Training in progress, step 600, checkpoint

checkpoint-600/adapter_config.json

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

checkpoint-600/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:07c173d6640b3cad5cb022664334d6a1bdc367e7798f133356250060f4277f67
+oid sha256:a460e7245c1902f3c055d5e3d959aab265720548811dc161626a10629db32dc4
 size 161533192

checkpoint-600/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1fb545a5f2038a483b3f52f22d16113dababb1dfccef4dd672b79b7844a4fbe7
+oid sha256:6e2541d76f1372d9ed1e746ee36219dcf4408354d680ba66e9db225eb3ee1c20
 size 323290986

checkpoint-600/trainer_state.json

@@ -10,518 +10,518 @@
   "log_history": [
     {
       "epoch": 0.010604453870625663,
-      "grad_norm": 2.0663416385650635,
+      "grad_norm": 2.276311159133911,
       "learning_rate": 4.982325910215624e-05,
-      "loss": 1.91,
+      "loss": 1.9038,
       "step": 10
     },
     {
       "epoch": 0.021208907741251327,
-      "grad_norm": 0.7459995746612549,
+      "grad_norm": 0.7678110003471375,
       "learning_rate": 4.964651820431248e-05,
-      "loss": 1.0094,
+      "loss": 0.997,
       "step": 20
     },
     {
       "epoch": 0.031813361611876985,
-      "grad_norm": 0.656954824924469,
+      "grad_norm": 0.6726565361022949,
       "learning_rate": 4.946977730646872e-05,
-      "loss": 0.8052,
+      "loss": 0.8021,
       "step": 30
     },
     {
       "epoch": 0.042417815482502653,
-      "grad_norm": 0.7273707389831543,
+      "grad_norm": 0.7318501472473145,
       "learning_rate": 4.929303640862496e-05,
-      "loss": 0.7145,
+      "loss": 0.7122,
       "step": 40
     },
     {
       "epoch": 0.053022269353128315,
-      "grad_norm": 0.6689164042472839,
+      "grad_norm": 0.6708294749259949,
       "learning_rate": 4.91162955107812e-05,
-      "loss": 0.6797,
+      "loss": 0.6783,
       "step": 50
     },
     {
       "epoch": 0.053022269353128315,
-      "eval_loss": 0.8142186403274536,
-      "eval_runtime": 11.794,
-      "eval_samples_per_second": 40.699,
-      "eval_steps_per_second": 2.544,
+      "eval_loss": 0.8075584173202515,
+      "eval_runtime": 11.7244,
+      "eval_samples_per_second": 40.94,
+      "eval_steps_per_second": 2.559,
       "step": 50
     },
     {
       "epoch": 0.06362672322375397,
-      "grad_norm": 0.6781293749809265,
+      "grad_norm": 0.7024317979812622,
       "learning_rate": 4.893955461293744e-05,
-      "loss": 0.6249,
+      "loss": 0.6244,
       "step": 60
     },
     {
       "epoch": 0.07423117709437964,
-      "grad_norm": 0.6776543259620667,
+      "grad_norm": 0.6959078907966614,
       "learning_rate": 4.876281371509368e-05,
-      "loss": 0.612,
+      "loss": 0.6117,
       "step": 70
     },
     {
       "epoch": 0.08483563096500531,
-      "grad_norm": 0.7603946924209595,
+      "grad_norm": 0.7824214696884155,
       "learning_rate": 4.858607281724991e-05,
-      "loss": 0.5998,
+      "loss": 0.5989,
       "step": 80
     },
     {
       "epoch": 0.09544008483563096,
-      "grad_norm": 0.7581377029418945,
+      "grad_norm": 0.7792642116546631,
       "learning_rate": 4.8409331919406155e-05,
-      "loss": 0.6032,
+      "loss": 0.6033,
       "step": 90
     },
     {
       "epoch": 0.10604453870625663,
-      "grad_norm": 0.7402133345603943,
+      "grad_norm": 0.7486705780029297,
       "learning_rate": 4.823259102156239e-05,
-      "loss": 0.5912,
+      "loss": 0.5904,
       "step": 100
     },
     {
       "epoch": 0.10604453870625663,
-      "eval_loss": 0.7294260859489441,
-      "eval_runtime": 11.8272,
-      "eval_samples_per_second": 40.585,
-      "eval_steps_per_second": 2.537,
+      "eval_loss": 0.7230511903762817,
+      "eval_runtime": 11.7589,
+      "eval_samples_per_second": 40.82,
+      "eval_steps_per_second": 2.551,
       "step": 100
     },
     {
       "epoch": 0.11664899257688228,
-      "grad_norm": 0.8224169015884399,
+      "grad_norm": 0.8214134573936462,
       "learning_rate": 4.8055850123718633e-05,
-      "loss": 0.578,
+      "loss": 0.5778,
       "step": 110
     },
     {
       "epoch": 0.12725344644750794,
-      "grad_norm": 0.7843881249427795,
+      "grad_norm": 0.7937195897102356,
       "learning_rate": 4.7879109225874866e-05,
-      "loss": 0.5751,
+      "loss": 0.5748,
       "step": 120
     },
     {
       "epoch": 0.1378579003181336,
-      "grad_norm": 0.772152304649353,
+      "grad_norm": 0.7781085968017578,
       "learning_rate": 4.770236832803111e-05,
-      "loss": 0.5471,
+      "loss": 0.5469,
       "step": 130
     },
     {
       "epoch": 0.14846235418875928,
-      "grad_norm": 0.6972452998161316,
+      "grad_norm": 0.695203423500061,
       "learning_rate": 4.7525627430187344e-05,
-      "loss": 0.5516,
+      "loss": 0.5511,
       "step": 140
     },
     {
       "epoch": 0.15906680805938495,
-      "grad_norm": 0.7806265354156494,
+      "grad_norm": 0.7824985384941101,
       "learning_rate": 4.734888653234359e-05,
-      "loss": 0.5263,
+      "loss": 0.5258,
       "step": 150
     },
     {
       "epoch": 0.15906680805938495,
-      "eval_loss": 0.6928939819335938,
-      "eval_runtime": 11.8093,
-      "eval_samples_per_second": 40.646,
-      "eval_steps_per_second": 2.54,
+      "eval_loss": 0.6867014765739441,
+      "eval_runtime": 11.7852,
+      "eval_samples_per_second": 40.729,
+      "eval_steps_per_second": 2.546,
       "step": 150
     },
     {
       "epoch": 0.16967126193001061,
-      "grad_norm": 0.7372023463249207,
+      "grad_norm": 0.7392640113830566,
       "learning_rate": 4.717214563449982e-05,
-      "loss": 0.5715,
+      "loss": 0.5714,
       "step": 160
     },
     {
       "epoch": 0.18027571580063625,
-      "grad_norm": 0.8236038684844971,
+      "grad_norm": 0.8316007852554321,
       "learning_rate": 4.699540473665606e-05,
-      "loss": 0.4987,
+      "loss": 0.4982,
       "step": 170
     },
     {
       "epoch": 0.19088016967126192,
-      "grad_norm": 0.8057310581207275,
+      "grad_norm": 0.8199877738952637,
       "learning_rate": 4.681866383881231e-05,
-      "loss": 0.5097,
+      "loss": 0.5092,
       "step": 180
     },
     {
       "epoch": 0.2014846235418876,
-      "grad_norm": 0.7124797105789185,
+      "grad_norm": 0.7089737057685852,
       "learning_rate": 4.664192294096854e-05,
-      "loss": 0.4952,
+      "loss": 0.4953,
       "step": 190
     },
     {
       "epoch": 0.21208907741251326,
-      "grad_norm": 0.7755710482597351,
+      "grad_norm": 0.7906930446624756,
       "learning_rate": 4.6465182043124786e-05,
-      "loss": 0.497,
+      "loss": 0.4974,
       "step": 200
     },
     {
       "epoch": 0.21208907741251326,
-      "eval_loss": 0.6748186349868774,
-      "eval_runtime": 11.8077,
-      "eval_samples_per_second": 40.651,
-      "eval_steps_per_second": 2.541,
+      "eval_loss": 0.669015645980835,
+      "eval_runtime": 11.8934,
+      "eval_samples_per_second": 40.358,
+      "eval_steps_per_second": 2.522,
       "step": 200
     },
     {
       "epoch": 0.22269353128313893,
-      "grad_norm": 0.912337064743042,
+      "grad_norm": 0.9085277318954468,
       "learning_rate": 4.628844114528102e-05,
-      "loss": 0.4997,
+      "loss": 0.4991,
       "step": 210
     },
     {
       "epoch": 0.23329798515376457,
-      "grad_norm": 0.771654486656189,
+      "grad_norm": 0.7775648832321167,
       "learning_rate": 4.6111700247437265e-05,
-      "loss": 0.5027,
+      "loss": 0.5024,
       "step": 220
     },
     {
       "epoch": 0.24390243902439024,
-      "grad_norm": 0.8241815567016602,
+      "grad_norm": 0.8324123620986938,
       "learning_rate": 4.59349593495935e-05,
-      "loss": 0.5196,
+      "loss": 0.5187,
       "step": 230
     },
     {
       "epoch": 0.2545068928950159,
-      "grad_norm": 0.7876004576683044,
+      "grad_norm": 0.8002191781997681,
       "learning_rate": 4.5758218451749736e-05,
-      "loss": 0.4791,
+      "loss": 0.4785,
       "step": 240
     },
     {
       "epoch": 0.2651113467656416,
-      "grad_norm": 0.7941911816596985,
+      "grad_norm": 0.8004704117774963,
       "learning_rate": 4.5581477553905975e-05,
-      "loss": 0.5078,
+      "loss": 0.5073,
       "step": 250
     },
     {
       "epoch": 0.2651113467656416,
-      "eval_loss": 0.653518795967102,
-      "eval_runtime": 11.7906,
-      "eval_samples_per_second": 40.71,
-      "eval_steps_per_second": 2.544,
+      "eval_loss": 0.6479213833808899,
+      "eval_runtime": 11.882,
+      "eval_samples_per_second": 40.397,
+      "eval_steps_per_second": 2.525,
       "step": 250
     },
     {
       "epoch": 0.2757158006362672,
-      "grad_norm": 0.8866105079650879,
+      "grad_norm": 0.893301248550415,
       "learning_rate": 4.5404736656062215e-05,
       "loss": 0.477,
       "step": 260
     },
     {
       "epoch": 0.2863202545068929,
-      "grad_norm": 0.8498968482017517,
+      "grad_norm": 0.8550834059715271,
       "learning_rate": 4.5227995758218454e-05,
-      "loss": 0.4926,
+      "loss": 0.4923,
       "step": 270
     },
     {
       "epoch": 0.29692470837751855,
-      "grad_norm": 0.8863232135772705,
+      "grad_norm": 0.8929152488708496,
       "learning_rate": 4.505125486037469e-05,
-      "loss": 0.463,
+      "loss": 0.4624,
       "step": 280
     },
     {
       "epoch": 0.3075291622481442,
-      "grad_norm": 0.9501034021377563,
+      "grad_norm": 0.9584974050521851,
       "learning_rate": 4.487451396253093e-05,
-      "loss": 0.4809,
+      "loss": 0.4801,
       "step": 290
     },
     {
       "epoch": 0.3181336161187699,
-      "grad_norm": 0.7933001518249512,
+      "grad_norm": 0.7924031615257263,
       "learning_rate": 4.469777306468717e-05,
-      "loss": 0.4941,
+      "loss": 0.4942,
       "step": 300
     },
     {
       "epoch": 0.3181336161187699,
-      "eval_loss": 0.6434358954429626,
-      "eval_runtime": 11.7969,
-      "eval_samples_per_second": 40.689,
-      "eval_steps_per_second": 2.543,
+      "eval_loss": 0.6371316313743591,
+      "eval_runtime": 11.8575,
+      "eval_samples_per_second": 40.481,
+      "eval_steps_per_second": 2.53,
       "step": 300
     },
     {
       "epoch": 0.32873806998939553,
-      "grad_norm": 0.9145284295082092,
+      "grad_norm": 0.9247742295265198,
       "learning_rate": 4.452103216684341e-05,
-      "loss": 0.5235,
+      "loss": 0.5234,
       "step": 310
     },
     {
       "epoch": 0.33934252386002123,
-      "grad_norm": 0.8557764887809753,
+      "grad_norm": 0.8617805242538452,
       "learning_rate": 4.434429126899965e-05,
-      "loss": 0.4809,
+      "loss": 0.4808,
       "step": 320
     },
     {
       "epoch": 0.34994697773064687,
-      "grad_norm": 0.8337067365646362,
+      "grad_norm": 0.8334285616874695,
       "learning_rate": 4.416755037115589e-05,
-      "loss": 0.4764,
+      "loss": 0.4758,
       "step": 330
     },
     {
       "epoch": 0.3605514316012725,
-      "grad_norm": 0.8928807377815247,
+      "grad_norm": 0.9079923629760742,
       "learning_rate": 4.399080947331213e-05,
-      "loss": 0.4634,
+      "loss": 0.463,
       "step": 340
     },
     {
       "epoch": 0.3711558854718982,
-      "grad_norm": 0.7639658451080322,
+      "grad_norm": 0.7632007002830505,
       "learning_rate": 4.381406857546837e-05,
-      "loss": 0.4829,
+      "loss": 0.4828,
       "step": 350
     },
     {
       "epoch": 0.3711558854718982,
-      "eval_loss": 0.6269502639770508,
-      "eval_runtime": 11.7838,
-      "eval_samples_per_second": 40.734,
-      "eval_steps_per_second": 2.546,
+      "eval_loss": 0.6212183237075806,
+      "eval_runtime": 11.8497,
+      "eval_samples_per_second": 40.507,
+      "eval_steps_per_second": 2.532,
       "step": 350
     },
     {
       "epoch": 0.38176033934252385,
-      "grad_norm": 0.8374601006507874,
+      "grad_norm": 0.8430343270301819,
       "learning_rate": 4.3637327677624606e-05,
-      "loss": 0.471,
+      "loss": 0.4711,
       "step": 360
     },
     {
       "epoch": 0.39236479321314954,
-      "grad_norm": 0.9471817016601562,
+      "grad_norm": 0.9495126605033875,
       "learning_rate": 4.3460586779780846e-05,
       "loss": 0.4528,
       "step": 370
     },
     {
       "epoch": 0.4029692470837752,
-      "grad_norm": 0.8860051035881042,
+      "grad_norm": 0.8903242945671082,
       "learning_rate": 4.328384588193708e-05,
-      "loss": 0.4684,
+      "loss": 0.468,
       "step": 380
     },
     {
       "epoch": 0.4135737009544008,
-      "grad_norm": 0.9099433422088623,
+      "grad_norm": 0.9306944608688354,
       "learning_rate": 4.3107104984093324e-05,
-      "loss": 0.4294,
+      "loss": 0.4299,
       "step": 390
     },
     {
       "epoch": 0.4241781548250265,
-      "grad_norm": 0.792826771736145,
+      "grad_norm": 0.7949309349060059,
       "learning_rate": 4.2930364086249556e-05,
-      "loss": 0.4597,
+      "loss": 0.4589,
       "step": 400
     },
     {
       "epoch": 0.4241781548250265,
-      "eval_loss": 0.6197967529296875,
-      "eval_runtime": 11.8062,
-      "eval_samples_per_second": 40.657,
-      "eval_steps_per_second": 2.541,
+      "eval_loss": 0.6148895621299744,
+      "eval_runtime": 11.8536,
+      "eval_samples_per_second": 40.494,
+      "eval_steps_per_second": 2.531,
       "step": 400
     },
     {
       "epoch": 0.43478260869565216,
-      "grad_norm": 0.8295171856880188,
+      "grad_norm": 0.8351686000823975,
       "learning_rate": 4.27536231884058e-05,
-      "loss": 0.4842,
+      "loss": 0.4836,
       "step": 410
     },
     {
       "epoch": 0.44538706256627786,
-      "grad_norm": 0.9126204252243042,
+      "grad_norm": 0.9184524416923523,
       "learning_rate": 4.2576882290562035e-05,
-      "loss": 0.4688,
+      "loss": 0.4686,
       "step": 420
     },
     {
       "epoch": 0.4559915164369035,
-      "grad_norm": 0.9485805630683899,
+      "grad_norm": 0.9610297679901123,
       "learning_rate": 4.240014139271828e-05,
-      "loss": 0.4803,
+      "loss": 0.4804,
       "step": 430
     },
     {
       "epoch": 0.46659597030752914,
-      "grad_norm": 0.8756628036499023,
+      "grad_norm": 0.8793377876281738,
       "learning_rate": 4.222340049487451e-05,
-      "loss": 0.4572,
+      "loss": 0.4571,
       "step": 440
     },
     {
       "epoch": 0.47720042417815484,
-      "grad_norm": 0.907244086265564,
+      "grad_norm": 0.913858950138092,
       "learning_rate": 4.204665959703076e-05,
-      "loss": 0.4394,
+      "loss": 0.4391,
       "step": 450
     },
     {
       "epoch": 0.47720042417815484,
-      "eval_loss": 0.6146622896194458,
-      "eval_runtime": 11.8005,
-      "eval_samples_per_second": 40.676,
-      "eval_steps_per_second": 2.542,
+      "eval_loss": 0.6096637845039368,
+      "eval_runtime": 11.8533,
+      "eval_samples_per_second": 40.495,
+      "eval_steps_per_second": 2.531,
       "step": 450
     },
     {
       "epoch": 0.4878048780487805,
-      "grad_norm": 0.9201627969741821,
+      "grad_norm": 0.9159062504768372,
       "learning_rate": 4.186991869918699e-05,
       "loss": 0.4294,
       "step": 460
     },
     {
       "epoch": 0.4984093319194062,
-      "grad_norm": 0.9242031574249268,
+      "grad_norm": 0.9346657395362854,
       "learning_rate": 4.169317780134323e-05,
-      "loss": 0.4335,
+      "loss": 0.433,
       "step": 470
     },
     {
       "epoch": 0.5090137857900318,
-      "grad_norm": 0.840977132320404,
+      "grad_norm": 0.8480401635169983,
       "learning_rate": 4.151643690349947e-05,
       "loss": 0.4639,
       "step": 480
     },
     {
       "epoch": 0.5196182396606575,
-      "grad_norm": 0.9487953782081604,
+      "grad_norm": 0.9547140002250671,
       "learning_rate": 4.133969600565571e-05,
       "loss": 0.4173,
       "step": 490
     },
     {
       "epoch": 0.5302226935312832,
-      "grad_norm": 0.8343156576156616,
+      "grad_norm": 0.8660142421722412,
       "learning_rate": 4.116295510781195e-05,
-      "loss": 0.4647,
+      "loss": 0.4644,
       "step": 500
     },
     {
       "epoch": 0.5302226935312832,
-      "eval_loss": 0.6104319095611572,
-      "eval_runtime": 11.8139,
-      "eval_samples_per_second": 40.63,
-      "eval_steps_per_second": 2.539,
+      "eval_loss": 0.6053425073623657,
+      "eval_runtime": 11.8341,
+      "eval_samples_per_second": 40.561,
+      "eval_steps_per_second": 2.535,
       "step": 500
     },
     {
       "epoch": 0.5408271474019088,
-      "grad_norm": 0.9216362237930298,
+      "grad_norm": 0.9299489855766296,
       "learning_rate": 4.098621420996819e-05,
-      "loss": 0.4499,
+      "loss": 0.4488,
       "step": 510
     },
     {
       "epoch": 0.5514316012725344,
-      "grad_norm": 0.957593560218811,
+      "grad_norm": 0.9637112021446228,
       "learning_rate": 4.080947331212443e-05,
-      "loss": 0.4202,
+      "loss": 0.4197,
       "step": 520
     },
     {
       "epoch": 0.5620360551431601,
-      "grad_norm": 1.0103106498718262,
+      "grad_norm": 0.9725180864334106,
       "learning_rate": 4.0632732414280666e-05,
-      "loss": 0.4315,
+      "loss": 0.4308,
       "step": 530
     },
     {
       "epoch": 0.5726405090137858,
-      "grad_norm": 1.0909500122070312,
+      "grad_norm": 1.095697045326233,
       "learning_rate": 4.0455991516436905e-05,
-      "loss": 0.4476,
+      "loss": 0.4473,
       "step": 540
     },
     {
       "epoch": 0.5832449628844114,
-      "grad_norm": 0.8890084028244019,
+      "grad_norm": 0.8940134048461914,
       "learning_rate": 4.0279250618593144e-05,
-      "loss": 0.4408,
+      "loss": 0.441,
       "step": 550
     },
     {
       "epoch": 0.5832449628844114,
-      "eval_loss": 0.6050440073013306,
-      "eval_runtime": 11.7965,
-      "eval_samples_per_second": 40.69,
-      "eval_steps_per_second": 2.543,
+      "eval_loss": 0.6002687215805054,
+      "eval_runtime": 11.8218,
+      "eval_samples_per_second": 40.603,
+      "eval_steps_per_second": 2.538,
       "step": 550
     },
     {
       "epoch": 0.5938494167550371,
-      "grad_norm": 1.038818359375,
+      "grad_norm": 1.051446795463562,
       "learning_rate": 4.0102509720749383e-05,
-      "loss": 0.4529,
+      "loss": 0.4531,
       "step": 560
     },
     {
       "epoch": 0.6044538706256628,
-      "grad_norm": 1.0575263500213623,
+      "grad_norm": 1.055331826210022,
       "learning_rate": 3.992576882290562e-05,
-      "loss": 0.4409,
+      "loss": 0.4405,
       "step": 570
     },
     {
       "epoch": 0.6150583244962884,
-      "grad_norm": 0.9827993512153625,
+      "grad_norm": 0.9812480211257935,
       "learning_rate": 3.974902792506186e-05,
-      "loss": 0.4132,
+      "loss": 0.4128,
       "step": 580
     },
     {
       "epoch": 0.6256627783669141,
-      "grad_norm": 0.9469496011734009,
+      "grad_norm": 0.9506167769432068,
       "learning_rate": 3.95722870272181e-05,
-      "loss": 0.4075,
+      "loss": 0.4077,
       "step": 590
     },
     {
       "epoch": 0.6362672322375398,
-      "grad_norm": 0.9693032503128052,
+      "grad_norm": 0.9744945764541626,
       "learning_rate": 3.939554612937434e-05,
-      "loss": 0.4533,
+      "loss": 0.4531,
       "step": 600
     },
     {
       "epoch": 0.6362672322375398,
-      "eval_loss": 0.5991437435150146,
-      "eval_runtime": 11.8081,
-      "eval_samples_per_second": 40.65,
-      "eval_steps_per_second": 2.541,
+      "eval_loss": 0.5941649079322815,
+      "eval_runtime": 11.8037,
+      "eval_samples_per_second": 40.665,
+      "eval_steps_per_second": 2.542,
       "step": 600
     }
   ],

checkpoint-600/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:872835407be2a0a95214d9cc3b7966660b3e25cbfda85f13ed9adad96c071793
+oid sha256:53a8c021c0523ab617f7d132f6e1d21f2075e97fdb7689e840960f38ef5cb15e
 size 5880