Training in progress, step 1200, checkpoint

checkpoint-1200/adapter_config.json

@@ -23,13 +23,13 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
-    "down_proj",
-    "v_proj",
-    "up_proj",
     "k_proj",
     "q_proj",
-    "gate_proj"
+    "v_proj",
+    "up_proj",
+    "gate_proj",
+    "down_proj",
+    "o_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

checkpoint-1200/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:00688c977a9d98c2d46875196b3600b45079b73b4b11c14a28b8df6440c22e96
+oid sha256:167d74b0bd171e2ea611e97e8f9726b37ee47bc5ded33a24652b7db2c68c35ba
 size 161533192

checkpoint-1200/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:402aa69fbc1e4f3e6e16a8465d1f4f00e980e91b642b4deb65f13e47a44ef447
+oid sha256:693ec9104c7fbd1f3baf5b80857ae6988d943e81fbfcf6a1205fa428081cdb3a
 size 323290986

checkpoint-1200/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:580dbc932d930ab938d71a62b495d495ead3b70c32d4f6ad57f7f3855c87d651
+oid sha256:5253d9c7b8c626f9223677d46fd02e270dfcbebc8ebb5d5c62ccc97d43898b17
 size 14244

checkpoint-1200/trainer_state.json

@@ -10,1034 +10,1034 @@
   "log_history": [
     {
       "epoch": 0.010604453870625663,
-      "grad_norm": 2.276311159133911,
+      "grad_norm": 2.4981472492218018,
       "learning_rate": 4.982325910215624e-05,
-      "loss": 1.9038,
+      "loss": 1.6893,
       "step": 10
     },
     {
       "epoch": 0.021208907741251327,
-      "grad_norm": 0.7678110003471375,
+      "grad_norm": 0.6994958519935608,
       "learning_rate": 4.964651820431248e-05,
-      "loss": 0.997,
+      "loss": 0.8541,
       "step": 20
     },
     {
       "epoch": 0.031813361611876985,
-      "grad_norm": 0.6726565361022949,
+      "grad_norm": 0.7477335333824158,
       "learning_rate": 4.946977730646872e-05,
-      "loss": 0.8021,
+      "loss": 0.7168,
       "step": 30
     },
     {
       "epoch": 0.042417815482502653,
-      "grad_norm": 0.7318501472473145,
+      "grad_norm": 0.6651986241340637,
       "learning_rate": 4.929303640862496e-05,
-      "loss": 0.7122,
+      "loss": 0.6438,
       "step": 40
     },
     {
       "epoch": 0.053022269353128315,
-      "grad_norm": 0.6708294749259949,
+      "grad_norm": 0.683933436870575,
       "learning_rate": 4.91162955107812e-05,
-      "loss": 0.6783,
+      "loss": 0.6729,
       "step": 50
     },
     {
       "epoch": 0.053022269353128315,
-      "eval_loss": 0.8075584173202515,
-      "eval_runtime": 11.7244,
-      "eval_samples_per_second": 40.94,
-      "eval_steps_per_second": 2.559,
+      "eval_loss": 0.7466596961021423,
+      "eval_runtime": 12.3447,
+      "eval_samples_per_second": 38.883,
+      "eval_steps_per_second": 2.43,
       "step": 50
     },
     {
       "epoch": 0.06362672322375397,
-      "grad_norm": 0.7024317979812622,
+      "grad_norm": 0.6808643341064453,
       "learning_rate": 4.893955461293744e-05,
-      "loss": 0.6244,
+      "loss": 0.6054,
       "step": 60
     },
     {
       "epoch": 0.07423117709437964,
-      "grad_norm": 0.6959078907966614,
+      "grad_norm": 0.6630373001098633,
       "learning_rate": 4.876281371509368e-05,
-      "loss": 0.6117,
+      "loss": 0.582,
       "step": 70
     },
     {
       "epoch": 0.08483563096500531,
-      "grad_norm": 0.7824214696884155,
+      "grad_norm": 0.5939441323280334,
       "learning_rate": 4.858607281724991e-05,
-      "loss": 0.5989,
+      "loss": 0.564,
       "step": 80
     },
     {
       "epoch": 0.09544008483563096,
-      "grad_norm": 0.7792642116546631,
+      "grad_norm": 0.756606936454773,
       "learning_rate": 4.8409331919406155e-05,
-      "loss": 0.6033,
+      "loss": 0.573,
       "step": 90
     },
     {
       "epoch": 0.10604453870625663,
-      "grad_norm": 0.7486705780029297,
+      "grad_norm": 0.7286397218704224,
       "learning_rate": 4.823259102156239e-05,
-      "loss": 0.5904,
+      "loss": 0.5886,
       "step": 100
     },
     {
       "epoch": 0.10604453870625663,
-      "eval_loss": 0.7230511903762817,
-      "eval_runtime": 11.7589,
-      "eval_samples_per_second": 40.82,
-      "eval_steps_per_second": 2.551,
+      "eval_loss": 0.6717663407325745,
+      "eval_runtime": 12.4167,
+      "eval_samples_per_second": 38.658,
+      "eval_steps_per_second": 2.416,
       "step": 100
     },
     {
       "epoch": 0.11664899257688228,
-      "grad_norm": 0.8214134573936462,
+      "grad_norm": 0.7108235359191895,
       "learning_rate": 4.8055850123718633e-05,
-      "loss": 0.5778,
+      "loss": 0.5759,
       "step": 110
     },
     {
       "epoch": 0.12725344644750794,
-      "grad_norm": 0.7937195897102356,
+      "grad_norm": 0.7061440944671631,
       "learning_rate": 4.7879109225874866e-05,
-      "loss": 0.5748,
+      "loss": 0.539,
       "step": 120
     },
     {
       "epoch": 0.1378579003181336,
-      "grad_norm": 0.7781085968017578,
+      "grad_norm": 0.8537445664405823,
       "learning_rate": 4.770236832803111e-05,
-      "loss": 0.5469,
+      "loss": 0.5294,
       "step": 130
     },
     {
       "epoch": 0.14846235418875928,
-      "grad_norm": 0.695203423500061,
+      "grad_norm": 0.740597128868103,
       "learning_rate": 4.7525627430187344e-05,
-      "loss": 0.5511,
+      "loss": 0.5384,
       "step": 140
     },
     {
       "epoch": 0.15906680805938495,
-      "grad_norm": 0.7824985384941101,
+      "grad_norm": 0.6992364525794983,
       "learning_rate": 4.734888653234359e-05,
-      "loss": 0.5258,
+      "loss": 0.5255,
       "step": 150
     },
     {
       "epoch": 0.15906680805938495,
-      "eval_loss": 0.6867014765739441,
-      "eval_runtime": 11.7852,
-      "eval_samples_per_second": 40.729,
-      "eval_steps_per_second": 2.546,
+      "eval_loss": 0.6394088864326477,
+      "eval_runtime": 12.37,
+      "eval_samples_per_second": 38.803,
+      "eval_steps_per_second": 2.425,
       "step": 150
     },
     {
       "epoch": 0.16967126193001061,
-      "grad_norm": 0.7392640113830566,
+      "grad_norm": 0.7426859736442566,
       "learning_rate": 4.717214563449982e-05,
-      "loss": 0.5714,
+      "loss": 0.5483,
       "step": 160
     },
     {
       "epoch": 0.18027571580063625,
-      "grad_norm": 0.8316007852554321,
+      "grad_norm": 0.7601574063301086,
       "learning_rate": 4.699540473665606e-05,
-      "loss": 0.4982,
+      "loss": 0.4993,
       "step": 170
     },
     {
       "epoch": 0.19088016967126192,
-      "grad_norm": 0.8199877738952637,
+      "grad_norm": 0.7388486862182617,
       "learning_rate": 4.681866383881231e-05,
-      "loss": 0.5092,
+      "loss": 0.5175,
       "step": 180
     },
     {
       "epoch": 0.2014846235418876,
-      "grad_norm": 0.7089737057685852,
+      "grad_norm": 0.7000683546066284,
       "learning_rate": 4.664192294096854e-05,
-      "loss": 0.4953,
+      "loss": 0.4936,
       "step": 190
     },
     {
       "epoch": 0.21208907741251326,
-      "grad_norm": 0.7906930446624756,
+      "grad_norm": 0.7684634923934937,
       "learning_rate": 4.6465182043124786e-05,
-      "loss": 0.4974,
+      "loss": 0.5049,
       "step": 200
     },
     {
       "epoch": 0.21208907741251326,
-      "eval_loss": 0.669015645980835,
-      "eval_runtime": 11.8934,
-      "eval_samples_per_second": 40.358,
-      "eval_steps_per_second": 2.522,
+      "eval_loss": 0.6269157528877258,
+      "eval_runtime": 12.3596,
+      "eval_samples_per_second": 38.836,
+      "eval_steps_per_second": 2.427,
       "step": 200
     },
     {
       "epoch": 0.22269353128313893,
-      "grad_norm": 0.9085277318954468,
+      "grad_norm": 0.7945635318756104,
       "learning_rate": 4.628844114528102e-05,
-      "loss": 0.4991,
+      "loss": 0.4957,
       "step": 210
     },
     {
       "epoch": 0.23329798515376457,
-      "grad_norm": 0.7775648832321167,
+      "grad_norm": 0.6900196671485901,
       "learning_rate": 4.6111700247437265e-05,
-      "loss": 0.5024,
+      "loss": 0.5009,
       "step": 220
     },
     {
       "epoch": 0.24390243902439024,
-      "grad_norm": 0.8324123620986938,
+      "grad_norm": 0.7506561875343323,
       "learning_rate": 4.59349593495935e-05,
-      "loss": 0.5187,
+      "loss": 0.509,
       "step": 230
     },
     {
       "epoch": 0.2545068928950159,
-      "grad_norm": 0.8002191781997681,
+      "grad_norm": 0.7476117014884949,
       "learning_rate": 4.5758218451749736e-05,
-      "loss": 0.4785,
+      "loss": 0.4755,
       "step": 240
     },
     {
       "epoch": 0.2651113467656416,
-      "grad_norm": 0.8004704117774963,
+      "grad_norm": 0.7749964594841003,
       "learning_rate": 4.5581477553905975e-05,
-      "loss": 0.5073,
+      "loss": 0.5163,
       "step": 250
     },
     {
       "epoch": 0.2651113467656416,
-      "eval_loss": 0.6479213833808899,
-      "eval_runtime": 11.882,
-      "eval_samples_per_second": 40.397,
-      "eval_steps_per_second": 2.525,
+      "eval_loss": 0.6084603667259216,
+      "eval_runtime": 12.3487,
+      "eval_samples_per_second": 38.87,
+      "eval_steps_per_second": 2.429,
       "step": 250
     },
     {
       "epoch": 0.2757158006362672,
-      "grad_norm": 0.893301248550415,
+      "grad_norm": 0.7424341440200806,
       "learning_rate": 4.5404736656062215e-05,
-      "loss": 0.477,
+      "loss": 0.4766,
       "step": 260
     },
     {
       "epoch": 0.2863202545068929,
-      "grad_norm": 0.8550834059715271,
+      "grad_norm": 0.7572893500328064,
       "learning_rate": 4.5227995758218454e-05,
-      "loss": 0.4923,
+      "loss": 0.4909,
       "step": 270
     },
     {
       "epoch": 0.29692470837751855,
-      "grad_norm": 0.8929152488708496,
+      "grad_norm": 0.7275112271308899,
       "learning_rate": 4.505125486037469e-05,
-      "loss": 0.4624,
+      "loss": 0.486,
       "step": 280
     },
     {
       "epoch": 0.3075291622481442,
-      "grad_norm": 0.9584974050521851,
+      "grad_norm": 0.7402988076210022,
       "learning_rate": 4.487451396253093e-05,
-      "loss": 0.4801,
+      "loss": 0.4891,
       "step": 290
     },
     {
       "epoch": 0.3181336161187699,
-      "grad_norm": 0.7924031615257263,
+      "grad_norm": 0.7442552447319031,
       "learning_rate": 4.469777306468717e-05,
-      "loss": 0.4942,
+      "loss": 0.4794,
       "step": 300
     },
     {
       "epoch": 0.3181336161187699,
-      "eval_loss": 0.6371316313743591,
-      "eval_runtime": 11.8575,
-      "eval_samples_per_second": 40.481,
-      "eval_steps_per_second": 2.53,
+      "eval_loss": 0.5998384952545166,
+      "eval_runtime": 12.3533,
+      "eval_samples_per_second": 38.856,
+      "eval_steps_per_second": 2.428,
       "step": 300
     },
     {
       "epoch": 0.32873806998939553,
-      "grad_norm": 0.9247742295265198,
+      "grad_norm": 0.7272119522094727,
       "learning_rate": 4.452103216684341e-05,
-      "loss": 0.5234,
+      "loss": 0.5141,
       "step": 310
     },
     {
       "epoch": 0.33934252386002123,
-      "grad_norm": 0.8617805242538452,
+      "grad_norm": 0.8105739951133728,
       "learning_rate": 4.434429126899965e-05,
-      "loss": 0.4808,
+      "loss": 0.4897,
       "step": 320
     },
     {
       "epoch": 0.34994697773064687,
-      "grad_norm": 0.8334285616874695,
+      "grad_norm": 0.7440537810325623,
       "learning_rate": 4.416755037115589e-05,
-      "loss": 0.4758,
+      "loss": 0.4721,
       "step": 330
     },
     {
       "epoch": 0.3605514316012725,
-      "grad_norm": 0.9079923629760742,
+      "grad_norm": 0.7973731756210327,
       "learning_rate": 4.399080947331213e-05,
-      "loss": 0.463,
+      "loss": 0.4643,
       "step": 340
     },
     {
       "epoch": 0.3711558854718982,
-      "grad_norm": 0.7632007002830505,
+      "grad_norm": 0.7304461598396301,
       "learning_rate": 4.381406857546837e-05,
-      "loss": 0.4828,
+      "loss": 0.4907,
       "step": 350
     },
     {
       "epoch": 0.3711558854718982,
-      "eval_loss": 0.6212183237075806,
-      "eval_runtime": 11.8497,
-      "eval_samples_per_second": 40.507,
-      "eval_steps_per_second": 2.532,
+      "eval_loss": 0.5852673649787903,
+      "eval_runtime": 12.3481,
+      "eval_samples_per_second": 38.872,
+      "eval_steps_per_second": 2.43,
       "step": 350
     },
     {
       "epoch": 0.38176033934252385,
-      "grad_norm": 0.8430343270301819,
+      "grad_norm": 0.68657386302948,
       "learning_rate": 4.3637327677624606e-05,
-      "loss": 0.4711,
+      "loss": 0.4734,
       "step": 360
     },
     {
       "epoch": 0.39236479321314954,
-      "grad_norm": 0.9495126605033875,
+      "grad_norm": 0.8211691379547119,
       "learning_rate": 4.3460586779780846e-05,
-      "loss": 0.4528,
+      "loss": 0.457,
       "step": 370
     },
     {
       "epoch": 0.4029692470837752,
-      "grad_norm": 0.8903242945671082,
+      "grad_norm": 0.7916038036346436,
       "learning_rate": 4.328384588193708e-05,
-      "loss": 0.468,
+      "loss": 0.4649,
       "step": 380
     },
     {
       "epoch": 0.4135737009544008,
-      "grad_norm": 0.9306944608688354,
+      "grad_norm": 0.7529716491699219,
       "learning_rate": 4.3107104984093324e-05,
-      "loss": 0.4299,
+      "loss": 0.4437,
       "step": 390
     },
     {
       "epoch": 0.4241781548250265,
-      "grad_norm": 0.7949309349060059,
+      "grad_norm": 0.7294536828994751,
       "learning_rate": 4.2930364086249556e-05,
-      "loss": 0.4589,
+      "loss": 0.4668,
       "step": 400
     },
     {
       "epoch": 0.4241781548250265,
-      "eval_loss": 0.6148895621299744,
-      "eval_runtime": 11.8536,
-      "eval_samples_per_second": 40.494,
-      "eval_steps_per_second": 2.531,
+      "eval_loss": 0.5796846747398376,
+      "eval_runtime": 12.3341,
+      "eval_samples_per_second": 38.916,
+      "eval_steps_per_second": 2.432,
       "step": 400
     },
     {
       "epoch": 0.43478260869565216,
-      "grad_norm": 0.8351686000823975,
+      "grad_norm": 0.7497897148132324,
       "learning_rate": 4.27536231884058e-05,
-      "loss": 0.4836,
+      "loss": 0.4944,
       "step": 410
     },
     {
       "epoch": 0.44538706256627786,
-      "grad_norm": 0.9184524416923523,
+      "grad_norm": 0.8800764083862305,
       "learning_rate": 4.2576882290562035e-05,
-      "loss": 0.4686,
+      "loss": 0.4614,
       "step": 420
     },
     {
       "epoch": 0.4559915164369035,
-      "grad_norm": 0.9610297679901123,
+      "grad_norm": 0.8081978559494019,
       "learning_rate": 4.240014139271828e-05,
-      "loss": 0.4804,
+      "loss": 0.4797,
       "step": 430
     },
     {
       "epoch": 0.46659597030752914,
-      "grad_norm": 0.8793377876281738,
+      "grad_norm": 0.7951847910881042,
       "learning_rate": 4.222340049487451e-05,
-      "loss": 0.4571,
+      "loss": 0.4628,
       "step": 440
     },
     {
       "epoch": 0.47720042417815484,
-      "grad_norm": 0.913858950138092,
+      "grad_norm": 0.7923071384429932,
       "learning_rate": 4.204665959703076e-05,
-      "loss": 0.4391,
+      "loss": 0.4445,
       "step": 450
     },
     {
       "epoch": 0.47720042417815484,
-      "eval_loss": 0.6096637845039368,
-      "eval_runtime": 11.8533,
-      "eval_samples_per_second": 40.495,
-      "eval_steps_per_second": 2.531,
+      "eval_loss": 0.5746996998786926,
+      "eval_runtime": 12.3033,
+      "eval_samples_per_second": 39.014,
+      "eval_steps_per_second": 2.438,
       "step": 450
     },
     {
       "epoch": 0.4878048780487805,
-      "grad_norm": 0.9159062504768372,
+      "grad_norm": 0.8088315725326538,
       "learning_rate": 4.186991869918699e-05,
-      "loss": 0.4294,
+      "loss": 0.4502,
       "step": 460
     },
     {
       "epoch": 0.4984093319194062,
-      "grad_norm": 0.9346657395362854,
+      "grad_norm": 0.7356102466583252,
       "learning_rate": 4.169317780134323e-05,
-      "loss": 0.433,
+      "loss": 0.4417,
       "step": 470
     },
     {
       "epoch": 0.5090137857900318,
-      "grad_norm": 0.8480401635169983,
+      "grad_norm": 0.8562701344490051,
       "learning_rate": 4.151643690349947e-05,
-      "loss": 0.4639,
+      "loss": 0.463,
       "step": 480
     },
     {
       "epoch": 0.5196182396606575,
-      "grad_norm": 0.9547140002250671,
+      "grad_norm": 0.8219832181930542,
       "learning_rate": 4.133969600565571e-05,
-      "loss": 0.4173,
+      "loss": 0.4217,
       "step": 490
     },
     {
       "epoch": 0.5302226935312832,
-      "grad_norm": 0.8660142421722412,
+      "grad_norm": 0.7816707491874695,
       "learning_rate": 4.116295510781195e-05,
-      "loss": 0.4644,
+      "loss": 0.4802,
       "step": 500
     },
     {
       "epoch": 0.5302226935312832,
-      "eval_loss": 0.6053425073623657,
-      "eval_runtime": 11.8341,
-      "eval_samples_per_second": 40.561,
-      "eval_steps_per_second": 2.535,
+      "eval_loss": 0.5699853301048279,
+      "eval_runtime": 12.2761,
+      "eval_samples_per_second": 39.1,
+      "eval_steps_per_second": 2.444,
       "step": 500
     },
     {
       "epoch": 0.5408271474019088,
-      "grad_norm": 0.9299489855766296,
+      "grad_norm": 0.7816687226295471,
       "learning_rate": 4.098621420996819e-05,
-      "loss": 0.4488,
+      "loss": 0.4503,
       "step": 510
     },
     {
       "epoch": 0.5514316012725344,
-      "grad_norm": 0.9637112021446228,
+      "grad_norm": 0.8369717597961426,
       "learning_rate": 4.080947331212443e-05,
-      "loss": 0.4197,
+      "loss": 0.439,
       "step": 520
     },
     {
       "epoch": 0.5620360551431601,
-      "grad_norm": 0.9725180864334106,
+      "grad_norm": 0.902275025844574,
       "learning_rate": 4.0632732414280666e-05,
-      "loss": 0.4308,
+      "loss": 0.4406,
       "step": 530
     },
     {
       "epoch": 0.5726405090137858,
-      "grad_norm": 1.095697045326233,
+      "grad_norm": 0.8947370648384094,
       "learning_rate": 4.0455991516436905e-05,
-      "loss": 0.4473,
+      "loss": 0.4698,
       "step": 540
     },
     {
       "epoch": 0.5832449628844114,
-      "grad_norm": 0.8940134048461914,
+      "grad_norm": 0.8357598185539246,
       "learning_rate": 4.0279250618593144e-05,
-      "loss": 0.441,
+      "loss": 0.4303,
       "step": 550
     },
     {
       "epoch": 0.5832449628844114,
-      "eval_loss": 0.6002687215805054,
-      "eval_runtime": 11.8218,
-      "eval_samples_per_second": 40.603,
-      "eval_steps_per_second": 2.538,
+      "eval_loss": 0.5633410215377808,
+      "eval_runtime": 12.2865,
+      "eval_samples_per_second": 39.067,
+      "eval_steps_per_second": 2.442,
       "step": 550
     },
     {
       "epoch": 0.5938494167550371,
-      "grad_norm": 1.051446795463562,
+      "grad_norm": 0.9880125522613525,
       "learning_rate": 4.0102509720749383e-05,
-      "loss": 0.4531,
+      "loss": 0.4708,
       "step": 560
     },
     {
       "epoch": 0.6044538706256628,
-      "grad_norm": 1.055331826210022,
+      "grad_norm": 0.8248680233955383,
       "learning_rate": 3.992576882290562e-05,
-      "loss": 0.4405,
+      "loss": 0.4434,
       "step": 570
     },
     {
       "epoch": 0.6150583244962884,
-      "grad_norm": 0.9812480211257935,
+      "grad_norm": 0.8025386333465576,
       "learning_rate": 3.974902792506186e-05,
-      "loss": 0.4128,
+      "loss": 0.4173,
       "step": 580
     },
     {
       "epoch": 0.6256627783669141,
-      "grad_norm": 0.9506167769432068,
+      "grad_norm": 0.8487368822097778,
       "learning_rate": 3.95722870272181e-05,
-      "loss": 0.4077,
+      "loss": 0.4452,
       "step": 590
     },
     {
       "epoch": 0.6362672322375398,
-      "grad_norm": 0.9744945764541626,
+      "grad_norm": 0.8240554928779602,
       "learning_rate": 3.939554612937434e-05,
-      "loss": 0.4531,
+      "loss": 0.4772,
       "step": 600
     },
     {
       "epoch": 0.6362672322375398,
-      "eval_loss": 0.5941649079322815,
-      "eval_runtime": 11.8037,
-      "eval_samples_per_second": 40.665,
-      "eval_steps_per_second": 2.542,
+      "eval_loss": 0.5596659183502197,
+      "eval_runtime": 12.2955,
+      "eval_samples_per_second": 39.039,
+      "eval_steps_per_second": 2.44,
       "step": 600
     },
     {
       "epoch": 0.6468716861081655,
-      "grad_norm": 0.9825407862663269,
+      "grad_norm": 0.798267126083374,
       "learning_rate": 3.921880523153057e-05,
-      "loss": 0.4281,
+      "loss": 0.4381,
       "step": 610
     },
     {
       "epoch": 0.6574761399787911,
-      "grad_norm": 0.9444342851638794,
+      "grad_norm": 0.8374388813972473,
       "learning_rate": 3.904206433368682e-05,
-      "loss": 0.4069,
+      "loss": 0.4408,
       "step": 620
     },
     {
       "epoch": 0.6680805938494168,
-      "grad_norm": 0.8893631100654602,
+      "grad_norm": 0.7530189156532288,
       "learning_rate": 3.886532343584305e-05,
-      "loss": 0.4067,
+      "loss": 0.4166,
       "step": 630
     },
     {
       "epoch": 0.6786850477200425,
-      "grad_norm": 0.9782839417457581,
+      "grad_norm": 0.8474013209342957,
       "learning_rate": 3.86885825379993e-05,
-      "loss": 0.4237,
+      "loss": 0.4266,
       "step": 640
     },
     {
       "epoch": 0.689289501590668,
-      "grad_norm": 0.9258303642272949,
+      "grad_norm": 0.8241029381752014,
       "learning_rate": 3.851184164015553e-05,
-      "loss": 0.4181,
+      "loss": 0.4365,
       "step": 650
     },
     {
       "epoch": 0.689289501590668,
-      "eval_loss": 0.5927689671516418,
-      "eval_runtime": 11.8222,
-      "eval_samples_per_second": 40.601,
-      "eval_steps_per_second": 2.538,
+      "eval_loss": 0.5578514337539673,
+      "eval_runtime": 12.2994,
+      "eval_samples_per_second": 39.026,
+      "eval_steps_per_second": 2.439,
       "step": 650
     },
     {
       "epoch": 0.6998939554612937,
-      "grad_norm": 0.9336357116699219,
+      "grad_norm": 0.8059120178222656,
       "learning_rate": 3.8335100742311775e-05,
-      "loss": 0.4264,
+      "loss": 0.4426,
       "step": 660
     },
     {
       "epoch": 0.7104984093319194,
-      "grad_norm": 1.0914337635040283,
+      "grad_norm": 0.8550363183021545,
       "learning_rate": 3.8158359844468015e-05,
-      "loss": 0.4163,
+      "loss": 0.4276,
       "step": 670
     },
     {
       "epoch": 0.721102863202545,
-      "grad_norm": 0.9416537284851074,
+      "grad_norm": 0.7503067851066589,
       "learning_rate": 3.7981618946624254e-05,
-      "loss": 0.4292,
+      "loss": 0.4256,
       "step": 680
     },
     {
       "epoch": 0.7317073170731707,
-      "grad_norm": 1.0038481950759888,
+      "grad_norm": 0.836462676525116,
       "learning_rate": 3.780487804878049e-05,
-      "loss": 0.4532,
+      "loss": 0.4748,
       "step": 690
     },
     {
       "epoch": 0.7423117709437964,
-      "grad_norm": 1.0638171434402466,
+      "grad_norm": 0.8518794178962708,
       "learning_rate": 3.7628137150936725e-05,
-      "loss": 0.437,
+      "loss": 0.4471,
       "step": 700
     },
     {
       "epoch": 0.7423117709437964,
-      "eval_loss": 0.5840710997581482,
-      "eval_runtime": 11.8245,
-      "eval_samples_per_second": 40.594,
-      "eval_steps_per_second": 2.537,
+      "eval_loss": 0.5494912266731262,
+      "eval_runtime": 12.3255,
+      "eval_samples_per_second": 38.944,
+      "eval_steps_per_second": 2.434,
       "step": 700
     },
     {
       "epoch": 0.7529162248144221,
-      "grad_norm": 1.091558814048767,
+      "grad_norm": 0.8614299297332764,
       "learning_rate": 3.745139625309297e-05,
-      "loss": 0.4171,
+      "loss": 0.4418,
       "step": 710
     },
     {
       "epoch": 0.7635206786850477,
-      "grad_norm": 1.052101731300354,
+      "grad_norm": 0.8816917538642883,
       "learning_rate": 3.7274655355249204e-05,
-      "loss": 0.425,
+      "loss": 0.4201,
       "step": 720
     },
     {
       "epoch": 0.7741251325556734,
-      "grad_norm": 0.9969995021820068,
+      "grad_norm": 0.8338310718536377,
       "learning_rate": 3.709791445740545e-05,
-      "loss": 0.4329,
+      "loss": 0.4694,
       "step": 730
     },
     {
       "epoch": 0.7847295864262991,
-      "grad_norm": 1.0250300168991089,
+      "grad_norm": 0.8940499424934387,
       "learning_rate": 3.692117355956168e-05,
-      "loss": 0.4221,
+      "loss": 0.4461,
       "step": 740
     },
     {
       "epoch": 0.7953340402969247,
-      "grad_norm": 1.0140836238861084,
+      "grad_norm": 0.7832528352737427,
       "learning_rate": 3.674443266171793e-05,
-      "loss": 0.4059,
+      "loss": 0.4085,
       "step": 750
     },
     {
       "epoch": 0.7953340402969247,
-      "eval_loss": 0.5901935696601868,
-      "eval_runtime": 11.8492,
-      "eval_samples_per_second": 40.509,
-      "eval_steps_per_second": 2.532,
+      "eval_loss": 0.5503777861595154,
+      "eval_runtime": 12.3484,
+      "eval_samples_per_second": 38.871,
+      "eval_steps_per_second": 2.429,
       "step": 750
     },
     {
       "epoch": 0.8059384941675504,
-      "grad_norm": 0.9665971994400024,
+      "grad_norm": 0.8207463622093201,
       "learning_rate": 3.656769176387416e-05,
-      "loss": 0.3824,
+      "loss": 0.4087,
       "step": 760
     },
     {
       "epoch": 0.8165429480381761,
-      "grad_norm": 0.9899901747703552,
+      "grad_norm": 0.8567345142364502,
       "learning_rate": 3.63909508660304e-05,
-      "loss": 0.402,
+      "loss": 0.4248,
       "step": 770
     },
     {
       "epoch": 0.8271474019088016,
-      "grad_norm": 0.9623307585716248,
+      "grad_norm": 0.797810971736908,
       "learning_rate": 3.621420996818664e-05,
-      "loss": 0.4078,
+      "loss": 0.4228,
       "step": 780
     },
     {
       "epoch": 0.8377518557794273,
-      "grad_norm": 1.0048012733459473,
+      "grad_norm": 0.8418610692024231,
       "learning_rate": 3.603746907034288e-05,
-      "loss": 0.4211,
+      "loss": 0.4517,
       "step": 790
     },
     {
       "epoch": 0.848356309650053,
-      "grad_norm": 1.04786217212677,
+      "grad_norm": 0.868778645992279,
       "learning_rate": 3.586072817249912e-05,
-      "loss": 0.4224,
+      "loss": 0.4521,
       "step": 800
     },
     {
       "epoch": 0.848356309650053,
-      "eval_loss": 0.5850113034248352,
-      "eval_runtime": 11.8518,
-      "eval_samples_per_second": 40.5,
-      "eval_steps_per_second": 2.531,
+      "eval_loss": 0.545513391494751,
+      "eval_runtime": 12.3617,
+      "eval_samples_per_second": 38.83,
+      "eval_steps_per_second": 2.427,
       "step": 800
     },
     {
       "epoch": 0.8589607635206787,
-      "grad_norm": 1.1098370552062988,
+      "grad_norm": 0.8869248628616333,
       "learning_rate": 3.5683987274655356e-05,
-      "loss": 0.4033,
+      "loss": 0.4324,
       "step": 810
     },
     {
       "epoch": 0.8695652173913043,
-      "grad_norm": 1.0828533172607422,
+      "grad_norm": 0.7634918689727783,
       "learning_rate": 3.5507246376811596e-05,
-      "loss": 0.3976,
+      "loss": 0.4085,
       "step": 820
     },
     {
       "epoch": 0.88016967126193,
-      "grad_norm": 1.0007745027542114,
+      "grad_norm": 0.8498929142951965,
       "learning_rate": 3.5330505478967835e-05,
-      "loss": 0.4101,
+      "loss": 0.4282,
       "step": 830
     },
     {
       "epoch": 0.8907741251325557,
-      "grad_norm": 1.0621625185012817,
+      "grad_norm": 0.9246194362640381,
       "learning_rate": 3.5153764581124074e-05,
-      "loss": 0.4113,
+      "loss": 0.4376,
       "step": 840
     },
     {
       "epoch": 0.9013785790031813,
-      "grad_norm": 1.0198171138763428,
+      "grad_norm": 0.8447480797767639,
       "learning_rate": 3.497702368328031e-05,
-      "loss": 0.3957,
+      "loss": 0.4139,
       "step": 850
     },
     {
       "epoch": 0.9013785790031813,
-      "eval_loss": 0.5811945199966431,
-      "eval_runtime": 11.8706,
-      "eval_samples_per_second": 40.436,
-      "eval_steps_per_second": 2.527,
+      "eval_loss": 0.5390809178352356,
+      "eval_runtime": 12.3471,
+      "eval_samples_per_second": 38.876,
+      "eval_steps_per_second": 2.43,
       "step": 850
     },
     {
       "epoch": 0.911983032873807,
-      "grad_norm": 1.035551905632019,
+      "grad_norm": 0.8634617924690247,
       "learning_rate": 3.480028278543655e-05,
-      "loss": 0.4047,
+      "loss": 0.4271,
       "step": 860
     },
     {
       "epoch": 0.9225874867444327,
-      "grad_norm": 1.0746926069259644,
+      "grad_norm": 0.8947970867156982,
       "learning_rate": 3.462354188759279e-05,
-      "loss": 0.4035,
+      "loss": 0.4294,
       "step": 870
     },
     {
       "epoch": 0.9331919406150583,
-      "grad_norm": 1.0593372583389282,
+      "grad_norm": 0.9222453236579895,
       "learning_rate": 3.444680098974903e-05,
-      "loss": 0.403,
+      "loss": 0.4303,
       "step": 880
     },
     {
       "epoch": 0.943796394485684,
-      "grad_norm": 1.0974407196044922,
+      "grad_norm": 0.9126210808753967,
       "learning_rate": 3.427006009190527e-05,
-      "loss": 0.3994,
+      "loss": 0.4172,
       "step": 890
     },
     {
       "epoch": 0.9544008483563097,
-      "grad_norm": 0.924757182598114,
+      "grad_norm": 0.7726302742958069,
       "learning_rate": 3.409331919406151e-05,
-      "loss": 0.3854,
+      "loss": 0.4092,
       "step": 900
     },
     {
       "epoch": 0.9544008483563097,
-      "eval_loss": 0.5818654894828796,
-      "eval_runtime": 11.8558,
-      "eval_samples_per_second": 40.486,
-      "eval_steps_per_second": 2.53,
+      "eval_loss": 0.5375456213951111,
+      "eval_runtime": 12.3333,
+      "eval_samples_per_second": 38.919,
+      "eval_steps_per_second": 2.432,
       "step": 900
     },
     {
       "epoch": 0.9650053022269353,
-      "grad_norm": 1.0254701375961304,
+      "grad_norm": 0.8598764538764954,
       "learning_rate": 3.391657829621775e-05,
-      "loss": 0.4214,
+      "loss": 0.4417,
       "step": 910
     },
     {
       "epoch": 0.975609756097561,
-      "grad_norm": 1.0513700246810913,
+      "grad_norm": 0.7907281517982483,
       "learning_rate": 3.373983739837399e-05,
-      "loss": 0.3934,
+      "loss": 0.4145,
       "step": 920
     },
     {
       "epoch": 0.9862142099681867,
-      "grad_norm": 1.0283734798431396,
+      "grad_norm": 0.8753092885017395,
       "learning_rate": 3.356309650053022e-05,
-      "loss": 0.3792,
+      "loss": 0.4026,
       "step": 930
     },
     {
       "epoch": 0.9968186638388123,
-      "grad_norm": 1.0671693086624146,
+      "grad_norm": 0.8522175550460815,
       "learning_rate": 3.3386355602686466e-05,
-      "loss": 0.3777,
+      "loss": 0.4208,
       "step": 940
     },
     {
       "epoch": 1.007423117709438,
-      "grad_norm": 1.0414798259735107,
+      "grad_norm": 0.8519699573516846,
       "learning_rate": 3.32096147048427e-05,
-      "loss": 0.3681,
+      "loss": 0.3799,
       "step": 950
     },
     {
       "epoch": 1.007423117709438,
-      "eval_loss": 0.5841290950775146,
-      "eval_runtime": 11.8527,
-      "eval_samples_per_second": 40.497,
-      "eval_steps_per_second": 2.531,
+      "eval_loss": 0.5359489321708679,
+      "eval_runtime": 12.3068,
+      "eval_samples_per_second": 39.003,
+      "eval_steps_per_second": 2.438,
       "step": 950
     },
     {
       "epoch": 1.0180275715800637,
-      "grad_norm": 1.1088826656341553,
+      "grad_norm": 0.9038631916046143,
       "learning_rate": 3.3032873806998944e-05,
-      "loss": 0.3536,
+      "loss": 0.3786,
       "step": 960
     },
     {
       "epoch": 1.0286320254506893,
-      "grad_norm": 1.2678877115249634,
+      "grad_norm": 0.897493302822113,
       "learning_rate": 3.285613290915518e-05,
-      "loss": 0.38,
+      "loss": 0.4039,
       "step": 970
     },
     {
       "epoch": 1.039236479321315,
-      "grad_norm": 1.0846420526504517,
+      "grad_norm": 0.8796167969703674,
       "learning_rate": 3.267939201131142e-05,
-      "loss": 0.3522,
+      "loss": 0.3817,
       "step": 980
     },
     {
       "epoch": 1.0498409331919407,
-      "grad_norm": 1.0741760730743408,
+      "grad_norm": 0.8639382719993591,
       "learning_rate": 3.2502651113467655e-05,
-      "loss": 0.3392,
+      "loss": 0.3613,
       "step": 990
     },
     {
       "epoch": 1.0604453870625663,
-      "grad_norm": 1.2564727067947388,
+      "grad_norm": 0.8916295170783997,
       "learning_rate": 3.2325910215623894e-05,
-      "loss": 0.3538,
+      "loss": 0.3759,
       "step": 1000
     },
     {
       "epoch": 1.0604453870625663,
-      "eval_loss": 0.5931693911552429,
-      "eval_runtime": 11.8583,
-      "eval_samples_per_second": 40.478,
-      "eval_steps_per_second": 2.53,
+      "eval_loss": 0.5398057103157043,
+      "eval_runtime": 12.2749,
+      "eval_samples_per_second": 39.104,
+      "eval_steps_per_second": 2.444,
       "step": 1000
     },
     {
       "epoch": 1.0710498409331919,
-      "grad_norm": 1.1698198318481445,
+      "grad_norm": 0.9307865500450134,
       "learning_rate": 3.2149169317780133e-05,
-      "loss": 0.3274,
+      "loss": 0.369,
       "step": 1010
     },
     {
       "epoch": 1.0816542948038177,
-      "grad_norm": 1.1250603199005127,
+      "grad_norm": 0.8993286490440369,
       "learning_rate": 3.197242841993637e-05,
-      "loss": 0.3519,
+      "loss": 0.3944,
       "step": 1020
     },
     {
       "epoch": 1.0922587486744433,
-      "grad_norm": 1.1354424953460693,
+      "grad_norm": 0.9066745638847351,
       "learning_rate": 3.179568752209262e-05,
-      "loss": 0.3474,
+      "loss": 0.4009,
       "step": 1030
     },
     {
       "epoch": 1.1028632025450689,
-      "grad_norm": 1.2070807218551636,
+      "grad_norm": 0.9609470963478088,
       "learning_rate": 3.161894662424885e-05,
-      "loss": 0.3426,
+      "loss": 0.3679,
       "step": 1040
     },
     {
       "epoch": 1.1134676564156947,
-      "grad_norm": 1.147916555404663,
+      "grad_norm": 0.853422999382019,
       "learning_rate": 3.14422057264051e-05,
-      "loss": 0.3449,
+      "loss": 0.3666,
       "step": 1050
     },
     {
       "epoch": 1.1134676564156947,
-      "eval_loss": 0.5951507687568665,
-      "eval_runtime": 11.8382,
-      "eval_samples_per_second": 40.547,
-      "eval_steps_per_second": 2.534,
+      "eval_loss": 0.5404828786849976,
+      "eval_runtime": 12.2658,
+      "eval_samples_per_second": 39.133,
+      "eval_steps_per_second": 2.446,
       "step": 1050
     },
     {
       "epoch": 1.1240721102863203,
-      "grad_norm": 1.1941380500793457,
+      "grad_norm": 0.9324142932891846,
       "learning_rate": 3.126546482856133e-05,
-      "loss": 0.3348,
+      "loss": 0.372,
       "step": 1060
     },
     {
       "epoch": 1.1346765641569458,
-      "grad_norm": 1.0945191383361816,
+      "grad_norm": 0.8521295189857483,
       "learning_rate": 3.108872393071757e-05,
-      "loss": 0.3724,
+      "loss": 0.4049,
       "step": 1070
     },
     {
       "epoch": 1.1452810180275717,
-      "grad_norm": 1.322505235671997,
+      "grad_norm": 0.8782608509063721,
       "learning_rate": 3.091198303287381e-05,
-      "loss": 0.3312,
+      "loss": 0.3637,
       "step": 1080
     },
     {
       "epoch": 1.1558854718981972,
-      "grad_norm": 1.0995663404464722,
+      "grad_norm": 0.9670329689979553,
       "learning_rate": 3.073524213503005e-05,
-      "loss": 0.351,
+      "loss": 0.3683,
       "step": 1090
     },
     {
       "epoch": 1.1664899257688228,
-      "grad_norm": 1.1347137689590454,
+      "grad_norm": 0.9527105093002319,
       "learning_rate": 3.0558501237186286e-05,
-      "loss": 0.343,
+      "loss": 0.3809,
       "step": 1100
     },
     {
       "epoch": 1.1664899257688228,
-      "eval_loss": 0.59864342212677,
-      "eval_runtime": 11.8201,
-      "eval_samples_per_second": 40.609,
-      "eval_steps_per_second": 2.538,
+      "eval_loss": 0.5395468473434448,
+      "eval_runtime": 12.2788,
+      "eval_samples_per_second": 39.092,
+      "eval_steps_per_second": 2.443,
       "step": 1100
     },
     {
       "epoch": 1.1770943796394486,
-      "grad_norm": 1.2727549076080322,
+      "grad_norm": 0.9289312362670898,
       "learning_rate": 3.0381760339342525e-05,
-      "loss": 0.342,
+      "loss": 0.374,
       "step": 1110
     },
     {
       "epoch": 1.1876988335100742,
-      "grad_norm": 1.1988881826400757,
+      "grad_norm": 0.9651899933815002,
       "learning_rate": 3.0205019441498768e-05,
-      "loss": 0.3238,
+      "loss": 0.3706,
       "step": 1120
     },
     {
       "epoch": 1.1983032873806998,
-      "grad_norm": 1.1220091581344604,
+      "grad_norm": 0.980241060256958,
       "learning_rate": 3.0028278543655004e-05,
-      "loss": 0.3281,
+      "loss": 0.3607,
       "step": 1130
     },
     {
       "epoch": 1.2089077412513256,
-      "grad_norm": 1.1029518842697144,
+      "grad_norm": 0.8908747434616089,
       "learning_rate": 2.985153764581124e-05,
-      "loss": 0.3566,
+      "loss": 0.3841,
       "step": 1140
     },
     {
       "epoch": 1.2195121951219512,
-      "grad_norm": 1.0658305883407593,
+      "grad_norm": 0.9470391869544983,
       "learning_rate": 2.9674796747967482e-05,
-      "loss": 0.3431,
+      "loss": 0.3945,
       "step": 1150
     },
     {
       "epoch": 1.2195121951219512,
-      "eval_loss": 0.5974192023277283,
-      "eval_runtime": 11.8028,
-      "eval_samples_per_second": 40.668,
-      "eval_steps_per_second": 2.542,
+      "eval_loss": 0.5362985134124756,
+      "eval_runtime": 12.3051,
+      "eval_samples_per_second": 39.008,
+      "eval_steps_per_second": 2.438,
       "step": 1150
     },
     {
       "epoch": 1.2301166489925768,
-      "grad_norm": 1.306966781616211,
+      "grad_norm": 0.952620267868042,
       "learning_rate": 2.9498055850123718e-05,
-      "loss": 0.3355,
+      "loss": 0.3531,
       "step": 1160
     },
     {
       "epoch": 1.2407211028632026,
-      "grad_norm": 1.370133638381958,
+      "grad_norm": 1.1877236366271973,
       "learning_rate": 2.932131495227996e-05,
-      "loss": 0.3475,
+      "loss": 0.3876,
       "step": 1170
     },
     {
       "epoch": 1.2513255567338282,
-      "grad_norm": 1.2001569271087646,
+      "grad_norm": 1.0492746829986572,
       "learning_rate": 2.9144574054436196e-05,
-      "loss": 0.331,
+      "loss": 0.375,
       "step": 1180
     },
     {
       "epoch": 1.2619300106044538,
-      "grad_norm": 1.27125883102417,
+      "grad_norm": 1.0071635246276855,
       "learning_rate": 2.896783315659244e-05,
-      "loss": 0.3461,
+      "loss": 0.3652,
       "step": 1190
     },
     {
       "epoch": 1.2725344644750796,
-      "grad_norm": 1.417680025100708,
+      "grad_norm": 0.9905943274497986,
       "learning_rate": 2.8791092258748675e-05,
-      "loss": 0.3298,
+      "loss": 0.3733,
       "step": 1200
     },
     {
       "epoch": 1.2725344644750796,
-      "eval_loss": 0.6026219129562378,
-      "eval_runtime": 11.8206,
-      "eval_samples_per_second": 40.607,
-      "eval_steps_per_second": 2.538,
+      "eval_loss": 0.5363767743110657,
+      "eval_runtime": 12.323,
+      "eval_samples_per_second": 38.952,
+      "eval_steps_per_second": 2.434,
       "step": 1200
     }
   ],
@@ -1058,7 +1058,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.2382115382161818e+17,
+  "total_flos": 1.232816857284526e+17,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

checkpoint-1200/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53a8c021c0523ab617f7d132f6e1d21f2075e97fdb7689e840960f38ef5cb15e
+oid sha256:c5186a048b8d2af8a229717dd049aeeb2812c6aef0d473c9569aaea2283f4a48
 size 5880