
{
|
| 3747 |
+
"epoch": 2.17,
|
| 3748 |
+
"learning_rate": 3.7440378366423e-06,
|
| 3749 |
+
"loss": 0.176,
|
| 3750 |
+
"step": 621
|
| 3751 |
+
},
|
| 3752 |
+
{
|
| 3753 |
+
"epoch": 2.17,
|
| 3754 |
+
"learning_rate": 3.714624522213681e-06,
|
| 3755 |
+
"loss": 0.1803,
|
| 3756 |
+
"step": 622
|
| 3757 |
+
},
|
| 3758 |
+
{
|
| 3759 |
+
"epoch": 2.18,
|
| 3760 |
+
"learning_rate": 3.6853008234236023e-06,
|
| 3761 |
+
"loss": 0.1704,
|
| 3762 |
+
"step": 623
|
| 3763 |
+
},
|
| 3764 |
+
{
|
| 3765 |
+
"epoch": 2.18,
|
| 3766 |
+
"learning_rate": 3.6560671583635467e-06,
|
| 3767 |
+
"loss": 0.1804,
|
| 3768 |
+
"step": 624
|
| 3769 |
+
},
|
| 3770 |
+
{
|
| 3771 |
+
"epoch": 2.19,
|
| 3772 |
+
"learning_rate": 3.626923943841325e-06,
|
| 3773 |
+
"loss": 0.1823,
|
| 3774 |
+
"step": 625
|
| 3775 |
+
},
|
| 3776 |
+
{
|
| 3777 |
+
"epoch": 2.19,
|
| 3778 |
+
"learning_rate": 3.5978715953751207e-06,
|
| 3779 |
+
"loss": 0.184,
|
| 3780 |
+
"step": 626
|
| 3781 |
+
},
|
| 3782 |
+
{
|
| 3783 |
+
"epoch": 2.19,
|
| 3784 |
+
"learning_rate": 3.568910527187557e-06,
|
| 3785 |
+
"loss": 0.1731,
|
| 3786 |
+
"step": 627
|
| 3787 |
+
},
|
| 3788 |
+
{
|
| 3789 |
+
"epoch": 2.2,
|
| 3790 |
+
"learning_rate": 3.5400411521998123e-06,
|
| 3791 |
+
"loss": 0.1812,
|
| 3792 |
+
"step": 628
|
| 3793 |
+
},
|
| 3794 |
+
{
|
| 3795 |
+
"epoch": 2.2,
|
| 3796 |
+
"learning_rate": 3.5112638820257115e-06,
|
| 3797 |
+
"loss": 0.1786,
|
| 3798 |
+
"step": 629
|
| 3799 |
+
},
|
| 3800 |
+
{
|
| 3801 |
+
"epoch": 2.2,
|
| 3802 |
+
"learning_rate": 3.4825791269658782e-06,
|
| 3803 |
+
"loss": 0.1806,
|
| 3804 |
+
"step": 630
|
| 3805 |
+
},
|
| 3806 |
+
{
|
| 3807 |
+
"epoch": 2.21,
|
| 3808 |
+
"learning_rate": 3.453987296001866e-06,
|
| 3809 |
+
"loss": 0.1803,
|
| 3810 |
+
"step": 631
|
| 3811 |
+
},
|
| 3812 |
+
{
|
| 3813 |
+
"epoch": 2.21,
|
| 3814 |
+
"learning_rate": 3.4254887967903373e-06,
|
| 3815 |
+
"loss": 0.1745,
|
| 3816 |
+
"step": 632
|
| 3817 |
+
},
|
| 3818 |
+
{
|
| 3819 |
+
"epoch": 2.21,
|
| 3820 |
+
"learning_rate": 3.397084035657243e-06,
|
| 3821 |
+
"loss": 0.1852,
|
| 3822 |
+
"step": 633
|
| 3823 |
+
},
|
| 3824 |
+
{
|
| 3825 |
+
"epoch": 2.22,
|
| 3826 |
+
"learning_rate": 3.3687734175920505e-06,
|
| 3827 |
+
"loss": 0.1852,
|
| 3828 |
+
"step": 634
|
| 3829 |
+
},
|
| 3830 |
+
{
|
| 3831 |
+
"epoch": 2.22,
|
| 3832 |
+
"learning_rate": 3.3405573462419362e-06,
|
| 3833 |
+
"loss": 0.1529,
|
| 3834 |
+
"step": 635
|
| 3835 |
+
},
|
| 3836 |
+
{
|
| 3837 |
+
"epoch": 2.22,
|
| 3838 |
+
"learning_rate": 3.3124362239060627e-06,
|
| 3839 |
+
"loss": 0.1798,
|
| 3840 |
+
"step": 636
|
| 3841 |
+
},
|
| 3842 |
+
{
|
| 3843 |
+
"epoch": 2.23,
|
| 3844 |
+
"learning_rate": 3.284410451529816e-06,
|
| 3845 |
+
"loss": 0.1723,
|
| 3846 |
+
"step": 637
|
| 3847 |
+
},
|
| 3848 |
+
{
|
| 3849 |
+
"epoch": 2.23,
|
| 3850 |
+
"learning_rate": 3.2564804286991137e-06,
|
| 3851 |
+
"loss": 0.1769,
|
| 3852 |
+
"step": 638
|
| 3853 |
+
},
|
| 3854 |
+
{
|
| 3855 |
+
"epoch": 2.23,
|
| 3856 |
+
"learning_rate": 3.2286465536346857e-06,
|
| 3857 |
+
"loss": 0.1831,
|
| 3858 |
+
"step": 639
|
| 3859 |
+
},
|
| 3860 |
+
{
|
| 3861 |
+
"epoch": 2.24,
|
| 3862 |
+
"learning_rate": 3.2009092231864047e-06,
|
| 3863 |
+
"loss": 0.1845,
|
| 3864 |
+
"step": 640
|
| 3865 |
+
},
|
| 3866 |
+
{
|
| 3867 |
+
"epoch": 2.24,
|
| 3868 |
+
"learning_rate": 3.173268832827643e-06,
|
| 3869 |
+
"loss": 0.1609,
|
| 3870 |
+
"step": 641
|
| 3871 |
+
},
|
| 3872 |
+
{
|
| 3873 |
+
"epoch": 2.24,
|
| 3874 |
+
"learning_rate": 3.145725776649602e-06,
|
| 3875 |
+
"loss": 0.1716,
|
| 3876 |
+
"step": 642
|
| 3877 |
+
},
|
| 3878 |
+
{
|
| 3879 |
+
"epoch": 2.25,
|
| 3880 |
+
"learning_rate": 3.1182804473557295e-06,
|
| 3881 |
+
"loss": 0.1853,
|
| 3882 |
+
"step": 643
|
| 3883 |
+
},
|
| 3884 |
+
{
|
| 3885 |
+
"epoch": 2.25,
|
| 3886 |
+
"learning_rate": 3.0909332362560875e-06,
|
| 3887 |
+
"loss": 0.172,
|
| 3888 |
+
"step": 644
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 2.26,
|
| 3892 |
+
"learning_rate": 3.0636845332617994e-06,
|
| 3893 |
+
"loss": 0.1818,
|
| 3894 |
+
"step": 645
|
| 3895 |
+
},
|
| 3896 |
+
{
|
| 3897 |
+
"epoch": 2.26,
|
| 3898 |
+
"learning_rate": 3.036534726879473e-06,
|
| 3899 |
+
"loss": 0.1801,
|
| 3900 |
+
"step": 646
|
| 3901 |
+
},
|
| 3902 |
+
{
|
| 3903 |
+
"epoch": 2.26,
|
| 3904 |
+
"learning_rate": 3.0094842042056706e-06,
|
| 3905 |
+
"loss": 0.1942,
|
| 3906 |
+
"step": 647
|
| 3907 |
+
},
|
| 3908 |
+
{
|
| 3909 |
+
"epoch": 2.27,
|
| 3910 |
+
"learning_rate": 2.982533350921383e-06,
|
| 3911 |
+
"loss": 0.177,
|
| 3912 |
+
"step": 648
|
| 3913 |
+
},
|
| 3914 |
+
{
|
| 3915 |
+
"epoch": 2.27,
|
| 3916 |
+
"learning_rate": 2.9556825512865418e-06,
|
| 3917 |
+
"loss": 0.1784,
|
| 3918 |
+
"step": 649
|
| 3919 |
+
},
|
| 3920 |
+
{
|
| 3921 |
+
"epoch": 2.27,
|
| 3922 |
+
"learning_rate": 2.9289321881345257e-06,
|
| 3923 |
+
"loss": 0.179,
|
| 3924 |
+
"step": 650
|
| 3925 |
+
},
|
| 3926 |
+
{
|
| 3927 |
+
"epoch": 2.28,
|
| 3928 |
+
"learning_rate": 2.902282642866716e-06,
|
| 3929 |
+
"loss": 0.1772,
|
| 3930 |
+
"step": 651
|
| 3931 |
+
},
|
| 3932 |
+
{
|
| 3933 |
+
"epoch": 2.28,
|
| 3934 |
+
"learning_rate": 2.8757342954470537e-06,
|
| 3935 |
+
"loss": 0.1745,
|
| 3936 |
+
"step": 652
|
| 3937 |
+
},
|
| 3938 |
+
{
|
| 3939 |
+
"epoch": 2.28,
|
| 3940 |
+
"learning_rate": 2.8492875243966114e-06,
|
| 3941 |
+
"loss": 0.1796,
|
| 3942 |
+
"step": 653
|
| 3943 |
+
},
|
| 3944 |
+
{
|
| 3945 |
+
"epoch": 2.29,
|
| 3946 |
+
"learning_rate": 2.8229427067882165e-06,
|
| 3947 |
+
"loss": 0.1744,
|
| 3948 |
+
"step": 654
|
| 3949 |
+
},
|
| 3950 |
+
{
|
| 3951 |
+
"epoch": 2.29,
|
| 3952 |
+
"learning_rate": 2.7967002182410596e-06,
|
| 3953 |
+
"loss": 0.177,
|
| 3954 |
+
"step": 655
|
| 3955 |
+
},
|
| 3956 |
+
{
|
| 3957 |
+
"epoch": 2.29,
|
| 3958 |
+
"learning_rate": 2.7705604329153434e-06,
|
| 3959 |
+
"loss": 0.177,
|
| 3960 |
+
"step": 656
|
| 3961 |
+
},
|
| 3962 |
+
{
|
| 3963 |
+
"epoch": 2.3,
|
| 3964 |
+
"learning_rate": 2.7445237235069453e-06,
|
| 3965 |
+
"loss": 0.1765,
|
| 3966 |
+
"step": 657
|
| 3967 |
+
},
|
| 3968 |
+
{
|
| 3969 |
+
"epoch": 2.3,
|
| 3970 |
+
"learning_rate": 2.7185904612421177e-06,
|
| 3971 |
+
"loss": 0.1784,
|
| 3972 |
+
"step": 658
|
| 3973 |
+
},
|
| 3974 |
+
{
|
| 3975 |
+
"epoch": 2.3,
|
| 3976 |
+
"learning_rate": 2.6927610158721708e-06,
|
| 3977 |
+
"loss": 0.1796,
|
| 3978 |
+
"step": 659
|
| 3979 |
+
},
|
| 3980 |
+
{
|
| 3981 |
+
"epoch": 2.31,
|
| 3982 |
+
"learning_rate": 2.6670357556682245e-06,
|
| 3983 |
+
"loss": 0.1842,
|
| 3984 |
+
"step": 660
|
| 3985 |
+
},
|
| 3986 |
+
{
|
| 3987 |
+
"epoch": 2.31,
|
| 3988 |
+
"learning_rate": 2.6414150474159405e-06,
|
| 3989 |
+
"loss": 0.1843,
|
| 3990 |
+
"step": 661
|
| 3991 |
+
},
|
| 3992 |
+
{
|
| 3993 |
+
"epoch": 2.31,
|
| 3994 |
+
"learning_rate": 2.615899256410306e-06,
|
| 3995 |
+
"loss": 0.1816,
|
| 3996 |
+
"step": 662
|
| 3997 |
+
},
|
| 3998 |
+
{
|
| 3999 |
+
"epoch": 2.32,
|
| 4000 |
+
"learning_rate": 2.5904887464504115e-06,
|
| 4001 |
+
"loss": 0.1872,
|
| 4002 |
+
"step": 663
|
| 4003 |
+
},
|
| 4004 |
+
{
|
| 4005 |
+
"epoch": 2.32,
|
| 4006 |
+
"learning_rate": 2.565183879834272e-06,
|
| 4007 |
+
"loss": 0.1793,
|
| 4008 |
+
"step": 664
|
| 4009 |
+
},
|
| 4010 |
+
{
|
| 4011 |
+
"epoch": 2.33,
|
| 4012 |
+
"learning_rate": 2.53998501735367e-06,
|
| 4013 |
+
"loss": 0.1846,
|
| 4014 |
+
"step": 665
|
| 4015 |
+
},
|
| 4016 |
+
{
|
| 4017 |
+
"epoch": 2.33,
|
| 4018 |
+
"learning_rate": 2.514892518288988e-06,
|
| 4019 |
+
"loss": 0.1696,
|
| 4020 |
+
"step": 666
|
| 4021 |
+
},
|
| 4022 |
+
{
|
| 4023 |
+
"epoch": 2.33,
|
| 4024 |
+
"learning_rate": 2.4899067404041156e-06,
|
| 4025 |
+
"loss": 0.1821,
|
| 4026 |
+
"step": 667
|
| 4027 |
+
},
|
| 4028 |
+
{
|
| 4029 |
+
"epoch": 2.34,
|
| 4030 |
+
"learning_rate": 2.465028039941316e-06,
|
| 4031 |
+
"loss": 0.1667,
|
| 4032 |
+
"step": 668
|
| 4033 |
+
},
|
| 4034 |
+
{
|
| 4035 |
+
"epoch": 2.34,
|
| 4036 |
+
"learning_rate": 2.4402567716161806e-06,
|
| 4037 |
+
"loss": 0.186,
|
| 4038 |
+
"step": 669
|
| 4039 |
+
},
|
| 4040 |
+
{
|
| 4041 |
+
"epoch": 2.34,
|
| 4042 |
+
"learning_rate": 2.415593288612541e-06,
|
| 4043 |
+
"loss": 0.1727,
|
| 4044 |
+
"step": 670
|
| 4045 |
+
},
|
| 4046 |
+
{
|
| 4047 |
+
"epoch": 2.35,
|
| 4048 |
+
"learning_rate": 2.3910379425774544e-06,
|
| 4049 |
+
"loss": 0.1783,
|
| 4050 |
+
"step": 671
|
| 4051 |
+
},
|
| 4052 |
+
{
|
| 4053 |
+
"epoch": 2.35,
|
| 4054 |
+
"learning_rate": 2.366591083616178e-06,
|
| 4055 |
+
"loss": 0.1595,
|
| 4056 |
+
"step": 672
|
| 4057 |
+
},
|
| 4058 |
+
{
|
| 4059 |
+
"epoch": 2.35,
|
| 4060 |
+
"learning_rate": 2.3422530602871874e-06,
|
| 4061 |
+
"loss": 0.1577,
|
| 4062 |
+
"step": 673
|
| 4063 |
+
},
|
| 4064 |
+
{
|
| 4065 |
+
"epoch": 2.36,
|
| 4066 |
+
"learning_rate": 2.318024219597196e-06,
|
| 4067 |
+
"loss": 0.1816,
|
| 4068 |
+
"step": 674
|
| 4069 |
+
},
|
| 4070 |
+
{
|
| 4071 |
+
"epoch": 2.36,
|
| 4072 |
+
"learning_rate": 2.2939049069962183e-06,
|
| 4073 |
+
"loss": 0.182,
|
| 4074 |
+
"step": 675
|
| 4075 |
+
},
|
| 4076 |
+
{
|
| 4077 |
+
"epoch": 2.36,
|
| 4078 |
+
"learning_rate": 2.26989546637263e-06,
|
| 4079 |
+
"loss": 0.1792,
|
| 4080 |
+
"step": 676
|
| 4081 |
+
},
|
| 4082 |
+
{
|
| 4083 |
+
"epoch": 2.37,
|
| 4084 |
+
"learning_rate": 2.2459962400482848e-06,
|
| 4085 |
+
"loss": 0.1826,
|
| 4086 |
+
"step": 677
|
| 4087 |
+
},
|
| 4088 |
+
{
|
| 4089 |
+
"epoch": 2.37,
|
| 4090 |
+
"learning_rate": 2.222207568773619e-06,
|
| 4091 |
+
"loss": 0.1798,
|
| 4092 |
+
"step": 678
|
| 4093 |
+
},
|
| 4094 |
+
{
|
| 4095 |
+
"epoch": 2.37,
|
| 4096 |
+
"learning_rate": 2.1985297917227922e-06,
|
| 4097 |
+
"loss": 0.1721,
|
| 4098 |
+
"step": 679
|
| 4099 |
+
},
|
| 4100 |
+
{
|
| 4101 |
+
"epoch": 2.38,
|
| 4102 |
+
"learning_rate": 2.1749632464888594e-06,
|
| 4103 |
+
"loss": 0.1786,
|
| 4104 |
+
"step": 680
|
| 4105 |
+
},
|
| 4106 |
+
{
|
| 4107 |
+
"epoch": 2.38,
|
| 4108 |
+
"learning_rate": 2.1515082690789535e-06,
|
| 4109 |
+
"loss": 0.1832,
|
| 4110 |
+
"step": 681
|
| 4111 |
+
},
|
| 4112 |
+
{
|
| 4113 |
+
"epoch": 2.38,
|
| 4114 |
+
"learning_rate": 2.1281651939094996e-06,
|
| 4115 |
+
"loss": 0.1741,
|
| 4116 |
+
"step": 682
|
| 4117 |
+
},
|
| 4118 |
+
{
|
| 4119 |
+
"epoch": 2.39,
|
| 4120 |
+
"learning_rate": 2.1049343538014354e-06,
|
| 4121 |
+
"loss": 0.1692,
|
| 4122 |
+
"step": 683
|
| 4123 |
+
},
|
| 4124 |
+
{
|
| 4125 |
+
"epoch": 2.39,
|
| 4126 |
+
"learning_rate": 2.0818160799754826e-06,
|
| 4127 |
+
"loss": 0.181,
|
| 4128 |
+
"step": 684
|
| 4129 |
+
},
|
| 4130 |
+
{
|
| 4131 |
+
"epoch": 2.4,
|
| 4132 |
+
"learning_rate": 2.0588107020474056e-06,
|
| 4133 |
+
"loss": 0.1675,
|
| 4134 |
+
"step": 685
|
| 4135 |
+
},
|
| 4136 |
+
{
|
| 4137 |
+
"epoch": 2.4,
|
| 4138 |
+
"learning_rate": 2.03591854802333e-06,
|
| 4139 |
+
"loss": 0.1842,
|
| 4140 |
+
"step": 686
|
| 4141 |
+
},
|
| 4142 |
+
{
|
| 4143 |
+
"epoch": 2.4,
|
| 4144 |
+
"learning_rate": 2.0131399442950507e-06,
|
| 4145 |
+
"loss": 0.1693,
|
| 4146 |
+
"step": 687
|
| 4147 |
+
},
|
| 4148 |
+
{
|
| 4149 |
+
"epoch": 2.41,
|
| 4150 |
+
"learning_rate": 1.990475215635388e-06,
|
| 4151 |
+
"loss": 0.1825,
|
| 4152 |
+
"step": 688
|
| 4153 |
+
},
|
| 4154 |
+
{
|
| 4155 |
+
"epoch": 2.41,
|
| 4156 |
+
"learning_rate": 1.967924685193552e-06,
|
| 4157 |
+
"loss": 0.1864,
|
| 4158 |
+
"step": 689
|
| 4159 |
+
},
|
| 4160 |
+
{
|
| 4161 |
+
"epoch": 2.41,
|
| 4162 |
+
"learning_rate": 1.94548867449054e-06,
|
| 4163 |
+
"loss": 0.1732,
|
| 4164 |
+
"step": 690
|
| 4165 |
+
},
|
| 4166 |
+
{
|
| 4167 |
+
"epoch": 2.42,
|
| 4168 |
+
"learning_rate": 1.9231675034145515e-06,
|
| 4169 |
+
"loss": 0.1752,
|
| 4170 |
+
"step": 691
|
| 4171 |
+
},
|
| 4172 |
+
{
|
| 4173 |
+
"epoch": 2.42,
|
| 4174 |
+
"learning_rate": 1.9009614902164175e-06,
|
| 4175 |
+
"loss": 0.1724,
|
| 4176 |
+
"step": 692
|
| 4177 |
+
},
|
| 4178 |
+
{
|
| 4179 |
+
"epoch": 2.42,
|
| 4180 |
+
"learning_rate": 1.8788709515050808e-06,
|
| 4181 |
+
"loss": 0.1729,
|
| 4182 |
+
"step": 693
|
| 4183 |
+
},
|
| 4184 |
+
{
|
| 4185 |
+
"epoch": 2.43,
|
| 4186 |
+
"learning_rate": 1.8568962022430637e-06,
|
| 4187 |
+
"loss": 0.177,
|
| 4188 |
+
"step": 694
|
| 4189 |
+
},
|
| 4190 |
+
{
|
| 4191 |
+
"epoch": 2.43,
|
| 4192 |
+
"learning_rate": 1.8350375557419875e-06,
|
| 4193 |
+
"loss": 0.186,
|
| 4194 |
+
"step": 695
|
| 4195 |
+
},
|
| 4196 |
+
{
|
| 4197 |
+
"epoch": 2.43,
|
| 4198 |
+
"learning_rate": 1.813295323658103e-06,
|
| 4199 |
+
"loss": 0.1835,
|
| 4200 |
+
"step": 696
|
| 4201 |
+
},
|
| 4202 |
+
{
|
| 4203 |
+
"epoch": 2.44,
|
| 4204 |
+
"learning_rate": 1.791669815987852e-06,
|
| 4205 |
+
"loss": 0.1777,
|
| 4206 |
+
"step": 697
|
| 4207 |
+
},
|
| 4208 |
+
{
|
| 4209 |
+
"epoch": 2.44,
|
| 4210 |
+
"learning_rate": 1.7701613410634367e-06,
|
| 4211 |
+
"loss": 0.1762,
|
| 4212 |
+
"step": 698
|
| 4213 |
+
},
|
| 4214 |
+
{
|
| 4215 |
+
"epoch": 2.44,
|
| 4216 |
+
"learning_rate": 1.7487702055484345e-06,
|
| 4217 |
+
"loss": 0.1754,
|
| 4218 |
+
"step": 699
|
| 4219 |
+
},
|
| 4220 |
+
{
|
| 4221 |
+
"epoch": 2.45,
|
| 4222 |
+
"learning_rate": 1.7274967144334155e-06,
|
| 4223 |
+
"loss": 0.1846,
|
| 4224 |
+
"step": 700
|
| 4225 |
+
},
|
| 4226 |
+
{
|
| 4227 |
+
"epoch": 2.45,
|
| 4228 |
+
"learning_rate": 1.7063411710316047e-06,
|
| 4229 |
+
"loss": 0.1677,
|
| 4230 |
+
"step": 701
|
| 4231 |
+
},
|
| 4232 |
+
{
|
| 4233 |
+
"epoch": 2.45,
|
| 4234 |
+
"learning_rate": 1.6853038769745466e-06,
|
| 4235 |
+
"loss": 0.1734,
|
| 4236 |
+
"step": 702
|
| 4237 |
+
},
|
| 4238 |
+
{
|
| 4239 |
+
"epoch": 2.46,
|
| 4240 |
+
"learning_rate": 1.6643851322078176e-06,
|
| 4241 |
+
"loss": 0.1817,
|
| 4242 |
+
"step": 703
|
| 4243 |
+
},
|
| 4244 |
+
{
|
| 4245 |
+
"epoch": 2.46,
|
| 4246 |
+
"learning_rate": 1.643585234986733e-06,
|
| 4247 |
+
"loss": 0.1695,
|
| 4248 |
+
"step": 704
|
| 4249 |
+
},
|
| 4250 |
+
{
|
| 4251 |
+
"epoch": 2.47,
|
| 4252 |
+
"learning_rate": 1.6229044818721062e-06,
|
| 4253 |
+
"loss": 0.1866,
|
| 4254 |
+
"step": 705
|
| 4255 |
+
},
|
| 4256 |
+
{
|
| 4257 |
+
"epoch": 2.47,
|
| 4258 |
+
"learning_rate": 1.6023431677260215e-06,
|
| 4259 |
+
"loss": 0.1666,
|
| 4260 |
+
"step": 706
|
| 4261 |
+
},
|
| 4262 |
+
{
|
| 4263 |
+
"epoch": 2.47,
|
| 4264 |
+
"learning_rate": 1.5819015857076215e-06,
|
| 4265 |
+
"loss": 0.1903,
|
| 4266 |
+
"step": 707
|
| 4267 |
+
},
|
| 4268 |
+
{
|
| 4269 |
+
"epoch": 2.48,
|
| 4270 |
+
"learning_rate": 1.5615800272689352e-06,
|
| 4271 |
+
"loss": 0.1652,
|
| 4272 |
+
"step": 708
|
| 4273 |
+
},
|
| 4274 |
+
{
|
| 4275 |
+
"epoch": 2.48,
|
| 4276 |
+
"learning_rate": 1.5413787821507141e-06,
|
| 4277 |
+
"loss": 0.1727,
|
| 4278 |
+
"step": 709
|
| 4279 |
+
},
|
| 4280 |
+
{
|
| 4281 |
+
"epoch": 2.48,
|
| 4282 |
+
"learning_rate": 1.5212981383783154e-06,
|
| 4283 |
+
"loss": 0.1794,
|
| 4284 |
+
"step": 710
|
| 4285 |
+
},
|
| 4286 |
+
{
|
| 4287 |
+
"epoch": 2.49,
|
| 4288 |
+
"learning_rate": 1.5013383822575766e-06,
|
| 4289 |
+
"loss": 0.18,
|
| 4290 |
+
"step": 711
|
| 4291 |
+
},
|
| 4292 |
+
{
|
| 4293 |
+
"epoch": 2.49,
|
| 4294 |
+
"learning_rate": 1.4814997983707457e-06,
|
| 4295 |
+
"loss": 0.1676,
|
| 4296 |
+
"step": 712
|
| 4297 |
+
},
|
| 4298 |
+
{
|
| 4299 |
+
"epoch": 2.49,
|
| 4300 |
+
"learning_rate": 1.4617826695724224e-06,
|
| 4301 |
+
"loss": 0.17,
|
| 4302 |
+
"step": 713
|
| 4303 |
+
},
|
| 4304 |
+
{
|
| 4305 |
+
"epoch": 2.5,
|
| 4306 |
+
"learning_rate": 1.4421872769855262e-06,
|
| 4307 |
+
"loss": 0.1875,
|
| 4308 |
+
"step": 714
|
| 4309 |
+
},
|
| 4310 |
+
{
|
| 4311 |
+
"epoch": 2.5,
|
| 4312 |
+
"learning_rate": 1.4227138999972801e-06,
|
| 4313 |
+
"loss": 0.1725,
|
| 4314 |
+
"step": 715
|
| 4315 |
+
},
|
| 4316 |
+
{
|
| 4317 |
+
"epoch": 2.5,
|
| 4318 |
+
"learning_rate": 1.403362816255236e-06,
|
| 4319 |
+
"loss": 0.183,
|
| 4320 |
+
"step": 716
|
| 4321 |
+
},
|
| 4322 |
+
{
|
| 4323 |
+
"epoch": 2.51,
|
| 4324 |
+
"learning_rate": 1.3841343016633168e-06,
|
| 4325 |
+
"loss": 0.1806,
|
| 4326 |
+
"step": 717
|
| 4327 |
+
},
|
| 4328 |
+
{
|
| 4329 |
+
"epoch": 2.51,
|
| 4330 |
+
"learning_rate": 1.3650286303778715e-06,
|
| 4331 |
+
"loss": 0.1901,
|
| 4332 |
+
"step": 718
|
| 4333 |
+
},
|
| 4334 |
+
{
|
| 4335 |
+
"epoch": 2.51,
|
| 4336 |
+
"learning_rate": 1.3460460748037773e-06,
|
| 4337 |
+
"loss": 0.1834,
|
| 4338 |
+
"step": 719
|
| 4339 |
+
},
|
| 4340 |
+
{
|
| 4341 |
+
"epoch": 2.52,
|
| 4342 |
+
"learning_rate": 1.3271869055905495e-06,
|
| 4343 |
+
"loss": 0.1769,
|
| 4344 |
+
"step": 720
|
| 4345 |
}
|
| 4346 |
],
|
| 4347 |
"max_steps": 858,
|
| 4348 |
"num_train_epochs": 3,
|
| 4349 |
+
"total_flos": 1.8324177213299098e+18,
|
| 4350 |
"trial_name": null,
|
| 4351 |
"trial_params": null
|
| 4352 |
}
|