Next training iteration (2.03m)

added_tokens.json

@@ -1 +1 @@
-{"NO_SEGMENT_TOKEN": 32112, "BETWEEN_SEGMENTS_TOKEN": 32111, "NUMBER_TOKEN": 32103, "START_SPONSOR_TOKEN": 32113, "START_SELFPROMO_TOKEN": 32115, "END_SPONSOR_TOKEN": 32114, "END_INTERACTION_TOKEN": 32118, "[Applause]": 32107, "HYPHENATED_URL_TOKEN": 32101, "EXTRACT_SEGMENTS: ": 32110, "SHORT_HYPHENATED_TOKEN": 32104, "END_SELFPROMO_TOKEN": 32116, "NUMBER_PERCENTAGE_TOKEN": 32102, "PROFANITY_TOKEN": 32109, "[Music]": 32106, "[Laughter]": 32108, "LONG_WORD_TOKEN": 32105, "URL_TOKEN": 32100, "START_INTERACTION_TOKEN": 32117}
+{"PROFANITY_TOKEN": 32109, "END_INTERACTION_TOKEN": 32118, "URL_TOKEN": 32100, "EXTRACT_SEGMENTS: ": 32110, "NUMBER_TOKEN": 32103, "NUMBER_PERCENTAGE_TOKEN": 32102, "LONG_WORD_TOKEN": 32105, "END_SPONSOR_TOKEN": 32114, "START_SELFPROMO_TOKEN": 32115, "END_SELFPROMO_TOKEN": 32116, "NO_SEGMENT_TOKEN": 32112, "HYPHENATED_URL_TOKEN": 32101, "START_SPONSOR_TOKEN": 32113, "[Music]": 32106, "[Laughter]": 32108, "BETWEEN_SEGMENTS_TOKEN": 32111, "[Applause]": 32107, "START_INTERACTION_TOKEN": 32117, "SHORT_HYPHENATED_TOKEN": 32104}

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "models/small-120k/",
+  "_name_or_path": "models/small-1712500/",
   "architectures": [
     "T5ForConditionalGeneration"
   ],

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2d8a0af26f7ff16bae5ef0583063c69ee47700be6a4793c994a3ef6e83d117b8
+oid sha256:49525bef46229b9dd14d256aab3b299a9abc1fe2f92316ca7465eda817803842
 size 307892869

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d8191c03018e0f166e6e407d71defd3cbfe5dcae86d404ff6a5903f7b765d8ac
+oid sha256:6019e275337380401ff115e5b151ec62b90d64182c9e491d19f52523b075b522
 size 14503

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:257198b9da6abece4de57c5b798cb130a628ec21c36fad1093e79d27e917992e
+oid sha256:aeafd2e8b1e45dc5c6795cc90be9dfd1344b28195e3f803fd812afbd480b4467
 size 623

tokenizer_config.json

@@ -1 +1 @@
-{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 100, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"], "model_max_length": 512, "name_or_path": "models/small-120k/", "special_tokens_map_file": "C:\\Users\\joshu/.cache\\huggingface\\transformers\\3ad6f8335c1b1ef8966245899d47dcf735abd134d21fd7d26f621fe45ac01184.c94798918c92ded6aeef2d2f0e666d2cc4145eca1aa6e1336fde07f2e13e2f46", "sp_model_kwargs": {}, "tokenizer_class": "T5Tokenizer"}
+{"eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>", "extra_ids": 100, "additional_special_tokens": ["<extra_id_0>", "<extra_id_1>", "<extra_id_2>", "<extra_id_3>", "<extra_id_4>", "<extra_id_5>", "<extra_id_6>", "<extra_id_7>", "<extra_id_8>", "<extra_id_9>", "<extra_id_10>", "<extra_id_11>", "<extra_id_12>", "<extra_id_13>", "<extra_id_14>", "<extra_id_15>", "<extra_id_16>", "<extra_id_17>", "<extra_id_18>", "<extra_id_19>", "<extra_id_20>", "<extra_id_21>", "<extra_id_22>", "<extra_id_23>", "<extra_id_24>", "<extra_id_25>", "<extra_id_26>", "<extra_id_27>", "<extra_id_28>", "<extra_id_29>", "<extra_id_30>", "<extra_id_31>", "<extra_id_32>", "<extra_id_33>", "<extra_id_34>", "<extra_id_35>", "<extra_id_36>", "<extra_id_37>", "<extra_id_38>", "<extra_id_39>", "<extra_id_40>", "<extra_id_41>", "<extra_id_42>", "<extra_id_43>", "<extra_id_44>", "<extra_id_45>", "<extra_id_46>", "<extra_id_47>", "<extra_id_48>", "<extra_id_49>", "<extra_id_50>", "<extra_id_51>", "<extra_id_52>", "<extra_id_53>", "<extra_id_54>", "<extra_id_55>", "<extra_id_56>", "<extra_id_57>", "<extra_id_58>", "<extra_id_59>", "<extra_id_60>", "<extra_id_61>", "<extra_id_62>", "<extra_id_63>", "<extra_id_64>", "<extra_id_65>", "<extra_id_66>", "<extra_id_67>", "<extra_id_68>", "<extra_id_69>", "<extra_id_70>", "<extra_id_71>", "<extra_id_72>", "<extra_id_73>", "<extra_id_74>", "<extra_id_75>", "<extra_id_76>", "<extra_id_77>", "<extra_id_78>", "<extra_id_79>", "<extra_id_80>", "<extra_id_81>", "<extra_id_82>", "<extra_id_83>", "<extra_id_84>", "<extra_id_85>", "<extra_id_86>", "<extra_id_87>", "<extra_id_88>", "<extra_id_89>", "<extra_id_90>", "<extra_id_91>", "<extra_id_92>", "<extra_id_93>", "<extra_id_94>", "<extra_id_95>", "<extra_id_96>", "<extra_id_97>", "<extra_id_98>", "<extra_id_99>"], "model_max_length": 512, "name_or_path": "models/small-1712500/", "special_tokens_map_file": "C:\\Users\\joshu/.cache\\huggingface\\transformers\\3ad6f8335c1b1ef8966245899d47dcf735abd134d21fd7d26f621fe45ac01184.c94798918c92ded6aeef2d2f0e666d2cc4145eca1aa6e1336fde07f2e13e2f46", "sp_model_kwargs": {}, "max_length": 512, "tokenizer_class": "T5Tokenizer"}

trainer_state.json

@@ -1,1116 +1,236 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 3.9405478794478936,
-  "global_step": 1100000,
+  "epoch": 3.678837487353996,
+  "global_step": 320000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.04,
-      "learning_rate": 4.8208841872978234e-05,
-      "loss": 0.0755,
-      "step": 10000
+      "epoch": 0.06,
+      "learning_rate": 4.971259082130047e-05,
+      "loss": 0.0475,
+      "step": 5000
     },
     {
-      "epoch": 0.07,
-      "learning_rate": 4.6417683745956466e-05,
-      "loss": 0.0721,
-      "step": 20000
-    },
-    {
-      "epoch": 0.07,
-      "eval_loss": 0.06494459509849548,
-      "eval_runtime": 754.7473,
-      "eval_samples_per_second": 82.19,
-      "eval_steps_per_second": 20.549,
-      "step": 20000
+      "epoch": 0.11,
+      "learning_rate": 4.9425181642600946e-05,
+      "loss": 0.0494,
+      "step": 10000
     },
     {
       "epoch": 0.11,
-      "learning_rate": 4.462652561893469e-05,
-      "loss": 0.0704,
-      "step": 30000
+      "eval_loss": 0.04334285110235214,
+      "eval_runtime": 233.2791,
+      "eval_samples_per_second": 82.862,
+      "eval_steps_per_second": 20.718,
+      "step": 10000
     },
     {
-      "epoch": 0.14,
-      "learning_rate": 4.2835367491912923e-05,
-      "loss": 0.0701,
-      "step": 40000
+      "epoch": 0.17,
+      "learning_rate": 4.913777246390141e-05,
+      "loss": 0.049,
+      "step": 15000
     },
     {
-      "epoch": 0.14,
-      "eval_loss": 0.06424280256032944,
-      "eval_runtime": 676.9024,
-      "eval_samples_per_second": 91.642,
-      "eval_steps_per_second": 22.912,
-      "step": 40000
+      "epoch": 0.23,
+      "learning_rate": 4.8850363285201876e-05,
+      "loss": 0.0477,
+      "step": 20000
     },
     {
-      "epoch": 0.18,
-      "learning_rate": 4.104420936489115e-05,
-      "loss": 0.0713,
-      "step": 50000
+      "epoch": 0.23,
+      "eval_loss": 0.04320817440748215,
+      "eval_runtime": 230.0333,
+      "eval_samples_per_second": 84.031,
+      "eval_steps_per_second": 21.01,
+      "step": 20000
     },
     {
-      "epoch": 0.21,
-      "learning_rate": 3.925305123786938e-05,
-      "loss": 0.0712,
-      "step": 60000
+      "epoch": 0.29,
+      "learning_rate": 4.856295410650235e-05,
+      "loss": 0.0494,
+      "step": 25000
     },
     {
-      "epoch": 0.21,
-      "eval_loss": 0.0624563954770565,
-      "eval_runtime": 678.8069,
-      "eval_samples_per_second": 91.385,
-      "eval_steps_per_second": 22.847,
-      "step": 60000
+      "epoch": 0.34,
+      "learning_rate": 4.827554492780282e-05,
+      "loss": 0.0464,
+      "step": 30000
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 3.746189311084761e-05,
-      "loss": 0.0698,
-      "step": 70000
+      "epoch": 0.34,
+      "eval_loss": 0.04396720230579376,
+      "eval_runtime": 235.2064,
+      "eval_samples_per_second": 82.183,
+      "eval_steps_per_second": 20.548,
+      "step": 30000
     },
     {
-      "epoch": 0.29,
-      "learning_rate": 3.5670734983825844e-05,
-      "loss": 0.0667,
-      "step": 80000
+      "epoch": 0.4,
+      "learning_rate": 4.7988135749103285e-05,
+      "loss": 0.0491,
+      "step": 35000
     },
     {
-      "epoch": 0.29,
-      "eval_loss": 0.06348983943462372,
-      "eval_runtime": 675.6836,
-      "eval_samples_per_second": 91.808,
-      "eval_steps_per_second": 22.953,
-      "step": 80000
+      "epoch": 0.46,
+      "learning_rate": 4.770072657040375e-05,
+      "loss": 0.0472,
+      "step": 40000
     },
     {
-      "epoch": 0.32,
-      "learning_rate": 3.3879576856804076e-05,
-      "loss": 0.0657,
-      "step": 90000
+      "epoch": 0.46,
+      "eval_loss": 0.04489925131201744,
+      "eval_runtime": 225.0238,
+      "eval_samples_per_second": 85.902,
+      "eval_steps_per_second": 21.478,
+      "step": 40000
     },
     {
-      "epoch": 0.36,
-      "learning_rate": 3.20884187297823e-05,
-      "loss": 0.0665,
-      "step": 100000
+      "epoch": 0.69,
+      "learning_rate": 4.655108985560563e-05,
+      "loss": 0.0484,
+      "step": 60000
     },
     {
-      "epoch": 0.36,
-      "eval_loss": 0.06186612322926521,
-      "eval_runtime": 676.1915,
-      "eval_samples_per_second": 91.739,
-      "eval_steps_per_second": 22.936,
-      "step": 100000
+      "epoch": 0.92,
+      "learning_rate": 4.540145314080751e-05,
+      "loss": 0.0481,
+      "step": 80000
     },
     {
-      "epoch": 0.39,
-      "learning_rate": 3.0297260602760537e-05,
-      "loss": 0.0657,
-      "step": 110000
+      "epoch": 0.92,
+      "eval_loss": 0.044077709317207336,
+      "eval_runtime": 226.2289,
+      "eval_samples_per_second": 85.444,
+      "eval_steps_per_second": 21.363,
+      "step": 80000
     },
     {
-      "epoch": 0.43,
-      "learning_rate": 2.8506102475738766e-05,
-      "loss": 0.0653,
-      "step": 120000
+      "epoch": 1.15,
+      "learning_rate": 4.4251816426009386e-05,
+      "loss": 0.0483,
+      "step": 100000
     },
     {
-      "epoch": 0.43,
-      "eval_loss": 0.061661478132009506,
-      "eval_runtime": 677.4648,
-      "eval_samples_per_second": 91.566,
-      "eval_steps_per_second": 22.893,
+      "epoch": 1.38,
+      "learning_rate": 4.310217971121126e-05,
+      "loss": 0.0463,
       "step": 120000
     },
     {
-      "epoch": 0.47,
-      "learning_rate": 2.6714944348716997e-05,
-      "loss": 0.0658,
-      "step": 130000
-    },
-    {
-      "epoch": 0.5,
-      "learning_rate": 2.4923786221695226e-05,
-      "loss": 0.0656,
-      "step": 140000
+      "epoch": 1.38,
+      "eval_loss": 0.04454144090414047,
+      "eval_runtime": 228.877,
+      "eval_samples_per_second": 84.456,
+      "eval_steps_per_second": 21.116,
+      "step": 120000
     },
     {
-      "epoch": 0.5,
-      "eval_loss": 0.06098250672221184,
-      "eval_runtime": 677.3485,
-      "eval_samples_per_second": 91.582,
-      "eval_steps_per_second": 22.897,
+      "epoch": 1.61,
+      "learning_rate": 4.1952542996413133e-05,
+      "loss": 0.0467,
       "step": 140000
     },
     {
-      "epoch": 0.54,
-      "learning_rate": 2.3132628094673455e-05,
-      "loss": 0.0649,
-      "step": 150000
-    },
-    {
-      "epoch": 0.57,
-      "learning_rate": 2.1341469967651687e-05,
-      "loss": 0.0648,
+      "epoch": 1.84,
+      "learning_rate": 4.0802906281615014e-05,
+      "loss": 0.0465,
       "step": 160000
     },
     {
-      "epoch": 0.57,
-      "eval_loss": 0.06077565252780914,
-      "eval_runtime": 677.1708,
-      "eval_samples_per_second": 91.606,
-      "eval_steps_per_second": 22.903,
+      "epoch": 1.84,
+      "eval_loss": 0.04413667321205139,
+      "eval_runtime": 228.9059,
+      "eval_samples_per_second": 84.445,
+      "eval_steps_per_second": 21.113,
       "step": 160000
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 1.9550311840629915e-05,
-      "loss": 0.0642,
-      "step": 170000
-    },
-    {
-      "epoch": 0.64,
-      "learning_rate": 1.7759153713608147e-05,
-      "loss": 0.0649,
-      "step": 180000
-    },
-    {
-      "epoch": 0.64,
-      "eval_loss": 0.05941811203956604,
-      "eval_runtime": 677.9109,
-      "eval_samples_per_second": 91.506,
-      "eval_steps_per_second": 22.878,
+      "epoch": 2.07,
+      "learning_rate": 3.965326956681689e-05,
+      "loss": 0.0469,
       "step": 180000
     },
     {
-      "epoch": 0.68,
-      "learning_rate": 1.5967995586586376e-05,
-      "loss": 0.0652,
-      "step": 190000
-    },
-    {
-      "epoch": 0.72,
-      "learning_rate": 1.4176837459564608e-05,
-      "loss": 0.0638,
+      "epoch": 2.3,
+      "learning_rate": 3.850363285201876e-05,
+      "loss": 0.0447,
       "step": 200000
     },
     {
-      "epoch": 0.72,
-      "eval_loss": 0.06042506918311119,
-      "eval_runtime": 678.1922,
-      "eval_samples_per_second": 91.468,
-      "eval_steps_per_second": 22.868,
+      "epoch": 2.3,
+      "eval_loss": 0.04485001042485237,
+      "eval_runtime": 237.9271,
+      "eval_samples_per_second": 81.243,
+      "eval_steps_per_second": 20.313,
       "step": 200000
     },
     {
-      "epoch": 0.75,
-      "learning_rate": 1.2385679332542836e-05,
-      "loss": 0.0621,
-      "step": 210000
-    },
-    {
-      "epoch": 0.79,
-      "learning_rate": 1.0594521205521066e-05,
-      "loss": 0.0643,
-      "step": 220000
-    },
-    {
-      "epoch": 0.79,
-      "eval_loss": 0.059569716453552246,
-      "eval_runtime": 688.4187,
-      "eval_samples_per_second": 90.109,
-      "eval_steps_per_second": 22.528,
+      "epoch": 2.53,
+      "learning_rate": 3.735399613722064e-05,
+      "loss": 0.0445,
       "step": 220000
     },
     {
-      "epoch": 0.82,
-      "learning_rate": 8.803363078499297e-06,
-      "loss": 0.0623,
-      "step": 230000
-    },
-    {
-      "epoch": 0.86,
-      "learning_rate": 7.012204951477527e-06,
-      "loss": 0.0646,
+      "epoch": 2.76,
+      "learning_rate": 3.620435942242252e-05,
+      "loss": 0.0435,
       "step": 240000
     },
     {
-      "epoch": 0.86,
-      "eval_loss": 0.059120796620845795,
-      "eval_runtime": 715.8659,
-      "eval_samples_per_second": 86.654,
-      "eval_steps_per_second": 21.665,
+      "epoch": 2.76,
+      "eval_loss": 0.04417626932263374,
+      "eval_runtime": 239.7656,
+      "eval_samples_per_second": 80.62,
+      "eval_steps_per_second": 20.157,
       "step": 240000
     },
     {
-      "epoch": 0.9,
-      "learning_rate": 4.5522104682445576e-05,
-      "loss": 0.0634,
-      "step": 250000
-    },
-    {
-      "epoch": 0.93,
-      "learning_rate": 4.53429888697434e-05,
-      "loss": 0.0648,
+      "epoch": 2.99,
+      "learning_rate": 3.505472270762439e-05,
+      "loss": 0.0443,
       "step": 260000
     },
     {
-      "epoch": 0.93,
-      "eval_loss": 0.06051425263285637,
-      "eval_runtime": 694.96,
-      "eval_samples_per_second": 89.261,
-      "eval_steps_per_second": 22.316,
-      "step": 260000
-    },
-    {
-      "epoch": 0.97,
-      "learning_rate": 4.516387305704122e-05,
-      "loss": 0.0644,
-      "step": 270000
-    },
-    {
-      "epoch": 1.0,
-      "learning_rate": 4.4984757244339046e-05,
-      "loss": 0.0655,
+      "epoch": 3.22,
+      "learning_rate": 3.390508599282627e-05,
+      "loss": 0.0425,
       "step": 280000
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 0.06069951504468918,
-      "eval_runtime": 679.0968,
-      "eval_samples_per_second": 91.346,
-      "eval_steps_per_second": 22.838,
+      "epoch": 3.22,
+      "eval_loss": 0.04555518552660942,
+      "eval_runtime": 239.0043,
+      "eval_samples_per_second": 80.877,
+      "eval_steps_per_second": 20.221,
       "step": 280000
     },
     {
-      "epoch": 1.04,
-      "learning_rate": 4.4805641431636866e-05,
-      "loss": 0.0652,
-      "step": 290000
-    },
-    {
-      "epoch": 1.07,
-      "learning_rate": 4.462652561893469e-05,
-      "loss": 0.0631,
-      "step": 300000
-    },
-    {
-      "epoch": 1.07,
-      "eval_loss": 0.06102127209305763,
-      "eval_runtime": 705.9715,
-      "eval_samples_per_second": 87.869,
-      "eval_steps_per_second": 21.968,
+      "epoch": 3.45,
+      "learning_rate": 3.2755449278028145e-05,
+      "loss": 0.0435,
       "step": 300000
     },
     {
-      "epoch": 1.11,
-      "learning_rate": 4.444740980623252e-05,
-      "loss": 0.064,
-      "step": 310000
-    },
-    {
-      "epoch": 1.15,
-      "learning_rate": 4.426829399353034e-05,
-      "loss": 0.0643,
+      "epoch": 3.68,
+      "learning_rate": 3.160581256323002e-05,
+      "loss": 0.0437,
       "step": 320000
     },
     {
-      "epoch": 1.15,
-      "eval_loss": 0.05971848592162132,
-      "eval_runtime": 698.4495,
-      "eval_samples_per_second": 88.815,
-      "eval_steps_per_second": 22.205,
+      "epoch": 3.68,
+      "eval_loss": 0.04474752023816109,
+      "eval_runtime": 291.4231,
+      "eval_samples_per_second": 66.33,
+      "eval_steps_per_second": 16.584,
       "step": 320000
-    },
-    {
-      "epoch": 1.18,
-      "learning_rate": 4.408917818082816e-05,
-      "loss": 0.0638,
-      "step": 330000
-    },
-    {
-      "epoch": 1.22,
-      "learning_rate": 4.391006236812599e-05,
-      "loss": 0.064,
-      "step": 340000
-    },
-    {
-      "epoch": 1.22,
-      "eval_loss": 0.05945688858628273,
-      "eval_runtime": 700.9514,
-      "eval_samples_per_second": 88.498,
-      "eval_steps_per_second": 22.126,
-      "step": 340000
-    },
-    {
-      "epoch": 1.25,
-      "learning_rate": 4.373094655542381e-05,
-      "loss": 0.0613,
-      "step": 350000
-    },
-    {
-      "epoch": 1.29,
-      "learning_rate": 4.355183074272163e-05,
-      "loss": 0.0635,
-      "step": 360000
-    },
-    {
-      "epoch": 1.29,
-      "eval_loss": 0.06002034246921539,
-      "eval_runtime": 697.4531,
-      "eval_samples_per_second": 88.942,
-      "eval_steps_per_second": 22.237,
-      "step": 360000
-    },
-    {
-      "epoch": 1.33,
-      "learning_rate": 4.337271493001945e-05,
-      "loss": 0.0635,
-      "step": 370000
-    },
-    {
-      "epoch": 1.36,
-      "learning_rate": 4.319359911731728e-05,
-      "loss": 0.0627,
-      "step": 380000
-    },
-    {
-      "epoch": 1.36,
-      "eval_loss": 0.0593414306640625,
-      "eval_runtime": 720.1404,
-      "eval_samples_per_second": 86.14,
-      "eval_steps_per_second": 21.536,
-      "step": 380000
-    },
-    {
-      "epoch": 1.4,
-      "learning_rate": 4.30144833046151e-05,
-      "loss": 0.0629,
-      "step": 390000
-    },
-    {
-      "epoch": 1.43,
-      "learning_rate": 4.2835367491912923e-05,
-      "loss": 0.0642,
-      "step": 400000
-    },
-    {
-      "epoch": 1.43,
-      "eval_loss": 0.059706129133701324,
-      "eval_runtime": 706.6838,
-      "eval_samples_per_second": 87.78,
-      "eval_steps_per_second": 21.946,
-      "step": 400000
-    },
-    {
-      "epoch": 1.47,
-      "learning_rate": 4.265625167921074e-05,
-      "loss": 0.062,
-      "step": 410000
-    },
-    {
-      "epoch": 1.5,
-      "learning_rate": 4.247713586650857e-05,
-      "loss": 0.0616,
-      "step": 420000
-    },
-    {
-      "epoch": 1.5,
-      "eval_loss": 0.058674536645412445,
-      "eval_runtime": 740.0142,
-      "eval_samples_per_second": 83.827,
-      "eval_steps_per_second": 20.958,
-      "step": 420000
-    },
-    {
-      "epoch": 1.54,
-      "learning_rate": 4.2298020053806394e-05,
-      "loss": 0.0611,
-      "step": 430000
-    },
-    {
-      "epoch": 1.58,
-      "learning_rate": 4.211890424110422e-05,
-      "loss": 0.0618,
-      "step": 440000
-    },
-    {
-      "epoch": 1.58,
-      "eval_loss": 0.05948682501912117,
-      "eval_runtime": 698.4731,
-      "eval_samples_per_second": 88.812,
-      "eval_steps_per_second": 22.204,
-      "step": 440000
-    },
-    {
-      "epoch": 1.61,
-      "learning_rate": 4.193978842840204e-05,
-      "loss": 0.062,
-      "step": 450000
-    },
-    {
-      "epoch": 1.65,
-      "learning_rate": 4.176067261569986e-05,
-      "loss": 0.0603,
-      "step": 460000
-    },
-    {
-      "epoch": 1.65,
-      "eval_loss": 0.059506043791770935,
-      "eval_runtime": 699.0665,
-      "eval_samples_per_second": 88.737,
-      "eval_steps_per_second": 22.185,
-      "step": 460000
-    },
-    {
-      "epoch": 1.68,
-      "learning_rate": 4.1581556802997684e-05,
-      "loss": 0.0606,
-      "step": 470000
-    },
-    {
-      "epoch": 1.72,
-      "learning_rate": 4.1402440990295504e-05,
-      "loss": 0.0596,
-      "step": 480000
-    },
-    {
-      "epoch": 1.72,
-      "eval_loss": 0.06113377958536148,
-      "eval_runtime": 687.5695,
-      "eval_samples_per_second": 90.221,
-      "eval_steps_per_second": 22.556,
-      "step": 480000
-    },
-    {
-      "epoch": 1.76,
-      "learning_rate": 4.122332517759333e-05,
-      "loss": 0.0604,
-      "step": 490000
-    },
-    {
-      "epoch": 1.79,
-      "learning_rate": 4.104420936489115e-05,
-      "loss": 0.0602,
-      "step": 500000
-    },
-    {
-      "epoch": 1.79,
-      "eval_loss": 0.058628179132938385,
-      "eval_runtime": 685.1917,
-      "eval_samples_per_second": 90.534,
-      "eval_steps_per_second": 22.635,
-      "step": 500000
-    },
-    {
-      "epoch": 1.83,
-      "learning_rate": 4.0865093552188975e-05,
-      "loss": 0.0588,
-      "step": 510000
-    },
-    {
-      "epoch": 1.86,
-      "learning_rate": 4.0685977739486794e-05,
-      "loss": 0.0603,
-      "step": 520000
-    },
-    {
-      "epoch": 1.86,
-      "eval_loss": 0.058195825666189194,
-      "eval_runtime": 684.8426,
-      "eval_samples_per_second": 90.58,
-      "eval_steps_per_second": 22.646,
-      "step": 520000
-    },
-    {
-      "epoch": 1.9,
-      "learning_rate": 4.050686192678462e-05,
-      "loss": 0.0606,
-      "step": 530000
-    },
-    {
-      "epoch": 1.93,
-      "learning_rate": 4.0327746114082445e-05,
-      "loss": 0.0603,
-      "step": 540000
-    },
-    {
-      "epoch": 1.93,
-      "eval_loss": 0.059178441762924194,
-      "eval_runtime": 686.71,
-      "eval_samples_per_second": 90.334,
-      "eval_steps_per_second": 22.584,
-      "step": 540000
-    },
-    {
-      "epoch": 1.97,
-      "learning_rate": 4.014863030138027e-05,
-      "loss": 0.0614,
-      "step": 550000
-    },
-    {
-      "epoch": 2.01,
-      "learning_rate": 3.996951448867809e-05,
-      "loss": 0.0596,
-      "step": 560000
-    },
-    {
-      "epoch": 2.01,
-      "eval_loss": 0.05809653550386429,
-      "eval_runtime": 686.1393,
-      "eval_samples_per_second": 90.409,
-      "eval_steps_per_second": 22.603,
-      "step": 560000
-    },
-    {
-      "epoch": 2.04,
-      "learning_rate": 3.9790398675975916e-05,
-      "loss": 0.0587,
-      "step": 570000
-    },
-    {
-      "epoch": 2.08,
-      "learning_rate": 3.9611282863273736e-05,
-      "loss": 0.0584,
-      "step": 580000
-    },
-    {
-      "epoch": 2.08,
-      "eval_loss": 0.05719467252492905,
-      "eval_runtime": 686.9168,
-      "eval_samples_per_second": 90.306,
-      "eval_steps_per_second": 22.578,
-      "step": 580000
-    },
-    {
-      "epoch": 2.11,
-      "learning_rate": 3.943216705057156e-05,
-      "loss": 0.0592,
-      "step": 590000
-    },
-    {
-      "epoch": 2.15,
-      "learning_rate": 3.925305123786938e-05,
-      "loss": 0.0592,
-      "step": 600000
-    },
-    {
-      "epoch": 2.15,
-      "eval_loss": 0.05839799717068672,
-      "eval_runtime": 691.5693,
-      "eval_samples_per_second": 89.699,
-      "eval_steps_per_second": 22.426,
-      "step": 600000
-    },
-    {
-      "epoch": 2.19,
-      "learning_rate": 3.9073935425167206e-05,
-      "loss": 0.0583,
-      "step": 610000
-    },
-    {
-      "epoch": 2.22,
-      "learning_rate": 3.8894819612465026e-05,
-      "loss": 0.0598,
-      "step": 620000
-    },
-    {
-      "epoch": 2.22,
-      "eval_loss": 0.05699307098984718,
-      "eval_runtime": 703.0844,
-      "eval_samples_per_second": 88.23,
-      "eval_steps_per_second": 22.059,
-      "step": 620000
-    },
-    {
-      "epoch": 2.26,
-      "learning_rate": 3.871570379976285e-05,
-      "loss": 0.0587,
-      "step": 630000
-    },
-    {
-      "epoch": 2.29,
-      "learning_rate": 3.853658798706067e-05,
-      "loss": 0.058,
-      "step": 640000
-    },
-    {
-      "epoch": 2.29,
-      "eval_loss": 0.05786525830626488,
-      "eval_runtime": 719.4112,
-      "eval_samples_per_second": 86.227,
-      "eval_steps_per_second": 21.558,
-      "step": 640000
-    },
-    {
-      "epoch": 2.33,
-      "learning_rate": 3.8357472174358497e-05,
-      "loss": 0.0579,
-      "step": 650000
-    },
-    {
-      "epoch": 2.36,
-      "learning_rate": 3.817835636165632e-05,
-      "loss": 0.0576,
-      "step": 660000
-    },
-    {
-      "epoch": 2.36,
-      "eval_loss": 0.05708477646112442,
-      "eval_runtime": 740.1092,
-      "eval_samples_per_second": 83.816,
-      "eval_steps_per_second": 20.955,
-      "step": 660000
-    },
-    {
-      "epoch": 2.4,
-      "learning_rate": 3.799924054895415e-05,
-      "loss": 0.0578,
-      "step": 670000
-    },
-    {
-      "epoch": 2.44,
-      "learning_rate": 3.782012473625197e-05,
-      "loss": 0.0586,
-      "step": 680000
-    },
-    {
-      "epoch": 2.44,
-      "eval_loss": 0.05821482464671135,
-      "eval_runtime": 697.0334,
-      "eval_samples_per_second": 88.996,
-      "eval_steps_per_second": 22.25,
-      "step": 680000
-    },
-    {
-      "epoch": 2.47,
-      "learning_rate": 3.764100892354979e-05,
-      "loss": 0.0579,
-      "step": 690000
-    },
-    {
-      "epoch": 2.51,
-      "learning_rate": 3.746189311084761e-05,
-      "loss": 0.0579,
-      "step": 700000
-    },
-    {
-      "epoch": 2.51,
-      "eval_loss": 0.057255882769823074,
-      "eval_runtime": 693.1254,
-      "eval_samples_per_second": 89.498,
-      "eval_steps_per_second": 22.375,
-      "step": 700000
-    },
-    {
-      "epoch": 2.54,
-      "learning_rate": 3.728277729814544e-05,
-      "loss": 0.0586,
-      "step": 710000
-    },
-    {
-      "epoch": 2.58,
-      "learning_rate": 3.710366148544326e-05,
-      "loss": 0.0583,
-      "step": 720000
-    },
-    {
-      "epoch": 2.58,
-      "eval_loss": 0.0572069026529789,
-      "eval_runtime": 691.2275,
-      "eval_samples_per_second": 89.743,
-      "eval_steps_per_second": 22.437,
-      "step": 720000
-    },
-    {
-      "epoch": 2.62,
-      "learning_rate": 3.6924545672741083e-05,
-      "loss": 0.0591,
-      "step": 730000
-    },
-    {
-      "epoch": 2.65,
-      "learning_rate": 3.67454298600389e-05,
-      "loss": 0.0573,
-      "step": 740000
-    },
-    {
-      "epoch": 2.65,
-      "eval_loss": 0.05713275447487831,
-      "eval_runtime": 696.2667,
-      "eval_samples_per_second": 89.094,
-      "eval_steps_per_second": 22.275,
-      "step": 740000
-    },
-    {
-      "epoch": 2.69,
-      "learning_rate": 3.656631404733673e-05,
-      "loss": 0.0559,
-      "step": 750000
-    },
-    {
-      "epoch": 2.72,
-      "learning_rate": 3.638719823463455e-05,
-      "loss": 0.0571,
-      "step": 760000
-    },
-    {
-      "epoch": 2.72,
-      "eval_loss": 0.056393299251794815,
-      "eval_runtime": 683.9753,
-      "eval_samples_per_second": 90.695,
-      "eval_steps_per_second": 22.675,
-      "step": 760000
-    },
-    {
-      "epoch": 2.76,
-      "learning_rate": 3.6208082421932374e-05,
-      "loss": 0.0572,
-      "step": 770000
-    },
-    {
-      "epoch": 2.79,
-      "learning_rate": 3.60289666092302e-05,
-      "loss": 0.0571,
-      "step": 780000
-    },
-    {
-      "epoch": 2.79,
-      "eval_loss": 0.057615023106336594,
-      "eval_runtime": 683.0895,
-      "eval_samples_per_second": 90.812,
-      "eval_steps_per_second": 22.704,
-      "step": 780000
-    },
-    {
-      "epoch": 2.83,
-      "learning_rate": 3.5849850796528025e-05,
-      "loss": 0.0575,
-      "step": 790000
-    },
-    {
-      "epoch": 2.87,
-      "learning_rate": 3.5670734983825844e-05,
-      "loss": 0.0556,
-      "step": 800000
-    },
-    {
-      "epoch": 2.87,
-      "eval_loss": 0.055679138749837875,
-      "eval_runtime": 682.7397,
-      "eval_samples_per_second": 90.859,
-      "eval_steps_per_second": 22.716,
-      "step": 800000
-    },
-    {
-      "epoch": 2.9,
-      "learning_rate": 3.549161917112367e-05,
-      "loss": 0.057,
-      "step": 810000
-    },
-    {
-      "epoch": 2.94,
-      "learning_rate": 3.531250335842149e-05,
-      "loss": 0.057,
-      "step": 820000
-    },
-    {
-      "epoch": 2.94,
-      "eval_loss": 0.056380968540906906,
-      "eval_runtime": 682.9026,
-      "eval_samples_per_second": 90.837,
-      "eval_steps_per_second": 22.71,
-      "step": 820000
-    },
-    {
-      "epoch": 2.97,
-      "learning_rate": 3.5133387545719315e-05,
-      "loss": 0.0558,
-      "step": 830000
-    },
-    {
-      "epoch": 3.01,
-      "learning_rate": 3.4954271733017135e-05,
-      "loss": 0.0561,
-      "step": 840000
-    },
-    {
-      "epoch": 3.01,
-      "eval_loss": 0.05610540881752968,
-      "eval_runtime": 684.8005,
-      "eval_samples_per_second": 90.586,
-      "eval_steps_per_second": 22.647,
-      "step": 840000
-    },
-    {
-      "epoch": 3.04,
-      "learning_rate": 3.477515592031496e-05,
-      "loss": 0.0558,
-      "step": 850000
-    },
-    {
-      "epoch": 3.08,
-      "learning_rate": 3.459604010761278e-05,
-      "loss": 0.0552,
-      "step": 860000
-    },
-    {
-      "epoch": 3.08,
-      "eval_loss": 0.05571727082133293,
-      "eval_runtime": 684.9047,
-      "eval_samples_per_second": 90.572,
-      "eval_steps_per_second": 22.644,
-      "step": 860000
-    },
-    {
-      "epoch": 3.12,
-      "learning_rate": 3.4416924294910605e-05,
-      "loss": 0.0566,
-      "step": 870000
-    },
-    {
-      "epoch": 3.15,
-      "learning_rate": 3.4237808482208425e-05,
-      "loss": 0.0572,
-      "step": 880000
-    },
-    {
-      "epoch": 3.15,
-      "eval_loss": 0.05569184198975563,
-      "eval_runtime": 685.6335,
-      "eval_samples_per_second": 90.475,
-      "eval_steps_per_second": 22.62,
-      "step": 880000
-    },
-    {
-      "epoch": 3.19,
-      "learning_rate": 3.405869266950625e-05,
-      "loss": 0.0558,
-      "step": 890000
-    },
-    {
-      "epoch": 3.22,
-      "learning_rate": 3.3879576856804076e-05,
-      "loss": 0.0559,
-      "step": 900000
-    },
-    {
-      "epoch": 3.22,
-      "eval_loss": 0.055385466665029526,
-      "eval_runtime": 689.1891,
-      "eval_samples_per_second": 90.009,
-      "eval_steps_per_second": 22.503,
-      "step": 900000
-    },
-    {
-      "epoch": 3.26,
-      "learning_rate": 3.37004610441019e-05,
-      "loss": 0.0557,
-      "step": 910000
-    },
-    {
-      "epoch": 3.3,
-      "learning_rate": 3.352134523139972e-05,
-      "loss": 0.0548,
-      "step": 920000
-    },
-    {
-      "epoch": 3.3,
-      "eval_loss": 0.05598163977265358,
-      "eval_runtime": 684.1191,
-      "eval_samples_per_second": 90.676,
-      "eval_steps_per_second": 22.67,
-      "step": 920000
-    },
-    {
-      "epoch": 3.33,
-      "learning_rate": 3.334222941869754e-05,
-      "loss": 0.0552,
-      "step": 930000
-    },
-    {
-      "epoch": 3.37,
-      "learning_rate": 3.3163113605995366e-05,
-      "loss": 0.0565,
-      "step": 940000
-    },
-    {
-      "epoch": 3.37,
-      "eval_loss": 0.055193424224853516,
-      "eval_runtime": 732.5988,
-      "eval_samples_per_second": 84.675,
-      "eval_steps_per_second": 21.17,
-      "step": 940000
-    },
-    {
-      "epoch": 3.4,
-      "learning_rate": 3.2983997793293186e-05,
-      "loss": 0.0557,
-      "step": 950000
-    },
-    {
-      "epoch": 3.44,
-      "learning_rate": 3.280488198059101e-05,
-      "loss": 0.055,
-      "step": 960000
-    },
-    {
-      "epoch": 3.44,
-      "eval_loss": 0.05571094900369644,
-      "eval_runtime": 724.8884,
-      "eval_samples_per_second": 85.576,
-      "eval_steps_per_second": 21.395,
-      "step": 960000
-    },
-    {
-      "epoch": 3.47,
-      "learning_rate": 3.262576616788883e-05,
-      "loss": 0.0556,
-      "step": 970000
-    },
-    {
-      "epoch": 3.51,
-      "learning_rate": 3.2446650355186657e-05,
-      "loss": 0.0563,
-      "step": 980000
-    },
-    {
-      "epoch": 3.51,
-      "eval_loss": 0.05553479492664337,
-      "eval_runtime": 714.2985,
-      "eval_samples_per_second": 86.845,
-      "eval_steps_per_second": 21.712,
-      "step": 980000
-    },
-    {
-      "epoch": 3.55,
-      "learning_rate": 3.2267534542484476e-05,
-      "loss": 0.0552,
-      "step": 990000
-    },
-    {
-      "epoch": 3.58,
-      "learning_rate": 3.20884187297823e-05,
-      "loss": 0.0553,
-      "step": 1000000
-    },
-    {
-      "epoch": 3.58,
-      "eval_loss": 0.05587638169527054,
-      "eval_runtime": 709.3024,
-      "eval_samples_per_second": 87.456,
-      "eval_steps_per_second": 21.865,
-      "step": 1000000
-    },
-    {
-      "epoch": 3.62,
-      "learning_rate": 3.190930291708013e-05,
-      "loss": 0.0544,
-      "step": 1010000
-    },
-    {
-      "epoch": 3.65,
-      "learning_rate": 3.1730187104377953e-05,
-      "loss": 0.0545,
-      "step": 1020000
-    },
-    {
-      "epoch": 3.65,
-      "eval_loss": 0.05536632239818573,
-      "eval_runtime": 696.888,
-      "eval_samples_per_second": 89.014,
-      "eval_steps_per_second": 22.255,
-      "step": 1020000
-    },
-    {
-      "epoch": 3.69,
-      "learning_rate": 3.155107129167577e-05,
-      "loss": 0.0543,
-      "step": 1030000
-    },
-    {
-      "epoch": 3.73,
-      "learning_rate": 3.13719554789736e-05,
-      "loss": 0.0544,
-      "step": 1040000
-    },
-    {
-      "epoch": 3.73,
-      "eval_loss": 0.05515935644507408,
-      "eval_runtime": 706.1745,
-      "eval_samples_per_second": 87.844,
-      "eval_steps_per_second": 21.962,
-      "step": 1040000
-    },
-    {
-      "epoch": 3.76,
-      "learning_rate": 3.119283966627142e-05,
-      "loss": 0.0534,
-      "step": 1050000
-    },
-    {
-      "epoch": 3.8,
-      "learning_rate": 3.1013723853569243e-05,
-      "loss": 0.0552,
-      "step": 1060000
-    },
-    {
-      "epoch": 3.8,
-      "eval_loss": 0.054923560470342636,
-      "eval_runtime": 716.4597,
-      "eval_samples_per_second": 86.583,
-      "eval_steps_per_second": 21.647,
-      "step": 1060000
-    },
-    {
-      "epoch": 3.83,
-      "learning_rate": 3.083460804086706e-05,
-      "loss": 0.055,
-      "step": 1070000
-    },
-    {
-      "epoch": 3.87,
-      "learning_rate": 3.065549222816489e-05,
-      "loss": 0.0547,
-      "step": 1080000
-    },
-    {
-      "epoch": 3.87,
-      "eval_loss": 0.05532824248075485,
-      "eval_runtime": 713.1552,
-      "eval_samples_per_second": 86.984,
-      "eval_steps_per_second": 21.747,
-      "step": 1080000
-    },
-    {
-      "epoch": 3.9,
-      "learning_rate": 3.047637641546271e-05,
-      "loss": 0.0526,
-      "step": 1090000
-    },
-    {
-      "epoch": 3.94,
-      "learning_rate": 3.0297260602760537e-05,
-      "loss": 0.0536,
-      "step": 1100000
-    },
-    {
-      "epoch": 3.94,
-      "eval_loss": 0.05594659596681595,
-      "eval_runtime": 765.7749,
-      "eval_samples_per_second": 81.007,
-      "eval_steps_per_second": 20.253,
-      "step": 1100000
     }
   ],
-  "max_steps": 2791490,
+  "max_steps": 869840,
   "num_train_epochs": 10,
-  "total_flos": 7.350678561491497e+17,
+  "total_flos": 2.2518483734502605e+17,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:01e951f41d1c838fb03d74c0997ade49e243ec0d07cb18ff2559ff234bae126c
+oid sha256:981e074dd006f0b82f1adae94afef8c0b53f496530dbead237c5195367cb5167
 size 3119