Farouk commited on
Commit ·
fa9e2bd
1
Parent(s): 94983f9
Training in progress, step 3800
Browse files- adapter_model.bin +1 -1
- {checkpoint-1600 → checkpoint-3600/adapter_model/adapter_model}/README.md +0 -0
- {checkpoint-1600 → checkpoint-3600/adapter_model/adapter_model}/adapter_config.json +0 -0
- {checkpoint-1600 → checkpoint-3600/adapter_model/adapter_model}/adapter_model.bin +1 -1
- {checkpoint-1600/adapter_model/adapter_model → checkpoint-3800}/README.md +0 -0
- {checkpoint-1600/adapter_model/adapter_model → checkpoint-3800}/adapter_config.json +0 -0
- {checkpoint-1600/adapter_model/adapter_model → checkpoint-3800}/adapter_model.bin +1 -1
- {checkpoint-1600 → checkpoint-3800}/added_tokens.json +0 -0
- {checkpoint-1600 → checkpoint-3800}/optimizer.pt +1 -1
- {checkpoint-1600 → checkpoint-3800}/rng_state.pth +1 -1
- {checkpoint-1600 → checkpoint-3800}/scheduler.pt +1 -1
- {checkpoint-1600 → checkpoint-3800}/special_tokens_map.json +0 -0
- {checkpoint-1600 → checkpoint-3800}/tokenizer.model +0 -0
- {checkpoint-1600 → checkpoint-3800}/tokenizer_config.json +0 -0
- {checkpoint-1600 → checkpoint-3800}/trainer_state.json +2106 -5
- {checkpoint-1600 → checkpoint-3800}/training_args.bin +0 -0
adapter_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63e6594e7065b5ddab685695820c8fc411772becad7644327c218c06705d934d
|
| 3 |
size 319977229
|
{checkpoint-1600 → checkpoint-3600/adapter_model/adapter_model}/README.md
RENAMED
|
File without changes
|
{checkpoint-1600 → checkpoint-3600/adapter_model/adapter_model}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-1600 → checkpoint-3600/adapter_model/adapter_model}/adapter_model.bin
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a44bb16c2131262914f9e8f6959aaea2f2ff70e5c22bb9d0488de268c42a2313
|
| 3 |
size 319977229
|
{checkpoint-1600/adapter_model/adapter_model → checkpoint-3800}/README.md
RENAMED
|
File without changes
|
{checkpoint-1600/adapter_model/adapter_model → checkpoint-3800}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-1600/adapter_model/adapter_model → checkpoint-3800}/adapter_model.bin
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 319977229
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63e6594e7065b5ddab685695820c8fc411772becad7644327c218c06705d934d
|
| 3 |
size 319977229
|
{checkpoint-1600 → checkpoint-3800}/added_tokens.json
RENAMED
|
File without changes
|
{checkpoint-1600 → checkpoint-3800}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1279539973
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eeff3c6a9327a8ae1df4007ba5c6e9f4dc86b1832048ba8b18476c7f996e53ac
|
| 3 |
size 1279539973
|
{checkpoint-1600 → checkpoint-3800}/rng_state.pth
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14511
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f70bc039340027e3e77f2cf1e0f4bf367543e8ba7196948de38bd59edd52845d
|
| 3 |
size 14511
|
{checkpoint-1600 → checkpoint-3800}/scheduler.pt
RENAMED
|
@@ -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 sha256:3961d2f201ce99426ee113655da78d8214f6ef72fc082855daa3c565989888c2
|
| 3 |
size 627
|
{checkpoint-1600 → checkpoint-3800}/special_tokens_map.json
RENAMED
|
File without changes
|
{checkpoint-1600 → checkpoint-3800}/tokenizer.model
RENAMED
|
File without changes
|
{checkpoint-1600 → checkpoint-3800}/tokenizer_config.json
RENAMED
|
File without changes
|
{checkpoint-1600 → checkpoint-3800}/trainer_state.json
RENAMED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "experts/expert-17/checkpoint-
|
| 4 |
-
"epoch": 0.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -1534,11 +1534,2112 @@
|
|
| 1534 |
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 1535 |
"mmlu_loss": 1.5103729819348837,
|
| 1536 |
"step": 1600
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1537 |
}
|
| 1538 |
],
|
| 1539 |
"max_steps": 10000,
|
| 1540 |
"num_train_epochs": 3,
|
| 1541 |
-
"total_flos":
|
| 1542 |
"trial_name": null,
|
| 1543 |
"trial_params": null
|
| 1544 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.38058969378471375,
|
| 3 |
+
"best_model_checkpoint": "experts/expert-17/checkpoint-3800",
|
| 4 |
+
"epoch": 0.7876055754184155,
|
| 5 |
+
"global_step": 3800,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 1534 |
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 1535 |
"mmlu_loss": 1.5103729819348837,
|
| 1536 |
"step": 1600
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.33,
|
| 1540 |
+
"learning_rate": 0.0002,
|
| 1541 |
+
"loss": 0.4223,
|
| 1542 |
+
"step": 1610
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.34,
|
| 1546 |
+
"learning_rate": 0.0002,
|
| 1547 |
+
"loss": 0.3835,
|
| 1548 |
+
"step": 1620
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 0.34,
|
| 1552 |
+
"learning_rate": 0.0002,
|
| 1553 |
+
"loss": 0.3958,
|
| 1554 |
+
"step": 1630
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.34,
|
| 1558 |
+
"learning_rate": 0.0002,
|
| 1559 |
+
"loss": 0.4213,
|
| 1560 |
+
"step": 1640
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.34,
|
| 1564 |
+
"learning_rate": 0.0002,
|
| 1565 |
+
"loss": 0.3742,
|
| 1566 |
+
"step": 1650
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.34,
|
| 1570 |
+
"learning_rate": 0.0002,
|
| 1571 |
+
"loss": 0.415,
|
| 1572 |
+
"step": 1660
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.35,
|
| 1576 |
+
"learning_rate": 0.0002,
|
| 1577 |
+
"loss": 0.4101,
|
| 1578 |
+
"step": 1670
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.35,
|
| 1582 |
+
"learning_rate": 0.0002,
|
| 1583 |
+
"loss": 0.45,
|
| 1584 |
+
"step": 1680
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.35,
|
| 1588 |
+
"learning_rate": 0.0002,
|
| 1589 |
+
"loss": 0.3688,
|
| 1590 |
+
"step": 1690
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 0.35,
|
| 1594 |
+
"learning_rate": 0.0002,
|
| 1595 |
+
"loss": 0.4063,
|
| 1596 |
+
"step": 1700
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 0.35,
|
| 1600 |
+
"learning_rate": 0.0002,
|
| 1601 |
+
"loss": 0.4455,
|
| 1602 |
+
"step": 1710
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.36,
|
| 1606 |
+
"learning_rate": 0.0002,
|
| 1607 |
+
"loss": 0.4502,
|
| 1608 |
+
"step": 1720
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.36,
|
| 1612 |
+
"learning_rate": 0.0002,
|
| 1613 |
+
"loss": 0.3972,
|
| 1614 |
+
"step": 1730
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.36,
|
| 1618 |
+
"learning_rate": 0.0002,
|
| 1619 |
+
"loss": 0.3691,
|
| 1620 |
+
"step": 1740
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.36,
|
| 1624 |
+
"learning_rate": 0.0002,
|
| 1625 |
+
"loss": 0.3852,
|
| 1626 |
+
"step": 1750
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.36,
|
| 1630 |
+
"learning_rate": 0.0002,
|
| 1631 |
+
"loss": 0.3858,
|
| 1632 |
+
"step": 1760
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 0.37,
|
| 1636 |
+
"learning_rate": 0.0002,
|
| 1637 |
+
"loss": 0.4137,
|
| 1638 |
+
"step": 1770
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 0.37,
|
| 1642 |
+
"learning_rate": 0.0002,
|
| 1643 |
+
"loss": 0.4141,
|
| 1644 |
+
"step": 1780
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 0.37,
|
| 1648 |
+
"learning_rate": 0.0002,
|
| 1649 |
+
"loss": 0.3969,
|
| 1650 |
+
"step": 1790
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.37,
|
| 1654 |
+
"learning_rate": 0.0002,
|
| 1655 |
+
"loss": 0.446,
|
| 1656 |
+
"step": 1800
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.37,
|
| 1660 |
+
"eval_loss": 0.4105340838432312,
|
| 1661 |
+
"eval_runtime": 150.5912,
|
| 1662 |
+
"eval_samples_per_second": 6.64,
|
| 1663 |
+
"eval_steps_per_second": 3.32,
|
| 1664 |
+
"step": 1800
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.37,
|
| 1668 |
+
"mmlu_eval_accuracy": 0.4912399811585959,
|
| 1669 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 1670 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 1671 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 1672 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 1673 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 1674 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 1675 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1676 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 1677 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 1678 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 1679 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 1680 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 1681 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 1682 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 1683 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 1684 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 1685 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 1686 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1687 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 1688 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 1689 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1690 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 1691 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 1692 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 1693 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 1694 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 1695 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 1696 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 1697 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 1698 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 1699 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 1700 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 1701 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 1702 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 1703 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 1704 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 1705 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 1706 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 1707 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 1708 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 1709 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 1710 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 1711 |
+
"mmlu_eval_accuracy_moral_disputes": 0.39473684210526316,
|
| 1712 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.23,
|
| 1713 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 1714 |
+
"mmlu_eval_accuracy_philosophy": 0.4411764705882353,
|
| 1715 |
+
"mmlu_eval_accuracy_prehistory": 0.45714285714285713,
|
| 1716 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 1717 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 1718 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 1719 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4927536231884058,
|
| 1720 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 1721 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1722 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 1723 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 1724 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 1725 |
+
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 1726 |
+
"mmlu_loss": 1.4265244243661666,
|
| 1727 |
+
"step": 1800
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.38,
|
| 1731 |
+
"learning_rate": 0.0002,
|
| 1732 |
+
"loss": 0.4121,
|
| 1733 |
+
"step": 1810
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.38,
|
| 1737 |
+
"learning_rate": 0.0002,
|
| 1738 |
+
"loss": 0.4347,
|
| 1739 |
+
"step": 1820
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 0.38,
|
| 1743 |
+
"learning_rate": 0.0002,
|
| 1744 |
+
"loss": 0.4474,
|
| 1745 |
+
"step": 1830
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.38,
|
| 1749 |
+
"learning_rate": 0.0002,
|
| 1750 |
+
"loss": 0.4006,
|
| 1751 |
+
"step": 1840
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 0.38,
|
| 1755 |
+
"learning_rate": 0.0002,
|
| 1756 |
+
"loss": 0.3988,
|
| 1757 |
+
"step": 1850
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 0.39,
|
| 1761 |
+
"learning_rate": 0.0002,
|
| 1762 |
+
"loss": 0.3596,
|
| 1763 |
+
"step": 1860
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.39,
|
| 1767 |
+
"learning_rate": 0.0002,
|
| 1768 |
+
"loss": 0.4356,
|
| 1769 |
+
"step": 1870
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.39,
|
| 1773 |
+
"learning_rate": 0.0002,
|
| 1774 |
+
"loss": 0.3909,
|
| 1775 |
+
"step": 1880
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 0.39,
|
| 1779 |
+
"learning_rate": 0.0002,
|
| 1780 |
+
"loss": 0.4389,
|
| 1781 |
+
"step": 1890
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 0.39,
|
| 1785 |
+
"learning_rate": 0.0002,
|
| 1786 |
+
"loss": 0.4139,
|
| 1787 |
+
"step": 1900
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.4,
|
| 1791 |
+
"learning_rate": 0.0002,
|
| 1792 |
+
"loss": 0.3697,
|
| 1793 |
+
"step": 1910
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 0.4,
|
| 1797 |
+
"learning_rate": 0.0002,
|
| 1798 |
+
"loss": 0.4176,
|
| 1799 |
+
"step": 1920
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 0.4,
|
| 1803 |
+
"learning_rate": 0.0002,
|
| 1804 |
+
"loss": 0.4291,
|
| 1805 |
+
"step": 1930
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 0.4,
|
| 1809 |
+
"learning_rate": 0.0002,
|
| 1810 |
+
"loss": 0.4733,
|
| 1811 |
+
"step": 1940
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 0.4,
|
| 1815 |
+
"learning_rate": 0.0002,
|
| 1816 |
+
"loss": 0.366,
|
| 1817 |
+
"step": 1950
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.41,
|
| 1821 |
+
"learning_rate": 0.0002,
|
| 1822 |
+
"loss": 0.3741,
|
| 1823 |
+
"step": 1960
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 0.41,
|
| 1827 |
+
"learning_rate": 0.0002,
|
| 1828 |
+
"loss": 0.4224,
|
| 1829 |
+
"step": 1970
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.41,
|
| 1833 |
+
"learning_rate": 0.0002,
|
| 1834 |
+
"loss": 0.3956,
|
| 1835 |
+
"step": 1980
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 0.41,
|
| 1839 |
+
"learning_rate": 0.0002,
|
| 1840 |
+
"loss": 0.3981,
|
| 1841 |
+
"step": 1990
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 0.41,
|
| 1845 |
+
"learning_rate": 0.0002,
|
| 1846 |
+
"loss": 0.3767,
|
| 1847 |
+
"step": 2000
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 0.41,
|
| 1851 |
+
"eval_loss": 0.4069203734397888,
|
| 1852 |
+
"eval_runtime": 150.4599,
|
| 1853 |
+
"eval_samples_per_second": 6.646,
|
| 1854 |
+
"eval_steps_per_second": 3.323,
|
| 1855 |
+
"step": 2000
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 0.41,
|
| 1859 |
+
"mmlu_eval_accuracy": 0.5015284318241597,
|
| 1860 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 1861 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 1862 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 1863 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 1864 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 1865 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 1866 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1867 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 1868 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 1869 |
+
"mmlu_eval_accuracy_college_medicine": 0.5,
|
| 1870 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 1871 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 1872 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 1873 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 1874 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 1875 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 1876 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 1877 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1878 |
+
"mmlu_eval_accuracy_high_school_biology": 0.375,
|
| 1879 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 1880 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1881 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 1882 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 1883 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 1884 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 1885 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 1886 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 1887 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 1888 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 1889 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 1890 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
|
| 1891 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 1892 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 1893 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 1894 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 1895 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 1896 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 1897 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 1898 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 1899 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 1900 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 1901 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 1902 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 1903 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 1904 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 1905 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 1906 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 1907 |
+
"mmlu_eval_accuracy_professional_accounting": 0.41935483870967744,
|
| 1908 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 1909 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 1910 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 1911 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 1912 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1913 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 1914 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 1915 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 1916 |
+
"mmlu_eval_accuracy_world_religions": 0.7894736842105263,
|
| 1917 |
+
"mmlu_loss": 1.5696932590194532,
|
| 1918 |
+
"step": 2000
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 0.42,
|
| 1922 |
+
"learning_rate": 0.0002,
|
| 1923 |
+
"loss": 0.4119,
|
| 1924 |
+
"step": 2010
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 0.42,
|
| 1928 |
+
"learning_rate": 0.0002,
|
| 1929 |
+
"loss": 0.4129,
|
| 1930 |
+
"step": 2020
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 0.42,
|
| 1934 |
+
"learning_rate": 0.0002,
|
| 1935 |
+
"loss": 0.4271,
|
| 1936 |
+
"step": 2030
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.42,
|
| 1940 |
+
"learning_rate": 0.0002,
|
| 1941 |
+
"loss": 0.393,
|
| 1942 |
+
"step": 2040
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 0.42,
|
| 1946 |
+
"learning_rate": 0.0002,
|
| 1947 |
+
"loss": 0.405,
|
| 1948 |
+
"step": 2050
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.43,
|
| 1952 |
+
"learning_rate": 0.0002,
|
| 1953 |
+
"loss": 0.4414,
|
| 1954 |
+
"step": 2060
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 0.43,
|
| 1958 |
+
"learning_rate": 0.0002,
|
| 1959 |
+
"loss": 0.3868,
|
| 1960 |
+
"step": 2070
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 0.43,
|
| 1964 |
+
"learning_rate": 0.0002,
|
| 1965 |
+
"loss": 0.3945,
|
| 1966 |
+
"step": 2080
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 0.43,
|
| 1970 |
+
"learning_rate": 0.0002,
|
| 1971 |
+
"loss": 0.3871,
|
| 1972 |
+
"step": 2090
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 0.44,
|
| 1976 |
+
"learning_rate": 0.0002,
|
| 1977 |
+
"loss": 0.3874,
|
| 1978 |
+
"step": 2100
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 0.44,
|
| 1982 |
+
"learning_rate": 0.0002,
|
| 1983 |
+
"loss": 0.394,
|
| 1984 |
+
"step": 2110
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.44,
|
| 1988 |
+
"learning_rate": 0.0002,
|
| 1989 |
+
"loss": 0.4386,
|
| 1990 |
+
"step": 2120
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.44,
|
| 1994 |
+
"learning_rate": 0.0002,
|
| 1995 |
+
"loss": 0.378,
|
| 1996 |
+
"step": 2130
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 0.44,
|
| 2000 |
+
"learning_rate": 0.0002,
|
| 2001 |
+
"loss": 0.4146,
|
| 2002 |
+
"step": 2140
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 0.45,
|
| 2006 |
+
"learning_rate": 0.0002,
|
| 2007 |
+
"loss": 0.3705,
|
| 2008 |
+
"step": 2150
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 0.45,
|
| 2012 |
+
"learning_rate": 0.0002,
|
| 2013 |
+
"loss": 0.3755,
|
| 2014 |
+
"step": 2160
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 0.45,
|
| 2018 |
+
"learning_rate": 0.0002,
|
| 2019 |
+
"loss": 0.43,
|
| 2020 |
+
"step": 2170
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 0.45,
|
| 2024 |
+
"learning_rate": 0.0002,
|
| 2025 |
+
"loss": 0.3936,
|
| 2026 |
+
"step": 2180
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 0.45,
|
| 2030 |
+
"learning_rate": 0.0002,
|
| 2031 |
+
"loss": 0.402,
|
| 2032 |
+
"step": 2190
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.46,
|
| 2036 |
+
"learning_rate": 0.0002,
|
| 2037 |
+
"loss": 0.4202,
|
| 2038 |
+
"step": 2200
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 0.46,
|
| 2042 |
+
"eval_loss": 0.40249407291412354,
|
| 2043 |
+
"eval_runtime": 150.6022,
|
| 2044 |
+
"eval_samples_per_second": 6.64,
|
| 2045 |
+
"eval_steps_per_second": 3.32,
|
| 2046 |
+
"step": 2200
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.46,
|
| 2050 |
+
"mmlu_eval_accuracy": 0.4967557687715033,
|
| 2051 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 2052 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2053 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 2054 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 2055 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 2056 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 2057 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2058 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 2059 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 2060 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 2061 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 2062 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2063 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
|
| 2064 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 2065 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 2066 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 2067 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 2068 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 2069 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 2070 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 2071 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2072 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 2073 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 2074 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 2075 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 2076 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2077 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
|
| 2078 |
+
"mmlu_eval_accuracy_high_school_physics": 0.058823529411764705,
|
| 2079 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 2080 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 2081 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2082 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 2083 |
+
"mmlu_eval_accuracy_human_aging": 0.6521739130434783,
|
| 2084 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 2085 |
+
"mmlu_eval_accuracy_international_law": 0.9230769230769231,
|
| 2086 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 2087 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 2088 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 2089 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 2090 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 2091 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 2092 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
|
| 2093 |
+
"mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
|
| 2094 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 2095 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 2096 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 2097 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2098 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 2099 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 2100 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 2101 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 2102 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 2103 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 2104 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 2105 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 2106 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 2107 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2108 |
+
"mmlu_loss": 1.5850450227294828,
|
| 2109 |
+
"step": 2200
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.46,
|
| 2113 |
+
"learning_rate": 0.0002,
|
| 2114 |
+
"loss": 0.362,
|
| 2115 |
+
"step": 2210
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 0.46,
|
| 2119 |
+
"learning_rate": 0.0002,
|
| 2120 |
+
"loss": 0.4067,
|
| 2121 |
+
"step": 2220
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.46,
|
| 2125 |
+
"learning_rate": 0.0002,
|
| 2126 |
+
"loss": 0.3897,
|
| 2127 |
+
"step": 2230
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 0.46,
|
| 2131 |
+
"learning_rate": 0.0002,
|
| 2132 |
+
"loss": 0.4179,
|
| 2133 |
+
"step": 2240
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 0.47,
|
| 2137 |
+
"learning_rate": 0.0002,
|
| 2138 |
+
"loss": 0.4071,
|
| 2139 |
+
"step": 2250
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 0.47,
|
| 2143 |
+
"learning_rate": 0.0002,
|
| 2144 |
+
"loss": 0.3906,
|
| 2145 |
+
"step": 2260
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.47,
|
| 2149 |
+
"learning_rate": 0.0002,
|
| 2150 |
+
"loss": 0.401,
|
| 2151 |
+
"step": 2270
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.47,
|
| 2155 |
+
"learning_rate": 0.0002,
|
| 2156 |
+
"loss": 0.3574,
|
| 2157 |
+
"step": 2280
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 0.47,
|
| 2161 |
+
"learning_rate": 0.0002,
|
| 2162 |
+
"loss": 0.4548,
|
| 2163 |
+
"step": 2290
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.48,
|
| 2167 |
+
"learning_rate": 0.0002,
|
| 2168 |
+
"loss": 0.4031,
|
| 2169 |
+
"step": 2300
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 0.48,
|
| 2173 |
+
"learning_rate": 0.0002,
|
| 2174 |
+
"loss": 0.4001,
|
| 2175 |
+
"step": 2310
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 0.48,
|
| 2179 |
+
"learning_rate": 0.0002,
|
| 2180 |
+
"loss": 0.378,
|
| 2181 |
+
"step": 2320
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 0.48,
|
| 2185 |
+
"learning_rate": 0.0002,
|
| 2186 |
+
"loss": 0.4307,
|
| 2187 |
+
"step": 2330
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.48,
|
| 2191 |
+
"learning_rate": 0.0002,
|
| 2192 |
+
"loss": 0.4357,
|
| 2193 |
+
"step": 2340
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.49,
|
| 2197 |
+
"learning_rate": 0.0002,
|
| 2198 |
+
"loss": 0.4138,
|
| 2199 |
+
"step": 2350
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 0.49,
|
| 2203 |
+
"learning_rate": 0.0002,
|
| 2204 |
+
"loss": 0.3981,
|
| 2205 |
+
"step": 2360
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 0.49,
|
| 2209 |
+
"learning_rate": 0.0002,
|
| 2210 |
+
"loss": 0.4251,
|
| 2211 |
+
"step": 2370
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 0.49,
|
| 2215 |
+
"learning_rate": 0.0002,
|
| 2216 |
+
"loss": 0.3681,
|
| 2217 |
+
"step": 2380
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 0.5,
|
| 2221 |
+
"learning_rate": 0.0002,
|
| 2222 |
+
"loss": 0.4395,
|
| 2223 |
+
"step": 2390
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 0.5,
|
| 2227 |
+
"learning_rate": 0.0002,
|
| 2228 |
+
"loss": 0.4343,
|
| 2229 |
+
"step": 2400
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.5,
|
| 2233 |
+
"eval_loss": 0.4002082943916321,
|
| 2234 |
+
"eval_runtime": 150.6447,
|
| 2235 |
+
"eval_samples_per_second": 6.638,
|
| 2236 |
+
"eval_steps_per_second": 3.319,
|
| 2237 |
+
"step": 2400
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 0.5,
|
| 2241 |
+
"mmlu_eval_accuracy": 0.4960865437362044,
|
| 2242 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
|
| 2243 |
+
"mmlu_eval_accuracy_anatomy": 0.5714285714285714,
|
| 2244 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 2245 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 2246 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 2247 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 2248 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2249 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 2250 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 2251 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 2252 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 2253 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2254 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 2255 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 2256 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 2257 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 2258 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 2259 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2260 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 2261 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 2262 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2263 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 2264 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 2265 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 2266 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 2267 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2268 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 2269 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 2270 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 2271 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 2272 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2273 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 2274 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 2275 |
+
"mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
|
| 2276 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 2277 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 2278 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 2279 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 2280 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 2281 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 2282 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 2283 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 2284 |
+
"mmlu_eval_accuracy_moral_disputes": 0.39473684210526316,
|
| 2285 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 2286 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 2287 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 2288 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2289 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 2290 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 2291 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 2292 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4492753623188406,
|
| 2293 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 2294 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 2295 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 2296 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 2297 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 2298 |
+
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 2299 |
+
"mmlu_loss": 1.4704214497116777,
|
| 2300 |
+
"step": 2400
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 0.5,
|
| 2304 |
+
"learning_rate": 0.0002,
|
| 2305 |
+
"loss": 0.3321,
|
| 2306 |
+
"step": 2410
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.5,
|
| 2310 |
+
"learning_rate": 0.0002,
|
| 2311 |
+
"loss": 0.4149,
|
| 2312 |
+
"step": 2420
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.5,
|
| 2316 |
+
"learning_rate": 0.0002,
|
| 2317 |
+
"loss": 0.3812,
|
| 2318 |
+
"step": 2430
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 0.51,
|
| 2322 |
+
"learning_rate": 0.0002,
|
| 2323 |
+
"loss": 0.3449,
|
| 2324 |
+
"step": 2440
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 0.51,
|
| 2328 |
+
"learning_rate": 0.0002,
|
| 2329 |
+
"loss": 0.4206,
|
| 2330 |
+
"step": 2450
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 0.51,
|
| 2334 |
+
"learning_rate": 0.0002,
|
| 2335 |
+
"loss": 0.3967,
|
| 2336 |
+
"step": 2460
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 0.51,
|
| 2340 |
+
"learning_rate": 0.0002,
|
| 2341 |
+
"loss": 0.4176,
|
| 2342 |
+
"step": 2470
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 0.51,
|
| 2346 |
+
"learning_rate": 0.0002,
|
| 2347 |
+
"loss": 0.3605,
|
| 2348 |
+
"step": 2480
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.52,
|
| 2352 |
+
"learning_rate": 0.0002,
|
| 2353 |
+
"loss": 0.424,
|
| 2354 |
+
"step": 2490
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.52,
|
| 2358 |
+
"learning_rate": 0.0002,
|
| 2359 |
+
"loss": 0.3737,
|
| 2360 |
+
"step": 2500
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 0.52,
|
| 2364 |
+
"learning_rate": 0.0002,
|
| 2365 |
+
"loss": 0.3723,
|
| 2366 |
+
"step": 2510
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 0.52,
|
| 2370 |
+
"learning_rate": 0.0002,
|
| 2371 |
+
"loss": 0.4035,
|
| 2372 |
+
"step": 2520
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 0.52,
|
| 2376 |
+
"learning_rate": 0.0002,
|
| 2377 |
+
"loss": 0.4206,
|
| 2378 |
+
"step": 2530
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 0.53,
|
| 2382 |
+
"learning_rate": 0.0002,
|
| 2383 |
+
"loss": 0.4349,
|
| 2384 |
+
"step": 2540
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 0.53,
|
| 2388 |
+
"learning_rate": 0.0002,
|
| 2389 |
+
"loss": 0.443,
|
| 2390 |
+
"step": 2550
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.53,
|
| 2394 |
+
"learning_rate": 0.0002,
|
| 2395 |
+
"loss": 0.4217,
|
| 2396 |
+
"step": 2560
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.53,
|
| 2400 |
+
"learning_rate": 0.0002,
|
| 2401 |
+
"loss": 0.447,
|
| 2402 |
+
"step": 2570
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 0.53,
|
| 2406 |
+
"learning_rate": 0.0002,
|
| 2407 |
+
"loss": 0.3828,
|
| 2408 |
+
"step": 2580
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 0.54,
|
| 2412 |
+
"learning_rate": 0.0002,
|
| 2413 |
+
"loss": 0.3725,
|
| 2414 |
+
"step": 2590
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 0.54,
|
| 2418 |
+
"learning_rate": 0.0002,
|
| 2419 |
+
"loss": 0.3936,
|
| 2420 |
+
"step": 2600
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 0.54,
|
| 2424 |
+
"eval_loss": 0.3957899808883667,
|
| 2425 |
+
"eval_runtime": 150.681,
|
| 2426 |
+
"eval_samples_per_second": 6.637,
|
| 2427 |
+
"eval_steps_per_second": 3.318,
|
| 2428 |
+
"step": 2600
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 0.54,
|
| 2432 |
+
"mmlu_eval_accuracy": 0.49736669283307794,
|
| 2433 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
|
| 2434 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2435 |
+
"mmlu_eval_accuracy_astronomy": 0.625,
|
| 2436 |
+
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
|
| 2437 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 2438 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 2439 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2440 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2441 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 2442 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 2443 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 2444 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 2445 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 2446 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 2447 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 2448 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
|
| 2449 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 2450 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2451 |
+
"mmlu_eval_accuracy_high_school_biology": 0.34375,
|
| 2452 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 2453 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2454 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 2455 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 2456 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 2457 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
|
| 2458 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2459 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 2460 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 2461 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 2462 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 2463 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2464 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 2465 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 2466 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 2467 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 2468 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 2469 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 2470 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 2471 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 2472 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 2473 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 2474 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 2475 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 2476 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 2477 |
+
"mmlu_eval_accuracy_nutrition": 0.6060606060606061,
|
| 2478 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 2479 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2480 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 2481 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 2482 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 2483 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 2484 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 2485 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 2486 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 2487 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 2488 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 2489 |
+
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 2490 |
+
"mmlu_loss": 1.5555474590850562,
|
| 2491 |
+
"step": 2600
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 0.54,
|
| 2495 |
+
"learning_rate": 0.0002,
|
| 2496 |
+
"loss": 0.4358,
|
| 2497 |
+
"step": 2610
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 0.54,
|
| 2501 |
+
"learning_rate": 0.0002,
|
| 2502 |
+
"loss": 0.391,
|
| 2503 |
+
"step": 2620
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 0.55,
|
| 2507 |
+
"learning_rate": 0.0002,
|
| 2508 |
+
"loss": 0.3766,
|
| 2509 |
+
"step": 2630
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.55,
|
| 2513 |
+
"learning_rate": 0.0002,
|
| 2514 |
+
"loss": 0.4347,
|
| 2515 |
+
"step": 2640
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 0.55,
|
| 2519 |
+
"learning_rate": 0.0002,
|
| 2520 |
+
"loss": 0.3926,
|
| 2521 |
+
"step": 2650
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 0.55,
|
| 2525 |
+
"learning_rate": 0.0002,
|
| 2526 |
+
"loss": 0.4063,
|
| 2527 |
+
"step": 2660
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 0.55,
|
| 2531 |
+
"learning_rate": 0.0002,
|
| 2532 |
+
"loss": 0.3824,
|
| 2533 |
+
"step": 2670
|
| 2534 |
+
},
|
| 2535 |
+
{
|
| 2536 |
+
"epoch": 0.56,
|
| 2537 |
+
"learning_rate": 0.0002,
|
| 2538 |
+
"loss": 0.4083,
|
| 2539 |
+
"step": 2680
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 0.56,
|
| 2543 |
+
"learning_rate": 0.0002,
|
| 2544 |
+
"loss": 0.389,
|
| 2545 |
+
"step": 2690
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 0.56,
|
| 2549 |
+
"learning_rate": 0.0002,
|
| 2550 |
+
"loss": 0.403,
|
| 2551 |
+
"step": 2700
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.56,
|
| 2555 |
+
"learning_rate": 0.0002,
|
| 2556 |
+
"loss": 0.427,
|
| 2557 |
+
"step": 2710
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 0.56,
|
| 2561 |
+
"learning_rate": 0.0002,
|
| 2562 |
+
"loss": 0.3788,
|
| 2563 |
+
"step": 2720
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 0.57,
|
| 2567 |
+
"learning_rate": 0.0002,
|
| 2568 |
+
"loss": 0.3701,
|
| 2569 |
+
"step": 2730
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.57,
|
| 2573 |
+
"learning_rate": 0.0002,
|
| 2574 |
+
"loss": 0.3759,
|
| 2575 |
+
"step": 2740
|
| 2576 |
+
},
|
| 2577 |
+
{
|
| 2578 |
+
"epoch": 0.57,
|
| 2579 |
+
"learning_rate": 0.0002,
|
| 2580 |
+
"loss": 0.3881,
|
| 2581 |
+
"step": 2750
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 0.57,
|
| 2585 |
+
"learning_rate": 0.0002,
|
| 2586 |
+
"loss": 0.3527,
|
| 2587 |
+
"step": 2760
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 0.57,
|
| 2591 |
+
"learning_rate": 0.0002,
|
| 2592 |
+
"loss": 0.3784,
|
| 2593 |
+
"step": 2770
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.58,
|
| 2597 |
+
"learning_rate": 0.0002,
|
| 2598 |
+
"loss": 0.4519,
|
| 2599 |
+
"step": 2780
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 0.58,
|
| 2603 |
+
"learning_rate": 0.0002,
|
| 2604 |
+
"loss": 0.3642,
|
| 2605 |
+
"step": 2790
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 0.58,
|
| 2609 |
+
"learning_rate": 0.0002,
|
| 2610 |
+
"loss": 0.3639,
|
| 2611 |
+
"step": 2800
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 0.58,
|
| 2615 |
+
"eval_loss": 0.3921334445476532,
|
| 2616 |
+
"eval_runtime": 150.5221,
|
| 2617 |
+
"eval_samples_per_second": 6.644,
|
| 2618 |
+
"eval_steps_per_second": 3.322,
|
| 2619 |
+
"step": 2800
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 0.58,
|
| 2623 |
+
"mmlu_eval_accuracy": 0.49279284075798974,
|
| 2624 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
|
| 2625 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2626 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 2627 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 2628 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 2629 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 2630 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2631 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2632 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 2633 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 2634 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 2635 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 2636 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
|
| 2637 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 2638 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 2639 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 2640 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 2641 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 2642 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 2643 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 2644 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2645 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 2646 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 2647 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
|
| 2648 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 2649 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 2650 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 2651 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 2652 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 2653 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 2654 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2655 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 2656 |
+
"mmlu_eval_accuracy_human_aging": 0.6956521739130435,
|
| 2657 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 2658 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 2659 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 2660 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 2661 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 2662 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 2663 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 2664 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 2665 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
|
| 2666 |
+
"mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
|
| 2667 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.23,
|
| 2668 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 2669 |
+
"mmlu_eval_accuracy_philosophy": 0.5882352941176471,
|
| 2670 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 2671 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 2672 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 2673 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 2674 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4492753623188406,
|
| 2675 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 2676 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 2677 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 2678 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 2679 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 2680 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2681 |
+
"mmlu_loss": 1.5415999452843678,
|
| 2682 |
+
"step": 2800
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 0.58,
|
| 2686 |
+
"learning_rate": 0.0002,
|
| 2687 |
+
"loss": 0.418,
|
| 2688 |
+
"step": 2810
|
| 2689 |
+
},
|
| 2690 |
+
{
|
| 2691 |
+
"epoch": 0.58,
|
| 2692 |
+
"learning_rate": 0.0002,
|
| 2693 |
+
"loss": 0.394,
|
| 2694 |
+
"step": 2820
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 0.59,
|
| 2698 |
+
"learning_rate": 0.0002,
|
| 2699 |
+
"loss": 0.3812,
|
| 2700 |
+
"step": 2830
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 0.59,
|
| 2704 |
+
"learning_rate": 0.0002,
|
| 2705 |
+
"loss": 0.4361,
|
| 2706 |
+
"step": 2840
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 0.59,
|
| 2710 |
+
"learning_rate": 0.0002,
|
| 2711 |
+
"loss": 0.3932,
|
| 2712 |
+
"step": 2850
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.59,
|
| 2716 |
+
"learning_rate": 0.0002,
|
| 2717 |
+
"loss": 0.3849,
|
| 2718 |
+
"step": 2860
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 0.59,
|
| 2722 |
+
"learning_rate": 0.0002,
|
| 2723 |
+
"loss": 0.3698,
|
| 2724 |
+
"step": 2870
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 0.6,
|
| 2728 |
+
"learning_rate": 0.0002,
|
| 2729 |
+
"loss": 0.3606,
|
| 2730 |
+
"step": 2880
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 0.6,
|
| 2734 |
+
"learning_rate": 0.0002,
|
| 2735 |
+
"loss": 0.4365,
|
| 2736 |
+
"step": 2890
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 0.6,
|
| 2740 |
+
"learning_rate": 0.0002,
|
| 2741 |
+
"loss": 0.3921,
|
| 2742 |
+
"step": 2900
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 0.6,
|
| 2746 |
+
"learning_rate": 0.0002,
|
| 2747 |
+
"loss": 0.4072,
|
| 2748 |
+
"step": 2910
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 0.61,
|
| 2752 |
+
"learning_rate": 0.0002,
|
| 2753 |
+
"loss": 0.4054,
|
| 2754 |
+
"step": 2920
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.61,
|
| 2758 |
+
"learning_rate": 0.0002,
|
| 2759 |
+
"loss": 0.3939,
|
| 2760 |
+
"step": 2930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 0.61,
|
| 2764 |
+
"learning_rate": 0.0002,
|
| 2765 |
+
"loss": 0.343,
|
| 2766 |
+
"step": 2940
|
| 2767 |
+
},
|
| 2768 |
+
{
|
| 2769 |
+
"epoch": 0.61,
|
| 2770 |
+
"learning_rate": 0.0002,
|
| 2771 |
+
"loss": 0.4022,
|
| 2772 |
+
"step": 2950
|
| 2773 |
+
},
|
| 2774 |
+
{
|
| 2775 |
+
"epoch": 0.61,
|
| 2776 |
+
"learning_rate": 0.0002,
|
| 2777 |
+
"loss": 0.424,
|
| 2778 |
+
"step": 2960
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 0.62,
|
| 2782 |
+
"learning_rate": 0.0002,
|
| 2783 |
+
"loss": 0.3498,
|
| 2784 |
+
"step": 2970
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 0.62,
|
| 2788 |
+
"learning_rate": 0.0002,
|
| 2789 |
+
"loss": 0.4219,
|
| 2790 |
+
"step": 2980
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 0.62,
|
| 2794 |
+
"learning_rate": 0.0002,
|
| 2795 |
+
"loss": 0.3896,
|
| 2796 |
+
"step": 2990
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.62,
|
| 2800 |
+
"learning_rate": 0.0002,
|
| 2801 |
+
"loss": 0.3983,
|
| 2802 |
+
"step": 3000
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 0.62,
|
| 2806 |
+
"eval_loss": 0.3904864192008972,
|
| 2807 |
+
"eval_runtime": 150.4927,
|
| 2808 |
+
"eval_samples_per_second": 6.645,
|
| 2809 |
+
"eval_steps_per_second": 3.322,
|
| 2810 |
+
"step": 3000
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.62,
|
| 2814 |
+
"mmlu_eval_accuracy": 0.5005240738793235,
|
| 2815 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 2816 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2817 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 2818 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 2819 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 2820 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 2821 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2822 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2823 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 2824 |
+
"mmlu_eval_accuracy_college_medicine": 0.5454545454545454,
|
| 2825 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 2826 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2827 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
|
| 2828 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 2829 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 2830 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 2831 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 2832 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2833 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 2834 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
|
| 2835 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2836 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 2837 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 2838 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 2839 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 2840 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 2841 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
|
| 2842 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 2843 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 2844 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 2845 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2846 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 2847 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 2848 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 2849 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 2850 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 2851 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 2852 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 2853 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 2854 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 2855 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 2856 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 2857 |
+
"mmlu_eval_accuracy_moral_disputes": 0.39473684210526316,
|
| 2858 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 2859 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 2860 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 2861 |
+
"mmlu_eval_accuracy_prehistory": 0.5428571428571428,
|
| 2862 |
+
"mmlu_eval_accuracy_professional_accounting": 0.45161290322580644,
|
| 2863 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 2864 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 2865 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4782608695652174,
|
| 2866 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 2867 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 2868 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 2869 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 2870 |
+
"mmlu_eval_accuracy_virology": 0.4444444444444444,
|
| 2871 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2872 |
+
"mmlu_loss": 1.651317346811917,
|
| 2873 |
+
"step": 3000
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.62,
|
| 2877 |
+
"learning_rate": 0.0002,
|
| 2878 |
+
"loss": 0.3804,
|
| 2879 |
+
"step": 3010
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 0.63,
|
| 2883 |
+
"learning_rate": 0.0002,
|
| 2884 |
+
"loss": 0.3682,
|
| 2885 |
+
"step": 3020
|
| 2886 |
+
},
|
| 2887 |
+
{
|
| 2888 |
+
"epoch": 0.63,
|
| 2889 |
+
"learning_rate": 0.0002,
|
| 2890 |
+
"loss": 0.4221,
|
| 2891 |
+
"step": 3030
|
| 2892 |
+
},
|
| 2893 |
+
{
|
| 2894 |
+
"epoch": 0.63,
|
| 2895 |
+
"learning_rate": 0.0002,
|
| 2896 |
+
"loss": 0.3889,
|
| 2897 |
+
"step": 3040
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 0.63,
|
| 2901 |
+
"learning_rate": 0.0002,
|
| 2902 |
+
"loss": 0.3717,
|
| 2903 |
+
"step": 3050
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 0.63,
|
| 2907 |
+
"learning_rate": 0.0002,
|
| 2908 |
+
"loss": 0.4352,
|
| 2909 |
+
"step": 3060
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 0.64,
|
| 2913 |
+
"learning_rate": 0.0002,
|
| 2914 |
+
"loss": 0.3694,
|
| 2915 |
+
"step": 3070
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.64,
|
| 2919 |
+
"learning_rate": 0.0002,
|
| 2920 |
+
"loss": 0.4021,
|
| 2921 |
+
"step": 3080
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 0.64,
|
| 2925 |
+
"learning_rate": 0.0002,
|
| 2926 |
+
"loss": 0.3549,
|
| 2927 |
+
"step": 3090
|
| 2928 |
+
},
|
| 2929 |
+
{
|
| 2930 |
+
"epoch": 0.64,
|
| 2931 |
+
"learning_rate": 0.0002,
|
| 2932 |
+
"loss": 0.3725,
|
| 2933 |
+
"step": 3100
|
| 2934 |
+
},
|
| 2935 |
+
{
|
| 2936 |
+
"epoch": 0.64,
|
| 2937 |
+
"learning_rate": 0.0002,
|
| 2938 |
+
"loss": 0.3616,
|
| 2939 |
+
"step": 3110
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 0.65,
|
| 2943 |
+
"learning_rate": 0.0002,
|
| 2944 |
+
"loss": 0.3776,
|
| 2945 |
+
"step": 3120
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 0.65,
|
| 2949 |
+
"learning_rate": 0.0002,
|
| 2950 |
+
"loss": 0.4128,
|
| 2951 |
+
"step": 3130
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 0.65,
|
| 2955 |
+
"learning_rate": 0.0002,
|
| 2956 |
+
"loss": 0.387,
|
| 2957 |
+
"step": 3140
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 0.65,
|
| 2961 |
+
"learning_rate": 0.0002,
|
| 2962 |
+
"loss": 0.3735,
|
| 2963 |
+
"step": 3150
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 0.65,
|
| 2967 |
+
"learning_rate": 0.0002,
|
| 2968 |
+
"loss": 0.3716,
|
| 2969 |
+
"step": 3160
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 0.66,
|
| 2973 |
+
"learning_rate": 0.0002,
|
| 2974 |
+
"loss": 0.3528,
|
| 2975 |
+
"step": 3170
|
| 2976 |
+
},
|
| 2977 |
+
{
|
| 2978 |
+
"epoch": 0.66,
|
| 2979 |
+
"learning_rate": 0.0002,
|
| 2980 |
+
"loss": 0.3752,
|
| 2981 |
+
"step": 3180
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 0.66,
|
| 2985 |
+
"learning_rate": 0.0002,
|
| 2986 |
+
"loss": 0.4204,
|
| 2987 |
+
"step": 3190
|
| 2988 |
+
},
|
| 2989 |
+
{
|
| 2990 |
+
"epoch": 0.66,
|
| 2991 |
+
"learning_rate": 0.0002,
|
| 2992 |
+
"loss": 0.4278,
|
| 2993 |
+
"step": 3200
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 0.66,
|
| 2997 |
+
"eval_loss": 0.38523146510124207,
|
| 2998 |
+
"eval_runtime": 150.465,
|
| 2999 |
+
"eval_samples_per_second": 6.646,
|
| 3000 |
+
"eval_steps_per_second": 3.323,
|
| 3001 |
+
"step": 3200
|
| 3002 |
+
},
|
| 3003 |
+
{
|
| 3004 |
+
"epoch": 0.66,
|
| 3005 |
+
"mmlu_eval_accuracy": 0.49974702582501596,
|
| 3006 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
|
| 3007 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 3008 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 3009 |
+
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
|
| 3010 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 3011 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 3012 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 3013 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 3014 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 3015 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 3016 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 3017 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 3018 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 3019 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 3020 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 3021 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 3022 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 3023 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 3024 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 3025 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 3026 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 3027 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 3028 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 3029 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 3030 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
|
| 3031 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 3032 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 3033 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 3034 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 3035 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 3036 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 3037 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 3038 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 3039 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 3040 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 3041 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 3042 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 3043 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 3044 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 3045 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 3046 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 3047 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
|
| 3048 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 3049 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 3050 |
+
"mmlu_eval_accuracy_nutrition": 0.6363636363636364,
|
| 3051 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 3052 |
+
"mmlu_eval_accuracy_prehistory": 0.5428571428571428,
|
| 3053 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 3054 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 3055 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 3056 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 3057 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 3058 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 3059 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 3060 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 3061 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 3062 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 3063 |
+
"mmlu_loss": 1.5921483102884368,
|
| 3064 |
+
"step": 3200
|
| 3065 |
+
},
|
| 3066 |
+
{
|
| 3067 |
+
"epoch": 0.67,
|
| 3068 |
+
"learning_rate": 0.0002,
|
| 3069 |
+
"loss": 0.3874,
|
| 3070 |
+
"step": 3210
|
| 3071 |
+
},
|
| 3072 |
+
{
|
| 3073 |
+
"epoch": 0.67,
|
| 3074 |
+
"learning_rate": 0.0002,
|
| 3075 |
+
"loss": 0.3961,
|
| 3076 |
+
"step": 3220
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.67,
|
| 3080 |
+
"learning_rate": 0.0002,
|
| 3081 |
+
"loss": 0.3815,
|
| 3082 |
+
"step": 3230
|
| 3083 |
+
},
|
| 3084 |
+
{
|
| 3085 |
+
"epoch": 0.67,
|
| 3086 |
+
"learning_rate": 0.0002,
|
| 3087 |
+
"loss": 0.3736,
|
| 3088 |
+
"step": 3240
|
| 3089 |
+
},
|
| 3090 |
+
{
|
| 3091 |
+
"epoch": 0.67,
|
| 3092 |
+
"learning_rate": 0.0002,
|
| 3093 |
+
"loss": 0.4002,
|
| 3094 |
+
"step": 3250
|
| 3095 |
+
},
|
| 3096 |
+
{
|
| 3097 |
+
"epoch": 0.68,
|
| 3098 |
+
"learning_rate": 0.0002,
|
| 3099 |
+
"loss": 0.3724,
|
| 3100 |
+
"step": 3260
|
| 3101 |
+
},
|
| 3102 |
+
{
|
| 3103 |
+
"epoch": 0.68,
|
| 3104 |
+
"learning_rate": 0.0002,
|
| 3105 |
+
"loss": 0.4034,
|
| 3106 |
+
"step": 3270
|
| 3107 |
+
},
|
| 3108 |
+
{
|
| 3109 |
+
"epoch": 0.68,
|
| 3110 |
+
"learning_rate": 0.0002,
|
| 3111 |
+
"loss": 0.3812,
|
| 3112 |
+
"step": 3280
|
| 3113 |
+
},
|
| 3114 |
+
{
|
| 3115 |
+
"epoch": 0.68,
|
| 3116 |
+
"learning_rate": 0.0002,
|
| 3117 |
+
"loss": 0.3777,
|
| 3118 |
+
"step": 3290
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.68,
|
| 3122 |
+
"learning_rate": 0.0002,
|
| 3123 |
+
"loss": 0.349,
|
| 3124 |
+
"step": 3300
|
| 3125 |
+
},
|
| 3126 |
+
{
|
| 3127 |
+
"epoch": 0.69,
|
| 3128 |
+
"learning_rate": 0.0002,
|
| 3129 |
+
"loss": 0.3678,
|
| 3130 |
+
"step": 3310
|
| 3131 |
+
},
|
| 3132 |
+
{
|
| 3133 |
+
"epoch": 0.69,
|
| 3134 |
+
"learning_rate": 0.0002,
|
| 3135 |
+
"loss": 0.4193,
|
| 3136 |
+
"step": 3320
|
| 3137 |
+
},
|
| 3138 |
+
{
|
| 3139 |
+
"epoch": 0.69,
|
| 3140 |
+
"learning_rate": 0.0002,
|
| 3141 |
+
"loss": 0.3874,
|
| 3142 |
+
"step": 3330
|
| 3143 |
+
},
|
| 3144 |
+
{
|
| 3145 |
+
"epoch": 0.69,
|
| 3146 |
+
"learning_rate": 0.0002,
|
| 3147 |
+
"loss": 0.394,
|
| 3148 |
+
"step": 3340
|
| 3149 |
+
},
|
| 3150 |
+
{
|
| 3151 |
+
"epoch": 0.69,
|
| 3152 |
+
"learning_rate": 0.0002,
|
| 3153 |
+
"loss": 0.3834,
|
| 3154 |
+
"step": 3350
|
| 3155 |
+
},
|
| 3156 |
+
{
|
| 3157 |
+
"epoch": 0.7,
|
| 3158 |
+
"learning_rate": 0.0002,
|
| 3159 |
+
"loss": 0.3857,
|
| 3160 |
+
"step": 3360
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 0.7,
|
| 3164 |
+
"learning_rate": 0.0002,
|
| 3165 |
+
"loss": 0.4303,
|
| 3166 |
+
"step": 3370
|
| 3167 |
+
},
|
| 3168 |
+
{
|
| 3169 |
+
"epoch": 0.7,
|
| 3170 |
+
"learning_rate": 0.0002,
|
| 3171 |
+
"loss": 0.3658,
|
| 3172 |
+
"step": 3380
|
| 3173 |
+
},
|
| 3174 |
+
{
|
| 3175 |
+
"epoch": 0.7,
|
| 3176 |
+
"learning_rate": 0.0002,
|
| 3177 |
+
"loss": 0.3801,
|
| 3178 |
+
"step": 3390
|
| 3179 |
+
},
|
| 3180 |
+
{
|
| 3181 |
+
"epoch": 0.7,
|
| 3182 |
+
"learning_rate": 0.0002,
|
| 3183 |
+
"loss": 0.378,
|
| 3184 |
+
"step": 3400
|
| 3185 |
+
},
|
| 3186 |
+
{
|
| 3187 |
+
"epoch": 0.7,
|
| 3188 |
+
"eval_loss": 0.38384097814559937,
|
| 3189 |
+
"eval_runtime": 150.4207,
|
| 3190 |
+
"eval_samples_per_second": 6.648,
|
| 3191 |
+
"eval_steps_per_second": 3.324,
|
| 3192 |
+
"step": 3400
|
| 3193 |
+
},
|
| 3194 |
+
{
|
| 3195 |
+
"epoch": 0.7,
|
| 3196 |
+
"mmlu_eval_accuracy": 0.5106483129629039,
|
| 3197 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 3198 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 3199 |
+
"mmlu_eval_accuracy_astronomy": 0.625,
|
| 3200 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 3201 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 3202 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 3203 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 3204 |
+
"mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
|
| 3205 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 3206 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 3207 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 3208 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 3209 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 3210 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 3211 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 3212 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 3213 |
+
"mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
|
| 3214 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 3215 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 3216 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.5,
|
| 3217 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 3218 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 3219 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 3220 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 3221 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 3222 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.13793103448275862,
|
| 3223 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 3224 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 3225 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 3226 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 3227 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 3228 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 3229 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 3230 |
+
"mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
|
| 3231 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 3232 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 3233 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 3234 |
+
"mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
|
| 3235 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 3236 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 3237 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 3238 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 3239 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 3240 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.23,
|
| 3241 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 3242 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 3243 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 3244 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 3245 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 3246 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 3247 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 3248 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 3249 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 3250 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 3251 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 3252 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 3253 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 3254 |
+
"mmlu_loss": 1.654249626413649,
|
| 3255 |
+
"step": 3400
|
| 3256 |
+
},
|
| 3257 |
+
{
|
| 3258 |
+
"epoch": 0.71,
|
| 3259 |
+
"learning_rate": 0.0002,
|
| 3260 |
+
"loss": 0.3978,
|
| 3261 |
+
"step": 3410
|
| 3262 |
+
},
|
| 3263 |
+
{
|
| 3264 |
+
"epoch": 0.71,
|
| 3265 |
+
"learning_rate": 0.0002,
|
| 3266 |
+
"loss": 0.3872,
|
| 3267 |
+
"step": 3420
|
| 3268 |
+
},
|
| 3269 |
+
{
|
| 3270 |
+
"epoch": 0.71,
|
| 3271 |
+
"learning_rate": 0.0002,
|
| 3272 |
+
"loss": 0.3881,
|
| 3273 |
+
"step": 3430
|
| 3274 |
+
},
|
| 3275 |
+
{
|
| 3276 |
+
"epoch": 0.71,
|
| 3277 |
+
"learning_rate": 0.0002,
|
| 3278 |
+
"loss": 0.3532,
|
| 3279 |
+
"step": 3440
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 0.72,
|
| 3283 |
+
"learning_rate": 0.0002,
|
| 3284 |
+
"loss": 0.3744,
|
| 3285 |
+
"step": 3450
|
| 3286 |
+
},
|
| 3287 |
+
{
|
| 3288 |
+
"epoch": 0.72,
|
| 3289 |
+
"learning_rate": 0.0002,
|
| 3290 |
+
"loss": 0.3656,
|
| 3291 |
+
"step": 3460
|
| 3292 |
+
},
|
| 3293 |
+
{
|
| 3294 |
+
"epoch": 0.72,
|
| 3295 |
+
"learning_rate": 0.0002,
|
| 3296 |
+
"loss": 0.3281,
|
| 3297 |
+
"step": 3470
|
| 3298 |
+
},
|
| 3299 |
+
{
|
| 3300 |
+
"epoch": 0.72,
|
| 3301 |
+
"learning_rate": 0.0002,
|
| 3302 |
+
"loss": 0.3753,
|
| 3303 |
+
"step": 3480
|
| 3304 |
+
},
|
| 3305 |
+
{
|
| 3306 |
+
"epoch": 0.72,
|
| 3307 |
+
"learning_rate": 0.0002,
|
| 3308 |
+
"loss": 0.3782,
|
| 3309 |
+
"step": 3490
|
| 3310 |
+
},
|
| 3311 |
+
{
|
| 3312 |
+
"epoch": 0.73,
|
| 3313 |
+
"learning_rate": 0.0002,
|
| 3314 |
+
"loss": 0.3618,
|
| 3315 |
+
"step": 3500
|
| 3316 |
+
},
|
| 3317 |
+
{
|
| 3318 |
+
"epoch": 0.73,
|
| 3319 |
+
"learning_rate": 0.0002,
|
| 3320 |
+
"loss": 0.3473,
|
| 3321 |
+
"step": 3510
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.73,
|
| 3325 |
+
"learning_rate": 0.0002,
|
| 3326 |
+
"loss": 0.3831,
|
| 3327 |
+
"step": 3520
|
| 3328 |
+
},
|
| 3329 |
+
{
|
| 3330 |
+
"epoch": 0.73,
|
| 3331 |
+
"learning_rate": 0.0002,
|
| 3332 |
+
"loss": 0.3715,
|
| 3333 |
+
"step": 3530
|
| 3334 |
+
},
|
| 3335 |
+
{
|
| 3336 |
+
"epoch": 0.73,
|
| 3337 |
+
"learning_rate": 0.0002,
|
| 3338 |
+
"loss": 0.347,
|
| 3339 |
+
"step": 3540
|
| 3340 |
+
},
|
| 3341 |
+
{
|
| 3342 |
+
"epoch": 0.74,
|
| 3343 |
+
"learning_rate": 0.0002,
|
| 3344 |
+
"loss": 0.355,
|
| 3345 |
+
"step": 3550
|
| 3346 |
+
},
|
| 3347 |
+
{
|
| 3348 |
+
"epoch": 0.74,
|
| 3349 |
+
"learning_rate": 0.0002,
|
| 3350 |
+
"loss": 0.3469,
|
| 3351 |
+
"step": 3560
|
| 3352 |
+
},
|
| 3353 |
+
{
|
| 3354 |
+
"epoch": 0.74,
|
| 3355 |
+
"learning_rate": 0.0002,
|
| 3356 |
+
"loss": 0.3703,
|
| 3357 |
+
"step": 3570
|
| 3358 |
+
},
|
| 3359 |
+
{
|
| 3360 |
+
"epoch": 0.74,
|
| 3361 |
+
"learning_rate": 0.0002,
|
| 3362 |
+
"loss": 0.3634,
|
| 3363 |
+
"step": 3580
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 0.74,
|
| 3367 |
+
"learning_rate": 0.0002,
|
| 3368 |
+
"loss": 0.4152,
|
| 3369 |
+
"step": 3590
|
| 3370 |
+
},
|
| 3371 |
+
{
|
| 3372 |
+
"epoch": 0.75,
|
| 3373 |
+
"learning_rate": 0.0002,
|
| 3374 |
+
"loss": 0.3602,
|
| 3375 |
+
"step": 3600
|
| 3376 |
+
},
|
| 3377 |
+
{
|
| 3378 |
+
"epoch": 0.75,
|
| 3379 |
+
"eval_loss": 0.3821766972541809,
|
| 3380 |
+
"eval_runtime": 150.3692,
|
| 3381 |
+
"eval_samples_per_second": 6.65,
|
| 3382 |
+
"eval_steps_per_second": 3.325,
|
| 3383 |
+
"step": 3600
|
| 3384 |
+
},
|
| 3385 |
+
{
|
| 3386 |
+
"epoch": 0.75,
|
| 3387 |
+
"mmlu_eval_accuracy": 0.5066592017748731,
|
| 3388 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 3389 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 3390 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 3391 |
+
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
|
| 3392 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 3393 |
+
"mmlu_eval_accuracy_college_biology": 0.5,
|
| 3394 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 3395 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 3396 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 3397 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 3398 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 3399 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 3400 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 3401 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 3402 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 3403 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 3404 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 3405 |
+
"mmlu_eval_accuracy_global_facts": 0.6,
|
| 3406 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 3407 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.5,
|
| 3408 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 3409 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 3410 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8636363636363636,
|
| 3411 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 3412 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 3413 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 3414 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 3415 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 3416 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 3417 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 3418 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 3419 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7307692307692307,
|
| 3420 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 3421 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 3422 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 3423 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 3424 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 3425 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 3426 |
+
"mmlu_eval_accuracy_management": 0.7272727272727273,
|
| 3427 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 3428 |
+
"mmlu_eval_accuracy_medical_genetics": 0.8181818181818182,
|
| 3429 |
+
"mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
|
| 3430 |
+
"mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
|
| 3431 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 3432 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 3433 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 3434 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 3435 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 3436 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 3437 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 3438 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 3439 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 3440 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 3441 |
+
"mmlu_eval_accuracy_sociology": 0.7272727272727273,
|
| 3442 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 3443 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 3444 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 3445 |
+
"mmlu_loss": 1.560965163433832,
|
| 3446 |
+
"step": 3600
|
| 3447 |
+
},
|
| 3448 |
+
{
|
| 3449 |
+
"epoch": 0.75,
|
| 3450 |
+
"learning_rate": 0.0002,
|
| 3451 |
+
"loss": 0.3308,
|
| 3452 |
+
"step": 3610
|
| 3453 |
+
},
|
| 3454 |
+
{
|
| 3455 |
+
"epoch": 0.75,
|
| 3456 |
+
"learning_rate": 0.0002,
|
| 3457 |
+
"loss": 0.393,
|
| 3458 |
+
"step": 3620
|
| 3459 |
+
},
|
| 3460 |
+
{
|
| 3461 |
+
"epoch": 0.75,
|
| 3462 |
+
"learning_rate": 0.0002,
|
| 3463 |
+
"loss": 0.3341,
|
| 3464 |
+
"step": 3630
|
| 3465 |
+
},
|
| 3466 |
+
{
|
| 3467 |
+
"epoch": 0.75,
|
| 3468 |
+
"learning_rate": 0.0002,
|
| 3469 |
+
"loss": 0.3789,
|
| 3470 |
+
"step": 3640
|
| 3471 |
+
},
|
| 3472 |
+
{
|
| 3473 |
+
"epoch": 0.76,
|
| 3474 |
+
"learning_rate": 0.0002,
|
| 3475 |
+
"loss": 0.3394,
|
| 3476 |
+
"step": 3650
|
| 3477 |
+
},
|
| 3478 |
+
{
|
| 3479 |
+
"epoch": 0.76,
|
| 3480 |
+
"learning_rate": 0.0002,
|
| 3481 |
+
"loss": 0.3975,
|
| 3482 |
+
"step": 3660
|
| 3483 |
+
},
|
| 3484 |
+
{
|
| 3485 |
+
"epoch": 0.76,
|
| 3486 |
+
"learning_rate": 0.0002,
|
| 3487 |
+
"loss": 0.3834,
|
| 3488 |
+
"step": 3670
|
| 3489 |
+
},
|
| 3490 |
+
{
|
| 3491 |
+
"epoch": 0.76,
|
| 3492 |
+
"learning_rate": 0.0002,
|
| 3493 |
+
"loss": 0.3597,
|
| 3494 |
+
"step": 3680
|
| 3495 |
+
},
|
| 3496 |
+
{
|
| 3497 |
+
"epoch": 0.76,
|
| 3498 |
+
"learning_rate": 0.0002,
|
| 3499 |
+
"loss": 0.3377,
|
| 3500 |
+
"step": 3690
|
| 3501 |
+
},
|
| 3502 |
+
{
|
| 3503 |
+
"epoch": 0.77,
|
| 3504 |
+
"learning_rate": 0.0002,
|
| 3505 |
+
"loss": 0.3229,
|
| 3506 |
+
"step": 3700
|
| 3507 |
+
},
|
| 3508 |
+
{
|
| 3509 |
+
"epoch": 0.77,
|
| 3510 |
+
"learning_rate": 0.0002,
|
| 3511 |
+
"loss": 0.3847,
|
| 3512 |
+
"step": 3710
|
| 3513 |
+
},
|
| 3514 |
+
{
|
| 3515 |
+
"epoch": 0.77,
|
| 3516 |
+
"learning_rate": 0.0002,
|
| 3517 |
+
"loss": 0.3477,
|
| 3518 |
+
"step": 3720
|
| 3519 |
+
},
|
| 3520 |
+
{
|
| 3521 |
+
"epoch": 0.77,
|
| 3522 |
+
"learning_rate": 0.0002,
|
| 3523 |
+
"loss": 0.4176,
|
| 3524 |
+
"step": 3730
|
| 3525 |
+
},
|
| 3526 |
+
{
|
| 3527 |
+
"epoch": 0.78,
|
| 3528 |
+
"learning_rate": 0.0002,
|
| 3529 |
+
"loss": 0.3849,
|
| 3530 |
+
"step": 3740
|
| 3531 |
+
},
|
| 3532 |
+
{
|
| 3533 |
+
"epoch": 0.78,
|
| 3534 |
+
"learning_rate": 0.0002,
|
| 3535 |
+
"loss": 0.4104,
|
| 3536 |
+
"step": 3750
|
| 3537 |
+
},
|
| 3538 |
+
{
|
| 3539 |
+
"epoch": 0.78,
|
| 3540 |
+
"learning_rate": 0.0002,
|
| 3541 |
+
"loss": 0.3364,
|
| 3542 |
+
"step": 3760
|
| 3543 |
+
},
|
| 3544 |
+
{
|
| 3545 |
+
"epoch": 0.78,
|
| 3546 |
+
"learning_rate": 0.0002,
|
| 3547 |
+
"loss": 0.3713,
|
| 3548 |
+
"step": 3770
|
| 3549 |
+
},
|
| 3550 |
+
{
|
| 3551 |
+
"epoch": 0.78,
|
| 3552 |
+
"learning_rate": 0.0002,
|
| 3553 |
+
"loss": 0.4152,
|
| 3554 |
+
"step": 3780
|
| 3555 |
+
},
|
| 3556 |
+
{
|
| 3557 |
+
"epoch": 0.79,
|
| 3558 |
+
"learning_rate": 0.0002,
|
| 3559 |
+
"loss": 0.3823,
|
| 3560 |
+
"step": 3790
|
| 3561 |
+
},
|
| 3562 |
+
{
|
| 3563 |
+
"epoch": 0.79,
|
| 3564 |
+
"learning_rate": 0.0002,
|
| 3565 |
+
"loss": 0.4243,
|
| 3566 |
+
"step": 3800
|
| 3567 |
+
},
|
| 3568 |
+
{
|
| 3569 |
+
"epoch": 0.79,
|
| 3570 |
+
"eval_loss": 0.38058969378471375,
|
| 3571 |
+
"eval_runtime": 150.3623,
|
| 3572 |
+
"eval_samples_per_second": 6.651,
|
| 3573 |
+
"eval_steps_per_second": 3.325,
|
| 3574 |
+
"step": 3800
|
| 3575 |
+
},
|
| 3576 |
+
{
|
| 3577 |
+
"epoch": 0.79,
|
| 3578 |
+
"mmlu_eval_accuracy": 0.49026848136160434,
|
| 3579 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
|
| 3580 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 3581 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 3582 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 3583 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 3584 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 3585 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 3586 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 3587 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 3588 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 3589 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 3590 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 3591 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 3592 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 3593 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 3594 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 3595 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 3596 |
+
"mmlu_eval_accuracy_global_facts": 0.3,
|
| 3597 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 3598 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
|
| 3599 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 3600 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 3601 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 3602 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 3603 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
|
| 3604 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.10344827586206896,
|
| 3605 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 3606 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 3607 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8833333333333333,
|
| 3608 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 3609 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 3610 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.6923076923076923,
|
| 3611 |
+
"mmlu_eval_accuracy_human_aging": 0.6956521739130435,
|
| 3612 |
+
"mmlu_eval_accuracy_human_sexuality": 0.5,
|
| 3613 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 3614 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 3615 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 3616 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 3617 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 3618 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 3619 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 3620 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 3621 |
+
"mmlu_eval_accuracy_moral_disputes": 0.39473684210526316,
|
| 3622 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 3623 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 3624 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 3625 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 3626 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3870967741935484,
|
| 3627 |
+
"mmlu_eval_accuracy_professional_law": 0.3352941176470588,
|
| 3628 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 3629 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4927536231884058,
|
| 3630 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 3631 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 3632 |
+
"mmlu_eval_accuracy_sociology": 0.7727272727272727,
|
| 3633 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 3634 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 3635 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 3636 |
+
"mmlu_loss": 1.425224914174167,
|
| 3637 |
+
"step": 3800
|
| 3638 |
}
|
| 3639 |
],
|
| 3640 |
"max_steps": 10000,
|
| 3641 |
"num_train_epochs": 3,
|
| 3642 |
+
"total_flos": 4.575699348904673e+17,
|
| 3643 |
"trial_name": null,
|
| 3644 |
"trial_params": null
|
| 3645 |
}
|
{checkpoint-1600 → checkpoint-3800}/training_args.bin
RENAMED
|
File without changes
|