Upload folder using huggingface_hub

adapter_config.json

@@ -29,10 +29,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q",
-    "k",
     "o",
-    "v"
+    "v",
+    "q",
+    "k"
   ],
   "target_parameters": null,
   "task_type": "SEQ_2_SEQ_LM",

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:640ecbfaac397ec3bfdcabb585e02a7c9d74cdca737214766274fbe3c6574c9f
+oid sha256:c76ca4b647b9b7722e05b59f65b4567da049cb2207ab0f9ad06a8ef472ed905d
 size 7117824

checkpoint-9213/adapter_config.json

@@ -29,10 +29,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q",
-    "k",
     "o",
-    "v"
+    "v",
+    "q",
+    "k"
   ],
   "target_parameters": null,
   "task_type": "SEQ_2_SEQ_LM",

checkpoint-9213/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:640ecbfaac397ec3bfdcabb585e02a7c9d74cdca737214766274fbe3c6574c9f
+oid sha256:c76ca4b647b9b7722e05b59f65b4567da049cb2207ab0f9ad06a8ef472ed905d
 size 7117824

checkpoint-9213/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89b9a84a299ec73f6f694fa211f58c1a71f1ed383ff10ddd9bdde45563b5b823
-size 3912788
+oid sha256:6bfb06864019c7baddd83f1f745b117d96c7d3ee900744262f3e09778553d71e
+size 14397434

checkpoint-9213/trainer_state.json

@@ -11,1329 +11,1329 @@
   "log_history": [
     {
       "epoch": 0.0162813415825464,
-      "grad_norm": 1.3473832607269287,
+      "grad_norm": 1.3684147596359253,
       "learning_rate": 4.973407142081841e-05,
-      "loss": 4.392664489746093,
+      "loss": 4.405337219238281,
       "step": 50
     },
     {
       "epoch": 0.0325626831650928,
-      "grad_norm": 1.8160868883132935,
+      "grad_norm": 2.8155977725982666,
       "learning_rate": 4.946271572777597e-05,
-      "loss": 3.211649169921875,
+      "loss": 3.2570425415039064,
       "step": 100
     },
     {
       "epoch": 0.04884402474763921,
-      "grad_norm": 1.1382832527160645,
+      "grad_norm": 1.134498119354248,
       "learning_rate": 4.919136003473353e-05,
-      "loss": 2.1162736511230467,
+      "loss": 2.13068603515625,
       "step": 150
     },
     {
       "epoch": 0.0651253663301856,
-      "grad_norm": 1.3651690483093262,
+      "grad_norm": 1.3554809093475342,
       "learning_rate": 4.892000434169109e-05,
-      "loss": 1.873338623046875,
+      "loss": 1.8829190063476562,
       "step": 200
     },
     {
       "epoch": 0.08140670791273201,
-      "grad_norm": 1.7120460271835327,
+      "grad_norm": 1.3557366132736206,
       "learning_rate": 4.8648648648648654e-05,
-      "loss": 1.8091192626953125,
+      "loss": 1.819012908935547,
       "step": 250
     },
     {
       "epoch": 0.09768804949527841,
-      "grad_norm": 1.4614651203155518,
+      "grad_norm": 1.4770574569702148,
       "learning_rate": 4.837729295560621e-05,
-      "loss": 1.7023934936523437,
+      "loss": 1.7141534423828124,
       "step": 300
     },
     {
       "epoch": 0.11396939107782482,
-      "grad_norm": 1.4845161437988281,
+      "grad_norm": 1.4904377460479736,
       "learning_rate": 4.810593726256377e-05,
-      "loss": 1.6588560485839843,
+      "loss": 1.6724388122558593,
       "step": 350
     },
     {
       "epoch": 0.1302507326603712,
-      "grad_norm": 1.4929733276367188,
+      "grad_norm": 1.4447667598724365,
       "learning_rate": 4.783458156952133e-05,
-      "loss": 1.5784068298339844,
+      "loss": 1.5954844665527343,
       "step": 400
     },
     {
       "epoch": 0.14653207424291761,
-      "grad_norm": 1.3720242977142334,
+      "grad_norm": 1.3709800243377686,
       "learning_rate": 4.756322587647889e-05,
-      "loss": 1.5646693420410156,
+      "loss": 1.585284423828125,
       "step": 450
     },
     {
       "epoch": 0.16281341582546402,
-      "grad_norm": 1.968331217765808,
+      "grad_norm": 1.9350112676620483,
       "learning_rate": 4.729187018343645e-05,
-      "loss": 1.424574737548828,
+      "loss": 1.4481756591796875,
       "step": 500
     },
     {
       "epoch": 0.17909475740801042,
-      "grad_norm": 1.8840748071670532,
+      "grad_norm": 1.8580759763717651,
       "learning_rate": 4.7020514490394014e-05,
-      "loss": 1.485799560546875,
+      "loss": 1.5084706115722657,
       "step": 550
     },
     {
       "epoch": 0.19537609899055683,
-      "grad_norm": 2.7317214012145996,
+      "grad_norm": 2.5280864238739014,
       "learning_rate": 4.674915879735157e-05,
-      "loss": 1.3331263732910157,
+      "loss": 1.3583644104003907,
       "step": 600
     },
     {
       "epoch": 0.21165744057310323,
-      "grad_norm": 1.773260474205017,
+      "grad_norm": 1.7610465288162231,
       "learning_rate": 4.647780310430913e-05,
-      "loss": 1.3437060546875,
+      "loss": 1.3752433776855468,
       "step": 650
     },
     {
       "epoch": 0.22793878215564964,
-      "grad_norm": 1.9822585582733154,
+      "grad_norm": 1.9367973804473877,
       "learning_rate": 4.620644741126669e-05,
-      "loss": 1.3029965209960936,
+      "loss": 1.3308279418945312,
       "step": 700
     },
     {
       "epoch": 0.24422012373819602,
-      "grad_norm": 1.9712088108062744,
+      "grad_norm": 1.8170645236968994,
       "learning_rate": 4.593509171822425e-05,
-      "loss": 1.1877630615234376,
+      "loss": 1.2124520111083985,
       "step": 750
     },
     {
       "epoch": 0.2605014653207424,
-      "grad_norm": 2.6370294094085693,
+      "grad_norm": 2.4546971321105957,
       "learning_rate": 4.566373602518181e-05,
-      "loss": 1.1992277526855468,
+      "loss": 1.2327165222167968,
       "step": 800
     },
     {
       "epoch": 0.2767828069032888,
-      "grad_norm": 2.512601852416992,
+      "grad_norm": 2.476984977722168,
       "learning_rate": 4.5392380332139374e-05,
-      "loss": 1.2441465759277344,
+      "loss": 1.2717761993408203,
       "step": 850
     },
     {
       "epoch": 0.29306414848583523,
-      "grad_norm": 2.6893482208251953,
+      "grad_norm": 2.4676401615142822,
       "learning_rate": 4.512102463909693e-05,
-      "loss": 1.1057972717285156,
+      "loss": 1.1377059173583985,
       "step": 900
     },
     {
       "epoch": 0.30934549006838163,
-      "grad_norm": 2.995448589324951,
+      "grad_norm": 3.8174469470977783,
       "learning_rate": 4.484966894605449e-05,
-      "loss": 1.1611495208740235,
+      "loss": 1.195997314453125,
       "step": 950
     },
     {
       "epoch": 0.32562683165092804,
-      "grad_norm": 2.9329564571380615,
+      "grad_norm": 2.879594564437866,
       "learning_rate": 4.457831325301205e-05,
-      "loss": 1.1252006530761718,
+      "loss": 1.160465316772461,
       "step": 1000
     },
     {
       "epoch": 0.34190817323347444,
-      "grad_norm": 2.749084234237671,
+      "grad_norm": 2.7231075763702393,
       "learning_rate": 4.430695755996961e-05,
-      "loss": 1.096492691040039,
+      "loss": 1.1307568359375,
       "step": 1050
     },
     {
       "epoch": 0.35818951481602085,
-      "grad_norm": 2.9316182136535645,
+      "grad_norm": 2.86122989654541,
       "learning_rate": 4.403560186692717e-05,
-      "loss": 1.0722020721435548,
+      "loss": 1.1008319854736328,
       "step": 1100
     },
     {
       "epoch": 0.37447085639856725,
-      "grad_norm": 2.6080806255340576,
+      "grad_norm": 5.22833251953125,
       "learning_rate": 4.3764246173884733e-05,
-      "loss": 1.0462689971923829,
+      "loss": 1.075364456176758,
       "step": 1150
     },
     {
       "epoch": 0.39075219798111366,
-      "grad_norm": 2.7301759719848633,
+      "grad_norm": 2.75107479095459,
       "learning_rate": 4.349289048084229e-05,
-      "loss": 0.9948210144042968,
+      "loss": 1.02243408203125,
       "step": 1200
     },
     {
       "epoch": 0.40703353956366006,
-      "grad_norm": 3.1961374282836914,
+      "grad_norm": 5.390212535858154,
       "learning_rate": 4.322153478779985e-05,
-      "loss": 0.9516075897216797,
+      "loss": 0.9791134643554688,
       "step": 1250
     },
     {
       "epoch": 0.42331488114620647,
-      "grad_norm": 3.144866704940796,
+      "grad_norm": 2.980733871459961,
       "learning_rate": 4.295017909475741e-05,
-      "loss": 0.9630168914794922,
+      "loss": 0.9889099884033203,
       "step": 1300
     },
     {
       "epoch": 0.4395962227287529,
-      "grad_norm": 2.025820255279541,
+      "grad_norm": 2.0358238220214844,
       "learning_rate": 4.267882340171497e-05,
-      "loss": 0.9979273223876953,
+      "loss": 1.0286465454101563,
       "step": 1350
     },
     {
       "epoch": 0.4558775643112993,
-      "grad_norm": 3.443812847137451,
+      "grad_norm": 3.465447187423706,
       "learning_rate": 4.2407467708672534e-05,
-      "loss": 0.8928343963623047,
+      "loss": 0.921280517578125,
       "step": 1400
     },
     {
       "epoch": 0.4721589058938456,
-      "grad_norm": 3.585447311401367,
+      "grad_norm": 3.2969017028808594,
       "learning_rate": 4.213611201563009e-05,
-      "loss": 0.9557633972167969,
+      "loss": 0.9863865661621094,
       "step": 1450
     },
     {
       "epoch": 0.48844024747639203,
-      "grad_norm": 2.1626651287078857,
+      "grad_norm": 2.22483229637146,
       "learning_rate": 4.186475632258765e-05,
-      "loss": 0.9039353179931641,
+      "loss": 0.9354537963867188,
       "step": 1500
     },
     {
       "epoch": 0.5047215890589385,
-      "grad_norm": 2.562255620956421,
+      "grad_norm": 2.548597812652588,
       "learning_rate": 4.159340062954521e-05,
-      "loss": 0.8657318115234375,
+      "loss": 0.9002230834960937,
       "step": 1550
     },
     {
       "epoch": 0.5210029306414848,
-      "grad_norm": 2.244072198867798,
+      "grad_norm": 2.0713655948638916,
       "learning_rate": 4.132204493650277e-05,
-      "loss": 0.9271598815917969,
+      "loss": 0.9603328704833984,
       "step": 1600
     },
     {
       "epoch": 0.5372842722240313,
-      "grad_norm": 2.8100831508636475,
+      "grad_norm": 2.6993308067321777,
       "learning_rate": 4.105068924346033e-05,
-      "loss": 0.8872251892089844,
+      "loss": 0.91706787109375,
       "step": 1650
     },
     {
       "epoch": 0.5535656138065776,
-      "grad_norm": 2.8497555255889893,
+      "grad_norm": 2.801042079925537,
       "learning_rate": 4.0779333550417894e-05,
-      "loss": 0.8983834838867187,
+      "loss": 0.9315050506591797,
       "step": 1700
     },
     {
       "epoch": 0.5698469553891241,
-      "grad_norm": 2.229484796524048,
+      "grad_norm": 2.2465991973876953,
       "learning_rate": 4.050797785737545e-05,
-      "loss": 0.9012404632568359,
+      "loss": 0.9338973236083984,
       "step": 1750
     },
     {
       "epoch": 0.5861282969716705,
-      "grad_norm": 2.551394462585449,
+      "grad_norm": 2.300447702407837,
       "learning_rate": 4.023662216433301e-05,
-      "loss": 0.8586711883544922,
+      "loss": 0.8962403106689453,
       "step": 1800
     },
     {
       "epoch": 0.6024096385542169,
-      "grad_norm": 3.706598997116089,
+      "grad_norm": 3.7357118129730225,
       "learning_rate": 3.996526647129057e-05,
-      "loss": 0.8685987091064453,
+      "loss": 0.8952316284179688,
       "step": 1850
     },
     {
       "epoch": 0.6186909801367633,
-      "grad_norm": 2.964557409286499,
+      "grad_norm": 2.987694263458252,
       "learning_rate": 3.969391077824813e-05,
-      "loss": 0.8568785095214844,
+      "loss": 0.8913738250732421,
       "step": 1900
     },
     {
       "epoch": 0.6349723217193096,
-      "grad_norm": 2.809697151184082,
+      "grad_norm": 2.678509473800659,
       "learning_rate": 3.942255508520569e-05,
-      "loss": 0.7817304992675781,
+      "loss": 0.8106181335449218,
       "step": 1950
     },
     {
       "epoch": 0.6512536633018561,
-      "grad_norm": 2.845287799835205,
+      "grad_norm": 2.694267511367798,
       "learning_rate": 3.9151199392163254e-05,
-      "loss": 0.8261214447021484,
+      "loss": 0.8525662994384766,
       "step": 2000
     },
     {
       "epoch": 0.6675350048844024,
-      "grad_norm": 2.3624467849731445,
+      "grad_norm": 2.4209225177764893,
       "learning_rate": 3.887984369912081e-05,
-      "loss": 0.7863189697265625,
+      "loss": 0.8165419769287109,
       "step": 2050
     },
     {
       "epoch": 0.6838163464669489,
-      "grad_norm": 3.5507824420928955,
+      "grad_norm": 3.41129994392395,
       "learning_rate": 3.860848800607837e-05,
-      "loss": 0.7708241271972657,
+      "loss": 0.7995121765136719,
       "step": 2100
     },
     {
       "epoch": 0.7000976880494952,
-      "grad_norm": 2.853994131088257,
+      "grad_norm": 2.8171026706695557,
       "learning_rate": 3.833713231303593e-05,
-      "loss": 0.7777981567382812,
+      "loss": 0.8041473388671875,
       "step": 2150
     },
     {
       "epoch": 0.7163790296320417,
-      "grad_norm": 3.099780321121216,
+      "grad_norm": 2.9240710735321045,
       "learning_rate": 3.806577661999349e-05,
-      "loss": 0.9427651214599609,
+      "loss": 0.9764547729492188,
       "step": 2200
     },
     {
       "epoch": 0.732660371214588,
-      "grad_norm": 2.143392562866211,
+      "grad_norm": 2.051103353500366,
       "learning_rate": 3.7794420926951055e-05,
-      "loss": 0.8011217498779297,
+      "loss": 0.8336094665527344,
       "step": 2250
     },
     {
       "epoch": 0.7489417127971345,
-      "grad_norm": 3.4226908683776855,
+      "grad_norm": 3.2116966247558594,
       "learning_rate": 3.7523065233908613e-05,
-      "loss": 0.7524455261230468,
+      "loss": 0.7777658081054688,
       "step": 2300
     },
     {
       "epoch": 0.7652230543796809,
-      "grad_norm": 3.735076904296875,
+      "grad_norm": 3.8302836418151855,
       "learning_rate": 3.725170954086617e-05,
-      "loss": 0.8192281341552734,
+      "loss": 0.848994140625,
       "step": 2350
     },
     {
       "epoch": 0.7815043959622273,
-      "grad_norm": 2.332482099533081,
+      "grad_norm": 2.0940141677856445,
       "learning_rate": 3.698035384782373e-05,
-      "loss": 0.6744720458984375,
+      "loss": 0.6995362091064453,
       "step": 2400
     },
     {
       "epoch": 0.7977857375447737,
-      "grad_norm": 3.1311633586883545,
+      "grad_norm": 3.155562400817871,
       "learning_rate": 3.670899815478129e-05,
-      "loss": 0.7469883728027343,
+      "loss": 0.7727601623535156,
       "step": 2450
     },
     {
       "epoch": 0.8140670791273201,
-      "grad_norm": 2.872544050216675,
+      "grad_norm": 2.886545419692993,
       "learning_rate": 3.643764246173885e-05,
-      "loss": 0.7605148315429687,
+      "loss": 0.7873483276367188,
       "step": 2500
     },
     {
       "epoch": 0.8303484207098665,
-      "grad_norm": 2.4674670696258545,
+      "grad_norm": 2.5638420581817627,
       "learning_rate": 3.6166286768696414e-05,
-      "loss": 0.7708393096923828,
+      "loss": 0.7987093353271484,
       "step": 2550
     },
     {
       "epoch": 0.8466297622924129,
-      "grad_norm": 2.391380548477173,
+      "grad_norm": 2.3368115425109863,
       "learning_rate": 3.589493107565397e-05,
-      "loss": 0.7390218353271485,
+      "loss": 0.7667920684814453,
       "step": 2600
     },
     {
       "epoch": 0.8629111038749593,
-      "grad_norm": 2.9628641605377197,
+      "grad_norm": 2.953003168106079,
       "learning_rate": 3.562357538261153e-05,
-      "loss": 0.660747299194336,
+      "loss": 0.6812684631347656,
       "step": 2650
     },
     {
       "epoch": 0.8791924454575057,
-      "grad_norm": 3.0942180156707764,
+      "grad_norm": 3.382375955581665,
       "learning_rate": 3.535221968956909e-05,
-      "loss": 0.725975570678711,
+      "loss": 0.7589307403564454,
       "step": 2700
     },
     {
       "epoch": 0.8954737870400521,
-      "grad_norm": 2.7016773223876953,
+      "grad_norm": 2.6374568939208984,
       "learning_rate": 3.508086399652665e-05,
-      "loss": 0.7020815277099609,
+      "loss": 0.7288359069824218,
       "step": 2750
     },
     {
       "epoch": 0.9117551286225986,
-      "grad_norm": 4.184632778167725,
+      "grad_norm": 2.135319709777832,
       "learning_rate": 3.480950830348421e-05,
-      "loss": 0.7197840118408203,
+      "loss": 0.7487136077880859,
       "step": 2800
     },
     {
       "epoch": 0.9280364702051449,
-      "grad_norm": 3.345290422439575,
+      "grad_norm": 3.235957145690918,
       "learning_rate": 3.4538152610441774e-05,
-      "loss": 0.7012833404541016,
+      "loss": 0.7287068176269531,
       "step": 2850
     },
     {
       "epoch": 0.9443178117876913,
-      "grad_norm": 3.3522071838378906,
+      "grad_norm": 3.254149913787842,
       "learning_rate": 3.426679691739933e-05,
-      "loss": 0.7069497680664063,
+      "loss": 0.7420393371582031,
       "step": 2900
     },
     {
       "epoch": 0.9605991533702377,
-      "grad_norm": 2.183042049407959,
+      "grad_norm": 2.4111130237579346,
       "learning_rate": 3.399544122435689e-05,
-      "loss": 0.6970455169677734,
+      "loss": 0.7257329559326172,
       "step": 2950
     },
     {
       "epoch": 0.9768804949527841,
-      "grad_norm": 3.836655616760254,
+      "grad_norm": 3.764932870864868,
       "learning_rate": 3.372408553131445e-05,
-      "loss": 0.6455158996582031,
+      "loss": 0.6783822631835937,
       "step": 3000
     },
     {
       "epoch": 0.9931618365353305,
-      "grad_norm": 2.9731621742248535,
+      "grad_norm": 3.1191906929016113,
       "learning_rate": 3.345272983827201e-05,
-      "loss": 0.7525621795654297,
+      "loss": 0.7785000610351562,
       "step": 3050
     },
     {
       "epoch": 1.0,
-      "eval_bertscore_f1": 0.9678319004297995,
-      "eval_bleu": 0.6471772455613484,
-      "eval_loss": 0.47463923692703247,
-      "eval_meteor": 0.7416422834916577,
-      "eval_rouge1": 0.855324288070084,
-      "eval_rouge2": 0.7532566990075111,
-      "eval_runtime": 127.1335,
-      "eval_samples_per_second": 10.163,
-      "eval_steps_per_second": 1.274,
+      "eval_bertscore_f1": 0.9666035647731817,
+      "eval_bleu": 0.6286042636210436,
+      "eval_loss": 0.4953671097755432,
+      "eval_meteor": 0.727270993632129,
+      "eval_rouge1": 0.8535427677296903,
+      "eval_rouge2": 0.7442740134144676,
+      "eval_runtime": 125.3536,
+      "eval_samples_per_second": 10.307,
+      "eval_steps_per_second": 1.292,
       "step": 3071
     },
     {
       "epoch": 1.009443178117877,
-      "grad_norm": 3.100937604904175,
+      "grad_norm": 3.273244857788086,
       "learning_rate": 3.318137414522957e-05,
-      "loss": 0.5857192230224609,
+      "loss": 0.6082234954833985,
       "step": 3100
     },
     {
       "epoch": 1.0257245197004232,
-      "grad_norm": 3.4073121547698975,
+      "grad_norm": 3.4152684211730957,
       "learning_rate": 3.291001845218713e-05,
-      "loss": 0.5697959518432617,
+      "loss": 0.5953511047363281,
       "step": 3150
     },
     {
       "epoch": 1.0420058612829697,
-      "grad_norm": 3.4600396156311035,
+      "grad_norm": 3.5387468338012695,
       "learning_rate": 3.263866275914469e-05,
-      "loss": 0.6803036499023437,
+      "loss": 0.6991500091552735,
       "step": 3200
     },
     {
       "epoch": 1.0582872028655161,
-      "grad_norm": 4.564633846282959,
+      "grad_norm": 4.7770280838012695,
       "learning_rate": 3.236730706610225e-05,
-      "loss": 0.6410063934326172,
+      "loss": 0.6681033325195312,
       "step": 3250
     },
     {
       "epoch": 1.0745685444480626,
-      "grad_norm": 3.4267613887786865,
+      "grad_norm": 3.492780923843384,
       "learning_rate": 3.209595137305981e-05,
-      "loss": 0.6364714813232422,
+      "loss": 0.6578762054443359,
       "step": 3300
     },
     {
       "epoch": 1.0908498860306088,
-      "grad_norm": 3.4027607440948486,
+      "grad_norm": 3.3742964267730713,
       "learning_rate": 3.182459568001737e-05,
-      "loss": 0.6122965240478515,
+      "loss": 0.6371722030639648,
       "step": 3350
     },
     {
       "epoch": 1.1071312276131553,
-      "grad_norm": 2.7794837951660156,
+      "grad_norm": 2.9794905185699463,
       "learning_rate": 3.155323998697493e-05,
-      "loss": 0.5932202529907227,
+      "loss": 0.6163125228881836,
       "step": 3400
     },
     {
       "epoch": 1.1234125691957018,
-      "grad_norm": 3.5116841793060303,
+      "grad_norm": 4.077469825744629,
       "learning_rate": 3.128188429393249e-05,
-      "loss": 0.6330239868164063,
+      "loss": 0.6614529418945313,
       "step": 3450
     },
     {
       "epoch": 1.1396939107782482,
-      "grad_norm": 2.800938129425049,
+      "grad_norm": 2.614112615585327,
       "learning_rate": 3.1010528600890045e-05,
-      "loss": 0.5926042938232422,
+      "loss": 0.6072891998291016,
       "step": 3500
     },
     {
       "epoch": 1.1559752523607945,
-      "grad_norm": 2.6473214626312256,
+      "grad_norm": 2.661522150039673,
       "learning_rate": 3.0739172907847604e-05,
-      "loss": 0.5683625793457031,
+      "loss": 0.5901923370361328,
       "step": 3550
     },
     {
       "epoch": 1.172256593943341,
-      "grad_norm": 3.2271904945373535,
+      "grad_norm": 3.5029804706573486,
       "learning_rate": 3.0467817214805167e-05,
-      "loss": 0.6833554077148437,
+      "loss": 0.7141423034667969,
       "step": 3600
     },
     {
       "epoch": 1.1885379355258874,
-      "grad_norm": 3.032439947128296,
+      "grad_norm": 2.75795578956604,
       "learning_rate": 3.0196461521762725e-05,
-      "loss": 0.6369483947753907,
+      "loss": 0.6581143951416015,
       "step": 3650
     },
     {
       "epoch": 1.2048192771084336,
-      "grad_norm": 3.3694541454315186,
+      "grad_norm": 3.335984230041504,
       "learning_rate": 2.992510582872029e-05,
-      "loss": 0.6112889480590821,
+      "loss": 0.6345047378540039,
       "step": 3700
     },
     {
       "epoch": 1.22110061869098,
-      "grad_norm": 2.811060905456543,
+      "grad_norm": 2.886397361755371,
       "learning_rate": 2.965375013567785e-05,
-      "loss": 0.5923568344116211,
+      "loss": 0.6183316802978516,
       "step": 3750
     },
     {
       "epoch": 1.2373819602735265,
-      "grad_norm": 3.533724069595337,
+      "grad_norm": 3.5690371990203857,
       "learning_rate": 2.938239444263541e-05,
-      "loss": 0.6127109909057618,
+      "loss": 0.6330132293701172,
       "step": 3800
     },
     {
       "epoch": 1.253663301856073,
-      "grad_norm": 3.5569841861724854,
+      "grad_norm": 2.857058525085449,
       "learning_rate": 2.9111038749592967e-05,
-      "loss": 0.554149169921875,
+      "loss": 0.5762019729614258,
       "step": 3850
     },
     {
       "epoch": 1.2699446434386195,
-      "grad_norm": 3.246440887451172,
+      "grad_norm": 3.4295766353607178,
       "learning_rate": 2.8839683056550526e-05,
-      "loss": 0.6123342514038086,
+      "loss": 0.6346803665161133,
       "step": 3900
     },
     {
       "epoch": 1.2862259850211657,
-      "grad_norm": 3.752815008163452,
+      "grad_norm": 3.578397274017334,
       "learning_rate": 2.8568327363508085e-05,
-      "loss": 0.596214942932129,
+      "loss": 0.6218410491943359,
       "step": 3950
     },
     {
       "epoch": 1.3025073266037122,
-      "grad_norm": 2.5075297355651855,
+      "grad_norm": 4.432509422302246,
       "learning_rate": 2.829697167046565e-05,
-      "loss": 0.6021640777587891,
+      "loss": 0.6320372009277344,
       "step": 4000
     },
     {
       "epoch": 1.3187886681862586,
-      "grad_norm": 2.307556629180908,
+      "grad_norm": 2.806671142578125,
       "learning_rate": 2.802561597742321e-05,
-      "loss": 0.628720359802246,
+      "loss": 0.652824478149414,
       "step": 4050
     },
     {
       "epoch": 1.3350700097688049,
-      "grad_norm": 3.7279775142669678,
+      "grad_norm": 3.3935513496398926,
       "learning_rate": 2.775426028438077e-05,
-      "loss": 0.6525094604492188,
+      "loss": 0.6719865417480468,
       "step": 4100
     },
     {
       "epoch": 1.3513513513513513,
-      "grad_norm": 3.003065347671509,
+      "grad_norm": 2.2870659828186035,
       "learning_rate": 2.7482904591338327e-05,
-      "loss": 0.5614350509643554,
+      "loss": 0.586540756225586,
       "step": 4150
     },
     {
       "epoch": 1.3676326929338978,
-      "grad_norm": 3.551547050476074,
+      "grad_norm": 3.7163562774658203,
       "learning_rate": 2.7211548898295886e-05,
-      "loss": 0.5608404922485352,
+      "loss": 0.5858817672729493,
       "step": 4200
     },
     {
       "epoch": 1.3839140345164442,
-      "grad_norm": 2.9098353385925293,
+      "grad_norm": 2.730625629425049,
       "learning_rate": 2.694019320525345e-05,
-      "loss": 0.5317767715454101,
+      "loss": 0.5543305206298829,
       "step": 4250
     },
     {
       "epoch": 1.4001953760989905,
-      "grad_norm": 5.120357036590576,
+      "grad_norm": 3.728361129760742,
       "learning_rate": 2.666883751221101e-05,
-      "loss": 0.6232344818115234,
+      "loss": 0.6442169952392578,
       "step": 4300
     },
     {
       "epoch": 1.416476717681537,
-      "grad_norm": 2.4433658123016357,
+      "grad_norm": 2.2651383876800537,
       "learning_rate": 2.639748181916857e-05,
-      "loss": 0.555752182006836,
+      "loss": 0.582794075012207,
       "step": 4350
     },
     {
       "epoch": 1.4327580592640834,
-      "grad_norm": 2.675360679626465,
+      "grad_norm": 2.430931806564331,
       "learning_rate": 2.6126126126126128e-05,
-      "loss": 0.5809175872802734,
+      "loss": 0.6022589874267578,
       "step": 4400
     },
     {
       "epoch": 1.4490394008466296,
-      "grad_norm": 3.008749485015869,
+      "grad_norm": 3.222708225250244,
       "learning_rate": 2.5854770433083687e-05,
-      "loss": 0.6393648910522461,
+      "loss": 0.6583724975585937,
       "step": 4450
     },
     {
       "epoch": 1.465320742429176,
-      "grad_norm": 4.671627998352051,
+      "grad_norm": 4.678814888000488,
       "learning_rate": 2.5583414740041246e-05,
-      "loss": 0.5763546371459961,
+      "loss": 0.5943536376953125,
       "step": 4500
     },
     {
       "epoch": 1.4816020840117226,
-      "grad_norm": 2.942686080932617,
+      "grad_norm": 3.0403647422790527,
       "learning_rate": 2.5312059046998808e-05,
-      "loss": 0.5711518478393555,
+      "loss": 0.5958936309814453,
       "step": 4550
     },
     {
       "epoch": 1.497883425594269,
-      "grad_norm": 2.2898240089416504,
+      "grad_norm": 2.538402795791626,
       "learning_rate": 2.504070335395637e-05,
-      "loss": 0.5915301132202149,
+      "loss": 0.6167163467407226,
       "step": 4600
     },
     {
       "epoch": 1.5141647671768155,
-      "grad_norm": 5.238982677459717,
+      "grad_norm": 5.17458438873291,
       "learning_rate": 2.476934766091393e-05,
-      "loss": 0.5306437683105468,
+      "loss": 0.5508831405639648,
       "step": 4650
     },
     {
       "epoch": 1.530446108759362,
-      "grad_norm": 2.6982979774475098,
+      "grad_norm": 2.4245738983154297,
       "learning_rate": 2.4497991967871488e-05,
-      "loss": 0.5979887008666992,
+      "loss": 0.6157744598388671,
       "step": 4700
     },
     {
       "epoch": 1.5467274503419082,
-      "grad_norm": 5.090776443481445,
+      "grad_norm": 2.481553792953491,
       "learning_rate": 2.4226636274829047e-05,
-      "loss": 0.5904266357421875,
+      "loss": 0.6096940231323242,
       "step": 4750
     },
     {
       "epoch": 1.5630087919244544,
-      "grad_norm": 2.6866397857666016,
+      "grad_norm": 2.797117233276367,
       "learning_rate": 2.395528058178661e-05,
-      "loss": 0.5932189178466797,
+      "loss": 0.6150970458984375,
       "step": 4800
     },
     {
       "epoch": 1.5792901335070009,
-      "grad_norm": 3.7417988777160645,
+      "grad_norm": 4.05893611907959,
       "learning_rate": 2.3683924888744168e-05,
-      "loss": 0.595960578918457,
+      "loss": 0.6198799133300781,
       "step": 4850
     },
     {
       "epoch": 1.5955714750895473,
-      "grad_norm": 3.4997494220733643,
+      "grad_norm": 3.5510406494140625,
       "learning_rate": 2.3412569195701726e-05,
-      "loss": 0.5230794143676758,
+      "loss": 0.5494784545898438,
       "step": 4900
     },
     {
       "epoch": 1.6118528166720938,
-      "grad_norm": 4.909201145172119,
+      "grad_norm": 4.9277424812316895,
       "learning_rate": 2.3141213502659285e-05,
-      "loss": 0.48100479125976564,
+      "loss": 0.500348129272461,
       "step": 4950
     },
     {
       "epoch": 1.6281341582546403,
-      "grad_norm": 2.8310112953186035,
+      "grad_norm": 2.930206775665283,
       "learning_rate": 2.2869857809616847e-05,
-      "loss": 0.4874506378173828,
+      "loss": 0.5013648986816406,
       "step": 5000
     },
     {
       "epoch": 1.6444154998371867,
-      "grad_norm": 2.82600736618042,
+      "grad_norm": 10.043802261352539,
       "learning_rate": 2.2598502116574406e-05,
-      "loss": 0.5806295776367187,
+      "loss": 0.5953195190429688,
       "step": 5050
     },
     {
       "epoch": 1.660696841419733,
-      "grad_norm": 3.411588430404663,
+      "grad_norm": 3.334196090698242,
       "learning_rate": 2.2327146423531965e-05,
-      "loss": 0.517121810913086,
+      "loss": 0.5374015426635742,
       "step": 5100
     },
     {
       "epoch": 1.6769781830022794,
-      "grad_norm": 3.2024784088134766,
+      "grad_norm": 3.292771339416504,
       "learning_rate": 2.2055790730489524e-05,
-      "loss": 0.46841278076171877,
+      "loss": 0.49036331176757814,
       "step": 5150
     },
     {
       "epoch": 1.6932595245848256,
-      "grad_norm": 3.798309803009033,
+      "grad_norm": 3.440549373626709,
       "learning_rate": 2.1784435037447086e-05,
-      "loss": 0.5157822418212891,
+      "loss": 0.5311320114135742,
       "step": 5200
     },
     {
       "epoch": 1.709540866167372,
-      "grad_norm": 3.463024854660034,
+      "grad_norm": 3.720184803009033,
       "learning_rate": 2.1513079344404645e-05,
-      "loss": 0.5623917007446289,
+      "loss": 0.5855265045166016,
       "step": 5250
     },
     {
       "epoch": 1.7258222077499186,
-      "grad_norm": 9.00114917755127,
+      "grad_norm": 3.9988839626312256,
       "learning_rate": 2.1241723651362204e-05,
-      "loss": 0.5178924560546875,
+      "loss": 0.5410572052001953,
       "step": 5300
     },
     {
       "epoch": 1.742103549332465,
-      "grad_norm": 3.134282350540161,
+      "grad_norm": 3.244662046432495,
       "learning_rate": 2.0970367958319766e-05,
-      "loss": 0.5335901641845703,
+      "loss": 0.553615951538086,
       "step": 5350
     },
     {
       "epoch": 1.7583848909150115,
-      "grad_norm": 3.1152961254119873,
+      "grad_norm": 3.0616097450256348,
       "learning_rate": 2.0699012265277325e-05,
-      "loss": 0.46617908477783204,
+      "loss": 0.4832605361938477,
       "step": 5400
     },
     {
       "epoch": 1.774666232497558,
-      "grad_norm": 3.5342652797698975,
+      "grad_norm": 3.586770534515381,
       "learning_rate": 2.0427656572234887e-05,
-      "loss": 0.48794960021972655,
+      "loss": 0.5060377883911132,
       "step": 5450
     },
     {
       "epoch": 1.7909475740801042,
-      "grad_norm": 2.6719322204589844,
+      "grad_norm": 2.99122953414917,
       "learning_rate": 2.0156300879192446e-05,
-      "loss": 0.48549480438232423,
+      "loss": 0.510421142578125,
       "step": 5500
     },
     {
       "epoch": 1.8072289156626506,
-      "grad_norm": 2.8136425018310547,
+      "grad_norm": 2.647064685821533,
       "learning_rate": 1.9884945186150005e-05,
-      "loss": 0.504596939086914,
+      "loss": 0.5255117034912109,
       "step": 5550
     },
     {
       "epoch": 1.8235102572451969,
-      "grad_norm": 2.985659599304199,
+      "grad_norm": 3.2572667598724365,
       "learning_rate": 1.9613589493107567e-05,
-      "loss": 0.46674480438232424,
+      "loss": 0.48496082305908206,
       "step": 5600
     },
     {
       "epoch": 1.8397915988277433,
-      "grad_norm": 4.119204521179199,
+      "grad_norm": 4.511913299560547,
       "learning_rate": 1.9342233800065126e-05,
-      "loss": 0.4856386566162109,
+      "loss": 0.5061034393310547,
       "step": 5650
     },
     {
       "epoch": 1.8560729404102898,
-      "grad_norm": 3.0939385890960693,
+      "grad_norm": 2.9536867141723633,
       "learning_rate": 1.9070878107022685e-05,
-      "loss": 0.5406102752685547,
+      "loss": 0.5593103790283203,
       "step": 5700
     },
     {
       "epoch": 1.8723542819928363,
-      "grad_norm": 2.096914052963257,
+      "grad_norm": 2.071770429611206,
       "learning_rate": 1.8799522413980247e-05,
-      "loss": 0.49306621551513674,
+      "loss": 0.5091780471801758,
       "step": 5750
     },
     {
       "epoch": 1.8886356235753827,
-      "grad_norm": 4.529496669769287,
+      "grad_norm": 4.718142986297607,
       "learning_rate": 1.8528166720937806e-05,
-      "loss": 0.4564324188232422,
+      "loss": 0.4754158401489258,
       "step": 5800
     },
     {
       "epoch": 1.904916965157929,
-      "grad_norm": 3.7081360816955566,
+      "grad_norm": 4.707818508148193,
       "learning_rate": 1.8256811027895364e-05,
-      "loss": 0.5462873458862305,
+      "loss": 0.5697180557250977,
       "step": 5850
     },
     {
       "epoch": 1.9211983067404754,
-      "grad_norm": 4.159782886505127,
+      "grad_norm": 4.631124973297119,
       "learning_rate": 1.7985455334852927e-05,
-      "loss": 0.47431507110595705,
+      "loss": 0.4917443084716797,
       "step": 5900
     },
     {
       "epoch": 1.9374796483230217,
-      "grad_norm": 3.467459201812744,
+      "grad_norm": 3.7037010192871094,
       "learning_rate": 1.7714099641810485e-05,
-      "loss": 0.5584798049926758,
+      "loss": 0.577647933959961,
       "step": 5950
     },
     {
       "epoch": 1.9537609899055681,
-      "grad_norm": 4.213109970092773,
+      "grad_norm": 4.4254302978515625,
       "learning_rate": 1.7442743948768044e-05,
-      "loss": 0.47076557159423826,
+      "loss": 0.49387569427490235,
       "step": 6000
     },
     {
       "epoch": 1.9700423314881146,
-      "grad_norm": 4.383584022521973,
+      "grad_norm": 4.429046630859375,
       "learning_rate": 1.7171388255725606e-05,
-      "loss": 0.5349767303466797,
+      "loss": 0.555143165588379,
       "step": 6050
     },
     {
       "epoch": 1.986323673070661,
-      "grad_norm": 3.1514296531677246,
+      "grad_norm": 3.211913824081421,
       "learning_rate": 1.6900032562683165e-05,
-      "loss": 0.5139558792114258,
+      "loss": 0.5266495132446289,
       "step": 6100
     },
     {
       "epoch": 2.0,
-      "eval_bertscore_f1": 0.9762323709095225,
-      "eval_bleu": 0.7430723352502658,
-      "eval_loss": 0.3539265990257263,
-      "eval_meteor": 0.8151086573537881,
-      "eval_rouge1": 0.8885716440737175,
-      "eval_rouge2": 0.8125576905491395,
-      "eval_runtime": 137.3289,
-      "eval_samples_per_second": 9.408,
-      "eval_steps_per_second": 1.18,
+      "eval_bertscore_f1": 0.9751379909253342,
+      "eval_bleu": 0.7283162211638717,
+      "eval_loss": 0.3667888939380646,
+      "eval_meteor": 0.806130448591728,
+      "eval_rouge1": 0.8830570311221204,
+      "eval_rouge2": 0.803848501332357,
+      "eval_runtime": 126.2611,
+      "eval_samples_per_second": 10.233,
+      "eval_steps_per_second": 1.283,
       "step": 6142
     },
     {
       "epoch": 2.0026050146532075,
-      "grad_norm": 2.005295991897583,
+      "grad_norm": 2.130920886993408,
       "learning_rate": 1.6628676869640724e-05,
-      "loss": 0.4775341796875,
+      "loss": 0.49579532623291017,
       "step": 6150
     },
     {
       "epoch": 2.018886356235754,
-      "grad_norm": 3.9823381900787354,
+      "grad_norm": 3.622119665145874,
       "learning_rate": 1.6357321176598286e-05,
-      "loss": 0.5708214950561523,
+      "loss": 0.5926795959472656,
       "step": 6200
     },
     {
       "epoch": 2.0351676978183004,
-      "grad_norm": 3.592066764831543,
+      "grad_norm": 4.096598148345947,
       "learning_rate": 1.6085965483555845e-05,
-      "loss": 0.494837646484375,
+      "loss": 0.511412124633789,
       "step": 6250
     },
     {
       "epoch": 2.0514490394008464,
-      "grad_norm": 3.034447193145752,
+      "grad_norm": 3.3541550636291504,
       "learning_rate": 1.5814609790513404e-05,
-      "loss": 0.48195556640625,
+      "loss": 0.5026980590820312,
       "step": 6300
     },
     {
       "epoch": 2.067730380983393,
-      "grad_norm": 3.7622528076171875,
+      "grad_norm": 3.797450304031372,
       "learning_rate": 1.5543254097470966e-05,
-      "loss": 0.46555606842041014,
+      "loss": 0.48105335235595703,
       "step": 6350
     },
     {
       "epoch": 2.0840117225659394,
-      "grad_norm": 4.787505626678467,
+      "grad_norm": 4.0247626304626465,
       "learning_rate": 1.5271898404428525e-05,
-      "loss": 0.4631903076171875,
+      "loss": 0.48235019683837893,
       "step": 6400
     },
     {
       "epoch": 2.100293064148486,
-      "grad_norm": 4.470603942871094,
+      "grad_norm": 4.341481685638428,
       "learning_rate": 1.5000542711386087e-05,
-      "loss": 0.4912979888916016,
+      "loss": 0.5079761123657227,
       "step": 6450
     },
     {
       "epoch": 2.1165744057310323,
-      "grad_norm": 2.7138943672180176,
+      "grad_norm": 2.415269374847412,
       "learning_rate": 1.4729187018343646e-05,
-      "loss": 0.5140954208374023,
+      "loss": 0.5332447814941407,
       "step": 6500
     },
     {
       "epoch": 2.1328557473135787,
-      "grad_norm": 4.439281463623047,
+      "grad_norm": 4.66851282119751,
       "learning_rate": 1.4457831325301205e-05,
-      "loss": 0.5343180465698242,
+      "loss": 0.5538092041015625,
       "step": 6550
     },
     {
       "epoch": 2.149137088896125,
-      "grad_norm": 3.9600956439971924,
+      "grad_norm": 3.448925495147705,
       "learning_rate": 1.4186475632258767e-05,
-      "loss": 0.5393027114868164,
+      "loss": 0.5555255889892579,
       "step": 6600
     },
     {
       "epoch": 2.165418430478671,
-      "grad_norm": 5.374337673187256,
+      "grad_norm": 4.12534236907959,
       "learning_rate": 1.3915119939216326e-05,
-      "loss": 0.4814667892456055,
+      "loss": 0.49463138580322263,
       "step": 6650
     },
     {
       "epoch": 2.1816997720612177,
-      "grad_norm": 2.706310987472534,
+      "grad_norm": 2.726137638092041,
       "learning_rate": 1.3643764246173885e-05,
-      "loss": 0.4734183883666992,
+      "loss": 0.4911256408691406,
       "step": 6700
     },
     {
       "epoch": 2.197981113643764,
-      "grad_norm": 3.1908249855041504,
+      "grad_norm": 2.9859631061553955,
       "learning_rate": 1.3372408553131447e-05,
-      "loss": 0.5154135131835937,
+      "loss": 0.5317694091796875,
       "step": 6750
     },
     {
       "epoch": 2.2142624552263106,
-      "grad_norm": 3.3044145107269287,
+      "grad_norm": 3.0363500118255615,
       "learning_rate": 1.3101052860089006e-05,
-      "loss": 0.4694767761230469,
+      "loss": 0.48204010009765624,
       "step": 6800
     },
     {
       "epoch": 2.230543796808857,
-      "grad_norm": 2.6275320053100586,
+      "grad_norm": 2.7553327083587646,
       "learning_rate": 1.2829697167046565e-05,
-      "loss": 0.5074032974243164,
+      "loss": 0.5275654602050781,
       "step": 6850
     },
     {
       "epoch": 2.2468251383914035,
-      "grad_norm": 3.4295599460601807,
+      "grad_norm": 3.37378191947937,
       "learning_rate": 1.2558341474004127e-05,
-      "loss": 0.47720569610595703,
+      "loss": 0.49252197265625,
       "step": 6900
     },
     {
       "epoch": 2.26310647997395,
-      "grad_norm": 3.5452775955200195,
+      "grad_norm": 3.5572731494903564,
       "learning_rate": 1.2286985780961686e-05,
-      "loss": 0.5167514038085937,
+      "loss": 0.5376744079589844,
       "step": 6950
     },
     {
       "epoch": 2.2793878215564964,
-      "grad_norm": 3.8189544677734375,
+      "grad_norm": 4.323084354400635,
       "learning_rate": 1.2015630087919246e-05,
-      "loss": 0.5443729400634766,
+      "loss": 0.5603115081787109,
       "step": 7000
     },
     {
       "epoch": 2.295669163139043,
-      "grad_norm": 3.834667682647705,
+      "grad_norm": 3.9723100662231445,
       "learning_rate": 1.1744274394876805e-05,
-      "loss": 0.43803905487060546,
+      "loss": 0.4530460739135742,
       "step": 7050
     },
     {
       "epoch": 2.311950504721589,
-      "grad_norm": 3.922828435897827,
+      "grad_norm": 4.0148491859436035,
       "learning_rate": 1.1472918701834365e-05,
-      "loss": 0.465465202331543,
+      "loss": 0.4808524703979492,
       "step": 7100
     },
     {
       "epoch": 2.3282318463041354,
-      "grad_norm": 3.319157838821411,
+      "grad_norm": 3.357252597808838,
       "learning_rate": 1.1201563008791926e-05,
-      "loss": 0.5172126388549805,
+      "loss": 0.5342596817016602,
       "step": 7150
     },
     {
       "epoch": 2.344513187886682,
-      "grad_norm": 1.2773399353027344,
+      "grad_norm": 1.4084240198135376,
       "learning_rate": 1.0930207315749485e-05,
-      "loss": 0.5022563552856445,
+      "loss": 0.5188829040527344,
       "step": 7200
     },
     {
       "epoch": 2.3607945294692283,
-      "grad_norm": 3.884661912918091,
+      "grad_norm": 4.223776817321777,
       "learning_rate": 1.0658851622707045e-05,
-      "loss": 0.5398352813720703,
+      "loss": 0.5551160049438476,
       "step": 7250
     },
     {
       "epoch": 2.3770758710517748,
-      "grad_norm": 3.0721209049224854,
+      "grad_norm": 3.2388432025909424,
       "learning_rate": 1.0387495929664604e-05,
-      "loss": 0.4502225875854492,
+      "loss": 0.4678123092651367,
       "step": 7300
     },
     {
       "epoch": 2.393357212634321,
-      "grad_norm": 2.1282992362976074,
+      "grad_norm": 2.2621917724609375,
       "learning_rate": 1.0116140236622165e-05,
-      "loss": 0.4745651626586914,
+      "loss": 0.49837650299072267,
       "step": 7350
     },
     {
       "epoch": 2.4096385542168672,
-      "grad_norm": 4.129814147949219,
+      "grad_norm": 3.9080207347869873,
       "learning_rate": 9.844784543579723e-06,
-      "loss": 0.4516494369506836,
+      "loss": 0.4679309844970703,
       "step": 7400
     },
     {
       "epoch": 2.4259198957994137,
-      "grad_norm": 2.759620428085327,
+      "grad_norm": 2.2795565128326416,
       "learning_rate": 9.573428850537284e-06,
-      "loss": 0.4808626174926758,
+      "loss": 0.4956610870361328,
       "step": 7450
     },
     {
       "epoch": 2.44220123738196,
-      "grad_norm": 2.358985424041748,
+      "grad_norm": 2.872938632965088,
       "learning_rate": 9.302073157494844e-06,
-      "loss": 0.46834888458251955,
+      "loss": 0.48609119415283203,
       "step": 7500
     },
     {
       "epoch": 2.4584825789645066,
-      "grad_norm": 4.089727401733398,
+      "grad_norm": 3.869072675704956,
       "learning_rate": 9.030717464452403e-06,
-      "loss": 0.5258768844604492,
+      "loss": 0.5468455886840821,
       "step": 7550
     },
     {
       "epoch": 2.474763920547053,
-      "grad_norm": 4.14641809463501,
+      "grad_norm": 3.9535272121429443,
       "learning_rate": 8.759361771409964e-06,
-      "loss": 0.43479076385498044,
+      "loss": 0.4581578826904297,
       "step": 7600
     },
     {
       "epoch": 2.4910452621295995,
-      "grad_norm": 2.1294007301330566,
+      "grad_norm": 2.2123703956604004,
       "learning_rate": 8.488006078367524e-06,
-      "loss": 0.4723660659790039,
+      "loss": 0.48407554626464844,
       "step": 7650
     },
     {
       "epoch": 2.507326603712146,
-      "grad_norm": 3.301629066467285,
+      "grad_norm": 3.5734164714813232,
       "learning_rate": 8.216650385325085e-06,
-      "loss": 0.39770748138427736,
+      "loss": 0.4096232986450195,
       "step": 7700
     },
     {
       "epoch": 2.5236079452946925,
-      "grad_norm": 2.9783337116241455,
+      "grad_norm": 5.056736946105957,
       "learning_rate": 7.945294692282644e-06,
-      "loss": 0.46555194854736326,
+      "loss": 0.48154861450195313,
       "step": 7750
     },
     {
       "epoch": 2.539889286877239,
-      "grad_norm": 4.9203619956970215,
+      "grad_norm": 5.742955207824707,
       "learning_rate": 7.673938999240204e-06,
-      "loss": 0.4547107696533203,
+      "loss": 0.46851539611816406,
       "step": 7800
     },
     {
       "epoch": 2.556170628459785,
-      "grad_norm": 4.045784950256348,
+      "grad_norm": 3.6929991245269775,
       "learning_rate": 7.402583306197765e-06,
-      "loss": 0.43739501953125,
+      "loss": 0.4527290725708008,
       "step": 7850
     },
     {
       "epoch": 2.5724519700423314,
-      "grad_norm": 3.234687566757202,
+      "grad_norm": 4.624508857727051,
       "learning_rate": 7.1312276131553235e-06,
-      "loss": 0.467929573059082,
+      "loss": 0.487193717956543,
       "step": 7900
     },
     {
       "epoch": 2.588733311624878,
-      "grad_norm": 3.6721928119659424,
+      "grad_norm": 3.526402711868286,
       "learning_rate": 6.859871920112884e-06,
-      "loss": 0.4593897247314453,
+      "loss": 0.4767793273925781,
       "step": 7950
     },
     {
       "epoch": 2.6050146532074243,
-      "grad_norm": 2.296928644180298,
+      "grad_norm": 2.557279586791992,
       "learning_rate": 6.5885162270704446e-06,
-      "loss": 0.46631996154785155,
+      "loss": 0.47788654327392577,
       "step": 8000
     },
     {
       "epoch": 2.6212959947899708,
-      "grad_norm": 3.3836448192596436,
+      "grad_norm": 3.1233720779418945,
       "learning_rate": 6.317160534028005e-06,
-      "loss": 0.4127694702148437,
+      "loss": 0.43118988037109374,
       "step": 8050
     },
     {
       "epoch": 2.6375773363725172,
-      "grad_norm": 3.1450395584106445,
+      "grad_norm": 3.1848602294921875,
       "learning_rate": 6.045804840985564e-06,
-      "loss": 0.4709606170654297,
+      "loss": 0.4824806594848633,
       "step": 8100
     },
     {
       "epoch": 2.6538586779550632,
-      "grad_norm": 2.056903600692749,
+      "grad_norm": 2.1851677894592285,
       "learning_rate": 5.774449147943124e-06,
-      "loss": 0.5080145263671875,
+      "loss": 0.5195888900756835,
       "step": 8150
     },
     {
       "epoch": 2.6701400195376097,
-      "grad_norm": 4.611896514892578,
+      "grad_norm": 4.459860801696777,
       "learning_rate": 5.503093454900684e-06,
-      "loss": 0.5028559875488281,
+      "loss": 0.5190325164794922,
       "step": 8200
     },
     {
       "epoch": 2.686421361120156,
-      "grad_norm": 2.8389217853546143,
+      "grad_norm": 2.845478057861328,
       "learning_rate": 5.231737761858245e-06,
-      "loss": 0.5155687713623047,
+      "loss": 0.5278813171386719,
       "step": 8250
     },
     {
       "epoch": 2.7027027027027026,
-      "grad_norm": 3.570791721343994,
+      "grad_norm": 3.6786937713623047,
       "learning_rate": 4.960382068815804e-06,
-      "loss": 0.46154869079589844,
+      "loss": 0.4795223236083984,
       "step": 8300
     },
     {
       "epoch": 2.718984044285249,
-      "grad_norm": 3.8014798164367676,
+      "grad_norm": 3.6952383518218994,
       "learning_rate": 4.689026375773364e-06,
-      "loss": 0.5015551376342774,
+      "loss": 0.5154057312011718,
       "step": 8350
     },
     {
       "epoch": 2.7352653858677956,
-      "grad_norm": 2.136251926422119,
+      "grad_norm": 2.0930325984954834,
       "learning_rate": 4.417670682730924e-06,
-      "loss": 0.4702875900268555,
+      "loss": 0.4847321319580078,
       "step": 8400
     },
     {
       "epoch": 2.751546727450342,
-      "grad_norm": 5.566357612609863,
+      "grad_norm": 6.89557409286499,
       "learning_rate": 4.146314989688484e-06,
-      "loss": 0.3770464324951172,
+      "loss": 0.39144508361816405,
       "step": 8450
     },
     {
       "epoch": 2.7678280690328885,
-      "grad_norm": 5.836922645568848,
+      "grad_norm": 4.23341703414917,
       "learning_rate": 3.874959296646044e-06,
-      "loss": 0.4574332809448242,
+      "loss": 0.47240074157714845,
       "step": 8500
     },
     {
       "epoch": 2.784109410615435,
-      "grad_norm": 16.352832794189453,
+      "grad_norm": 5.129818439483643,
       "learning_rate": 3.603603603603604e-06,
-      "loss": 0.48586685180664063,
+      "loss": 0.4966455841064453,
       "step": 8550
     },
     {
       "epoch": 2.800390752197981,
-      "grad_norm": 3.018529176712036,
+      "grad_norm": 2.9496281147003174,
       "learning_rate": 3.3322479105611635e-06,
-      "loss": 0.4619815063476562,
+      "loss": 0.48110599517822267,
       "step": 8600
     },
     {
       "epoch": 2.8166720937805274,
-      "grad_norm": 5.671543598175049,
+      "grad_norm": 5.308284759521484,
       "learning_rate": 3.0608922175187236e-06,
-      "loss": 0.45566730499267577,
+      "loss": 0.4717050552368164,
       "step": 8650
     },
     {
       "epoch": 2.832953435363074,
-      "grad_norm": 3.141303062438965,
+      "grad_norm": 3.275146722793579,
       "learning_rate": 2.7895365244762837e-06,
-      "loss": 0.423536491394043,
+      "loss": 0.4382866668701172,
       "step": 8700
     },
     {
       "epoch": 2.8492347769456203,
-      "grad_norm": 2.619455337524414,
+      "grad_norm": 2.5721969604492188,
       "learning_rate": 2.5181808314338434e-06,
-      "loss": 0.4463621520996094,
+      "loss": 0.46353874206542967,
       "step": 8750
     },
     {
       "epoch": 2.865516118528167,
-      "grad_norm": 3.0434818267822266,
+      "grad_norm": 3.426837921142578,
       "learning_rate": 2.2468251383914035e-06,
-      "loss": 0.4879434967041016,
+      "loss": 0.5103521347045898,
       "step": 8800
     },
     {
       "epoch": 2.8817974601107132,
-      "grad_norm": 2.338992118835449,
+      "grad_norm": 2.3799993991851807,
       "learning_rate": 1.9754694453489636e-06,
-      "loss": 0.4577774429321289,
+      "loss": 0.47350223541259767,
       "step": 8850
     },
     {
       "epoch": 2.8980788016932593,
-      "grad_norm": 3.1424238681793213,
+      "grad_norm": 3.316828966140747,
       "learning_rate": 1.7041137523065235e-06,
-      "loss": 0.4456531524658203,
+      "loss": 0.4602887725830078,
       "step": 8900
     },
     {
       "epoch": 2.9143601432758057,
-      "grad_norm": 4.704187870025635,
+      "grad_norm": 5.089015007019043,
       "learning_rate": 1.4327580592640834e-06,
-      "loss": 0.4674592590332031,
+      "loss": 0.4908058547973633,
       "step": 8950
     },
     {
       "epoch": 2.930641484858352,
-      "grad_norm": 5.090559959411621,
+      "grad_norm": 5.589182376861572,
       "learning_rate": 1.1614023662216434e-06,
-      "loss": 0.43730361938476564,
+      "loss": 0.4539414215087891,
       "step": 9000
     },
     {
       "epoch": 2.9469228264408986,
-      "grad_norm": 2.395810604095459,
+      "grad_norm": 2.4032421112060547,
       "learning_rate": 8.900466731792033e-07,
-      "loss": 0.4117137908935547,
+      "loss": 0.42325592041015625,
       "step": 9050
     },
     {
       "epoch": 2.963204168023445,
-      "grad_norm": 2.933120012283325,
+      "grad_norm": 3.4075100421905518,
       "learning_rate": 6.186909801367633e-07,
-      "loss": 0.38719585418701175,
+      "loss": 0.4070619201660156,
       "step": 9100
     },
     {
       "epoch": 2.9794855096059916,
-      "grad_norm": 1.5927964448928833,
+      "grad_norm": 1.6805068254470825,
       "learning_rate": 3.4733528709432325e-07,
-      "loss": 0.42159496307373046,
+      "loss": 0.4350212860107422,
       "step": 9150
     },
     {
       "epoch": 2.995766851188538,
-      "grad_norm": 4.6336774826049805,
+      "grad_norm": 3.386293888092041,
       "learning_rate": 7.597959405188321e-08,
-      "loss": 0.4757699966430664,
+      "loss": 0.4917270660400391,
       "step": 9200
     },
     {
       "epoch": 3.0,
-      "eval_bertscore_f1": 0.9787175891107819,
-      "eval_bleu": 0.7693596969484261,
-      "eval_loss": 0.32842469215393066,
-      "eval_meteor": 0.8357277347287969,
-      "eval_rouge1": 0.8992664738649058,
-      "eval_rouge2": 0.828427573552017,
-      "eval_runtime": 129.1375,
-      "eval_samples_per_second": 10.005,
-      "eval_steps_per_second": 1.254,
+      "eval_bertscore_f1": 0.9777762908776847,
+      "eval_bleu": 0.7589549870918858,
+      "eval_loss": 0.3381657600402832,
+      "eval_meteor": 0.8307886222747983,
+      "eval_rouge1": 0.893313778543083,
+      "eval_rouge2": 0.821781747009253,
+      "eval_runtime": 130.6802,
+      "eval_samples_per_second": 9.887,
+      "eval_steps_per_second": 1.24,
       "step": 9213
     }
   ],

checkpoint-9213/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8945bcdf4317e2f7559f3b2665cbaf54f85caa090c8d212b42fe7b6b9307fa8e
+oid sha256:21e7988a724f50ab4d65854a8bd4de7b4eceb713886a6a05ceeaeae7fb11b713
 size 4920