Model save

README.md

@@ -27,7 +27,7 @@ print(output["generated_text"])
 
 ## Training procedure
 
-[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/is-bang-an/huggingface/runs/ebvrh0v0) 
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/is-bang-an/huggingface/runs/yu9go9ug) 
 
 
 This model was trained with SFT.

all_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 0.999259807549963,
     "total_flos": 76888336760832.0,
-    "train_loss": 0.7676037124351219,
-    "train_runtime": 4810.4535,
+    "train_loss": 0.7676166036393908,
+    "train_runtime": 4800.2526,
     "train_samples": 16610,
-    "train_samples_per_second": 4.492,
-    "train_steps_per_second": 0.14
+    "train_samples_per_second": 4.502,
+    "train_steps_per_second": 0.141
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b5e862a573b95f6b408a790d71c3b1a5cc6f2b1fa51522456a41bdedebea2060
+oid sha256:9ebfb6105b435026630d68e72bb3eff23933338abb8a196c9e7b6f4903ecfc57
 size 3087467144

train_results.json

@@ -1,9 +1,9 @@
 {
     "epoch": 0.999259807549963,
     "total_flos": 76888336760832.0,
-    "train_loss": 0.7676037124351219,
-    "train_runtime": 4810.4535,
+    "train_loss": 0.7676166036393908,
+    "train_runtime": 4800.2526,
     "train_samples": 16610,
-    "train_samples_per_second": 4.492,
-    "train_steps_per_second": 0.14
+    "train_samples_per_second": 4.502,
+    "train_steps_per_second": 0.141
 }

trainer_state.json

@@ -10,993 +10,993 @@
   "log_history": [
     {
       "epoch": 0.007401924500370096,
-      "grad_norm": 2.6741910876480133,
+      "grad_norm": 2.6709367750052144,
       "learning_rate": 1.4705882352941177e-06,
       "loss": 1.0786,
       "step": 5
     },
     {
       "epoch": 0.014803849000740192,
-      "grad_norm": 2.2890043514790124,
+      "grad_norm": 2.2871070474818787,
       "learning_rate": 2.9411764705882355e-06,
-      "loss": 1.0901,
+      "loss": 1.0902,
       "step": 10
     },
     {
       "epoch": 0.02220577350111029,
-      "grad_norm": 1.4158383642752441,
+      "grad_norm": 1.4182795521636087,
       "learning_rate": 4.411764705882353e-06,
-      "loss": 1.0619,
+      "loss": 1.062,
       "step": 15
     },
     {
       "epoch": 0.029607698001480384,
-      "grad_norm": 1.3744246179214759,
+      "grad_norm": 1.3764280404457665,
       "learning_rate": 5.882352941176471e-06,
-      "loss": 1.0087,
+      "loss": 1.0086,
       "step": 20
     },
     {
       "epoch": 0.037009622501850484,
-      "grad_norm": 1.1138825641334238,
+      "grad_norm": 1.1156261369748648,
       "learning_rate": 7.352941176470589e-06,
       "loss": 0.9684,
       "step": 25
     },
     {
       "epoch": 0.04441154700222058,
-      "grad_norm": 0.9533582253776373,
+      "grad_norm": 0.9545532440344103,
       "learning_rate": 8.823529411764707e-06,
       "loss": 0.9217,
       "step": 30
     },
     {
       "epoch": 0.05181347150259067,
-      "grad_norm": 0.7615937992294558,
+      "grad_norm": 0.7611871553908859,
       "learning_rate": 1.0294117647058823e-05,
       "loss": 0.8858,
       "step": 35
     },
     {
       "epoch": 0.05921539600296077,
-      "grad_norm": 0.685895935004526,
+      "grad_norm": 0.6854484767024068,
       "learning_rate": 1.1764705882352942e-05,
       "loss": 0.8631,
       "step": 40
     },
     {
       "epoch": 0.06661732050333087,
-      "grad_norm": 0.7458205571085653,
+      "grad_norm": 0.750666680519459,
       "learning_rate": 1.323529411764706e-05,
-      "loss": 0.8486,
+      "loss": 0.8487,
       "step": 45
     },
     {
       "epoch": 0.07401924500370097,
-      "grad_norm": 0.81838531500722,
+      "grad_norm": 0.8172244317532135,
       "learning_rate": 1.4705882352941179e-05,
       "loss": 0.861,
       "step": 50
     },
     {
       "epoch": 0.08142116950407106,
-      "grad_norm": 0.6242801536100675,
+      "grad_norm": 0.6258688920241917,
       "learning_rate": 1.6176470588235296e-05,
       "loss": 0.8545,
       "step": 55
     },
     {
       "epoch": 0.08882309400444116,
-      "grad_norm": 0.7316076425219573,
+      "grad_norm": 0.7311294156696475,
       "learning_rate": 1.7647058823529414e-05,
       "loss": 0.8293,
       "step": 60
     },
     {
       "epoch": 0.09622501850481126,
-      "grad_norm": 0.6844867257520654,
+      "grad_norm": 0.6829997052596827,
       "learning_rate": 1.911764705882353e-05,
-      "loss": 0.8258,
+      "loss": 0.8259,
       "step": 65
     },
     {
       "epoch": 0.10362694300518134,
-      "grad_norm": 0.765785250790924,
+      "grad_norm": 0.769333158518139,
       "learning_rate": 1.9999464266898485e-05,
       "loss": 0.821,
       "step": 70
     },
     {
       "epoch": 0.11102886750555144,
-      "grad_norm": 0.9729016810538598,
+      "grad_norm": 0.9661291332835485,
       "learning_rate": 1.9993437928712977e-05,
       "loss": 0.8164,
       "step": 75
     },
     {
       "epoch": 0.11843079200592153,
-      "grad_norm": 0.8225241303829622,
+      "grad_norm": 0.8313345878160039,
       "learning_rate": 1.998071963486563e-05,
       "loss": 0.8061,
       "step": 80
     },
     {
       "epoch": 0.12583271650629163,
-      "grad_norm": 0.6568777162044267,
+      "grad_norm": 0.6599625325000631,
       "learning_rate": 1.9961317901970953e-05,
-      "loss": 0.7946,
+      "loss": 0.7945,
       "step": 85
     },
     {
       "epoch": 0.13323464100666174,
-      "grad_norm": 0.7526781560851372,
+      "grad_norm": 0.7556971029869796,
       "learning_rate": 1.993524572210807e-05,
-      "loss": 0.7947,
+      "loss": 0.7946,
       "step": 90
     },
     {
       "epoch": 0.14063656550703182,
-      "grad_norm": 0.789352241305715,
+      "grad_norm": 0.8000011001473601,
       "learning_rate": 1.990252055412077e-05,
-      "loss": 0.7906,
+      "loss": 0.7905,
       "step": 95
     },
     {
       "epoch": 0.14803849000740193,
-      "grad_norm": 0.7824165232548723,
+      "grad_norm": 0.7843797614751532,
       "learning_rate": 1.9863164311926433e-05,
       "loss": 0.8172,
       "step": 100
     },
     {
       "epoch": 0.14803849000740193,
-      "eval_loss": 0.8197923302650452,
-      "eval_runtime": 7.1026,
-      "eval_samples_per_second": 18.022,
-      "eval_steps_per_second": 2.253,
+      "eval_loss": 0.8196982145309448,
+      "eval_runtime": 7.0817,
+      "eval_samples_per_second": 18.075,
+      "eval_steps_per_second": 2.259,
       "step": 100
     },
     {
       "epoch": 0.15544041450777202,
-      "grad_norm": 0.7146335889959788,
+      "grad_norm": 0.713372656728166,
       "learning_rate": 1.981720334984174e-05,
       "loss": 0.7921,
       "step": 105
     },
     {
       "epoch": 0.16284233900814213,
-      "grad_norm": 0.7105261534052547,
+      "grad_norm": 0.7107830515856771,
       "learning_rate": 1.9764668444934853e-05,
       "loss": 0.7858,
       "step": 110
     },
     {
       "epoch": 0.1702442635085122,
-      "grad_norm": 0.7208660385802352,
+      "grad_norm": 0.7210451915786353,
       "learning_rate": 1.970559477641606e-05,
       "loss": 0.7631,
       "step": 115
     },
     {
       "epoch": 0.17764618800888232,
-      "grad_norm": 0.7351900176129011,
+      "grad_norm": 0.7328353666424424,
       "learning_rate": 1.9640021902080523e-05,
       "loss": 0.7929,
       "step": 120
     },
     {
       "epoch": 0.1850481125092524,
-      "grad_norm": 0.6269075384999976,
+      "grad_norm": 0.625857980018156,
       "learning_rate": 1.9567993731818988e-05,
-      "loss": 0.7914,
+      "loss": 0.7916,
       "step": 125
     },
     {
       "epoch": 0.19245003700962252,
-      "grad_norm": 0.7491156358068426,
+      "grad_norm": 0.7455061248180267,
       "learning_rate": 1.9489558498214197e-05,
-      "loss": 0.7841,
+      "loss": 0.7842,
       "step": 130
     },
     {
       "epoch": 0.1998519615099926,
-      "grad_norm": 0.6493807633307159,
+      "grad_norm": 0.6529537900630853,
       "learning_rate": 1.9404768724242667e-05,
-      "loss": 0.7704,
+      "loss": 0.7703,
       "step": 135
     },
     {
       "epoch": 0.20725388601036268,
-      "grad_norm": 0.7309098161737758,
+      "grad_norm": 0.7294370711408014,
       "learning_rate": 1.931368118810346e-05,
       "loss": 0.7947,
       "step": 140
     },
     {
       "epoch": 0.2146558105107328,
-      "grad_norm": 0.7526989843144094,
+      "grad_norm": 0.7485205785730842,
       "learning_rate": 1.92163568851975e-05,
-      "loss": 0.7757,
+      "loss": 0.7756,
       "step": 145
     },
     {
       "epoch": 0.22205773501110287,
-      "grad_norm": 0.8027859753874074,
+      "grad_norm": 0.8058314493735931,
       "learning_rate": 1.911286098728296e-05,
       "loss": 0.772,
       "step": 150
     },
     {
       "epoch": 0.22945965951147299,
-      "grad_norm": 0.6979819030971295,
+      "grad_norm": 0.700914939266579,
       "learning_rate": 1.900326279883392e-05,
-      "loss": 0.8015,
+      "loss": 0.8014,
       "step": 155
     },
     {
       "epoch": 0.23686158401184307,
-      "grad_norm": 0.7033001242795566,
+      "grad_norm": 0.7020963336241166,
       "learning_rate": 1.8887635710631716e-05,
-      "loss": 0.8045,
+      "loss": 0.8046,
       "step": 160
     },
     {
       "epoch": 0.24426350851221318,
-      "grad_norm": 0.6977045359860305,
+      "grad_norm": 0.7005897864198626,
       "learning_rate": 1.8766057150619865e-05,
       "loss": 0.7775,
       "step": 165
     },
     {
       "epoch": 0.25166543301258326,
-      "grad_norm": 0.7079499649707683,
+      "grad_norm": 0.7058185749313685,
       "learning_rate": 1.8638608532055635e-05,
       "loss": 0.7947,
       "step": 170
     },
     {
       "epoch": 0.25906735751295334,
-      "grad_norm": 0.6635507596196774,
+      "grad_norm": 0.664038868478026,
       "learning_rate": 1.8505375198992856e-05,
-      "loss": 0.783,
+      "loss": 0.7831,
       "step": 175
     },
     {
       "epoch": 0.2664692820133235,
-      "grad_norm": 0.7152344314350637,
+      "grad_norm": 0.72014538282932,
       "learning_rate": 1.836644636913258e-05,
       "loss": 0.7542,
       "step": 180
     },
     {
       "epoch": 0.27387120651369357,
-      "grad_norm": 0.7762491808113644,
+      "grad_norm": 0.7770026346024808,
       "learning_rate": 1.8221915074079764e-05,
-      "loss": 0.7778,
+      "loss": 0.7779,
       "step": 185
     },
     {
       "epoch": 0.28127313101406365,
-      "grad_norm": 0.8629438062163043,
+      "grad_norm": 0.865663637067605,
       "learning_rate": 1.8071878097046064e-05,
-      "loss": 0.7564,
+      "loss": 0.7563,
       "step": 190
     },
     {
       "epoch": 0.28867505551443373,
-      "grad_norm": 0.6204773442950375,
+      "grad_norm": 0.6184476258823719,
       "learning_rate": 1.7916435908040413e-05,
-      "loss": 0.7723,
+      "loss": 0.7724,
       "step": 195
     },
     {
       "epoch": 0.29607698001480387,
-      "grad_norm": 0.6428760976160367,
+      "grad_norm": 0.6494917102142564,
       "learning_rate": 1.7755692596590778e-05,
-      "loss": 0.7745,
+      "loss": 0.7746,
       "step": 200
     },
     {
       "epoch": 0.29607698001480387,
-      "eval_loss": 0.7898457050323486,
-      "eval_runtime": 7.0928,
-      "eval_samples_per_second": 18.047,
-      "eval_steps_per_second": 2.256,
+      "eval_loss": 0.7898926138877869,
+      "eval_runtime": 7.0802,
+      "eval_samples_per_second": 18.079,
+      "eval_steps_per_second": 2.26,
       "step": 200
     },
     {
       "epoch": 0.30347890451517395,
-      "grad_norm": 0.6966664120903373,
+      "grad_norm": 0.6973010604211403,
       "learning_rate": 1.7589755802042188e-05,
-      "loss": 0.7729,
+      "loss": 0.773,
       "step": 205
     },
     {
       "epoch": 0.31088082901554404,
-      "grad_norm": 0.6917393371799728,
+      "grad_norm": 0.6949606493395256,
       "learning_rate": 1.7418736641477636e-05,
       "loss": 0.7561,
       "step": 210
     },
     {
       "epoch": 0.3182827535159141,
-      "grad_norm": 0.6286907442457873,
+      "grad_norm": 0.6262569109444988,
       "learning_rate": 1.7242749635310222e-05,
       "loss": 0.758,
       "step": 215
     },
     {
       "epoch": 0.32568467801628426,
-      "grad_norm": 0.6585809174451455,
+      "grad_norm": 0.6545508953720698,
       "learning_rate": 1.7061912630596252e-05,
-      "loss": 0.7605,
+      "loss": 0.7604,
       "step": 220
     },
     {
       "epoch": 0.33308660251665434,
-      "grad_norm": 0.7168433708612332,
+      "grad_norm": 0.7178597609986006,
       "learning_rate": 1.6876346722120747e-05,
-      "loss": 0.7752,
+      "loss": 0.7753,
       "step": 225
     },
     {
       "epoch": 0.3404885270170244,
-      "grad_norm": 0.688720551345225,
+      "grad_norm": 0.6912584764916417,
       "learning_rate": 1.6686176171308125e-05,
       "loss": 0.7977,
       "step": 230
     },
     {
       "epoch": 0.3478904515173945,
-      "grad_norm": 0.6673718182403205,
+      "grad_norm": 0.6748238681034779,
       "learning_rate": 1.6491528323012412e-05,
       "loss": 0.7595,
       "step": 235
     },
     {
       "epoch": 0.35529237601776464,
-      "grad_norm": 0.6501486801291453,
+      "grad_norm": 0.6568455110309562,
       "learning_rate": 1.6292533520242663e-05,
       "loss": 0.7624,
       "step": 240
     },
     {
       "epoch": 0.3626943005181347,
-      "grad_norm": 0.6681605003188212,
+      "grad_norm": 0.6696394227198411,
       "learning_rate": 1.6089325016880737e-05,
-      "loss": 0.7532,
+      "loss": 0.7533,
       "step": 245
     },
     {
       "epoch": 0.3700962250185048,
-      "grad_norm": 0.6827453032262573,
+      "grad_norm": 0.6863856043690244,
       "learning_rate": 1.588203888844982e-05,
-      "loss": 0.7682,
+      "loss": 0.7684,
       "step": 250
     },
     {
       "epoch": 0.3774981495188749,
-      "grad_norm": 0.6452082624974447,
+      "grad_norm": 0.6413183265768512,
       "learning_rate": 1.5670813940993504e-05,
-      "loss": 0.741,
+      "loss": 0.7409,
       "step": 255
     },
     {
       "epoch": 0.38490007401924503,
-      "grad_norm": 0.5933793013042101,
+      "grad_norm": 0.592394054979568,
       "learning_rate": 1.5455791618126407e-05,
-      "loss": 0.7335,
+      "loss": 0.7336,
       "step": 260
     },
     {
       "epoch": 0.3923019985196151,
-      "grad_norm": 0.6241768840270675,
+      "grad_norm": 0.6278449742084541,
       "learning_rate": 1.5237115906318565e-05,
       "loss": 0.7572,
       "step": 265
     },
     {
       "epoch": 0.3997039230199852,
-      "grad_norm": 0.6624256460480871,
+      "grad_norm": 0.6678910847724688,
       "learning_rate": 1.5014933238477069e-05,
       "loss": 0.7379,
       "step": 270
     },
     {
       "epoch": 0.4071058475203553,
-      "grad_norm": 0.6926850270930469,
+      "grad_norm": 0.6936079894912379,
       "learning_rate": 1.4789392395889468e-05,
       "loss": 0.7633,
       "step": 275
     },
     {
       "epoch": 0.41450777202072536,
-      "grad_norm": 0.6446075865715499,
+      "grad_norm": 0.6427795800840255,
       "learning_rate": 1.4560644408594602e-05,
       "loss": 0.7442,
       "step": 280
     },
     {
       "epoch": 0.4219096965210955,
-      "grad_norm": 0.6587998534583325,
+      "grad_norm": 0.6507557085189994,
       "learning_rate": 1.432884245424761e-05,
       "loss": 0.7556,
       "step": 285
     },
     {
       "epoch": 0.4293116210214656,
-      "grad_norm": 0.6634427746618408,
+      "grad_norm": 0.6624582442770054,
       "learning_rate": 1.4094141755546816e-05,
       "loss": 0.7833,
       "step": 290
     },
     {
       "epoch": 0.43671354552183567,
-      "grad_norm": 0.6279777480504406,
+      "grad_norm": 0.6260844530638364,
       "learning_rate": 1.3856699476291176e-05,
-      "loss": 0.7428,
+      "loss": 0.7429,
       "step": 295
     },
     {
       "epoch": 0.44411547002220575,
-      "grad_norm": 0.6828209851102316,
+      "grad_norm": 0.6817749702455657,
       "learning_rate": 1.3616674616137902e-05,
       "loss": 0.7644,
       "step": 300
     },
     {
       "epoch": 0.44411547002220575,
-      "eval_loss": 0.7755329608917236,
-      "eval_runtime": 7.1011,
-      "eval_samples_per_second": 18.025,
-      "eval_steps_per_second": 2.253,
+      "eval_loss": 0.7755764722824097,
+      "eval_runtime": 7.0841,
+      "eval_samples_per_second": 18.069,
+      "eval_steps_per_second": 2.259,
       "step": 300
     },
     {
       "epoch": 0.4515173945225759,
-      "grad_norm": 0.6545192109146697,
+      "grad_norm": 0.6557824164506009,
       "learning_rate": 1.3374227904130724e-05,
-      "loss": 0.7549,
+      "loss": 0.755,
       "step": 305
     },
     {
       "epoch": 0.45891931902294597,
-      "grad_norm": 0.6442815860325689,
+      "grad_norm": 0.6429161176286008,
       "learning_rate": 1.3129521691070108e-05,
-      "loss": 0.7328,
+      "loss": 0.7327,
       "step": 310
     },
     {
       "epoch": 0.46632124352331605,
-      "grad_norm": 0.7168191522252756,
+      "grad_norm": 0.7172336735771104,
       "learning_rate": 1.2882719840797473e-05,
-      "loss": 0.7513,
+      "loss": 0.7514,
       "step": 315
     },
     {
       "epoch": 0.47372316802368614,
-      "grad_norm": 0.6859678587218556,
+      "grad_norm": 0.6888969569785534,
       "learning_rate": 1.2633987620466229e-05,
-      "loss": 0.7354,
+      "loss": 0.7353,
       "step": 320
     },
     {
       "epoch": 0.4811250925240563,
-      "grad_norm": 0.6366532146199427,
+      "grad_norm": 0.6341647872044919,
       "learning_rate": 1.2383491589873122e-05,
-      "loss": 0.7407,
+      "loss": 0.7406,
       "step": 325
     },
     {
       "epoch": 0.48852701702442636,
-      "grad_norm": 0.6079119402423303,
+      "grad_norm": 0.6076464907196722,
       "learning_rate": 1.213139948992394e-05,
-      "loss": 0.7499,
+      "loss": 0.7498,
       "step": 330
     },
     {
       "epoch": 0.49592894152479644,
-      "grad_norm": 0.7008221821337609,
+      "grad_norm": 0.7070131265743488,
       "learning_rate": 1.187788013030837e-05,
       "loss": 0.7469,
       "step": 335
     },
     {
       "epoch": 0.5033308660251665,
-      "grad_norm": 0.6152939544821733,
+      "grad_norm": 0.6137043611077291,
       "learning_rate": 1.1623103276459086e-05,
-      "loss": 0.7506,
+      "loss": 0.7507,
       "step": 340
     },
     {
       "epoch": 0.5107327905255367,
-      "grad_norm": 0.644010855379666,
+      "grad_norm": 0.6456952973092422,
       "learning_rate": 1.1367239535870913e-05,
       "loss": 0.7426,
       "step": 345
     },
     {
       "epoch": 0.5181347150259067,
-      "grad_norm": 0.6874332341347237,
+      "grad_norm": 0.6874472223982817,
       "learning_rate": 1.1110460243856051e-05,
       "loss": 0.7302,
       "step": 350
     },
     {
       "epoch": 0.5255366395262768,
-      "grad_norm": 0.6588126352184485,
+      "grad_norm": 0.6585897080191178,
       "learning_rate": 1.085293734881197e-05,
-      "loss": 0.7468,
+      "loss": 0.7469,
       "step": 355
     },
     {
       "epoch": 0.532938564026647,
-      "grad_norm": 0.6021377543856489,
+      "grad_norm": 0.597859239299925,
       "learning_rate": 1.0594843297078736e-05,
       "loss": 0.766,
       "step": 360
     },
     {
       "epoch": 0.540340488527017,
-      "grad_norm": 0.6899936187904021,
+      "grad_norm": 0.6867943839454653,
       "learning_rate": 1.0336350917462925e-05,
-      "loss": 0.7559,
+      "loss": 0.7558,
       "step": 365
     },
     {
       "epoch": 0.5477424130273871,
-      "grad_norm": 0.6023466178513429,
+      "grad_norm": 0.6059202493597124,
       "learning_rate": 1.0077633305505402e-05,
-      "loss": 0.7433,
+      "loss": 0.7434,
       "step": 370
     },
     {
       "epoch": 0.5551443375277573,
-      "grad_norm": 0.63503275535569,
+      "grad_norm": 0.6312185196007014,
       "learning_rate": 9.818863707570476e-06,
-      "loss": 0.7607,
+      "loss": 0.7608,
       "step": 375
     },
     {
       "epoch": 0.5625462620281273,
-      "grad_norm": 0.6588845716695092,
+      "grad_norm": 0.6583669975631296,
       "learning_rate": 9.560215404834094e-06,
-      "loss": 0.7515,
+      "loss": 0.7516,
       "step": 380
     },
     {
       "epoch": 0.5699481865284974,
-      "grad_norm": 0.6339041649100026,
+      "grad_norm": 0.6337654329864656,
       "learning_rate": 9.30186159724869e-06,
       "loss": 0.7145,
       "step": 385
     },
     {
       "epoch": 0.5773501110288675,
-      "grad_norm": 0.6045706355205543,
+      "grad_norm": 0.6063835380133286,
       "learning_rate": 9.043975287562443e-06,
-      "loss": 0.7471,
+      "loss": 0.747,
       "step": 390
     },
     {
       "epoch": 0.5847520355292376,
-      "grad_norm": 0.6764548092814746,
+      "grad_norm": 0.6801968160814837,
       "learning_rate": 8.786729165470584e-06,
-      "loss": 0.7253,
+      "loss": 0.7254,
       "step": 395
     },
     {
       "epoch": 0.5921539600296077,
-      "grad_norm": 0.6885907129135019,
+      "grad_norm": 0.6884878033137862,
       "learning_rate": 8.530295491976338e-06,
-      "loss": 0.7307,
+      "loss": 0.7308,
       "step": 400
     },
     {
       "epoch": 0.5921539600296077,
-      "eval_loss": 0.7637214660644531,
-      "eval_runtime": 7.0915,
-      "eval_samples_per_second": 18.05,
-      "eval_steps_per_second": 2.256,
+      "eval_loss": 0.7637947797775269,
+      "eval_runtime": 7.0881,
+      "eval_samples_per_second": 18.059,
+      "eval_steps_per_second": 2.257,
       "step": 400
     },
     {
       "epoch": 0.5995558845299778,
-      "grad_norm": 0.5887801704512137,
+      "grad_norm": 0.5905316945269669,
       "learning_rate": 8.274845984038916e-06,
       "loss": 0.7173,
       "step": 405
     },
     {
       "epoch": 0.6069578090303479,
-      "grad_norm": 0.656410885558084,
+      "grad_norm": 0.652102126972111,
       "learning_rate": 8.020551699585843e-06,
-      "loss": 0.7468,
+      "loss": 0.7469,
       "step": 410
     },
     {
       "epoch": 0.6143597335307179,
-      "grad_norm": 0.6065347988971304,
+      "grad_norm": 0.6046273530879036,
       "learning_rate": 7.76758292296659e-06,
-      "loss": 0.7264,
+      "loss": 0.7265,
       "step": 415
     },
     {
       "epoch": 0.6217616580310881,
-      "grad_norm": 0.649848842204615,
+      "grad_norm": 0.6506612501289583,
       "learning_rate": 7.5161090509242005e-06,
-      "loss": 0.7418,
+      "loss": 0.742,
       "step": 420
     },
     {
       "epoch": 0.6291635825314582,
-      "grad_norm": 0.6488869069591486,
+      "grad_norm": 0.6445503530456065,
       "learning_rate": 7.2662984791613186e-06,
       "loss": 0.7345,
       "step": 425
     },
     {
       "epoch": 0.6365655070318282,
-      "grad_norm": 0.6516433245100297,
+      "grad_norm": 0.6524566274539656,
       "learning_rate": 7.01831848957653e-06,
       "loss": 0.749,
       "step": 430
     },
     {
       "epoch": 0.6439674315321984,
-      "grad_norm": 0.6026797161760676,
+      "grad_norm": 0.6006427158853699,
       "learning_rate": 6.772335138246548e-06,
       "loss": 0.7469,
       "step": 435
     },
     {
       "epoch": 0.6513693560325685,
-      "grad_norm": 0.6277455528232541,
+      "grad_norm": 0.6253883357593474,
       "learning_rate": 6.528513144229256e-06,
       "loss": 0.7427,
       "step": 440
     },
     {
       "epoch": 0.6587712805329385,
-      "grad_norm": 0.6141125946978384,
+      "grad_norm": 0.6149011526684693,
       "learning_rate": 6.287015779262064e-06,
-      "loss": 0.749,
+      "loss": 0.7489,
       "step": 445
     },
     {
       "epoch": 0.6661732050333087,
-      "grad_norm": 0.6515237468882386,
+      "grad_norm": 0.655303715198759,
       "learning_rate": 6.048004758429451e-06,
       "loss": 0.7275,
       "step": 450
     },
     {
       "epoch": 0.6735751295336787,
-      "grad_norm": 0.605340859153651,
+      "grad_norm": 0.60174347779124,
       "learning_rate": 5.811640131872867e-06,
-      "loss": 0.7499,
+      "loss": 0.7498,
       "step": 455
     },
     {
       "epoch": 0.6809770540340488,
-      "grad_norm": 0.6062934511176497,
+      "grad_norm": 0.6063626940584474,
       "learning_rate": 5.578080177615575e-06,
-      "loss": 0.7202,
+      "loss": 0.7201,
       "step": 460
     },
     {
       "epoch": 0.688378978534419,
-      "grad_norm": 0.6215783046194415,
+      "grad_norm": 0.6218180992736176,
       "learning_rate": 5.347481295574141e-06,
       "loss": 0.7172,
       "step": 465
     },
     {
       "epoch": 0.695780903034789,
-      "grad_norm": 0.6098446448612718,
+      "grad_norm": 0.6109254076692403,
       "learning_rate": 5.119997902827584e-06,
       "loss": 0.7287,
       "step": 470
     },
     {
       "epoch": 0.7031828275351591,
-      "grad_norm": 0.6071896928919045,
+      "grad_norm": 0.6117825683298314,
       "learning_rate": 4.8957823302142916e-06,
-      "loss": 0.7354,
+      "loss": 0.7355,
       "step": 475
     },
     {
       "epoch": 0.7105847520355293,
-      "grad_norm": 0.5709732809231792,
+      "grad_norm": 0.5644874141920769,
       "learning_rate": 4.674984720325961e-06,
       "loss": 0.7208,
       "step": 480
     },
     {
       "epoch": 0.7179866765358993,
-      "grad_norm": 0.5856635778208634,
+      "grad_norm": 0.5866643065055667,
       "learning_rate": 4.457752926966888e-06,
-      "loss": 0.715,
+      "loss": 0.7151,
       "step": 485
     },
     {
       "epoch": 0.7253886010362695,
-      "grad_norm": 0.5807525667797813,
+      "grad_norm": 0.5815137410742766,
       "learning_rate": 4.244232416145839e-06,
-      "loss": 0.7338,
+      "loss": 0.7337,
       "step": 490
     },
     {
       "epoch": 0.7327905255366395,
-      "grad_norm": 0.5899767021260597,
+      "grad_norm": 0.5893669437648787,
       "learning_rate": 4.0345661686669745e-06,
       "loss": 0.7269,
       "step": 495
     },
     {
       "epoch": 0.7401924500370096,
-      "grad_norm": 0.620833320044676,
+      "grad_norm": 0.6231119397765088,
       "learning_rate": 3.828894584384867e-06,
       "loss": 0.7355,
       "step": 500
     },
     {
       "epoch": 0.7401924500370096,
-      "eval_loss": 0.7563157081604004,
-      "eval_runtime": 7.1093,
-      "eval_samples_per_second": 18.005,
-      "eval_steps_per_second": 2.251,
+      "eval_loss": 0.7561960220336914,
+      "eval_runtime": 7.0784,
+      "eval_samples_per_second": 18.083,
+      "eval_steps_per_second": 2.26,
       "step": 500
     },
     {
       "epoch": 0.7475943745373798,
-      "grad_norm": 0.5584124590432081,
+      "grad_norm": 0.5547460159002005,
       "learning_rate": 3.62735538818787e-06,
       "loss": 0.7196,
       "step": 505
     },
     {
       "epoch": 0.7549962990377498,
-      "grad_norm": 0.6290566212809122,
+      "grad_norm": 0.6311442463920164,
       "learning_rate": 3.4300835377726904e-06,
-      "loss": 0.7233,
+      "loss": 0.7234,
       "step": 510
     },
     {
       "epoch": 0.7623982235381199,
-      "grad_norm": 0.6147581914210406,
+      "grad_norm": 0.6161479513086646,
       "learning_rate": 3.2372111332720045e-06,
       "loss": 0.7588,
       "step": 515
     },
     {
       "epoch": 0.7698001480384901,
-      "grad_norm": 0.609969440330252,
+      "grad_norm": 0.6108773239475987,
       "learning_rate": 3.048867328795588e-06,
-      "loss": 0.7156,
+      "loss": 0.7157,
       "step": 520
     },
     {
       "epoch": 0.7772020725388601,
-      "grad_norm": 0.5943507423692393,
+      "grad_norm": 0.602925688413523,
       "learning_rate": 2.865178245944218e-06,
       "loss": 0.7145,
       "step": 525
     },
     {
       "epoch": 0.7846039970392302,
-      "grad_norm": 0.5735693750532326,
+      "grad_norm": 0.5701021420362337,
       "learning_rate": 2.686266889354211e-06,
-      "loss": 0.7374,
+      "loss": 0.7375,
       "step": 530
     },
     {
       "epoch": 0.7920059215396003,
-      "grad_norm": 0.5879362568428208,
+      "grad_norm": 0.5887011145957908,
       "learning_rate": 2.5122530643292274e-06,
       "loss": 0.7429,
       "step": 535
     },
     {
       "epoch": 0.7994078460399704,
-      "grad_norm": 0.6291447773815294,
+      "grad_norm": 0.6312752747025913,
       "learning_rate": 2.3432532966144526e-06,
       "loss": 0.7323,
       "step": 540
     },
     {
       "epoch": 0.8068097705403405,
-      "grad_norm": 0.5851868491818125,
+      "grad_norm": 0.5823788876767197,
       "learning_rate": 2.1793807543668857e-06,
       "loss": 0.734,
       "step": 545
     },
     {
       "epoch": 0.8142116950407106,
-      "grad_norm": 0.5514287987527494,
+      "grad_norm": 0.5500197273786551,
       "learning_rate": 2.0207451723739633e-06,
       "loss": 0.7258,
       "step": 550
     },
     {
       "epoch": 0.8216136195410807,
-      "grad_norm": 0.5597971075762194,
+      "grad_norm": 0.5605928225621666,
       "learning_rate": 1.8674527785713247e-06,
-      "loss": 0.7324,
+      "loss": 0.7323,
       "step": 555
     },
     {
       "epoch": 0.8290155440414507,
-      "grad_norm": 0.6153642015397675,
+      "grad_norm": 0.6153548429599589,
       "learning_rate": 1.7196062229088606e-06,
-      "loss": 0.6998,
+      "loss": 0.6996,
       "step": 560
     },
     {
       "epoch": 0.8364174685418209,
-      "grad_norm": 0.6286712970030335,
+      "grad_norm": 0.6241569667891514,
       "learning_rate": 1.577304508612717e-06,
       "loss": 0.7298,
       "step": 565
     },
     {
       "epoch": 0.843819393042191,
-      "grad_norm": 0.5202130313277985,
+      "grad_norm": 0.5193118067897788,
       "learning_rate": 1.4406429258892762e-06,
       "loss": 0.7503,
       "step": 570
     },
     {
       "epoch": 0.851221317542561,
-      "grad_norm": 0.6140488221729622,
+      "grad_norm": 0.6129769104488084,
       "learning_rate": 1.3097129881154936e-06,
       "loss": 0.72,
       "step": 575
     },
     {
       "epoch": 0.8586232420429312,
-      "grad_norm": 0.5851314946850364,
+      "grad_norm": 0.5835803211279808,
       "learning_rate": 1.1846023705583442e-06,
       "loss": 0.7164,
       "step": 580
     },
     {
       "epoch": 0.8660251665433013,
-      "grad_norm": 0.5793114473346295,
+      "grad_norm": 0.5805836101347949,
       "learning_rate": 1.065394851664394e-06,
-      "loss": 0.7344,
+      "loss": 0.7345,
       "step": 585
     },
     {
       "epoch": 0.8734270910436713,
-      "grad_norm": 0.5480172795627596,
+      "grad_norm": 0.5488458370278894,
       "learning_rate": 9.521702569588199e-07,
       "loss": 0.7537,
       "step": 590
     },
     {
       "epoch": 0.8808290155440415,
-      "grad_norm": 0.5747689076626602,
+      "grad_norm": 0.574169045086665,
       "learning_rate": 8.450044055914497e-07,
       "loss": 0.7221,
       "step": 595
     },
     {
       "epoch": 0.8882309400444115,
-      "grad_norm": 0.5614205673986494,
+      "grad_norm": 0.5603143384229617,
       "learning_rate": 7.439690595656013e-07,
       "loss": 0.7446,
       "step": 600
     },
     {
       "epoch": 0.8882309400444115,
-      "eval_loss": 0.7532215118408203,
-      "eval_runtime": 7.0965,
-      "eval_samples_per_second": 18.037,
-      "eval_steps_per_second": 2.255,
+      "eval_loss": 0.753212571144104,
+      "eval_runtime": 7.0895,
+      "eval_samples_per_second": 18.055,
+      "eval_steps_per_second": 2.257,
       "step": 600
     },
     {
       "epoch": 0.8956328645447816,
-      "grad_norm": 0.6279531134162607,
+      "grad_norm": 0.6329369985344253,
       "learning_rate": 6.491318756837417e-07,
-      "loss": 0.7298,
+      "loss": 0.7299,
       "step": 605
     },
     {
       "epoch": 0.9030347890451518,
-      "grad_norm": 0.5118890379128042,
+      "grad_norm": 0.512657762146579,
       "learning_rate": 5.605563602421149e-07,
       "loss": 0.7059,
       "step": 610
     },
     {
       "epoch": 0.9104367135455218,
-      "grad_norm": 0.5548165195544706,
+      "grad_norm": 0.555802795710424,
       "learning_rate": 4.783018265047179e-07,
-      "loss": 0.7557,
+      "loss": 0.7556,
       "step": 615
     },
     {
       "epoch": 0.9178386380458919,
-      "grad_norm": 0.5841126339690919,
+      "grad_norm": 0.5850038916370969,
       "learning_rate": 4.024233549850509e-07,
       "loss": 0.7437,
       "step": 620
     },
     {
       "epoch": 0.9252405625462621,
-      "grad_norm": 0.5500866498461655,
+      "grad_norm": 0.5486195860835535,
       "learning_rate": 3.329717565622825e-07,
-      "loss": 0.7405,
+      "loss": 0.7404,
       "step": 625
     },
     {
       "epoch": 0.9326424870466321,
-      "grad_norm": 0.5499805117486685,
+      "grad_norm": 0.5514548625215937,
       "learning_rate": 2.6999353845651113e-07,
-      "loss": 0.724,
+      "loss": 0.7241,
       "step": 630
     },
     {
       "epoch": 0.9400444115470022,
-      "grad_norm": 0.5850424082936563,
+      "grad_norm": 0.5878396775004657,
       "learning_rate": 2.1353087308590314e-07,
-      "loss": 0.7391,
+      "loss": 0.739,
       "step": 635
     },
     {
       "epoch": 0.9474463360473723,
-      "grad_norm": 0.5584988807196664,
+      "grad_norm": 0.5583557984778206,
       "learning_rate": 1.6362156982656085e-07,
-      "loss": 0.7292,
+      "loss": 0.7291,
       "step": 640
     },
     {
       "epoch": 0.9548482605477424,
-      "grad_norm": 0.5230733641745596,
+      "grad_norm": 0.5235524601826119,
       "learning_rate": 1.2029904969404482e-07,
       "loss": 0.7127,
       "step": 645
     },
     {
       "epoch": 0.9622501850481125,
-      "grad_norm": 0.6003490232316495,
+      "grad_norm": 0.6023587868318137,
       "learning_rate": 8.359232296349163e-08,
-      "loss": 0.7164,
+      "loss": 0.7165,
       "step": 650
     },
     {
       "epoch": 0.9696521095484826,
-      "grad_norm": 0.5697422776830529,
+      "grad_norm": 0.5670622600030921,
       "learning_rate": 5.3525969743324356e-08,
       "loss": 0.7321,
       "step": 655
     },
     {
       "epoch": 0.9770540340488527,
-      "grad_norm": 0.5462820150416294,
+      "grad_norm": 0.5473289868827732,
       "learning_rate": 3.012012351554017e-08,
       "loss": 0.7065,
       "step": 660
     },
     {
       "epoch": 0.9844559585492227,
-      "grad_norm": 0.5799454156121296,
+      "grad_norm": 0.5803068194369324,
       "learning_rate": 1.3390457653639221e-08,
-      "loss": 0.7353,
+      "loss": 0.7354,
       "step": 665
     },
     {
       "epoch": 0.9918578830495929,
-      "grad_norm": 0.6443892875748711,
+      "grad_norm": 0.641603033929549,
       "learning_rate": 3.3481749271768726e-09,
-      "loss": 0.7462,
+      "loss": 0.7463,
       "step": 670
     },
     {
       "epoch": 0.999259807549963,
-      "grad_norm": 0.5744857640888457,
+      "grad_norm": 0.5695984279665459,
       "learning_rate": 0.0,
       "loss": 0.7159,
       "step": 675
@@ -1005,10 +1005,10 @@
       "epoch": 0.999259807549963,
       "step": 675,
       "total_flos": 76888336760832.0,
-      "train_loss": 0.7676037124351219,
-      "train_runtime": 4810.4535,
-      "train_samples_per_second": 4.492,
-      "train_steps_per_second": 0.14
+      "train_loss": 0.7676166036393908,
+      "train_runtime": 4800.2526,
+      "train_samples_per_second": 4.502,
+      "train_steps_per_second": 0.141
     }
   ],
   "logging_steps": 5,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8bd1100246b04db3818cf08f698c257d9b18f8121b2073ed2af3fe6b1be64523
+oid sha256:8e1cd235b31e53cf1a1187712302fdd0362be459682d5e0f36cf27c3ec15f60a
 size 7352