End of training

Browse files

Files changed (10) hide show

README.md +26 -36
added_tokens.json +102 -0
config.json +1 -1
generation_config.json +1 -1
model.safetensors +3 -0
runs/Feb09_09-32-14_53af0da22b49/events.out.tfevents.1707471138.53af0da22b49.7344.6 +3 -0
special_tokens_map.json +21 -3
tokenizer_config.json +826 -0
trainer_state.json +924 -974
training_args.bin +2 -2

README.md CHANGED Viewed

@@ -15,9 +15,9 @@ should probably proofread and complete it, then remove this comment. -->
 This model is a fine-tuned version of [alexue4/text-normalization-ru-new](https://huggingface.co/alexue4/text-normalization-ru-new) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.0003
 - Mean Distance: 0
-- Max Distance: 0
 ## Model description
@@ -37,48 +37,38 @@ More information needed
 The following hyperparameters were used during training:
 - learning_rate: 0.0001
-- train_batch_size: 30
-- eval_batch_size: 30
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 25
 ### Training results
-| Training Loss | Epoch | Step | Validation Loss | Mean Distance | Max Distance |
-|:-------------:|:-----:|:----:|:---------------:|:-------------:|:------------:|
-| 0.0013        | 1.0   | 69   | 0.0028          | 0             | 2            |
-| 0.0006        | 2.0   | 138  | 0.0026          | 0             | 3            |
-| 0.0025        | 3.0   | 207  | 0.0039          | 0             | 3            |
-| 0.0004        | 4.0   | 276  | 0.0037          | 0             | 3            |
-| 0.0005        | 5.0   | 345  | 0.0091          | 0             | 3            |
-| 0.0009        | 6.0   | 414  | 0.0006          | 0             | 0            |
-| 0.0016        | 7.0   | 483  | 0.0003          | 0             | 0            |
-| 0.0012        | 8.0   | 552  | 0.0111          | 0             | 5            |
-| 0.0008        | 9.0   | 621  | 0.0004          | 0             | 0            |
-| 0.0018        | 10.0  | 690  | 0.0003          | 0             | 0            |
-| 0.0028        | 11.0  | 759  | 0.0003          | 0             | 0            |
-| 0.0008        | 12.0  | 828  | 0.0003          | 0             | 0            |
-| 0.001         | 13.0  | 897  | 0.0004          | 0             | 2            |
-| 0.0026        | 14.0  | 966  | 0.0005          | 0             | 2            |
-| 0.0015        | 15.0  | 1035 | 0.0007          | 0             | 3            |
-| 0.0009        | 16.0  | 1104 | 0.0007          | 0             | 3            |
-| 0.0014        | 17.0  | 1173 | 0.0003          | 0             | 0            |
-| 0.001         | 18.0  | 1242 | 0.0004          | 0             | 0            |
-| 0.0007        | 19.0  | 1311 | 0.0013          | 0             | 3            |
-| 0.0013        | 20.0  | 1380 | 0.0013          | 0             | 3            |
-| 0.0007        | 21.0  | 1449 | 0.0003          | 0             | 0            |
-| 0.0016        | 22.0  | 1518 | 0.0003          | 0             | 0            |
-| 0.0013        | 23.0  | 1587 | 0.0003          | 0             | 0            |
-| 0.0004        | 24.0  | 1656 | 0.0003          | 0             | 0            |
-| 0.001         | 25.0  | 1725 | 0.0003          | 0             | 0            |
 ### Framework versions
-- Transformers 4.32.1
-- Pytorch 2.0.1+cu117
-- Datasets 2.14.4
-- Tokenizers 0.13.3

 This model is a fine-tuned version of [alexue4/text-normalization-ru-new](https://huggingface.co/alexue4/text-normalization-ru-new) on the None dataset.
 It achieves the following results on the evaluation set:
+- Loss: 0.0008
 - Mean Distance: 0
+- Max Distance: 3
 ## Model description
 The following hyperparameters were used during training:
 - learning_rate: 0.0001
+- train_batch_size: 150
+- eval_batch_size: 150
 - seed: 42
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
+- num_epochs: 15
 ### Training results
+| Training Loss | Epoch | Step  | Validation Loss | Mean Distance | Max Distance |
+|:-------------:|:-----:|:-----:|:---------------:|:-------------:|:------------:|
+| 0.0004        | 1.0   | 3059  | 0.0022          | 0             | 12           |
+| 0.0003        | 2.0   | 6118  | 0.0011          | 0             | 5            |
+| 0.0003        | 3.0   | 9177  | 0.0010          | 0             | 6            |
+| 0.0003        | 4.0   | 12236 | 0.0012          | 0             | 3            |
+| 0.0003        | 5.0   | 15295 | 0.0008          | 0             | 3            |
+| 0.0002        | 6.0   | 18354 | 0.0009          | 0             | 3            |
+| 0.0002        | 7.0   | 21413 | 0.0008          | 0             | 3            |
+| 0.0002        | 8.0   | 24472 | 0.0008          | 0             | 3            |
+| 0.0002        | 9.0   | 27531 | 0.0007          | 0             | 3            |
+| 0.0002        | 10.0  | 30590 | 0.0008          | 0             | 3            |
+| 0.0002        | 11.0  | 33649 | 0.0008          | 0             | 3            |
+| 0.0002        | 12.0  | 36708 | 0.0008          | 0             | 3            |
+| 0.0002        | 13.0  | 39767 | 0.0008          | 0             | 3            |
+| 0.0002        | 14.0  | 42826 | 0.0008          | 0             | 3            |
+| 0.0002        | 15.0  | 45885 | 0.0008          | 0             | 3            |
 ### Framework versions
+- Transformers 4.37.2
+- Pytorch 2.1.0+cu121
+- Datasets 2.16.1
+- Tokenizers 0.15.1

added_tokens.json ADDED Viewed

	@@ -0,0 +1,102 @@

+{
+  "<extra_id_0>": 20199,
+  "<extra_id_10>": 20189,
+  "<extra_id_11>": 20188,
+  "<extra_id_12>": 20187,
+  "<extra_id_13>": 20186,
+  "<extra_id_14>": 20185,
+  "<extra_id_15>": 20184,
+  "<extra_id_16>": 20183,
+  "<extra_id_17>": 20182,
+  "<extra_id_18>": 20181,
+  "<extra_id_19>": 20180,
+  "<extra_id_1>": 20198,
+  "<extra_id_20>": 20179,
+  "<extra_id_21>": 20178,
+  "<extra_id_22>": 20177,
+  "<extra_id_23>": 20176,
+  "<extra_id_24>": 20175,
+  "<extra_id_25>": 20174,
+  "<extra_id_26>": 20173,
+  "<extra_id_27>": 20172,
+  "<extra_id_28>": 20171,
+  "<extra_id_29>": 20170,
+  "<extra_id_2>": 20197,
+  "<extra_id_30>": 20169,
+  "<extra_id_31>": 20168,
+  "<extra_id_32>": 20167,
+  "<extra_id_33>": 20166,
+  "<extra_id_34>": 20165,
+  "<extra_id_35>": 20164,
+  "<extra_id_36>": 20163,
+  "<extra_id_37>": 20162,
+  "<extra_id_38>": 20161,
+  "<extra_id_39>": 20160,
+  "<extra_id_3>": 20196,
+  "<extra_id_40>": 20159,
+  "<extra_id_41>": 20158,
+  "<extra_id_42>": 20157,
+  "<extra_id_43>": 20156,
+  "<extra_id_44>": 20155,
+  "<extra_id_45>": 20154,
+  "<extra_id_46>": 20153,
+  "<extra_id_47>": 20152,
+  "<extra_id_48>": 20151,
+  "<extra_id_49>": 20150,
+  "<extra_id_4>": 20195,
+  "<extra_id_50>": 20149,
+  "<extra_id_51>": 20148,
+  "<extra_id_52>": 20147,
+  "<extra_id_53>": 20146,
+  "<extra_id_54>": 20145,
+  "<extra_id_55>": 20144,
+  "<extra_id_56>": 20143,
+  "<extra_id_57>": 20142,
+  "<extra_id_58>": 20141,
+  "<extra_id_59>": 20140,
+  "<extra_id_5>": 20194,
+  "<extra_id_60>": 20139,
+  "<extra_id_61>": 20138,
+  "<extra_id_62>": 20137,
+  "<extra_id_63>": 20136,
+  "<extra_id_64>": 20135,
+  "<extra_id_65>": 20134,
+  "<extra_id_66>": 20133,
+  "<extra_id_67>": 20132,
+  "<extra_id_68>": 20131,
+  "<extra_id_69>": 20130,
+  "<extra_id_6>": 20193,
+  "<extra_id_70>": 20129,
+  "<extra_id_71>": 20128,
+  "<extra_id_72>": 20127,
+  "<extra_id_73>": 20126,
+  "<extra_id_74>": 20125,
+  "<extra_id_75>": 20124,
+  "<extra_id_76>": 20123,
+  "<extra_id_77>": 20122,
+  "<extra_id_78>": 20121,
+  "<extra_id_79>": 20120,
+  "<extra_id_7>": 20192,
+  "<extra_id_80>": 20119,
+  "<extra_id_81>": 20118,
+  "<extra_id_82>": 20117,
+  "<extra_id_83>": 20116,
+  "<extra_id_84>": 20115,
+  "<extra_id_85>": 20114,
+  "<extra_id_86>": 20113,
+  "<extra_id_87>": 20112,
+  "<extra_id_88>": 20111,
+  "<extra_id_89>": 20110,
+  "<extra_id_8>": 20191,
+  "<extra_id_90>": 20109,
+  "<extra_id_91>": 20108,
+  "<extra_id_92>": 20107,
+  "<extra_id_93>": 20106,
+  "<extra_id_94>": 20105,
+  "<extra_id_95>": 20104,
+  "<extra_id_96>": 20103,
+  "<extra_id_97>": 20102,
+  "<extra_id_98>": 20101,
+  "<extra_id_99>": 20100,
+  "<extra_id_9>": 20190
+}

config.json CHANGED Viewed

@@ -26,7 +26,7 @@
   "tie_word_embeddings": false,
   "tokenizer_class": "T5Tokenizer",
   "torch_dtype": "float32",
-  "transformers_version": "4.32.1",
   "use_cache": true,
   "vocab_size": 20100
 }

   "tie_word_embeddings": false,
   "tokenizer_class": "T5Tokenizer",
   "torch_dtype": "float32",
+  "transformers_version": "4.37.2",
   "use_cache": true,
   "vocab_size": 20100
 }

generation_config.json CHANGED Viewed

@@ -3,5 +3,5 @@
   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,
-  "transformers_version": "4.32.1"
 }

   "decoder_start_token_id": 0,
   "eos_token_id": 1,
   "pad_token_id": 0,
+  "transformers_version": "4.37.2"
 }

model.safetensors ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:6110aaf653c8f40296e4b2ecfada2f35edce407a76dc3ea7e879357e5b60b1fb
+size 258600360

runs/Feb09_09-32-14_53af0da22b49/events.out.tfevents.1707471138.53af0da22b49.7344.6 ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:2710126dbff402c0f7c727617f35948c443e13334be9e4d553f6fd3f73aa4b19
+size 47147

special_tokens_map.json CHANGED Viewed

@@ -101,7 +101,25 @@
     "<extra_id_98>",
     "<extra_id_99>"
   ],
-  "eos_token": "</s>",
-  "pad_token": "<pad>",
-  "unk_token": "<unk>"
 }

     "<extra_id_98>",
     "<extra_id_99>"
   ],
+  "eos_token": {
+    "content": "</s>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "pad_token": {
+    "content": "<pad>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  },
+  "unk_token": {
+    "content": "<unk>",
+    "lstrip": false,
+    "normalized": false,
+    "rstrip": false,
+    "single_word": false
+  }
 }

tokenizer_config.json CHANGED Viewed

@@ -1,4 +1,830 @@
 {
   "additional_special_tokens": [
     "<extra_id_0>",
     "<extra_id_1>",

 {
+  "added_tokens_decoder": {
+    "0": {
+      "content": "<pad>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "1": {
+      "content": "</s>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "2": {
+      "content": "<unk>",
+      "lstrip": false,
+      "normalized": false,
+      "rstrip": false,
+      "single_word": false,
+      "special": true
+    },
+    "20100": {
+      "content": "<extra_id_99>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20101": {
+      "content": "<extra_id_98>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20102": {
+      "content": "<extra_id_97>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20103": {
+      "content": "<extra_id_96>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20104": {
+      "content": "<extra_id_95>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20105": {
+      "content": "<extra_id_94>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20106": {
+      "content": "<extra_id_93>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20107": {
+      "content": "<extra_id_92>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20108": {
+      "content": "<extra_id_91>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20109": {
+      "content": "<extra_id_90>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20110": {
+      "content": "<extra_id_89>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20111": {
+      "content": "<extra_id_88>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20112": {
+      "content": "<extra_id_87>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20113": {
+      "content": "<extra_id_86>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20114": {
+      "content": "<extra_id_85>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20115": {
+      "content": "<extra_id_84>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20116": {
+      "content": "<extra_id_83>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20117": {
+      "content": "<extra_id_82>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20118": {
+      "content": "<extra_id_81>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20119": {
+      "content": "<extra_id_80>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20120": {
+      "content": "<extra_id_79>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20121": {
+      "content": "<extra_id_78>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20122": {
+      "content": "<extra_id_77>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20123": {
+      "content": "<extra_id_76>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20124": {
+      "content": "<extra_id_75>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20125": {
+      "content": "<extra_id_74>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20126": {
+      "content": "<extra_id_73>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20127": {
+      "content": "<extra_id_72>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20128": {
+      "content": "<extra_id_71>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20129": {
+      "content": "<extra_id_70>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20130": {
+      "content": "<extra_id_69>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20131": {
+      "content": "<extra_id_68>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20132": {
+      "content": "<extra_id_67>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20133": {
+      "content": "<extra_id_66>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20134": {
+      "content": "<extra_id_65>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20135": {
+      "content": "<extra_id_64>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20136": {
+      "content": "<extra_id_63>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20137": {
+      "content": "<extra_id_62>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20138": {
+      "content": "<extra_id_61>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20139": {
+      "content": "<extra_id_60>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20140": {
+      "content": "<extra_id_59>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20141": {
+      "content": "<extra_id_58>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20142": {
+      "content": "<extra_id_57>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20143": {
+      "content": "<extra_id_56>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20144": {
+      "content": "<extra_id_55>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20145": {
+      "content": "<extra_id_54>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20146": {
+      "content": "<extra_id_53>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20147": {
+      "content": "<extra_id_52>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20148": {
+      "content": "<extra_id_51>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20149": {
+      "content": "<extra_id_50>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20150": {
+      "content": "<extra_id_49>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20151": {
+      "content": "<extra_id_48>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20152": {
+      "content": "<extra_id_47>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20153": {
+      "content": "<extra_id_46>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20154": {
+      "content": "<extra_id_45>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20155": {
+      "content": "<extra_id_44>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20156": {
+      "content": "<extra_id_43>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20157": {
+      "content": "<extra_id_42>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20158": {
+      "content": "<extra_id_41>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20159": {
+      "content": "<extra_id_40>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20160": {
+      "content": "<extra_id_39>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20161": {
+      "content": "<extra_id_38>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20162": {
+      "content": "<extra_id_37>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20163": {
+      "content": "<extra_id_36>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20164": {
+      "content": "<extra_id_35>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20165": {
+      "content": "<extra_id_34>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20166": {
+      "content": "<extra_id_33>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20167": {
+      "content": "<extra_id_32>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20168": {
+      "content": "<extra_id_31>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20169": {
+      "content": "<extra_id_30>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20170": {
+      "content": "<extra_id_29>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20171": {
+      "content": "<extra_id_28>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20172": {
+      "content": "<extra_id_27>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20173": {
+      "content": "<extra_id_26>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20174": {
+      "content": "<extra_id_25>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20175": {
+      "content": "<extra_id_24>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20176": {
+      "content": "<extra_id_23>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20177": {
+      "content": "<extra_id_22>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20178": {
+      "content": "<extra_id_21>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20179": {
+      "content": "<extra_id_20>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20180": {
+      "content": "<extra_id_19>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20181": {
+      "content": "<extra_id_18>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20182": {
+      "content": "<extra_id_17>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20183": {
+      "content": "<extra_id_16>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20184": {
+      "content": "<extra_id_15>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20185": {
+      "content": "<extra_id_14>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20186": {
+      "content": "<extra_id_13>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20187": {
+      "content": "<extra_id_12>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20188": {
+      "content": "<extra_id_11>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20189": {
+      "content": "<extra_id_10>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20190": {
+      "content": "<extra_id_9>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20191": {
+      "content": "<extra_id_8>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20192": {
+      "content": "<extra_id_7>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20193": {
+      "content": "<extra_id_6>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20194": {
+      "content": "<extra_id_5>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20195": {
+      "content": "<extra_id_4>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20196": {
+      "content": "<extra_id_3>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20197": {
+      "content": "<extra_id_2>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20198": {
+      "content": "<extra_id_1>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    },
+    "20199": {
+      "content": "<extra_id_0>",
+      "lstrip": true,
+      "normalized": false,
+      "rstrip": true,
+      "single_word": false,
+      "special": true
+    }
+  },
   "additional_special_tokens": [
     "<extra_id_0>",
     "<extra_id_1>",

trainer_state.json CHANGED Viewed

@@ -1,1430 +1,1380 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 25.0,
   "eval_steps": 500,
-  "global_step": 1725,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "learning_rate": 5.780346820809248e-07,
-      "loss": 0.005,
       "step": 1
     },
     {
-      "epoch": 0.13,
-      "learning_rate": 5.202312138728324e-06,
-      "loss": 0.0008,
-      "step": 9
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 1.0404624277456647e-05,
-      "loss": 0.0019,
-      "step": 18
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 1.5606936416184973e-05,
-      "loss": 0.0019,
-      "step": 27
     },
     {
-      "epoch": 0.52,
-      "learning_rate": 2.0809248554913295e-05,
-      "loss": 0.0016,
-      "step": 36
     },
     {
-      "epoch": 0.65,
-      "learning_rate": 2.6011560693641617e-05,
-      "loss": 0.0027,
-      "step": 45
     },
     {
-      "epoch": 0.78,
-      "learning_rate": 3.1213872832369946e-05,
-      "loss": 0.0032,
-      "step": 54
     },
     {
-      "epoch": 0.91,
-      "learning_rate": 3.6416184971098265e-05,
-      "loss": 0.0013,
-      "step": 63
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 0.002811400219798088,
-      "eval_max_distance": 2,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9415,
-      "eval_samples_per_second": 42.235,
-      "eval_steps_per_second": 1.545,
-      "step": 69
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.161849710982659e-05,
-      "loss": 0.0033,
-      "step": 72
     },
     {
-      "epoch": 1.17,
-      "learning_rate": 4.6820809248554915e-05,
-      "loss": 0.0034,
-      "step": 81
     },
     {
-      "epoch": 1.3,
-      "learning_rate": 5.2023121387283234e-05,
-      "loss": 0.0015,
-      "step": 90
     },
     {
-      "epoch": 1.43,
-      "learning_rate": 5.722543352601156e-05,
-      "loss": 0.0042,
-      "step": 99
     },
     {
-      "epoch": 1.57,
-      "learning_rate": 6.242774566473989e-05,
-      "loss": 0.0025,
-      "step": 108
     },
     {
-      "epoch": 1.7,
-      "learning_rate": 6.763005780346822e-05,
-      "loss": 0.0012,
-      "step": 117
     },
     {
-      "epoch": 1.83,
-      "learning_rate": 7.283236994219653e-05,
-      "loss": 0.002,
-      "step": 126
     },
     {
-      "epoch": 1.96,
-      "learning_rate": 7.803468208092485e-05,
-      "loss": 0.0006,
-      "step": 135
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 0.0026117784436792135,
-      "eval_max_distance": 3,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9167,
-      "eval_samples_per_second": 42.781,
-      "eval_steps_per_second": 1.565,
-      "step": 138
     },
     {
-      "epoch": 2.09,
-      "learning_rate": 8.323699421965318e-05,
-      "loss": 0.001,
-      "step": 144
     },
     {
-      "epoch": 2.22,
-      "learning_rate": 8.84393063583815e-05,
-      "loss": 0.0016,
-      "step": 153
     },
     {
-      "epoch": 2.35,
-      "learning_rate": 9.364161849710983e-05,
-      "loss": 0.0011,
-      "step": 162
     },
     {
-      "epoch": 2.48,
-      "learning_rate": 9.884393063583816e-05,
-      "loss": 0.001,
-      "step": 171
     },
     {
-      "epoch": 2.61,
-      "learning_rate": 9.954896907216495e-05,
-      "loss": 0.0026,
-      "step": 180
     },
     {
-      "epoch": 2.74,
-      "learning_rate": 9.896907216494846e-05,
-      "loss": 0.0008,
-      "step": 189
     },
     {
-      "epoch": 2.87,
-      "learning_rate": 9.838917525773196e-05,
-      "loss": 0.0022,
-      "step": 198
     },
     {
-      "epoch": 3.0,
-      "learning_rate": 9.780927835051546e-05,
-      "loss": 0.0025,
-      "step": 207
     },
     {
-      "epoch": 3.0,
-      "eval_loss": 0.003930480219423771,
-      "eval_max_distance": 3,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9341,
-      "eval_samples_per_second": 42.396,
-      "eval_steps_per_second": 1.551,
-      "step": 207
     },
     {
-      "epoch": 3.13,
-      "learning_rate": 9.722938144329897e-05,
-      "loss": 0.0019,
-      "step": 216
     },
     {
-      "epoch": 3.26,
-      "learning_rate": 9.664948453608248e-05,
-      "loss": 0.0024,
-      "step": 225
     },
     {
-      "epoch": 3.39,
-      "learning_rate": 9.606958762886598e-05,
-      "loss": 0.0007,
-      "step": 234
     },
     {
-      "epoch": 3.52,
-      "learning_rate": 9.54896907216495e-05,
-      "loss": 0.0026,
-      "step": 243
     },
     {
-      "epoch": 3.65,
-      "learning_rate": 9.490979381443299e-05,
-      "loss": 0.0009,
-      "step": 252
     },
     {
-      "epoch": 3.78,
-      "learning_rate": 9.43298969072165e-05,
-      "loss": 0.0012,
-      "step": 261
     },
     {
-      "epoch": 3.91,
-      "learning_rate": 9.375e-05,
       "loss": 0.0004,
-      "step": 270
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 0.0036886015441268682,
-      "eval_max_distance": 3,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9271,
-      "eval_samples_per_second": 42.551,
-      "eval_steps_per_second": 1.557,
-      "step": 276
     },
     {
-      "epoch": 4.04,
-      "learning_rate": 9.317010309278351e-05,
-      "loss": 0.0018,
-      "step": 279
     },
     {
-      "epoch": 4.17,
-      "learning_rate": 9.259020618556701e-05,
-      "loss": 0.0017,
-      "step": 288
     },
     {
-      "epoch": 4.3,
-      "learning_rate": 9.201030927835051e-05,
-      "loss": 0.0014,
-      "step": 297
     },
     {
-      "epoch": 4.43,
-      "learning_rate": 9.143041237113402e-05,
-      "loss": 0.0005,
-      "step": 306
     },
     {
-      "epoch": 4.57,
-      "learning_rate": 9.085051546391753e-05,
-      "loss": 0.001,
-      "step": 315
     },
     {
-      "epoch": 4.7,
-      "learning_rate": 9.027061855670103e-05,
-      "loss": 0.002,
-      "step": 324
     },
     {
-      "epoch": 4.83,
-      "learning_rate": 8.969072164948454e-05,
-      "loss": 0.0016,
-      "step": 333
     },
     {
-      "epoch": 4.96,
-      "learning_rate": 8.911082474226806e-05,
-      "loss": 0.0005,
-      "step": 342
     },
     {
-      "epoch": 5.0,
-      "eval_loss": 0.009117466397583485,
-      "eval_max_distance": 3,
       "eval_mean_distance": 0,
-      "eval_runtime": 1.9694,
-      "eval_samples_per_second": 41.638,
-      "eval_steps_per_second": 1.523,
-      "step": 345
     },
     {
-      "epoch": 5.09,
-      "learning_rate": 8.853092783505154e-05,
-      "loss": 0.0005,
-      "step": 351
     },
     {
-      "epoch": 5.22,
-      "learning_rate": 8.795103092783505e-05,
-      "loss": 0.0051,
-      "step": 360
     },
     {
-      "epoch": 5.35,
-      "learning_rate": 8.737113402061856e-05,
-      "loss": 0.0004,
-      "step": 369
     },
     {
-      "epoch": 5.48,
-      "learning_rate": 8.679123711340206e-05,
-      "loss": 0.0012,
-      "step": 378
     },
     {
-      "epoch": 5.61,
-      "learning_rate": 8.621134020618558e-05,
-      "loss": 0.001,
-      "step": 387
     },
     {
-      "epoch": 5.74,
-      "learning_rate": 8.563144329896907e-05,
-      "loss": 0.0015,
-      "step": 396
     },
     {
-      "epoch": 5.87,
-      "learning_rate": 8.505154639175259e-05,
-      "loss": 0.0016,
-      "step": 405
     },
     {
-      "epoch": 6.0,
-      "learning_rate": 8.447164948453608e-05,
-      "loss": 0.0009,
-      "step": 414
     },
     {
-      "epoch": 6.0,
-      "eval_loss": 0.0005720060435123742,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9399,
-      "eval_samples_per_second": 42.271,
-      "eval_steps_per_second": 1.546,
-      "step": 414
     },
     {
-      "epoch": 6.13,
-      "learning_rate": 8.38917525773196e-05,
-      "loss": 0.0005,
-      "step": 423
     },
     {
-      "epoch": 6.26,
-      "learning_rate": 8.331185567010311e-05,
-      "loss": 0.0006,
-      "step": 432
     },
     {
-      "epoch": 6.39,
-      "learning_rate": 8.273195876288659e-05,
-      "loss": 0.0014,
-      "step": 441
     },
     {
-      "epoch": 6.52,
-      "learning_rate": 8.21520618556701e-05,
-      "loss": 0.0019,
-      "step": 450
     },
     {
-      "epoch": 6.65,
-      "learning_rate": 8.157216494845362e-05,
-      "loss": 0.001,
-      "step": 459
     },
     {
-      "epoch": 6.78,
-      "learning_rate": 8.099226804123711e-05,
-      "loss": 0.0008,
-      "step": 468
     },
     {
-      "epoch": 6.91,
-      "learning_rate": 8.041237113402063e-05,
-      "loss": 0.0016,
-      "step": 477
     },
     {
-      "epoch": 7.0,
-      "eval_loss": 0.00027213190332986414,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9188,
-      "eval_samples_per_second": 42.735,
-      "eval_steps_per_second": 1.563,
-      "step": 483
     },
     {
-      "epoch": 7.04,
-      "learning_rate": 7.983247422680414e-05,
-      "loss": 0.0006,
-      "step": 486
     },
     {
-      "epoch": 7.17,
-      "learning_rate": 7.925257731958762e-05,
-      "loss": 0.007,
-      "step": 495
     },
     {
-      "epoch": 7.3,
-      "learning_rate": 7.867268041237113e-05,
-      "loss": 0.0004,
-      "step": 504
     },
     {
-      "epoch": 7.43,
-      "learning_rate": 7.809278350515465e-05,
-      "loss": 0.0016,
-      "step": 513
     },
     {
-      "epoch": 7.57,
-      "learning_rate": 7.751288659793814e-05,
-      "loss": 0.0006,
-      "step": 522
     },
     {
-      "epoch": 7.7,
-      "learning_rate": 7.693298969072166e-05,
-      "loss": 0.0011,
-      "step": 531
     },
     {
-      "epoch": 7.83,
-      "learning_rate": 7.635309278350515e-05,
-      "loss": 0.0014,
-      "step": 540
     },
     {
-      "epoch": 7.96,
-      "learning_rate": 7.577319587628867e-05,
-      "loss": 0.0012,
-      "step": 549
     },
     {
-      "epoch": 8.0,
-      "eval_loss": 0.011139851063489914,
-      "eval_max_distance": 5,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9435,
-      "eval_samples_per_second": 42.193,
-      "eval_steps_per_second": 1.544,
-      "step": 552
     },
     {
-      "epoch": 8.09,
-      "learning_rate": 7.519329896907217e-05,
-      "loss": 0.0008,
-      "step": 558
     },
     {
-      "epoch": 8.22,
-      "learning_rate": 7.461340206185568e-05,
-      "loss": 0.0011,
-      "step": 567
     },
     {
-      "epoch": 8.35,
-      "learning_rate": 7.403350515463919e-05,
-      "loss": 0.0025,
-      "step": 576
     },
     {
-      "epoch": 8.48,
-      "learning_rate": 7.345360824742269e-05,
-      "loss": 0.003,
-      "step": 585
     },
     {
-      "epoch": 8.61,
-      "learning_rate": 7.287371134020619e-05,
-      "loss": 0.004,
-      "step": 594
     },
     {
-      "epoch": 8.74,
-      "learning_rate": 7.22938144329897e-05,
-      "loss": 0.002,
-      "step": 603
     },
     {
-      "epoch": 8.87,
-      "learning_rate": 7.17139175257732e-05,
-      "loss": 0.0007,
-      "step": 612
     },
     {
-      "epoch": 9.0,
-      "learning_rate": 7.113402061855671e-05,
-      "loss": 0.0008,
-      "step": 621
     },
     {
-      "epoch": 9.0,
-      "eval_loss": 0.0003953798732254654,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.995,
-      "eval_samples_per_second": 41.102,
-      "eval_steps_per_second": 1.504,
-      "step": 621
     },
     {
-      "epoch": 9.13,
-      "learning_rate": 7.055412371134022e-05,
-      "loss": 0.0005,
-      "step": 630
     },
     {
-      "epoch": 9.26,
-      "learning_rate": 6.99742268041237e-05,
-      "loss": 0.0004,
-      "step": 639
     },
     {
-      "epoch": 9.39,
-      "learning_rate": 6.939432989690722e-05,
-      "loss": 0.0013,
-      "step": 648
     },
     {
-      "epoch": 9.52,
-      "learning_rate": 6.881443298969073e-05,
       "loss": 0.0002,
-      "step": 657
     },
     {
-      "epoch": 9.65,
-      "learning_rate": 6.823453608247423e-05,
-      "loss": 0.0011,
-      "step": 666
     },
     {
-      "epoch": 9.78,
-      "learning_rate": 6.765463917525774e-05,
-      "loss": 0.0018,
-      "step": 675
     },
     {
-      "epoch": 9.91,
-      "learning_rate": 6.707474226804124e-05,
-      "loss": 0.0018,
-      "step": 684
     },
     {
-      "epoch": 10.0,
-      "eval_loss": 0.00027754431357607245,
-      "eval_max_distance": 0,
       "eval_mean_distance": 0,
-      "eval_runtime": 1.9222,
-      "eval_samples_per_second": 42.659,
-      "eval_steps_per_second": 1.561,
-      "step": 690
-    },
-    {
-      "epoch": 10.04,
-      "learning_rate": 6.649484536082475e-05,
-      "loss": 0.0011,
-      "step": 693
     },
     {
-      "epoch": 10.17,
-      "learning_rate": 6.591494845360825e-05,
-      "loss": 0.0006,
-      "step": 702
     },
     {
-      "epoch": 10.3,
-      "learning_rate": 6.533505154639176e-05,
-      "loss": 0.0011,
-      "step": 711
     },
     {
-      "epoch": 10.43,
-      "learning_rate": 6.475515463917527e-05,
-      "loss": 0.0013,
-      "step": 720
     },
     {
-      "epoch": 10.57,
-      "learning_rate": 6.417525773195877e-05,
-      "loss": 0.0006,
-      "step": 729
     },
     {
-      "epoch": 10.7,
-      "learning_rate": 6.359536082474227e-05,
-      "loss": 0.0018,
-      "step": 738
     },
     {
-      "epoch": 10.83,
-      "learning_rate": 6.301546391752578e-05,
-      "loss": 0.0016,
-      "step": 747
     },
     {
-      "epoch": 10.96,
-      "learning_rate": 6.243556701030928e-05,
-      "loss": 0.0028,
-      "step": 756
     },
     {
-      "epoch": 11.0,
-      "eval_loss": 0.00033258015173487365,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9385,
-      "eval_samples_per_second": 42.301,
-      "eval_steps_per_second": 1.548,
-      "step": 759
     },
     {
-      "epoch": 11.09,
-      "learning_rate": 6.185567010309279e-05,
-      "loss": 0.0009,
-      "step": 765
     },
     {
-      "epoch": 11.22,
-      "learning_rate": 6.12757731958763e-05,
-      "loss": 0.0005,
-      "step": 774
     },
     {
-      "epoch": 11.35,
-      "learning_rate": 6.069587628865979e-05,
-      "loss": 0.0011,
-      "step": 783
     },
     {
-      "epoch": 11.48,
-      "learning_rate": 6.01159793814433e-05,
-      "loss": 0.0007,
-      "step": 792
     },
     {
-      "epoch": 11.61,
-      "learning_rate": 5.953608247422681e-05,
-      "loss": 0.0012,
-      "step": 801
     },
     {
-      "epoch": 11.74,
-      "learning_rate": 5.8956185567010315e-05,
-      "loss": 0.0021,
-      "step": 810
     },
     {
-      "epoch": 11.87,
-      "learning_rate": 5.837628865979382e-05,
-      "loss": 0.0006,
-      "step": 819
     },
     {
-      "epoch": 12.0,
-      "learning_rate": 5.779639175257732e-05,
-      "loss": 0.0008,
-      "step": 828
     },
     {
-      "epoch": 12.0,
-      "eval_loss": 0.0002690624096430838,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9448,
-      "eval_samples_per_second": 42.163,
-      "eval_steps_per_second": 1.543,
-      "step": 828
     },
     {
-      "epoch": 12.13,
-      "learning_rate": 5.721649484536082e-05,
-      "loss": 0.001,
-      "step": 837
     },
     {
-      "epoch": 12.26,
-      "learning_rate": 5.663659793814433e-05,
-      "loss": 0.0012,
-      "step": 846
     },
     {
-      "epoch": 12.39,
-      "learning_rate": 5.605670103092784e-05,
-      "loss": 0.0005,
-      "step": 855
     },
     {
-      "epoch": 12.52,
-      "learning_rate": 5.5476804123711345e-05,
-      "loss": 0.0011,
-      "step": 864
     },
     {
-      "epoch": 12.65,
-      "learning_rate": 5.489690721649485e-05,
-      "loss": 0.0026,
-      "step": 873
     },
     {
-      "epoch": 12.78,
-      "learning_rate": 5.431701030927835e-05,
-      "loss": 0.0009,
-      "step": 882
     },
     {
-      "epoch": 12.91,
-      "learning_rate": 5.3737113402061854e-05,
-      "loss": 0.001,
-      "step": 891
     },
     {
-      "epoch": 13.0,
-      "eval_loss": 0.0004277784610167146,
-      "eval_max_distance": 2,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9315,
-      "eval_samples_per_second": 42.454,
-      "eval_steps_per_second": 1.553,
-      "step": 897
     },
     {
-      "epoch": 13.04,
-      "learning_rate": 5.3157216494845366e-05,
-      "loss": 0.0007,
-      "step": 900
     },
     {
-      "epoch": 13.17,
-      "learning_rate": 5.257731958762887e-05,
-      "loss": 0.0008,
-      "step": 909
     },
     {
-      "epoch": 13.3,
-      "learning_rate": 5.1997422680412376e-05,
-      "loss": 0.0005,
-      "step": 918
     },
     {
-      "epoch": 13.43,
-      "learning_rate": 5.1417525773195874e-05,
-      "loss": 0.0004,
-      "step": 927
     },
     {
-      "epoch": 13.57,
-      "learning_rate": 5.083762886597938e-05,
-      "loss": 0.0006,
-      "step": 936
     },
     {
-      "epoch": 13.7,
-      "learning_rate": 5.025773195876289e-05,
-      "loss": 0.0025,
-      "step": 945
     },
     {
-      "epoch": 13.83,
-      "learning_rate": 4.9677835051546396e-05,
-      "loss": 0.0009,
-      "step": 954
     },
     {
-      "epoch": 13.96,
-      "learning_rate": 4.9097938144329895e-05,
-      "loss": 0.0026,
-      "step": 963
     },
     {
-      "epoch": 14.0,
-      "eval_loss": 0.0005385838449001312,
-      "eval_max_distance": 2,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.993,
-      "eval_samples_per_second": 41.144,
-      "eval_steps_per_second": 1.505,
-      "step": 966
     },
     {
-      "epoch": 14.09,
-      "learning_rate": 4.8518041237113407e-05,
-      "loss": 0.0016,
-      "step": 972
     },
     {
-      "epoch": 14.22,
-      "learning_rate": 4.793814432989691e-05,
-      "loss": 0.0014,
-      "step": 981
     },
     {
-      "epoch": 14.35,
-      "learning_rate": 4.735824742268041e-05,
-      "loss": 0.0007,
-      "step": 990
     },
     {
-      "epoch": 14.48,
-      "learning_rate": 4.677835051546392e-05,
-      "loss": 0.0031,
-      "step": 999
     },
     {
-      "epoch": 14.61,
-      "learning_rate": 4.619845360824743e-05,
-      "loss": 0.0008,
-      "step": 1008
     },
     {
-      "epoch": 14.74,
-      "learning_rate": 4.561855670103093e-05,
-      "loss": 0.0028,
-      "step": 1017
     },
     {
-      "epoch": 14.87,
-      "learning_rate": 4.503865979381444e-05,
-      "loss": 0.0004,
-      "step": 1026
     },
     {
-      "epoch": 15.0,
-      "learning_rate": 4.4458762886597936e-05,
-      "loss": 0.0015,
-      "step": 1035
     },
     {
-      "epoch": 15.0,
-      "eval_loss": 0.0007138837827369571,
       "eval_max_distance": 3,
       "eval_mean_distance": 0,
-      "eval_runtime": 1.9688,
-      "eval_samples_per_second": 41.651,
-      "eval_steps_per_second": 1.524,
-      "step": 1035
     },
     {
-      "epoch": 15.13,
-      "learning_rate": 4.387886597938145e-05,
-      "loss": 0.0005,
-      "step": 1044
     },
     {
-      "epoch": 15.26,
-      "learning_rate": 4.329896907216495e-05,
-      "loss": 0.0011,
-      "step": 1053
     },
     {
-      "epoch": 15.39,
-      "learning_rate": 4.271907216494845e-05,
-      "loss": 0.0001,
-      "step": 1062
     },
     {
-      "epoch": 15.52,
-      "learning_rate": 4.213917525773196e-05,
-      "loss": 0.0009,
-      "step": 1071
     },
     {
-      "epoch": 15.65,
-      "learning_rate": 4.155927835051547e-05,
-      "loss": 0.0017,
-      "step": 1080
     },
     {
-      "epoch": 15.78,
-      "learning_rate": 4.097938144329897e-05,
-      "loss": 0.0013,
-      "step": 1089
     },
     {
-      "epoch": 15.91,
-      "learning_rate": 4.039948453608248e-05,
-      "loss": 0.0009,
-      "step": 1098
     },
     {
-      "epoch": 16.0,
-      "eval_loss": 0.0006717974320054054,
-      "eval_max_distance": 3,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9244,
-      "eval_samples_per_second": 42.612,
-      "eval_steps_per_second": 1.559,
-      "step": 1104
     },
     {
-      "epoch": 16.04,
-      "learning_rate": 3.9819587628865976e-05,
-      "loss": 0.0015,
-      "step": 1107
     },
     {
-      "epoch": 16.17,
-      "learning_rate": 3.923969072164949e-05,
       "loss": 0.0003,
-      "step": 1116
     },
     {
-      "epoch": 16.3,
-      "learning_rate": 3.865979381443299e-05,
-      "loss": 0.0006,
-      "step": 1125
     },
     {
-      "epoch": 16.43,
-      "learning_rate": 3.807989690721649e-05,
-      "loss": 0.0013,
-      "step": 1134
     },
     {
-      "epoch": 16.57,
-      "learning_rate": 3.7500000000000003e-05,
-      "loss": 0.0009,
-      "step": 1143
     },
     {
-      "epoch": 16.7,
-      "learning_rate": 3.692010309278351e-05,
-      "loss": 0.0004,
-      "step": 1152
     },
     {
-      "epoch": 16.83,
-      "learning_rate": 3.6340206185567014e-05,
-      "loss": 0.0008,
-      "step": 1161
     },
     {
-      "epoch": 16.96,
-      "learning_rate": 3.576030927835052e-05,
-      "loss": 0.0014,
-      "step": 1170
     },
     {
-      "epoch": 17.0,
-      "eval_loss": 0.00033988503855653107,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9136,
-      "eval_samples_per_second": 42.851,
-      "eval_steps_per_second": 1.568,
-      "step": 1173
     },
     {
-      "epoch": 17.09,
-      "learning_rate": 3.5180412371134024e-05,
-      "loss": 0.0016,
-      "step": 1179
     },
     {
-      "epoch": 17.22,
-      "learning_rate": 3.460051546391753e-05,
-      "loss": 0.0004,
-      "step": 1188
     },
     {
-      "epoch": 17.35,
-      "learning_rate": 3.4020618556701034e-05,
-      "loss": 0.002,
-      "step": 1197
     },
     {
-      "epoch": 17.48,
-      "learning_rate": 3.344072164948453e-05,
-      "loss": 0.0012,
-      "step": 1206
     },
     {
-      "epoch": 17.61,
-      "learning_rate": 3.2860824742268044e-05,
       "loss": 0.0002,
-      "step": 1215
     },
     {
-      "epoch": 17.74,
-      "learning_rate": 3.228092783505155e-05,
-      "loss": 0.0006,
-      "step": 1224
     },
     {
-      "epoch": 17.87,
-      "learning_rate": 3.1701030927835054e-05,
-      "loss": 0.0044,
-      "step": 1233
     },
     {
-      "epoch": 18.0,
-      "learning_rate": 3.112113402061856e-05,
-      "loss": 0.001,
-      "step": 1242
     },
     {
-      "epoch": 18.0,
-      "eval_loss": 0.00037691937177442014,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9496,
-      "eval_samples_per_second": 42.059,
-      "eval_steps_per_second": 1.539,
-      "step": 1242
     },
     {
-      "epoch": 18.13,
-      "learning_rate": 3.0541237113402065e-05,
-      "loss": 0.0005,
-      "step": 1251
     },
     {
-      "epoch": 18.26,
-      "learning_rate": 2.9961340206185566e-05,
-      "loss": 0.0009,
-      "step": 1260
     },
     {
-      "epoch": 18.39,
-      "learning_rate": 2.9381443298969075e-05,
-      "loss": 0.0008,
-      "step": 1269
     },
     {
-      "epoch": 18.52,
-      "learning_rate": 2.8801546391752577e-05,
-      "loss": 0.0007,
-      "step": 1278
     },
     {
-      "epoch": 18.65,
-      "learning_rate": 2.8221649484536085e-05,
-      "loss": 0.0016,
-      "step": 1287
     },
     {
-      "epoch": 18.78,
-      "learning_rate": 2.764175257731959e-05,
-      "loss": 0.0012,
-      "step": 1296
     },
     {
-      "epoch": 18.91,
-      "learning_rate": 2.7061855670103092e-05,
-      "loss": 0.0007,
-      "step": 1305
     },
     {
-      "epoch": 19.0,
-      "eval_loss": 0.001327142701484263,
-      "eval_max_distance": 3,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9144,
-      "eval_samples_per_second": 42.834,
-      "eval_steps_per_second": 1.567,
-      "step": 1311
     },
     {
-      "epoch": 19.04,
-      "learning_rate": 2.64819587628866e-05,
-      "loss": 0.0005,
-      "step": 1314
     },
     {
-      "epoch": 19.17,
-      "learning_rate": 2.5902061855670106e-05,
-      "loss": 0.0011,
-      "step": 1323
     },
     {
-      "epoch": 19.3,
-      "learning_rate": 2.5322164948453607e-05,
-      "loss": 0.0006,
-      "step": 1332
     },
     {
-      "epoch": 19.43,
-      "learning_rate": 2.4742268041237116e-05,
-      "loss": 0.0004,
-      "step": 1341
     },
     {
-      "epoch": 19.57,
-      "learning_rate": 2.416237113402062e-05,
-      "loss": 0.002,
-      "step": 1350
     },
     {
-      "epoch": 19.7,
-      "learning_rate": 2.3582474226804126e-05,
-      "loss": 0.0003,
-      "step": 1359
     },
     {
-      "epoch": 19.83,
-      "learning_rate": 2.3002577319587628e-05,
-      "loss": 0.0011,
-      "step": 1368
     },
     {
-      "epoch": 19.96,
-      "learning_rate": 2.2422680412371136e-05,
-      "loss": 0.0013,
-      "step": 1377
     },
     {
-      "epoch": 20.0,
-      "eval_loss": 0.0012958323350176215,
       "eval_max_distance": 3,
       "eval_mean_distance": 0,
-      "eval_runtime": 1.9255,
-      "eval_samples_per_second": 42.587,
-      "eval_steps_per_second": 1.558,
-      "step": 1380
     },
     {
-      "epoch": 20.09,
-      "learning_rate": 2.184278350515464e-05,
-      "loss": 0.0017,
-      "step": 1386
     },
     {
-      "epoch": 20.22,
-      "learning_rate": 2.1262886597938146e-05,
-      "loss": 0.0011,
-      "step": 1395
     },
     {
-      "epoch": 20.35,
-      "learning_rate": 2.0682989690721648e-05,
-      "loss": 0.0014,
-      "step": 1404
     },
     {
-      "epoch": 20.48,
-      "learning_rate": 2.0103092783505157e-05,
-      "loss": 0.0006,
-      "step": 1413
     },
     {
-      "epoch": 20.61,
-      "learning_rate": 1.952319587628866e-05,
-      "loss": 0.0003,
-      "step": 1422
     },
     {
-      "epoch": 20.74,
-      "learning_rate": 1.8943298969072167e-05,
       "loss": 0.0002,
-      "step": 1431
     },
     {
-      "epoch": 20.87,
-      "learning_rate": 1.8363402061855672e-05,
-      "loss": 0.0024,
-      "step": 1440
     },
     {
-      "epoch": 21.0,
-      "learning_rate": 1.7783505154639177e-05,
-      "loss": 0.0007,
-      "step": 1449
     },
     {
-      "epoch": 21.0,
-      "eval_loss": 0.0002567500632721931,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9342,
-      "eval_samples_per_second": 42.395,
-      "eval_steps_per_second": 1.551,
-      "step": 1449
     },
     {
-      "epoch": 21.13,
-      "learning_rate": 1.7203608247422682e-05,
-      "loss": 0.0006,
-      "step": 1458
     },
     {
-      "epoch": 21.26,
-      "learning_rate": 1.6623711340206187e-05,
-      "loss": 0.002,
-      "step": 1467
     },
     {
-      "epoch": 21.39,
-      "learning_rate": 1.6043814432989692e-05,
-      "loss": 0.0007,
-      "step": 1476
     },
     {
-      "epoch": 21.52,
-      "learning_rate": 1.5463917525773197e-05,
-      "loss": 0.0007,
-      "step": 1485
     },
     {
-      "epoch": 21.65,
-      "learning_rate": 1.4884020618556702e-05,
-      "loss": 0.0016,
-      "step": 1494
     },
     {
-      "epoch": 21.78,
-      "learning_rate": 1.4304123711340206e-05,
-      "loss": 0.0008,
-      "step": 1503
     },
     {
-      "epoch": 21.91,
-      "learning_rate": 1.3724226804123713e-05,
-      "loss": 0.0016,
-      "step": 1512
     },
     {
-      "epoch": 22.0,
-      "eval_loss": 0.0002821955131366849,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9113,
-      "eval_samples_per_second": 42.902,
-      "eval_steps_per_second": 1.57,
-      "step": 1518
     },
     {
-      "epoch": 22.04,
-      "learning_rate": 1.3144329896907218e-05,
-      "loss": 0.0018,
-      "step": 1521
     },
     {
-      "epoch": 22.17,
-      "learning_rate": 1.2564432989690723e-05,
-      "loss": 0.0027,
-      "step": 1530
     },
     {
-      "epoch": 22.3,
-      "learning_rate": 1.1984536082474228e-05,
       "loss": 0.0002,
-      "step": 1539
     },
     {
-      "epoch": 22.43,
-      "learning_rate": 1.1404639175257733e-05,
-      "loss": 0.001,
-      "step": 1548
     },
     {
-      "epoch": 22.57,
-      "learning_rate": 1.0824742268041238e-05,
-      "loss": 0.0007,
-      "step": 1557
     },
     {
-      "epoch": 22.7,
-      "learning_rate": 1.0244845360824743e-05,
-      "loss": 0.0009,
-      "step": 1566
     },
     {
-      "epoch": 22.83,
-      "learning_rate": 9.664948453608248e-06,
-      "loss": 0.0012,
-      "step": 1575
     },
     {
-      "epoch": 22.96,
-      "learning_rate": 9.085051546391753e-06,
-      "loss": 0.0013,
-      "step": 1584
     },
     {
-      "epoch": 23.0,
-      "eval_loss": 0.00030223012436181307,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9675,
-      "eval_samples_per_second": 41.677,
-      "eval_steps_per_second": 1.525,
-      "step": 1587
     },
     {
-      "epoch": 23.09,
-      "learning_rate": 8.505154639175259e-06,
-      "loss": 0.0025,
-      "step": 1593
     },
     {
-      "epoch": 23.22,
-      "learning_rate": 7.925257731958764e-06,
-      "loss": 0.001,
-      "step": 1602
     },
     {
-      "epoch": 23.35,
-      "learning_rate": 7.345360824742269e-06,
-      "loss": 0.0004,
-      "step": 1611
     },
     {
-      "epoch": 23.48,
-      "learning_rate": 6.765463917525773e-06,
-      "loss": 0.0006,
-      "step": 1620
     },
     {
-      "epoch": 23.61,
-      "learning_rate": 6.185567010309279e-06,
-      "loss": 0.001,
-      "step": 1629
     },
     {
-      "epoch": 23.74,
-      "learning_rate": 5.605670103092784e-06,
-      "loss": 0.0012,
-      "step": 1638
     },
     {
-      "epoch": 23.87,
-      "learning_rate": 5.025773195876289e-06,
-      "loss": 0.0013,
-      "step": 1647
     },
     {
-      "epoch": 24.0,
-      "learning_rate": 4.445876288659794e-06,
-      "loss": 0.0004,
-      "step": 1656
     },
     {
-      "epoch": 24.0,
-      "eval_loss": 0.0002631743554957211,
-      "eval_max_distance": 0,
-      "eval_mean_distance": 0,
-      "eval_runtime": 1.9278,
-      "eval_samples_per_second": 42.536,
-      "eval_steps_per_second": 1.556,
-      "step": 1656
     },
     {
-      "epoch": 24.13,
-      "learning_rate": 3.865979381443299e-06,
-      "loss": 0.0018,
-      "step": 1665
     },
     {
-      "epoch": 24.26,
-      "learning_rate": 3.2860824742268044e-06,
-      "loss": 0.0006,
-      "step": 1674
     },
     {
-      "epoch": 24.39,
-      "learning_rate": 2.7061855670103095e-06,
-      "loss": 0.0012,
-      "step": 1683
     },
     {
-      "epoch": 24.52,
-      "learning_rate": 2.1262886597938146e-06,
-      "loss": 0.0009,
-      "step": 1692
     },
     {
-      "epoch": 24.65,
-      "learning_rate": 1.5463917525773197e-06,
-      "loss": 0.0007,
-      "step": 1701
     },
     {
-      "epoch": 24.78,
-      "learning_rate": 9.664948453608248e-07,
-      "loss": 0.0009,
-      "step": 1710
     },
     {
-      "epoch": 24.91,
-      "learning_rate": 3.8659793814432993e-07,
-      "loss": 0.001,
-      "step": 1719
     },
     {
-      "epoch": 25.0,
-      "eval_loss": 0.0002593309909570962,
-      "eval_max_distance": 0,
       "eval_mean_distance": 0,
-      "eval_runtime": 2.0449,
-      "eval_samples_per_second": 40.1,
-      "eval_steps_per_second": 1.467,
-      "step": 1725
-    },
-    {
-      "epoch": 25.0,
-      "step": 1725,
-      "total_flos": 459342194208768.0,
-      "train_loss": 0.0013090899151048043,
-      "train_runtime": 199.817,
-      "train_samples_per_second": 256.61,
-      "train_steps_per_second": 8.633
     }
   ],
-  "logging_steps": 9,
-  "max_steps": 1725,
-  "num_train_epochs": 25,
-  "save_steps": 18,
-  "total_flos": 459342194208768.0,
   "trial_name": null,
   "trial_params": null
 }

 {
   "best_metric": null,
   "best_model_checkpoint": null,
+  "epoch": 15.0,
   "eval_steps": 500,
+  "global_step": 45885,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
+      "epoch": 0.0,
+      "learning_rate": 2.1791239921551537e-08,
+      "loss": 0.0002,
       "step": 1
     },
     {
+      "epoch": 0.08,
+      "learning_rate": 5.011985181956853e-06,
+      "loss": 0.0017,
+      "step": 230
     },
     {
+      "epoch": 0.15,
+      "learning_rate": 1.0023970363913707e-05,
+      "loss": 0.0012,
+      "step": 460
     },
     {
+      "epoch": 0.23,
+      "learning_rate": 1.503595554587056e-05,
+      "loss": 0.0008,
+      "step": 690
     },
     {
+      "epoch": 0.3,
+      "learning_rate": 2.0047940727827413e-05,
+      "loss": 0.001,
+      "step": 920
     },
     {
+      "epoch": 0.38,
+      "learning_rate": 2.5059925909784264e-05,
+      "loss": 0.0008,
+      "step": 1150
     },
     {
+      "epoch": 0.45,
+      "learning_rate": 3.007191109174112e-05,
+      "loss": 0.0007,
+      "step": 1380
     },
     {
+      "epoch": 0.53,
+      "learning_rate": 3.5083896273697975e-05,
+      "loss": 0.0007,
+      "step": 1610
     },
     {
+      "epoch": 0.6,
+      "learning_rate": 4.0095881455654826e-05,
+      "loss": 0.0006,
+      "step": 1840
     },
     {
+      "epoch": 0.68,
+      "learning_rate": 4.5107866637611684e-05,
+      "loss": 0.0006,
+      "step": 2070
     },
     {
+      "epoch": 0.75,
+      "learning_rate": 5.011985181956853e-05,
+      "loss": 0.0005,
+      "step": 2300
     },
     {
+      "epoch": 0.83,
+      "learning_rate": 5.5131837001525385e-05,
+      "loss": 0.0004,
+      "step": 2530
     },
     {
+      "epoch": 0.9,
+      "learning_rate": 6.014382218348224e-05,
+      "loss": 0.0003,
+      "step": 2760
     },
     {
+      "epoch": 0.98,
+      "learning_rate": 6.51558073654391e-05,
+      "loss": 0.0004,
+      "step": 2990
     },
     {
+      "epoch": 1.0,
+      "eval_loss": 0.002156034577637911,
+      "eval_max_distance": 12,
+      "eval_mean_distance": 0,
+      "eval_runtime": 31.367,
+      "eval_samples_per_second": 18.778,
+      "eval_steps_per_second": 0.128,
+      "step": 3059
     },
     {
+      "epoch": 1.05,
+      "learning_rate": 7.016779254739595e-05,
+      "loss": 0.0004,
+      "step": 3220
     },
     {
+      "epoch": 1.13,
+      "learning_rate": 7.517977772935281e-05,
+      "loss": 0.0003,
+      "step": 3450
     },
     {
+      "epoch": 1.2,
+      "learning_rate": 8.019176291130965e-05,
+      "loss": 0.0004,
+      "step": 3680
     },
     {
+      "epoch": 1.28,
+      "learning_rate": 8.520374809326651e-05,
+      "loss": 0.0003,
+      "step": 3910
     },
     {
+      "epoch": 1.35,
+      "learning_rate": 9.021573327522337e-05,
+      "loss": 0.0004,
+      "step": 4140
     },
     {
+      "epoch": 1.43,
+      "learning_rate": 9.522771845718021e-05,
+      "loss": 0.0003,
+      "step": 4370
     },
     {
+      "epoch": 1.5,
+      "learning_rate": 9.997336303758234e-05,
+      "loss": 0.0003,
+      "step": 4600
     },
     {
+      "epoch": 1.58,
+      "learning_rate": 9.941640836884929e-05,
+      "loss": 0.0003,
+      "step": 4830
     },
     {
+      "epoch": 1.65,
+      "learning_rate": 9.885945370011624e-05,
+      "loss": 0.0004,
+      "step": 5060
     },
     {
+      "epoch": 1.73,
+      "learning_rate": 9.830249903138319e-05,
+      "loss": 0.0003,
+      "step": 5290
     },
     {
+      "epoch": 1.8,
+      "learning_rate": 9.774554436265014e-05,
+      "loss": 0.0004,
+      "step": 5520
     },
     {
+      "epoch": 1.88,
+      "learning_rate": 9.718858969391709e-05,
+      "loss": 0.0003,
+      "step": 5750
     },
     {
+      "epoch": 1.95,
+      "learning_rate": 9.663163502518404e-05,
+      "loss": 0.0003,
+      "step": 5980
     },
     {
+      "epoch": 2.0,
+      "eval_loss": 0.0011327904649078846,
+      "eval_max_distance": 5,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.6053,
+      "eval_samples_per_second": 19.245,
+      "eval_steps_per_second": 0.131,
+      "step": 6118
     },
     {
+      "epoch": 2.03,
+      "learning_rate": 9.6074680356451e-05,
+      "loss": 0.0003,
+      "step": 6210
     },
     {
+      "epoch": 2.11,
+      "learning_rate": 9.551772568771794e-05,
+      "loss": 0.0003,
+      "step": 6440
     },
     {
+      "epoch": 2.18,
+      "learning_rate": 9.496077101898489e-05,
+      "loss": 0.0003,
+      "step": 6670
     },
     {
+      "epoch": 2.26,
+      "learning_rate": 9.440381635025185e-05,
+      "loss": 0.0003,
+      "step": 6900
     },
     {
+      "epoch": 2.33,
+      "learning_rate": 9.384686168151879e-05,
       "loss": 0.0004,
+      "step": 7130
     },
     {
+      "epoch": 2.41,
+      "learning_rate": 9.328990701278575e-05,
+      "loss": 0.0003,
+      "step": 7360
     },
     {
+      "epoch": 2.48,
+      "learning_rate": 9.273295234405269e-05,
+      "loss": 0.0003,
+      "step": 7590
     },
     {
+      "epoch": 2.56,
+      "learning_rate": 9.217599767531964e-05,
+      "loss": 0.0003,
+      "step": 7820
     },
     {
+      "epoch": 2.63,
+      "learning_rate": 9.16190430065866e-05,
+      "loss": 0.0003,
+      "step": 8050
     },
     {
+      "epoch": 2.71,
+      "learning_rate": 9.106208833785354e-05,
+      "loss": 0.0003,
+      "step": 8280
     },
     {
+      "epoch": 2.78,
+      "learning_rate": 9.05051336691205e-05,
+      "loss": 0.0003,
+      "step": 8510
     },
     {
+      "epoch": 2.86,
+      "learning_rate": 8.994817900038746e-05,
+      "loss": 0.0003,
+      "step": 8740
     },
     {
+      "epoch": 2.93,
+      "learning_rate": 8.939122433165439e-05,
+      "loss": 0.0003,
+      "step": 8970
     },
     {
+      "epoch": 3.0,
+      "eval_loss": 0.0010147562716156244,
+      "eval_max_distance": 6,
       "eval_mean_distance": 0,
+      "eval_runtime": 30.7374,
+      "eval_samples_per_second": 19.162,
+      "eval_steps_per_second": 0.13,
+      "step": 9177
     },
     {
+      "epoch": 3.01,
+      "learning_rate": 8.883426966292136e-05,
+      "loss": 0.0003,
+      "step": 9200
     },
     {
+      "epoch": 3.08,
+      "learning_rate": 8.82773149941883e-05,
+      "loss": 0.0002,
+      "step": 9430
     },
     {
+      "epoch": 3.16,
+      "learning_rate": 8.772036032545526e-05,
+      "loss": 0.0003,
+      "step": 9660
     },
     {
+      "epoch": 3.23,
+      "learning_rate": 8.716340565672221e-05,
+      "loss": 0.0002,
+      "step": 9890
     },
     {
+      "epoch": 3.31,
+      "learning_rate": 8.660645098798916e-05,
+      "loss": 0.0003,
+      "step": 10120
     },
     {
+      "epoch": 3.38,
+      "learning_rate": 8.604949631925611e-05,
+      "loss": 0.0003,
+      "step": 10350
     },
     {
+      "epoch": 3.46,
+      "learning_rate": 8.549254165052306e-05,
+      "loss": 0.0003,
+      "step": 10580
     },
     {
+      "epoch": 3.53,
+      "learning_rate": 8.493558698179001e-05,
+      "loss": 0.0003,
+      "step": 10810
     },
     {
+      "epoch": 3.61,
+      "learning_rate": 8.437863231305696e-05,
+      "loss": 0.0003,
+      "step": 11040
     },
     {
+      "epoch": 3.68,
+      "learning_rate": 8.382167764432391e-05,
+      "loss": 0.0003,
+      "step": 11270
     },
     {
+      "epoch": 3.76,
+      "learning_rate": 8.326472297559086e-05,
+      "loss": 0.0003,
+      "step": 11500
     },
     {
+      "epoch": 3.83,
+      "learning_rate": 8.270776830685781e-05,
+      "loss": 0.0003,
+      "step": 11730
     },
     {
+      "epoch": 3.91,
+      "learning_rate": 8.215081363812476e-05,
+      "loss": 0.0003,
+      "step": 11960
     },
     {
+      "epoch": 3.98,
+      "learning_rate": 8.159385896939171e-05,
+      "loss": 0.0003,
+      "step": 12190
     },
     {
+      "epoch": 4.0,
+      "eval_loss": 0.0011835404438897967,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.6791,
+      "eval_samples_per_second": 19.199,
+      "eval_steps_per_second": 0.13,
+      "step": 12236
     },
     {
+      "epoch": 4.06,
+      "learning_rate": 8.103690430065866e-05,
+      "loss": 0.0003,
+      "step": 12420
     },
     {
+      "epoch": 4.14,
+      "learning_rate": 8.047994963192561e-05,
+      "loss": 0.0003,
+      "step": 12650
     },
     {
+      "epoch": 4.21,
+      "learning_rate": 7.992299496319256e-05,
+      "loss": 0.0003,
+      "step": 12880
     },
     {
+      "epoch": 4.29,
+      "learning_rate": 7.936604029445953e-05,
+      "loss": 0.0002,
+      "step": 13110
     },
     {
+      "epoch": 4.36,
+      "learning_rate": 7.880908562572646e-05,
+      "loss": 0.0003,
+      "step": 13340
     },
     {
+      "epoch": 4.44,
+      "learning_rate": 7.825213095699341e-05,
+      "loss": 0.0003,
+      "step": 13570
     },
     {
+      "epoch": 4.51,
+      "learning_rate": 7.769517628826036e-05,
+      "loss": 0.0003,
+      "step": 13800
     },
     {
+      "epoch": 4.59,
+      "learning_rate": 7.713822161952731e-05,
+      "loss": 0.0003,
+      "step": 14030
     },
     {
+      "epoch": 4.66,
+      "learning_rate": 7.658126695079428e-05,
+      "loss": 0.0002,
+      "step": 14260
     },
     {
+      "epoch": 4.74,
+      "learning_rate": 7.602431228206122e-05,
+      "loss": 0.0003,
+      "step": 14490
     },
     {
+      "epoch": 4.81,
+      "learning_rate": 7.546735761332817e-05,
+      "loss": 0.0003,
+      "step": 14720
     },
     {
+      "epoch": 4.89,
+      "learning_rate": 7.491040294459513e-05,
+      "loss": 0.0002,
+      "step": 14950
     },
     {
+      "epoch": 4.96,
+      "learning_rate": 7.435344827586207e-05,
+      "loss": 0.0003,
+      "step": 15180
     },
     {
+      "epoch": 5.0,
+      "eval_loss": 0.000761075527407229,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.6946,
+      "eval_samples_per_second": 19.189,
+      "eval_steps_per_second": 0.13,
+      "step": 15295
     },
     {
+      "epoch": 5.04,
+      "learning_rate": 7.379649360712903e-05,
+      "loss": 0.0002,
+      "step": 15410
     },
     {
+      "epoch": 5.11,
+      "learning_rate": 7.323953893839598e-05,
+      "loss": 0.0002,
+      "step": 15640
     },
     {
+      "epoch": 5.19,
+      "learning_rate": 7.268258426966292e-05,
+      "loss": 0.0003,
+      "step": 15870
     },
     {
+      "epoch": 5.26,
+      "learning_rate": 7.212562960092988e-05,
+      "loss": 0.0002,
+      "step": 16100
     },
     {
+      "epoch": 5.34,
+      "learning_rate": 7.156867493219682e-05,
+      "loss": 0.0003,
+      "step": 16330
     },
     {
+      "epoch": 5.41,
+      "learning_rate": 7.101172026346378e-05,
+      "loss": 0.0002,
+      "step": 16560
     },
     {
+      "epoch": 5.49,
+      "learning_rate": 7.045476559473073e-05,
+      "loss": 0.0002,
+      "step": 16790
     },
     {
+      "epoch": 5.56,
+      "learning_rate": 6.989781092599767e-05,
+      "loss": 0.0003,
+      "step": 17020
     },
     {
+      "epoch": 5.64,
+      "learning_rate": 6.934085625726463e-05,
+      "loss": 0.0003,
+      "step": 17250
     },
     {
+      "epoch": 5.71,
+      "learning_rate": 6.878390158853158e-05,
       "loss": 0.0002,
+      "step": 17480
     },
     {
+      "epoch": 5.79,
+      "learning_rate": 6.822694691979853e-05,
+      "loss": 0.0003,
+      "step": 17710
     },
     {
+      "epoch": 5.86,
+      "learning_rate": 6.766999225106548e-05,
+      "loss": 0.0003,
+      "step": 17940
     },
     {
+      "epoch": 5.94,
+      "learning_rate": 6.711303758233242e-05,
+      "loss": 0.0002,
+      "step": 18170
     },
     {
+      "epoch": 6.0,
+      "eval_loss": 0.0009165782830677927,
+      "eval_max_distance": 3,
       "eval_mean_distance": 0,
+      "eval_runtime": 30.7001,
+      "eval_samples_per_second": 19.186,
+      "eval_steps_per_second": 0.13,
+      "step": 18354
     },
     {
+      "epoch": 6.02,
+      "learning_rate": 6.655608291359938e-05,
+      "loss": 0.0003,
+      "step": 18400
     },
     {
+      "epoch": 6.09,
+      "learning_rate": 6.599912824486634e-05,
+      "loss": 0.0002,
+      "step": 18630
     },
     {
+      "epoch": 6.17,
+      "learning_rate": 6.544217357613329e-05,
+      "loss": 0.0002,
+      "step": 18860
     },
     {
+      "epoch": 6.24,
+      "learning_rate": 6.488521890740024e-05,
+      "loss": 0.0003,
+      "step": 19090
     },
     {
+      "epoch": 6.32,
+      "learning_rate": 6.432826423866719e-05,
+      "loss": 0.0002,
+      "step": 19320
     },
     {
+      "epoch": 6.39,
+      "learning_rate": 6.377130956993414e-05,
+      "loss": 0.0002,
+      "step": 19550
     },
     {
+      "epoch": 6.47,
+      "learning_rate": 6.321435490120109e-05,
+      "loss": 0.0003,
+      "step": 19780
     },
     {
+      "epoch": 6.54,
+      "learning_rate": 6.265740023246804e-05,
+      "loss": 0.0002,
+      "step": 20010
     },
     {
+      "epoch": 6.62,
+      "learning_rate": 6.210044556373499e-05,
+      "loss": 0.0002,
+      "step": 20240
     },
     {
+      "epoch": 6.69,
+      "learning_rate": 6.154349089500194e-05,
+      "loss": 0.0003,
+      "step": 20470
     },
     {
+      "epoch": 6.77,
+      "learning_rate": 6.098653622626889e-05,
+      "loss": 0.0002,
+      "step": 20700
     },
     {
+      "epoch": 6.84,
+      "learning_rate": 6.042958155753584e-05,
+      "loss": 0.0002,
+      "step": 20930
     },
     {
+      "epoch": 6.92,
+      "learning_rate": 5.9872626888802796e-05,
+      "loss": 0.0003,
+      "step": 21160
     },
     {
+      "epoch": 6.99,
+      "learning_rate": 5.931567222006974e-05,
+      "loss": 0.0002,
+      "step": 21390
     },
     {
+      "epoch": 7.0,
+      "eval_loss": 0.0007532148738391697,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 31.1332,
+      "eval_samples_per_second": 18.919,
+      "eval_steps_per_second": 0.128,
+      "step": 21413
     },
     {
+      "epoch": 7.07,
+      "learning_rate": 5.87587175513367e-05,
+      "loss": 0.0003,
+      "step": 21620
     },
     {
+      "epoch": 7.14,
+      "learning_rate": 5.820176288260365e-05,
+      "loss": 0.0002,
+      "step": 21850
     },
     {
+      "epoch": 7.22,
+      "learning_rate": 5.764480821387059e-05,
+      "loss": 0.0002,
+      "step": 22080
     },
     {
+      "epoch": 7.29,
+      "learning_rate": 5.708785354513755e-05,
+      "loss": 0.0002,
+      "step": 22310
     },
     {
+      "epoch": 7.37,
+      "learning_rate": 5.653089887640449e-05,
+      "loss": 0.0003,
+      "step": 22540
     },
     {
+      "epoch": 7.44,
+      "learning_rate": 5.597394420767145e-05,
+      "loss": 0.0003,
+      "step": 22770
     },
     {
+      "epoch": 7.52,
+      "learning_rate": 5.54169895389384e-05,
+      "loss": 0.0002,
+      "step": 23000
     },
     {
+      "epoch": 7.59,
+      "learning_rate": 5.486003487020534e-05,
+      "loss": 0.0002,
+      "step": 23230
     },
     {
+      "epoch": 7.67,
+      "learning_rate": 5.43030802014723e-05,
+      "loss": 0.0003,
+      "step": 23460
     },
     {
+      "epoch": 7.74,
+      "learning_rate": 5.374612553273926e-05,
+      "loss": 0.0002,
+      "step": 23690
     },
     {
+      "epoch": 7.82,
+      "learning_rate": 5.31891708640062e-05,
+      "loss": 0.0002,
+      "step": 23920
     },
     {
+      "epoch": 7.89,
+      "learning_rate": 5.263221619527316e-05,
+      "loss": 0.0002,
+      "step": 24150
     },
     {
+      "epoch": 7.97,
+      "learning_rate": 5.20752615265401e-05,
+      "loss": 0.0002,
+      "step": 24380
     },
     {
+      "epoch": 8.0,
+      "eval_loss": 0.0007926285616122186,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.7407,
+      "eval_samples_per_second": 19.16,
+      "eval_steps_per_second": 0.13,
+      "step": 24472
     },
     {
+      "epoch": 8.05,
+      "learning_rate": 5.151830685780705e-05,
+      "loss": 0.0002,
+      "step": 24610
     },
     {
+      "epoch": 8.12,
+      "learning_rate": 5.096135218907401e-05,
+      "loss": 0.0002,
+      "step": 24840
     },
     {
+      "epoch": 8.2,
+      "learning_rate": 5.040439752034095e-05,
+      "loss": 0.0002,
+      "step": 25070
     },
     {
+      "epoch": 8.27,
+      "learning_rate": 4.984744285160791e-05,
+      "loss": 0.0002,
+      "step": 25300
     },
     {
+      "epoch": 8.35,
+      "learning_rate": 4.929048818287485e-05,
+      "loss": 0.0002,
+      "step": 25530
     },
     {
+      "epoch": 8.42,
+      "learning_rate": 4.873353351414181e-05,
+      "loss": 0.0002,
+      "step": 25760
     },
     {
+      "epoch": 8.5,
+      "learning_rate": 4.817657884540876e-05,
+      "loss": 0.0002,
+      "step": 25990
     },
     {
+      "epoch": 8.57,
+      "learning_rate": 4.761962417667571e-05,
+      "loss": 0.0002,
+      "step": 26220
     },
     {
+      "epoch": 8.65,
+      "learning_rate": 4.706266950794266e-05,
+      "loss": 0.0002,
+      "step": 26450
     },
     {
+      "epoch": 8.72,
+      "learning_rate": 4.650571483920961e-05,
+      "loss": 0.0002,
+      "step": 26680
     },
     {
+      "epoch": 8.8,
+      "learning_rate": 4.594876017047656e-05,
+      "loss": 0.0002,
+      "step": 26910
     },
     {
+      "epoch": 8.87,
+      "learning_rate": 4.539180550174351e-05,
+      "loss": 0.0003,
+      "step": 27140
     },
     {
+      "epoch": 8.95,
+      "learning_rate": 4.483485083301046e-05,
+      "loss": 0.0002,
+      "step": 27370
     },
     {
+      "epoch": 9.0,
+      "eval_loss": 0.0007057118928059936,
       "eval_max_distance": 3,
       "eval_mean_distance": 0,
+      "eval_runtime": 30.795,
+      "eval_samples_per_second": 19.127,
+      "eval_steps_per_second": 0.13,
+      "step": 27531
     },
     {
+      "epoch": 9.02,
+      "learning_rate": 4.427789616427741e-05,
+      "loss": 0.0002,
+      "step": 27600
     },
     {
+      "epoch": 9.1,
+      "learning_rate": 4.372094149554436e-05,
+      "loss": 0.0002,
+      "step": 27830
     },
     {
+      "epoch": 9.17,
+      "learning_rate": 4.3163986826811313e-05,
+      "loss": 0.0002,
+      "step": 28060
     },
     {
+      "epoch": 9.25,
+      "learning_rate": 4.2607032158078264e-05,
+      "loss": 0.0002,
+      "step": 28290
     },
     {
+      "epoch": 9.32,
+      "learning_rate": 4.205007748934522e-05,
+      "loss": 0.0002,
+      "step": 28520
     },
     {
+      "epoch": 9.4,
+      "learning_rate": 4.149312282061217e-05,
+      "loss": 0.0002,
+      "step": 28750
     },
     {
+      "epoch": 9.47,
+      "learning_rate": 4.0936168151879115e-05,
+      "loss": 0.0002,
+      "step": 28980
     },
     {
+      "epoch": 9.55,
+      "learning_rate": 4.0379213483146065e-05,
+      "loss": 0.0002,
+      "step": 29210
     },
     {
+      "epoch": 9.62,
+      "learning_rate": 3.982225881441302e-05,
+      "loss": 0.0002,
+      "step": 29440
     },
     {
+      "epoch": 9.7,
+      "learning_rate": 3.926530414567997e-05,
       "loss": 0.0003,
+      "step": 29670
     },
     {
+      "epoch": 9.77,
+      "learning_rate": 3.870834947694692e-05,
+      "loss": 0.0002,
+      "step": 29900
     },
     {
+      "epoch": 9.85,
+      "learning_rate": 3.8151394808213873e-05,
+      "loss": 0.0002,
+      "step": 30130
     },
     {
+      "epoch": 9.92,
+      "learning_rate": 3.7594440139480824e-05,
+      "loss": 0.0002,
+      "step": 30360
     },
     {
+      "epoch": 10.0,
+      "learning_rate": 3.7037485470747774e-05,
+      "loss": 0.0002,
+      "step": 30590
     },
     {
+      "epoch": 10.0,
+      "eval_loss": 0.000787499884609133,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.7843,
+      "eval_samples_per_second": 19.133,
+      "eval_steps_per_second": 0.13,
+      "step": 30590
     },
     {
+      "epoch": 10.08,
+      "learning_rate": 3.6480530802014724e-05,
+      "loss": 0.0002,
+      "step": 30820
     },
     {
+      "epoch": 10.15,
+      "learning_rate": 3.5923576133281675e-05,
+      "loss": 0.0002,
+      "step": 31050
     },
     {
+      "epoch": 10.23,
+      "learning_rate": 3.5366621464548625e-05,
+      "loss": 0.0002,
+      "step": 31280
     },
     {
+      "epoch": 10.3,
+      "learning_rate": 3.4809666795815576e-05,
+      "loss": 0.0002,
+      "step": 31510
     },
     {
+      "epoch": 10.38,
+      "learning_rate": 3.4252712127082526e-05,
+      "loss": 0.0002,
+      "step": 31740
     },
     {
+      "epoch": 10.45,
+      "learning_rate": 3.369575745834948e-05,
+      "loss": 0.0002,
+      "step": 31970
     },
     {
+      "epoch": 10.53,
+      "learning_rate": 3.313880278961643e-05,
       "loss": 0.0002,
+      "step": 32200
     },
     {
+      "epoch": 10.6,
+      "learning_rate": 3.258184812088338e-05,
+      "loss": 0.0002,
+      "step": 32430
     },
     {
+      "epoch": 10.68,
+      "learning_rate": 3.202489345215033e-05,
+      "loss": 0.0002,
+      "step": 32660
     },
     {
+      "epoch": 10.75,
+      "learning_rate": 3.1467938783417284e-05,
+      "loss": 0.0002,
+      "step": 32890
     },
     {
+      "epoch": 10.83,
+      "learning_rate": 3.0910984114684235e-05,
+      "loss": 0.0002,
+      "step": 33120
     },
     {
+      "epoch": 10.9,
+      "learning_rate": 3.0354029445951182e-05,
+      "loss": 0.0002,
+      "step": 33350
     },
     {
+      "epoch": 10.98,
+      "learning_rate": 2.9797074777218132e-05,
+      "loss": 0.0002,
+      "step": 33580
     },
     {
+      "epoch": 11.0,
+      "eval_loss": 0.000802784226834774,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.8756,
+      "eval_samples_per_second": 19.077,
+      "eval_steps_per_second": 0.13,
+      "step": 33649
     },
     {
+      "epoch": 11.05,
+      "learning_rate": 2.9240120108485086e-05,
+      "loss": 0.0002,
+      "step": 33810
     },
     {
+      "epoch": 11.13,
+      "learning_rate": 2.8683165439752036e-05,
+      "loss": 0.0002,
+      "step": 34040
     },
     {
+      "epoch": 11.2,
+      "learning_rate": 2.8126210771018983e-05,
+      "loss": 0.0002,
+      "step": 34270
     },
     {
+      "epoch": 11.28,
+      "learning_rate": 2.756925610228594e-05,
+      "loss": 0.0002,
+      "step": 34500
     },
     {
+      "epoch": 11.35,
+      "learning_rate": 2.7012301433552887e-05,
+      "loss": 0.0002,
+      "step": 34730
     },
     {
+      "epoch": 11.43,
+      "learning_rate": 2.6455346764819838e-05,
+      "loss": 0.0002,
+      "step": 34960
     },
     {
+      "epoch": 11.5,
+      "learning_rate": 2.5898392096086788e-05,
+      "loss": 0.0002,
+      "step": 35190
     },
     {
+      "epoch": 11.58,
+      "learning_rate": 2.5341437427353742e-05,
+      "loss": 0.0002,
+      "step": 35420
     },
     {
+      "epoch": 11.65,
+      "learning_rate": 2.4784482758620692e-05,
+      "loss": 0.0002,
+      "step": 35650
     },
     {
+      "epoch": 11.73,
+      "learning_rate": 2.4227528089887643e-05,
+      "loss": 0.0002,
+      "step": 35880
     },
     {
+      "epoch": 11.8,
+      "learning_rate": 2.3670573421154593e-05,
+      "loss": 0.0002,
+      "step": 36110
     },
     {
+      "epoch": 11.88,
+      "learning_rate": 2.3113618752421543e-05,
+      "loss": 0.0002,
+      "step": 36340
     },
     {
+      "epoch": 11.95,
+      "learning_rate": 2.2556664083688494e-05,
+      "loss": 0.0002,
+      "step": 36570
     },
     {
+      "epoch": 12.0,
+      "eval_loss": 0.0007881763740442693,
       "eval_max_distance": 3,
       "eval_mean_distance": 0,
+      "eval_runtime": 30.7194,
+      "eval_samples_per_second": 19.174,
+      "eval_steps_per_second": 0.13,
+      "step": 36708
     },
     {
+      "epoch": 12.03,
+      "learning_rate": 2.1999709414955444e-05,
+      "loss": 0.0002,
+      "step": 36800
     },
     {
+      "epoch": 12.11,
+      "learning_rate": 2.1442754746222398e-05,
+      "loss": 0.0002,
+      "step": 37030
     },
     {
+      "epoch": 12.18,
+      "learning_rate": 2.0885800077489345e-05,
+      "loss": 0.0002,
+      "step": 37260
     },
     {
+      "epoch": 12.26,
+      "learning_rate": 2.03288454087563e-05,
+      "loss": 0.0002,
+      "step": 37490
     },
     {
+      "epoch": 12.33,
+      "learning_rate": 1.9771890740023245e-05,
+      "loss": 0.0002,
+      "step": 37720
     },
     {
+      "epoch": 12.41,
+      "learning_rate": 1.92149360712902e-05,
       "loss": 0.0002,
+      "step": 37950
     },
     {
+      "epoch": 12.48,
+      "learning_rate": 1.865798140255715e-05,
+      "loss": 0.0002,
+      "step": 38180
     },
     {
+      "epoch": 12.56,
+      "learning_rate": 1.81010267338241e-05,
+      "loss": 0.0002,
+      "step": 38410
     },
     {
+      "epoch": 12.63,
+      "learning_rate": 1.754407206509105e-05,
+      "loss": 0.0002,
+      "step": 38640
     },
     {
+      "epoch": 12.71,
+      "learning_rate": 1.6987117396358e-05,
+      "loss": 0.0002,
+      "step": 38870
     },
     {
+      "epoch": 12.78,
+      "learning_rate": 1.6430162727624954e-05,
+      "loss": 0.0002,
+      "step": 39100
     },
     {
+      "epoch": 12.86,
+      "learning_rate": 1.5873208058891905e-05,
+      "loss": 0.0002,
+      "step": 39330
     },
     {
+      "epoch": 12.93,
+      "learning_rate": 1.5316253390158855e-05,
+      "loss": 0.0002,
+      "step": 39560
     },
     {
+      "epoch": 13.0,
+      "eval_loss": 0.000838827807456255,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.7487,
+      "eval_samples_per_second": 19.155,
+      "eval_steps_per_second": 0.13,
+      "step": 39767
     },
     {
+      "epoch": 13.01,
+      "learning_rate": 1.4759298721425804e-05,
+      "loss": 0.0002,
+      "step": 39790
     },
     {
+      "epoch": 13.08,
+      "learning_rate": 1.4202344052692756e-05,
+      "loss": 0.0002,
+      "step": 40020
     },
     {
+      "epoch": 13.16,
+      "learning_rate": 1.3645389383959706e-05,
+      "loss": 0.0002,
+      "step": 40250
     },
     {
+      "epoch": 13.23,
+      "learning_rate": 1.3088434715226658e-05,
+      "loss": 0.0002,
+      "step": 40480
     },
     {
+      "epoch": 13.31,
+      "learning_rate": 1.2531480046493607e-05,
+      "loss": 0.0002,
+      "step": 40710
     },
     {
+      "epoch": 13.38,
+      "learning_rate": 1.1974525377760557e-05,
       "loss": 0.0002,
+      "step": 40940
     },
     {
+      "epoch": 13.46,
+      "learning_rate": 1.141757070902751e-05,
+      "loss": 0.0002,
+      "step": 41170
     },
     {
+      "epoch": 13.53,
+      "learning_rate": 1.0860616040294461e-05,
+      "loss": 0.0002,
+      "step": 41400
     },
     {
+      "epoch": 13.61,
+      "learning_rate": 1.0303661371561412e-05,
+      "loss": 0.0002,
+      "step": 41630
     },
     {
+      "epoch": 13.68,
+      "learning_rate": 9.746706702828362e-06,
+      "loss": 0.0002,
+      "step": 41860
     },
     {
+      "epoch": 13.76,
+      "learning_rate": 9.189752034095312e-06,
+      "loss": 0.0002,
+      "step": 42090
     },
     {
+      "epoch": 13.83,
+      "learning_rate": 8.632797365362263e-06,
+      "loss": 0.0002,
+      "step": 42320
     },
     {
+      "epoch": 13.91,
+      "learning_rate": 8.075842696629215e-06,
+      "loss": 0.0002,
+      "step": 42550
     },
     {
+      "epoch": 13.98,
+      "learning_rate": 7.518888027896165e-06,
+      "loss": 0.0002,
+      "step": 42780
     },
     {
+      "epoch": 14.0,
+      "eval_loss": 0.0008361997315660119,
+      "eval_max_distance": 3,
+      "eval_mean_distance": 0,
+      "eval_runtime": 30.9876,
+      "eval_samples_per_second": 19.008,
+      "eval_steps_per_second": 0.129,
+      "step": 42826
     },
     {
+      "epoch": 14.06,
+      "learning_rate": 6.9619333591631155e-06,
+      "loss": 0.0002,
+      "step": 43010
     },
     {
+      "epoch": 14.14,
+      "learning_rate": 6.404978690430067e-06,
+      "loss": 0.0002,
+      "step": 43240
     },
     {
+      "epoch": 14.21,
+      "learning_rate": 5.848024021697017e-06,
+      "loss": 0.0002,
+      "step": 43470
     },
     {
+      "epoch": 14.29,
+      "learning_rate": 5.291069352963967e-06,
+      "loss": 0.0002,
+      "step": 43700
     },
     {
+      "epoch": 14.36,
+      "learning_rate": 4.7341146842309186e-06,
+      "loss": 0.0002,
+      "step": 43930
     },
     {
+      "epoch": 14.44,
+      "learning_rate": 4.177160015497869e-06,
+      "loss": 0.0002,
+      "step": 44160
     },
     {
+      "epoch": 14.51,
+      "learning_rate": 3.6202053467648197e-06,
+      "loss": 0.0002,
+      "step": 44390
     },
     {
+      "epoch": 14.59,
+      "learning_rate": 3.063250678031771e-06,
+      "loss": 0.0002,
+      "step": 44620
     },
     {
+      "epoch": 14.66,
+      "learning_rate": 2.5062960092987217e-06,
+      "loss": 0.0002,
+      "step": 44850
     },
     {
+      "epoch": 14.74,
+      "learning_rate": 1.949341340565672e-06,
+      "loss": 0.0002,
+      "step": 45080
     },
     {
+      "epoch": 14.81,
+      "learning_rate": 1.392386671832623e-06,
+      "loss": 0.0002,
+      "step": 45310
     },
     {
+      "epoch": 14.89,
+      "learning_rate": 8.354320030995738e-07,
+      "loss": 0.0002,
+      "step": 45540
     },
     {
+      "epoch": 14.96,
+      "learning_rate": 2.784773343665246e-07,
+      "loss": 0.0002,
+      "step": 45770
     },
     {
+      "epoch": 15.0,
+      "eval_loss": 0.0008401814266107976,
+      "eval_max_distance": 3,
       "eval_mean_distance": 0,
+      "eval_runtime": 30.8673,
+      "eval_samples_per_second": 19.082,
+      "eval_steps_per_second": 0.13,
+      "step": 45885
+    },
+    {
+      "epoch": 15.0,
+      "step": 45885,
+      "total_flos": 7.988818341631795e+16,
+      "train_loss": 0.0002746593983470409,
+      "train_runtime": 6476.6732,
+      "train_samples_per_second": 531.266,
+      "train_steps_per_second": 7.085
     }
   ],
+  "logging_steps": 230,
+  "max_steps": 45885,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 15,
+  "save_steps": 4589,
+  "total_flos": 7.988818341631795e+16,
+  "train_batch_size": 75,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c1715be697a9be6fe0784bf769267cffcc821d071738c8fdd99dec6f06f39db0
-size 4091

 version https://git-lfs.github.com/spec/v1
+oid sha256:1eb73bbd8d63a4a4a349732132a9ced8e0528dece7669b9d0a3ba1f248d5ec37
+size 4792