Upload folder using huggingface_hub

checkpoint-7075/adapter_config.json

@@ -7,14 +7,14 @@
   "exclude_modules": null,
   "fan_in_fan_out": false,
   "inference_mode": true,
-  "init_lora_weights": true,
+  "init_lora_weights": "gaussian",
   "layer_replication": null,
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
   "lora_alpha": 16,
   "lora_bias": false,
-  "lora_dropout": 0.1,
+  "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
@@ -23,8 +23,8 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "q",
-    "v"
+    "v",
+    "q"
   ],
   "task_type": "SEQ_2_SEQ_LM",
   "use_dora": false,

checkpoint-7075/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:626aff5650b6dcfb6222236d07bcc5b0b087d26230196dfb68526f0dc5e7275a
+oid sha256:7ec74429cd88425595c28644ccdcb6ba72c1587c51b304c636b2c63f3c665d4d
 size 3558888

checkpoint-7075/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:799f271151d88084c967000bbb5c74f1dae3634d4b2f7c37c86fe58ad77d1e20
+oid sha256:b5ff62a64be4676061af7cd8f46ce20a726f85268cf60c5c00c2fd0848ead8ca
 size 7198906

checkpoint-7075/trainer_state.json

@@ -1,7 +1,7 @@
 {
   "best_global_step": 7075,
-  "best_metric": 0.9583620159785347,
-  "best_model_checkpoint": "/kaggle/working/codet5-k8s-lora/checkpoint-7075",
+  "best_metric": 0.9626803387159316,
+  "best_model_checkpoint": "./output-codet5-k8s-lora/checkpoint-7075",
   "epoch": 5.0,
   "eval_steps": 500,
   "global_step": 7075,
@@ -11,1054 +11,1054 @@
   "log_history": [
     {
       "epoch": 0.0353356890459364,
-      "grad_norm": 88194.484375,
+      "grad_norm": 132030.078125,
       "learning_rate": 4.9653710247349825e-05,
-      "loss": 4.5256,
+      "loss": 4.4129,
       "step": 50
     },
     {
       "epoch": 0.0706713780918728,
-      "grad_norm": 53482.01171875,
+      "grad_norm": 259212.703125,
       "learning_rate": 4.930035335689046e-05,
-      "loss": 3.5832,
+      "loss": 3.4742,
       "step": 100
     },
     {
       "epoch": 0.10600706713780919,
-      "grad_norm": 88556.421875,
+      "grad_norm": 135731.703125,
       "learning_rate": 4.89469964664311e-05,
-      "loss": 2.9383,
+      "loss": 2.4982,
       "step": 150
     },
     {
       "epoch": 0.1413427561837456,
-      "grad_norm": 53563.4609375,
+      "grad_norm": 253683.546875,
       "learning_rate": 4.859363957597173e-05,
-      "loss": 2.0823,
+      "loss": 2.2733,
       "step": 200
     },
     {
       "epoch": 0.17667844522968199,
-      "grad_norm": 69098.1875,
+      "grad_norm": 447967.6875,
       "learning_rate": 4.8240282685512374e-05,
-      "loss": 1.9462,
+      "loss": 2.1743,
       "step": 250
     },
     {
       "epoch": 0.21201413427561838,
-      "grad_norm": 137658.15625,
+      "grad_norm": 248496.0,
       "learning_rate": 4.788692579505301e-05,
-      "loss": 1.7835,
+      "loss": 2.0272,
       "step": 300
     },
     {
       "epoch": 0.24734982332155478,
-      "grad_norm": 77329.25,
+      "grad_norm": 137680.71875,
       "learning_rate": 4.753356890459364e-05,
-      "loss": 1.7243,
+      "loss": 1.9573,
       "step": 350
     },
     {
       "epoch": 0.2826855123674912,
-      "grad_norm": 79922.1796875,
+      "grad_norm": 165613.9375,
       "learning_rate": 4.718021201413428e-05,
-      "loss": 1.7295,
+      "loss": 1.9568,
       "step": 400
     },
     {
       "epoch": 0.31802120141342755,
-      "grad_norm": 329205.75,
+      "grad_norm": 139130.828125,
       "learning_rate": 4.682685512367491e-05,
-      "loss": 1.5981,
+      "loss": 1.8366,
       "step": 450
     },
     {
       "epoch": 0.35335689045936397,
-      "grad_norm": 79152.3671875,
+      "grad_norm": 141737.296875,
       "learning_rate": 4.647349823321555e-05,
-      "loss": 1.5299,
+      "loss": 1.7679,
       "step": 500
     },
     {
       "epoch": 0.38869257950530034,
-      "grad_norm": 96051.7109375,
+      "grad_norm": 151085.828125,
       "learning_rate": 4.612014134275618e-05,
-      "loss": 1.4665,
+      "loss": 1.7186,
       "step": 550
     },
     {
       "epoch": 0.42402826855123676,
-      "grad_norm": 84361.9921875,
+      "grad_norm": 143353.3125,
       "learning_rate": 4.5766784452296823e-05,
-      "loss": 1.4434,
+      "loss": 1.6871,
       "step": 600
     },
     {
       "epoch": 0.45936395759717313,
-      "grad_norm": 99281.8984375,
+      "grad_norm": 129146.640625,
       "learning_rate": 4.541342756183746e-05,
-      "loss": 1.4238,
+      "loss": 1.6727,
       "step": 650
     },
     {
       "epoch": 0.49469964664310956,
-      "grad_norm": 1462805.5,
+      "grad_norm": 318427.25,
       "learning_rate": 4.50600706713781e-05,
-      "loss": 1.3647,
+      "loss": 1.6139,
       "step": 700
     },
     {
       "epoch": 0.5300353356890459,
-      "grad_norm": 114605.0390625,
+      "grad_norm": 167751.1875,
       "learning_rate": 4.470671378091873e-05,
-      "loss": 1.3974,
+      "loss": 1.6485,
       "step": 750
     },
     {
       "epoch": 0.5653710247349824,
-      "grad_norm": 114158.0,
+      "grad_norm": 185479.96875,
       "learning_rate": 4.435335689045937e-05,
-      "loss": 1.3204,
+      "loss": 1.5644,
       "step": 800
     },
     {
       "epoch": 0.6007067137809188,
-      "grad_norm": 98117.0078125,
+      "grad_norm": 162241.875,
       "learning_rate": 4.4000000000000006e-05,
-      "loss": 1.267,
+      "loss": 1.5162,
       "step": 850
     },
     {
       "epoch": 0.6360424028268551,
-      "grad_norm": 91731.5859375,
+      "grad_norm": 220409.78125,
       "learning_rate": 4.364664310954063e-05,
-      "loss": 1.2922,
+      "loss": 1.5365,
       "step": 900
     },
     {
       "epoch": 0.6713780918727915,
-      "grad_norm": 126348.9609375,
+      "grad_norm": 219547.015625,
       "learning_rate": 4.329328621908127e-05,
-      "loss": 1.2541,
+      "loss": 1.4956,
       "step": 950
     },
     {
       "epoch": 0.7067137809187279,
-      "grad_norm": 120821.4296875,
+      "grad_norm": 151856.1875,
       "learning_rate": 4.293992932862191e-05,
-      "loss": 1.2063,
+      "loss": 1.4474,
       "step": 1000
     },
     {
       "epoch": 0.7420494699646644,
-      "grad_norm": 116547.875,
+      "grad_norm": 139005.203125,
       "learning_rate": 4.258657243816255e-05,
-      "loss": 1.1705,
+      "loss": 1.4196,
       "step": 1050
     },
     {
       "epoch": 0.7773851590106007,
-      "grad_norm": 142100.546875,
+      "grad_norm": 207968.453125,
       "learning_rate": 4.223321554770318e-05,
-      "loss": 1.1996,
+      "loss": 1.4448,
       "step": 1100
     },
     {
       "epoch": 0.8127208480565371,
-      "grad_norm": 437601.09375,
+      "grad_norm": 226858.75,
       "learning_rate": 4.187985865724382e-05,
-      "loss": 1.1202,
+      "loss": 1.3649,
       "step": 1150
     },
     {
       "epoch": 0.8480565371024735,
-      "grad_norm": 113456.03125,
+      "grad_norm": 139849.125,
       "learning_rate": 4.1526501766784455e-05,
-      "loss": 1.0699,
+      "loss": 1.3146,
       "step": 1200
     },
     {
       "epoch": 0.8833922261484098,
-      "grad_norm": 135482.171875,
+      "grad_norm": 191754.421875,
       "learning_rate": 4.1173144876325096e-05,
-      "loss": 1.1186,
+      "loss": 1.3667,
       "step": 1250
     },
     {
       "epoch": 0.9187279151943463,
-      "grad_norm": 108526.2421875,
+      "grad_norm": 399802.3125,
       "learning_rate": 4.081978798586573e-05,
-      "loss": 1.1162,
+      "loss": 1.3588,
       "step": 1300
     },
     {
       "epoch": 0.9540636042402827,
-      "grad_norm": 104943.265625,
+      "grad_norm": 163376.953125,
       "learning_rate": 4.0466431095406356e-05,
-      "loss": 1.0548,
+      "loss": 1.305,
       "step": 1350
     },
     {
       "epoch": 0.9893992932862191,
-      "grad_norm": 985795.1875,
+      "grad_norm": 235141.46875,
       "learning_rate": 4.0113074204947e-05,
-      "loss": 1.0657,
+      "loss": 1.3144,
       "step": 1400
     },
     {
       "epoch": 1.0,
-      "eval_bertscore_f1": 0.9434455523617787,
-      "eval_bleu": 0.34195108703106286,
-      "eval_loss": 0.8279675245285034,
-      "eval_meteor": 0.552436877090645,
-      "eval_rouge1": 0.7415447899379005,
-      "eval_rouge2": 0.5638244221936095,
-      "eval_runtime": 459.8688,
-      "eval_samples_per_second": 14.056,
-      "eval_steps_per_second": 0.879,
+      "eval_bertscore_f1": 0.9499767605989876,
+      "eval_bleu": 0.41483915022966605,
+      "eval_loss": 0.6510708332061768,
+      "eval_meteor": 0.6007343806481693,
+      "eval_rouge1": 0.7698895071809828,
+      "eval_rouge2": 0.6286208265783073,
+      "eval_runtime": 476.7994,
+      "eval_samples_per_second": 13.557,
+      "eval_steps_per_second": 0.847,
       "step": 1415
     },
     {
       "epoch": 1.0247349823321554,
-      "grad_norm": 193677.953125,
+      "grad_norm": 154460.578125,
       "learning_rate": 3.975971731448763e-05,
-      "loss": 1.0568,
+      "loss": 1.3006,
       "step": 1450
     },
     {
       "epoch": 1.0600706713780919,
-      "grad_norm": 135512.28125,
+      "grad_norm": 152772.09375,
       "learning_rate": 3.940636042402827e-05,
-      "loss": 1.0642,
+      "loss": 1.3061,
       "step": 1500
     },
     {
       "epoch": 1.0954063604240283,
-      "grad_norm": 127916.6953125,
+      "grad_norm": 157520.671875,
       "learning_rate": 3.9053003533568905e-05,
-      "loss": 1.0061,
+      "loss": 1.2585,
       "step": 1550
     },
     {
       "epoch": 1.1307420494699647,
-      "grad_norm": 135308.234375,
+      "grad_norm": 159070.796875,
       "learning_rate": 3.8699646643109545e-05,
-      "loss": 1.005,
+      "loss": 1.2477,
       "step": 1600
     },
     {
       "epoch": 1.1660777385159011,
-      "grad_norm": 165271.25,
+      "grad_norm": 162802.515625,
       "learning_rate": 3.834628975265018e-05,
-      "loss": 1.0272,
+      "loss": 1.2697,
       "step": 1650
     },
     {
       "epoch": 1.2014134275618376,
-      "grad_norm": 147234.546875,
+      "grad_norm": 218109.6875,
       "learning_rate": 3.799293286219082e-05,
-      "loss": 0.9993,
+      "loss": 1.2485,
       "step": 1700
     },
     {
       "epoch": 1.2367491166077738,
-      "grad_norm": 752427.875,
+      "grad_norm": 201347.4375,
       "learning_rate": 3.763957597173145e-05,
-      "loss": 0.9749,
+      "loss": 1.218,
       "step": 1750
     },
     {
       "epoch": 1.2720848056537102,
-      "grad_norm": 135767.078125,
+      "grad_norm": 171368.71875,
       "learning_rate": 3.728621908127209e-05,
-      "loss": 1.0294,
+      "loss": 1.2826,
       "step": 1800
     },
     {
       "epoch": 1.3074204946996466,
-      "grad_norm": 148665.4375,
+      "grad_norm": 178375.859375,
       "learning_rate": 3.693286219081272e-05,
-      "loss": 1.0313,
+      "loss": 1.2763,
       "step": 1850
     },
     {
       "epoch": 1.342756183745583,
-      "grad_norm": 130839.71875,
+      "grad_norm": 163735.65625,
       "learning_rate": 3.6579505300353355e-05,
-      "loss": 0.9474,
+      "loss": 1.2008,
       "step": 1900
     },
     {
       "epoch": 1.3780918727915195,
-      "grad_norm": 150283.3125,
+      "grad_norm": 233520.015625,
       "learning_rate": 3.6226148409893995e-05,
-      "loss": 0.9838,
+      "loss": 1.2451,
       "step": 1950
     },
     {
       "epoch": 1.4134275618374559,
-      "grad_norm": 141787.671875,
+      "grad_norm": 416601.09375,
       "learning_rate": 3.587279151943463e-05,
-      "loss": 0.9388,
+      "loss": 1.189,
       "step": 2000
     },
     {
       "epoch": 1.4487632508833923,
-      "grad_norm": 166732.296875,
+      "grad_norm": 242072.71875,
       "learning_rate": 3.551943462897527e-05,
-      "loss": 0.9291,
+      "loss": 1.1835,
       "step": 2050
     },
     {
       "epoch": 1.4840989399293285,
-      "grad_norm": 120619.9609375,
+      "grad_norm": 165171.296875,
       "learning_rate": 3.51660777385159e-05,
-      "loss": 0.8662,
+      "loss": 1.1288,
       "step": 2100
     },
     {
       "epoch": 1.5194346289752652,
-      "grad_norm": 171583.796875,
+      "grad_norm": 178379.890625,
       "learning_rate": 3.4812720848056544e-05,
-      "loss": 0.8699,
+      "loss": 1.1147,
       "step": 2150
     },
     {
       "epoch": 1.5547703180212014,
-      "grad_norm": 144918.625,
+      "grad_norm": 189740.796875,
       "learning_rate": 3.445936395759717e-05,
-      "loss": 0.837,
+      "loss": 1.0938,
       "step": 2200
     },
     {
       "epoch": 1.5901060070671378,
-      "grad_norm": 132313.546875,
+      "grad_norm": 130276.1171875,
       "learning_rate": 3.410600706713781e-05,
-      "loss": 0.8758,
+      "loss": 1.1391,
       "step": 2250
     },
     {
       "epoch": 1.6254416961130742,
-      "grad_norm": 233730.875,
+      "grad_norm": 183882.953125,
       "learning_rate": 3.3752650176678445e-05,
-      "loss": 0.8976,
+      "loss": 1.154,
       "step": 2300
     },
     {
       "epoch": 1.6607773851590106,
-      "grad_norm": 145260.0625,
+      "grad_norm": 161704.109375,
       "learning_rate": 3.3399293286219085e-05,
-      "loss": 0.9026,
+      "loss": 1.1451,
       "step": 2350
     },
     {
       "epoch": 1.696113074204947,
-      "grad_norm": 139912.078125,
+      "grad_norm": 191092.828125,
       "learning_rate": 3.304593639575972e-05,
-      "loss": 0.9291,
+      "loss": 1.189,
       "step": 2400
     },
     {
       "epoch": 1.7314487632508833,
-      "grad_norm": 148225.671875,
+      "grad_norm": 628473.625,
       "learning_rate": 3.269257950530035e-05,
-      "loss": 0.8501,
+      "loss": 1.1158,
       "step": 2450
     },
     {
       "epoch": 1.76678445229682,
-      "grad_norm": 152729.78125,
+      "grad_norm": 170168.171875,
       "learning_rate": 3.233922261484099e-05,
-      "loss": 0.8378,
+      "loss": 1.112,
       "step": 2500
     },
     {
       "epoch": 1.802120141342756,
-      "grad_norm": 144360.765625,
+      "grad_norm": 176820.375,
       "learning_rate": 3.198586572438163e-05,
-      "loss": 0.8709,
+      "loss": 1.1347,
       "step": 2550
     },
     {
       "epoch": 1.8374558303886925,
-      "grad_norm": 181913.359375,
+      "grad_norm": 184860.6875,
       "learning_rate": 3.163250883392227e-05,
-      "loss": 0.8747,
+      "loss": 1.1363,
       "step": 2600
     },
     {
       "epoch": 1.872791519434629,
-      "grad_norm": 281330.1875,
+      "grad_norm": 188585.625,
       "learning_rate": 3.1279151943462894e-05,
-      "loss": 0.818,
+      "loss": 1.076,
       "step": 2650
     },
     {
       "epoch": 1.9081272084805654,
-      "grad_norm": 182152.09375,
+      "grad_norm": 210377.1875,
       "learning_rate": 3.0925795053003535e-05,
-      "loss": 0.7932,
+      "loss": 1.0597,
       "step": 2700
     },
     {
       "epoch": 1.9434628975265018,
-      "grad_norm": 206569.15625,
+      "grad_norm": 190550.59375,
       "learning_rate": 3.057243816254417e-05,
-      "loss": 0.851,
+      "loss": 1.117,
       "step": 2750
     },
     {
       "epoch": 1.978798586572438,
-      "grad_norm": 214797.078125,
+      "grad_norm": 252235.53125,
       "learning_rate": 3.021908127208481e-05,
-      "loss": 0.8445,
+      "loss": 1.1122,
       "step": 2800
     },
     {
       "epoch": 2.0,
-      "eval_bertscore_f1": 0.9522738781490243,
-      "eval_bleu": 0.4393754784053474,
-      "eval_loss": 0.6055822968482971,
-      "eval_meteor": 0.6140149788880958,
-      "eval_rouge1": 0.7802981035017684,
-      "eval_rouge2": 0.6476425566779656,
-      "eval_runtime": 432.9039,
-      "eval_samples_per_second": 14.932,
-      "eval_steps_per_second": 0.933,
+      "eval_bertscore_f1": 0.9582845463629553,
+      "eval_bleu": 0.49983486433792085,
+      "eval_loss": 0.4666912853717804,
+      "eval_meteor": 0.669274930453781,
+      "eval_rouge1": 0.8072540196527545,
+      "eval_rouge2": 0.70742760483214,
+      "eval_runtime": 458.6727,
+      "eval_samples_per_second": 14.093,
+      "eval_steps_per_second": 0.881,
       "step": 2830
     },
     {
       "epoch": 2.0141342756183747,
-      "grad_norm": 171307.796875,
+      "grad_norm": 257279.75,
       "learning_rate": 2.9865724381625443e-05,
-      "loss": 0.8134,
+      "loss": 1.0781,
       "step": 2850
     },
     {
       "epoch": 2.049469964664311,
-      "grad_norm": 165611.75,
+      "grad_norm": 189275.46875,
       "learning_rate": 2.9512367491166083e-05,
-      "loss": 0.8406,
+      "loss": 1.1029,
       "step": 2900
     },
     {
       "epoch": 2.0848056537102475,
-      "grad_norm": 214883.765625,
+      "grad_norm": 194805.359375,
       "learning_rate": 2.9159010600706717e-05,
-      "loss": 0.7745,
+      "loss": 1.0469,
       "step": 2950
     },
     {
       "epoch": 2.1201413427561837,
-      "grad_norm": 137894.703125,
+      "grad_norm": 201691.625,
       "learning_rate": 2.8805653710247347e-05,
-      "loss": 0.7598,
+      "loss": 1.0245,
       "step": 3000
     },
     {
       "epoch": 2.1554770318021204,
-      "grad_norm": 159089.390625,
+      "grad_norm": 235752.0,
       "learning_rate": 2.8452296819787988e-05,
-      "loss": 0.7726,
+      "loss": 1.0363,
       "step": 3050
     },
     {
       "epoch": 2.1908127208480566,
-      "grad_norm": 165668.390625,
+      "grad_norm": 173017.640625,
       "learning_rate": 2.809893992932862e-05,
-      "loss": 0.7863,
+      "loss": 1.0579,
       "step": 3100
     },
     {
       "epoch": 2.2261484098939928,
-      "grad_norm": 138427.421875,
+      "grad_norm": 184098.203125,
       "learning_rate": 2.7745583038869262e-05,
-      "loss": 0.8514,
+      "loss": 1.1153,
       "step": 3150
     },
     {
       "epoch": 2.2614840989399294,
-      "grad_norm": 187050.6875,
+      "grad_norm": 240071.0,
       "learning_rate": 2.7392226148409892e-05,
-      "loss": 0.8563,
+      "loss": 1.1175,
       "step": 3200
     },
     {
       "epoch": 2.2968197879858656,
-      "grad_norm": 173597.140625,
+      "grad_norm": 243769.109375,
       "learning_rate": 2.7038869257950533e-05,
-      "loss": 0.754,
+      "loss": 1.027,
       "step": 3250
     },
     {
       "epoch": 2.3321554770318023,
-      "grad_norm": 178735.671875,
+      "grad_norm": 464112.78125,
       "learning_rate": 2.6685512367491167e-05,
-      "loss": 0.8188,
+      "loss": 1.0759,
       "step": 3300
     },
     {
       "epoch": 2.3674911660777385,
-      "grad_norm": 165290.921875,
+      "grad_norm": 169012.796875,
       "learning_rate": 2.6332155477031807e-05,
-      "loss": 0.8194,
+      "loss": 1.0839,
       "step": 3350
     },
     {
       "epoch": 2.402826855123675,
-      "grad_norm": 142451.59375,
+      "grad_norm": 166011.828125,
       "learning_rate": 2.597879858657244e-05,
-      "loss": 0.7399,
+      "loss": 1.0121,
       "step": 3400
     },
     {
       "epoch": 2.4381625441696113,
-      "grad_norm": 159937.9375,
+      "grad_norm": 158634.28125,
       "learning_rate": 2.562544169611307e-05,
-      "loss": 0.7892,
+      "loss": 1.06,
       "step": 3450
     },
     {
       "epoch": 2.4734982332155475,
-      "grad_norm": 161491.390625,
+      "grad_norm": 162840.34375,
       "learning_rate": 2.5272084805653712e-05,
-      "loss": 0.7216,
+      "loss": 0.9904,
       "step": 3500
     },
     {
       "epoch": 2.508833922261484,
-      "grad_norm": 142754.046875,
+      "grad_norm": 162062.4375,
       "learning_rate": 2.491872791519435e-05,
-      "loss": 0.7567,
+      "loss": 1.0173,
       "step": 3550
     },
     {
       "epoch": 2.5441696113074204,
-      "grad_norm": 172165.828125,
+      "grad_norm": 193363.75,
       "learning_rate": 2.4565371024734986e-05,
-      "loss": 0.7714,
+      "loss": 1.0344,
       "step": 3600
     },
     {
       "epoch": 2.579505300353357,
-      "grad_norm": 149934.609375,
+      "grad_norm": 190665.90625,
       "learning_rate": 2.421201413427562e-05,
-      "loss": 0.7358,
+      "loss": 1.0007,
       "step": 3650
     },
     {
       "epoch": 2.614840989399293,
-      "grad_norm": 158452.328125,
+      "grad_norm": 175752.640625,
       "learning_rate": 2.3858657243816253e-05,
-      "loss": 0.7984,
+      "loss": 1.0578,
       "step": 3700
     },
     {
       "epoch": 2.65017667844523,
-      "grad_norm": 160516.578125,
+      "grad_norm": 175319.953125,
       "learning_rate": 2.350530035335689e-05,
-      "loss": 0.7578,
+      "loss": 1.0294,
       "step": 3750
     },
     {
       "epoch": 2.685512367491166,
-      "grad_norm": 154526.3125,
+      "grad_norm": 169006.34375,
       "learning_rate": 2.3151943462897528e-05,
-      "loss": 0.7347,
+      "loss": 1.0066,
       "step": 3800
     },
     {
       "epoch": 2.7208480565371023,
-      "grad_norm": 131827.875,
+      "grad_norm": 192368.734375,
       "learning_rate": 2.2798586572438165e-05,
-      "loss": 0.6999,
+      "loss": 0.9715,
       "step": 3850
     },
     {
       "epoch": 2.756183745583039,
-      "grad_norm": 152252.828125,
+      "grad_norm": 167620.265625,
       "learning_rate": 2.24452296819788e-05,
-      "loss": 0.699,
+      "loss": 0.9757,
       "step": 3900
     },
     {
       "epoch": 2.791519434628975,
-      "grad_norm": 200477.015625,
+      "grad_norm": 216407.484375,
       "learning_rate": 2.2091872791519436e-05,
-      "loss": 0.7385,
+      "loss": 1.0102,
       "step": 3950
     },
     {
       "epoch": 2.8268551236749118,
-      "grad_norm": 185108.78125,
+      "grad_norm": 169598.328125,
       "learning_rate": 2.1738515901060073e-05,
-      "loss": 0.7255,
+      "loss": 1.0011,
       "step": 4000
     },
     {
       "epoch": 2.862190812720848,
-      "grad_norm": 164106.140625,
+      "grad_norm": 447852.53125,
       "learning_rate": 2.138515901060071e-05,
-      "loss": 0.7056,
+      "loss": 0.9803,
       "step": 4050
     },
     {
       "epoch": 2.8975265017667846,
-      "grad_norm": 150298.546875,
+      "grad_norm": 183687.234375,
       "learning_rate": 2.1031802120141344e-05,
-      "loss": 0.7292,
+      "loss": 1.0082,
       "step": 4100
     },
     {
       "epoch": 2.932862190812721,
-      "grad_norm": 143401.09375,
+      "grad_norm": 150918.265625,
       "learning_rate": 2.067844522968198e-05,
-      "loss": 0.7063,
+      "loss": 0.9742,
       "step": 4150
     },
     {
       "epoch": 2.968197879858657,
-      "grad_norm": 143880.890625,
+      "grad_norm": 174224.703125,
       "learning_rate": 2.0325088339222618e-05,
-      "loss": 0.7133,
+      "loss": 0.9893,
       "step": 4200
     },
     {
       "epoch": 3.0,
-      "eval_bertscore_f1": 0.9557588405925597,
-      "eval_bleu": 0.4786301159708564,
-      "eval_loss": 0.5168418288230896,
-      "eval_meteor": 0.647062164805781,
-      "eval_rouge1": 0.7974368056119322,
-      "eval_rouge2": 0.6847280282295476,
-      "eval_runtime": 430.3015,
-      "eval_samples_per_second": 15.022,
-      "eval_steps_per_second": 0.939,
+      "eval_bertscore_f1": 0.9605225993430998,
+      "eval_bleu": 0.5316084480437336,
+      "eval_loss": 0.40197497606277466,
+      "eval_meteor": 0.6927241095474356,
+      "eval_rouge1": 0.8163514632828656,
+      "eval_rouge2": 0.7309898153084421,
+      "eval_runtime": 453.3627,
+      "eval_samples_per_second": 14.258,
+      "eval_steps_per_second": 0.891,
       "step": 4245
     },
     {
       "epoch": 3.0035335689045937,
-      "grad_norm": 137463.609375,
+      "grad_norm": 155634.3125,
       "learning_rate": 1.997173144876325e-05,
-      "loss": 0.6865,
+      "loss": 0.9659,
       "step": 4250
     },
     {
       "epoch": 3.03886925795053,
-      "grad_norm": 161362.03125,
+      "grad_norm": 441207.90625,
       "learning_rate": 1.9618374558303885e-05,
-      "loss": 0.7003,
+      "loss": 0.9841,
       "step": 4300
     },
     {
       "epoch": 3.0742049469964665,
-      "grad_norm": 152229.84375,
+      "grad_norm": 171151.390625,
       "learning_rate": 1.9265017667844522e-05,
-      "loss": 0.6962,
+      "loss": 0.9685,
       "step": 4350
     },
     {
       "epoch": 3.1095406360424027,
-      "grad_norm": 154452.28125,
+      "grad_norm": 178025.75,
       "learning_rate": 1.891166077738516e-05,
-      "loss": 0.7069,
+      "loss": 0.9823,
       "step": 4400
     },
     {
       "epoch": 3.1448763250883394,
-      "grad_norm": 353214.9375,
+      "grad_norm": 194535.046875,
       "learning_rate": 1.8558303886925797e-05,
-      "loss": 0.6871,
+      "loss": 0.9673,
       "step": 4450
     },
     {
       "epoch": 3.1802120141342756,
-      "grad_norm": 158643.328125,
+      "grad_norm": 161927.78125,
       "learning_rate": 1.8204946996466434e-05,
-      "loss": 0.7152,
+      "loss": 0.9857,
       "step": 4500
     },
     {
       "epoch": 3.215547703180212,
-      "grad_norm": 155058.890625,
+      "grad_norm": 162012.375,
       "learning_rate": 1.7851590106007067e-05,
-      "loss": 0.721,
+      "loss": 0.9981,
       "step": 4550
     },
     {
       "epoch": 3.2508833922261484,
-      "grad_norm": 144921.78125,
+      "grad_norm": 189190.890625,
       "learning_rate": 1.7498233215547705e-05,
-      "loss": 0.7333,
+      "loss": 0.9984,
       "step": 4600
     },
     {
       "epoch": 3.2862190812720846,
-      "grad_norm": 141372.734375,
+      "grad_norm": 160611.28125,
       "learning_rate": 1.714487632508834e-05,
-      "loss": 0.6832,
+      "loss": 0.962,
       "step": 4650
     },
     {
       "epoch": 3.3215547703180213,
-      "grad_norm": 137484.8125,
+      "grad_norm": 160583.796875,
       "learning_rate": 1.679151943462898e-05,
-      "loss": 0.6588,
+      "loss": 0.9382,
       "step": 4700
     },
     {
       "epoch": 3.3568904593639575,
-      "grad_norm": 190962.765625,
+      "grad_norm": 241355.578125,
       "learning_rate": 1.643816254416961e-05,
-      "loss": 0.7555,
+      "loss": 1.0251,
       "step": 4750
     },
     {
       "epoch": 3.392226148409894,
-      "grad_norm": 167470.15625,
+      "grad_norm": 186881.078125,
       "learning_rate": 1.6084805653710246e-05,
-      "loss": 0.7017,
+      "loss": 0.9726,
       "step": 4800
     },
     {
       "epoch": 3.4275618374558303,
-      "grad_norm": 169313.171875,
+      "grad_norm": 189630.578125,
       "learning_rate": 1.5731448763250883e-05,
-      "loss": 0.6365,
+      "loss": 0.912,
       "step": 4850
     },
     {
       "epoch": 3.462897526501767,
-      "grad_norm": 186077.046875,
+      "grad_norm": 427752.875,
       "learning_rate": 1.537809187279152e-05,
-      "loss": 0.7116,
+      "loss": 0.9862,
       "step": 4900
     },
     {
       "epoch": 3.498233215547703,
-      "grad_norm": 298409.65625,
+      "grad_norm": 221891.625,
       "learning_rate": 1.5024734982332156e-05,
-      "loss": 0.7189,
+      "loss": 0.9943,
       "step": 4950
     },
     {
       "epoch": 3.53356890459364,
-      "grad_norm": 160374.9375,
+      "grad_norm": 167730.796875,
       "learning_rate": 1.4671378091872793e-05,
-      "loss": 0.6522,
+      "loss": 0.926,
       "step": 5000
     },
     {
       "epoch": 3.568904593639576,
-      "grad_norm": 162222.75,
+      "grad_norm": 193771.1875,
       "learning_rate": 1.4318021201413428e-05,
-      "loss": 0.7109,
+      "loss": 0.993,
       "step": 5050
     },
     {
       "epoch": 3.604240282685512,
-      "grad_norm": 321612.25,
+      "grad_norm": 187069.09375,
       "learning_rate": 1.3964664310954066e-05,
-      "loss": 0.7548,
+      "loss": 1.0317,
       "step": 5100
     },
     {
       "epoch": 3.639575971731449,
-      "grad_norm": 158497.75,
+      "grad_norm": 188425.546875,
       "learning_rate": 1.3611307420494701e-05,
-      "loss": 0.7098,
+      "loss": 0.9729,
       "step": 5150
     },
     {
       "epoch": 3.674911660777385,
-      "grad_norm": 143490.0,
+      "grad_norm": 163470.703125,
       "learning_rate": 1.3257950530035338e-05,
-      "loss": 0.6907,
+      "loss": 0.9688,
       "step": 5200
     },
     {
       "epoch": 3.7102473498233217,
-      "grad_norm": 161849.75,
+      "grad_norm": 187701.09375,
       "learning_rate": 1.2904593639575973e-05,
-      "loss": 0.6155,
+      "loss": 0.9105,
       "step": 5250
     },
     {
       "epoch": 3.745583038869258,
-      "grad_norm": 139143.828125,
+      "grad_norm": 144018.484375,
       "learning_rate": 1.2551236749116607e-05,
-      "loss": 0.6978,
+      "loss": 0.9754,
       "step": 5300
     },
     {
       "epoch": 3.7809187279151946,
-      "grad_norm": 160687.375,
+      "grad_norm": 154232.65625,
       "learning_rate": 1.2197879858657244e-05,
-      "loss": 0.6624,
+      "loss": 0.9472,
       "step": 5350
     },
     {
       "epoch": 3.8162544169611308,
-      "grad_norm": 178367.625,
+      "grad_norm": 170624.640625,
       "learning_rate": 1.184452296819788e-05,
-      "loss": 0.7098,
+      "loss": 0.9886,
       "step": 5400
     },
     {
       "epoch": 3.851590106007067,
-      "grad_norm": 166319.734375,
+      "grad_norm": 1157734.625,
       "learning_rate": 1.1491166077738517e-05,
-      "loss": 0.6305,
+      "loss": 0.915,
       "step": 5450
     },
     {
       "epoch": 3.8869257950530036,
-      "grad_norm": 182385.0625,
+      "grad_norm": 179182.0,
       "learning_rate": 1.1137809187279152e-05,
-      "loss": 0.6361,
+      "loss": 0.9216,
       "step": 5500
     },
     {
       "epoch": 3.92226148409894,
-      "grad_norm": 146347.28125,
+      "grad_norm": 872790.1875,
       "learning_rate": 1.078445229681979e-05,
-      "loss": 0.6609,
+      "loss": 0.9351,
       "step": 5550
     },
     {
       "epoch": 3.9575971731448765,
-      "grad_norm": 167555.078125,
+      "grad_norm": 269436.21875,
       "learning_rate": 1.0431095406360425e-05,
-      "loss": 0.6351,
+      "loss": 0.9223,
       "step": 5600
     },
     {
       "epoch": 3.9929328621908127,
-      "grad_norm": 230608.890625,
+      "grad_norm": 197819.828125,
       "learning_rate": 1.0077738515901062e-05,
-      "loss": 0.6894,
+      "loss": 0.9707,
       "step": 5650
     },
     {
       "epoch": 4.0,
-      "eval_bertscore_f1": 0.9578082369695796,
-      "eval_bleu": 0.4998358619352557,
-      "eval_loss": 0.4754031300544739,
-      "eval_meteor": 0.6658311501198411,
-      "eval_rouge1": 0.8054192335793489,
-      "eval_rouge2": 0.7038693380378069,
-      "eval_runtime": 428.7031,
-      "eval_samples_per_second": 15.078,
-      "eval_steps_per_second": 0.942,
+      "eval_bertscore_f1": 0.9621182246373432,
+      "eval_bleu": 0.5466384314770388,
+      "eval_loss": 0.3729407489299774,
+      "eval_meteor": 0.706948756410978,
+      "eval_rouge1": 0.8231910064206467,
+      "eval_rouge2": 0.7462398210944181,
+      "eval_runtime": 455.8029,
+      "eval_samples_per_second": 14.182,
+      "eval_steps_per_second": 0.886,
       "step": 5660
     },
     {
       "epoch": 4.028268551236749,
-      "grad_norm": 215650.8125,
+      "grad_norm": 211690.5,
       "learning_rate": 9.724381625441696e-06,
-      "loss": 0.6677,
+      "loss": 0.9475,
       "step": 5700
     },
     {
       "epoch": 4.063604240282685,
-      "grad_norm": 166426.796875,
+      "grad_norm": 1410444.25,
       "learning_rate": 9.371024734982333e-06,
-      "loss": 0.6579,
+      "loss": 0.943,
       "step": 5750
     },
     {
       "epoch": 4.098939929328622,
-      "grad_norm": 133948.8125,
+      "grad_norm": 158585.84375,
       "learning_rate": 9.017667844522968e-06,
-      "loss": 0.6721,
+      "loss": 0.9525,
       "step": 5800
     },
     {
       "epoch": 4.134275618374558,
-      "grad_norm": 121285.546875,
+      "grad_norm": 127129.171875,
       "learning_rate": 8.664310954063605e-06,
-      "loss": 0.6635,
+      "loss": 0.9423,
       "step": 5850
     },
     {
       "epoch": 4.169611307420495,
-      "grad_norm": 151956.71875,
+      "grad_norm": 158647.09375,
       "learning_rate": 8.31095406360424e-06,
-      "loss": 0.6525,
+      "loss": 0.9317,
       "step": 5900
     },
     {
       "epoch": 4.204946996466431,
-      "grad_norm": 133390.5625,
+      "grad_norm": 151794.578125,
       "learning_rate": 7.957597173144876e-06,
-      "loss": 0.6485,
+      "loss": 0.9252,
       "step": 5950
     },
     {
       "epoch": 4.240282685512367,
-      "grad_norm": 139297.71875,
+      "grad_norm": 134228.625,
       "learning_rate": 7.6042402826855124e-06,
-      "loss": 0.6646,
+      "loss": 0.9466,
       "step": 6000
     },
     {
       "epoch": 4.275618374558304,
-      "grad_norm": 163106.578125,
+      "grad_norm": 153596.578125,
       "learning_rate": 7.250883392226149e-06,
-      "loss": 0.6336,
+      "loss": 0.9165,
       "step": 6050
     },
     {
       "epoch": 4.310954063604241,
-      "grad_norm": 218559.828125,
+      "grad_norm": 205703.125,
       "learning_rate": 6.897526501766785e-06,
-      "loss": 0.6386,
+      "loss": 0.9143,
       "step": 6100
     },
     {
       "epoch": 4.3462897526501765,
-      "grad_norm": 132456.375,
+      "grad_norm": 133143.015625,
       "learning_rate": 6.544169611307421e-06,
-      "loss": 0.7033,
+      "loss": 0.9926,
       "step": 6150
     },
     {
       "epoch": 4.381625441696113,
-      "grad_norm": 140776.359375,
+      "grad_norm": 131342.953125,
       "learning_rate": 6.190812720848057e-06,
-      "loss": 0.6751,
+      "loss": 0.9561,
       "step": 6200
     },
     {
       "epoch": 4.41696113074205,
-      "grad_norm": 157513.3125,
+      "grad_norm": 167654.875,
       "learning_rate": 5.837455830388693e-06,
-      "loss": 0.6483,
+      "loss": 0.9183,
       "step": 6250
     },
     {
       "epoch": 4.4522968197879855,
-      "grad_norm": 130282.9765625,
+      "grad_norm": 160178.71875,
       "learning_rate": 5.484098939929329e-06,
-      "loss": 0.663,
+      "loss": 0.9432,
       "step": 6300
     },
     {
       "epoch": 4.487632508833922,
-      "grad_norm": 142280.046875,
+      "grad_norm": 169846.84375,
       "learning_rate": 5.1307420494699655e-06,
-      "loss": 0.683,
+      "loss": 0.9598,
       "step": 6350
     },
     {
       "epoch": 4.522968197879859,
-      "grad_norm": 155985.96875,
+      "grad_norm": 159304.1875,
       "learning_rate": 4.777385159010601e-06,
-      "loss": 0.6307,
+      "loss": 0.9163,
       "step": 6400
     },
     {
       "epoch": 4.5583038869257955,
-      "grad_norm": 147858.40625,
+      "grad_norm": 159722.328125,
       "learning_rate": 4.424028268551237e-06,
-      "loss": 0.6431,
+      "loss": 0.9269,
       "step": 6450
     },
     {
       "epoch": 4.593639575971731,
-      "grad_norm": 152688.6875,
+      "grad_norm": 134695.21875,
       "learning_rate": 4.070671378091873e-06,
-      "loss": 0.6788,
+      "loss": 0.9521,
       "step": 6500
     },
     {
       "epoch": 4.628975265017668,
-      "grad_norm": 203074.90625,
+      "grad_norm": 214522.46875,
       "learning_rate": 3.717314487632509e-06,
-      "loss": 0.6703,
+      "loss": 0.9487,
       "step": 6550
     },
     {
       "epoch": 4.6643109540636045,
-      "grad_norm": 196964.53125,
+      "grad_norm": 277130.375,
       "learning_rate": 3.363957597173145e-06,
-      "loss": 0.6619,
+      "loss": 0.9512,
       "step": 6600
     },
     {
       "epoch": 4.69964664310954,
-      "grad_norm": 199824.15625,
+      "grad_norm": 229889.265625,
       "learning_rate": 3.010600706713781e-06,
-      "loss": 0.6759,
+      "loss": 0.9568,
       "step": 6650
     },
     {
       "epoch": 4.734982332155477,
-      "grad_norm": 157021.34375,
+      "grad_norm": 160132.65625,
       "learning_rate": 2.6572438162544172e-06,
-      "loss": 0.6314,
+      "loss": 0.9169,
       "step": 6700
     },
     {
       "epoch": 4.770318021201414,
-      "grad_norm": 145809.40625,
+      "grad_norm": 153591.515625,
       "learning_rate": 2.303886925795053e-06,
-      "loss": 0.6452,
+      "loss": 0.9247,
       "step": 6750
     },
     {
       "epoch": 4.80565371024735,
-      "grad_norm": 143137.765625,
+      "grad_norm": 198496.34375,
       "learning_rate": 1.9505300353356893e-06,
-      "loss": 0.6813,
+      "loss": 0.963,
       "step": 6800
     },
     {
       "epoch": 4.840989399293286,
-      "grad_norm": 195221.671875,
+      "grad_norm": 217991.0625,
       "learning_rate": 1.5971731448763252e-06,
-      "loss": 0.6832,
+      "loss": 0.9591,
       "step": 6850
     },
     {
       "epoch": 4.876325088339223,
-      "grad_norm": 152621.546875,
+      "grad_norm": 170718.59375,
       "learning_rate": 1.2438162544169612e-06,
-      "loss": 0.6295,
+      "loss": 0.9099,
       "step": 6900
     },
     {
       "epoch": 4.911660777385159,
-      "grad_norm": 140083.28125,
+      "grad_norm": 138367.578125,
       "learning_rate": 8.904593639575972e-07,
-      "loss": 0.6375,
+      "loss": 0.9267,
       "step": 6950
     },
     {
       "epoch": 4.946996466431095,
-      "grad_norm": 140951.5625,
+      "grad_norm": 170947.796875,
       "learning_rate": 5.371024734982332e-07,
-      "loss": 0.6901,
+      "loss": 0.9668,
       "step": 7000
     },
     {
       "epoch": 4.982332155477032,
-      "grad_norm": 159365.84375,
+      "grad_norm": 127770.6484375,
       "learning_rate": 1.8374558303886926e-07,
-      "loss": 0.6547,
+      "loss": 0.9397,
       "step": 7050
     },
     {
       "epoch": 5.0,
-      "eval_bertscore_f1": 0.9583620159785347,
-      "eval_bleu": 0.5038629414956816,
-      "eval_loss": 0.4653092622756958,
-      "eval_meteor": 0.6703550407484931,
-      "eval_rouge1": 0.8078007642628657,
-      "eval_rouge2": 0.7083775833633327,
-      "eval_runtime": 428.55,
-      "eval_samples_per_second": 15.083,
-      "eval_steps_per_second": 0.943,
+      "eval_bertscore_f1": 0.9626803387159316,
+      "eval_bleu": 0.5519723913292562,
+      "eval_loss": 0.36424726247787476,
+      "eval_meteor": 0.7112807593232308,
+      "eval_rouge1": 0.8253264091986198,
+      "eval_rouge2": 0.750184658379115,
+      "eval_runtime": 454.9282,
+      "eval_samples_per_second": 14.209,
+      "eval_steps_per_second": 0.888,
       "step": 7075
     }
   ],

checkpoint-7075/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c763ff9d225f7827b3812ee2353510a3bce94f41c8e401d9ea1bd41910731347
+oid sha256:d2165abc25380192b370a5e4b0328e2b882051447b7b8cc7d6c4043ff001ffa8
 size 5432