Farouk commited on
Commit ·
c1908a9
1
Parent(s): 16a482d
Training in progress, step 3200
Browse files- adapter_model.bin +1 -1
- {checkpoint-1000 → checkpoint-3000/adapter_model/adapter_model}/README.md +0 -0
- {checkpoint-1000 → checkpoint-3000/adapter_model/adapter_model}/adapter_config.json +0 -0
- {checkpoint-1000 → checkpoint-3000/adapter_model/adapter_model}/adapter_model.bin +1 -1
- {checkpoint-1000/adapter_model/adapter_model → checkpoint-3200}/README.md +0 -0
- {checkpoint-1000/adapter_model/adapter_model → checkpoint-3200}/adapter_config.json +0 -0
- {checkpoint-1000/adapter_model/adapter_model → checkpoint-3200}/adapter_model.bin +1 -1
- {checkpoint-1000 → checkpoint-3200}/added_tokens.json +0 -0
- {checkpoint-1000 → checkpoint-3200}/optimizer.pt +1 -1
- {checkpoint-1000 → checkpoint-3200}/rng_state.pth +1 -1
- {checkpoint-1000 → checkpoint-3200}/scheduler.pt +1 -1
- {checkpoint-1000 → checkpoint-3200}/special_tokens_map.json +0 -0
- {checkpoint-1000 → checkpoint-3200}/tokenizer.model +0 -0
- {checkpoint-1000 → checkpoint-3200}/tokenizer_config.json +0 -0
- {checkpoint-1000 → checkpoint-3200}/trainer_state.json +2106 -5
- {checkpoint-1000 → checkpoint-3200}/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:f741d1f59b3587625e5a113d94b16eee6f694bf6f8d28858703c327bb2c9b901
|
| 3 |
size 319977229
|
{checkpoint-1000 → checkpoint-3000/adapter_model/adapter_model}/README.md
RENAMED
|
File without changes
|
{checkpoint-1000 → checkpoint-3000/adapter_model/adapter_model}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-1000 → checkpoint-3000/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 sha256:549cc04dfe3704d3b602edbc710988eba51d7047b8371004ad5d5d4e1cc79b7c
|
| 3 |
size 319977229
|
{checkpoint-1000/adapter_model/adapter_model → checkpoint-3200}/README.md
RENAMED
|
File without changes
|
{checkpoint-1000/adapter_model/adapter_model → checkpoint-3200}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-1000/adapter_model/adapter_model → checkpoint-3200}/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:f741d1f59b3587625e5a113d94b16eee6f694bf6f8d28858703c327bb2c9b901
|
| 3 |
size 319977229
|
{checkpoint-1000 → checkpoint-3200}/added_tokens.json
RENAMED
|
File without changes
|
{checkpoint-1000 → checkpoint-3200}/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:e1a8e2087a43035cecae8424d572f1e54f7a9d9499d139bc2e8b192168e39ed9
|
| 3 |
size 1279539973
|
{checkpoint-1000 → checkpoint-3200}/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:a5463493212120226f2c8927321113cc3034c18a25590bd7c26adc7f1e518b6b
|
| 3 |
size 14511
|
{checkpoint-1000 → checkpoint-3200}/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 sha256:50b57a34df83b700e2c13775ff734b4569b74ce7e20da3479db76577bb4e906e
|
| 3 |
size 627
|
{checkpoint-1000 → checkpoint-3200}/special_tokens_map.json
RENAMED
|
File without changes
|
{checkpoint-1000 → checkpoint-3200}/tokenizer.model
RENAMED
|
File without changes
|
{checkpoint-1000 → checkpoint-3200}/tokenizer_config.json
RENAMED
|
File without changes
|
{checkpoint-1000 → checkpoint-3200}/trainer_state.json
RENAMED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "experts/expert-6/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -961,11 +961,2112 @@
|
|
| 961 |
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 962 |
"mmlu_loss": 1.33750198186222,
|
| 963 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 964 |
}
|
| 965 |
],
|
| 966 |
"max_steps": 10000,
|
| 967 |
"num_train_epochs": 6,
|
| 968 |
-
"total_flos":
|
| 969 |
"trial_name": null,
|
| 970 |
"trial_params": null
|
| 971 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.5174126029014587,
|
| 3 |
+
"best_model_checkpoint": "experts/expert-6/checkpoint-3200",
|
| 4 |
+
"epoch": 1.754626456477039,
|
| 5 |
+
"global_step": 3200,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 961 |
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 962 |
"mmlu_loss": 1.33750198186222,
|
| 963 |
"step": 1000
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.55,
|
| 967 |
+
"learning_rate": 0.0002,
|
| 968 |
+
"loss": 0.6046,
|
| 969 |
+
"step": 1010
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.56,
|
| 973 |
+
"learning_rate": 0.0002,
|
| 974 |
+
"loss": 0.5002,
|
| 975 |
+
"step": 1020
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.56,
|
| 979 |
+
"learning_rate": 0.0002,
|
| 980 |
+
"loss": 0.5698,
|
| 981 |
+
"step": 1030
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.57,
|
| 985 |
+
"learning_rate": 0.0002,
|
| 986 |
+
"loss": 0.5725,
|
| 987 |
+
"step": 1040
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.58,
|
| 991 |
+
"learning_rate": 0.0002,
|
| 992 |
+
"loss": 0.4983,
|
| 993 |
+
"step": 1050
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.58,
|
| 997 |
+
"learning_rate": 0.0002,
|
| 998 |
+
"loss": 0.5334,
|
| 999 |
+
"step": 1060
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 0.59,
|
| 1003 |
+
"learning_rate": 0.0002,
|
| 1004 |
+
"loss": 0.56,
|
| 1005 |
+
"step": 1070
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.59,
|
| 1009 |
+
"learning_rate": 0.0002,
|
| 1010 |
+
"loss": 0.5172,
|
| 1011 |
+
"step": 1080
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 0.6,
|
| 1015 |
+
"learning_rate": 0.0002,
|
| 1016 |
+
"loss": 0.5881,
|
| 1017 |
+
"step": 1090
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.6,
|
| 1021 |
+
"learning_rate": 0.0002,
|
| 1022 |
+
"loss": 0.5451,
|
| 1023 |
+
"step": 1100
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.61,
|
| 1027 |
+
"learning_rate": 0.0002,
|
| 1028 |
+
"loss": 0.5509,
|
| 1029 |
+
"step": 1110
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.61,
|
| 1033 |
+
"learning_rate": 0.0002,
|
| 1034 |
+
"loss": 0.5885,
|
| 1035 |
+
"step": 1120
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.62,
|
| 1039 |
+
"learning_rate": 0.0002,
|
| 1040 |
+
"loss": 0.5413,
|
| 1041 |
+
"step": 1130
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.63,
|
| 1045 |
+
"learning_rate": 0.0002,
|
| 1046 |
+
"loss": 0.5581,
|
| 1047 |
+
"step": 1140
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.63,
|
| 1051 |
+
"learning_rate": 0.0002,
|
| 1052 |
+
"loss": 0.5196,
|
| 1053 |
+
"step": 1150
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 0.64,
|
| 1057 |
+
"learning_rate": 0.0002,
|
| 1058 |
+
"loss": 0.5483,
|
| 1059 |
+
"step": 1160
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.64,
|
| 1063 |
+
"learning_rate": 0.0002,
|
| 1064 |
+
"loss": 0.5608,
|
| 1065 |
+
"step": 1170
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.65,
|
| 1069 |
+
"learning_rate": 0.0002,
|
| 1070 |
+
"loss": 0.5433,
|
| 1071 |
+
"step": 1180
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.65,
|
| 1075 |
+
"learning_rate": 0.0002,
|
| 1076 |
+
"loss": 0.5425,
|
| 1077 |
+
"step": 1190
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 0.66,
|
| 1081 |
+
"learning_rate": 0.0002,
|
| 1082 |
+
"loss": 0.5552,
|
| 1083 |
+
"step": 1200
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.66,
|
| 1087 |
+
"eval_loss": 0.5514205098152161,
|
| 1088 |
+
"eval_runtime": 109.0129,
|
| 1089 |
+
"eval_samples_per_second": 9.173,
|
| 1090 |
+
"eval_steps_per_second": 4.587,
|
| 1091 |
+
"step": 1200
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 0.66,
|
| 1095 |
+
"mmlu_eval_accuracy": 0.49572621607362827,
|
| 1096 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 1097 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 1098 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 1099 |
+
"mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
|
| 1100 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 1101 |
+
"mmlu_eval_accuracy_college_biology": 0.3125,
|
| 1102 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1103 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 1104 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 1105 |
+
"mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
|
| 1106 |
+
"mmlu_eval_accuracy_college_physics": 0.5454545454545454,
|
| 1107 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 1108 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 1109 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 1110 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 1111 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
|
| 1112 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 1113 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1114 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 1115 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 1116 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1117 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 1118 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 1119 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 1120 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 1121 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
|
| 1122 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 1123 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 1124 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 1125 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 1126 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 1127 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 1128 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 1129 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 1130 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 1131 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 1132 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 1133 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 1134 |
+
"mmlu_eval_accuracy_management": 0.7272727272727273,
|
| 1135 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 1136 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 1137 |
+
"mmlu_eval_accuracy_miscellaneous": 0.627906976744186,
|
| 1138 |
+
"mmlu_eval_accuracy_moral_disputes": 0.39473684210526316,
|
| 1139 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 1140 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 1141 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 1142 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 1143 |
+
"mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
|
| 1144 |
+
"mmlu_eval_accuracy_professional_law": 0.3235294117647059,
|
| 1145 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
|
| 1146 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 1147 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 1148 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1149 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 1150 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 1151 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 1152 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 1153 |
+
"mmlu_loss": 1.246683247367956,
|
| 1154 |
+
"step": 1200
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.66,
|
| 1158 |
+
"learning_rate": 0.0002,
|
| 1159 |
+
"loss": 0.5701,
|
| 1160 |
+
"step": 1210
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.67,
|
| 1164 |
+
"learning_rate": 0.0002,
|
| 1165 |
+
"loss": 0.5689,
|
| 1166 |
+
"step": 1220
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.67,
|
| 1170 |
+
"learning_rate": 0.0002,
|
| 1171 |
+
"loss": 0.5589,
|
| 1172 |
+
"step": 1230
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.68,
|
| 1176 |
+
"learning_rate": 0.0002,
|
| 1177 |
+
"loss": 0.5228,
|
| 1178 |
+
"step": 1240
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.69,
|
| 1182 |
+
"learning_rate": 0.0002,
|
| 1183 |
+
"loss": 0.5,
|
| 1184 |
+
"step": 1250
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.69,
|
| 1188 |
+
"learning_rate": 0.0002,
|
| 1189 |
+
"loss": 0.5536,
|
| 1190 |
+
"step": 1260
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.7,
|
| 1194 |
+
"learning_rate": 0.0002,
|
| 1195 |
+
"loss": 0.5597,
|
| 1196 |
+
"step": 1270
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.7,
|
| 1200 |
+
"learning_rate": 0.0002,
|
| 1201 |
+
"loss": 0.541,
|
| 1202 |
+
"step": 1280
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.71,
|
| 1206 |
+
"learning_rate": 0.0002,
|
| 1207 |
+
"loss": 0.5468,
|
| 1208 |
+
"step": 1290
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.71,
|
| 1212 |
+
"learning_rate": 0.0002,
|
| 1213 |
+
"loss": 0.5692,
|
| 1214 |
+
"step": 1300
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.72,
|
| 1218 |
+
"learning_rate": 0.0002,
|
| 1219 |
+
"loss": 0.5645,
|
| 1220 |
+
"step": 1310
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.72,
|
| 1224 |
+
"learning_rate": 0.0002,
|
| 1225 |
+
"loss": 0.5462,
|
| 1226 |
+
"step": 1320
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.73,
|
| 1230 |
+
"learning_rate": 0.0002,
|
| 1231 |
+
"loss": 0.5102,
|
| 1232 |
+
"step": 1330
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.73,
|
| 1236 |
+
"learning_rate": 0.0002,
|
| 1237 |
+
"loss": 0.5537,
|
| 1238 |
+
"step": 1340
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.74,
|
| 1242 |
+
"learning_rate": 0.0002,
|
| 1243 |
+
"loss": 0.5233,
|
| 1244 |
+
"step": 1350
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.75,
|
| 1248 |
+
"learning_rate": 0.0002,
|
| 1249 |
+
"loss": 0.5764,
|
| 1250 |
+
"step": 1360
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.75,
|
| 1254 |
+
"learning_rate": 0.0002,
|
| 1255 |
+
"loss": 0.5333,
|
| 1256 |
+
"step": 1370
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.76,
|
| 1260 |
+
"learning_rate": 0.0002,
|
| 1261 |
+
"loss": 0.5871,
|
| 1262 |
+
"step": 1380
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.76,
|
| 1266 |
+
"learning_rate": 0.0002,
|
| 1267 |
+
"loss": 0.5579,
|
| 1268 |
+
"step": 1390
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 0.77,
|
| 1272 |
+
"learning_rate": 0.0002,
|
| 1273 |
+
"loss": 0.5446,
|
| 1274 |
+
"step": 1400
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.77,
|
| 1278 |
+
"eval_loss": 0.5456764101982117,
|
| 1279 |
+
"eval_runtime": 103.6888,
|
| 1280 |
+
"eval_samples_per_second": 9.644,
|
| 1281 |
+
"eval_steps_per_second": 4.822,
|
| 1282 |
+
"step": 1400
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 0.77,
|
| 1286 |
+
"mmlu_eval_accuracy": 0.49496437436898577,
|
| 1287 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 1288 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 1289 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 1290 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 1291 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
|
| 1292 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 1293 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1294 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 1295 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 1296 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 1297 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 1298 |
+
"mmlu_eval_accuracy_computer_security": 0.36363636363636365,
|
| 1299 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 1300 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 1301 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 1302 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 1303 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 1304 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1305 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 1306 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 1307 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1308 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 1309 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 1310 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 1311 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
|
| 1312 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
|
| 1313 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 1314 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 1315 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 1316 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 1317 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 1318 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 1319 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 1320 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 1321 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 1322 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 1323 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 1324 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 1325 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 1326 |
+
"mmlu_eval_accuracy_marketing": 0.88,
|
| 1327 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 1328 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
|
| 1329 |
+
"mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
|
| 1330 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 1331 |
+
"mmlu_eval_accuracy_nutrition": 0.7272727272727273,
|
| 1332 |
+
"mmlu_eval_accuracy_philosophy": 0.5588235294117647,
|
| 1333 |
+
"mmlu_eval_accuracy_prehistory": 0.42857142857142855,
|
| 1334 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 1335 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 1336 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 1337 |
+
"mmlu_eval_accuracy_professional_psychology": 0.5072463768115942,
|
| 1338 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 1339 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1340 |
+
"mmlu_eval_accuracy_sociology": 0.5909090909090909,
|
| 1341 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 1342 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 1343 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 1344 |
+
"mmlu_loss": 1.308620565242618,
|
| 1345 |
+
"step": 1400
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 0.77,
|
| 1349 |
+
"learning_rate": 0.0002,
|
| 1350 |
+
"loss": 0.5309,
|
| 1351 |
+
"step": 1410
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 0.78,
|
| 1355 |
+
"learning_rate": 0.0002,
|
| 1356 |
+
"loss": 0.5283,
|
| 1357 |
+
"step": 1420
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 0.78,
|
| 1361 |
+
"learning_rate": 0.0002,
|
| 1362 |
+
"loss": 0.5341,
|
| 1363 |
+
"step": 1430
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.79,
|
| 1367 |
+
"learning_rate": 0.0002,
|
| 1368 |
+
"loss": 0.5299,
|
| 1369 |
+
"step": 1440
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 0.8,
|
| 1373 |
+
"learning_rate": 0.0002,
|
| 1374 |
+
"loss": 0.5495,
|
| 1375 |
+
"step": 1450
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 0.8,
|
| 1379 |
+
"learning_rate": 0.0002,
|
| 1380 |
+
"loss": 0.5273,
|
| 1381 |
+
"step": 1460
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.81,
|
| 1385 |
+
"learning_rate": 0.0002,
|
| 1386 |
+
"loss": 0.5587,
|
| 1387 |
+
"step": 1470
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 0.81,
|
| 1391 |
+
"learning_rate": 0.0002,
|
| 1392 |
+
"loss": 0.5268,
|
| 1393 |
+
"step": 1480
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 0.82,
|
| 1397 |
+
"learning_rate": 0.0002,
|
| 1398 |
+
"loss": 0.5117,
|
| 1399 |
+
"step": 1490
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 0.82,
|
| 1403 |
+
"learning_rate": 0.0002,
|
| 1404 |
+
"loss": 0.5744,
|
| 1405 |
+
"step": 1500
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.83,
|
| 1409 |
+
"learning_rate": 0.0002,
|
| 1410 |
+
"loss": 0.5194,
|
| 1411 |
+
"step": 1510
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 0.83,
|
| 1415 |
+
"learning_rate": 0.0002,
|
| 1416 |
+
"loss": 0.5755,
|
| 1417 |
+
"step": 1520
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.84,
|
| 1421 |
+
"learning_rate": 0.0002,
|
| 1422 |
+
"loss": 0.5433,
|
| 1423 |
+
"step": 1530
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.84,
|
| 1427 |
+
"learning_rate": 0.0002,
|
| 1428 |
+
"loss": 0.5632,
|
| 1429 |
+
"step": 1540
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 0.85,
|
| 1433 |
+
"learning_rate": 0.0002,
|
| 1434 |
+
"loss": 0.5372,
|
| 1435 |
+
"step": 1550
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 0.86,
|
| 1439 |
+
"learning_rate": 0.0002,
|
| 1440 |
+
"loss": 0.5226,
|
| 1441 |
+
"step": 1560
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 0.86,
|
| 1445 |
+
"learning_rate": 0.0002,
|
| 1446 |
+
"loss": 0.5105,
|
| 1447 |
+
"step": 1570
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 0.87,
|
| 1451 |
+
"learning_rate": 0.0002,
|
| 1452 |
+
"loss": 0.5838,
|
| 1453 |
+
"step": 1580
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 0.87,
|
| 1457 |
+
"learning_rate": 0.0002,
|
| 1458 |
+
"loss": 0.5393,
|
| 1459 |
+
"step": 1590
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 0.88,
|
| 1463 |
+
"learning_rate": 0.0002,
|
| 1464 |
+
"loss": 0.5643,
|
| 1465 |
+
"step": 1600
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.88,
|
| 1469 |
+
"eval_loss": 0.5385324954986572,
|
| 1470 |
+
"eval_runtime": 104.2039,
|
| 1471 |
+
"eval_samples_per_second": 9.597,
|
| 1472 |
+
"eval_steps_per_second": 4.798,
|
| 1473 |
+
"step": 1600
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 0.88,
|
| 1477 |
+
"mmlu_eval_accuracy": 0.4970869163651323,
|
| 1478 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 1479 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 1480 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 1481 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 1482 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 1483 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 1484 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1485 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 1486 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 1487 |
+
"mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
|
| 1488 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 1489 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 1490 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 1491 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 1492 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 1493 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
|
| 1494 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 1495 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 1496 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 1497 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
|
| 1498 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1499 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 1500 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 1501 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 1502 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
|
| 1503 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 1504 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 1505 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 1506 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 1507 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 1508 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 1509 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 1510 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 1511 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 1512 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 1513 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 1514 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 1515 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 1516 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 1517 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 1518 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 1519 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
|
| 1520 |
+
"mmlu_eval_accuracy_moral_disputes": 0.3684210526315789,
|
| 1521 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.24,
|
| 1522 |
+
"mmlu_eval_accuracy_nutrition": 0.7575757575757576,
|
| 1523 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 1524 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 1525 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
|
| 1526 |
+
"mmlu_eval_accuracy_professional_law": 0.32941176470588235,
|
| 1527 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 1528 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 1529 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 1530 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1531 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 1532 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 1533 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 1534 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 1535 |
+
"mmlu_loss": 1.238472634328252,
|
| 1536 |
+
"step": 1600
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.88,
|
| 1540 |
+
"learning_rate": 0.0002,
|
| 1541 |
+
"loss": 0.5145,
|
| 1542 |
+
"step": 1610
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.89,
|
| 1546 |
+
"learning_rate": 0.0002,
|
| 1547 |
+
"loss": 0.5653,
|
| 1548 |
+
"step": 1620
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 0.89,
|
| 1552 |
+
"learning_rate": 0.0002,
|
| 1553 |
+
"loss": 0.5438,
|
| 1554 |
+
"step": 1630
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 0.9,
|
| 1558 |
+
"learning_rate": 0.0002,
|
| 1559 |
+
"loss": 0.5182,
|
| 1560 |
+
"step": 1640
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 0.9,
|
| 1564 |
+
"learning_rate": 0.0002,
|
| 1565 |
+
"loss": 0.5302,
|
| 1566 |
+
"step": 1650
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.91,
|
| 1570 |
+
"learning_rate": 0.0002,
|
| 1571 |
+
"loss": 0.5761,
|
| 1572 |
+
"step": 1660
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.92,
|
| 1576 |
+
"learning_rate": 0.0002,
|
| 1577 |
+
"loss": 0.5034,
|
| 1578 |
+
"step": 1670
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 0.92,
|
| 1582 |
+
"learning_rate": 0.0002,
|
| 1583 |
+
"loss": 0.549,
|
| 1584 |
+
"step": 1680
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.93,
|
| 1588 |
+
"learning_rate": 0.0002,
|
| 1589 |
+
"loss": 0.5062,
|
| 1590 |
+
"step": 1690
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 0.93,
|
| 1594 |
+
"learning_rate": 0.0002,
|
| 1595 |
+
"loss": 0.5182,
|
| 1596 |
+
"step": 1700
|
| 1597 |
+
},
|
| 1598 |
+
{
|
| 1599 |
+
"epoch": 0.94,
|
| 1600 |
+
"learning_rate": 0.0002,
|
| 1601 |
+
"loss": 0.5078,
|
| 1602 |
+
"step": 1710
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 0.94,
|
| 1606 |
+
"learning_rate": 0.0002,
|
| 1607 |
+
"loss": 0.5673,
|
| 1608 |
+
"step": 1720
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.95,
|
| 1612 |
+
"learning_rate": 0.0002,
|
| 1613 |
+
"loss": 0.5387,
|
| 1614 |
+
"step": 1730
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.95,
|
| 1618 |
+
"learning_rate": 0.0002,
|
| 1619 |
+
"loss": 0.5543,
|
| 1620 |
+
"step": 1740
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 0.96,
|
| 1624 |
+
"learning_rate": 0.0002,
|
| 1625 |
+
"loss": 0.5063,
|
| 1626 |
+
"step": 1750
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.97,
|
| 1630 |
+
"learning_rate": 0.0002,
|
| 1631 |
+
"loss": 0.556,
|
| 1632 |
+
"step": 1760
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 0.97,
|
| 1636 |
+
"learning_rate": 0.0002,
|
| 1637 |
+
"loss": 0.5683,
|
| 1638 |
+
"step": 1770
|
| 1639 |
+
},
|
| 1640 |
+
{
|
| 1641 |
+
"epoch": 0.98,
|
| 1642 |
+
"learning_rate": 0.0002,
|
| 1643 |
+
"loss": 0.5028,
|
| 1644 |
+
"step": 1780
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 0.98,
|
| 1648 |
+
"learning_rate": 0.0002,
|
| 1649 |
+
"loss": 0.5215,
|
| 1650 |
+
"step": 1790
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 0.99,
|
| 1654 |
+
"learning_rate": 0.0002,
|
| 1655 |
+
"loss": 0.6093,
|
| 1656 |
+
"step": 1800
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.99,
|
| 1660 |
+
"eval_loss": 0.533049464225769,
|
| 1661 |
+
"eval_runtime": 103.8125,
|
| 1662 |
+
"eval_samples_per_second": 9.633,
|
| 1663 |
+
"eval_steps_per_second": 4.816,
|
| 1664 |
+
"step": 1800
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.99,
|
| 1668 |
+
"mmlu_eval_accuracy": 0.48692903861568687,
|
| 1669 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 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": 0.5517241379310345,
|
| 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": 0.16666666666666666,
|
| 1683 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 1684 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
|
| 1685 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 1686 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1687 |
+
"mmlu_eval_accuracy_high_school_biology": 0.5,
|
| 1688 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
|
| 1689 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1690 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 1691 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 1692 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 1693 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 1694 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 1695 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
|
| 1696 |
+
"mmlu_eval_accuracy_high_school_physics": 0.17647058823529413,
|
| 1697 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 1698 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
|
| 1699 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 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": 0.8461538461538461,
|
| 1704 |
+
"mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
|
| 1705 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6666666666666666,
|
| 1706 |
+
"mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
|
| 1707 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 1708 |
+
"mmlu_eval_accuracy_marketing": 0.76,
|
| 1709 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 1710 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
|
| 1711 |
+
"mmlu_eval_accuracy_moral_disputes": 0.34210526315789475,
|
| 1712 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.26,
|
| 1713 |
+
"mmlu_eval_accuracy_nutrition": 0.7575757575757576,
|
| 1714 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 1715 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 1716 |
+
"mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
|
| 1717 |
+
"mmlu_eval_accuracy_professional_law": 0.3411764705882353,
|
| 1718 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
|
| 1719 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4927536231884058,
|
| 1720 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 1721 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 1722 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 1723 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 1724 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 1725 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 1726 |
+
"mmlu_loss": 1.2787463259603584,
|
| 1727 |
+
"step": 1800
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 0.99,
|
| 1731 |
+
"learning_rate": 0.0002,
|
| 1732 |
+
"loss": 0.5148,
|
| 1733 |
+
"step": 1810
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 1.0,
|
| 1737 |
+
"learning_rate": 0.0002,
|
| 1738 |
+
"loss": 0.5756,
|
| 1739 |
+
"step": 1820
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 1.0,
|
| 1743 |
+
"learning_rate": 0.0002,
|
| 1744 |
+
"loss": 0.4839,
|
| 1745 |
+
"step": 1830
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 1.01,
|
| 1749 |
+
"learning_rate": 0.0002,
|
| 1750 |
+
"loss": 0.5187,
|
| 1751 |
+
"step": 1840
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 1.01,
|
| 1755 |
+
"learning_rate": 0.0002,
|
| 1756 |
+
"loss": 0.4551,
|
| 1757 |
+
"step": 1850
|
| 1758 |
+
},
|
| 1759 |
+
{
|
| 1760 |
+
"epoch": 1.02,
|
| 1761 |
+
"learning_rate": 0.0002,
|
| 1762 |
+
"loss": 0.4728,
|
| 1763 |
+
"step": 1860
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.03,
|
| 1767 |
+
"learning_rate": 0.0002,
|
| 1768 |
+
"loss": 0.4363,
|
| 1769 |
+
"step": 1870
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 1.03,
|
| 1773 |
+
"learning_rate": 0.0002,
|
| 1774 |
+
"loss": 0.505,
|
| 1775 |
+
"step": 1880
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.04,
|
| 1779 |
+
"learning_rate": 0.0002,
|
| 1780 |
+
"loss": 0.432,
|
| 1781 |
+
"step": 1890
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.04,
|
| 1785 |
+
"learning_rate": 0.0002,
|
| 1786 |
+
"loss": 0.5109,
|
| 1787 |
+
"step": 1900
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 1.05,
|
| 1791 |
+
"learning_rate": 0.0002,
|
| 1792 |
+
"loss": 0.4998,
|
| 1793 |
+
"step": 1910
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 1.05,
|
| 1797 |
+
"learning_rate": 0.0002,
|
| 1798 |
+
"loss": 0.4783,
|
| 1799 |
+
"step": 1920
|
| 1800 |
+
},
|
| 1801 |
+
{
|
| 1802 |
+
"epoch": 1.06,
|
| 1803 |
+
"learning_rate": 0.0002,
|
| 1804 |
+
"loss": 0.4621,
|
| 1805 |
+
"step": 1930
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 1.06,
|
| 1809 |
+
"learning_rate": 0.0002,
|
| 1810 |
+
"loss": 0.4886,
|
| 1811 |
+
"step": 1940
|
| 1812 |
+
},
|
| 1813 |
+
{
|
| 1814 |
+
"epoch": 1.07,
|
| 1815 |
+
"learning_rate": 0.0002,
|
| 1816 |
+
"loss": 0.4442,
|
| 1817 |
+
"step": 1950
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 1.07,
|
| 1821 |
+
"learning_rate": 0.0002,
|
| 1822 |
+
"loss": 0.4949,
|
| 1823 |
+
"step": 1960
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 1.08,
|
| 1827 |
+
"learning_rate": 0.0002,
|
| 1828 |
+
"loss": 0.4921,
|
| 1829 |
+
"step": 1970
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 1.09,
|
| 1833 |
+
"learning_rate": 0.0002,
|
| 1834 |
+
"loss": 0.4902,
|
| 1835 |
+
"step": 1980
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 1.09,
|
| 1839 |
+
"learning_rate": 0.0002,
|
| 1840 |
+
"loss": 0.4557,
|
| 1841 |
+
"step": 1990
|
| 1842 |
+
},
|
| 1843 |
+
{
|
| 1844 |
+
"epoch": 1.1,
|
| 1845 |
+
"learning_rate": 0.0002,
|
| 1846 |
+
"loss": 0.4776,
|
| 1847 |
+
"step": 2000
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 1.1,
|
| 1851 |
+
"eval_loss": 0.5350334644317627,
|
| 1852 |
+
"eval_runtime": 103.7048,
|
| 1853 |
+
"eval_samples_per_second": 9.643,
|
| 1854 |
+
"eval_steps_per_second": 4.821,
|
| 1855 |
+
"step": 2000
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 1.1,
|
| 1859 |
+
"mmlu_eval_accuracy": 0.4966352630822314,
|
| 1860 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
|
| 1861 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 1862 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 1863 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 1864 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 1865 |
+
"mmlu_eval_accuracy_college_biology": 0.3125,
|
| 1866 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 1867 |
+
"mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
|
| 1868 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 1869 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 1870 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 1871 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 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": 0.07142857142857142,
|
| 1877 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 1878 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 1879 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.5,
|
| 1880 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 1881 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
|
| 1882 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 1883 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 1884 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 1885 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
|
| 1886 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
|
| 1887 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 1888 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.85,
|
| 1889 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
|
| 1890 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 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": 0.36363636363636365,
|
| 1898 |
+
"mmlu_eval_accuracy_management": 0.7272727272727273,
|
| 1899 |
+
"mmlu_eval_accuracy_marketing": 0.8,
|
| 1900 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 1901 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 1902 |
+
"mmlu_eval_accuracy_moral_disputes": 0.3684210526315789,
|
| 1903 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 1904 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 1905 |
+
"mmlu_eval_accuracy_philosophy": 0.47058823529411764,
|
| 1906 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 1907 |
+
"mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
|
| 1908 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 1909 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 1910 |
+
"mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
|
| 1911 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 1912 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 1913 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 1914 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 1915 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 1916 |
+
"mmlu_eval_accuracy_world_religions": 0.7368421052631579,
|
| 1917 |
+
"mmlu_loss": 1.1903044548342807,
|
| 1918 |
+
"step": 2000
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 1.1,
|
| 1922 |
+
"learning_rate": 0.0002,
|
| 1923 |
+
"loss": 0.4911,
|
| 1924 |
+
"step": 2010
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 1.11,
|
| 1928 |
+
"learning_rate": 0.0002,
|
| 1929 |
+
"loss": 0.4564,
|
| 1930 |
+
"step": 2020
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 1.11,
|
| 1934 |
+
"learning_rate": 0.0002,
|
| 1935 |
+
"loss": 0.4534,
|
| 1936 |
+
"step": 2030
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 1.12,
|
| 1940 |
+
"learning_rate": 0.0002,
|
| 1941 |
+
"loss": 0.48,
|
| 1942 |
+
"step": 2040
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 1.12,
|
| 1946 |
+
"learning_rate": 0.0002,
|
| 1947 |
+
"loss": 0.5122,
|
| 1948 |
+
"step": 2050
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 1.13,
|
| 1952 |
+
"learning_rate": 0.0002,
|
| 1953 |
+
"loss": 0.4804,
|
| 1954 |
+
"step": 2060
|
| 1955 |
+
},
|
| 1956 |
+
{
|
| 1957 |
+
"epoch": 1.14,
|
| 1958 |
+
"learning_rate": 0.0002,
|
| 1959 |
+
"loss": 0.4901,
|
| 1960 |
+
"step": 2070
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 1.14,
|
| 1964 |
+
"learning_rate": 0.0002,
|
| 1965 |
+
"loss": 0.4926,
|
| 1966 |
+
"step": 2080
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 1.15,
|
| 1970 |
+
"learning_rate": 0.0002,
|
| 1971 |
+
"loss": 0.4472,
|
| 1972 |
+
"step": 2090
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 1.15,
|
| 1976 |
+
"learning_rate": 0.0002,
|
| 1977 |
+
"loss": 0.4732,
|
| 1978 |
+
"step": 2100
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 1.16,
|
| 1982 |
+
"learning_rate": 0.0002,
|
| 1983 |
+
"loss": 0.4895,
|
| 1984 |
+
"step": 2110
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.16,
|
| 1988 |
+
"learning_rate": 0.0002,
|
| 1989 |
+
"loss": 0.4603,
|
| 1990 |
+
"step": 2120
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.17,
|
| 1994 |
+
"learning_rate": 0.0002,
|
| 1995 |
+
"loss": 0.4726,
|
| 1996 |
+
"step": 2130
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 1.17,
|
| 2000 |
+
"learning_rate": 0.0002,
|
| 2001 |
+
"loss": 0.4474,
|
| 2002 |
+
"step": 2140
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 1.18,
|
| 2006 |
+
"learning_rate": 0.0002,
|
| 2007 |
+
"loss": 0.4242,
|
| 2008 |
+
"step": 2150
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 1.18,
|
| 2012 |
+
"learning_rate": 0.0002,
|
| 2013 |
+
"loss": 0.5283,
|
| 2014 |
+
"step": 2160
|
| 2015 |
+
},
|
| 2016 |
+
{
|
| 2017 |
+
"epoch": 1.19,
|
| 2018 |
+
"learning_rate": 0.0002,
|
| 2019 |
+
"loss": 0.4318,
|
| 2020 |
+
"step": 2170
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 1.2,
|
| 2024 |
+
"learning_rate": 0.0002,
|
| 2025 |
+
"loss": 0.4811,
|
| 2026 |
+
"step": 2180
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 1.2,
|
| 2030 |
+
"learning_rate": 0.0002,
|
| 2031 |
+
"loss": 0.4754,
|
| 2032 |
+
"step": 2190
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 1.21,
|
| 2036 |
+
"learning_rate": 0.0002,
|
| 2037 |
+
"loss": 0.473,
|
| 2038 |
+
"step": 2200
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 1.21,
|
| 2042 |
+
"eval_loss": 0.5315611958503723,
|
| 2043 |
+
"eval_runtime": 103.69,
|
| 2044 |
+
"eval_samples_per_second": 9.644,
|
| 2045 |
+
"eval_steps_per_second": 4.822,
|
| 2046 |
+
"step": 2200
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 1.21,
|
| 2050 |
+
"mmlu_eval_accuracy": 0.49373812079675494,
|
| 2051 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 2052 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2053 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 2054 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 2055 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5862068965517241,
|
| 2056 |
+
"mmlu_eval_accuracy_college_biology": 0.3125,
|
| 2057 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2058 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2059 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 2060 |
+
"mmlu_eval_accuracy_college_medicine": 0.5,
|
| 2061 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 2062 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2063 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 2064 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 2065 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 2066 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 2067 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 2068 |
+
"mmlu_eval_accuracy_global_facts": 0.4,
|
| 2069 |
+
"mmlu_eval_accuracy_high_school_biology": 0.46875,
|
| 2070 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.5,
|
| 2071 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2072 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 2073 |
+
"mmlu_eval_accuracy_high_school_geography": 0.9090909090909091,
|
| 2074 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
|
| 2075 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 2076 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
|
| 2077 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
|
| 2078 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 2079 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 2080 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 2081 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
|
| 2082 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 2083 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 2084 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 2085 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 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": 0.7272727272727273,
|
| 2090 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 2091 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 2092 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 2093 |
+
"mmlu_eval_accuracy_moral_disputes": 0.34210526315789475,
|
| 2094 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 2095 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 2096 |
+
"mmlu_eval_accuracy_philosophy": 0.5294117647058824,
|
| 2097 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2098 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 2099 |
+
"mmlu_eval_accuracy_professional_law": 0.3588235294117647,
|
| 2100 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 2101 |
+
"mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
|
| 2102 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 2103 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 2104 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 2105 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 2106 |
+
"mmlu_eval_accuracy_virology": 0.5,
|
| 2107 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2108 |
+
"mmlu_loss": 1.3697273570159085,
|
| 2109 |
+
"step": 2200
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 1.21,
|
| 2113 |
+
"learning_rate": 0.0002,
|
| 2114 |
+
"loss": 0.4855,
|
| 2115 |
+
"step": 2210
|
| 2116 |
+
},
|
| 2117 |
+
{
|
| 2118 |
+
"epoch": 1.22,
|
| 2119 |
+
"learning_rate": 0.0002,
|
| 2120 |
+
"loss": 0.5047,
|
| 2121 |
+
"step": 2220
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 1.22,
|
| 2125 |
+
"learning_rate": 0.0002,
|
| 2126 |
+
"loss": 0.4993,
|
| 2127 |
+
"step": 2230
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 1.23,
|
| 2131 |
+
"learning_rate": 0.0002,
|
| 2132 |
+
"loss": 0.4539,
|
| 2133 |
+
"step": 2240
|
| 2134 |
+
},
|
| 2135 |
+
{
|
| 2136 |
+
"epoch": 1.23,
|
| 2137 |
+
"learning_rate": 0.0002,
|
| 2138 |
+
"loss": 0.4327,
|
| 2139 |
+
"step": 2250
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 1.24,
|
| 2143 |
+
"learning_rate": 0.0002,
|
| 2144 |
+
"loss": 0.4786,
|
| 2145 |
+
"step": 2260
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 1.24,
|
| 2149 |
+
"learning_rate": 0.0002,
|
| 2150 |
+
"loss": 0.4849,
|
| 2151 |
+
"step": 2270
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 1.25,
|
| 2155 |
+
"learning_rate": 0.0002,
|
| 2156 |
+
"loss": 0.4395,
|
| 2157 |
+
"step": 2280
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 1.26,
|
| 2161 |
+
"learning_rate": 0.0002,
|
| 2162 |
+
"loss": 0.4213,
|
| 2163 |
+
"step": 2290
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 1.26,
|
| 2167 |
+
"learning_rate": 0.0002,
|
| 2168 |
+
"loss": 0.5263,
|
| 2169 |
+
"step": 2300
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 1.27,
|
| 2173 |
+
"learning_rate": 0.0002,
|
| 2174 |
+
"loss": 0.4284,
|
| 2175 |
+
"step": 2310
|
| 2176 |
+
},
|
| 2177 |
+
{
|
| 2178 |
+
"epoch": 1.27,
|
| 2179 |
+
"learning_rate": 0.0002,
|
| 2180 |
+
"loss": 0.4465,
|
| 2181 |
+
"step": 2320
|
| 2182 |
+
},
|
| 2183 |
+
{
|
| 2184 |
+
"epoch": 1.28,
|
| 2185 |
+
"learning_rate": 0.0002,
|
| 2186 |
+
"loss": 0.4766,
|
| 2187 |
+
"step": 2330
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 1.28,
|
| 2191 |
+
"learning_rate": 0.0002,
|
| 2192 |
+
"loss": 0.4801,
|
| 2193 |
+
"step": 2340
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 1.29,
|
| 2197 |
+
"learning_rate": 0.0002,
|
| 2198 |
+
"loss": 0.4689,
|
| 2199 |
+
"step": 2350
|
| 2200 |
+
},
|
| 2201 |
+
{
|
| 2202 |
+
"epoch": 1.29,
|
| 2203 |
+
"learning_rate": 0.0002,
|
| 2204 |
+
"loss": 0.4823,
|
| 2205 |
+
"step": 2360
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 1.3,
|
| 2209 |
+
"learning_rate": 0.0002,
|
| 2210 |
+
"loss": 0.5108,
|
| 2211 |
+
"step": 2370
|
| 2212 |
+
},
|
| 2213 |
+
{
|
| 2214 |
+
"epoch": 1.31,
|
| 2215 |
+
"learning_rate": 0.0002,
|
| 2216 |
+
"loss": 0.4787,
|
| 2217 |
+
"step": 2380
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 1.31,
|
| 2221 |
+
"learning_rate": 0.0002,
|
| 2222 |
+
"loss": 0.4892,
|
| 2223 |
+
"step": 2390
|
| 2224 |
+
},
|
| 2225 |
+
{
|
| 2226 |
+
"epoch": 1.32,
|
| 2227 |
+
"learning_rate": 0.0002,
|
| 2228 |
+
"loss": 0.5474,
|
| 2229 |
+
"step": 2400
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 1.32,
|
| 2233 |
+
"eval_loss": 0.527065634727478,
|
| 2234 |
+
"eval_runtime": 103.7154,
|
| 2235 |
+
"eval_samples_per_second": 9.642,
|
| 2236 |
+
"eval_steps_per_second": 4.821,
|
| 2237 |
+
"step": 2400
|
| 2238 |
+
},
|
| 2239 |
+
{
|
| 2240 |
+
"epoch": 1.32,
|
| 2241 |
+
"mmlu_eval_accuracy": 0.490837312752586,
|
| 2242 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
|
| 2243 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2244 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 2245 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 2246 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 2247 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 2248 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2249 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2250 |
+
"mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
|
| 2251 |
+
"mmlu_eval_accuracy_college_medicine": 0.5,
|
| 2252 |
+
"mmlu_eval_accuracy_college_physics": 0.2727272727272727,
|
| 2253 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2254 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 2255 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 2256 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 2257 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 2258 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 2259 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2260 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 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": 0.8181818181818182,
|
| 2265 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 2266 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
|
| 2267 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2268 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
|
| 2269 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 2270 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 2271 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 2272 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2273 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 2274 |
+
"mmlu_eval_accuracy_human_aging": 0.7391304347826086,
|
| 2275 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 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": 0.88,
|
| 2282 |
+
"mmlu_eval_accuracy_medical_genetics": 0.9090909090909091,
|
| 2283 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 2284 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 2285 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 2286 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 2287 |
+
"mmlu_eval_accuracy_philosophy": 0.47058823529411764,
|
| 2288 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2289 |
+
"mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
|
| 2290 |
+
"mmlu_eval_accuracy_professional_law": 0.3764705882352941,
|
| 2291 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 2292 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4492753623188406,
|
| 2293 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 2294 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 2295 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 2296 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 2297 |
+
"mmlu_eval_accuracy_virology": 0.6111111111111112,
|
| 2298 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2299 |
+
"mmlu_loss": 1.3973398435053563,
|
| 2300 |
+
"step": 2400
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 1.32,
|
| 2304 |
+
"learning_rate": 0.0002,
|
| 2305 |
+
"loss": 0.4455,
|
| 2306 |
+
"step": 2410
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 1.33,
|
| 2310 |
+
"learning_rate": 0.0002,
|
| 2311 |
+
"loss": 0.488,
|
| 2312 |
+
"step": 2420
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 1.33,
|
| 2316 |
+
"learning_rate": 0.0002,
|
| 2317 |
+
"loss": 0.4228,
|
| 2318 |
+
"step": 2430
|
| 2319 |
+
},
|
| 2320 |
+
{
|
| 2321 |
+
"epoch": 1.34,
|
| 2322 |
+
"learning_rate": 0.0002,
|
| 2323 |
+
"loss": 0.4649,
|
| 2324 |
+
"step": 2440
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 1.34,
|
| 2328 |
+
"learning_rate": 0.0002,
|
| 2329 |
+
"loss": 0.4978,
|
| 2330 |
+
"step": 2450
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 1.35,
|
| 2334 |
+
"learning_rate": 0.0002,
|
| 2335 |
+
"loss": 0.4979,
|
| 2336 |
+
"step": 2460
|
| 2337 |
+
},
|
| 2338 |
+
{
|
| 2339 |
+
"epoch": 1.35,
|
| 2340 |
+
"learning_rate": 0.0002,
|
| 2341 |
+
"loss": 0.4918,
|
| 2342 |
+
"step": 2470
|
| 2343 |
+
},
|
| 2344 |
+
{
|
| 2345 |
+
"epoch": 1.36,
|
| 2346 |
+
"learning_rate": 0.0002,
|
| 2347 |
+
"loss": 0.4598,
|
| 2348 |
+
"step": 2480
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 1.37,
|
| 2352 |
+
"learning_rate": 0.0002,
|
| 2353 |
+
"loss": 0.4814,
|
| 2354 |
+
"step": 2490
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 1.37,
|
| 2358 |
+
"learning_rate": 0.0002,
|
| 2359 |
+
"loss": 0.4902,
|
| 2360 |
+
"step": 2500
|
| 2361 |
+
},
|
| 2362 |
+
{
|
| 2363 |
+
"epoch": 1.38,
|
| 2364 |
+
"learning_rate": 0.0002,
|
| 2365 |
+
"loss": 0.4968,
|
| 2366 |
+
"step": 2510
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 1.38,
|
| 2370 |
+
"learning_rate": 0.0002,
|
| 2371 |
+
"loss": 0.4729,
|
| 2372 |
+
"step": 2520
|
| 2373 |
+
},
|
| 2374 |
+
{
|
| 2375 |
+
"epoch": 1.39,
|
| 2376 |
+
"learning_rate": 0.0002,
|
| 2377 |
+
"loss": 0.5053,
|
| 2378 |
+
"step": 2530
|
| 2379 |
+
},
|
| 2380 |
+
{
|
| 2381 |
+
"epoch": 1.39,
|
| 2382 |
+
"learning_rate": 0.0002,
|
| 2383 |
+
"loss": 0.4799,
|
| 2384 |
+
"step": 2540
|
| 2385 |
+
},
|
| 2386 |
+
{
|
| 2387 |
+
"epoch": 1.4,
|
| 2388 |
+
"learning_rate": 0.0002,
|
| 2389 |
+
"loss": 0.4841,
|
| 2390 |
+
"step": 2550
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 1.4,
|
| 2394 |
+
"learning_rate": 0.0002,
|
| 2395 |
+
"loss": 0.5059,
|
| 2396 |
+
"step": 2560
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.41,
|
| 2400 |
+
"learning_rate": 0.0002,
|
| 2401 |
+
"loss": 0.4522,
|
| 2402 |
+
"step": 2570
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 1.41,
|
| 2406 |
+
"learning_rate": 0.0002,
|
| 2407 |
+
"loss": 0.4599,
|
| 2408 |
+
"step": 2580
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 1.42,
|
| 2412 |
+
"learning_rate": 0.0002,
|
| 2413 |
+
"loss": 0.5075,
|
| 2414 |
+
"step": 2590
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 1.43,
|
| 2418 |
+
"learning_rate": 0.0002,
|
| 2419 |
+
"loss": 0.4774,
|
| 2420 |
+
"step": 2600
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 1.43,
|
| 2424 |
+
"eval_loss": 0.5256263613700867,
|
| 2425 |
+
"eval_runtime": 103.8048,
|
| 2426 |
+
"eval_samples_per_second": 9.633,
|
| 2427 |
+
"eval_steps_per_second": 4.817,
|
| 2428 |
+
"step": 2600
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 1.43,
|
| 2432 |
+
"mmlu_eval_accuracy": 0.4964072558531528,
|
| 2433 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 2434 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2435 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 2436 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 2437 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 2438 |
+
"mmlu_eval_accuracy_college_biology": 0.375,
|
| 2439 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2440 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2441 |
+
"mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
|
| 2442 |
+
"mmlu_eval_accuracy_college_medicine": 0.5909090909090909,
|
| 2443 |
+
"mmlu_eval_accuracy_college_physics": 0.36363636363636365,
|
| 2444 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2445 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 2446 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 2447 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 2448 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
|
| 2449 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 2450 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2451 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 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": 0.6111111111111112,
|
| 2455 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 2456 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
|
| 2457 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
|
| 2458 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2459 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
|
| 2460 |
+
"mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
|
| 2461 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 2462 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
|
| 2463 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
|
| 2464 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 2465 |
+
"mmlu_eval_accuracy_human_aging": 0.782608695652174,
|
| 2466 |
+
"mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
|
| 2467 |
+
"mmlu_eval_accuracy_international_law": 0.8461538461538461,
|
| 2468 |
+
"mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
|
| 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": 0.8,
|
| 2473 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 2474 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
|
| 2475 |
+
"mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
|
| 2476 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 2477 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 2478 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 2479 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2480 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 2481 |
+
"mmlu_eval_accuracy_professional_law": 0.34705882352941175,
|
| 2482 |
+
"mmlu_eval_accuracy_professional_medicine": 0.6129032258064516,
|
| 2483 |
+
"mmlu_eval_accuracy_professional_psychology": 0.4492753623188406,
|
| 2484 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 2485 |
+
"mmlu_eval_accuracy_security_studies": 0.5185185185185185,
|
| 2486 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 2487 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.7272727272727273,
|
| 2488 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 2489 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2490 |
+
"mmlu_loss": 1.3019743448916055,
|
| 2491 |
+
"step": 2600
|
| 2492 |
+
},
|
| 2493 |
+
{
|
| 2494 |
+
"epoch": 1.43,
|
| 2495 |
+
"learning_rate": 0.0002,
|
| 2496 |
+
"loss": 0.4856,
|
| 2497 |
+
"step": 2610
|
| 2498 |
+
},
|
| 2499 |
+
{
|
| 2500 |
+
"epoch": 1.44,
|
| 2501 |
+
"learning_rate": 0.0002,
|
| 2502 |
+
"loss": 0.4981,
|
| 2503 |
+
"step": 2620
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 1.44,
|
| 2507 |
+
"learning_rate": 0.0002,
|
| 2508 |
+
"loss": 0.4695,
|
| 2509 |
+
"step": 2630
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 1.45,
|
| 2513 |
+
"learning_rate": 0.0002,
|
| 2514 |
+
"loss": 0.484,
|
| 2515 |
+
"step": 2640
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 1.45,
|
| 2519 |
+
"learning_rate": 0.0002,
|
| 2520 |
+
"loss": 0.4417,
|
| 2521 |
+
"step": 2650
|
| 2522 |
+
},
|
| 2523 |
+
{
|
| 2524 |
+
"epoch": 1.46,
|
| 2525 |
+
"learning_rate": 0.0002,
|
| 2526 |
+
"loss": 0.4862,
|
| 2527 |
+
"step": 2660
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 1.46,
|
| 2531 |
+
"learning_rate": 0.0002,
|
| 2532 |
+
"loss": 0.4458,
|
| 2533 |
+
"step": 2670
|
| 2534 |
+
},
|
| 2535 |
+
{
|
| 2536 |
+
"epoch": 1.47,
|
| 2537 |
+
"learning_rate": 0.0002,
|
| 2538 |
+
"loss": 0.4836,
|
| 2539 |
+
"step": 2680
|
| 2540 |
+
},
|
| 2541 |
+
{
|
| 2542 |
+
"epoch": 1.47,
|
| 2543 |
+
"learning_rate": 0.0002,
|
| 2544 |
+
"loss": 0.4855,
|
| 2545 |
+
"step": 2690
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 1.48,
|
| 2549 |
+
"learning_rate": 0.0002,
|
| 2550 |
+
"loss": 0.4548,
|
| 2551 |
+
"step": 2700
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 1.49,
|
| 2555 |
+
"learning_rate": 0.0002,
|
| 2556 |
+
"loss": 0.418,
|
| 2557 |
+
"step": 2710
|
| 2558 |
+
},
|
| 2559 |
+
{
|
| 2560 |
+
"epoch": 1.49,
|
| 2561 |
+
"learning_rate": 0.0002,
|
| 2562 |
+
"loss": 0.444,
|
| 2563 |
+
"step": 2720
|
| 2564 |
+
},
|
| 2565 |
+
{
|
| 2566 |
+
"epoch": 1.5,
|
| 2567 |
+
"learning_rate": 0.0002,
|
| 2568 |
+
"loss": 0.486,
|
| 2569 |
+
"step": 2730
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 1.5,
|
| 2573 |
+
"learning_rate": 0.0002,
|
| 2574 |
+
"loss": 0.43,
|
| 2575 |
+
"step": 2740
|
| 2576 |
+
},
|
| 2577 |
+
{
|
| 2578 |
+
"epoch": 1.51,
|
| 2579 |
+
"learning_rate": 0.0002,
|
| 2580 |
+
"loss": 0.4872,
|
| 2581 |
+
"step": 2750
|
| 2582 |
+
},
|
| 2583 |
+
{
|
| 2584 |
+
"epoch": 1.51,
|
| 2585 |
+
"learning_rate": 0.0002,
|
| 2586 |
+
"loss": 0.476,
|
| 2587 |
+
"step": 2760
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 1.52,
|
| 2591 |
+
"learning_rate": 0.0002,
|
| 2592 |
+
"loss": 0.5366,
|
| 2593 |
+
"step": 2770
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 1.52,
|
| 2597 |
+
"learning_rate": 0.0002,
|
| 2598 |
+
"loss": 0.511,
|
| 2599 |
+
"step": 2780
|
| 2600 |
+
},
|
| 2601 |
+
{
|
| 2602 |
+
"epoch": 1.53,
|
| 2603 |
+
"learning_rate": 0.0002,
|
| 2604 |
+
"loss": 0.4935,
|
| 2605 |
+
"step": 2790
|
| 2606 |
+
},
|
| 2607 |
+
{
|
| 2608 |
+
"epoch": 1.54,
|
| 2609 |
+
"learning_rate": 0.0002,
|
| 2610 |
+
"loss": 0.4932,
|
| 2611 |
+
"step": 2800
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 1.54,
|
| 2615 |
+
"eval_loss": 0.5219016075134277,
|
| 2616 |
+
"eval_runtime": 103.5568,
|
| 2617 |
+
"eval_samples_per_second": 9.657,
|
| 2618 |
+
"eval_steps_per_second": 4.828,
|
| 2619 |
+
"step": 2800
|
| 2620 |
+
},
|
| 2621 |
+
{
|
| 2622 |
+
"epoch": 1.54,
|
| 2623 |
+
"mmlu_eval_accuracy": 0.5042748824180435,
|
| 2624 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
|
| 2625 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2626 |
+
"mmlu_eval_accuracy_astronomy": 0.5625,
|
| 2627 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 2628 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5172413793103449,
|
| 2629 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 2630 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2631 |
+
"mmlu_eval_accuracy_college_computer_science": 0.18181818181818182,
|
| 2632 |
+
"mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
|
| 2633 |
+
"mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
|
| 2634 |
+
"mmlu_eval_accuracy_college_physics": 0.45454545454545453,
|
| 2635 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 2636 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
|
| 2637 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 2638 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.375,
|
| 2639 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 2640 |
+
"mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
|
| 2641 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2642 |
+
"mmlu_eval_accuracy_high_school_biology": 0.4375,
|
| 2643 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454,
|
| 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": 0.8181818181818182,
|
| 2647 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
|
| 2648 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
|
| 2649 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2650 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5,
|
| 2651 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 2652 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 2653 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
|
| 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": 0.782608695652174,
|
| 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": 0.2727272727272727,
|
| 2662 |
+
"mmlu_eval_accuracy_management": 0.7272727272727273,
|
| 2663 |
+
"mmlu_eval_accuracy_marketing": 0.84,
|
| 2664 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 2665 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 2666 |
+
"mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
|
| 2667 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 2668 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 2669 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 2670 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2671 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 2672 |
+
"mmlu_eval_accuracy_professional_law": 0.34705882352941175,
|
| 2673 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 2674 |
+
"mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
|
| 2675 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 2676 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 2677 |
+
"mmlu_eval_accuracy_sociology": 0.6818181818181818,
|
| 2678 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
|
| 2679 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 2680 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2681 |
+
"mmlu_loss": 1.424923150797737,
|
| 2682 |
+
"step": 2800
|
| 2683 |
+
},
|
| 2684 |
+
{
|
| 2685 |
+
"epoch": 1.54,
|
| 2686 |
+
"learning_rate": 0.0002,
|
| 2687 |
+
"loss": 0.4884,
|
| 2688 |
+
"step": 2810
|
| 2689 |
+
},
|
| 2690 |
+
{
|
| 2691 |
+
"epoch": 1.55,
|
| 2692 |
+
"learning_rate": 0.0002,
|
| 2693 |
+
"loss": 0.48,
|
| 2694 |
+
"step": 2820
|
| 2695 |
+
},
|
| 2696 |
+
{
|
| 2697 |
+
"epoch": 1.55,
|
| 2698 |
+
"learning_rate": 0.0002,
|
| 2699 |
+
"loss": 0.4884,
|
| 2700 |
+
"step": 2830
|
| 2701 |
+
},
|
| 2702 |
+
{
|
| 2703 |
+
"epoch": 1.56,
|
| 2704 |
+
"learning_rate": 0.0002,
|
| 2705 |
+
"loss": 0.4609,
|
| 2706 |
+
"step": 2840
|
| 2707 |
+
},
|
| 2708 |
+
{
|
| 2709 |
+
"epoch": 1.56,
|
| 2710 |
+
"learning_rate": 0.0002,
|
| 2711 |
+
"loss": 0.4654,
|
| 2712 |
+
"step": 2850
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 1.57,
|
| 2716 |
+
"learning_rate": 0.0002,
|
| 2717 |
+
"loss": 0.4916,
|
| 2718 |
+
"step": 2860
|
| 2719 |
+
},
|
| 2720 |
+
{
|
| 2721 |
+
"epoch": 1.57,
|
| 2722 |
+
"learning_rate": 0.0002,
|
| 2723 |
+
"loss": 0.4665,
|
| 2724 |
+
"step": 2870
|
| 2725 |
+
},
|
| 2726 |
+
{
|
| 2727 |
+
"epoch": 1.58,
|
| 2728 |
+
"learning_rate": 0.0002,
|
| 2729 |
+
"loss": 0.4697,
|
| 2730 |
+
"step": 2880
|
| 2731 |
+
},
|
| 2732 |
+
{
|
| 2733 |
+
"epoch": 1.58,
|
| 2734 |
+
"learning_rate": 0.0002,
|
| 2735 |
+
"loss": 0.5023,
|
| 2736 |
+
"step": 2890
|
| 2737 |
+
},
|
| 2738 |
+
{
|
| 2739 |
+
"epoch": 1.59,
|
| 2740 |
+
"learning_rate": 0.0002,
|
| 2741 |
+
"loss": 0.4947,
|
| 2742 |
+
"step": 2900
|
| 2743 |
+
},
|
| 2744 |
+
{
|
| 2745 |
+
"epoch": 1.6,
|
| 2746 |
+
"learning_rate": 0.0002,
|
| 2747 |
+
"loss": 0.4731,
|
| 2748 |
+
"step": 2910
|
| 2749 |
+
},
|
| 2750 |
+
{
|
| 2751 |
+
"epoch": 1.6,
|
| 2752 |
+
"learning_rate": 0.0002,
|
| 2753 |
+
"loss": 0.4735,
|
| 2754 |
+
"step": 2920
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 1.61,
|
| 2758 |
+
"learning_rate": 0.0002,
|
| 2759 |
+
"loss": 0.4335,
|
| 2760 |
+
"step": 2930
|
| 2761 |
+
},
|
| 2762 |
+
{
|
| 2763 |
+
"epoch": 1.61,
|
| 2764 |
+
"learning_rate": 0.0002,
|
| 2765 |
+
"loss": 0.4311,
|
| 2766 |
+
"step": 2940
|
| 2767 |
+
},
|
| 2768 |
+
{
|
| 2769 |
+
"epoch": 1.62,
|
| 2770 |
+
"learning_rate": 0.0002,
|
| 2771 |
+
"loss": 0.4756,
|
| 2772 |
+
"step": 2950
|
| 2773 |
+
},
|
| 2774 |
+
{
|
| 2775 |
+
"epoch": 1.62,
|
| 2776 |
+
"learning_rate": 0.0002,
|
| 2777 |
+
"loss": 0.4819,
|
| 2778 |
+
"step": 2960
|
| 2779 |
+
},
|
| 2780 |
+
{
|
| 2781 |
+
"epoch": 1.63,
|
| 2782 |
+
"learning_rate": 0.0002,
|
| 2783 |
+
"loss": 0.4611,
|
| 2784 |
+
"step": 2970
|
| 2785 |
+
},
|
| 2786 |
+
{
|
| 2787 |
+
"epoch": 1.63,
|
| 2788 |
+
"learning_rate": 0.0002,
|
| 2789 |
+
"loss": 0.4982,
|
| 2790 |
+
"step": 2980
|
| 2791 |
+
},
|
| 2792 |
+
{
|
| 2793 |
+
"epoch": 1.64,
|
| 2794 |
+
"learning_rate": 0.0002,
|
| 2795 |
+
"loss": 0.4626,
|
| 2796 |
+
"step": 2990
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 1.64,
|
| 2800 |
+
"learning_rate": 0.0002,
|
| 2801 |
+
"loss": 0.453,
|
| 2802 |
+
"step": 3000
|
| 2803 |
+
},
|
| 2804 |
+
{
|
| 2805 |
+
"epoch": 1.64,
|
| 2806 |
+
"eval_loss": 0.5194316506385803,
|
| 2807 |
+
"eval_runtime": 103.6672,
|
| 2808 |
+
"eval_samples_per_second": 9.646,
|
| 2809 |
+
"eval_steps_per_second": 4.823,
|
| 2810 |
+
"step": 3000
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 1.64,
|
| 2814 |
+
"mmlu_eval_accuracy": 0.49891729256300676,
|
| 2815 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 2816 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 2817 |
+
"mmlu_eval_accuracy_astronomy": 0.4375,
|
| 2818 |
+
"mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
|
| 2819 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 2820 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 2821 |
+
"mmlu_eval_accuracy_college_chemistry": 0.25,
|
| 2822 |
+
"mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
|
| 2823 |
+
"mmlu_eval_accuracy_college_mathematics": 0.09090909090909091,
|
| 2824 |
+
"mmlu_eval_accuracy_college_medicine": 0.5,
|
| 2825 |
+
"mmlu_eval_accuracy_college_physics": 0.6363636363636364,
|
| 2826 |
+
"mmlu_eval_accuracy_computer_security": 0.5454545454545454,
|
| 2827 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
|
| 2828 |
+
"mmlu_eval_accuracy_econometrics": 0.16666666666666666,
|
| 2829 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.3125,
|
| 2830 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
|
| 2831 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 2832 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 2833 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 2834 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.5454545454545454,
|
| 2835 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 2836 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
|
| 2837 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 2838 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
|
| 2839 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.46511627906976744,
|
| 2840 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
|
| 2841 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.5,
|
| 2842 |
+
"mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
|
| 2843 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 2844 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
|
| 2845 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 2846 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 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": 1.0,
|
| 2856 |
+
"mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
|
| 2857 |
+
"mmlu_eval_accuracy_moral_disputes": 0.3684210526315789,
|
| 2858 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.25,
|
| 2859 |
+
"mmlu_eval_accuracy_nutrition": 0.6666666666666666,
|
| 2860 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 2861 |
+
"mmlu_eval_accuracy_prehistory": 0.5142857142857142,
|
| 2862 |
+
"mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
|
| 2863 |
+
"mmlu_eval_accuracy_professional_law": 0.3588235294117647,
|
| 2864 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5806451612903226,
|
| 2865 |
+
"mmlu_eval_accuracy_professional_psychology": 0.391304347826087,
|
| 2866 |
+
"mmlu_eval_accuracy_public_relations": 0.5833333333333334,
|
| 2867 |
+
"mmlu_eval_accuracy_security_studies": 0.48148148148148145,
|
| 2868 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 2869 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
|
| 2870 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 2871 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 2872 |
+
"mmlu_loss": 1.350923795807455,
|
| 2873 |
+
"step": 3000
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 1.65,
|
| 2877 |
+
"learning_rate": 0.0002,
|
| 2878 |
+
"loss": 0.4553,
|
| 2879 |
+
"step": 3010
|
| 2880 |
+
},
|
| 2881 |
+
{
|
| 2882 |
+
"epoch": 1.66,
|
| 2883 |
+
"learning_rate": 0.0002,
|
| 2884 |
+
"loss": 0.4995,
|
| 2885 |
+
"step": 3020
|
| 2886 |
+
},
|
| 2887 |
+
{
|
| 2888 |
+
"epoch": 1.66,
|
| 2889 |
+
"learning_rate": 0.0002,
|
| 2890 |
+
"loss": 0.5203,
|
| 2891 |
+
"step": 3030
|
| 2892 |
+
},
|
| 2893 |
+
{
|
| 2894 |
+
"epoch": 1.67,
|
| 2895 |
+
"learning_rate": 0.0002,
|
| 2896 |
+
"loss": 0.4755,
|
| 2897 |
+
"step": 3040
|
| 2898 |
+
},
|
| 2899 |
+
{
|
| 2900 |
+
"epoch": 1.67,
|
| 2901 |
+
"learning_rate": 0.0002,
|
| 2902 |
+
"loss": 0.4353,
|
| 2903 |
+
"step": 3050
|
| 2904 |
+
},
|
| 2905 |
+
{
|
| 2906 |
+
"epoch": 1.68,
|
| 2907 |
+
"learning_rate": 0.0002,
|
| 2908 |
+
"loss": 0.4401,
|
| 2909 |
+
"step": 3060
|
| 2910 |
+
},
|
| 2911 |
+
{
|
| 2912 |
+
"epoch": 1.68,
|
| 2913 |
+
"learning_rate": 0.0002,
|
| 2914 |
+
"loss": 0.4387,
|
| 2915 |
+
"step": 3070
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 1.69,
|
| 2919 |
+
"learning_rate": 0.0002,
|
| 2920 |
+
"loss": 0.4298,
|
| 2921 |
+
"step": 3080
|
| 2922 |
+
},
|
| 2923 |
+
{
|
| 2924 |
+
"epoch": 1.69,
|
| 2925 |
+
"learning_rate": 0.0002,
|
| 2926 |
+
"loss": 0.4794,
|
| 2927 |
+
"step": 3090
|
| 2928 |
+
},
|
| 2929 |
+
{
|
| 2930 |
+
"epoch": 1.7,
|
| 2931 |
+
"learning_rate": 0.0002,
|
| 2932 |
+
"loss": 0.4355,
|
| 2933 |
+
"step": 3100
|
| 2934 |
+
},
|
| 2935 |
+
{
|
| 2936 |
+
"epoch": 1.71,
|
| 2937 |
+
"learning_rate": 0.0002,
|
| 2938 |
+
"loss": 0.433,
|
| 2939 |
+
"step": 3110
|
| 2940 |
+
},
|
| 2941 |
+
{
|
| 2942 |
+
"epoch": 1.71,
|
| 2943 |
+
"learning_rate": 0.0002,
|
| 2944 |
+
"loss": 0.5218,
|
| 2945 |
+
"step": 3120
|
| 2946 |
+
},
|
| 2947 |
+
{
|
| 2948 |
+
"epoch": 1.72,
|
| 2949 |
+
"learning_rate": 0.0002,
|
| 2950 |
+
"loss": 0.4604,
|
| 2951 |
+
"step": 3130
|
| 2952 |
+
},
|
| 2953 |
+
{
|
| 2954 |
+
"epoch": 1.72,
|
| 2955 |
+
"learning_rate": 0.0002,
|
| 2956 |
+
"loss": 0.4624,
|
| 2957 |
+
"step": 3140
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 1.73,
|
| 2961 |
+
"learning_rate": 0.0002,
|
| 2962 |
+
"loss": 0.4368,
|
| 2963 |
+
"step": 3150
|
| 2964 |
+
},
|
| 2965 |
+
{
|
| 2966 |
+
"epoch": 1.73,
|
| 2967 |
+
"learning_rate": 0.0002,
|
| 2968 |
+
"loss": 0.5012,
|
| 2969 |
+
"step": 3160
|
| 2970 |
+
},
|
| 2971 |
+
{
|
| 2972 |
+
"epoch": 1.74,
|
| 2973 |
+
"learning_rate": 0.0002,
|
| 2974 |
+
"loss": 0.4795,
|
| 2975 |
+
"step": 3170
|
| 2976 |
+
},
|
| 2977 |
+
{
|
| 2978 |
+
"epoch": 1.74,
|
| 2979 |
+
"learning_rate": 0.0002,
|
| 2980 |
+
"loss": 0.4809,
|
| 2981 |
+
"step": 3180
|
| 2982 |
+
},
|
| 2983 |
+
{
|
| 2984 |
+
"epoch": 1.75,
|
| 2985 |
+
"learning_rate": 0.0002,
|
| 2986 |
+
"loss": 0.4331,
|
| 2987 |
+
"step": 3190
|
| 2988 |
+
},
|
| 2989 |
+
{
|
| 2990 |
+
"epoch": 1.75,
|
| 2991 |
+
"learning_rate": 0.0002,
|
| 2992 |
+
"loss": 0.4311,
|
| 2993 |
+
"step": 3200
|
| 2994 |
+
},
|
| 2995 |
+
{
|
| 2996 |
+
"epoch": 1.75,
|
| 2997 |
+
"eval_loss": 0.5174126029014587,
|
| 2998 |
+
"eval_runtime": 103.6744,
|
| 2999 |
+
"eval_samples_per_second": 9.646,
|
| 3000 |
+
"eval_steps_per_second": 4.823,
|
| 3001 |
+
"step": 3200
|
| 3002 |
+
},
|
| 3003 |
+
{
|
| 3004 |
+
"epoch": 1.75,
|
| 3005 |
+
"mmlu_eval_accuracy": 0.4920535786873217,
|
| 3006 |
+
"mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
|
| 3007 |
+
"mmlu_eval_accuracy_anatomy": 0.6428571428571429,
|
| 3008 |
+
"mmlu_eval_accuracy_astronomy": 0.5,
|
| 3009 |
+
"mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
|
| 3010 |
+
"mmlu_eval_accuracy_clinical_knowledge": 0.5517241379310345,
|
| 3011 |
+
"mmlu_eval_accuracy_college_biology": 0.4375,
|
| 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": 0.5,
|
| 3016 |
+
"mmlu_eval_accuracy_college_physics": 0.5454545454545454,
|
| 3017 |
+
"mmlu_eval_accuracy_computer_security": 0.45454545454545453,
|
| 3018 |
+
"mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
|
| 3019 |
+
"mmlu_eval_accuracy_econometrics": 0.25,
|
| 3020 |
+
"mmlu_eval_accuracy_electrical_engineering": 0.25,
|
| 3021 |
+
"mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
|
| 3022 |
+
"mmlu_eval_accuracy_formal_logic": 0.07142857142857142,
|
| 3023 |
+
"mmlu_eval_accuracy_global_facts": 0.5,
|
| 3024 |
+
"mmlu_eval_accuracy_high_school_biology": 0.40625,
|
| 3025 |
+
"mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
|
| 3026 |
+
"mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
|
| 3027 |
+
"mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
|
| 3028 |
+
"mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
|
| 3029 |
+
"mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
|
| 3030 |
+
"mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
|
| 3031 |
+
"mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
|
| 3032 |
+
"mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
|
| 3033 |
+
"mmlu_eval_accuracy_high_school_physics": 0.11764705882352941,
|
| 3034 |
+
"mmlu_eval_accuracy_high_school_psychology": 0.8666666666666667,
|
| 3035 |
+
"mmlu_eval_accuracy_high_school_statistics": 0.5217391304347826,
|
| 3036 |
+
"mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
|
| 3037 |
+
"mmlu_eval_accuracy_high_school_world_history": 0.7692307692307693,
|
| 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": 0.36363636363636365,
|
| 3042 |
+
"mmlu_eval_accuracy_logical_fallacies": 0.6111111111111112,
|
| 3043 |
+
"mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
|
| 3044 |
+
"mmlu_eval_accuracy_management": 0.6363636363636364,
|
| 3045 |
+
"mmlu_eval_accuracy_marketing": 0.88,
|
| 3046 |
+
"mmlu_eval_accuracy_medical_genetics": 1.0,
|
| 3047 |
+
"mmlu_eval_accuracy_miscellaneous": 0.627906976744186,
|
| 3048 |
+
"mmlu_eval_accuracy_moral_disputes": 0.3684210526315789,
|
| 3049 |
+
"mmlu_eval_accuracy_moral_scenarios": 0.26,
|
| 3050 |
+
"mmlu_eval_accuracy_nutrition": 0.696969696969697,
|
| 3051 |
+
"mmlu_eval_accuracy_philosophy": 0.5,
|
| 3052 |
+
"mmlu_eval_accuracy_prehistory": 0.4857142857142857,
|
| 3053 |
+
"mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
|
| 3054 |
+
"mmlu_eval_accuracy_professional_law": 0.35294117647058826,
|
| 3055 |
+
"mmlu_eval_accuracy_professional_medicine": 0.5483870967741935,
|
| 3056 |
+
"mmlu_eval_accuracy_professional_psychology": 0.463768115942029,
|
| 3057 |
+
"mmlu_eval_accuracy_public_relations": 0.6666666666666666,
|
| 3058 |
+
"mmlu_eval_accuracy_security_studies": 0.4444444444444444,
|
| 3059 |
+
"mmlu_eval_accuracy_sociology": 0.6363636363636364,
|
| 3060 |
+
"mmlu_eval_accuracy_us_foreign_policy": 0.45454545454545453,
|
| 3061 |
+
"mmlu_eval_accuracy_virology": 0.5555555555555556,
|
| 3062 |
+
"mmlu_eval_accuracy_world_religions": 0.6842105263157895,
|
| 3063 |
+
"mmlu_loss": 1.3065739216443453,
|
| 3064 |
+
"step": 3200
|
| 3065 |
}
|
| 3066 |
],
|
| 3067 |
"max_steps": 10000,
|
| 3068 |
"num_train_epochs": 6,
|
| 3069 |
+
"total_flos": 9.210373571942646e+17,
|
| 3070 |
"trial_name": null,
|
| 3071 |
"trial_params": null
|
| 3072 |
}
|
{checkpoint-1000 → checkpoint-3200}/training_args.bin
RENAMED
|
File without changes
|