Farouk commited on
Commit ·
1fa1300
1
Parent(s): 1d4e441
Training in progress, step 600
Browse files- adapter_model.bin +1 -1
- {checkpoint-200 → checkpoint-400/adapter_model/adapter_model}/README.md +0 -0
- {checkpoint-200 → checkpoint-400/adapter_model/adapter_model}/adapter_config.json +0 -0
- {checkpoint-200 → checkpoint-400/adapter_model/adapter_model}/adapter_model.bin +1 -1
- {checkpoint-200/adapter_model/adapter_model → checkpoint-600}/README.md +0 -0
- {checkpoint-200/adapter_model/adapter_model → checkpoint-600}/adapter_config.json +0 -0
- {checkpoint-200/adapter_model/adapter_model → checkpoint-600}/adapter_model.bin +1 -1
- {checkpoint-200 → checkpoint-600}/added_tokens.json +0 -0
- {checkpoint-200 → checkpoint-600}/optimizer.pt +2 -2
- {checkpoint-200 → checkpoint-600}/rng_state.pth +1 -1
- {checkpoint-200 → checkpoint-600}/scheduler.pt +1 -1
- {checkpoint-200 → checkpoint-600}/special_tokens_map.json +0 -0
- {checkpoint-200 → checkpoint-600}/tokenizer.model +0 -0
- {checkpoint-200 → checkpoint-600}/tokenizer_config.json +0 -0
- {checkpoint-200 → checkpoint-600}/trainer_state.json +2773 -5
- {checkpoint-200 → checkpoint-600}/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:bde6f04612dc19db50ddda3a658e889e1c30d546abe7ad825b1aff70c5c11175
|
| 3 |
size 319977229
|
{checkpoint-200 → checkpoint-400/adapter_model/adapter_model}/README.md
RENAMED
|
File without changes
|
{checkpoint-200 → checkpoint-400/adapter_model/adapter_model}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-200 → checkpoint-400/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 sha256:2ea9607443b0ff3ee7637f15cdb79ac3f93543ff2572cf1ee53654822049e53c
|
| 3 |
size 319977229
|
{checkpoint-200/adapter_model/adapter_model → checkpoint-600}/README.md
RENAMED
|
File without changes
|
{checkpoint-200/adapter_model/adapter_model → checkpoint-600}/adapter_config.json
RENAMED
|
File without changes
|
{checkpoint-200/adapter_model/adapter_model → checkpoint-600}/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:bde6f04612dc19db50ddda3a658e889e1c30d546abe7ad825b1aff70c5c11175
|
| 3 |
size 319977229
|
{checkpoint-200 → checkpoint-600}/added_tokens.json
RENAMED
|
File without changes
|
{checkpoint-200 → checkpoint-600}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ed42fb1f94e4f5a70bd119411e355ee7336303b88a4a60c3924604aebc5f95a1
|
| 3 |
+
size 1279539973
|
{checkpoint-200 → checkpoint-600}/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 sha256:428800511b314d28f9b52a9e7f3e52507c9b3807682a1f0b719700359060bcc4
|
| 3 |
size 14511
|
{checkpoint-200 → checkpoint-600}/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:ec86fe6259605d8ef886cd8fc711f042f8248ddd86839f789c7a9c4f568fa3a5
|
| 3 |
size 627
|
{checkpoint-200 → checkpoint-600}/special_tokens_map.json
RENAMED
|
File without changes
|
{checkpoint-200 → checkpoint-600}/tokenizer.model
RENAMED
|
File without changes
|
{checkpoint-200 → checkpoint-600}/tokenizer_config.json
RENAMED
|
File without changes
|
{checkpoint-200 → checkpoint-600}/trainer_state.json
RENAMED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 0.
|
| 3 |
-
"best_model_checkpoint": "./output_v2/7b_cluster00_Nous-Hermes-llama-2-7b_codellama_blob_1/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -1390,11 +1390,2779 @@
|
|
| 1390 |
"dharma_loss": 2.4329242930412294,
|
| 1391 |
"epoch": 0.56,
|
| 1392 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1393 |
}
|
| 1394 |
],
|
| 1395 |
"max_steps": 1000,
|
| 1396 |
"num_train_epochs": 3,
|
| 1397 |
-
"total_flos":
|
| 1398 |
"trial_name": null,
|
| 1399 |
"trial_params": null
|
| 1400 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 0.5050536394119263,
|
| 3 |
+
"best_model_checkpoint": "./output_v2/7b_cluster00_Nous-Hermes-llama-2-7b_codellama_blob_1/checkpoint-600",
|
| 4 |
+
"epoch": 1.6842105263157894,
|
| 5 |
+
"global_step": 600,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 1390 |
"dharma_loss": 2.4329242930412294,
|
| 1391 |
"epoch": 0.56,
|
| 1392 |
"step": 200
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.56,
|
| 1396 |
+
"learning_rate": 0.00027757768148179275,
|
| 1397 |
+
"loss": 0.4509,
|
| 1398 |
+
"step": 201
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.57,
|
| 1402 |
+
"learning_rate": 0.00027732150118568017,
|
| 1403 |
+
"loss": 0.562,
|
| 1404 |
+
"step": 202
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.57,
|
| 1408 |
+
"learning_rate": 0.00027706398534726757,
|
| 1409 |
+
"loss": 0.4544,
|
| 1410 |
+
"step": 203
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.57,
|
| 1414 |
+
"learning_rate": 0.00027680513666777435,
|
| 1415 |
+
"loss": 0.4827,
|
| 1416 |
+
"step": 204
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.58,
|
| 1420 |
+
"learning_rate": 0.00027654495786240066,
|
| 1421 |
+
"loss": 0.4569,
|
| 1422 |
+
"step": 205
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 0.58,
|
| 1426 |
+
"learning_rate": 0.00027628345166029906,
|
| 1427 |
+
"loss": 0.5267,
|
| 1428 |
+
"step": 206
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.58,
|
| 1432 |
+
"learning_rate": 0.00027602062080454577,
|
| 1433 |
+
"loss": 0.6012,
|
| 1434 |
+
"step": 207
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.58,
|
| 1438 |
+
"learning_rate": 0.0002757564680521122,
|
| 1439 |
+
"loss": 0.5415,
|
| 1440 |
+
"step": 208
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.59,
|
| 1444 |
+
"learning_rate": 0.0002754909961738357,
|
| 1445 |
+
"loss": 0.6135,
|
| 1446 |
+
"step": 209
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.59,
|
| 1450 |
+
"learning_rate": 0.00027522420795439065,
|
| 1451 |
+
"loss": 0.7098,
|
| 1452 |
+
"step": 210
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.59,
|
| 1456 |
+
"learning_rate": 0.0002749561061922592,
|
| 1457 |
+
"loss": 0.5442,
|
| 1458 |
+
"step": 211
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.6,
|
| 1462 |
+
"learning_rate": 0.00027468669369970207,
|
| 1463 |
+
"loss": 0.5521,
|
| 1464 |
+
"step": 212
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 0.6,
|
| 1468 |
+
"learning_rate": 0.0002744159733027287,
|
| 1469 |
+
"loss": 0.6381,
|
| 1470 |
+
"step": 213
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.6,
|
| 1474 |
+
"learning_rate": 0.0002741439478410681,
|
| 1475 |
+
"loss": 0.6088,
|
| 1476 |
+
"step": 214
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.6,
|
| 1480 |
+
"learning_rate": 0.0002738706201681384,
|
| 1481 |
+
"loss": 0.6892,
|
| 1482 |
+
"step": 215
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.61,
|
| 1486 |
+
"learning_rate": 0.00027359599315101787,
|
| 1487 |
+
"loss": 0.5922,
|
| 1488 |
+
"step": 216
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.61,
|
| 1492 |
+
"learning_rate": 0.00027332006967041366,
|
| 1493 |
+
"loss": 0.618,
|
| 1494 |
+
"step": 217
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 0.61,
|
| 1498 |
+
"learning_rate": 0.0002730428526206327,
|
| 1499 |
+
"loss": 0.4905,
|
| 1500 |
+
"step": 218
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.61,
|
| 1504 |
+
"learning_rate": 0.0002727643449095507,
|
| 1505 |
+
"loss": 0.7306,
|
| 1506 |
+
"step": 219
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 0.62,
|
| 1510 |
+
"learning_rate": 0.0002724845494585816,
|
| 1511 |
+
"loss": 0.605,
|
| 1512 |
+
"step": 220
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 0.62,
|
| 1516 |
+
"learning_rate": 0.0002722034692026474,
|
| 1517 |
+
"loss": 0.7353,
|
| 1518 |
+
"step": 221
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 0.62,
|
| 1522 |
+
"learning_rate": 0.00027192110709014697,
|
| 1523 |
+
"loss": 0.5488,
|
| 1524 |
+
"step": 222
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.63,
|
| 1528 |
+
"learning_rate": 0.0002716374660829252,
|
| 1529 |
+
"loss": 0.5709,
|
| 1530 |
+
"step": 223
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.63,
|
| 1534 |
+
"learning_rate": 0.0002713525491562421,
|
| 1535 |
+
"loss": 0.6261,
|
| 1536 |
+
"step": 224
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.63,
|
| 1540 |
+
"learning_rate": 0.00027106635929874135,
|
| 1541 |
+
"loss": 0.6109,
|
| 1542 |
+
"step": 225
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.63,
|
| 1546 |
+
"eval_loss": 0.5185112357139587,
|
| 1547 |
+
"eval_runtime": 26.3526,
|
| 1548 |
+
"eval_samples_per_second": 7.589,
|
| 1549 |
+
"eval_steps_per_second": 1.897,
|
| 1550 |
+
"step": 225
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"dharma_eval_accuracy": 0.43693206960781283,
|
| 1554 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.5555555555555556,
|
| 1555 |
+
"dharma_eval_accuracy_ARC-Easy": 0.6296296296296297,
|
| 1556 |
+
"dharma_eval_accuracy_BoolQ": 0.7592592592592593,
|
| 1557 |
+
"dharma_eval_accuracy_MMLU": 0.4444444444444444,
|
| 1558 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 1559 |
+
"dharma_eval_accuracy_bigbench": 0.32786885245901637,
|
| 1560 |
+
"dharma_eval_accuracy_openbookqa": 0.1111111111111111,
|
| 1561 |
+
"dharma_eval_accuracy_truthful_qa": 0.3888888888888889,
|
| 1562 |
+
"dharma_eval_accuracy_winogrande": 0.4444444444444444,
|
| 1563 |
+
"dharma_loss": 2.3577894673347473,
|
| 1564 |
+
"epoch": 0.63,
|
| 1565 |
+
"step": 225
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.63,
|
| 1569 |
+
"learning_rate": 0.0002707788995124192,
|
| 1570 |
+
"loss": 0.6372,
|
| 1571 |
+
"step": 226
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 0.64,
|
| 1575 |
+
"learning_rate": 0.00027049017281259276,
|
| 1576 |
+
"loss": 0.5911,
|
| 1577 |
+
"step": 227
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.64,
|
| 1581 |
+
"learning_rate": 0.0002702001822278685,
|
| 1582 |
+
"loss": 0.411,
|
| 1583 |
+
"step": 228
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 0.64,
|
| 1587 |
+
"learning_rate": 0.00026990893080011035,
|
| 1588 |
+
"loss": 0.5805,
|
| 1589 |
+
"step": 229
|
| 1590 |
+
},
|
| 1591 |
+
{
|
| 1592 |
+
"epoch": 0.65,
|
| 1593 |
+
"learning_rate": 0.0002696164215844081,
|
| 1594 |
+
"loss": 0.5452,
|
| 1595 |
+
"step": 230
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 0.65,
|
| 1599 |
+
"learning_rate": 0.00026932265764904493,
|
| 1600 |
+
"loss": 0.5327,
|
| 1601 |
+
"step": 231
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.65,
|
| 1605 |
+
"learning_rate": 0.0002690276420754655,
|
| 1606 |
+
"loss": 0.4257,
|
| 1607 |
+
"step": 232
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 0.65,
|
| 1611 |
+
"learning_rate": 0.0002687313779582436,
|
| 1612 |
+
"loss": 0.5536,
|
| 1613 |
+
"step": 233
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 0.66,
|
| 1617 |
+
"learning_rate": 0.0002684338684050497,
|
| 1618 |
+
"loss": 0.6901,
|
| 1619 |
+
"step": 234
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.66,
|
| 1623 |
+
"learning_rate": 0.00026813511653661816,
|
| 1624 |
+
"loss": 0.5755,
|
| 1625 |
+
"step": 235
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 0.66,
|
| 1629 |
+
"learning_rate": 0.0002678351254867147,
|
| 1630 |
+
"loss": 0.5879,
|
| 1631 |
+
"step": 236
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 0.67,
|
| 1635 |
+
"learning_rate": 0.0002675338984021035,
|
| 1636 |
+
"loss": 0.2712,
|
| 1637 |
+
"step": 237
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 0.67,
|
| 1641 |
+
"learning_rate": 0.00026723143844251416,
|
| 1642 |
+
"loss": 0.6444,
|
| 1643 |
+
"step": 238
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.67,
|
| 1647 |
+
"learning_rate": 0.0002669277487806085,
|
| 1648 |
+
"loss": 0.4874,
|
| 1649 |
+
"step": 239
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.67,
|
| 1653 |
+
"learning_rate": 0.0002666228326019474,
|
| 1654 |
+
"loss": 0.5012,
|
| 1655 |
+
"step": 240
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 0.68,
|
| 1659 |
+
"learning_rate": 0.0002663166931049572,
|
| 1660 |
+
"loss": 0.5705,
|
| 1661 |
+
"step": 241
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.68,
|
| 1665 |
+
"learning_rate": 0.00026600933350089654,
|
| 1666 |
+
"loss": 0.5436,
|
| 1667 |
+
"step": 242
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 0.68,
|
| 1671 |
+
"learning_rate": 0.00026570075701382213,
|
| 1672 |
+
"loss": 0.7598,
|
| 1673 |
+
"step": 243
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 0.68,
|
| 1677 |
+
"learning_rate": 0.0002653909668805553,
|
| 1678 |
+
"loss": 0.6408,
|
| 1679 |
+
"step": 244
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 0.69,
|
| 1683 |
+
"learning_rate": 0.00026507996635064787,
|
| 1684 |
+
"loss": 0.5418,
|
| 1685 |
+
"step": 245
|
| 1686 |
+
},
|
| 1687 |
+
{
|
| 1688 |
+
"epoch": 0.69,
|
| 1689 |
+
"learning_rate": 0.0002647677586863484,
|
| 1690 |
+
"loss": 0.612,
|
| 1691 |
+
"step": 246
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.69,
|
| 1695 |
+
"learning_rate": 0.0002644543471625675,
|
| 1696 |
+
"loss": 0.6403,
|
| 1697 |
+
"step": 247
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 0.7,
|
| 1701 |
+
"learning_rate": 0.00026413973506684364,
|
| 1702 |
+
"loss": 0.5907,
|
| 1703 |
+
"step": 248
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.7,
|
| 1707 |
+
"learning_rate": 0.000263823925699309,
|
| 1708 |
+
"loss": 0.5865,
|
| 1709 |
+
"step": 249
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 0.7,
|
| 1713 |
+
"learning_rate": 0.00026350692237265427,
|
| 1714 |
+
"loss": 0.5661,
|
| 1715 |
+
"step": 250
|
| 1716 |
+
},
|
| 1717 |
+
{
|
| 1718 |
+
"epoch": 0.7,
|
| 1719 |
+
"eval_loss": 0.5193732976913452,
|
| 1720 |
+
"eval_runtime": 26.3567,
|
| 1721 |
+
"eval_samples_per_second": 7.588,
|
| 1722 |
+
"eval_steps_per_second": 1.897,
|
| 1723 |
+
"step": 250
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"dharma_eval_accuracy": 0.3625658476234253,
|
| 1727 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.5,
|
| 1728 |
+
"dharma_eval_accuracy_ARC-Easy": 0.42592592592592593,
|
| 1729 |
+
"dharma_eval_accuracy_BoolQ": 0.7777777777777778,
|
| 1730 |
+
"dharma_eval_accuracy_MMLU": 0.46296296296296297,
|
| 1731 |
+
"dharma_eval_accuracy_agieval": 0.23728813559322035,
|
| 1732 |
+
"dharma_eval_accuracy_bigbench": 0.22950819672131148,
|
| 1733 |
+
"dharma_eval_accuracy_openbookqa": 0.09259259259259259,
|
| 1734 |
+
"dharma_eval_accuracy_truthful_qa": 0.48148148148148145,
|
| 1735 |
+
"dharma_eval_accuracy_winogrande": 0.05555555555555555,
|
| 1736 |
+
"dharma_loss": 2.5813710255622864,
|
| 1737 |
+
"epoch": 0.7,
|
| 1738 |
+
"step": 250
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.7,
|
| 1742 |
+
"learning_rate": 0.00026318872841209445,
|
| 1743 |
+
"loss": 0.3614,
|
| 1744 |
+
"step": 251
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 0.71,
|
| 1748 |
+
"learning_rate": 0.0002628693471553335,
|
| 1749 |
+
"loss": 0.5204,
|
| 1750 |
+
"step": 252
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 0.71,
|
| 1754 |
+
"learning_rate": 0.00026254878195252983,
|
| 1755 |
+
"loss": 0.4298,
|
| 1756 |
+
"step": 253
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.71,
|
| 1760 |
+
"learning_rate": 0.00026222703616626055,
|
| 1761 |
+
"loss": 0.4459,
|
| 1762 |
+
"step": 254
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 0.72,
|
| 1766 |
+
"learning_rate": 0.00026190411317148686,
|
| 1767 |
+
"loss": 0.5273,
|
| 1768 |
+
"step": 255
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 0.72,
|
| 1772 |
+
"learning_rate": 0.00026158001635551815,
|
| 1773 |
+
"loss": 0.622,
|
| 1774 |
+
"step": 256
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 0.72,
|
| 1778 |
+
"learning_rate": 0.00026125474911797657,
|
| 1779 |
+
"loss": 0.6702,
|
| 1780 |
+
"step": 257
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.72,
|
| 1784 |
+
"learning_rate": 0.0002609283148707616,
|
| 1785 |
+
"loss": 0.5812,
|
| 1786 |
+
"step": 258
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 0.73,
|
| 1790 |
+
"learning_rate": 0.000260600717038014,
|
| 1791 |
+
"loss": 0.886,
|
| 1792 |
+
"step": 259
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.73,
|
| 1796 |
+
"learning_rate": 0.00026027195905608006,
|
| 1797 |
+
"loss": 0.4844,
|
| 1798 |
+
"step": 260
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 0.73,
|
| 1802 |
+
"learning_rate": 0.00025994204437347537,
|
| 1803 |
+
"loss": 0.5818,
|
| 1804 |
+
"step": 261
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.74,
|
| 1808 |
+
"learning_rate": 0.00025961097645084885,
|
| 1809 |
+
"loss": 0.629,
|
| 1810 |
+
"step": 262
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 0.74,
|
| 1814 |
+
"learning_rate": 0.0002592787587609465,
|
| 1815 |
+
"loss": 0.423,
|
| 1816 |
+
"step": 263
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 0.74,
|
| 1820 |
+
"learning_rate": 0.00025894539478857445,
|
| 1821 |
+
"loss": 0.6513,
|
| 1822 |
+
"step": 264
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.74,
|
| 1826 |
+
"learning_rate": 0.00025861088803056325,
|
| 1827 |
+
"loss": 0.4615,
|
| 1828 |
+
"step": 265
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 0.75,
|
| 1832 |
+
"learning_rate": 0.0002582752419957303,
|
| 1833 |
+
"loss": 0.4516,
|
| 1834 |
+
"step": 266
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"epoch": 0.75,
|
| 1838 |
+
"learning_rate": 0.0002579384602048438,
|
| 1839 |
+
"loss": 0.5692,
|
| 1840 |
+
"step": 267
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 0.75,
|
| 1844 |
+
"learning_rate": 0.00025760054619058535,
|
| 1845 |
+
"loss": 0.6714,
|
| 1846 |
+
"step": 268
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 0.76,
|
| 1850 |
+
"learning_rate": 0.00025726150349751305,
|
| 1851 |
+
"loss": 0.533,
|
| 1852 |
+
"step": 269
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 0.76,
|
| 1856 |
+
"learning_rate": 0.0002569213356820244,
|
| 1857 |
+
"loss": 0.6656,
|
| 1858 |
+
"step": 270
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 0.76,
|
| 1862 |
+
"learning_rate": 0.0002565800463123187,
|
| 1863 |
+
"loss": 0.6265,
|
| 1864 |
+
"step": 271
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.76,
|
| 1868 |
+
"learning_rate": 0.0002562376389683599,
|
| 1869 |
+
"loss": 0.5127,
|
| 1870 |
+
"step": 272
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 0.77,
|
| 1874 |
+
"learning_rate": 0.00025589411724183923,
|
| 1875 |
+
"loss": 0.5243,
|
| 1876 |
+
"step": 273
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 0.77,
|
| 1880 |
+
"learning_rate": 0.0002555494847361369,
|
| 1881 |
+
"loss": 0.5455,
|
| 1882 |
+
"step": 274
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 0.77,
|
| 1886 |
+
"learning_rate": 0.0002552037450662849,
|
| 1887 |
+
"loss": 0.5457,
|
| 1888 |
+
"step": 275
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.77,
|
| 1892 |
+
"eval_loss": 0.5239390730857849,
|
| 1893 |
+
"eval_runtime": 26.3626,
|
| 1894 |
+
"eval_samples_per_second": 7.587,
|
| 1895 |
+
"eval_steps_per_second": 1.897,
|
| 1896 |
+
"step": 275
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"dharma_eval_accuracy": 0.46902203058234054,
|
| 1900 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.5555555555555556,
|
| 1901 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5555555555555556,
|
| 1902 |
+
"dharma_eval_accuracy_BoolQ": 0.7222222222222222,
|
| 1903 |
+
"dharma_eval_accuracy_MMLU": 0.4444444444444444,
|
| 1904 |
+
"dharma_eval_accuracy_agieval": 0.23728813559322035,
|
| 1905 |
+
"dharma_eval_accuracy_bigbench": 0.4098360655737705,
|
| 1906 |
+
"dharma_eval_accuracy_openbookqa": 0.25925925925925924,
|
| 1907 |
+
"dharma_eval_accuracy_truthful_qa": 0.6481481481481481,
|
| 1908 |
+
"dharma_eval_accuracy_winogrande": 0.3888888888888889,
|
| 1909 |
+
"dharma_loss": 2.3993593831062316,
|
| 1910 |
+
"epoch": 0.77,
|
| 1911 |
+
"step": 275
|
| 1912 |
+
},
|
| 1913 |
+
{
|
| 1914 |
+
"epoch": 0.77,
|
| 1915 |
+
"learning_rate": 0.0002548569018589286,
|
| 1916 |
+
"loss": 0.5044,
|
| 1917 |
+
"step": 276
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 0.78,
|
| 1921 |
+
"learning_rate": 0.0002545089587522893,
|
| 1922 |
+
"loss": 0.672,
|
| 1923 |
+
"step": 277
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 0.78,
|
| 1927 |
+
"learning_rate": 0.0002541599193961254,
|
| 1928 |
+
"loss": 0.5391,
|
| 1929 |
+
"step": 278
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.78,
|
| 1933 |
+
"learning_rate": 0.0002538097874516947,
|
| 1934 |
+
"loss": 0.5886,
|
| 1935 |
+
"step": 279
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 0.79,
|
| 1939 |
+
"learning_rate": 0.00025345856659171563,
|
| 1940 |
+
"loss": 0.4727,
|
| 1941 |
+
"step": 280
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.79,
|
| 1945 |
+
"learning_rate": 0.0002531062605003287,
|
| 1946 |
+
"loss": 0.4911,
|
| 1947 |
+
"step": 281
|
| 1948 |
+
},
|
| 1949 |
+
{
|
| 1950 |
+
"epoch": 0.79,
|
| 1951 |
+
"learning_rate": 0.00025275287287305814,
|
| 1952 |
+
"loss": 0.4916,
|
| 1953 |
+
"step": 282
|
| 1954 |
+
},
|
| 1955 |
+
{
|
| 1956 |
+
"epoch": 0.79,
|
| 1957 |
+
"learning_rate": 0.00025239840741677304,
|
| 1958 |
+
"loss": 0.4675,
|
| 1959 |
+
"step": 283
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 0.8,
|
| 1963 |
+
"learning_rate": 0.00025204286784964817,
|
| 1964 |
+
"loss": 0.8109,
|
| 1965 |
+
"step": 284
|
| 1966 |
+
},
|
| 1967 |
+
{
|
| 1968 |
+
"epoch": 0.8,
|
| 1969 |
+
"learning_rate": 0.00025168625790112547,
|
| 1970 |
+
"loss": 0.4635,
|
| 1971 |
+
"step": 285
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 0.8,
|
| 1975 |
+
"learning_rate": 0.00025132858131187444,
|
| 1976 |
+
"loss": 0.549,
|
| 1977 |
+
"step": 286
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.81,
|
| 1981 |
+
"learning_rate": 0.00025096984183375334,
|
| 1982 |
+
"loss": 0.5386,
|
| 1983 |
+
"step": 287
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.81,
|
| 1987 |
+
"learning_rate": 0.0002506100432297695,
|
| 1988 |
+
"loss": 0.5898,
|
| 1989 |
+
"step": 288
|
| 1990 |
+
},
|
| 1991 |
+
{
|
| 1992 |
+
"epoch": 0.81,
|
| 1993 |
+
"learning_rate": 0.00025024918927404004,
|
| 1994 |
+
"loss": 0.6002,
|
| 1995 |
+
"step": 289
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.81,
|
| 1999 |
+
"learning_rate": 0.00024988728375175214,
|
| 2000 |
+
"loss": 0.5995,
|
| 2001 |
+
"step": 290
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.82,
|
| 2005 |
+
"learning_rate": 0.00024952433045912354,
|
| 2006 |
+
"loss": 0.4781,
|
| 2007 |
+
"step": 291
|
| 2008 |
+
},
|
| 2009 |
+
{
|
| 2010 |
+
"epoch": 0.82,
|
| 2011 |
+
"learning_rate": 0.00024916033320336264,
|
| 2012 |
+
"loss": 0.6746,
|
| 2013 |
+
"step": 292
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 0.82,
|
| 2017 |
+
"learning_rate": 0.00024879529580262817,
|
| 2018 |
+
"loss": 0.5565,
|
| 2019 |
+
"step": 293
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.83,
|
| 2023 |
+
"learning_rate": 0.00024842922208598994,
|
| 2024 |
+
"loss": 0.6523,
|
| 2025 |
+
"step": 294
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.83,
|
| 2029 |
+
"learning_rate": 0.0002480621158933879,
|
| 2030 |
+
"loss": 0.5657,
|
| 2031 |
+
"step": 295
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 0.83,
|
| 2035 |
+
"learning_rate": 0.00024769398107559224,
|
| 2036 |
+
"loss": 0.5042,
|
| 2037 |
+
"step": 296
|
| 2038 |
+
},
|
| 2039 |
+
{
|
| 2040 |
+
"epoch": 0.83,
|
| 2041 |
+
"learning_rate": 0.0002473248214941632,
|
| 2042 |
+
"loss": 0.5771,
|
| 2043 |
+
"step": 297
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 0.84,
|
| 2047 |
+
"learning_rate": 0.00024695464102141,
|
| 2048 |
+
"loss": 0.4739,
|
| 2049 |
+
"step": 298
|
| 2050 |
+
},
|
| 2051 |
+
{
|
| 2052 |
+
"epoch": 0.84,
|
| 2053 |
+
"learning_rate": 0.0002465834435403506,
|
| 2054 |
+
"loss": 0.4495,
|
| 2055 |
+
"step": 299
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 0.84,
|
| 2059 |
+
"learning_rate": 0.00024621123294467096,
|
| 2060 |
+
"loss": 0.5886,
|
| 2061 |
+
"step": 300
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 0.84,
|
| 2065 |
+
"eval_loss": 0.515432596206665,
|
| 2066 |
+
"eval_runtime": 26.3157,
|
| 2067 |
+
"eval_samples_per_second": 7.6,
|
| 2068 |
+
"eval_steps_per_second": 1.9,
|
| 2069 |
+
"step": 300
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"dharma_eval_accuracy": 0.478252989799407,
|
| 2073 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.5555555555555556,
|
| 2074 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
|
| 2075 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 2076 |
+
"dharma_eval_accuracy_MMLU": 0.4444444444444444,
|
| 2077 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 2078 |
+
"dharma_eval_accuracy_bigbench": 0.45901639344262296,
|
| 2079 |
+
"dharma_eval_accuracy_openbookqa": 0.2777777777777778,
|
| 2080 |
+
"dharma_eval_accuracy_truthful_qa": 0.5740740740740741,
|
| 2081 |
+
"dharma_eval_accuracy_winogrande": 0.5,
|
| 2082 |
+
"dharma_loss": 2.5394882402420045,
|
| 2083 |
+
"epoch": 0.84,
|
| 2084 |
+
"step": 300
|
| 2085 |
+
},
|
| 2086 |
+
{
|
| 2087 |
+
"epoch": 0.84,
|
| 2088 |
+
"learning_rate": 0.00024583801313868416,
|
| 2089 |
+
"loss": 0.4195,
|
| 2090 |
+
"step": 301
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 0.85,
|
| 2094 |
+
"learning_rate": 0.0002454637880372892,
|
| 2095 |
+
"loss": 0.4727,
|
| 2096 |
+
"step": 302
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.85,
|
| 2100 |
+
"learning_rate": 0.0002450885615659305,
|
| 2101 |
+
"loss": 0.4463,
|
| 2102 |
+
"step": 303
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.85,
|
| 2106 |
+
"learning_rate": 0.00024471233766055605,
|
| 2107 |
+
"loss": 0.6394,
|
| 2108 |
+
"step": 304
|
| 2109 |
+
},
|
| 2110 |
+
{
|
| 2111 |
+
"epoch": 0.86,
|
| 2112 |
+
"learning_rate": 0.00024433512026757664,
|
| 2113 |
+
"loss": 0.6439,
|
| 2114 |
+
"step": 305
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.86,
|
| 2118 |
+
"learning_rate": 0.00024395691334382427,
|
| 2119 |
+
"loss": 0.5627,
|
| 2120 |
+
"step": 306
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 0.86,
|
| 2124 |
+
"learning_rate": 0.00024357772085651055,
|
| 2125 |
+
"loss": 0.4879,
|
| 2126 |
+
"step": 307
|
| 2127 |
+
},
|
| 2128 |
+
{
|
| 2129 |
+
"epoch": 0.86,
|
| 2130 |
+
"learning_rate": 0.00024319754678318528,
|
| 2131 |
+
"loss": 0.4786,
|
| 2132 |
+
"step": 308
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 0.87,
|
| 2136 |
+
"learning_rate": 0.00024281639511169456,
|
| 2137 |
+
"loss": 0.393,
|
| 2138 |
+
"step": 309
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.87,
|
| 2142 |
+
"learning_rate": 0.0002424342698401391,
|
| 2143 |
+
"loss": 0.7039,
|
| 2144 |
+
"step": 310
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.87,
|
| 2148 |
+
"learning_rate": 0.00024205117497683209,
|
| 2149 |
+
"loss": 0.5914,
|
| 2150 |
+
"step": 311
|
| 2151 |
+
},
|
| 2152 |
+
{
|
| 2153 |
+
"epoch": 0.88,
|
| 2154 |
+
"learning_rate": 0.0002416671145402575,
|
| 2155 |
+
"loss": 0.595,
|
| 2156 |
+
"step": 312
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.88,
|
| 2160 |
+
"learning_rate": 0.0002412820925590275,
|
| 2161 |
+
"loss": 0.758,
|
| 2162 |
+
"step": 313
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 0.88,
|
| 2166 |
+
"learning_rate": 0.00024089611307184048,
|
| 2167 |
+
"loss": 0.4295,
|
| 2168 |
+
"step": 314
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 0.88,
|
| 2172 |
+
"learning_rate": 0.0002405091801274387,
|
| 2173 |
+
"loss": 0.4564,
|
| 2174 |
+
"step": 315
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 0.89,
|
| 2178 |
+
"learning_rate": 0.00024012129778456552,
|
| 2179 |
+
"loss": 0.5362,
|
| 2180 |
+
"step": 316
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.89,
|
| 2184 |
+
"learning_rate": 0.00023973247011192325,
|
| 2185 |
+
"loss": 0.6515,
|
| 2186 |
+
"step": 317
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.89,
|
| 2190 |
+
"learning_rate": 0.0002393427011881302,
|
| 2191 |
+
"loss": 0.7295,
|
| 2192 |
+
"step": 318
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 0.9,
|
| 2196 |
+
"learning_rate": 0.0002389519951016779,
|
| 2197 |
+
"loss": 0.751,
|
| 2198 |
+
"step": 319
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 0.9,
|
| 2202 |
+
"learning_rate": 0.00023856035595088839,
|
| 2203 |
+
"loss": 0.895,
|
| 2204 |
+
"step": 320
|
| 2205 |
+
},
|
| 2206 |
+
{
|
| 2207 |
+
"epoch": 0.9,
|
| 2208 |
+
"learning_rate": 0.00023816778784387094,
|
| 2209 |
+
"loss": 0.6754,
|
| 2210 |
+
"step": 321
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 0.9,
|
| 2214 |
+
"learning_rate": 0.00023777429489847934,
|
| 2215 |
+
"loss": 0.6156,
|
| 2216 |
+
"step": 322
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 0.91,
|
| 2220 |
+
"learning_rate": 0.0002373798812422683,
|
| 2221 |
+
"loss": 0.4012,
|
| 2222 |
+
"step": 323
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.91,
|
| 2226 |
+
"learning_rate": 0.0002369845510124505,
|
| 2227 |
+
"loss": 0.3156,
|
| 2228 |
+
"step": 324
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.91,
|
| 2232 |
+
"learning_rate": 0.0002365883083558529,
|
| 2233 |
+
"loss": 0.5232,
|
| 2234 |
+
"step": 325
|
| 2235 |
+
},
|
| 2236 |
+
{
|
| 2237 |
+
"epoch": 0.91,
|
| 2238 |
+
"eval_loss": 0.5132270455360413,
|
| 2239 |
+
"eval_runtime": 26.337,
|
| 2240 |
+
"eval_samples_per_second": 7.594,
|
| 2241 |
+
"eval_steps_per_second": 1.898,
|
| 2242 |
+
"step": 325
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"dharma_eval_accuracy": 0.38590223393100737,
|
| 2246 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.46296296296296297,
|
| 2247 |
+
"dharma_eval_accuracy_ARC-Easy": 0.4444444444444444,
|
| 2248 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 2249 |
+
"dharma_eval_accuracy_MMLU": 0.42592592592592593,
|
| 2250 |
+
"dharma_eval_accuracy_agieval": 0.2542372881355932,
|
| 2251 |
+
"dharma_eval_accuracy_bigbench": 0.3114754098360656,
|
| 2252 |
+
"dharma_eval_accuracy_openbookqa": 0.2777777777777778,
|
| 2253 |
+
"dharma_eval_accuracy_truthful_qa": 0.5555555555555556,
|
| 2254 |
+
"dharma_eval_accuracy_winogrande": 0.037037037037037035,
|
| 2255 |
+
"dharma_loss": 2.2708842940330505,
|
| 2256 |
+
"epoch": 0.91,
|
| 2257 |
+
"step": 325
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.92,
|
| 2261 |
+
"learning_rate": 0.00023619115742887356,
|
| 2262 |
+
"loss": 0.3488,
|
| 2263 |
+
"step": 326
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.92,
|
| 2267 |
+
"learning_rate": 0.00023579310239743773,
|
| 2268 |
+
"loss": 0.5628,
|
| 2269 |
+
"step": 327
|
| 2270 |
+
},
|
| 2271 |
+
{
|
| 2272 |
+
"epoch": 0.92,
|
| 2273 |
+
"learning_rate": 0.00023539414743695437,
|
| 2274 |
+
"loss": 0.4697,
|
| 2275 |
+
"step": 328
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 0.92,
|
| 2279 |
+
"learning_rate": 0.00023499429673227222,
|
| 2280 |
+
"loss": 0.6077,
|
| 2281 |
+
"step": 329
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 0.93,
|
| 2285 |
+
"learning_rate": 0.00023459355447763596,
|
| 2286 |
+
"loss": 0.5314,
|
| 2287 |
+
"step": 330
|
| 2288 |
+
},
|
| 2289 |
+
{
|
| 2290 |
+
"epoch": 0.93,
|
| 2291 |
+
"learning_rate": 0.00023419192487664217,
|
| 2292 |
+
"loss": 0.5117,
|
| 2293 |
+
"step": 331
|
| 2294 |
+
},
|
| 2295 |
+
{
|
| 2296 |
+
"epoch": 0.93,
|
| 2297 |
+
"learning_rate": 0.0002337894121421954,
|
| 2298 |
+
"loss": 0.4915,
|
| 2299 |
+
"step": 332
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.93,
|
| 2303 |
+
"learning_rate": 0.0002333860204964637,
|
| 2304 |
+
"loss": 0.5615,
|
| 2305 |
+
"step": 333
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.94,
|
| 2309 |
+
"learning_rate": 0.00023298175417083459,
|
| 2310 |
+
"loss": 0.598,
|
| 2311 |
+
"step": 334
|
| 2312 |
+
},
|
| 2313 |
+
{
|
| 2314 |
+
"epoch": 0.94,
|
| 2315 |
+
"learning_rate": 0.0002325766174058705,
|
| 2316 |
+
"loss": 0.5588,
|
| 2317 |
+
"step": 335
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 0.94,
|
| 2321 |
+
"learning_rate": 0.0002321706144512644,
|
| 2322 |
+
"loss": 0.4396,
|
| 2323 |
+
"step": 336
|
| 2324 |
+
},
|
| 2325 |
+
{
|
| 2326 |
+
"epoch": 0.95,
|
| 2327 |
+
"learning_rate": 0.0002317637495657952,
|
| 2328 |
+
"loss": 0.4851,
|
| 2329 |
+
"step": 337
|
| 2330 |
+
},
|
| 2331 |
+
{
|
| 2332 |
+
"epoch": 0.95,
|
| 2333 |
+
"learning_rate": 0.000231356027017283,
|
| 2334 |
+
"loss": 0.5265,
|
| 2335 |
+
"step": 338
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 0.95,
|
| 2339 |
+
"learning_rate": 0.00023094745108254435,
|
| 2340 |
+
"loss": 0.5465,
|
| 2341 |
+
"step": 339
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.95,
|
| 2345 |
+
"learning_rate": 0.00023053802604734757,
|
| 2346 |
+
"loss": 0.5982,
|
| 2347 |
+
"step": 340
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.96,
|
| 2351 |
+
"learning_rate": 0.00023012775620636746,
|
| 2352 |
+
"loss": 0.5515,
|
| 2353 |
+
"step": 341
|
| 2354 |
+
},
|
| 2355 |
+
{
|
| 2356 |
+
"epoch": 0.96,
|
| 2357 |
+
"learning_rate": 0.00022971664586314054,
|
| 2358 |
+
"loss": 0.6549,
|
| 2359 |
+
"step": 342
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 0.96,
|
| 2363 |
+
"learning_rate": 0.00022930469933001975,
|
| 2364 |
+
"loss": 0.5104,
|
| 2365 |
+
"step": 343
|
| 2366 |
+
},
|
| 2367 |
+
{
|
| 2368 |
+
"epoch": 0.97,
|
| 2369 |
+
"learning_rate": 0.00022889192092812938,
|
| 2370 |
+
"loss": 0.4687,
|
| 2371 |
+
"step": 344
|
| 2372 |
+
},
|
| 2373 |
+
{
|
| 2374 |
+
"epoch": 0.97,
|
| 2375 |
+
"learning_rate": 0.0002284783149873195,
|
| 2376 |
+
"loss": 0.6088,
|
| 2377 |
+
"step": 345
|
| 2378 |
+
},
|
| 2379 |
+
{
|
| 2380 |
+
"epoch": 0.97,
|
| 2381 |
+
"learning_rate": 0.00022806388584612066,
|
| 2382 |
+
"loss": 0.6173,
|
| 2383 |
+
"step": 346
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.97,
|
| 2387 |
+
"learning_rate": 0.00022764863785169854,
|
| 2388 |
+
"loss": 0.5955,
|
| 2389 |
+
"step": 347
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.98,
|
| 2393 |
+
"learning_rate": 0.000227232575359808,
|
| 2394 |
+
"loss": 0.5085,
|
| 2395 |
+
"step": 348
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 0.98,
|
| 2399 |
+
"learning_rate": 0.0002268157027347478,
|
| 2400 |
+
"loss": 0.3806,
|
| 2401 |
+
"step": 349
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 0.98,
|
| 2405 |
+
"learning_rate": 0.00022639802434931444,
|
| 2406 |
+
"loss": 0.6013,
|
| 2407 |
+
"step": 350
|
| 2408 |
+
},
|
| 2409 |
+
{
|
| 2410 |
+
"epoch": 0.98,
|
| 2411 |
+
"eval_loss": 0.511376678943634,
|
| 2412 |
+
"eval_runtime": 26.3778,
|
| 2413 |
+
"eval_samples_per_second": 7.582,
|
| 2414 |
+
"eval_steps_per_second": 1.896,
|
| 2415 |
+
"step": 350
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"dharma_eval_accuracy": 0.45162236423697943,
|
| 2419 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.4444444444444444,
|
| 2420 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
|
| 2421 |
+
"dharma_eval_accuracy_BoolQ": 0.7592592592592593,
|
| 2422 |
+
"dharma_eval_accuracy_MMLU": 0.37037037037037035,
|
| 2423 |
+
"dharma_eval_accuracy_agieval": 0.22033898305084745,
|
| 2424 |
+
"dharma_eval_accuracy_bigbench": 0.3442622950819672,
|
| 2425 |
+
"dharma_eval_accuracy_openbookqa": 0.37037037037037035,
|
| 2426 |
+
"dharma_eval_accuracy_truthful_qa": 0.5370370370370371,
|
| 2427 |
+
"dharma_eval_accuracy_winogrande": 0.48148148148148145,
|
| 2428 |
+
"dharma_loss": 2.6632097277641296,
|
| 2429 |
+
"epoch": 0.98,
|
| 2430 |
+
"step": 350
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 0.99,
|
| 2434 |
+
"learning_rate": 0.0002259795445847566,
|
| 2435 |
+
"loss": 0.5341,
|
| 2436 |
+
"step": 351
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 0.99,
|
| 2440 |
+
"learning_rate": 0.00022556026783072895,
|
| 2441 |
+
"loss": 0.3715,
|
| 2442 |
+
"step": 352
|
| 2443 |
+
},
|
| 2444 |
+
{
|
| 2445 |
+
"epoch": 0.99,
|
| 2446 |
+
"learning_rate": 0.00022514019848524628,
|
| 2447 |
+
"loss": 0.5229,
|
| 2448 |
+
"step": 353
|
| 2449 |
+
},
|
| 2450 |
+
{
|
| 2451 |
+
"epoch": 0.99,
|
| 2452 |
+
"learning_rate": 0.00022471934095463723,
|
| 2453 |
+
"loss": 0.4048,
|
| 2454 |
+
"step": 354
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 1.0,
|
| 2458 |
+
"learning_rate": 0.00022429769965349815,
|
| 2459 |
+
"loss": 0.48,
|
| 2460 |
+
"step": 355
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 1.0,
|
| 2464 |
+
"learning_rate": 0.00022387527900464673,
|
| 2465 |
+
"loss": 0.4543,
|
| 2466 |
+
"step": 356
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 1.0,
|
| 2470 |
+
"learning_rate": 0.00022345208343907576,
|
| 2471 |
+
"loss": 0.579,
|
| 2472 |
+
"step": 357
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 1.0,
|
| 2476 |
+
"learning_rate": 0.00022302811739590642,
|
| 2477 |
+
"loss": 0.488,
|
| 2478 |
+
"step": 358
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 1.01,
|
| 2482 |
+
"learning_rate": 0.00022260338532234194,
|
| 2483 |
+
"loss": 0.4045,
|
| 2484 |
+
"step": 359
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 1.01,
|
| 2488 |
+
"learning_rate": 0.00022217789167362073,
|
| 2489 |
+
"loss": 0.5919,
|
| 2490 |
+
"step": 360
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 1.01,
|
| 2494 |
+
"learning_rate": 0.00022175164091296987,
|
| 2495 |
+
"loss": 0.6644,
|
| 2496 |
+
"step": 361
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 1.02,
|
| 2500 |
+
"learning_rate": 0.00022132463751155812,
|
| 2501 |
+
"loss": 0.6124,
|
| 2502 |
+
"step": 362
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 1.02,
|
| 2506 |
+
"learning_rate": 0.00022089688594844915,
|
| 2507 |
+
"loss": 0.5964,
|
| 2508 |
+
"step": 363
|
| 2509 |
+
},
|
| 2510 |
+
{
|
| 2511 |
+
"epoch": 1.02,
|
| 2512 |
+
"learning_rate": 0.00022046839071055435,
|
| 2513 |
+
"loss": 0.3682,
|
| 2514 |
+
"step": 364
|
| 2515 |
+
},
|
| 2516 |
+
{
|
| 2517 |
+
"epoch": 1.02,
|
| 2518 |
+
"learning_rate": 0.00022003915629258605,
|
| 2519 |
+
"loss": 0.4854,
|
| 2520 |
+
"step": 365
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 1.03,
|
| 2524 |
+
"learning_rate": 0.00021960918719701026,
|
| 2525 |
+
"loss": 0.5045,
|
| 2526 |
+
"step": 366
|
| 2527 |
+
},
|
| 2528 |
+
{
|
| 2529 |
+
"epoch": 1.03,
|
| 2530 |
+
"learning_rate": 0.00021917848793399924,
|
| 2531 |
+
"loss": 0.4382,
|
| 2532 |
+
"step": 367
|
| 2533 |
+
},
|
| 2534 |
+
{
|
| 2535 |
+
"epoch": 1.03,
|
| 2536 |
+
"learning_rate": 0.00021874706302138448,
|
| 2537 |
+
"loss": 0.5718,
|
| 2538 |
+
"step": 368
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 1.04,
|
| 2542 |
+
"learning_rate": 0.00021831491698460923,
|
| 2543 |
+
"loss": 0.4179,
|
| 2544 |
+
"step": 369
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 1.04,
|
| 2548 |
+
"learning_rate": 0.00021788205435668083,
|
| 2549 |
+
"loss": 0.4178,
|
| 2550 |
+
"step": 370
|
| 2551 |
+
},
|
| 2552 |
+
{
|
| 2553 |
+
"epoch": 1.04,
|
| 2554 |
+
"learning_rate": 0.0002174484796781235,
|
| 2555 |
+
"loss": 0.4802,
|
| 2556 |
+
"step": 371
|
| 2557 |
+
},
|
| 2558 |
+
{
|
| 2559 |
+
"epoch": 1.04,
|
| 2560 |
+
"learning_rate": 0.00021701419749693034,
|
| 2561 |
+
"loss": 0.4051,
|
| 2562 |
+
"step": 372
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 1.05,
|
| 2566 |
+
"learning_rate": 0.00021657921236851603,
|
| 2567 |
+
"loss": 0.3847,
|
| 2568 |
+
"step": 373
|
| 2569 |
+
},
|
| 2570 |
+
{
|
| 2571 |
+
"epoch": 1.05,
|
| 2572 |
+
"learning_rate": 0.00021614352885566872,
|
| 2573 |
+
"loss": 0.5093,
|
| 2574 |
+
"step": 374
|
| 2575 |
+
},
|
| 2576 |
+
{
|
| 2577 |
+
"epoch": 1.05,
|
| 2578 |
+
"learning_rate": 0.00021570715152850232,
|
| 2579 |
+
"loss": 0.4866,
|
| 2580 |
+
"step": 375
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 1.05,
|
| 2584 |
+
"eval_loss": 0.5142313838005066,
|
| 2585 |
+
"eval_runtime": 26.3986,
|
| 2586 |
+
"eval_samples_per_second": 7.576,
|
| 2587 |
+
"eval_steps_per_second": 1.894,
|
| 2588 |
+
"step": 375
|
| 2589 |
+
},
|
| 2590 |
+
{
|
| 2591 |
+
"dharma_eval_accuracy": 0.44273958129658786,
|
| 2592 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.25925925925925924,
|
| 2593 |
+
"dharma_eval_accuracy_ARC-Easy": 0.35185185185185186,
|
| 2594 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 2595 |
+
"dharma_eval_accuracy_MMLU": 0.2222222222222222,
|
| 2596 |
+
"dharma_eval_accuracy_agieval": 0.2542372881355932,
|
| 2597 |
+
"dharma_eval_accuracy_bigbench": 0.5081967213114754,
|
| 2598 |
+
"dharma_eval_accuracy_openbookqa": 0.37037037037037035,
|
| 2599 |
+
"dharma_eval_accuracy_truthful_qa": 0.7777777777777778,
|
| 2600 |
+
"dharma_eval_accuracy_winogrande": 0.5370370370370371,
|
| 2601 |
+
"dharma_loss": 3.0705994338989258,
|
| 2602 |
+
"epoch": 1.05,
|
| 2603 |
+
"step": 375
|
| 2604 |
+
},
|
| 2605 |
+
{
|
| 2606 |
+
"epoch": 1.06,
|
| 2607 |
+
"learning_rate": 0.00021527008496440846,
|
| 2608 |
+
"loss": 0.6129,
|
| 2609 |
+
"step": 376
|
| 2610 |
+
},
|
| 2611 |
+
{
|
| 2612 |
+
"epoch": 1.06,
|
| 2613 |
+
"learning_rate": 0.00021483233374800862,
|
| 2614 |
+
"loss": 0.5435,
|
| 2615 |
+
"step": 377
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 1.06,
|
| 2619 |
+
"learning_rate": 0.00021439390247110588,
|
| 2620 |
+
"loss": 0.3336,
|
| 2621 |
+
"step": 378
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 1.06,
|
| 2625 |
+
"learning_rate": 0.00021395479573263688,
|
| 2626 |
+
"loss": 0.6667,
|
| 2627 |
+
"step": 379
|
| 2628 |
+
},
|
| 2629 |
+
{
|
| 2630 |
+
"epoch": 1.07,
|
| 2631 |
+
"learning_rate": 0.00021351501813862356,
|
| 2632 |
+
"loss": 0.479,
|
| 2633 |
+
"step": 380
|
| 2634 |
+
},
|
| 2635 |
+
{
|
| 2636 |
+
"epoch": 1.07,
|
| 2637 |
+
"learning_rate": 0.00021307457430212468,
|
| 2638 |
+
"loss": 0.4249,
|
| 2639 |
+
"step": 381
|
| 2640 |
+
},
|
| 2641 |
+
{
|
| 2642 |
+
"epoch": 1.07,
|
| 2643 |
+
"learning_rate": 0.00021263346884318777,
|
| 2644 |
+
"loss": 0.5989,
|
| 2645 |
+
"step": 382
|
| 2646 |
+
},
|
| 2647 |
+
{
|
| 2648 |
+
"epoch": 1.08,
|
| 2649 |
+
"learning_rate": 0.00021219170638880037,
|
| 2650 |
+
"loss": 0.3444,
|
| 2651 |
+
"step": 383
|
| 2652 |
+
},
|
| 2653 |
+
{
|
| 2654 |
+
"epoch": 1.08,
|
| 2655 |
+
"learning_rate": 0.00021174929157284156,
|
| 2656 |
+
"loss": 0.3373,
|
| 2657 |
+
"step": 384
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 1.08,
|
| 2661 |
+
"learning_rate": 0.00021130622903603344,
|
| 2662 |
+
"loss": 0.4317,
|
| 2663 |
+
"step": 385
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 1.08,
|
| 2667 |
+
"learning_rate": 0.0002108625234258923,
|
| 2668 |
+
"loss": 0.4541,
|
| 2669 |
+
"step": 386
|
| 2670 |
+
},
|
| 2671 |
+
{
|
| 2672 |
+
"epoch": 1.09,
|
| 2673 |
+
"learning_rate": 0.00021041817939668005,
|
| 2674 |
+
"loss": 0.5544,
|
| 2675 |
+
"step": 387
|
| 2676 |
+
},
|
| 2677 |
+
{
|
| 2678 |
+
"epoch": 1.09,
|
| 2679 |
+
"learning_rate": 0.00020997320160935534,
|
| 2680 |
+
"loss": 0.3578,
|
| 2681 |
+
"step": 388
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 1.09,
|
| 2685 |
+
"learning_rate": 0.00020952759473152457,
|
| 2686 |
+
"loss": 0.3263,
|
| 2687 |
+
"step": 389
|
| 2688 |
+
},
|
| 2689 |
+
{
|
| 2690 |
+
"epoch": 1.09,
|
| 2691 |
+
"learning_rate": 0.00020908136343739307,
|
| 2692 |
+
"loss": 0.4524,
|
| 2693 |
+
"step": 390
|
| 2694 |
+
},
|
| 2695 |
+
{
|
| 2696 |
+
"epoch": 1.1,
|
| 2697 |
+
"learning_rate": 0.00020863451240771592,
|
| 2698 |
+
"loss": 0.4767,
|
| 2699 |
+
"step": 391
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 1.1,
|
| 2703 |
+
"learning_rate": 0.00020818704632974896,
|
| 2704 |
+
"loss": 0.4093,
|
| 2705 |
+
"step": 392
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 1.1,
|
| 2709 |
+
"learning_rate": 0.00020773896989719965,
|
| 2710 |
+
"loss": 0.4865,
|
| 2711 |
+
"step": 393
|
| 2712 |
+
},
|
| 2713 |
+
{
|
| 2714 |
+
"epoch": 1.11,
|
| 2715 |
+
"learning_rate": 0.00020729028781017777,
|
| 2716 |
+
"loss": 0.4853,
|
| 2717 |
+
"step": 394
|
| 2718 |
+
},
|
| 2719 |
+
{
|
| 2720 |
+
"epoch": 1.11,
|
| 2721 |
+
"learning_rate": 0.000206841004775146,
|
| 2722 |
+
"loss": 0.3023,
|
| 2723 |
+
"step": 395
|
| 2724 |
+
},
|
| 2725 |
+
{
|
| 2726 |
+
"epoch": 1.11,
|
| 2727 |
+
"learning_rate": 0.00020639112550487084,
|
| 2728 |
+
"loss": 0.5841,
|
| 2729 |
+
"step": 396
|
| 2730 |
+
},
|
| 2731 |
+
{
|
| 2732 |
+
"epoch": 1.11,
|
| 2733 |
+
"learning_rate": 0.000205940654718373,
|
| 2734 |
+
"loss": 0.5222,
|
| 2735 |
+
"step": 397
|
| 2736 |
+
},
|
| 2737 |
+
{
|
| 2738 |
+
"epoch": 1.12,
|
| 2739 |
+
"learning_rate": 0.0002054895971408778,
|
| 2740 |
+
"loss": 0.4549,
|
| 2741 |
+
"step": 398
|
| 2742 |
+
},
|
| 2743 |
+
{
|
| 2744 |
+
"epoch": 1.12,
|
| 2745 |
+
"learning_rate": 0.0002050379575037658,
|
| 2746 |
+
"loss": 0.4911,
|
| 2747 |
+
"step": 399
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 1.12,
|
| 2751 |
+
"learning_rate": 0.00020458574054452313,
|
| 2752 |
+
"loss": 0.4955,
|
| 2753 |
+
"step": 400
|
| 2754 |
+
},
|
| 2755 |
+
{
|
| 2756 |
+
"epoch": 1.12,
|
| 2757 |
+
"eval_loss": 0.5140312910079956,
|
| 2758 |
+
"eval_runtime": 26.3702,
|
| 2759 |
+
"eval_samples_per_second": 7.584,
|
| 2760 |
+
"eval_steps_per_second": 1.896,
|
| 2761 |
+
"step": 400
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"dharma_eval_accuracy": 0.42833628911551785,
|
| 2765 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.2777777777777778,
|
| 2766 |
+
"dharma_eval_accuracy_ARC-Easy": 0.4074074074074074,
|
| 2767 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 2768 |
+
"dharma_eval_accuracy_MMLU": 0.24074074074074073,
|
| 2769 |
+
"dharma_eval_accuracy_agieval": 0.2542372881355932,
|
| 2770 |
+
"dharma_eval_accuracy_bigbench": 0.5081967213114754,
|
| 2771 |
+
"dharma_eval_accuracy_openbookqa": 0.25925925925925924,
|
| 2772 |
+
"dharma_eval_accuracy_truthful_qa": 0.6666666666666666,
|
| 2773 |
+
"dharma_eval_accuracy_winogrande": 0.5370370370370371,
|
| 2774 |
+
"dharma_loss": 2.6535617146492005,
|
| 2775 |
+
"epoch": 1.12,
|
| 2776 |
+
"step": 400
|
| 2777 |
+
},
|
| 2778 |
+
{
|
| 2779 |
+
"epoch": 1.13,
|
| 2780 |
+
"learning_rate": 0.00020413295100669165,
|
| 2781 |
+
"loss": 0.3082,
|
| 2782 |
+
"step": 401
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 1.13,
|
| 2786 |
+
"learning_rate": 0.00020367959363981936,
|
| 2787 |
+
"loss": 0.5595,
|
| 2788 |
+
"step": 402
|
| 2789 |
+
},
|
| 2790 |
+
{
|
| 2791 |
+
"epoch": 1.13,
|
| 2792 |
+
"learning_rate": 0.0002032256731994106,
|
| 2793 |
+
"loss": 0.618,
|
| 2794 |
+
"step": 403
|
| 2795 |
+
},
|
| 2796 |
+
{
|
| 2797 |
+
"epoch": 1.13,
|
| 2798 |
+
"learning_rate": 0.00020277119444687584,
|
| 2799 |
+
"loss": 0.4732,
|
| 2800 |
+
"step": 404
|
| 2801 |
+
},
|
| 2802 |
+
{
|
| 2803 |
+
"epoch": 1.14,
|
| 2804 |
+
"learning_rate": 0.0002023161621494823,
|
| 2805 |
+
"loss": 0.4542,
|
| 2806 |
+
"step": 405
|
| 2807 |
+
},
|
| 2808 |
+
{
|
| 2809 |
+
"epoch": 1.14,
|
| 2810 |
+
"learning_rate": 0.00020186058108030342,
|
| 2811 |
+
"loss": 0.4732,
|
| 2812 |
+
"step": 406
|
| 2813 |
+
},
|
| 2814 |
+
{
|
| 2815 |
+
"epoch": 1.14,
|
| 2816 |
+
"learning_rate": 0.00020140445601816897,
|
| 2817 |
+
"loss": 0.4085,
|
| 2818 |
+
"step": 407
|
| 2819 |
+
},
|
| 2820 |
+
{
|
| 2821 |
+
"epoch": 1.15,
|
| 2822 |
+
"learning_rate": 0.00020094779174761505,
|
| 2823 |
+
"loss": 0.4966,
|
| 2824 |
+
"step": 408
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 1.15,
|
| 2828 |
+
"learning_rate": 0.00020049059305883379,
|
| 2829 |
+
"loss": 0.4264,
|
| 2830 |
+
"step": 409
|
| 2831 |
+
},
|
| 2832 |
+
{
|
| 2833 |
+
"epoch": 1.15,
|
| 2834 |
+
"learning_rate": 0.0002000328647476231,
|
| 2835 |
+
"loss": 0.6945,
|
| 2836 |
+
"step": 410
|
| 2837 |
+
},
|
| 2838 |
+
{
|
| 2839 |
+
"epoch": 1.15,
|
| 2840 |
+
"learning_rate": 0.00019957461161533628,
|
| 2841 |
+
"loss": 0.3748,
|
| 2842 |
+
"step": 411
|
| 2843 |
+
},
|
| 2844 |
+
{
|
| 2845 |
+
"epoch": 1.16,
|
| 2846 |
+
"learning_rate": 0.00019911583846883197,
|
| 2847 |
+
"loss": 0.4211,
|
| 2848 |
+
"step": 412
|
| 2849 |
+
},
|
| 2850 |
+
{
|
| 2851 |
+
"epoch": 1.16,
|
| 2852 |
+
"learning_rate": 0.00019865655012042335,
|
| 2853 |
+
"loss": 0.421,
|
| 2854 |
+
"step": 413
|
| 2855 |
+
},
|
| 2856 |
+
{
|
| 2857 |
+
"epoch": 1.16,
|
| 2858 |
+
"learning_rate": 0.00019819675138782784,
|
| 2859 |
+
"loss": 0.3954,
|
| 2860 |
+
"step": 414
|
| 2861 |
+
},
|
| 2862 |
+
{
|
| 2863 |
+
"epoch": 1.16,
|
| 2864 |
+
"learning_rate": 0.0001977364470941166,
|
| 2865 |
+
"loss": 0.4557,
|
| 2866 |
+
"step": 415
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 1.17,
|
| 2870 |
+
"learning_rate": 0.0001972756420676638,
|
| 2871 |
+
"loss": 0.5205,
|
| 2872 |
+
"step": 416
|
| 2873 |
+
},
|
| 2874 |
+
{
|
| 2875 |
+
"epoch": 1.17,
|
| 2876 |
+
"learning_rate": 0.00019681434114209614,
|
| 2877 |
+
"loss": 0.6422,
|
| 2878 |
+
"step": 417
|
| 2879 |
+
},
|
| 2880 |
+
{
|
| 2881 |
+
"epoch": 1.17,
|
| 2882 |
+
"learning_rate": 0.0001963525491562421,
|
| 2883 |
+
"loss": 0.3991,
|
| 2884 |
+
"step": 418
|
| 2885 |
+
},
|
| 2886 |
+
{
|
| 2887 |
+
"epoch": 1.18,
|
| 2888 |
+
"learning_rate": 0.00019589027095408105,
|
| 2889 |
+
"loss": 0.4185,
|
| 2890 |
+
"step": 419
|
| 2891 |
+
},
|
| 2892 |
+
{
|
| 2893 |
+
"epoch": 1.18,
|
| 2894 |
+
"learning_rate": 0.0001954275113846926,
|
| 2895 |
+
"loss": 0.4448,
|
| 2896 |
+
"step": 420
|
| 2897 |
+
},
|
| 2898 |
+
{
|
| 2899 |
+
"epoch": 1.18,
|
| 2900 |
+
"learning_rate": 0.00019496427530220565,
|
| 2901 |
+
"loss": 0.5212,
|
| 2902 |
+
"step": 421
|
| 2903 |
+
},
|
| 2904 |
+
{
|
| 2905 |
+
"epoch": 1.18,
|
| 2906 |
+
"learning_rate": 0.0001945005675657475,
|
| 2907 |
+
"loss": 0.5918,
|
| 2908 |
+
"step": 422
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 1.19,
|
| 2912 |
+
"learning_rate": 0.0001940363930393929,
|
| 2913 |
+
"loss": 0.3506,
|
| 2914 |
+
"step": 423
|
| 2915 |
+
},
|
| 2916 |
+
{
|
| 2917 |
+
"epoch": 1.19,
|
| 2918 |
+
"learning_rate": 0.00019357175659211298,
|
| 2919 |
+
"loss": 0.6174,
|
| 2920 |
+
"step": 424
|
| 2921 |
+
},
|
| 2922 |
+
{
|
| 2923 |
+
"epoch": 1.19,
|
| 2924 |
+
"learning_rate": 0.00019310666309772424,
|
| 2925 |
+
"loss": 0.3454,
|
| 2926 |
+
"step": 425
|
| 2927 |
+
},
|
| 2928 |
+
{
|
| 2929 |
+
"epoch": 1.19,
|
| 2930 |
+
"eval_loss": 0.5155076384544373,
|
| 2931 |
+
"eval_runtime": 26.4029,
|
| 2932 |
+
"eval_samples_per_second": 7.575,
|
| 2933 |
+
"eval_steps_per_second": 1.894,
|
| 2934 |
+
"step": 425
|
| 2935 |
+
},
|
| 2936 |
+
{
|
| 2937 |
+
"dharma_eval_accuracy": 0.46314133898648113,
|
| 2938 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.4074074074074074,
|
| 2939 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5,
|
| 2940 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 2941 |
+
"dharma_eval_accuracy_MMLU": 0.2962962962962963,
|
| 2942 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 2943 |
+
"dharma_eval_accuracy_bigbench": 0.5081967213114754,
|
| 2944 |
+
"dharma_eval_accuracy_openbookqa": 0.2777777777777778,
|
| 2945 |
+
"dharma_eval_accuracy_truthful_qa": 0.6481481481481481,
|
| 2946 |
+
"dharma_eval_accuracy_winogrande": 0.5555555555555556,
|
| 2947 |
+
"dharma_loss": 2.406665512084961,
|
| 2948 |
+
"epoch": 1.19,
|
| 2949 |
+
"step": 425
|
| 2950 |
+
},
|
| 2951 |
+
{
|
| 2952 |
+
"epoch": 1.2,
|
| 2953 |
+
"learning_rate": 0.0001926411174348373,
|
| 2954 |
+
"loss": 0.525,
|
| 2955 |
+
"step": 426
|
| 2956 |
+
},
|
| 2957 |
+
{
|
| 2958 |
+
"epoch": 1.2,
|
| 2959 |
+
"learning_rate": 0.00019217512448680584,
|
| 2960 |
+
"loss": 0.4988,
|
| 2961 |
+
"step": 427
|
| 2962 |
+
},
|
| 2963 |
+
{
|
| 2964 |
+
"epoch": 1.2,
|
| 2965 |
+
"learning_rate": 0.0001917086891416755,
|
| 2966 |
+
"loss": 0.3246,
|
| 2967 |
+
"step": 428
|
| 2968 |
+
},
|
| 2969 |
+
{
|
| 2970 |
+
"epoch": 1.2,
|
| 2971 |
+
"learning_rate": 0.00019124181629213226,
|
| 2972 |
+
"loss": 0.6274,
|
| 2973 |
+
"step": 429
|
| 2974 |
+
},
|
| 2975 |
+
{
|
| 2976 |
+
"epoch": 1.21,
|
| 2977 |
+
"learning_rate": 0.0001907745108354514,
|
| 2978 |
+
"loss": 0.8056,
|
| 2979 |
+
"step": 430
|
| 2980 |
+
},
|
| 2981 |
+
{
|
| 2982 |
+
"epoch": 1.21,
|
| 2983 |
+
"learning_rate": 0.0001903067776734461,
|
| 2984 |
+
"loss": 0.4925,
|
| 2985 |
+
"step": 431
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 1.21,
|
| 2989 |
+
"learning_rate": 0.00018983862171241577,
|
| 2990 |
+
"loss": 0.5953,
|
| 2991 |
+
"step": 432
|
| 2992 |
+
},
|
| 2993 |
+
{
|
| 2994 |
+
"epoch": 1.22,
|
| 2995 |
+
"learning_rate": 0.000189370047863095,
|
| 2996 |
+
"loss": 0.5673,
|
| 2997 |
+
"step": 433
|
| 2998 |
+
},
|
| 2999 |
+
{
|
| 3000 |
+
"epoch": 1.22,
|
| 3001 |
+
"learning_rate": 0.00018890106104060174,
|
| 3002 |
+
"loss": 0.5765,
|
| 3003 |
+
"step": 434
|
| 3004 |
+
},
|
| 3005 |
+
{
|
| 3006 |
+
"epoch": 1.22,
|
| 3007 |
+
"learning_rate": 0.0001884316661643858,
|
| 3008 |
+
"loss": 0.453,
|
| 3009 |
+
"step": 435
|
| 3010 |
+
},
|
| 3011 |
+
{
|
| 3012 |
+
"epoch": 1.22,
|
| 3013 |
+
"learning_rate": 0.00018796186815817743,
|
| 3014 |
+
"loss": 0.4503,
|
| 3015 |
+
"step": 436
|
| 3016 |
+
},
|
| 3017 |
+
{
|
| 3018 |
+
"epoch": 1.23,
|
| 3019 |
+
"learning_rate": 0.00018749167194993528,
|
| 3020 |
+
"loss": 0.4895,
|
| 3021 |
+
"step": 437
|
| 3022 |
+
},
|
| 3023 |
+
{
|
| 3024 |
+
"epoch": 1.23,
|
| 3025 |
+
"learning_rate": 0.0001870210824717951,
|
| 3026 |
+
"loss": 0.432,
|
| 3027 |
+
"step": 438
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 1.23,
|
| 3031 |
+
"learning_rate": 0.0001865501046600179,
|
| 3032 |
+
"loss": 0.3827,
|
| 3033 |
+
"step": 439
|
| 3034 |
+
},
|
| 3035 |
+
{
|
| 3036 |
+
"epoch": 1.24,
|
| 3037 |
+
"learning_rate": 0.00018607874345493805,
|
| 3038 |
+
"loss": 0.5148,
|
| 3039 |
+
"step": 440
|
| 3040 |
+
},
|
| 3041 |
+
{
|
| 3042 |
+
"epoch": 1.24,
|
| 3043 |
+
"learning_rate": 0.0001856070038009115,
|
| 3044 |
+
"loss": 0.5425,
|
| 3045 |
+
"step": 441
|
| 3046 |
+
},
|
| 3047 |
+
{
|
| 3048 |
+
"epoch": 1.24,
|
| 3049 |
+
"learning_rate": 0.00018513489064626398,
|
| 3050 |
+
"loss": 0.6838,
|
| 3051 |
+
"step": 442
|
| 3052 |
+
},
|
| 3053 |
+
{
|
| 3054 |
+
"epoch": 1.24,
|
| 3055 |
+
"learning_rate": 0.00018466240894323918,
|
| 3056 |
+
"loss": 0.4326,
|
| 3057 |
+
"step": 443
|
| 3058 |
+
},
|
| 3059 |
+
{
|
| 3060 |
+
"epoch": 1.25,
|
| 3061 |
+
"learning_rate": 0.00018418956364794653,
|
| 3062 |
+
"loss": 0.4965,
|
| 3063 |
+
"step": 444
|
| 3064 |
+
},
|
| 3065 |
+
{
|
| 3066 |
+
"epoch": 1.25,
|
| 3067 |
+
"learning_rate": 0.00018371635972030942,
|
| 3068 |
+
"loss": 0.4713,
|
| 3069 |
+
"step": 445
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 1.25,
|
| 3073 |
+
"learning_rate": 0.00018324280212401314,
|
| 3074 |
+
"loss": 0.5163,
|
| 3075 |
+
"step": 446
|
| 3076 |
+
},
|
| 3077 |
+
{
|
| 3078 |
+
"epoch": 1.25,
|
| 3079 |
+
"learning_rate": 0.00018276889582645277,
|
| 3080 |
+
"loss": 0.3899,
|
| 3081 |
+
"step": 447
|
| 3082 |
+
},
|
| 3083 |
+
{
|
| 3084 |
+
"epoch": 1.26,
|
| 3085 |
+
"learning_rate": 0.00018229464579868121,
|
| 3086 |
+
"loss": 0.3494,
|
| 3087 |
+
"step": 448
|
| 3088 |
+
},
|
| 3089 |
+
{
|
| 3090 |
+
"epoch": 1.26,
|
| 3091 |
+
"learning_rate": 0.00018182005701535678,
|
| 3092 |
+
"loss": 0.4522,
|
| 3093 |
+
"step": 449
|
| 3094 |
+
},
|
| 3095 |
+
{
|
| 3096 |
+
"epoch": 1.26,
|
| 3097 |
+
"learning_rate": 0.00018134513445469127,
|
| 3098 |
+
"loss": 0.4883,
|
| 3099 |
+
"step": 450
|
| 3100 |
+
},
|
| 3101 |
+
{
|
| 3102 |
+
"epoch": 1.26,
|
| 3103 |
+
"eval_loss": 0.5140238404273987,
|
| 3104 |
+
"eval_runtime": 26.3956,
|
| 3105 |
+
"eval_samples_per_second": 7.577,
|
| 3106 |
+
"eval_steps_per_second": 1.894,
|
| 3107 |
+
"step": 450
|
| 3108 |
+
},
|
| 3109 |
+
{
|
| 3110 |
+
"dharma_eval_accuracy": 0.4671948197773273,
|
| 3111 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.3888888888888889,
|
| 3112 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
|
| 3113 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 3114 |
+
"dharma_eval_accuracy_MMLU": 0.3148148148148148,
|
| 3115 |
+
"dharma_eval_accuracy_agieval": 0.2542372881355932,
|
| 3116 |
+
"dharma_eval_accuracy_bigbench": 0.5245901639344263,
|
| 3117 |
+
"dharma_eval_accuracy_openbookqa": 0.3148148148148148,
|
| 3118 |
+
"dharma_eval_accuracy_truthful_qa": 0.7592592592592593,
|
| 3119 |
+
"dharma_eval_accuracy_winogrande": 0.42592592592592593,
|
| 3120 |
+
"dharma_loss": 2.752000973701477,
|
| 3121 |
+
"epoch": 1.26,
|
| 3122 |
+
"step": 450
|
| 3123 |
+
},
|
| 3124 |
+
{
|
| 3125 |
+
"epoch": 1.27,
|
| 3126 |
+
"learning_rate": 0.00018086988309839753,
|
| 3127 |
+
"loss": 0.5255,
|
| 3128 |
+
"step": 451
|
| 3129 |
+
},
|
| 3130 |
+
{
|
| 3131 |
+
"epoch": 1.27,
|
| 3132 |
+
"learning_rate": 0.00018039430793163753,
|
| 3133 |
+
"loss": 0.6,
|
| 3134 |
+
"step": 452
|
| 3135 |
+
},
|
| 3136 |
+
{
|
| 3137 |
+
"epoch": 1.27,
|
| 3138 |
+
"learning_rate": 0.0001799184139429696,
|
| 3139 |
+
"loss": 0.214,
|
| 3140 |
+
"step": 453
|
| 3141 |
+
},
|
| 3142 |
+
{
|
| 3143 |
+
"epoch": 1.27,
|
| 3144 |
+
"learning_rate": 0.00017944220612429662,
|
| 3145 |
+
"loss": 0.4462,
|
| 3146 |
+
"step": 454
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 1.28,
|
| 3150 |
+
"learning_rate": 0.0001789656894708132,
|
| 3151 |
+
"loss": 0.3911,
|
| 3152 |
+
"step": 455
|
| 3153 |
+
},
|
| 3154 |
+
{
|
| 3155 |
+
"epoch": 1.28,
|
| 3156 |
+
"learning_rate": 0.00017848886898095352,
|
| 3157 |
+
"loss": 0.5309,
|
| 3158 |
+
"step": 456
|
| 3159 |
+
},
|
| 3160 |
+
{
|
| 3161 |
+
"epoch": 1.28,
|
| 3162 |
+
"learning_rate": 0.00017801174965633893,
|
| 3163 |
+
"loss": 0.3596,
|
| 3164 |
+
"step": 457
|
| 3165 |
+
},
|
| 3166 |
+
{
|
| 3167 |
+
"epoch": 1.29,
|
| 3168 |
+
"learning_rate": 0.00017753433650172538,
|
| 3169 |
+
"loss": 0.427,
|
| 3170 |
+
"step": 458
|
| 3171 |
+
},
|
| 3172 |
+
{
|
| 3173 |
+
"epoch": 1.29,
|
| 3174 |
+
"learning_rate": 0.00017705663452495098,
|
| 3175 |
+
"loss": 0.5063,
|
| 3176 |
+
"step": 459
|
| 3177 |
+
},
|
| 3178 |
+
{
|
| 3179 |
+
"epoch": 1.29,
|
| 3180 |
+
"learning_rate": 0.00017657864873688343,
|
| 3181 |
+
"loss": 0.5837,
|
| 3182 |
+
"step": 460
|
| 3183 |
+
},
|
| 3184 |
+
{
|
| 3185 |
+
"epoch": 1.29,
|
| 3186 |
+
"learning_rate": 0.00017610038415136734,
|
| 3187 |
+
"loss": 0.5938,
|
| 3188 |
+
"step": 461
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 1.3,
|
| 3192 |
+
"learning_rate": 0.00017562184578517203,
|
| 3193 |
+
"loss": 0.5135,
|
| 3194 |
+
"step": 462
|
| 3195 |
+
},
|
| 3196 |
+
{
|
| 3197 |
+
"epoch": 1.3,
|
| 3198 |
+
"learning_rate": 0.00017514303865793848,
|
| 3199 |
+
"loss": 0.3718,
|
| 3200 |
+
"step": 463
|
| 3201 |
+
},
|
| 3202 |
+
{
|
| 3203 |
+
"epoch": 1.3,
|
| 3204 |
+
"learning_rate": 0.00017466396779212693,
|
| 3205 |
+
"loss": 0.5354,
|
| 3206 |
+
"step": 464
|
| 3207 |
+
},
|
| 3208 |
+
{
|
| 3209 |
+
"epoch": 1.31,
|
| 3210 |
+
"learning_rate": 0.00017418463821296395,
|
| 3211 |
+
"loss": 0.4169,
|
| 3212 |
+
"step": 465
|
| 3213 |
+
},
|
| 3214 |
+
{
|
| 3215 |
+
"epoch": 1.31,
|
| 3216 |
+
"learning_rate": 0.00017370505494839012,
|
| 3217 |
+
"loss": 0.383,
|
| 3218 |
+
"step": 466
|
| 3219 |
+
},
|
| 3220 |
+
{
|
| 3221 |
+
"epoch": 1.31,
|
| 3222 |
+
"learning_rate": 0.00017322522302900696,
|
| 3223 |
+
"loss": 0.6588,
|
| 3224 |
+
"step": 467
|
| 3225 |
+
},
|
| 3226 |
+
{
|
| 3227 |
+
"epoch": 1.31,
|
| 3228 |
+
"learning_rate": 0.00017274514748802418,
|
| 3229 |
+
"loss": 0.4773,
|
| 3230 |
+
"step": 468
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 1.32,
|
| 3234 |
+
"learning_rate": 0.00017226483336120707,
|
| 3235 |
+
"loss": 0.4327,
|
| 3236 |
+
"step": 469
|
| 3237 |
+
},
|
| 3238 |
+
{
|
| 3239 |
+
"epoch": 1.32,
|
| 3240 |
+
"learning_rate": 0.00017178428568682353,
|
| 3241 |
+
"loss": 0.5411,
|
| 3242 |
+
"step": 470
|
| 3243 |
+
},
|
| 3244 |
+
{
|
| 3245 |
+
"epoch": 1.32,
|
| 3246 |
+
"learning_rate": 0.00017130350950559132,
|
| 3247 |
+
"loss": 0.4514,
|
| 3248 |
+
"step": 471
|
| 3249 |
+
},
|
| 3250 |
+
{
|
| 3251 |
+
"epoch": 1.32,
|
| 3252 |
+
"learning_rate": 0.00017082250986062502,
|
| 3253 |
+
"loss": 0.4935,
|
| 3254 |
+
"step": 472
|
| 3255 |
+
},
|
| 3256 |
+
{
|
| 3257 |
+
"epoch": 1.33,
|
| 3258 |
+
"learning_rate": 0.00017034129179738328,
|
| 3259 |
+
"loss": 0.4042,
|
| 3260 |
+
"step": 473
|
| 3261 |
+
},
|
| 3262 |
+
{
|
| 3263 |
+
"epoch": 1.33,
|
| 3264 |
+
"learning_rate": 0.00016985986036361598,
|
| 3265 |
+
"loss": 0.3173,
|
| 3266 |
+
"step": 474
|
| 3267 |
+
},
|
| 3268 |
+
{
|
| 3269 |
+
"epoch": 1.33,
|
| 3270 |
+
"learning_rate": 0.00016937822060931096,
|
| 3271 |
+
"loss": 0.4456,
|
| 3272 |
+
"step": 475
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 1.33,
|
| 3276 |
+
"eval_loss": 0.5098786354064941,
|
| 3277 |
+
"eval_runtime": 26.3977,
|
| 3278 |
+
"eval_samples_per_second": 7.576,
|
| 3279 |
+
"eval_steps_per_second": 1.894,
|
| 3280 |
+
"step": 475
|
| 3281 |
+
},
|
| 3282 |
+
{
|
| 3283 |
+
"dharma_eval_accuracy": 0.4939437909707429,
|
| 3284 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.48148148148148145,
|
| 3285 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5185185185185185,
|
| 3286 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 3287 |
+
"dharma_eval_accuracy_MMLU": 0.35185185185185186,
|
| 3288 |
+
"dharma_eval_accuracy_agieval": 0.2542372881355932,
|
| 3289 |
+
"dharma_eval_accuracy_bigbench": 0.5245901639344263,
|
| 3290 |
+
"dharma_eval_accuracy_openbookqa": 0.3148148148148148,
|
| 3291 |
+
"dharma_eval_accuracy_truthful_qa": 0.7407407407407407,
|
| 3292 |
+
"dharma_eval_accuracy_winogrande": 0.5555555555555556,
|
| 3293 |
+
"dharma_loss": 2.942282904624939,
|
| 3294 |
+
"epoch": 1.33,
|
| 3295 |
+
"step": 475
|
| 3296 |
+
},
|
| 3297 |
+
{
|
| 3298 |
+
"epoch": 1.34,
|
| 3299 |
+
"learning_rate": 0.00016889637758664137,
|
| 3300 |
+
"loss": 0.3904,
|
| 3301 |
+
"step": 476
|
| 3302 |
+
},
|
| 3303 |
+
{
|
| 3304 |
+
"epoch": 1.34,
|
| 3305 |
+
"learning_rate": 0.0001684143363499125,
|
| 3306 |
+
"loss": 0.4446,
|
| 3307 |
+
"step": 477
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 1.34,
|
| 3311 |
+
"learning_rate": 0.00016793210195550882,
|
| 3312 |
+
"loss": 0.4849,
|
| 3313 |
+
"step": 478
|
| 3314 |
+
},
|
| 3315 |
+
{
|
| 3316 |
+
"epoch": 1.34,
|
| 3317 |
+
"learning_rate": 0.00016744967946184097,
|
| 3318 |
+
"loss": 0.4961,
|
| 3319 |
+
"step": 479
|
| 3320 |
+
},
|
| 3321 |
+
{
|
| 3322 |
+
"epoch": 1.35,
|
| 3323 |
+
"learning_rate": 0.00016696707392929266,
|
| 3324 |
+
"loss": 0.5676,
|
| 3325 |
+
"step": 480
|
| 3326 |
+
},
|
| 3327 |
+
{
|
| 3328 |
+
"epoch": 1.35,
|
| 3329 |
+
"learning_rate": 0.0001664842904201674,
|
| 3330 |
+
"loss": 0.4996,
|
| 3331 |
+
"step": 481
|
| 3332 |
+
},
|
| 3333 |
+
{
|
| 3334 |
+
"epoch": 1.35,
|
| 3335 |
+
"learning_rate": 0.00016600133399863594,
|
| 3336 |
+
"loss": 0.2721,
|
| 3337 |
+
"step": 482
|
| 3338 |
+
},
|
| 3339 |
+
{
|
| 3340 |
+
"epoch": 1.36,
|
| 3341 |
+
"learning_rate": 0.00016551820973068255,
|
| 3342 |
+
"loss": 0.3858,
|
| 3343 |
+
"step": 483
|
| 3344 |
+
},
|
| 3345 |
+
{
|
| 3346 |
+
"epoch": 1.36,
|
| 3347 |
+
"learning_rate": 0.00016503492268405215,
|
| 3348 |
+
"loss": 0.4972,
|
| 3349 |
+
"step": 484
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 1.36,
|
| 3353 |
+
"learning_rate": 0.0001645514779281972,
|
| 3354 |
+
"loss": 0.5966,
|
| 3355 |
+
"step": 485
|
| 3356 |
+
},
|
| 3357 |
+
{
|
| 3358 |
+
"epoch": 1.36,
|
| 3359 |
+
"learning_rate": 0.00016406788053422445,
|
| 3360 |
+
"loss": 0.3846,
|
| 3361 |
+
"step": 486
|
| 3362 |
+
},
|
| 3363 |
+
{
|
| 3364 |
+
"epoch": 1.37,
|
| 3365 |
+
"learning_rate": 0.00016358413557484167,
|
| 3366 |
+
"loss": 0.4582,
|
| 3367 |
+
"step": 487
|
| 3368 |
+
},
|
| 3369 |
+
{
|
| 3370 |
+
"epoch": 1.37,
|
| 3371 |
+
"learning_rate": 0.00016310024812430458,
|
| 3372 |
+
"loss": 0.5007,
|
| 3373 |
+
"step": 488
|
| 3374 |
+
},
|
| 3375 |
+
{
|
| 3376 |
+
"epoch": 1.37,
|
| 3377 |
+
"learning_rate": 0.00016261622325836362,
|
| 3378 |
+
"loss": 0.3344,
|
| 3379 |
+
"step": 489
|
| 3380 |
+
},
|
| 3381 |
+
{
|
| 3382 |
+
"epoch": 1.38,
|
| 3383 |
+
"learning_rate": 0.00016213206605421063,
|
| 3384 |
+
"loss": 0.483,
|
| 3385 |
+
"step": 490
|
| 3386 |
+
},
|
| 3387 |
+
{
|
| 3388 |
+
"epoch": 1.38,
|
| 3389 |
+
"learning_rate": 0.0001616477815904255,
|
| 3390 |
+
"loss": 0.4295,
|
| 3391 |
+
"step": 491
|
| 3392 |
+
},
|
| 3393 |
+
{
|
| 3394 |
+
"epoch": 1.38,
|
| 3395 |
+
"learning_rate": 0.0001611633749469231,
|
| 3396 |
+
"loss": 0.4847,
|
| 3397 |
+
"step": 492
|
| 3398 |
+
},
|
| 3399 |
+
{
|
| 3400 |
+
"epoch": 1.38,
|
| 3401 |
+
"learning_rate": 0.0001606788512048999,
|
| 3402 |
+
"loss": 0.3394,
|
| 3403 |
+
"step": 493
|
| 3404 |
+
},
|
| 3405 |
+
{
|
| 3406 |
+
"epoch": 1.39,
|
| 3407 |
+
"learning_rate": 0.00016019421544678073,
|
| 3408 |
+
"loss": 0.4139,
|
| 3409 |
+
"step": 494
|
| 3410 |
+
},
|
| 3411 |
+
{
|
| 3412 |
+
"epoch": 1.39,
|
| 3413 |
+
"learning_rate": 0.00015970947275616536,
|
| 3414 |
+
"loss": 0.3148,
|
| 3415 |
+
"step": 495
|
| 3416 |
+
},
|
| 3417 |
+
{
|
| 3418 |
+
"epoch": 1.39,
|
| 3419 |
+
"learning_rate": 0.00015922462821777524,
|
| 3420 |
+
"loss": 0.4729,
|
| 3421 |
+
"step": 496
|
| 3422 |
+
},
|
| 3423 |
+
{
|
| 3424 |
+
"epoch": 1.4,
|
| 3425 |
+
"learning_rate": 0.00015873968691740005,
|
| 3426 |
+
"loss": 0.3881,
|
| 3427 |
+
"step": 497
|
| 3428 |
+
},
|
| 3429 |
+
{
|
| 3430 |
+
"epoch": 1.4,
|
| 3431 |
+
"learning_rate": 0.00015825465394184472,
|
| 3432 |
+
"loss": 0.4628,
|
| 3433 |
+
"step": 498
|
| 3434 |
+
},
|
| 3435 |
+
{
|
| 3436 |
+
"epoch": 1.4,
|
| 3437 |
+
"learning_rate": 0.00015776953437887553,
|
| 3438 |
+
"loss": 0.5961,
|
| 3439 |
+
"step": 499
|
| 3440 |
+
},
|
| 3441 |
+
{
|
| 3442 |
+
"epoch": 1.4,
|
| 3443 |
+
"learning_rate": 0.00015728433331716724,
|
| 3444 |
+
"loss": 0.6399,
|
| 3445 |
+
"step": 500
|
| 3446 |
+
},
|
| 3447 |
+
{
|
| 3448 |
+
"epoch": 1.4,
|
| 3449 |
+
"eval_loss": 0.50834059715271,
|
| 3450 |
+
"eval_runtime": 26.4232,
|
| 3451 |
+
"eval_samples_per_second": 7.569,
|
| 3452 |
+
"eval_steps_per_second": 1.892,
|
| 3453 |
+
"step": 500
|
| 3454 |
+
},
|
| 3455 |
+
{
|
| 3456 |
+
"dharma_eval_accuracy": 0.4811876184170958,
|
| 3457 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.46296296296296297,
|
| 3458 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
|
| 3459 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 3460 |
+
"dharma_eval_accuracy_MMLU": 0.37037037037037035,
|
| 3461 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 3462 |
+
"dharma_eval_accuracy_bigbench": 0.5409836065573771,
|
| 3463 |
+
"dharma_eval_accuracy_openbookqa": 0.2962962962962963,
|
| 3464 |
+
"dharma_eval_accuracy_truthful_qa": 0.6851851851851852,
|
| 3465 |
+
"dharma_eval_accuracy_winogrande": 0.46296296296296297,
|
| 3466 |
+
"dharma_loss": 2.7791563968658446,
|
| 3467 |
+
"epoch": 1.4,
|
| 3468 |
+
"step": 500
|
| 3469 |
+
},
|
| 3470 |
+
{
|
| 3471 |
+
"epoch": 1.41,
|
| 3472 |
+
"learning_rate": 0.0001567990558462493,
|
| 3473 |
+
"loss": 0.3633,
|
| 3474 |
+
"step": 501
|
| 3475 |
+
},
|
| 3476 |
+
{
|
| 3477 |
+
"epoch": 1.41,
|
| 3478 |
+
"learning_rate": 0.0001563137070564528,
|
| 3479 |
+
"loss": 0.6472,
|
| 3480 |
+
"step": 502
|
| 3481 |
+
},
|
| 3482 |
+
{
|
| 3483 |
+
"epoch": 1.41,
|
| 3484 |
+
"learning_rate": 0.00015582829203885688,
|
| 3485 |
+
"loss": 0.3952,
|
| 3486 |
+
"step": 503
|
| 3487 |
+
},
|
| 3488 |
+
{
|
| 3489 |
+
"epoch": 1.41,
|
| 3490 |
+
"learning_rate": 0.00015534281588523534,
|
| 3491 |
+
"loss": 0.3335,
|
| 3492 |
+
"step": 504
|
| 3493 |
+
},
|
| 3494 |
+
{
|
| 3495 |
+
"epoch": 1.42,
|
| 3496 |
+
"learning_rate": 0.0001548572836880034,
|
| 3497 |
+
"loss": 0.6748,
|
| 3498 |
+
"step": 505
|
| 3499 |
+
},
|
| 3500 |
+
{
|
| 3501 |
+
"epoch": 1.42,
|
| 3502 |
+
"learning_rate": 0.00015437170054016395,
|
| 3503 |
+
"loss": 0.5213,
|
| 3504 |
+
"step": 506
|
| 3505 |
+
},
|
| 3506 |
+
{
|
| 3507 |
+
"epoch": 1.42,
|
| 3508 |
+
"learning_rate": 0.00015388607153525452,
|
| 3509 |
+
"loss": 0.5454,
|
| 3510 |
+
"step": 507
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 1.43,
|
| 3514 |
+
"learning_rate": 0.00015340040176729347,
|
| 3515 |
+
"loss": 0.605,
|
| 3516 |
+
"step": 508
|
| 3517 |
+
},
|
| 3518 |
+
{
|
| 3519 |
+
"epoch": 1.43,
|
| 3520 |
+
"learning_rate": 0.000152914696330727,
|
| 3521 |
+
"loss": 0.5786,
|
| 3522 |
+
"step": 509
|
| 3523 |
+
},
|
| 3524 |
+
{
|
| 3525 |
+
"epoch": 1.43,
|
| 3526 |
+
"learning_rate": 0.00015242896032037522,
|
| 3527 |
+
"loss": 0.5918,
|
| 3528 |
+
"step": 510
|
| 3529 |
+
},
|
| 3530 |
+
{
|
| 3531 |
+
"epoch": 1.43,
|
| 3532 |
+
"learning_rate": 0.00015194319883137905,
|
| 3533 |
+
"loss": 0.7016,
|
| 3534 |
+
"step": 511
|
| 3535 |
+
},
|
| 3536 |
+
{
|
| 3537 |
+
"epoch": 1.44,
|
| 3538 |
+
"learning_rate": 0.0001514574169591466,
|
| 3539 |
+
"loss": 0.3418,
|
| 3540 |
+
"step": 512
|
| 3541 |
+
},
|
| 3542 |
+
{
|
| 3543 |
+
"epoch": 1.44,
|
| 3544 |
+
"learning_rate": 0.0001509716197992999,
|
| 3545 |
+
"loss": 0.4958,
|
| 3546 |
+
"step": 513
|
| 3547 |
+
},
|
| 3548 |
+
{
|
| 3549 |
+
"epoch": 1.44,
|
| 3550 |
+
"learning_rate": 0.0001504858124476213,
|
| 3551 |
+
"loss": 0.3953,
|
| 3552 |
+
"step": 514
|
| 3553 |
+
},
|
| 3554 |
+
{
|
| 3555 |
+
"epoch": 1.45,
|
| 3556 |
+
"learning_rate": 0.00015,
|
| 3557 |
+
"loss": 0.5772,
|
| 3558 |
+
"step": 515
|
| 3559 |
+
},
|
| 3560 |
+
{
|
| 3561 |
+
"epoch": 1.45,
|
| 3562 |
+
"learning_rate": 0.0001495141875523787,
|
| 3563 |
+
"loss": 0.4234,
|
| 3564 |
+
"step": 516
|
| 3565 |
+
},
|
| 3566 |
+
{
|
| 3567 |
+
"epoch": 1.45,
|
| 3568 |
+
"learning_rate": 0.00014902838020070008,
|
| 3569 |
+
"loss": 0.2827,
|
| 3570 |
+
"step": 517
|
| 3571 |
+
},
|
| 3572 |
+
{
|
| 3573 |
+
"epoch": 1.45,
|
| 3574 |
+
"learning_rate": 0.0001485425830408534,
|
| 3575 |
+
"loss": 0.4302,
|
| 3576 |
+
"step": 518
|
| 3577 |
+
},
|
| 3578 |
+
{
|
| 3579 |
+
"epoch": 1.46,
|
| 3580 |
+
"learning_rate": 0.00014805680116862095,
|
| 3581 |
+
"loss": 0.4149,
|
| 3582 |
+
"step": 519
|
| 3583 |
+
},
|
| 3584 |
+
{
|
| 3585 |
+
"epoch": 1.46,
|
| 3586 |
+
"learning_rate": 0.00014757103967962475,
|
| 3587 |
+
"loss": 0.4118,
|
| 3588 |
+
"step": 520
|
| 3589 |
+
},
|
| 3590 |
+
{
|
| 3591 |
+
"epoch": 1.46,
|
| 3592 |
+
"learning_rate": 0.00014708530366927297,
|
| 3593 |
+
"loss": 0.3965,
|
| 3594 |
+
"step": 521
|
| 3595 |
+
},
|
| 3596 |
+
{
|
| 3597 |
+
"epoch": 1.47,
|
| 3598 |
+
"learning_rate": 0.0001465995982327065,
|
| 3599 |
+
"loss": 0.5043,
|
| 3600 |
+
"step": 522
|
| 3601 |
+
},
|
| 3602 |
+
{
|
| 3603 |
+
"epoch": 1.47,
|
| 3604 |
+
"learning_rate": 0.0001461139284647455,
|
| 3605 |
+
"loss": 0.4457,
|
| 3606 |
+
"step": 523
|
| 3607 |
+
},
|
| 3608 |
+
{
|
| 3609 |
+
"epoch": 1.47,
|
| 3610 |
+
"learning_rate": 0.00014562829945983607,
|
| 3611 |
+
"loss": 0.5667,
|
| 3612 |
+
"step": 524
|
| 3613 |
+
},
|
| 3614 |
+
{
|
| 3615 |
+
"epoch": 1.47,
|
| 3616 |
+
"learning_rate": 0.00014514271631199663,
|
| 3617 |
+
"loss": 0.4982,
|
| 3618 |
+
"step": 525
|
| 3619 |
+
},
|
| 3620 |
+
{
|
| 3621 |
+
"epoch": 1.47,
|
| 3622 |
+
"eval_loss": 0.5073952674865723,
|
| 3623 |
+
"eval_runtime": 26.2985,
|
| 3624 |
+
"eval_samples_per_second": 7.605,
|
| 3625 |
+
"eval_steps_per_second": 1.901,
|
| 3626 |
+
"step": 525
|
| 3627 |
+
},
|
| 3628 |
+
{
|
| 3629 |
+
"dharma_eval_accuracy": 0.4803106029681313,
|
| 3630 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.5,
|
| 3631 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
|
| 3632 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 3633 |
+
"dharma_eval_accuracy_MMLU": 0.4444444444444444,
|
| 3634 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 3635 |
+
"dharma_eval_accuracy_bigbench": 0.45901639344262296,
|
| 3636 |
+
"dharma_eval_accuracy_openbookqa": 0.3148148148148148,
|
| 3637 |
+
"dharma_eval_accuracy_truthful_qa": 0.5925925925925926,
|
| 3638 |
+
"dharma_eval_accuracy_winogrande": 0.5,
|
| 3639 |
+
"dharma_loss": 2.697408908367157,
|
| 3640 |
+
"epoch": 1.47,
|
| 3641 |
+
"step": 525
|
| 3642 |
+
},
|
| 3643 |
+
{
|
| 3644 |
+
"epoch": 1.48,
|
| 3645 |
+
"learning_rate": 0.00014465718411476463,
|
| 3646 |
+
"loss": 0.3422,
|
| 3647 |
+
"step": 526
|
| 3648 |
+
},
|
| 3649 |
+
{
|
| 3650 |
+
"epoch": 1.48,
|
| 3651 |
+
"learning_rate": 0.00014417170796114312,
|
| 3652 |
+
"loss": 0.3436,
|
| 3653 |
+
"step": 527
|
| 3654 |
+
},
|
| 3655 |
+
{
|
| 3656 |
+
"epoch": 1.48,
|
| 3657 |
+
"learning_rate": 0.0001436862929435472,
|
| 3658 |
+
"loss": 0.3001,
|
| 3659 |
+
"step": 528
|
| 3660 |
+
},
|
| 3661 |
+
{
|
| 3662 |
+
"epoch": 1.48,
|
| 3663 |
+
"learning_rate": 0.00014320094415375072,
|
| 3664 |
+
"loss": 0.4839,
|
| 3665 |
+
"step": 529
|
| 3666 |
+
},
|
| 3667 |
+
{
|
| 3668 |
+
"epoch": 1.49,
|
| 3669 |
+
"learning_rate": 0.0001427156666828328,
|
| 3670 |
+
"loss": 0.3702,
|
| 3671 |
+
"step": 530
|
| 3672 |
+
},
|
| 3673 |
+
{
|
| 3674 |
+
"epoch": 1.49,
|
| 3675 |
+
"learning_rate": 0.00014223046562112442,
|
| 3676 |
+
"loss": 0.5617,
|
| 3677 |
+
"step": 531
|
| 3678 |
+
},
|
| 3679 |
+
{
|
| 3680 |
+
"epoch": 1.49,
|
| 3681 |
+
"learning_rate": 0.00014174534605815525,
|
| 3682 |
+
"loss": 0.5076,
|
| 3683 |
+
"step": 532
|
| 3684 |
+
},
|
| 3685 |
+
{
|
| 3686 |
+
"epoch": 1.5,
|
| 3687 |
+
"learning_rate": 0.00014126031308259992,
|
| 3688 |
+
"loss": 0.5191,
|
| 3689 |
+
"step": 533
|
| 3690 |
+
},
|
| 3691 |
+
{
|
| 3692 |
+
"epoch": 1.5,
|
| 3693 |
+
"learning_rate": 0.0001407753717822248,
|
| 3694 |
+
"loss": 0.6035,
|
| 3695 |
+
"step": 534
|
| 3696 |
+
},
|
| 3697 |
+
{
|
| 3698 |
+
"epoch": 1.5,
|
| 3699 |
+
"learning_rate": 0.00014029052724383464,
|
| 3700 |
+
"loss": 0.4406,
|
| 3701 |
+
"step": 535
|
| 3702 |
+
},
|
| 3703 |
+
{
|
| 3704 |
+
"epoch": 1.5,
|
| 3705 |
+
"learning_rate": 0.00013980578455321922,
|
| 3706 |
+
"loss": 0.4745,
|
| 3707 |
+
"step": 536
|
| 3708 |
+
},
|
| 3709 |
+
{
|
| 3710 |
+
"epoch": 1.51,
|
| 3711 |
+
"learning_rate": 0.00013932114879510005,
|
| 3712 |
+
"loss": 0.4057,
|
| 3713 |
+
"step": 537
|
| 3714 |
+
},
|
| 3715 |
+
{
|
| 3716 |
+
"epoch": 1.51,
|
| 3717 |
+
"learning_rate": 0.0001388366250530769,
|
| 3718 |
+
"loss": 0.5748,
|
| 3719 |
+
"step": 538
|
| 3720 |
+
},
|
| 3721 |
+
{
|
| 3722 |
+
"epoch": 1.51,
|
| 3723 |
+
"learning_rate": 0.0001383522184095745,
|
| 3724 |
+
"loss": 0.6466,
|
| 3725 |
+
"step": 539
|
| 3726 |
+
},
|
| 3727 |
+
{
|
| 3728 |
+
"epoch": 1.52,
|
| 3729 |
+
"learning_rate": 0.00013786793394578937,
|
| 3730 |
+
"loss": 0.4808,
|
| 3731 |
+
"step": 540
|
| 3732 |
+
},
|
| 3733 |
+
{
|
| 3734 |
+
"epoch": 1.52,
|
| 3735 |
+
"learning_rate": 0.00013738377674163635,
|
| 3736 |
+
"loss": 0.4348,
|
| 3737 |
+
"step": 541
|
| 3738 |
+
},
|
| 3739 |
+
{
|
| 3740 |
+
"epoch": 1.52,
|
| 3741 |
+
"learning_rate": 0.0001368997518756954,
|
| 3742 |
+
"loss": 0.3255,
|
| 3743 |
+
"step": 542
|
| 3744 |
+
},
|
| 3745 |
+
{
|
| 3746 |
+
"epoch": 1.52,
|
| 3747 |
+
"learning_rate": 0.00013641586442515833,
|
| 3748 |
+
"loss": 0.3578,
|
| 3749 |
+
"step": 543
|
| 3750 |
+
},
|
| 3751 |
+
{
|
| 3752 |
+
"epoch": 1.53,
|
| 3753 |
+
"learning_rate": 0.00013593211946577555,
|
| 3754 |
+
"loss": 0.4295,
|
| 3755 |
+
"step": 544
|
| 3756 |
+
},
|
| 3757 |
+
{
|
| 3758 |
+
"epoch": 1.53,
|
| 3759 |
+
"learning_rate": 0.0001354485220718028,
|
| 3760 |
+
"loss": 0.4566,
|
| 3761 |
+
"step": 545
|
| 3762 |
+
},
|
| 3763 |
+
{
|
| 3764 |
+
"epoch": 1.53,
|
| 3765 |
+
"learning_rate": 0.00013496507731594783,
|
| 3766 |
+
"loss": 0.5032,
|
| 3767 |
+
"step": 546
|
| 3768 |
+
},
|
| 3769 |
+
{
|
| 3770 |
+
"epoch": 1.54,
|
| 3771 |
+
"learning_rate": 0.00013448179026931745,
|
| 3772 |
+
"loss": 0.5479,
|
| 3773 |
+
"step": 547
|
| 3774 |
+
},
|
| 3775 |
+
{
|
| 3776 |
+
"epoch": 1.54,
|
| 3777 |
+
"learning_rate": 0.00013399866600136403,
|
| 3778 |
+
"loss": 0.4946,
|
| 3779 |
+
"step": 548
|
| 3780 |
+
},
|
| 3781 |
+
{
|
| 3782 |
+
"epoch": 1.54,
|
| 3783 |
+
"learning_rate": 0.00013351570957983257,
|
| 3784 |
+
"loss": 0.5054,
|
| 3785 |
+
"step": 549
|
| 3786 |
+
},
|
| 3787 |
+
{
|
| 3788 |
+
"epoch": 1.54,
|
| 3789 |
+
"learning_rate": 0.00013303292607070737,
|
| 3790 |
+
"loss": 0.2585,
|
| 3791 |
+
"step": 550
|
| 3792 |
+
},
|
| 3793 |
+
{
|
| 3794 |
+
"epoch": 1.54,
|
| 3795 |
+
"eval_loss": 0.5078052282333374,
|
| 3796 |
+
"eval_runtime": 26.3852,
|
| 3797 |
+
"eval_samples_per_second": 7.58,
|
| 3798 |
+
"eval_steps_per_second": 1.895,
|
| 3799 |
+
"step": 550
|
| 3800 |
+
},
|
| 3801 |
+
{
|
| 3802 |
+
"dharma_eval_accuracy": 0.46749268486788176,
|
| 3803 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.48148148148148145,
|
| 3804 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
|
| 3805 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 3806 |
+
"dharma_eval_accuracy_MMLU": 0.42592592592592593,
|
| 3807 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 3808 |
+
"dharma_eval_accuracy_bigbench": 0.4918032786885246,
|
| 3809 |
+
"dharma_eval_accuracy_openbookqa": 0.2777777777777778,
|
| 3810 |
+
"dharma_eval_accuracy_truthful_qa": 0.5370370370370371,
|
| 3811 |
+
"dharma_eval_accuracy_winogrande": 0.48148148148148145,
|
| 3812 |
+
"dharma_loss": 2.9045387988090514,
|
| 3813 |
+
"epoch": 1.54,
|
| 3814 |
+
"step": 550
|
| 3815 |
+
},
|
| 3816 |
+
{
|
| 3817 |
+
"epoch": 1.55,
|
| 3818 |
+
"learning_rate": 0.000132550320538159,
|
| 3819 |
+
"loss": 0.5333,
|
| 3820 |
+
"step": 551
|
| 3821 |
+
},
|
| 3822 |
+
{
|
| 3823 |
+
"epoch": 1.55,
|
| 3824 |
+
"learning_rate": 0.00013206789804449116,
|
| 3825 |
+
"loss": 0.3957,
|
| 3826 |
+
"step": 552
|
| 3827 |
+
},
|
| 3828 |
+
{
|
| 3829 |
+
"epoch": 1.55,
|
| 3830 |
+
"learning_rate": 0.0001315856636500875,
|
| 3831 |
+
"loss": 0.5521,
|
| 3832 |
+
"step": 553
|
| 3833 |
+
},
|
| 3834 |
+
{
|
| 3835 |
+
"epoch": 1.56,
|
| 3836 |
+
"learning_rate": 0.00013110362241335865,
|
| 3837 |
+
"loss": 0.5326,
|
| 3838 |
+
"step": 554
|
| 3839 |
+
},
|
| 3840 |
+
{
|
| 3841 |
+
"epoch": 1.56,
|
| 3842 |
+
"learning_rate": 0.00013062177939068907,
|
| 3843 |
+
"loss": 0.5668,
|
| 3844 |
+
"step": 555
|
| 3845 |
+
},
|
| 3846 |
+
{
|
| 3847 |
+
"epoch": 1.56,
|
| 3848 |
+
"learning_rate": 0.00013014013963638408,
|
| 3849 |
+
"loss": 0.3726,
|
| 3850 |
+
"step": 556
|
| 3851 |
+
},
|
| 3852 |
+
{
|
| 3853 |
+
"epoch": 1.56,
|
| 3854 |
+
"learning_rate": 0.0001296587082026167,
|
| 3855 |
+
"loss": 0.3974,
|
| 3856 |
+
"step": 557
|
| 3857 |
+
},
|
| 3858 |
+
{
|
| 3859 |
+
"epoch": 1.57,
|
| 3860 |
+
"learning_rate": 0.000129177490139375,
|
| 3861 |
+
"loss": 0.3974,
|
| 3862 |
+
"step": 558
|
| 3863 |
+
},
|
| 3864 |
+
{
|
| 3865 |
+
"epoch": 1.57,
|
| 3866 |
+
"learning_rate": 0.0001286964904944087,
|
| 3867 |
+
"loss": 0.4792,
|
| 3868 |
+
"step": 559
|
| 3869 |
+
},
|
| 3870 |
+
{
|
| 3871 |
+
"epoch": 1.57,
|
| 3872 |
+
"learning_rate": 0.00012821571431317647,
|
| 3873 |
+
"loss": 0.4565,
|
| 3874 |
+
"step": 560
|
| 3875 |
+
},
|
| 3876 |
+
{
|
| 3877 |
+
"epoch": 1.57,
|
| 3878 |
+
"learning_rate": 0.0001277351666387929,
|
| 3879 |
+
"loss": 0.3991,
|
| 3880 |
+
"step": 561
|
| 3881 |
+
},
|
| 3882 |
+
{
|
| 3883 |
+
"epoch": 1.58,
|
| 3884 |
+
"learning_rate": 0.0001272548525119758,
|
| 3885 |
+
"loss": 0.3742,
|
| 3886 |
+
"step": 562
|
| 3887 |
+
},
|
| 3888 |
+
{
|
| 3889 |
+
"epoch": 1.58,
|
| 3890 |
+
"learning_rate": 0.000126774776970993,
|
| 3891 |
+
"loss": 0.3932,
|
| 3892 |
+
"step": 563
|
| 3893 |
+
},
|
| 3894 |
+
{
|
| 3895 |
+
"epoch": 1.58,
|
| 3896 |
+
"learning_rate": 0.00012629494505160982,
|
| 3897 |
+
"loss": 0.5469,
|
| 3898 |
+
"step": 564
|
| 3899 |
+
},
|
| 3900 |
+
{
|
| 3901 |
+
"epoch": 1.59,
|
| 3902 |
+
"learning_rate": 0.00012581536178703603,
|
| 3903 |
+
"loss": 0.2917,
|
| 3904 |
+
"step": 565
|
| 3905 |
+
},
|
| 3906 |
+
{
|
| 3907 |
+
"epoch": 1.59,
|
| 3908 |
+
"learning_rate": 0.0001253360322078731,
|
| 3909 |
+
"loss": 0.5076,
|
| 3910 |
+
"step": 566
|
| 3911 |
+
},
|
| 3912 |
+
{
|
| 3913 |
+
"epoch": 1.59,
|
| 3914 |
+
"learning_rate": 0.00012485696134206147,
|
| 3915 |
+
"loss": 0.4549,
|
| 3916 |
+
"step": 567
|
| 3917 |
+
},
|
| 3918 |
+
{
|
| 3919 |
+
"epoch": 1.59,
|
| 3920 |
+
"learning_rate": 0.0001243781542148279,
|
| 3921 |
+
"loss": 0.4924,
|
| 3922 |
+
"step": 568
|
| 3923 |
+
},
|
| 3924 |
+
{
|
| 3925 |
+
"epoch": 1.6,
|
| 3926 |
+
"learning_rate": 0.00012389961584863263,
|
| 3927 |
+
"loss": 0.4153,
|
| 3928 |
+
"step": 569
|
| 3929 |
+
},
|
| 3930 |
+
{
|
| 3931 |
+
"epoch": 1.6,
|
| 3932 |
+
"learning_rate": 0.0001234213512631166,
|
| 3933 |
+
"loss": 0.5995,
|
| 3934 |
+
"step": 570
|
| 3935 |
+
},
|
| 3936 |
+
{
|
| 3937 |
+
"epoch": 1.6,
|
| 3938 |
+
"learning_rate": 0.000122943365475049,
|
| 3939 |
+
"loss": 0.38,
|
| 3940 |
+
"step": 571
|
| 3941 |
+
},
|
| 3942 |
+
{
|
| 3943 |
+
"epoch": 1.61,
|
| 3944 |
+
"learning_rate": 0.0001224656634982746,
|
| 3945 |
+
"loss": 0.5892,
|
| 3946 |
+
"step": 572
|
| 3947 |
+
},
|
| 3948 |
+
{
|
| 3949 |
+
"epoch": 1.61,
|
| 3950 |
+
"learning_rate": 0.00012198825034366104,
|
| 3951 |
+
"loss": 0.3141,
|
| 3952 |
+
"step": 573
|
| 3953 |
+
},
|
| 3954 |
+
{
|
| 3955 |
+
"epoch": 1.61,
|
| 3956 |
+
"learning_rate": 0.00012151113101904645,
|
| 3957 |
+
"loss": 0.4494,
|
| 3958 |
+
"step": 574
|
| 3959 |
+
},
|
| 3960 |
+
{
|
| 3961 |
+
"epoch": 1.61,
|
| 3962 |
+
"learning_rate": 0.0001210343105291868,
|
| 3963 |
+
"loss": 0.4938,
|
| 3964 |
+
"step": 575
|
| 3965 |
+
},
|
| 3966 |
+
{
|
| 3967 |
+
"epoch": 1.61,
|
| 3968 |
+
"eval_loss": 0.504688560962677,
|
| 3969 |
+
"eval_runtime": 26.3824,
|
| 3970 |
+
"eval_samples_per_second": 7.581,
|
| 3971 |
+
"eval_steps_per_second": 1.895,
|
| 3972 |
+
"step": 575
|
| 3973 |
+
},
|
| 3974 |
+
{
|
| 3975 |
+
"dharma_eval_accuracy": 0.4820703510463012,
|
| 3976 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.48148148148148145,
|
| 3977 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
|
| 3978 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 3979 |
+
"dharma_eval_accuracy_MMLU": 0.42592592592592593,
|
| 3980 |
+
"dharma_eval_accuracy_agieval": 0.2542372881355932,
|
| 3981 |
+
"dharma_eval_accuracy_bigbench": 0.4918032786885246,
|
| 3982 |
+
"dharma_eval_accuracy_openbookqa": 0.2962962962962963,
|
| 3983 |
+
"dharma_eval_accuracy_truthful_qa": 0.6481481481481481,
|
| 3984 |
+
"dharma_eval_accuracy_winogrande": 0.5,
|
| 3985 |
+
"dharma_loss": 2.8464001512527464,
|
| 3986 |
+
"epoch": 1.61,
|
| 3987 |
+
"step": 575
|
| 3988 |
+
},
|
| 3989 |
+
{
|
| 3990 |
+
"epoch": 1.62,
|
| 3991 |
+
"learning_rate": 0.00012055779387570338,
|
| 3992 |
+
"loss": 0.4101,
|
| 3993 |
+
"step": 576
|
| 3994 |
+
},
|
| 3995 |
+
{
|
| 3996 |
+
"epoch": 1.62,
|
| 3997 |
+
"learning_rate": 0.00012008158605703037,
|
| 3998 |
+
"loss": 0.4743,
|
| 3999 |
+
"step": 577
|
| 4000 |
+
},
|
| 4001 |
+
{
|
| 4002 |
+
"epoch": 1.62,
|
| 4003 |
+
"learning_rate": 0.00011960569206836249,
|
| 4004 |
+
"loss": 0.5178,
|
| 4005 |
+
"step": 578
|
| 4006 |
+
},
|
| 4007 |
+
{
|
| 4008 |
+
"epoch": 1.63,
|
| 4009 |
+
"learning_rate": 0.00011913011690160246,
|
| 4010 |
+
"loss": 0.5462,
|
| 4011 |
+
"step": 579
|
| 4012 |
+
},
|
| 4013 |
+
{
|
| 4014 |
+
"epoch": 1.63,
|
| 4015 |
+
"learning_rate": 0.00011865486554530873,
|
| 4016 |
+
"loss": 0.3295,
|
| 4017 |
+
"step": 580
|
| 4018 |
+
},
|
| 4019 |
+
{
|
| 4020 |
+
"epoch": 1.63,
|
| 4021 |
+
"learning_rate": 0.00011817994298464319,
|
| 4022 |
+
"loss": 0.4082,
|
| 4023 |
+
"step": 581
|
| 4024 |
+
},
|
| 4025 |
+
{
|
| 4026 |
+
"epoch": 1.63,
|
| 4027 |
+
"learning_rate": 0.00011770535420131876,
|
| 4028 |
+
"loss": 0.495,
|
| 4029 |
+
"step": 582
|
| 4030 |
+
},
|
| 4031 |
+
{
|
| 4032 |
+
"epoch": 1.64,
|
| 4033 |
+
"learning_rate": 0.00011723110417354719,
|
| 4034 |
+
"loss": 0.5554,
|
| 4035 |
+
"step": 583
|
| 4036 |
+
},
|
| 4037 |
+
{
|
| 4038 |
+
"epoch": 1.64,
|
| 4039 |
+
"learning_rate": 0.00011675719787598688,
|
| 4040 |
+
"loss": 0.4386,
|
| 4041 |
+
"step": 584
|
| 4042 |
+
},
|
| 4043 |
+
{
|
| 4044 |
+
"epoch": 1.64,
|
| 4045 |
+
"learning_rate": 0.00011628364027969062,
|
| 4046 |
+
"loss": 0.4684,
|
| 4047 |
+
"step": 585
|
| 4048 |
+
},
|
| 4049 |
+
{
|
| 4050 |
+
"epoch": 1.64,
|
| 4051 |
+
"learning_rate": 0.00011581043635205352,
|
| 4052 |
+
"loss": 0.3877,
|
| 4053 |
+
"step": 586
|
| 4054 |
+
},
|
| 4055 |
+
{
|
| 4056 |
+
"epoch": 1.65,
|
| 4057 |
+
"learning_rate": 0.00011533759105676082,
|
| 4058 |
+
"loss": 0.6039,
|
| 4059 |
+
"step": 587
|
| 4060 |
+
},
|
| 4061 |
+
{
|
| 4062 |
+
"epoch": 1.65,
|
| 4063 |
+
"learning_rate": 0.000114865109353736,
|
| 4064 |
+
"loss": 0.4679,
|
| 4065 |
+
"step": 588
|
| 4066 |
+
},
|
| 4067 |
+
{
|
| 4068 |
+
"epoch": 1.65,
|
| 4069 |
+
"learning_rate": 0.00011439299619908854,
|
| 4070 |
+
"loss": 0.5623,
|
| 4071 |
+
"step": 589
|
| 4072 |
+
},
|
| 4073 |
+
{
|
| 4074 |
+
"epoch": 1.66,
|
| 4075 |
+
"learning_rate": 0.00011392125654506198,
|
| 4076 |
+
"loss": 0.4549,
|
| 4077 |
+
"step": 590
|
| 4078 |
+
},
|
| 4079 |
+
{
|
| 4080 |
+
"epoch": 1.66,
|
| 4081 |
+
"learning_rate": 0.00011344989533998204,
|
| 4082 |
+
"loss": 0.4881,
|
| 4083 |
+
"step": 591
|
| 4084 |
+
},
|
| 4085 |
+
{
|
| 4086 |
+
"epoch": 1.66,
|
| 4087 |
+
"learning_rate": 0.00011297891752820484,
|
| 4088 |
+
"loss": 0.483,
|
| 4089 |
+
"step": 592
|
| 4090 |
+
},
|
| 4091 |
+
{
|
| 4092 |
+
"epoch": 1.66,
|
| 4093 |
+
"learning_rate": 0.0001125083280500647,
|
| 4094 |
+
"loss": 0.6766,
|
| 4095 |
+
"step": 593
|
| 4096 |
+
},
|
| 4097 |
+
{
|
| 4098 |
+
"epoch": 1.67,
|
| 4099 |
+
"learning_rate": 0.00011203813184182255,
|
| 4100 |
+
"loss": 0.4011,
|
| 4101 |
+
"step": 594
|
| 4102 |
+
},
|
| 4103 |
+
{
|
| 4104 |
+
"epoch": 1.67,
|
| 4105 |
+
"learning_rate": 0.00011156833383561413,
|
| 4106 |
+
"loss": 0.5004,
|
| 4107 |
+
"step": 595
|
| 4108 |
+
},
|
| 4109 |
+
{
|
| 4110 |
+
"epoch": 1.67,
|
| 4111 |
+
"learning_rate": 0.00011109893895939823,
|
| 4112 |
+
"loss": 0.4021,
|
| 4113 |
+
"step": 596
|
| 4114 |
+
},
|
| 4115 |
+
{
|
| 4116 |
+
"epoch": 1.68,
|
| 4117 |
+
"learning_rate": 0.00011062995213690494,
|
| 4118 |
+
"loss": 0.6189,
|
| 4119 |
+
"step": 597
|
| 4120 |
+
},
|
| 4121 |
+
{
|
| 4122 |
+
"epoch": 1.68,
|
| 4123 |
+
"learning_rate": 0.00011016137828758422,
|
| 4124 |
+
"loss": 0.5078,
|
| 4125 |
+
"step": 598
|
| 4126 |
+
},
|
| 4127 |
+
{
|
| 4128 |
+
"epoch": 1.68,
|
| 4129 |
+
"learning_rate": 0.0001096932223265539,
|
| 4130 |
+
"loss": 0.637,
|
| 4131 |
+
"step": 599
|
| 4132 |
+
},
|
| 4133 |
+
{
|
| 4134 |
+
"epoch": 1.68,
|
| 4135 |
+
"learning_rate": 0.00010922548916454855,
|
| 4136 |
+
"loss": 0.5294,
|
| 4137 |
+
"step": 600
|
| 4138 |
+
},
|
| 4139 |
+
{
|
| 4140 |
+
"epoch": 1.68,
|
| 4141 |
+
"eval_loss": 0.5050536394119263,
|
| 4142 |
+
"eval_runtime": 26.3547,
|
| 4143 |
+
"eval_samples_per_second": 7.589,
|
| 4144 |
+
"eval_steps_per_second": 1.897,
|
| 4145 |
+
"step": 600
|
| 4146 |
+
},
|
| 4147 |
+
{
|
| 4148 |
+
"dharma_eval_accuracy": 0.4757231375427789,
|
| 4149 |
+
"dharma_eval_accuracy_ARC-Challenge": 0.46296296296296297,
|
| 4150 |
+
"dharma_eval_accuracy_ARC-Easy": 0.5370370370370371,
|
| 4151 |
+
"dharma_eval_accuracy_BoolQ": 0.7037037037037037,
|
| 4152 |
+
"dharma_eval_accuracy_MMLU": 0.42592592592592593,
|
| 4153 |
+
"dharma_eval_accuracy_agieval": 0.2711864406779661,
|
| 4154 |
+
"dharma_eval_accuracy_bigbench": 0.4918032786885246,
|
| 4155 |
+
"dharma_eval_accuracy_openbookqa": 0.2962962962962963,
|
| 4156 |
+
"dharma_eval_accuracy_truthful_qa": 0.6111111111111112,
|
| 4157 |
+
"dharma_eval_accuracy_winogrande": 0.48148148148148145,
|
| 4158 |
+
"dharma_loss": 2.8082506337165833,
|
| 4159 |
+
"epoch": 1.68,
|
| 4160 |
+
"step": 600
|
| 4161 |
}
|
| 4162 |
],
|
| 4163 |
"max_steps": 1000,
|
| 4164 |
"num_train_epochs": 3,
|
| 4165 |
+
"total_flos": 1.9307426648459674e+17,
|
| 4166 |
"trial_name": null,
|
| 4167 |
"trial_params": null
|
| 4168 |
}
|
{checkpoint-200 → checkpoint-600}/training_args.bin
RENAMED
|
File without changes
|