Add 1850000
Browse files- added_tokens.json +1 -1
- config.json +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- tokenizer_config.json +1 -1
- trainer_state.json +1124 -11
- training_args.bin +1 -1
added_tokens.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"
|
|
|
|
| 1 |
+
{"START_SPONSOR_TOKEN": 32112, "LONG_WORD_TOKEN": 32106, "BETWEEN_SEGMENTS_TOKEN": 32118, "SHORT_HYPHENATED_TOKEN": 32105, "[Laughter]": 32109, "END_SPONSOR_TOKEN": 32113, "NUMBER_PERCENTAGE_TOKEN": 32103, "NO_SEGMENT_TOKEN": 32111, "START_SELFPROMO_TOKEN": 32114, "[Applause]": 32108, "URL_TOKEN": 32101, "EXTRACT_SEGMENTS: ": 32100, "START_INTERACTION_TOKEN": 32116, "PROFANITY_TOKEN": 32110, "HYPHENATED_URL_TOKEN": 32102, "[Music]": 32107, "END_INTERACTION_TOKEN": 32117, "END_SELFPROMO_TOKEN": 32115, "NUMBER_TOKEN": 32104}
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "/1TB_SSD/SB_AI/
|
| 3 |
"architectures": [
|
| 4 |
"T5ForConditionalGeneration"
|
| 5 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "/1TB_SSD/SB_AI/out_epoch1/out/checkpoint-1115000/",
|
| 3 |
"architectures": [
|
| 4 |
"T5ForConditionalGeneration"
|
| 5 |
],
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 891703231
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afdfb877d569756c5d3e589de624b065735445d4431398f5ec538b4f3ee17e99
|
| 3 |
size 891703231
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14439
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:867ce08b496f6dc7dd44b00318be2bd40fa0c1f470100e78a8ccca16dc0eb97f
|
| 3 |
size 14439
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c476f8e287d72e82b2fe676d2e42ba4bc778d9bb83ac44801844f0c68ef1e65d
|
| 3 |
size 623
|
tokenizer_config.json
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 100, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"], "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "/1TB_SSD/SB_AI/
|
|
|
|
| 1 |
+
{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 100, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"], "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "/1TB_SSD/SB_AI/out_epoch1/out/checkpoint-1115000/", "tokenizer_class": "T5Tokenizer"}
|
trainer_state.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 1.
|
| 5 |
-
"global_step":
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
@@ -1459,17 +1459,1130 @@
|
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 1.0,
|
| 1462 |
-
"
|
| 1463 |
-
"
|
| 1464 |
-
"
|
| 1465 |
-
|
| 1466 |
-
|
| 1467 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1468 |
}
|
| 1469 |
],
|
| 1470 |
-
"max_steps":
|
| 1471 |
-
"num_train_epochs":
|
| 1472 |
-
"total_flos":
|
| 1473 |
"trial_name": null,
|
| 1474 |
"trial_params": null
|
| 1475 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.656824235129331,
|
| 5 |
+
"global_step": 1850000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
|
|
|
| 1459 |
},
|
| 1460 |
{
|
| 1461 |
"epoch": 1.0,
|
| 1462 |
+
"learning_rate": 2.492374130615067e-05,
|
| 1463 |
+
"loss": 0.0633,
|
| 1464 |
+
"step": 1120000
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 1.01,
|
| 1468 |
+
"learning_rate": 2.4811793722695988e-05,
|
| 1469 |
+
"loss": 0.0625,
|
| 1470 |
+
"step": 1125000
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 1.01,
|
| 1474 |
+
"eval_loss": 0.06666136533021927,
|
| 1475 |
+
"eval_runtime": 1682.1519,
|
| 1476 |
+
"eval_samples_per_second": 36.877,
|
| 1477 |
+
"eval_steps_per_second": 36.877,
|
| 1478 |
+
"step": 1125000
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 1.01,
|
| 1482 |
+
"learning_rate": 2.46998461392413e-05,
|
| 1483 |
+
"loss": 0.0645,
|
| 1484 |
+
"step": 1130000
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 1.02,
|
| 1488 |
+
"learning_rate": 2.4587898555786618e-05,
|
| 1489 |
+
"loss": 0.058,
|
| 1490 |
+
"step": 1135000
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.02,
|
| 1494 |
+
"learning_rate": 2.4475950972331933e-05,
|
| 1495 |
+
"loss": 0.0644,
|
| 1496 |
+
"step": 1140000
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 1.03,
|
| 1500 |
+
"learning_rate": 2.4364003388877245e-05,
|
| 1501 |
+
"loss": 0.0599,
|
| 1502 |
+
"step": 1145000
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.03,
|
| 1506 |
+
"learning_rate": 2.4252055805422564e-05,
|
| 1507 |
+
"loss": 0.0614,
|
| 1508 |
+
"step": 1150000
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.03,
|
| 1512 |
+
"eval_loss": 0.06576403230428696,
|
| 1513 |
+
"eval_runtime": 1686.8754,
|
| 1514 |
+
"eval_samples_per_second": 36.774,
|
| 1515 |
+
"eval_steps_per_second": 36.774,
|
| 1516 |
+
"step": 1150000
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.03,
|
| 1520 |
+
"learning_rate": 2.414010822196788e-05,
|
| 1521 |
+
"loss": 0.0637,
|
| 1522 |
+
"step": 1155000
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 1.04,
|
| 1526 |
+
"learning_rate": 2.4028160638513194e-05,
|
| 1527 |
+
"loss": 0.058,
|
| 1528 |
+
"step": 1160000
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 1.04,
|
| 1532 |
+
"learning_rate": 2.391621305505851e-05,
|
| 1533 |
+
"loss": 0.0623,
|
| 1534 |
+
"step": 1165000
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 1.05,
|
| 1538 |
+
"learning_rate": 2.3804265471603828e-05,
|
| 1539 |
+
"loss": 0.0628,
|
| 1540 |
+
"step": 1170000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.05,
|
| 1544 |
+
"learning_rate": 2.369231788814914e-05,
|
| 1545 |
+
"loss": 0.0597,
|
| 1546 |
+
"step": 1175000
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 1.05,
|
| 1550 |
+
"eval_loss": 0.06830067932605743,
|
| 1551 |
+
"eval_runtime": 1683.3372,
|
| 1552 |
+
"eval_samples_per_second": 36.851,
|
| 1553 |
+
"eval_steps_per_second": 36.851,
|
| 1554 |
+
"step": 1175000
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.06,
|
| 1558 |
+
"learning_rate": 2.3580370304694455e-05,
|
| 1559 |
+
"loss": 0.0622,
|
| 1560 |
+
"step": 1180000
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.06,
|
| 1564 |
+
"learning_rate": 2.3468422721239773e-05,
|
| 1565 |
+
"loss": 0.0579,
|
| 1566 |
+
"step": 1185000
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 1.07,
|
| 1570 |
+
"learning_rate": 2.3356475137785085e-05,
|
| 1571 |
+
"loss": 0.0644,
|
| 1572 |
+
"step": 1190000
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 1.07,
|
| 1576 |
+
"learning_rate": 2.3244527554330404e-05,
|
| 1577 |
+
"loss": 0.063,
|
| 1578 |
+
"step": 1195000
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 1.07,
|
| 1582 |
+
"learning_rate": 2.313257997087572e-05,
|
| 1583 |
+
"loss": 0.0629,
|
| 1584 |
+
"step": 1200000
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 1.07,
|
| 1588 |
+
"eval_loss": 0.0691303089261055,
|
| 1589 |
+
"eval_runtime": 1683.681,
|
| 1590 |
+
"eval_samples_per_second": 36.844,
|
| 1591 |
+
"eval_steps_per_second": 36.844,
|
| 1592 |
+
"step": 1200000
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 1.08,
|
| 1596 |
+
"learning_rate": 2.3020632387421034e-05,
|
| 1597 |
+
"loss": 0.0647,
|
| 1598 |
+
"step": 1205000
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.08,
|
| 1602 |
+
"learning_rate": 2.290868480396635e-05,
|
| 1603 |
+
"loss": 0.0645,
|
| 1604 |
+
"step": 1210000
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 1.09,
|
| 1608 |
+
"learning_rate": 2.2796737220511664e-05,
|
| 1609 |
+
"loss": 0.063,
|
| 1610 |
+
"step": 1215000
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 1.09,
|
| 1614 |
+
"learning_rate": 2.268478963705698e-05,
|
| 1615 |
+
"loss": 0.0638,
|
| 1616 |
+
"step": 1220000
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.1,
|
| 1620 |
+
"learning_rate": 2.2572842053602295e-05,
|
| 1621 |
+
"loss": 0.0603,
|
| 1622 |
+
"step": 1225000
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.1,
|
| 1626 |
+
"eval_loss": 0.06777703762054443,
|
| 1627 |
+
"eval_runtime": 1680.6968,
|
| 1628 |
+
"eval_samples_per_second": 36.909,
|
| 1629 |
+
"eval_steps_per_second": 36.909,
|
| 1630 |
+
"step": 1225000
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.1,
|
| 1634 |
+
"learning_rate": 2.246089447014761e-05,
|
| 1635 |
+
"loss": 0.0612,
|
| 1636 |
+
"step": 1230000
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 1.11,
|
| 1640 |
+
"learning_rate": 2.2348946886692925e-05,
|
| 1641 |
+
"loss": 0.0586,
|
| 1642 |
+
"step": 1235000
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 1.11,
|
| 1646 |
+
"learning_rate": 2.2236999303238244e-05,
|
| 1647 |
+
"loss": 0.0607,
|
| 1648 |
+
"step": 1240000
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 1.11,
|
| 1652 |
+
"learning_rate": 2.212505171978356e-05,
|
| 1653 |
+
"loss": 0.0653,
|
| 1654 |
+
"step": 1245000
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 1.12,
|
| 1658 |
+
"learning_rate": 2.201310413632887e-05,
|
| 1659 |
+
"loss": 0.0601,
|
| 1660 |
+
"step": 1250000
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 1.12,
|
| 1664 |
+
"eval_loss": 0.07459407299757004,
|
| 1665 |
+
"eval_runtime": 1683.3206,
|
| 1666 |
+
"eval_samples_per_second": 36.852,
|
| 1667 |
+
"eval_steps_per_second": 36.852,
|
| 1668 |
+
"step": 1250000
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 1.12,
|
| 1672 |
+
"learning_rate": 2.190115655287419e-05,
|
| 1673 |
+
"loss": 0.0638,
|
| 1674 |
+
"step": 1255000
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 1.13,
|
| 1678 |
+
"learning_rate": 2.1789208969419504e-05,
|
| 1679 |
+
"loss": 0.0634,
|
| 1680 |
+
"step": 1260000
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 1.13,
|
| 1684 |
+
"learning_rate": 2.167726138596482e-05,
|
| 1685 |
+
"loss": 0.0622,
|
| 1686 |
+
"step": 1265000
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.14,
|
| 1690 |
+
"learning_rate": 2.1565313802510135e-05,
|
| 1691 |
+
"loss": 0.0593,
|
| 1692 |
+
"step": 1270000
|
| 1693 |
+
},
|
| 1694 |
+
{
|
| 1695 |
+
"epoch": 1.14,
|
| 1696 |
+
"learning_rate": 2.145336621905545e-05,
|
| 1697 |
+
"loss": 0.0606,
|
| 1698 |
+
"step": 1275000
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 1.14,
|
| 1702 |
+
"eval_loss": 0.06908420473337173,
|
| 1703 |
+
"eval_runtime": 1680.5541,
|
| 1704 |
+
"eval_samples_per_second": 36.912,
|
| 1705 |
+
"eval_steps_per_second": 36.912,
|
| 1706 |
+
"step": 1275000
|
| 1707 |
+
},
|
| 1708 |
+
{
|
| 1709 |
+
"epoch": 1.15,
|
| 1710 |
+
"learning_rate": 2.1341418635600765e-05,
|
| 1711 |
+
"loss": 0.0612,
|
| 1712 |
+
"step": 1280000
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 1.15,
|
| 1716 |
+
"learning_rate": 2.122947105214608e-05,
|
| 1717 |
+
"loss": 0.066,
|
| 1718 |
+
"step": 1285000
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 1.16,
|
| 1722 |
+
"learning_rate": 2.1117523468691396e-05,
|
| 1723 |
+
"loss": 0.0562,
|
| 1724 |
+
"step": 1290000
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 1.16,
|
| 1728 |
+
"learning_rate": 2.100557588523671e-05,
|
| 1729 |
+
"loss": 0.0617,
|
| 1730 |
+
"step": 1295000
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 1.16,
|
| 1734 |
+
"learning_rate": 2.089362830178203e-05,
|
| 1735 |
+
"loss": 0.0671,
|
| 1736 |
+
"step": 1300000
|
| 1737 |
+
},
|
| 1738 |
+
{
|
| 1739 |
+
"epoch": 1.16,
|
| 1740 |
+
"eval_loss": 0.07024173438549042,
|
| 1741 |
+
"eval_runtime": 1683.3132,
|
| 1742 |
+
"eval_samples_per_second": 36.852,
|
| 1743 |
+
"eval_steps_per_second": 36.852,
|
| 1744 |
+
"step": 1300000
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 1.17,
|
| 1748 |
+
"learning_rate": 2.078168071832734e-05,
|
| 1749 |
+
"loss": 0.0578,
|
| 1750 |
+
"step": 1305000
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 1.17,
|
| 1754 |
+
"learning_rate": 2.066973313487266e-05,
|
| 1755 |
+
"loss": 0.0592,
|
| 1756 |
+
"step": 1310000
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 1.18,
|
| 1760 |
+
"learning_rate": 2.0557785551417975e-05,
|
| 1761 |
+
"loss": 0.0607,
|
| 1762 |
+
"step": 1315000
|
| 1763 |
+
},
|
| 1764 |
+
{
|
| 1765 |
+
"epoch": 1.18,
|
| 1766 |
+
"learning_rate": 2.0445837967963287e-05,
|
| 1767 |
+
"loss": 0.0645,
|
| 1768 |
+
"step": 1320000
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 1.19,
|
| 1772 |
+
"learning_rate": 2.0333890384508605e-05,
|
| 1773 |
+
"loss": 0.0625,
|
| 1774 |
+
"step": 1325000
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.19,
|
| 1778 |
+
"eval_loss": 0.06607282906770706,
|
| 1779 |
+
"eval_runtime": 1680.5346,
|
| 1780 |
+
"eval_samples_per_second": 36.913,
|
| 1781 |
+
"eval_steps_per_second": 36.913,
|
| 1782 |
+
"step": 1325000
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 1.19,
|
| 1786 |
+
"learning_rate": 2.022194280105392e-05,
|
| 1787 |
+
"loss": 0.0625,
|
| 1788 |
+
"step": 1330000
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.2,
|
| 1792 |
+
"learning_rate": 2.0109995217599236e-05,
|
| 1793 |
+
"loss": 0.0605,
|
| 1794 |
+
"step": 1335000
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 1.2,
|
| 1798 |
+
"learning_rate": 1.999804763414455e-05,
|
| 1799 |
+
"loss": 0.0592,
|
| 1800 |
+
"step": 1340000
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 1.2,
|
| 1804 |
+
"learning_rate": 1.9886100050689866e-05,
|
| 1805 |
+
"loss": 0.0652,
|
| 1806 |
+
"step": 1345000
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 1.21,
|
| 1810 |
+
"learning_rate": 1.977415246723518e-05,
|
| 1811 |
+
"loss": 0.0617,
|
| 1812 |
+
"step": 1350000
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 1.21,
|
| 1816 |
+
"eval_loss": 0.0687505304813385,
|
| 1817 |
+
"eval_runtime": 1680.1242,
|
| 1818 |
+
"eval_samples_per_second": 36.922,
|
| 1819 |
+
"eval_steps_per_second": 36.922,
|
| 1820 |
+
"step": 1350000
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 1.21,
|
| 1824 |
+
"learning_rate": 1.9662204883780496e-05,
|
| 1825 |
+
"loss": 0.0607,
|
| 1826 |
+
"step": 1355000
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 1.22,
|
| 1830 |
+
"learning_rate": 1.9550257300325815e-05,
|
| 1831 |
+
"loss": 0.0619,
|
| 1832 |
+
"step": 1360000
|
| 1833 |
+
},
|
| 1834 |
+
{
|
| 1835 |
+
"epoch": 1.22,
|
| 1836 |
+
"learning_rate": 1.9438309716871127e-05,
|
| 1837 |
+
"loss": 0.0629,
|
| 1838 |
+
"step": 1365000
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 1.23,
|
| 1842 |
+
"learning_rate": 1.9326362133416445e-05,
|
| 1843 |
+
"loss": 0.0637,
|
| 1844 |
+
"step": 1370000
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 1.23,
|
| 1848 |
+
"learning_rate": 1.921441454996176e-05,
|
| 1849 |
+
"loss": 0.0579,
|
| 1850 |
+
"step": 1375000
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 1.23,
|
| 1854 |
+
"eval_loss": 0.06793049722909927,
|
| 1855 |
+
"eval_runtime": 1680.8053,
|
| 1856 |
+
"eval_samples_per_second": 36.907,
|
| 1857 |
+
"eval_steps_per_second": 36.907,
|
| 1858 |
+
"step": 1375000
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 1.24,
|
| 1862 |
+
"learning_rate": 1.9102466966507072e-05,
|
| 1863 |
+
"loss": 0.0632,
|
| 1864 |
+
"step": 1380000
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 1.24,
|
| 1868 |
+
"learning_rate": 1.899051938305239e-05,
|
| 1869 |
+
"loss": 0.0593,
|
| 1870 |
+
"step": 1385000
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 1.24,
|
| 1874 |
+
"learning_rate": 1.8878571799597706e-05,
|
| 1875 |
+
"loss": 0.0622,
|
| 1876 |
+
"step": 1390000
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"epoch": 1.25,
|
| 1880 |
+
"learning_rate": 1.876662421614302e-05,
|
| 1881 |
+
"loss": 0.064,
|
| 1882 |
+
"step": 1395000
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 1.25,
|
| 1886 |
+
"learning_rate": 1.8654676632688336e-05,
|
| 1887 |
+
"loss": 0.0663,
|
| 1888 |
+
"step": 1400000
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 1.25,
|
| 1892 |
+
"eval_loss": 0.0633900910615921,
|
| 1893 |
+
"eval_runtime": 1680.7989,
|
| 1894 |
+
"eval_samples_per_second": 36.907,
|
| 1895 |
+
"eval_steps_per_second": 36.907,
|
| 1896 |
+
"step": 1400000
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 1.26,
|
| 1900 |
+
"learning_rate": 1.8542729049233655e-05,
|
| 1901 |
+
"loss": 0.0606,
|
| 1902 |
+
"step": 1405000
|
| 1903 |
+
},
|
| 1904 |
+
{
|
| 1905 |
+
"epoch": 1.26,
|
| 1906 |
+
"learning_rate": 1.8430781465778967e-05,
|
| 1907 |
+
"loss": 0.0633,
|
| 1908 |
+
"step": 1410000
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 1.27,
|
| 1912 |
+
"learning_rate": 1.8318833882324282e-05,
|
| 1913 |
+
"loss": 0.064,
|
| 1914 |
+
"step": 1415000
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 1.27,
|
| 1918 |
+
"learning_rate": 1.82068862988696e-05,
|
| 1919 |
+
"loss": 0.0573,
|
| 1920 |
+
"step": 1420000
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 1.28,
|
| 1924 |
+
"learning_rate": 1.8094938715414912e-05,
|
| 1925 |
+
"loss": 0.0583,
|
| 1926 |
+
"step": 1425000
|
| 1927 |
+
},
|
| 1928 |
+
{
|
| 1929 |
+
"epoch": 1.28,
|
| 1930 |
+
"eval_loss": 0.063847616314888,
|
| 1931 |
+
"eval_runtime": 1680.6703,
|
| 1932 |
+
"eval_samples_per_second": 36.91,
|
| 1933 |
+
"eval_steps_per_second": 36.91,
|
| 1934 |
+
"step": 1425000
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 1.28,
|
| 1938 |
+
"learning_rate": 1.798299113196023e-05,
|
| 1939 |
+
"loss": 0.0669,
|
| 1940 |
+
"step": 1430000
|
| 1941 |
+
},
|
| 1942 |
+
{
|
| 1943 |
+
"epoch": 1.29,
|
| 1944 |
+
"learning_rate": 1.7871043548505546e-05,
|
| 1945 |
+
"loss": 0.0576,
|
| 1946 |
+
"step": 1435000
|
| 1947 |
+
},
|
| 1948 |
+
{
|
| 1949 |
+
"epoch": 1.29,
|
| 1950 |
+
"learning_rate": 1.775909596505086e-05,
|
| 1951 |
+
"loss": 0.0603,
|
| 1952 |
+
"step": 1440000
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 1.29,
|
| 1956 |
+
"learning_rate": 1.7647148381596177e-05,
|
| 1957 |
+
"loss": 0.0628,
|
| 1958 |
+
"step": 1445000
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 1.3,
|
| 1962 |
+
"learning_rate": 1.7535200798141492e-05,
|
| 1963 |
+
"loss": 0.0623,
|
| 1964 |
+
"step": 1450000
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 1.3,
|
| 1968 |
+
"eval_loss": 0.06811905652284622,
|
| 1969 |
+
"eval_runtime": 1680.0308,
|
| 1970 |
+
"eval_samples_per_second": 36.924,
|
| 1971 |
+
"eval_steps_per_second": 36.924,
|
| 1972 |
+
"step": 1450000
|
| 1973 |
+
},
|
| 1974 |
+
{
|
| 1975 |
+
"epoch": 1.3,
|
| 1976 |
+
"learning_rate": 1.7423253214686807e-05,
|
| 1977 |
+
"loss": 0.066,
|
| 1978 |
+
"step": 1455000
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 1.31,
|
| 1982 |
+
"learning_rate": 1.7311305631232122e-05,
|
| 1983 |
+
"loss": 0.0645,
|
| 1984 |
+
"step": 1460000
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 1.31,
|
| 1988 |
+
"learning_rate": 1.719935804777744e-05,
|
| 1989 |
+
"loss": 0.0673,
|
| 1990 |
+
"step": 1465000
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.32,
|
| 1994 |
+
"learning_rate": 1.7087410464322753e-05,
|
| 1995 |
+
"loss": 0.0601,
|
| 1996 |
+
"step": 1470000
|
| 1997 |
+
},
|
| 1998 |
+
{
|
| 1999 |
+
"epoch": 1.32,
|
| 2000 |
+
"learning_rate": 1.697546288086807e-05,
|
| 2001 |
+
"loss": 0.0615,
|
| 2002 |
+
"step": 1475000
|
| 2003 |
+
},
|
| 2004 |
+
{
|
| 2005 |
+
"epoch": 1.32,
|
| 2006 |
+
"eval_loss": 0.06700597703456879,
|
| 2007 |
+
"eval_runtime": 1680.6996,
|
| 2008 |
+
"eval_samples_per_second": 36.909,
|
| 2009 |
+
"eval_steps_per_second": 36.909,
|
| 2010 |
+
"step": 1475000
|
| 2011 |
+
},
|
| 2012 |
+
{
|
| 2013 |
+
"epoch": 1.33,
|
| 2014 |
+
"learning_rate": 1.6863515297413386e-05,
|
| 2015 |
+
"loss": 0.0651,
|
| 2016 |
+
"step": 1480000
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 1.33,
|
| 2020 |
+
"learning_rate": 1.6751567713958698e-05,
|
| 2021 |
+
"loss": 0.0596,
|
| 2022 |
+
"step": 1485000
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 1.33,
|
| 2026 |
+
"learning_rate": 1.6639620130504017e-05,
|
| 2027 |
+
"loss": 0.0616,
|
| 2028 |
+
"step": 1490000
|
| 2029 |
+
},
|
| 2030 |
+
{
|
| 2031 |
+
"epoch": 1.34,
|
| 2032 |
+
"learning_rate": 1.6527672547049332e-05,
|
| 2033 |
+
"loss": 0.0609,
|
| 2034 |
+
"step": 1495000
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 1.34,
|
| 2038 |
+
"learning_rate": 1.6415724963594647e-05,
|
| 2039 |
+
"loss": 0.0592,
|
| 2040 |
+
"step": 1500000
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.34,
|
| 2044 |
+
"eval_loss": 0.06664443016052246,
|
| 2045 |
+
"eval_runtime": 1681.9546,
|
| 2046 |
+
"eval_samples_per_second": 36.881,
|
| 2047 |
+
"eval_steps_per_second": 36.881,
|
| 2048 |
+
"step": 1500000
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 1.35,
|
| 2052 |
+
"learning_rate": 1.6303777380139962e-05,
|
| 2053 |
+
"loss": 0.0618,
|
| 2054 |
+
"step": 1505000
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.35,
|
| 2058 |
+
"learning_rate": 1.6191829796685277e-05,
|
| 2059 |
+
"loss": 0.0586,
|
| 2060 |
+
"step": 1510000
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 1.36,
|
| 2064 |
+
"learning_rate": 1.6079882213230593e-05,
|
| 2065 |
+
"loss": 0.0611,
|
| 2066 |
+
"step": 1515000
|
| 2067 |
+
},
|
| 2068 |
+
{
|
| 2069 |
+
"epoch": 1.36,
|
| 2070 |
+
"learning_rate": 1.5967934629775908e-05,
|
| 2071 |
+
"loss": 0.0635,
|
| 2072 |
+
"step": 1520000
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 1.37,
|
| 2076 |
+
"learning_rate": 1.5855987046321223e-05,
|
| 2077 |
+
"loss": 0.0626,
|
| 2078 |
+
"step": 1525000
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 1.37,
|
| 2082 |
+
"eval_loss": 0.06663180142641068,
|
| 2083 |
+
"eval_runtime": 1679.9392,
|
| 2084 |
+
"eval_samples_per_second": 36.926,
|
| 2085 |
+
"eval_steps_per_second": 36.926,
|
| 2086 |
+
"step": 1525000
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 1.37,
|
| 2090 |
+
"learning_rate": 1.5744039462866538e-05,
|
| 2091 |
+
"loss": 0.0637,
|
| 2092 |
+
"step": 1530000
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 1.37,
|
| 2096 |
+
"learning_rate": 1.5632091879411857e-05,
|
| 2097 |
+
"loss": 0.0582,
|
| 2098 |
+
"step": 1535000
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 1.38,
|
| 2102 |
+
"learning_rate": 1.5520144295957172e-05,
|
| 2103 |
+
"loss": 0.0618,
|
| 2104 |
+
"step": 1540000
|
| 2105 |
+
},
|
| 2106 |
+
{
|
| 2107 |
+
"epoch": 1.38,
|
| 2108 |
+
"learning_rate": 1.5408196712502487e-05,
|
| 2109 |
+
"loss": 0.0594,
|
| 2110 |
+
"step": 1545000
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.39,
|
| 2114 |
+
"learning_rate": 1.5296249129047802e-05,
|
| 2115 |
+
"loss": 0.063,
|
| 2116 |
+
"step": 1550000
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 1.39,
|
| 2120 |
+
"eval_loss": 0.06467917561531067,
|
| 2121 |
+
"eval_runtime": 1681.2275,
|
| 2122 |
+
"eval_samples_per_second": 36.897,
|
| 2123 |
+
"eval_steps_per_second": 36.897,
|
| 2124 |
+
"step": 1550000
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 1.39,
|
| 2128 |
+
"learning_rate": 1.5184301545593116e-05,
|
| 2129 |
+
"loss": 0.0637,
|
| 2130 |
+
"step": 1555000
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 1.4,
|
| 2134 |
+
"learning_rate": 1.5072353962138433e-05,
|
| 2135 |
+
"loss": 0.0641,
|
| 2136 |
+
"step": 1560000
|
| 2137 |
+
},
|
| 2138 |
+
{
|
| 2139 |
+
"epoch": 1.4,
|
| 2140 |
+
"learning_rate": 1.4960406378683748e-05,
|
| 2141 |
+
"loss": 0.0615,
|
| 2142 |
+
"step": 1565000
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 1.41,
|
| 2146 |
+
"learning_rate": 1.4848458795229065e-05,
|
| 2147 |
+
"loss": 0.061,
|
| 2148 |
+
"step": 1570000
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 1.41,
|
| 2152 |
+
"learning_rate": 1.4736511211774378e-05,
|
| 2153 |
+
"loss": 0.0648,
|
| 2154 |
+
"step": 1575000
|
| 2155 |
+
},
|
| 2156 |
+
{
|
| 2157 |
+
"epoch": 1.41,
|
| 2158 |
+
"eval_loss": 0.06533137708902359,
|
| 2159 |
+
"eval_runtime": 1677.1597,
|
| 2160 |
+
"eval_samples_per_second": 36.987,
|
| 2161 |
+
"eval_steps_per_second": 36.987,
|
| 2162 |
+
"step": 1575000
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 1.42,
|
| 2166 |
+
"learning_rate": 1.4624563628319693e-05,
|
| 2167 |
+
"loss": 0.0623,
|
| 2168 |
+
"step": 1580000
|
| 2169 |
+
},
|
| 2170 |
+
{
|
| 2171 |
+
"epoch": 1.42,
|
| 2172 |
+
"learning_rate": 1.451261604486501e-05,
|
| 2173 |
+
"loss": 0.0631,
|
| 2174 |
+
"step": 1585000
|
| 2175 |
+
},
|
| 2176 |
+
{
|
| 2177 |
+
"epoch": 1.42,
|
| 2178 |
+
"learning_rate": 1.4400668461410324e-05,
|
| 2179 |
+
"loss": 0.0613,
|
| 2180 |
+
"step": 1590000
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 1.43,
|
| 2184 |
+
"learning_rate": 1.428872087795564e-05,
|
| 2185 |
+
"loss": 0.0611,
|
| 2186 |
+
"step": 1595000
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 1.43,
|
| 2190 |
+
"learning_rate": 1.4176773294500956e-05,
|
| 2191 |
+
"loss": 0.0611,
|
| 2192 |
+
"step": 1600000
|
| 2193 |
+
},
|
| 2194 |
+
{
|
| 2195 |
+
"epoch": 1.43,
|
| 2196 |
+
"eval_loss": 0.06996028870344162,
|
| 2197 |
+
"eval_runtime": 1678.7591,
|
| 2198 |
+
"eval_samples_per_second": 36.952,
|
| 2199 |
+
"eval_steps_per_second": 36.952,
|
| 2200 |
+
"step": 1600000
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.44,
|
| 2204 |
+
"learning_rate": 1.4064825711046273e-05,
|
| 2205 |
+
"loss": 0.0629,
|
| 2206 |
+
"step": 1605000
|
| 2207 |
+
},
|
| 2208 |
+
{
|
| 2209 |
+
"epoch": 1.44,
|
| 2210 |
+
"learning_rate": 1.3952878127591586e-05,
|
| 2211 |
+
"loss": 0.0607,
|
| 2212 |
+
"step": 1610000
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 1.45,
|
| 2216 |
+
"learning_rate": 1.3840930544136901e-05,
|
| 2217 |
+
"loss": 0.0653,
|
| 2218 |
+
"step": 1615000
|
| 2219 |
+
},
|
| 2220 |
+
{
|
| 2221 |
+
"epoch": 1.45,
|
| 2222 |
+
"learning_rate": 1.3728982960682218e-05,
|
| 2223 |
+
"loss": 0.0608,
|
| 2224 |
+
"step": 1620000
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"epoch": 1.46,
|
| 2228 |
+
"learning_rate": 1.3617035377227534e-05,
|
| 2229 |
+
"loss": 0.0622,
|
| 2230 |
+
"step": 1625000
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 1.46,
|
| 2234 |
+
"eval_loss": 0.0634424015879631,
|
| 2235 |
+
"eval_runtime": 1677.9992,
|
| 2236 |
+
"eval_samples_per_second": 36.968,
|
| 2237 |
+
"eval_steps_per_second": 36.968,
|
| 2238 |
+
"step": 1625000
|
| 2239 |
+
},
|
| 2240 |
+
{
|
| 2241 |
+
"epoch": 1.46,
|
| 2242 |
+
"learning_rate": 1.350508779377285e-05,
|
| 2243 |
+
"loss": 0.0622,
|
| 2244 |
+
"step": 1630000
|
| 2245 |
+
},
|
| 2246 |
+
{
|
| 2247 |
+
"epoch": 1.46,
|
| 2248 |
+
"learning_rate": 1.3393140210318164e-05,
|
| 2249 |
+
"loss": 0.0634,
|
| 2250 |
+
"step": 1635000
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 1.47,
|
| 2254 |
+
"learning_rate": 1.328119262686348e-05,
|
| 2255 |
+
"loss": 0.059,
|
| 2256 |
+
"step": 1640000
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.47,
|
| 2260 |
+
"learning_rate": 1.3169245043408796e-05,
|
| 2261 |
+
"loss": 0.0608,
|
| 2262 |
+
"step": 1645000
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 1.48,
|
| 2266 |
+
"learning_rate": 1.305729745995411e-05,
|
| 2267 |
+
"loss": 0.0617,
|
| 2268 |
+
"step": 1650000
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 1.48,
|
| 2272 |
+
"eval_loss": 0.06513845920562744,
|
| 2273 |
+
"eval_runtime": 1682.3756,
|
| 2274 |
+
"eval_samples_per_second": 36.872,
|
| 2275 |
+
"eval_steps_per_second": 36.872,
|
| 2276 |
+
"step": 1650000
|
| 2277 |
+
},
|
| 2278 |
+
{
|
| 2279 |
+
"epoch": 1.48,
|
| 2280 |
+
"learning_rate": 1.2945349876499426e-05,
|
| 2281 |
+
"loss": 0.0622,
|
| 2282 |
+
"step": 1655000
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 1.49,
|
| 2286 |
+
"learning_rate": 1.2833402293044742e-05,
|
| 2287 |
+
"loss": 0.0603,
|
| 2288 |
+
"step": 1660000
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 1.49,
|
| 2292 |
+
"learning_rate": 1.2721454709590058e-05,
|
| 2293 |
+
"loss": 0.0647,
|
| 2294 |
+
"step": 1665000
|
| 2295 |
+
},
|
| 2296 |
+
{
|
| 2297 |
+
"epoch": 1.5,
|
| 2298 |
+
"learning_rate": 1.2609507126135372e-05,
|
| 2299 |
+
"loss": 0.0579,
|
| 2300 |
+
"step": 1670000
|
| 2301 |
+
},
|
| 2302 |
+
{
|
| 2303 |
+
"epoch": 1.5,
|
| 2304 |
+
"learning_rate": 1.2497559542680689e-05,
|
| 2305 |
+
"loss": 0.0613,
|
| 2306 |
+
"step": 1675000
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 1.5,
|
| 2310 |
+
"eval_loss": 0.0634496882557869,
|
| 2311 |
+
"eval_runtime": 1679.6133,
|
| 2312 |
+
"eval_samples_per_second": 36.933,
|
| 2313 |
+
"eval_steps_per_second": 36.933,
|
| 2314 |
+
"step": 1675000
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 1.5,
|
| 2318 |
+
"learning_rate": 1.2385611959226004e-05,
|
| 2319 |
+
"loss": 0.0574,
|
| 2320 |
+
"step": 1680000
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 1.51,
|
| 2324 |
+
"learning_rate": 1.227366437577132e-05,
|
| 2325 |
+
"loss": 0.0574,
|
| 2326 |
+
"step": 1685000
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 1.51,
|
| 2330 |
+
"learning_rate": 1.2161716792316636e-05,
|
| 2331 |
+
"loss": 0.0616,
|
| 2332 |
+
"step": 1690000
|
| 2333 |
+
},
|
| 2334 |
+
{
|
| 2335 |
+
"epoch": 1.52,
|
| 2336 |
+
"learning_rate": 1.2049769208861951e-05,
|
| 2337 |
+
"loss": 0.06,
|
| 2338 |
+
"step": 1695000
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 1.52,
|
| 2342 |
+
"learning_rate": 1.1937821625407265e-05,
|
| 2343 |
+
"loss": 0.0639,
|
| 2344 |
+
"step": 1700000
|
| 2345 |
+
},
|
| 2346 |
+
{
|
| 2347 |
+
"epoch": 1.52,
|
| 2348 |
+
"eval_loss": 0.0661427304148674,
|
| 2349 |
+
"eval_runtime": 1685.0546,
|
| 2350 |
+
"eval_samples_per_second": 36.814,
|
| 2351 |
+
"eval_steps_per_second": 36.814,
|
| 2352 |
+
"step": 1700000
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 1.53,
|
| 2356 |
+
"learning_rate": 1.1825874041952582e-05,
|
| 2357 |
+
"loss": 0.0638,
|
| 2358 |
+
"step": 1705000
|
| 2359 |
+
},
|
| 2360 |
+
{
|
| 2361 |
+
"epoch": 1.53,
|
| 2362 |
+
"learning_rate": 1.1713926458497897e-05,
|
| 2363 |
+
"loss": 0.0618,
|
| 2364 |
+
"step": 1710000
|
| 2365 |
+
},
|
| 2366 |
+
{
|
| 2367 |
+
"epoch": 1.54,
|
| 2368 |
+
"learning_rate": 1.1601978875043212e-05,
|
| 2369 |
+
"loss": 0.0616,
|
| 2370 |
+
"step": 1715000
|
| 2371 |
+
},
|
| 2372 |
+
{
|
| 2373 |
+
"epoch": 1.54,
|
| 2374 |
+
"learning_rate": 1.1490031291588529e-05,
|
| 2375 |
+
"loss": 0.0597,
|
| 2376 |
+
"step": 1720000
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 1.54,
|
| 2380 |
+
"learning_rate": 1.1378083708133844e-05,
|
| 2381 |
+
"loss": 0.0615,
|
| 2382 |
+
"step": 1725000
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 1.54,
|
| 2386 |
+
"eval_loss": 0.06442756205797195,
|
| 2387 |
+
"eval_runtime": 1688.2768,
|
| 2388 |
+
"eval_samples_per_second": 36.743,
|
| 2389 |
+
"eval_steps_per_second": 36.743,
|
| 2390 |
+
"step": 1725000
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 1.55,
|
| 2394 |
+
"learning_rate": 1.126613612467916e-05,
|
| 2395 |
+
"loss": 0.0627,
|
| 2396 |
+
"step": 1730000
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 1.55,
|
| 2400 |
+
"learning_rate": 1.1154188541224474e-05,
|
| 2401 |
+
"loss": 0.0599,
|
| 2402 |
+
"step": 1735000
|
| 2403 |
+
},
|
| 2404 |
+
{
|
| 2405 |
+
"epoch": 1.56,
|
| 2406 |
+
"learning_rate": 1.104224095776979e-05,
|
| 2407 |
+
"loss": 0.0609,
|
| 2408 |
+
"step": 1740000
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 1.56,
|
| 2412 |
+
"learning_rate": 1.0930293374315105e-05,
|
| 2413 |
+
"loss": 0.067,
|
| 2414 |
+
"step": 1745000
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 1.57,
|
| 2418 |
+
"learning_rate": 1.081834579086042e-05,
|
| 2419 |
+
"loss": 0.0605,
|
| 2420 |
+
"step": 1750000
|
| 2421 |
+
},
|
| 2422 |
+
{
|
| 2423 |
+
"epoch": 1.57,
|
| 2424 |
+
"eval_loss": 0.06615401804447174,
|
| 2425 |
+
"eval_runtime": 1682.0319,
|
| 2426 |
+
"eval_samples_per_second": 36.88,
|
| 2427 |
+
"eval_steps_per_second": 36.88,
|
| 2428 |
+
"step": 1750000
|
| 2429 |
+
},
|
| 2430 |
+
{
|
| 2431 |
+
"epoch": 1.57,
|
| 2432 |
+
"learning_rate": 1.0706398207405737e-05,
|
| 2433 |
+
"loss": 0.0603,
|
| 2434 |
+
"step": 1755000
|
| 2435 |
+
},
|
| 2436 |
+
{
|
| 2437 |
+
"epoch": 1.58,
|
| 2438 |
+
"learning_rate": 1.0594450623951052e-05,
|
| 2439 |
+
"loss": 0.0578,
|
| 2440 |
+
"step": 1760000
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 1.58,
|
| 2444 |
+
"learning_rate": 1.0482503040496367e-05,
|
| 2445 |
+
"loss": 0.0658,
|
| 2446 |
+
"step": 1765000
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 1.59,
|
| 2450 |
+
"learning_rate": 1.0370555457041682e-05,
|
| 2451 |
+
"loss": 0.0594,
|
| 2452 |
+
"step": 1770000
|
| 2453 |
+
},
|
| 2454 |
+
{
|
| 2455 |
+
"epoch": 1.59,
|
| 2456 |
+
"learning_rate": 1.0258607873586998e-05,
|
| 2457 |
+
"loss": 0.0622,
|
| 2458 |
+
"step": 1775000
|
| 2459 |
+
},
|
| 2460 |
+
{
|
| 2461 |
+
"epoch": 1.59,
|
| 2462 |
+
"eval_loss": 0.06558772176504135,
|
| 2463 |
+
"eval_runtime": 1679.2015,
|
| 2464 |
+
"eval_samples_per_second": 36.942,
|
| 2465 |
+
"eval_steps_per_second": 36.942,
|
| 2466 |
+
"step": 1775000
|
| 2467 |
+
},
|
| 2468 |
+
{
|
| 2469 |
+
"epoch": 1.59,
|
| 2470 |
+
"learning_rate": 1.0146660290132313e-05,
|
| 2471 |
+
"loss": 0.0654,
|
| 2472 |
+
"step": 1780000
|
| 2473 |
+
},
|
| 2474 |
+
{
|
| 2475 |
+
"epoch": 1.6,
|
| 2476 |
+
"learning_rate": 1.003471270667763e-05,
|
| 2477 |
+
"loss": 0.0629,
|
| 2478 |
+
"step": 1785000
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 1.6,
|
| 2482 |
+
"learning_rate": 9.922765123222945e-06,
|
| 2483 |
+
"loss": 0.0541,
|
| 2484 |
+
"step": 1790000
|
| 2485 |
+
},
|
| 2486 |
+
{
|
| 2487 |
+
"epoch": 1.61,
|
| 2488 |
+
"learning_rate": 9.81081753976826e-06,
|
| 2489 |
+
"loss": 0.0605,
|
| 2490 |
+
"step": 1795000
|
| 2491 |
+
},
|
| 2492 |
+
{
|
| 2493 |
+
"epoch": 1.61,
|
| 2494 |
+
"learning_rate": 9.698869956313575e-06,
|
| 2495 |
+
"loss": 0.0585,
|
| 2496 |
+
"step": 1800000
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 1.61,
|
| 2500 |
+
"eval_loss": 0.0633106529712677,
|
| 2501 |
+
"eval_runtime": 1681.6891,
|
| 2502 |
+
"eval_samples_per_second": 36.887,
|
| 2503 |
+
"eval_steps_per_second": 36.887,
|
| 2504 |
+
"step": 1800000
|
| 2505 |
+
},
|
| 2506 |
+
{
|
| 2507 |
+
"epoch": 1.62,
|
| 2508 |
+
"learning_rate": 9.58692237285889e-06,
|
| 2509 |
+
"loss": 0.0654,
|
| 2510 |
+
"step": 1805000
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 1.62,
|
| 2514 |
+
"learning_rate": 9.474974789404206e-06,
|
| 2515 |
+
"loss": 0.0622,
|
| 2516 |
+
"step": 1810000
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 1.63,
|
| 2520 |
+
"learning_rate": 9.363027205949523e-06,
|
| 2521 |
+
"loss": 0.0612,
|
| 2522 |
+
"step": 1815000
|
| 2523 |
+
},
|
| 2524 |
+
{
|
| 2525 |
+
"epoch": 1.63,
|
| 2526 |
+
"learning_rate": 9.251079622494838e-06,
|
| 2527 |
+
"loss": 0.056,
|
| 2528 |
+
"step": 1820000
|
| 2529 |
+
},
|
| 2530 |
+
{
|
| 2531 |
+
"epoch": 1.63,
|
| 2532 |
+
"learning_rate": 9.139132039040153e-06,
|
| 2533 |
+
"loss": 0.0628,
|
| 2534 |
+
"step": 1825000
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 1.63,
|
| 2538 |
+
"eval_loss": 0.06252375990152359,
|
| 2539 |
+
"eval_runtime": 1681.1292,
|
| 2540 |
+
"eval_samples_per_second": 36.9,
|
| 2541 |
+
"eval_steps_per_second": 36.9,
|
| 2542 |
+
"step": 1825000
|
| 2543 |
+
},
|
| 2544 |
+
{
|
| 2545 |
+
"epoch": 1.64,
|
| 2546 |
+
"learning_rate": 9.02718445558547e-06,
|
| 2547 |
+
"loss": 0.0619,
|
| 2548 |
+
"step": 1830000
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 1.64,
|
| 2552 |
+
"learning_rate": 8.915236872130783e-06,
|
| 2553 |
+
"loss": 0.063,
|
| 2554 |
+
"step": 1835000
|
| 2555 |
+
},
|
| 2556 |
+
{
|
| 2557 |
+
"epoch": 1.65,
|
| 2558 |
+
"learning_rate": 8.803289288676098e-06,
|
| 2559 |
+
"loss": 0.0607,
|
| 2560 |
+
"step": 1840000
|
| 2561 |
+
},
|
| 2562 |
+
{
|
| 2563 |
+
"epoch": 1.65,
|
| 2564 |
+
"learning_rate": 8.691341705221415e-06,
|
| 2565 |
+
"loss": 0.0633,
|
| 2566 |
+
"step": 1845000
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 1.66,
|
| 2570 |
+
"learning_rate": 8.57939412176673e-06,
|
| 2571 |
+
"loss": 0.0638,
|
| 2572 |
+
"step": 1850000
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 1.66,
|
| 2576 |
+
"eval_loss": 0.06624045222997665,
|
| 2577 |
+
"eval_runtime": 1680.5873,
|
| 2578 |
+
"eval_samples_per_second": 36.912,
|
| 2579 |
+
"eval_steps_per_second": 36.912,
|
| 2580 |
+
"step": 1850000
|
| 2581 |
}
|
| 2582 |
],
|
| 2583 |
+
"max_steps": 2233188,
|
| 2584 |
+
"num_train_epochs": 2,
|
| 2585 |
+
"total_flos": 5.556070184057856e+17,
|
| 2586 |
"trial_name": null,
|
| 2587 |
"trial_params": null
|
| 2588 |
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3119
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:70b443918dd5518fff4a7556be44b80bcda3039466b050dbf86c95c0ba51c348
|
| 3 |
size 3119
|