Training in progress, step 16, checkpoint

last-checkpoint/config.json

@@ -1,12 +1,12 @@
 {
-  "_name_or_path": "google-t5/t5-small",
+  "_name_or_path": "VietAI/vit5-base",
   "architectures": [
     "T5ForConditionalGeneration"
   ],
   "classifier_dropout": 0.0,
-  "d_ff": 2048,
+  "d_ff": 3072,
   "d_kv": 64,
-  "d_model": 512,
+  "d_model": 768,
   "decoder_start_token_id": 0,
   "dense_act_fn": "relu",
   "dropout_rate": 0.1,
@@ -18,44 +18,15 @@
   "layer_norm_epsilon": 1e-06,
   "model_type": "t5",
   "n_positions": 512,
-  "num_decoder_layers": 6,
-  "num_heads": 8,
-  "num_layers": 6,
+  "num_decoder_layers": 12,
+  "num_heads": 12,
+  "num_layers": 12,
   "output_past": true,
   "pad_token_id": 0,
   "relative_attention_max_distance": 128,
   "relative_attention_num_buckets": 32,
-  "task_specific_params": {
-    "summarization": {
-      "early_stopping": true,
-      "length_penalty": 2.0,
-      "max_length": 200,
-      "min_length": 30,
-      "no_repeat_ngram_size": 3,
-      "num_beams": 4,
-      "prefix": "summarize: "
-    },
-    "translation_en_to_de": {
-      "early_stopping": true,
-      "max_length": 300,
-      "num_beams": 4,
-      "prefix": "translate English to German: "
-    },
-    "translation_en_to_fr": {
-      "early_stopping": true,
-      "max_length": 300,
-      "num_beams": 4,
-      "prefix": "translate English to French: "
-    },
-    "translation_en_to_ro": {
-      "early_stopping": true,
-      "max_length": 300,
-      "num_beams": 4,
-      "prefix": "translate English to Romanian: "
-    }
-  },
   "torch_dtype": "float32",
   "transformers_version": "4.42.3",
   "use_cache": true,
-  "vocab_size": 32128
+  "vocab_size": 36096
 }

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:67cde00778fd9b31e92769f2fa5fb22957b3c18857b8fa15489d34bf37fd4392
-size 242041896
+oid sha256:119eea453bd7b4029fcdef03fe0a7106f3b8552ddeebbf478f83f4f695d0bc27
+size 903834408

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a268ea2d87ebd5586132b48f679290592ca529d7cd515b00d80780d58f6b8ca
-size 484163514
+oid sha256:097f2da6ca0542c33e534858431ad08295fb5b03025fd8e94bda6c3ada974000
+size 1807824186

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:369ba992403565a78cd72e1427835ccc8b05eba0f71afd8552e75e7e6ea369c4
+oid sha256:2fd40d17872e1d1f44fd19772092ee3ba49e037d722a9306b6de7d846efa46dc
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ba478b7a054120870328b778675384e38480f575773ac8971e03a9ab54918ff3
+oid sha256:e37c62b1a4bff2525d39b9bf2be166315395310f82e667777248052632e706f7
 size 1064

last-checkpoint/special_tokens_map.json

@@ -95,31 +95,9 @@
     "<extra_id_92>",
     "<extra_id_93>",
     "<extra_id_94>",
-    "<extra_id_95>",
-    "<extra_id_96>",
-    "<extra_id_97>",
-    "<extra_id_98>",
-    "<extra_id_99>"
+    "<extra_id_95>"
   ],
-  "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
-  }
+  "eos_token": "</s>",
+  "pad_token": "<pad>",
+  "unk_token": "<unk>"
 }

last-checkpoint/spiece.model

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d60acb128cf7b7f2536e8f38a5b18a05535c9e14c7a355904270e15b0945ea86
-size 791656
+oid sha256:59986b62f9f0b90edafb9b073ea7b93d21114a5841219a1ea2399ade73f729c6
+size 820370

last-checkpoint/tokenizer_config.json

@@ -24,39 +24,7 @@
       "single_word": false,
       "special": true
     },
-    "32000": {
-      "content": "<extra_id_99>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
-    },
-    "32001": {
-      "content": "<extra_id_98>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
-    },
-    "32002": {
-      "content": "<extra_id_97>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
-    },
-    "32003": {
-      "content": "<extra_id_96>",
-      "lstrip": false,
-      "normalized": false,
-      "rstrip": false,
-      "single_word": false,
-      "special": true
-    },
-    "32004": {
+    "36000": {
       "content": "<extra_id_95>",
       "lstrip": false,
       "normalized": false,
@@ -64,7 +32,7 @@
       "single_word": false,
       "special": true
     },
-    "32005": {
+    "36001": {
       "content": "<extra_id_94>",
       "lstrip": false,
       "normalized": false,
@@ -72,7 +40,7 @@
       "single_word": false,
       "special": true
     },
-    "32006": {
+    "36002": {
       "content": "<extra_id_93>",
       "lstrip": false,
       "normalized": false,
@@ -80,7 +48,7 @@
       "single_word": false,
       "special": true
     },
-    "32007": {
+    "36003": {
       "content": "<extra_id_92>",
       "lstrip": false,
       "normalized": false,
@@ -88,7 +56,7 @@
       "single_word": false,
       "special": true
     },
-    "32008": {
+    "36004": {
       "content": "<extra_id_91>",
       "lstrip": false,
       "normalized": false,
@@ -96,7 +64,7 @@
       "single_word": false,
       "special": true
     },
-    "32009": {
+    "36005": {
       "content": "<extra_id_90>",
       "lstrip": false,
       "normalized": false,
@@ -104,7 +72,7 @@
       "single_word": false,
       "special": true
     },
-    "32010": {
+    "36006": {
       "content": "<extra_id_89>",
       "lstrip": false,
       "normalized": false,
@@ -112,7 +80,7 @@
       "single_word": false,
       "special": true
     },
-    "32011": {
+    "36007": {
       "content": "<extra_id_88>",
       "lstrip": false,
       "normalized": false,
@@ -120,7 +88,7 @@
       "single_word": false,
       "special": true
     },
-    "32012": {
+    "36008": {
       "content": "<extra_id_87>",
       "lstrip": false,
       "normalized": false,
@@ -128,7 +96,7 @@
       "single_word": false,
       "special": true
     },
-    "32013": {
+    "36009": {
       "content": "<extra_id_86>",
       "lstrip": false,
       "normalized": false,
@@ -136,7 +104,7 @@
       "single_word": false,
       "special": true
     },
-    "32014": {
+    "36010": {
       "content": "<extra_id_85>",
       "lstrip": false,
       "normalized": false,
@@ -144,7 +112,7 @@
       "single_word": false,
       "special": true
     },
-    "32015": {
+    "36011": {
       "content": "<extra_id_84>",
       "lstrip": false,
       "normalized": false,
@@ -152,7 +120,7 @@
       "single_word": false,
       "special": true
     },
-    "32016": {
+    "36012": {
       "content": "<extra_id_83>",
       "lstrip": false,
       "normalized": false,
@@ -160,7 +128,7 @@
       "single_word": false,
       "special": true
     },
-    "32017": {
+    "36013": {
       "content": "<extra_id_82>",
       "lstrip": false,
       "normalized": false,
@@ -168,7 +136,7 @@
       "single_word": false,
       "special": true
     },
-    "32018": {
+    "36014": {
       "content": "<extra_id_81>",
       "lstrip": false,
       "normalized": false,
@@ -176,7 +144,7 @@
       "single_word": false,
       "special": true
     },
-    "32019": {
+    "36015": {
       "content": "<extra_id_80>",
       "lstrip": false,
       "normalized": false,
@@ -184,7 +152,7 @@
       "single_word": false,
       "special": true
     },
-    "32020": {
+    "36016": {
       "content": "<extra_id_79>",
       "lstrip": false,
       "normalized": false,
@@ -192,7 +160,7 @@
       "single_word": false,
       "special": true
     },
-    "32021": {
+    "36017": {
       "content": "<extra_id_78>",
       "lstrip": false,
       "normalized": false,
@@ -200,7 +168,7 @@
       "single_word": false,
       "special": true
     },
-    "32022": {
+    "36018": {
       "content": "<extra_id_77>",
       "lstrip": false,
       "normalized": false,
@@ -208,7 +176,7 @@
       "single_word": false,
       "special": true
     },
-    "32023": {
+    "36019": {
       "content": "<extra_id_76>",
       "lstrip": false,
       "normalized": false,
@@ -216,7 +184,7 @@
       "single_word": false,
       "special": true
     },
-    "32024": {
+    "36020": {
       "content": "<extra_id_75>",
       "lstrip": false,
       "normalized": false,
@@ -224,7 +192,7 @@
       "single_word": false,
       "special": true
     },
-    "32025": {
+    "36021": {
       "content": "<extra_id_74>",
       "lstrip": false,
       "normalized": false,
@@ -232,7 +200,7 @@
       "single_word": false,
       "special": true
     },
-    "32026": {
+    "36022": {
       "content": "<extra_id_73>",
       "lstrip": false,
       "normalized": false,
@@ -240,7 +208,7 @@
       "single_word": false,
       "special": true
     },
-    "32027": {
+    "36023": {
       "content": "<extra_id_72>",
       "lstrip": false,
       "normalized": false,
@@ -248,7 +216,7 @@
       "single_word": false,
       "special": true
     },
-    "32028": {
+    "36024": {
       "content": "<extra_id_71>",
       "lstrip": false,
       "normalized": false,
@@ -256,7 +224,7 @@
       "single_word": false,
       "special": true
     },
-    "32029": {
+    "36025": {
       "content": "<extra_id_70>",
       "lstrip": false,
       "normalized": false,
@@ -264,7 +232,7 @@
       "single_word": false,
       "special": true
     },
-    "32030": {
+    "36026": {
       "content": "<extra_id_69>",
       "lstrip": false,
       "normalized": false,
@@ -272,7 +240,7 @@
       "single_word": false,
       "special": true
     },
-    "32031": {
+    "36027": {
       "content": "<extra_id_68>",
       "lstrip": false,
       "normalized": false,
@@ -280,7 +248,7 @@
       "single_word": false,
       "special": true
     },
-    "32032": {
+    "36028": {
       "content": "<extra_id_67>",
       "lstrip": false,
       "normalized": false,
@@ -288,7 +256,7 @@
       "single_word": false,
       "special": true
     },
-    "32033": {
+    "36029": {
       "content": "<extra_id_66>",
       "lstrip": false,
       "normalized": false,
@@ -296,7 +264,7 @@
       "single_word": false,
       "special": true
     },
-    "32034": {
+    "36030": {
       "content": "<extra_id_65>",
       "lstrip": false,
       "normalized": false,
@@ -304,7 +272,7 @@
       "single_word": false,
       "special": true
     },
-    "32035": {
+    "36031": {
       "content": "<extra_id_64>",
       "lstrip": false,
       "normalized": false,
@@ -312,7 +280,7 @@
       "single_word": false,
       "special": true
     },
-    "32036": {
+    "36032": {
       "content": "<extra_id_63>",
       "lstrip": false,
       "normalized": false,
@@ -320,7 +288,7 @@
       "single_word": false,
       "special": true
     },
-    "32037": {
+    "36033": {
       "content": "<extra_id_62>",
       "lstrip": false,
       "normalized": false,
@@ -328,7 +296,7 @@
       "single_word": false,
       "special": true
     },
-    "32038": {
+    "36034": {
       "content": "<extra_id_61>",
       "lstrip": false,
       "normalized": false,
@@ -336,7 +304,7 @@
       "single_word": false,
       "special": true
     },
-    "32039": {
+    "36035": {
       "content": "<extra_id_60>",
       "lstrip": false,
       "normalized": false,
@@ -344,7 +312,7 @@
       "single_word": false,
       "special": true
     },
-    "32040": {
+    "36036": {
       "content": "<extra_id_59>",
       "lstrip": false,
       "normalized": false,
@@ -352,7 +320,7 @@
       "single_word": false,
       "special": true
     },
-    "32041": {
+    "36037": {
       "content": "<extra_id_58>",
       "lstrip": false,
       "normalized": false,
@@ -360,7 +328,7 @@
       "single_word": false,
       "special": true
     },
-    "32042": {
+    "36038": {
       "content": "<extra_id_57>",
       "lstrip": false,
       "normalized": false,
@@ -368,7 +336,7 @@
       "single_word": false,
       "special": true
     },
-    "32043": {
+    "36039": {
       "content": "<extra_id_56>",
       "lstrip": false,
       "normalized": false,
@@ -376,7 +344,7 @@
       "single_word": false,
       "special": true
     },
-    "32044": {
+    "36040": {
       "content": "<extra_id_55>",
       "lstrip": false,
       "normalized": false,
@@ -384,7 +352,7 @@
       "single_word": false,
       "special": true
     },
-    "32045": {
+    "36041": {
       "content": "<extra_id_54>",
       "lstrip": false,
       "normalized": false,
@@ -392,7 +360,7 @@
       "single_word": false,
       "special": true
     },
-    "32046": {
+    "36042": {
       "content": "<extra_id_53>",
       "lstrip": false,
       "normalized": false,
@@ -400,7 +368,7 @@
       "single_word": false,
       "special": true
     },
-    "32047": {
+    "36043": {
       "content": "<extra_id_52>",
       "lstrip": false,
       "normalized": false,
@@ -408,7 +376,7 @@
       "single_word": false,
       "special": true
     },
-    "32048": {
+    "36044": {
       "content": "<extra_id_51>",
       "lstrip": false,
       "normalized": false,
@@ -416,7 +384,7 @@
       "single_word": false,
       "special": true
     },
-    "32049": {
+    "36045": {
       "content": "<extra_id_50>",
       "lstrip": false,
       "normalized": false,
@@ -424,7 +392,7 @@
       "single_word": false,
       "special": true
     },
-    "32050": {
+    "36046": {
       "content": "<extra_id_49>",
       "lstrip": false,
       "normalized": false,
@@ -432,7 +400,7 @@
       "single_word": false,
       "special": true
     },
-    "32051": {
+    "36047": {
       "content": "<extra_id_48>",
       "lstrip": false,
       "normalized": false,
@@ -440,7 +408,7 @@
       "single_word": false,
       "special": true
     },
-    "32052": {
+    "36048": {
       "content": "<extra_id_47>",
       "lstrip": false,
       "normalized": false,
@@ -448,7 +416,7 @@
       "single_word": false,
       "special": true
     },
-    "32053": {
+    "36049": {
       "content": "<extra_id_46>",
       "lstrip": false,
       "normalized": false,
@@ -456,7 +424,7 @@
       "single_word": false,
       "special": true
     },
-    "32054": {
+    "36050": {
       "content": "<extra_id_45>",
       "lstrip": false,
       "normalized": false,
@@ -464,7 +432,7 @@
       "single_word": false,
       "special": true
     },
-    "32055": {
+    "36051": {
       "content": "<extra_id_44>",
       "lstrip": false,
       "normalized": false,
@@ -472,7 +440,7 @@
       "single_word": false,
       "special": true
     },
-    "32056": {
+    "36052": {
       "content": "<extra_id_43>",
       "lstrip": false,
       "normalized": false,
@@ -480,7 +448,7 @@
       "single_word": false,
       "special": true
     },
-    "32057": {
+    "36053": {
       "content": "<extra_id_42>",
       "lstrip": false,
       "normalized": false,
@@ -488,7 +456,7 @@
       "single_word": false,
       "special": true
     },
-    "32058": {
+    "36054": {
       "content": "<extra_id_41>",
       "lstrip": false,
       "normalized": false,
@@ -496,7 +464,7 @@
       "single_word": false,
       "special": true
     },
-    "32059": {
+    "36055": {
       "content": "<extra_id_40>",
       "lstrip": false,
       "normalized": false,
@@ -504,7 +472,7 @@
       "single_word": false,
       "special": true
     },
-    "32060": {
+    "36056": {
       "content": "<extra_id_39>",
       "lstrip": false,
       "normalized": false,
@@ -512,7 +480,7 @@
       "single_word": false,
       "special": true
     },
-    "32061": {
+    "36057": {
       "content": "<extra_id_38>",
       "lstrip": false,
       "normalized": false,
@@ -520,7 +488,7 @@
       "single_word": false,
       "special": true
     },
-    "32062": {
+    "36058": {
       "content": "<extra_id_37>",
       "lstrip": false,
       "normalized": false,
@@ -528,7 +496,7 @@
       "single_word": false,
       "special": true
     },
-    "32063": {
+    "36059": {
       "content": "<extra_id_36>",
       "lstrip": false,
       "normalized": false,
@@ -536,7 +504,7 @@
       "single_word": false,
       "special": true
     },
-    "32064": {
+    "36060": {
       "content": "<extra_id_35>",
       "lstrip": false,
       "normalized": false,
@@ -544,7 +512,7 @@
       "single_word": false,
       "special": true
     },
-    "32065": {
+    "36061": {
       "content": "<extra_id_34>",
       "lstrip": false,
       "normalized": false,
@@ -552,7 +520,7 @@
       "single_word": false,
       "special": true
     },
-    "32066": {
+    "36062": {
       "content": "<extra_id_33>",
       "lstrip": false,
       "normalized": false,
@@ -560,7 +528,7 @@
       "single_word": false,
       "special": true
     },
-    "32067": {
+    "36063": {
       "content": "<extra_id_32>",
       "lstrip": false,
       "normalized": false,
@@ -568,7 +536,7 @@
       "single_word": false,
       "special": true
     },
-    "32068": {
+    "36064": {
       "content": "<extra_id_31>",
       "lstrip": false,
       "normalized": false,
@@ -576,7 +544,7 @@
       "single_word": false,
       "special": true
     },
-    "32069": {
+    "36065": {
       "content": "<extra_id_30>",
       "lstrip": false,
       "normalized": false,
@@ -584,7 +552,7 @@
       "single_word": false,
       "special": true
     },
-    "32070": {
+    "36066": {
       "content": "<extra_id_29>",
       "lstrip": false,
       "normalized": false,
@@ -592,7 +560,7 @@
       "single_word": false,
       "special": true
     },
-    "32071": {
+    "36067": {
       "content": "<extra_id_28>",
       "lstrip": false,
       "normalized": false,
@@ -600,7 +568,7 @@
       "single_word": false,
       "special": true
     },
-    "32072": {
+    "36068": {
       "content": "<extra_id_27>",
       "lstrip": false,
       "normalized": false,
@@ -608,7 +576,7 @@
       "single_word": false,
       "special": true
     },
-    "32073": {
+    "36069": {
       "content": "<extra_id_26>",
       "lstrip": false,
       "normalized": false,
@@ -616,7 +584,7 @@
       "single_word": false,
       "special": true
     },
-    "32074": {
+    "36070": {
       "content": "<extra_id_25>",
       "lstrip": false,
       "normalized": false,
@@ -624,7 +592,7 @@
       "single_word": false,
       "special": true
     },
-    "32075": {
+    "36071": {
       "content": "<extra_id_24>",
       "lstrip": false,
       "normalized": false,
@@ -632,7 +600,7 @@
       "single_word": false,
       "special": true
     },
-    "32076": {
+    "36072": {
       "content": "<extra_id_23>",
       "lstrip": false,
       "normalized": false,
@@ -640,7 +608,7 @@
       "single_word": false,
       "special": true
     },
-    "32077": {
+    "36073": {
       "content": "<extra_id_22>",
       "lstrip": false,
       "normalized": false,
@@ -648,7 +616,7 @@
       "single_word": false,
       "special": true
     },
-    "32078": {
+    "36074": {
       "content": "<extra_id_21>",
       "lstrip": false,
       "normalized": false,
@@ -656,7 +624,7 @@
       "single_word": false,
       "special": true
     },
-    "32079": {
+    "36075": {
       "content": "<extra_id_20>",
       "lstrip": false,
       "normalized": false,
@@ -664,7 +632,7 @@
       "single_word": false,
       "special": true
     },
-    "32080": {
+    "36076": {
       "content": "<extra_id_19>",
       "lstrip": false,
       "normalized": false,
@@ -672,7 +640,7 @@
       "single_word": false,
       "special": true
     },
-    "32081": {
+    "36077": {
       "content": "<extra_id_18>",
       "lstrip": false,
       "normalized": false,
@@ -680,7 +648,7 @@
       "single_word": false,
       "special": true
     },
-    "32082": {
+    "36078": {
       "content": "<extra_id_17>",
       "lstrip": false,
       "normalized": false,
@@ -688,7 +656,7 @@
       "single_word": false,
       "special": true
     },
-    "32083": {
+    "36079": {
       "content": "<extra_id_16>",
       "lstrip": false,
       "normalized": false,
@@ -696,7 +664,7 @@
       "single_word": false,
       "special": true
     },
-    "32084": {
+    "36080": {
       "content": "<extra_id_15>",
       "lstrip": false,
       "normalized": false,
@@ -704,7 +672,7 @@
       "single_word": false,
       "special": true
     },
-    "32085": {
+    "36081": {
       "content": "<extra_id_14>",
       "lstrip": false,
       "normalized": false,
@@ -712,7 +680,7 @@
       "single_word": false,
       "special": true
     },
-    "32086": {
+    "36082": {
       "content": "<extra_id_13>",
       "lstrip": false,
       "normalized": false,
@@ -720,7 +688,7 @@
       "single_word": false,
       "special": true
     },
-    "32087": {
+    "36083": {
       "content": "<extra_id_12>",
       "lstrip": false,
       "normalized": false,
@@ -728,7 +696,7 @@
       "single_word": false,
       "special": true
     },
-    "32088": {
+    "36084": {
       "content": "<extra_id_11>",
       "lstrip": false,
       "normalized": false,
@@ -736,7 +704,7 @@
       "single_word": false,
       "special": true
     },
-    "32089": {
+    "36085": {
       "content": "<extra_id_10>",
       "lstrip": false,
       "normalized": false,
@@ -744,7 +712,7 @@
       "single_word": false,
       "special": true
     },
-    "32090": {
+    "36086": {
       "content": "<extra_id_9>",
       "lstrip": false,
       "normalized": false,
@@ -752,7 +720,7 @@
       "single_word": false,
       "special": true
     },
-    "32091": {
+    "36087": {
       "content": "<extra_id_8>",
       "lstrip": false,
       "normalized": false,
@@ -760,7 +728,7 @@
       "single_word": false,
       "special": true
     },
-    "32092": {
+    "36088": {
       "content": "<extra_id_7>",
       "lstrip": false,
       "normalized": false,
@@ -768,7 +736,7 @@
       "single_word": false,
       "special": true
     },
-    "32093": {
+    "36089": {
       "content": "<extra_id_6>",
       "lstrip": false,
       "normalized": false,
@@ -776,7 +744,7 @@
       "single_word": false,
       "special": true
     },
-    "32094": {
+    "36090": {
       "content": "<extra_id_5>",
       "lstrip": false,
       "normalized": false,
@@ -784,7 +752,7 @@
       "single_word": false,
       "special": true
     },
-    "32095": {
+    "36091": {
       "content": "<extra_id_4>",
       "lstrip": false,
       "normalized": false,
@@ -792,7 +760,7 @@
       "single_word": false,
       "special": true
     },
-    "32096": {
+    "36092": {
       "content": "<extra_id_3>",
       "lstrip": false,
       "normalized": false,
@@ -800,7 +768,7 @@
       "single_word": false,
       "special": true
     },
-    "32097": {
+    "36093": {
       "content": "<extra_id_2>",
       "lstrip": false,
       "normalized": false,
@@ -808,7 +776,7 @@
       "single_word": false,
       "special": true
     },
-    "32098": {
+    "36094": {
       "content": "<extra_id_1>",
       "lstrip": false,
       "normalized": false,
@@ -816,7 +784,7 @@
       "single_word": false,
       "special": true
     },
-    "32099": {
+    "36095": {
       "content": "<extra_id_0>",
       "lstrip": false,
       "normalized": false,
@@ -921,17 +889,14 @@
     "<extra_id_92>",
     "<extra_id_93>",
     "<extra_id_94>",
-    "<extra_id_95>",
-    "<extra_id_96>",
-    "<extra_id_97>",
-    "<extra_id_98>",
-    "<extra_id_99>"
+    "<extra_id_95>"
   ],
   "clean_up_tokenization_spaces": true,
   "eos_token": "</s>",
-  "extra_ids": 100,
-  "model_max_length": 512,
+  "extra_ids": 96,
+  "model_max_length": 1000000000000000019884624838656,
   "pad_token": "<pad>",
+  "sp_model_kwargs": {},
   "tokenizer_class": "T5Tokenizer",
   "unk_token": "<unk>"
 }

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.013446464646464646,
+  "epoch": 0.0010343434343434343,
   "eval_steps": 16,
-  "global_step": 208,
+  "global_step": 16,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -12,1580 +12,128 @@
       "epoch": 6.464646464646465e-05,
       "grad_norm": NaN,
       "learning_rate": 0.0,
-      "loss": 12.4995,
+      "loss": 29.6213,
       "step": 1
     },
     {
       "epoch": 0.0001292929292929293,
       "grad_norm": NaN,
       "learning_rate": 0.0,
-      "loss": 12.5063,
+      "loss": 29.6208,
       "step": 2
     },
     {
       "epoch": 0.00019393939393939395,
-      "grad_norm": 68.83582305908203,
-      "learning_rate": 1.2903225806451614e-06,
-      "loss": 12.8615,
+      "grad_norm": 129.40235900878906,
+      "learning_rate": 4.3010752688172043e-07,
+      "loss": 29.5846,
       "step": 3
     },
     {
       "epoch": 0.0002585858585858586,
-      "grad_norm": 57.923553466796875,
-      "learning_rate": 2.580645161290323e-06,
-      "loss": 12.6171,
+      "grad_norm": Infinity,
+      "learning_rate": 4.3010752688172043e-07,
+      "loss": 29.7161,
       "step": 4
     },
     {
       "epoch": 0.00032323232323232324,
-      "grad_norm": NaN,
-      "learning_rate": 2.580645161290323e-06,
-      "loss": 12.4297,
+      "grad_norm": 130.79031372070312,
+      "learning_rate": 8.602150537634409e-07,
+      "loss": 29.7196,
       "step": 5
     },
     {
       "epoch": 0.0003878787878787879,
-      "grad_norm": 61.84542465209961,
-      "learning_rate": 3.870967741935484e-06,
-      "loss": 13.0576,
+      "grad_norm": 123.62369537353516,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 29.2487,
       "step": 6
     },
     {
       "epoch": 0.0004525252525252525,
-      "grad_norm": 54.698020935058594,
-      "learning_rate": 5.161290322580646e-06,
-      "loss": 12.7474,
+      "grad_norm": 135.1348876953125,
+      "learning_rate": 1.7204301075268817e-06,
+      "loss": 29.6055,
       "step": 7
     },
     {
       "epoch": 0.0005171717171717172,
-      "grad_norm": NaN,
-      "learning_rate": 5.161290322580646e-06,
-      "loss": 12.7983,
+      "grad_norm": Infinity,
+      "learning_rate": 1.7204301075268817e-06,
+      "loss": 28.568,
       "step": 8
     },
     {
       "epoch": 0.0005818181818181818,
-      "grad_norm": 60.97978591918945,
-      "learning_rate": 6.451612903225806e-06,
-      "loss": 12.524,
+      "grad_norm": 270.385498046875,
+      "learning_rate": 2.1505376344086023e-06,
+      "loss": 29.7127,
       "step": 9
     },
     {
       "epoch": 0.0006464646464646465,
-      "grad_norm": 54.93958282470703,
-      "learning_rate": 7.741935483870968e-06,
-      "loss": 13.3508,
+      "grad_norm": 109.7217788696289,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 29.2279,
       "step": 10
     },
     {
       "epoch": 0.0007111111111111111,
-      "grad_norm": 95.04920196533203,
-      "learning_rate": 9.03225806451613e-06,
-      "loss": 12.7488,
+      "grad_norm": 403.46337890625,
+      "learning_rate": 3.0107526881720433e-06,
+      "loss": 28.7925,
       "step": 11
     },
     {
       "epoch": 0.0007757575757575758,
-      "grad_norm": 146.7698516845703,
-      "learning_rate": 1.0322580645161291e-05,
-      "loss": 12.3513,
+      "grad_norm": 114.81087493896484,
+      "learning_rate": 3.4408602150537635e-06,
+      "loss": 28.0663,
       "step": 12
     },
     {
       "epoch": 0.0008404040404040404,
-      "grad_norm": 61.88846206665039,
-      "learning_rate": 1.1612903225806453e-05,
-      "loss": 12.392,
+      "grad_norm": 240.21282958984375,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 27.3475,
       "step": 13
     },
     {
       "epoch": 0.000905050505050505,
-      "grad_norm": 84.30684661865234,
-      "learning_rate": 1.2903225806451613e-05,
-      "loss": 12.0805,
+      "grad_norm": 236.48675537109375,
+      "learning_rate": 4.3010752688172045e-06,
+      "loss": 26.9302,
       "step": 14
     },
     {
       "epoch": 0.0009696969696969697,
-      "grad_norm": 166.33363342285156,
-      "learning_rate": 1.4193548387096774e-05,
-      "loss": 11.7633,
+      "grad_norm": 203.8461456298828,
+      "learning_rate": 4.731182795698925e-06,
+      "loss": 24.0135,
       "step": 15
     },
     {
       "epoch": 0.0010343434343434343,
-      "grad_norm": 70.05492401123047,
-      "learning_rate": 1.5483870967741936e-05,
-      "loss": 11.2654,
+      "grad_norm": 156.15663146972656,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 22.7445,
       "step": 16
     },
     {
       "epoch": 0.0010343434343434343,
-      "eval_bleu": 4.01633632443273,
-      "eval_loss": 12.072832107543945,
-      "eval_runtime": 1.3625,
-      "eval_samples_per_second": 11.743,
-      "eval_steps_per_second": 2.936,
+      "eval_bleu": 0.11919568898736486,
+      "eval_loss": 21.227251052856445,
+      "eval_runtime": 2.9098,
+      "eval_samples_per_second": 10.997,
+      "eval_steps_per_second": 1.375,
       "step": 16
-    },
-    {
-      "epoch": 0.0010989898989898989,
-      "grad_norm": NaN,
-      "learning_rate": 1.5483870967741936e-05,
-      "loss": 11.3409,
-      "step": 17
-    },
-    {
-      "epoch": 0.0011636363636363637,
-      "grad_norm": NaN,
-      "learning_rate": 1.5483870967741936e-05,
-      "loss": 10.8596,
-      "step": 18
-    },
-    {
-      "epoch": 0.0012282828282828282,
-      "grad_norm": 191.51080322265625,
-      "learning_rate": 1.6774193548387098e-05,
-      "loss": 11.4091,
-      "step": 19
-    },
-    {
-      "epoch": 0.001292929292929293,
-      "grad_norm": 60.890865325927734,
-      "learning_rate": 1.806451612903226e-05,
-      "loss": 11.6678,
-      "step": 20
-    },
-    {
-      "epoch": 0.0013575757575757575,
-      "grad_norm": 86.45341491699219,
-      "learning_rate": 1.935483870967742e-05,
-      "loss": 11.0757,
-      "step": 21
-    },
-    {
-      "epoch": 0.0014222222222222223,
-      "grad_norm": 157.23501586914062,
-      "learning_rate": 2.0645161290322582e-05,
-      "loss": 10.7504,
-      "step": 22
-    },
-    {
-      "epoch": 0.0014868686868686868,
-      "grad_norm": 63.56692123413086,
-      "learning_rate": 2.1935483870967744e-05,
-      "loss": 10.3107,
-      "step": 23
-    },
-    {
-      "epoch": 0.0015515151515151516,
-      "grad_norm": 134.14984130859375,
-      "learning_rate": 2.3225806451612906e-05,
-      "loss": 9.6532,
-      "step": 24
-    },
-    {
-      "epoch": 0.0016161616161616162,
-      "grad_norm": 58.10708999633789,
-      "learning_rate": 2.4516129032258064e-05,
-      "loss": 9.6044,
-      "step": 25
-    },
-    {
-      "epoch": 0.0016808080808080807,
-      "grad_norm": 66.19347381591797,
-      "learning_rate": 2.5806451612903226e-05,
-      "loss": 9.1942,
-      "step": 26
-    },
-    {
-      "epoch": 0.0017454545454545455,
-      "grad_norm": 60.57972717285156,
-      "learning_rate": 2.709677419354839e-05,
-      "loss": 9.2875,
-      "step": 27
-    },
-    {
-      "epoch": 0.00181010101010101,
-      "grad_norm": 58.872520446777344,
-      "learning_rate": 2.838709677419355e-05,
-      "loss": 7.7579,
-      "step": 28
-    },
-    {
-      "epoch": 0.0018747474747474748,
-      "grad_norm": 63.92591094970703,
-      "learning_rate": 2.967741935483871e-05,
-      "loss": 7.4737,
-      "step": 29
-    },
-    {
-      "epoch": 0.0019393939393939393,
-      "grad_norm": 59.86642837524414,
-      "learning_rate": 3.096774193548387e-05,
-      "loss": 6.8395,
-      "step": 30
-    },
-    {
-      "epoch": 0.002004040404040404,
-      "grad_norm": 61.897117614746094,
-      "learning_rate": 3.2258064516129034e-05,
-      "loss": 6.788,
-      "step": 31
-    },
-    {
-      "epoch": 0.0020686868686868687,
-      "grad_norm": 68.02660369873047,
-      "learning_rate": 3.3548387096774195e-05,
-      "loss": 6.5021,
-      "step": 32
-    },
-    {
-      "epoch": 0.0020686868686868687,
-      "eval_bleu": 3.747450816674144,
-      "eval_loss": 3.330573558807373,
-      "eval_runtime": 1.3643,
-      "eval_samples_per_second": 11.728,
-      "eval_steps_per_second": 2.932,
-      "step": 32
-    },
-    {
-      "epoch": 0.0021333333333333334,
-      "grad_norm": 59.149505615234375,
-      "learning_rate": 3.483870967741936e-05,
-      "loss": 5.7595,
-      "step": 33
-    },
-    {
-      "epoch": 0.0021979797979797978,
-      "grad_norm": 62.66898727416992,
-      "learning_rate": 3.612903225806452e-05,
-      "loss": 5.3139,
-      "step": 34
-    },
-    {
-      "epoch": 0.0022626262626262625,
-      "grad_norm": 59.711551666259766,
-      "learning_rate": 3.741935483870968e-05,
-      "loss": 5.1228,
-      "step": 35
-    },
-    {
-      "epoch": 0.0023272727272727273,
-      "grad_norm": 55.77004623413086,
-      "learning_rate": 3.870967741935484e-05,
-      "loss": 4.1742,
-      "step": 36
-    },
-    {
-      "epoch": 0.002391919191919192,
-      "grad_norm": 52.294471740722656,
-      "learning_rate": 4e-05,
-      "loss": 4.2128,
-      "step": 37
-    },
-    {
-      "epoch": 0.0024565656565656564,
-      "grad_norm": 51.9980583190918,
-      "learning_rate": 4.1290322580645165e-05,
-      "loss": 3.3269,
-      "step": 38
-    },
-    {
-      "epoch": 0.002521212121212121,
-      "grad_norm": 40.03384017944336,
-      "learning_rate": 4.258064516129032e-05,
-      "loss": 2.7555,
-      "step": 39
-    },
-    {
-      "epoch": 0.002585858585858586,
-      "grad_norm": 33.953269958496094,
-      "learning_rate": 4.387096774193549e-05,
-      "loss": 2.4034,
-      "step": 40
-    },
-    {
-      "epoch": 0.0026505050505050507,
-      "grad_norm": 29.490108489990234,
-      "learning_rate": 4.516129032258064e-05,
-      "loss": 2.1995,
-      "step": 41
-    },
-    {
-      "epoch": 0.002715151515151515,
-      "grad_norm": 56.527854919433594,
-      "learning_rate": 4.645161290322581e-05,
-      "loss": 1.7884,
-      "step": 42
-    },
-    {
-      "epoch": 0.00277979797979798,
-      "grad_norm": 21.921388626098633,
-      "learning_rate": 4.774193548387097e-05,
-      "loss": 1.8252,
-      "step": 43
-    },
-    {
-      "epoch": 0.0028444444444444446,
-      "grad_norm": 18.03729820251465,
-      "learning_rate": 4.903225806451613e-05,
-      "loss": 1.706,
-      "step": 44
-    },
-    {
-      "epoch": 0.002909090909090909,
-      "grad_norm": 11.620859146118164,
-      "learning_rate": 5.032258064516129e-05,
-      "loss": 1.4186,
-      "step": 45
-    },
-    {
-      "epoch": 0.0029737373737373737,
-      "grad_norm": 10.846900939941406,
-      "learning_rate": 5.161290322580645e-05,
-      "loss": 1.3585,
-      "step": 46
-    },
-    {
-      "epoch": 0.0030383838383838385,
-      "grad_norm": 9.786290168762207,
-      "learning_rate": 5.290322580645162e-05,
-      "loss": 1.345,
-      "step": 47
-    },
-    {
-      "epoch": 0.0031030303030303032,
-      "grad_norm": 27.742897033691406,
-      "learning_rate": 5.419354838709678e-05,
-      "loss": 1.1557,
-      "step": 48
-    },
-    {
-      "epoch": 0.0031030303030303032,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.3357515335083008,
-      "eval_runtime": 1.3804,
-      "eval_samples_per_second": 11.591,
-      "eval_steps_per_second": 2.898,
-      "step": 48
-    },
-    {
-      "epoch": 0.0031676767676767676,
-      "grad_norm": 3.792893409729004,
-      "learning_rate": 5.5483870967741936e-05,
-      "loss": 1.0484,
-      "step": 49
-    },
-    {
-      "epoch": 0.0032323232323232323,
-      "grad_norm": 6.693092346191406,
-      "learning_rate": 5.67741935483871e-05,
-      "loss": 1.1635,
-      "step": 50
-    },
-    {
-      "epoch": 0.003296969696969697,
-      "grad_norm": 3.844968795776367,
-      "learning_rate": 5.8064516129032266e-05,
-      "loss": 1.0061,
-      "step": 51
-    },
-    {
-      "epoch": 0.0033616161616161614,
-      "grad_norm": 4.14786958694458,
-      "learning_rate": 5.935483870967742e-05,
-      "loss": 0.9577,
-      "step": 52
-    },
-    {
-      "epoch": 0.003426262626262626,
-      "grad_norm": 4.455865859985352,
-      "learning_rate": 6.064516129032258e-05,
-      "loss": 0.8501,
-      "step": 53
-    },
-    {
-      "epoch": 0.003490909090909091,
-      "grad_norm": 8.088818550109863,
-      "learning_rate": 6.193548387096774e-05,
-      "loss": 0.7826,
-      "step": 54
-    },
-    {
-      "epoch": 0.0035555555555555557,
-      "grad_norm": 3.6797592639923096,
-      "learning_rate": 6.32258064516129e-05,
-      "loss": 0.7374,
-      "step": 55
-    },
-    {
-      "epoch": 0.00362020202020202,
-      "grad_norm": 3.0284504890441895,
-      "learning_rate": 6.451612903225807e-05,
-      "loss": 0.6646,
-      "step": 56
-    },
-    {
-      "epoch": 0.003684848484848485,
-      "grad_norm": 2.8636531829833984,
-      "learning_rate": 6.580645161290323e-05,
-      "loss": 0.6316,
-      "step": 57
-    },
-    {
-      "epoch": 0.0037494949494949496,
-      "grad_norm": 2.4601149559020996,
-      "learning_rate": 6.709677419354839e-05,
-      "loss": 0.5411,
-      "step": 58
-    },
-    {
-      "epoch": 0.003814141414141414,
-      "grad_norm": 2.3120172023773193,
-      "learning_rate": 6.838709677419355e-05,
-      "loss": 0.4993,
-      "step": 59
-    },
-    {
-      "epoch": 0.0038787878787878787,
-      "grad_norm": 1.8615421056747437,
-      "learning_rate": 6.967741935483871e-05,
-      "loss": 0.4766,
-      "step": 60
-    },
-    {
-      "epoch": 0.0039434343434343435,
-      "grad_norm": 1.5349754095077515,
-      "learning_rate": 7.096774193548388e-05,
-      "loss": 0.4203,
-      "step": 61
-    },
-    {
-      "epoch": 0.004008080808080808,
-      "grad_norm": 1.7214694023132324,
-      "learning_rate": 7.225806451612904e-05,
-      "loss": 0.4294,
-      "step": 62
-    },
-    {
-      "epoch": 0.004072727272727273,
-      "grad_norm": 1.5352543592453003,
-      "learning_rate": 7.35483870967742e-05,
-      "loss": 0.3648,
-      "step": 63
-    },
-    {
-      "epoch": 0.004137373737373737,
-      "grad_norm": 0.8122034072875977,
-      "learning_rate": 7.483870967741936e-05,
-      "loss": 0.3224,
-      "step": 64
-    },
-    {
-      "epoch": 0.004137373737373737,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.24285614490509033,
-      "eval_runtime": 1.3711,
-      "eval_samples_per_second": 11.67,
-      "eval_steps_per_second": 2.917,
-      "step": 64
-    },
-    {
-      "epoch": 0.004202020202020202,
-      "grad_norm": 0.8955532312393188,
-      "learning_rate": 7.612903225806451e-05,
-      "loss": 0.347,
-      "step": 65
-    },
-    {
-      "epoch": 0.004266666666666667,
-      "grad_norm": 0.5511079430580139,
-      "learning_rate": 7.741935483870968e-05,
-      "loss": 0.3468,
-      "step": 66
-    },
-    {
-      "epoch": 0.004331313131313131,
-      "grad_norm": 0.8366074562072754,
-      "learning_rate": 7.870967741935484e-05,
-      "loss": 0.2938,
-      "step": 67
-    },
-    {
-      "epoch": 0.0043959595959595955,
-      "grad_norm": 0.43903565406799316,
-      "learning_rate": 8e-05,
-      "loss": 0.2681,
-      "step": 68
-    },
-    {
-      "epoch": 0.004460606060606061,
-      "grad_norm": 0.5412452816963196,
-      "learning_rate": 8.129032258064517e-05,
-      "loss": 0.2725,
-      "step": 69
-    },
-    {
-      "epoch": 0.004525252525252525,
-      "grad_norm": 0.4017622470855713,
-      "learning_rate": 8.258064516129033e-05,
-      "loss": 0.2347,
-      "step": 70
-    },
-    {
-      "epoch": 0.00458989898989899,
-      "grad_norm": 22.646089553833008,
-      "learning_rate": 8.387096774193549e-05,
-      "loss": 0.2836,
-      "step": 71
-    },
-    {
-      "epoch": 0.004654545454545455,
-      "grad_norm": 0.3233143091201782,
-      "learning_rate": 8.516129032258064e-05,
-      "loss": 0.2306,
-      "step": 72
-    },
-    {
-      "epoch": 0.004719191919191919,
-      "grad_norm": 0.3327657878398895,
-      "learning_rate": 8.645161290322581e-05,
-      "loss": 0.2516,
-      "step": 73
-    },
-    {
-      "epoch": 0.004783838383838384,
-      "grad_norm": 0.32695600390434265,
-      "learning_rate": 8.774193548387098e-05,
-      "loss": 0.2525,
-      "step": 74
-    },
-    {
-      "epoch": 0.0048484848484848485,
-      "grad_norm": 0.2796344459056854,
-      "learning_rate": 8.903225806451614e-05,
-      "loss": 0.2679,
-      "step": 75
-    },
-    {
-      "epoch": 0.004913131313131313,
-      "grad_norm": 0.26281794905662537,
-      "learning_rate": 9.032258064516129e-05,
-      "loss": 0.2468,
-      "step": 76
-    },
-    {
-      "epoch": 0.004977777777777778,
-      "grad_norm": 0.2546544075012207,
-      "learning_rate": 9.161290322580646e-05,
-      "loss": 0.237,
-      "step": 77
-    },
-    {
-      "epoch": 0.005042424242424242,
-      "grad_norm": 0.24231921136379242,
-      "learning_rate": 9.290322580645162e-05,
-      "loss": 0.2529,
-      "step": 78
-    },
-    {
-      "epoch": 0.005107070707070707,
-      "grad_norm": 0.24682262539863586,
-      "learning_rate": 9.419354838709677e-05,
-      "loss": 0.2292,
-      "step": 79
-    },
-    {
-      "epoch": 0.005171717171717172,
-      "grad_norm": 0.30323857069015503,
-      "learning_rate": 9.548387096774195e-05,
-      "loss": 0.2191,
-      "step": 80
-    },
-    {
-      "epoch": 0.005171717171717172,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.19482067227363586,
-      "eval_runtime": 1.3813,
-      "eval_samples_per_second": 11.584,
-      "eval_steps_per_second": 2.896,
-      "step": 80
-    },
-    {
-      "epoch": 0.005236363636363636,
-      "grad_norm": 0.21761713922023773,
-      "learning_rate": 9.677419354838711e-05,
-      "loss": 0.2414,
-      "step": 81
-    },
-    {
-      "epoch": 0.005301010101010101,
-      "grad_norm": 0.23040293157100677,
-      "learning_rate": 9.806451612903226e-05,
-      "loss": 0.2274,
-      "step": 82
-    },
-    {
-      "epoch": 0.005365656565656566,
-      "grad_norm": 0.23759864270687103,
-      "learning_rate": 9.935483870967742e-05,
-      "loss": 0.2126,
-      "step": 83
-    },
-    {
-      "epoch": 0.00543030303030303,
-      "grad_norm": 0.19173021614551544,
-      "learning_rate": 0.00010064516129032258,
-      "loss": 0.2219,
-      "step": 84
-    },
-    {
-      "epoch": 0.005494949494949495,
-      "grad_norm": 0.19772595167160034,
-      "learning_rate": 0.00010193548387096774,
-      "loss": 0.2046,
-      "step": 85
-    },
-    {
-      "epoch": 0.00555959595959596,
-      "grad_norm": 0.19209372997283936,
-      "learning_rate": 0.0001032258064516129,
-      "loss": 0.2202,
-      "step": 86
-    },
-    {
-      "epoch": 0.005624242424242424,
-      "grad_norm": 0.18713383376598358,
-      "learning_rate": 0.00010451612903225806,
-      "loss": 0.2279,
-      "step": 87
-    },
-    {
-      "epoch": 0.005688888888888889,
-      "grad_norm": 0.20889417827129364,
-      "learning_rate": 0.00010580645161290324,
-      "loss": 0.1787,
-      "step": 88
-    },
-    {
-      "epoch": 0.0057535353535353535,
-      "grad_norm": 0.17321723699569702,
-      "learning_rate": 0.0001070967741935484,
-      "loss": 0.2047,
-      "step": 89
-    },
-    {
-      "epoch": 0.005818181818181818,
-      "grad_norm": 0.25413277745246887,
-      "learning_rate": 0.00010838709677419356,
-      "loss": 0.2226,
-      "step": 90
-    },
-    {
-      "epoch": 0.005882828282828283,
-      "grad_norm": 0.1873357743024826,
-      "learning_rate": 0.00010967741935483871,
-      "loss": 0.1974,
-      "step": 91
-    },
-    {
-      "epoch": 0.005947474747474747,
-      "grad_norm": 0.14157669246196747,
-      "learning_rate": 0.00011096774193548387,
-      "loss": 0.1757,
-      "step": 92
-    },
-    {
-      "epoch": 0.006012121212121212,
-      "grad_norm": 0.18100616335868835,
-      "learning_rate": 0.00011225806451612903,
-      "loss": 0.193,
-      "step": 93
-    },
-    {
-      "epoch": 0.006076767676767677,
-      "grad_norm": 0.17187540233135223,
-      "learning_rate": 0.0001135483870967742,
-      "loss": 0.1996,
-      "step": 94
-    },
-    {
-      "epoch": 0.006141414141414141,
-      "grad_norm": 0.25643497705459595,
-      "learning_rate": 0.00011483870967741937,
-      "loss": 0.1944,
-      "step": 95
-    },
-    {
-      "epoch": 0.0062060606060606064,
-      "grad_norm": 0.3475594222545624,
-      "learning_rate": 0.00011612903225806453,
-      "loss": 0.1821,
-      "step": 96
-    },
-    {
-      "epoch": 0.0062060606060606064,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.15860861539840698,
-      "eval_runtime": 1.3903,
-      "eval_samples_per_second": 11.508,
-      "eval_steps_per_second": 2.877,
-      "step": 96
-    },
-    {
-      "epoch": 0.006270707070707071,
-      "grad_norm": 0.17301534116268158,
-      "learning_rate": 0.00011741935483870967,
-      "loss": 0.1964,
-      "step": 97
-    },
-    {
-      "epoch": 0.006335353535353535,
-      "grad_norm": 0.13940928876399994,
-      "learning_rate": 0.00011870967741935484,
-      "loss": 0.1702,
-      "step": 98
-    },
-    {
-      "epoch": 0.0064,
-      "grad_norm": 0.1584329903125763,
-      "learning_rate": 0.00012,
-      "loss": 0.1667,
-      "step": 99
-    },
-    {
-      "epoch": 0.006464646464646465,
-      "grad_norm": 0.19155238568782806,
-      "learning_rate": 0.00012129032258064516,
-      "loss": 0.1941,
-      "step": 100
-    },
-    {
-      "epoch": 0.006529292929292929,
-      "grad_norm": 0.1830209642648697,
-      "learning_rate": 0.00012258064516129034,
-      "loss": 0.1794,
-      "step": 101
-    },
-    {
-      "epoch": 0.006593939393939394,
-      "grad_norm": 0.9416115880012512,
-      "learning_rate": 0.0001238709677419355,
-      "loss": 0.1799,
-      "step": 102
-    },
-    {
-      "epoch": 0.0066585858585858585,
-      "grad_norm": 0.13209928572177887,
-      "learning_rate": 0.00012516129032258066,
-      "loss": 0.1715,
-      "step": 103
-    },
-    {
-      "epoch": 0.006723232323232323,
-      "grad_norm": 0.15749603509902954,
-      "learning_rate": 0.0001264516129032258,
-      "loss": 0.1809,
-      "step": 104
-    },
-    {
-      "epoch": 0.006787878787878788,
-      "grad_norm": 0.1440904438495636,
-      "learning_rate": 0.00012774193548387096,
-      "loss": 0.1883,
-      "step": 105
-    },
-    {
-      "epoch": 0.006852525252525252,
-      "grad_norm": 0.13838624954223633,
-      "learning_rate": 0.00012903225806451613,
-      "loss": 0.1835,
-      "step": 106
-    },
-    {
-      "epoch": 0.006917171717171718,
-      "grad_norm": 0.13904741406440735,
-      "learning_rate": 0.0001303225806451613,
-      "loss": 0.1723,
-      "step": 107
-    },
-    {
-      "epoch": 0.006981818181818182,
-      "grad_norm": 0.18367218971252441,
-      "learning_rate": 0.00013161290322580646,
-      "loss": 0.1888,
-      "step": 108
-    },
-    {
-      "epoch": 0.007046464646464646,
-      "grad_norm": 0.12487432360649109,
-      "learning_rate": 0.00013290322580645163,
-      "loss": 0.1821,
-      "step": 109
-    },
-    {
-      "epoch": 0.0071111111111111115,
-      "grad_norm": 0.13827675580978394,
-      "learning_rate": 0.00013419354838709678,
-      "loss": 0.1733,
-      "step": 110
-    },
-    {
-      "epoch": 0.007175757575757576,
-      "grad_norm": 0.12497523427009583,
-      "learning_rate": 0.00013548387096774193,
-      "loss": 0.1804,
-      "step": 111
-    },
-    {
-      "epoch": 0.00724040404040404,
-      "grad_norm": 0.11209689825773239,
-      "learning_rate": 0.0001367741935483871,
-      "loss": 0.1782,
-      "step": 112
-    },
-    {
-      "epoch": 0.00724040404040404,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.14075997471809387,
-      "eval_runtime": 1.3802,
-      "eval_samples_per_second": 11.593,
-      "eval_steps_per_second": 2.898,
-      "step": 112
-    },
-    {
-      "epoch": 0.007305050505050505,
-      "grad_norm": 0.1200064942240715,
-      "learning_rate": 0.00013806451612903225,
-      "loss": 0.1665,
-      "step": 113
-    },
-    {
-      "epoch": 0.00736969696969697,
-      "grad_norm": 0.2869766652584076,
-      "learning_rate": 0.00013935483870967743,
-      "loss": 0.164,
-      "step": 114
-    },
-    {
-      "epoch": 0.007434343434343434,
-      "grad_norm": 0.12517796456813812,
-      "learning_rate": 0.0001406451612903226,
-      "loss": 0.1584,
-      "step": 115
-    },
-    {
-      "epoch": 0.007498989898989899,
-      "grad_norm": 0.11323254555463791,
-      "learning_rate": 0.00014193548387096775,
-      "loss": 0.1593,
-      "step": 116
-    },
-    {
-      "epoch": 0.0075636363636363635,
-      "grad_norm": 0.23315002024173737,
-      "learning_rate": 0.00014322580645161293,
-      "loss": 0.1725,
-      "step": 117
-    },
-    {
-      "epoch": 0.007628282828282828,
-      "grad_norm": 0.11487537622451782,
-      "learning_rate": 0.00014451612903225807,
-      "loss": 0.1554,
-      "step": 118
-    },
-    {
-      "epoch": 0.007692929292929293,
-      "grad_norm": 0.11991633474826813,
-      "learning_rate": 0.00014580645161290322,
-      "loss": 0.1525,
-      "step": 119
-    },
-    {
-      "epoch": 0.007757575757575757,
-      "grad_norm": 0.17173829674720764,
-      "learning_rate": 0.0001470967741935484,
-      "loss": 0.1961,
-      "step": 120
-    },
-    {
-      "epoch": 0.007822222222222222,
-      "grad_norm": 0.22421815991401672,
-      "learning_rate": 0.00014838709677419355,
-      "loss": 0.1491,
-      "step": 121
-    },
-    {
-      "epoch": 0.007886868686868687,
-      "grad_norm": 0.10590796172618866,
-      "learning_rate": 0.00014967741935483872,
-      "loss": 0.1466,
-      "step": 122
-    },
-    {
-      "epoch": 0.007951515151515152,
-      "grad_norm": 0.15446847677230835,
-      "learning_rate": 0.0001509677419354839,
-      "loss": 0.1506,
-      "step": 123
-    },
-    {
-      "epoch": 0.008016161616161616,
-      "grad_norm": 0.1181936264038086,
-      "learning_rate": 0.00015225806451612902,
-      "loss": 0.1483,
-      "step": 124
-    },
-    {
-      "epoch": 0.00808080808080808,
-      "grad_norm": 0.12403552234172821,
-      "learning_rate": 0.0001535483870967742,
-      "loss": 0.1602,
-      "step": 125
-    },
-    {
-      "epoch": 0.008145454545454546,
-      "grad_norm": 0.12927326560020447,
-      "learning_rate": 0.00015483870967741937,
-      "loss": 0.173,
-      "step": 126
-    },
-    {
-      "epoch": 0.00821010101010101,
-      "grad_norm": 0.7120084762573242,
-      "learning_rate": 0.00015612903225806451,
-      "loss": 0.1561,
-      "step": 127
-    },
-    {
-      "epoch": 0.008274747474747475,
-      "grad_norm": 0.09934462606906891,
-      "learning_rate": 0.0001574193548387097,
-      "loss": 0.1433,
-      "step": 128
-    },
-    {
-      "epoch": 0.008274747474747475,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.12477699667215347,
-      "eval_runtime": 1.3776,
-      "eval_samples_per_second": 11.614,
-      "eval_steps_per_second": 2.904,
-      "step": 128
-    },
-    {
-      "epoch": 0.00833939393939394,
-      "grad_norm": 0.11033691465854645,
-      "learning_rate": 0.00015870967741935487,
-      "loss": 0.1615,
-      "step": 129
-    },
-    {
-      "epoch": 0.008404040404040403,
-      "grad_norm": 0.12420456856489182,
-      "learning_rate": 0.00016,
-      "loss": 0.1477,
-      "step": 130
-    },
-    {
-      "epoch": 0.008468686868686869,
-      "grad_norm": 0.1626136600971222,
-      "learning_rate": 0.00016129032258064516,
-      "loss": 0.1624,
-      "step": 131
-    },
-    {
-      "epoch": 0.008533333333333334,
-      "grad_norm": 0.1007506251335144,
-      "learning_rate": 0.00016258064516129034,
-      "loss": 0.1499,
-      "step": 132
-    },
-    {
-      "epoch": 0.008597979797979797,
-      "grad_norm": 0.35909000039100647,
-      "learning_rate": 0.00016387096774193548,
-      "loss": 0.1646,
-      "step": 133
-    },
-    {
-      "epoch": 0.008662626262626262,
-      "grad_norm": 0.139847531914711,
-      "learning_rate": 0.00016516129032258066,
-      "loss": 0.1531,
-      "step": 134
-    },
-    {
-      "epoch": 0.008727272727272728,
-      "grad_norm": 0.14874647557735443,
-      "learning_rate": 0.0001664516129032258,
-      "loss": 0.1518,
-      "step": 135
-    },
-    {
-      "epoch": 0.008791919191919191,
-      "grad_norm": 0.0868111178278923,
-      "learning_rate": 0.00016774193548387098,
-      "loss": 0.1429,
-      "step": 136
-    },
-    {
-      "epoch": 0.008856565656565656,
-      "grad_norm": 0.13478216528892517,
-      "learning_rate": 0.00016903225806451616,
-      "loss": 0.1591,
-      "step": 137
-    },
-    {
-      "epoch": 0.008921212121212121,
-      "grad_norm": 0.27142342925071716,
-      "learning_rate": 0.00017032258064516128,
-      "loss": 0.1421,
-      "step": 138
-    },
-    {
-      "epoch": 0.008985858585858587,
-      "grad_norm": 0.10351862758398056,
-      "learning_rate": 0.00017161290322580645,
-      "loss": 0.1314,
-      "step": 139
-    },
-    {
-      "epoch": 0.00905050505050505,
-      "grad_norm": 0.2667485177516937,
-      "learning_rate": 0.00017290322580645163,
-      "loss": 0.1383,
-      "step": 140
-    },
-    {
-      "epoch": 0.009115151515151515,
-      "grad_norm": 0.12030935287475586,
-      "learning_rate": 0.00017419354838709678,
-      "loss": 0.1532,
-      "step": 141
-    },
-    {
-      "epoch": 0.00917979797979798,
-      "grad_norm": 0.11514189839363098,
-      "learning_rate": 0.00017548387096774195,
-      "loss": 0.1478,
-      "step": 142
-    },
-    {
-      "epoch": 0.009244444444444444,
-      "grad_norm": 0.09005405008792877,
-      "learning_rate": 0.0001767741935483871,
-      "loss": 0.1323,
-      "step": 143
-    },
-    {
-      "epoch": 0.00930909090909091,
-      "grad_norm": 0.09564518928527832,
-      "learning_rate": 0.00017806451612903228,
-      "loss": 0.1591,
-      "step": 144
-    },
-    {
-      "epoch": 0.00930909090909091,
-      "eval_bleu": 0.0,
-      "eval_loss": 0.11460547149181366,
-      "eval_runtime": 1.359,
-      "eval_samples_per_second": 11.773,
-      "eval_steps_per_second": 2.943,
-      "step": 144
-    },
-    {
-      "epoch": 0.009373737373737374,
-      "grad_norm": 0.0995207279920578,
-      "learning_rate": 0.00017935483870967742,
-      "loss": 0.1577,
-      "step": 145
-    },
-    {
-      "epoch": 0.009438383838383838,
-      "grad_norm": 0.4307728707790375,
-      "learning_rate": 0.00018064516129032257,
-      "loss": 0.1381,
-      "step": 146
-    },
-    {
-      "epoch": 0.009503030303030303,
-      "grad_norm": 0.10841380804777145,
-      "learning_rate": 0.00018193548387096775,
-      "loss": 0.1695,
-      "step": 147
-    },
-    {
-      "epoch": 0.009567676767676768,
-      "grad_norm": 0.08941018581390381,
-      "learning_rate": 0.00018322580645161292,
-      "loss": 0.1407,
-      "step": 148
-    },
-    {
-      "epoch": 0.009632323232323232,
-      "grad_norm": 0.09527455270290375,
-      "learning_rate": 0.00018451612903225807,
-      "loss": 0.1515,
-      "step": 149
-    },
-    {
-      "epoch": 0.009696969696969697,
-      "grad_norm": 0.07641109079122543,
-      "learning_rate": 0.00018580645161290325,
-      "loss": 0.1433,
-      "step": 150
-    },
-    {
-      "epoch": 0.009761616161616162,
-      "grad_norm": 0.1487646847963333,
-      "learning_rate": 0.0001870967741935484,
-      "loss": 0.1474,
-      "step": 151
-    },
-    {
-      "epoch": 0.009826262626262626,
-      "grad_norm": 0.08308811485767365,
-      "learning_rate": 0.00018838709677419354,
-      "loss": 0.1323,
-      "step": 152
-    },
-    {
-      "epoch": 0.00989090909090909,
-      "grad_norm": 0.06572406738996506,
-      "learning_rate": 0.00018967741935483872,
-      "loss": 0.1407,
-      "step": 153
-    },
-    {
-      "epoch": 0.009955555555555556,
-      "grad_norm": 0.24972431361675262,
-      "learning_rate": 0.0001909677419354839,
-      "loss": 0.1385,
-      "step": 154
-    },
-    {
-      "epoch": 0.01002020202020202,
-      "grad_norm": 0.07581052929162979,
-      "learning_rate": 0.00019225806451612904,
-      "loss": 0.1507,
-      "step": 155
-    },
-    {
-      "epoch": 0.010084848484848485,
-      "grad_norm": 0.1980135142803192,
-      "learning_rate": 0.00019354838709677422,
-      "loss": 0.1419,
-      "step": 156
-    },
-    {
-      "epoch": 0.01014949494949495,
-      "grad_norm": 0.06879571825265884,
-      "learning_rate": 0.00019483870967741936,
-      "loss": 0.1371,
-      "step": 157
-    },
-    {
-      "epoch": 0.010214141414141413,
-      "grad_norm": 0.08168785274028778,
-      "learning_rate": 0.0001961290322580645,
-      "loss": 0.143,
-      "step": 158
-    },
-    {
-      "epoch": 0.010278787878787879,
-      "grad_norm": 0.13444702327251434,
-      "learning_rate": 0.00019741935483870969,
-      "loss": 0.1458,
-      "step": 159
-    },
-    {
-      "epoch": 0.010343434343434344,
-      "grad_norm": 0.09177995473146439,
-      "learning_rate": 0.00019870967741935483,
-      "loss": 0.1302,
-      "step": 160
-    },
-    {
-      "epoch": 0.010343434343434344,
-      "eval_bleu": 0.1344542592045913,
-      "eval_loss": 0.10595569759607315,
-      "eval_runtime": 1.353,
-      "eval_samples_per_second": 11.825,
-      "eval_steps_per_second": 2.956,
-      "step": 160
-    },
-    {
-      "epoch": 0.010408080808080807,
-      "grad_norm": 0.059201959520578384,
-      "learning_rate": 0.0002,
-      "loss": 0.1317,
-      "step": 161
-    },
-    {
-      "epoch": 0.010472727272727272,
-      "grad_norm": 0.1885806769132614,
-      "learning_rate": 0.00019999999789549876,
-      "loss": 0.1319,
-      "step": 162
-    },
-    {
-      "epoch": 0.010537373737373738,
-      "grad_norm": 0.06697044521570206,
-      "learning_rate": 0.0001999999915819952,
-      "loss": 0.1136,
-      "step": 163
-    },
-    {
-      "epoch": 0.010602020202020203,
-      "grad_norm": 0.0689595639705658,
-      "learning_rate": 0.00019999998105948953,
-      "loss": 0.1189,
-      "step": 164
-    },
-    {
-      "epoch": 0.010666666666666666,
-      "grad_norm": 0.07707302272319794,
-      "learning_rate": 0.00019999996632798217,
-      "loss": 0.1412,
-      "step": 165
-    },
-    {
-      "epoch": 0.010731313131313132,
-      "grad_norm": 0.22306282818317413,
-      "learning_rate": 0.00019999994738747378,
-      "loss": 0.1657,
-      "step": 166
-    },
-    {
-      "epoch": 0.010795959595959597,
-      "grad_norm": 0.09084911644458771,
-      "learning_rate": 0.00019999992423796515,
-      "loss": 0.126,
-      "step": 167
-    },
-    {
-      "epoch": 0.01086060606060606,
-      "grad_norm": 0.09681031852960587,
-      "learning_rate": 0.00019999989687945728,
-      "loss": 0.1303,
-      "step": 168
-    },
-    {
-      "epoch": 0.010925252525252525,
-      "grad_norm": 0.12961797416210175,
-      "learning_rate": 0.0001999998653119513,
-      "loss": 0.1231,
-      "step": 169
-    },
-    {
-      "epoch": 0.01098989898989899,
-      "grad_norm": 0.07255159318447113,
-      "learning_rate": 0.00019999982953544852,
-      "loss": 0.1324,
-      "step": 170
-    },
-    {
-      "epoch": 0.011054545454545454,
-      "grad_norm": 0.07213090360164642,
-      "learning_rate": 0.00019999978954995045,
-      "loss": 0.1243,
-      "step": 171
-    },
-    {
-      "epoch": 0.01111919191919192,
-      "grad_norm": 0.058742836117744446,
-      "learning_rate": 0.0001999997453554588,
-      "loss": 0.1196,
-      "step": 172
-    },
-    {
-      "epoch": 0.011183838383838384,
-      "grad_norm": 0.06557495146989822,
-      "learning_rate": 0.00019999969695197543,
-      "loss": 0.1234,
-      "step": 173
-    },
-    {
-      "epoch": 0.011248484848484848,
-      "grad_norm": 0.05962904542684555,
-      "learning_rate": 0.00019999964433950235,
-      "loss": 0.1465,
-      "step": 174
-    },
-    {
-      "epoch": 0.011313131313131313,
-      "grad_norm": 0.11403318494558334,
-      "learning_rate": 0.00019999958751804178,
-      "loss": 0.1147,
-      "step": 175
-    },
-    {
-      "epoch": 0.011377777777777778,
-      "grad_norm": 0.06283015757799149,
-      "learning_rate": 0.0001999995264875961,
-      "loss": 0.1051,
-      "step": 176
-    },
-    {
-      "epoch": 0.011377777777777778,
-      "eval_bleu": 0.34108656655271324,
-      "eval_loss": 0.10132479667663574,
-      "eval_runtime": 1.4086,
-      "eval_samples_per_second": 11.359,
-      "eval_steps_per_second": 2.84,
-      "step": 176
-    },
-    {
-      "epoch": 0.011442424242424242,
-      "grad_norm": 0.07057774811983109,
-      "learning_rate": 0.00019999946124816794,
-      "loss": 0.1159,
-      "step": 177
-    },
-    {
-      "epoch": 0.011507070707070707,
-      "grad_norm": 0.1344994753599167,
-      "learning_rate": 0.00019999939179975997,
-      "loss": 0.1123,
-      "step": 178
-    },
-    {
-      "epoch": 0.011571717171717172,
-      "grad_norm": 0.053567882627248764,
-      "learning_rate": 0.00019999931814237515,
-      "loss": 0.1319,
-      "step": 179
-    },
-    {
-      "epoch": 0.011636363636363636,
-      "grad_norm": 0.05020461976528168,
-      "learning_rate": 0.0001999992402760166,
-      "loss": 0.1315,
-      "step": 180
-    },
-    {
-      "epoch": 0.0117010101010101,
-      "grad_norm": 0.09403225779533386,
-      "learning_rate": 0.00019999915820068757,
-      "loss": 0.1275,
-      "step": 181
-    },
-    {
-      "epoch": 0.011765656565656566,
-      "grad_norm": 0.07833687961101532,
-      "learning_rate": 0.0001999990719163915,
-      "loss": 0.1216,
-      "step": 182
-    },
-    {
-      "epoch": 0.01183030303030303,
-      "grad_norm": 0.03991740942001343,
-      "learning_rate": 0.00019999898142313206,
-      "loss": 0.1142,
-      "step": 183
-    },
-    {
-      "epoch": 0.011894949494949495,
-      "grad_norm": 0.04619375616312027,
-      "learning_rate": 0.00019999888672091304,
-      "loss": 0.1103,
-      "step": 184
-    },
-    {
-      "epoch": 0.01195959595959596,
-      "grad_norm": 0.037316370755434036,
-      "learning_rate": 0.0001999987878097384,
-      "loss": 0.1168,
-      "step": 185
-    },
-    {
-      "epoch": 0.012024242424242423,
-      "grad_norm": 0.04291122406721115,
-      "learning_rate": 0.00019999868468961233,
-      "loss": 0.1198,
-      "step": 186
-    },
-    {
-      "epoch": 0.012088888888888889,
-      "grad_norm": 0.07236277312040329,
-      "learning_rate": 0.00019999857736053918,
-      "loss": 0.12,
-      "step": 187
-    },
-    {
-      "epoch": 0.012153535353535354,
-      "grad_norm": 0.04166350141167641,
-      "learning_rate": 0.0001999984658225235,
-      "loss": 0.1312,
-      "step": 188
-    },
-    {
-      "epoch": 0.012218181818181819,
-      "grad_norm": 0.04327237978577614,
-      "learning_rate": 0.00019999835007556986,
-      "loss": 0.138,
-      "step": 189
-    },
-    {
-      "epoch": 0.012282828282828282,
-      "grad_norm": 0.13315382599830627,
-      "learning_rate": 0.00019999823011968327,
-      "loss": 0.1395,
-      "step": 190
-    },
-    {
-      "epoch": 0.012347474747474748,
-      "grad_norm": 0.04486257955431938,
-      "learning_rate": 0.0001999981059548687,
-      "loss": 0.1281,
-      "step": 191
-    },
-    {
-      "epoch": 0.012412121212121213,
-      "grad_norm": 0.06127722188830376,
-      "learning_rate": 0.0001999979775811314,
-      "loss": 0.1295,
-      "step": 192
-    },
-    {
-      "epoch": 0.012412121212121213,
-      "eval_bleu": 2.797660142073947,
-      "eval_loss": 0.09842301905155182,
-      "eval_runtime": 1.3435,
-      "eval_samples_per_second": 11.909,
-      "eval_steps_per_second": 2.977,
-      "step": 192
-    },
-    {
-      "epoch": 0.012476767676767676,
-      "grad_norm": 0.10934247821569443,
-      "learning_rate": 0.00019999784499847678,
-      "loss": 0.119,
-      "step": 193
-    },
-    {
-      "epoch": 0.012541414141414142,
-      "grad_norm": 0.04266177862882614,
-      "learning_rate": 0.0001999977082069104,
-      "loss": 0.1094,
-      "step": 194
-    },
-    {
-      "epoch": 0.012606060606060607,
-      "grad_norm": 0.10852430015802383,
-      "learning_rate": 0.00019999756720643803,
-      "loss": 0.1118,
-      "step": 195
-    },
-    {
-      "epoch": 0.01267070707070707,
-      "grad_norm": 0.06190445274114609,
-      "learning_rate": 0.0001999974219970656,
-      "loss": 0.129,
-      "step": 196
-    },
-    {
-      "epoch": 0.012735353535353535,
-      "grad_norm": 0.04268389567732811,
-      "learning_rate": 0.00019999727257879923,
-      "loss": 0.1149,
-      "step": 197
-    },
-    {
-      "epoch": 0.0128,
-      "grad_norm": 0.04210319742560387,
-      "learning_rate": 0.0001999971189516452,
-      "loss": 0.1231,
-      "step": 198
-    },
-    {
-      "epoch": 0.012864646464646464,
-      "grad_norm": 0.07176094502210617,
-      "learning_rate": 0.00019999696111561,
-      "loss": 0.1123,
-      "step": 199
-    },
-    {
-      "epoch": 0.01292929292929293,
-      "grad_norm": 0.04062803462147713,
-      "learning_rate": 0.00019999679907070023,
-      "loss": 0.1225,
-      "step": 200
-    },
-    {
-      "epoch": 0.012993939393939394,
-      "grad_norm": 0.04266968369483948,
-      "learning_rate": 0.00019999663281692275,
-      "loss": 0.1259,
-      "step": 201
-    },
-    {
-      "epoch": 0.013058585858585858,
-      "grad_norm": 0.045373089611530304,
-      "learning_rate": 0.00019999646235428452,
-      "loss": 0.1353,
-      "step": 202
-    },
-    {
-      "epoch": 0.013123232323232323,
-      "grad_norm": 0.04623784124851227,
-      "learning_rate": 0.00019999628768279276,
-      "loss": 0.1224,
-      "step": 203
-    },
-    {
-      "epoch": 0.013187878787878788,
-      "grad_norm": 0.03664301335811615,
-      "learning_rate": 0.0001999961088024548,
-      "loss": 0.1361,
-      "step": 204
-    },
-    {
-      "epoch": 0.013252525252525252,
-      "grad_norm": 0.03849755972623825,
-      "learning_rate": 0.00019999592571327815,
-      "loss": 0.1307,
-      "step": 205
-    },
-    {
-      "epoch": 0.013317171717171717,
-      "grad_norm": 0.03995022922754288,
-      "learning_rate": 0.00019999573841527054,
-      "loss": 0.1079,
-      "step": 206
-    },
-    {
-      "epoch": 0.013381818181818182,
-      "grad_norm": 0.039675675332546234,
-      "learning_rate": 0.00019999554690843988,
-      "loss": 0.1212,
-      "step": 207
-    },
-    {
-      "epoch": 0.013446464646464646,
-      "grad_norm": 0.05080877244472504,
-      "learning_rate": 0.00019999535119279415,
-      "loss": 0.0991,
-      "step": 208
-    },
-    {
-      "epoch": 0.013446464646464646,
-      "eval_bleu": 3.66705872401506,
-      "eval_loss": 0.09652489423751831,
-      "eval_runtime": 1.3632,
-      "eval_samples_per_second": 11.737,
-      "eval_steps_per_second": 2.934,
-      "step": 208
     }
   ],
   "logging_steps": 1,
-  "max_steps": 15468,
+  "max_steps": 46404,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 1,
+  "num_train_epochs": 3,
   "save_steps": 16,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -1599,8 +147,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 1801670061195264.0,
-  "train_batch_size": 4,
+  "total_flos": 311786439966720.0,
+  "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc65473306b4a3f5fae99fa380d17d29a861d4b0c3d748ffda65119a04290231
+oid sha256:3f3a0792a04fef5ea618963940e094b8c3947590c45a58f3ea50f2fde6e0f3e8
 size 5240