Upload folder using huggingface_hub

adapter_config.json

@@ -23,10 +23,10 @@
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
+    "qkv_proj",
     "gate_up_proj",
-    "down_proj",
-    "qkv_proj"
+    "o_proj",
+    "down_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b27a8f8f54ccc708af8722eb0dcfe9eb01e94d238fbfce8dbdc0281dac482d6a
+oid sha256:6f59f1a583963d421ac23948e79a9a6d7a43d73f7425dc3f81a1c3fba0234b14
 size 738232680

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e49313aaf2f278db459177c84bc07eae986e38e54dfc6aaeb7631ba961b9eb17
-size 738413771
+oid sha256:62dee84323e49bce2894834a449efb717fb4f5c0e43b0fc53908d8314f0fea2d
+size 1476611275

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9e1228c554e741eb261b7a2778cf9c596738b56bf6824b377cad0dd9c0e056fe
+oid sha256:adb5558d81e37e22ad7ef00a5b98b4bbabdb72c5cf774602127ab6ff043ccfad
 size 14645

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21c858dbf1ed6fb7ffe87cb1d0d39648f4c7ed2952d68adcb7d4391a9350ff94
+oid sha256:3105ebe8471f9890c3eb1f20cc0f2520fa5fdb0128474bbc87e607b2ec7c53dc
 size 1465

trainer_state.json

@@ -1,298 +1,523 @@
 {
-  "best_global_step": 285,
-  "best_metric": 0.002059183781966567,
-  "best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-285",
-  "epoch": 19.0,
+  "best_global_step": 500,
+  "best_metric": 0.0006410721107386053,
+  "best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-500",
+  "epoch": 33.333333333333336,
   "eval_steps": 500,
-  "global_step": 285,
+  "global_step": 500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 2.0235157012939453,
+      "grad_norm": 2.040301561355591,
       "learning_rate": 4.666666666666667e-05,
-      "loss": 2.4322,
+      "loss": 2.4299,
       "step": 15
     },
     {
       "epoch": 1.0,
-      "eval_loss": 1.4220346212387085,
-      "eval_runtime": 3.483,
-      "eval_samples_per_second": 8.613,
-      "eval_steps_per_second": 1.148,
+      "eval_loss": 1.4176747798919678,
+      "eval_runtime": 3.3557,
+      "eval_samples_per_second": 8.94,
+      "eval_steps_per_second": 1.192,
       "step": 15
     },
     {
       "epoch": 2.0,
-      "grad_norm": 1.0887768268585205,
+      "grad_norm": 1.0678095817565918,
       "learning_rate": 4.989727297688797e-05,
-      "loss": 1.1734,
+      "loss": 1.173,
       "step": 30
     },
     {
       "epoch": 2.0,
-      "eval_loss": 0.9683536887168884,
-      "eval_runtime": 3.5192,
-      "eval_samples_per_second": 8.525,
-      "eval_steps_per_second": 1.137,
+      "eval_loss": 0.9685609340667725,
+      "eval_runtime": 3.4367,
+      "eval_samples_per_second": 8.729,
+      "eval_steps_per_second": 1.164,
       "step": 30
     },
     {
       "epoch": 3.0,
-      "grad_norm": 0.9712529182434082,
+      "grad_norm": 1.092650055885315,
       "learning_rate": 4.956021052727731e-05,
-      "loss": 0.9344,
+      "loss": 0.9354,
       "step": 45
     },
     {
       "epoch": 3.0,
-      "eval_loss": 0.8322282433509827,
-      "eval_runtime": 3.5008,
-      "eval_samples_per_second": 8.57,
-      "eval_steps_per_second": 1.143,
+      "eval_loss": 0.8357031345367432,
+      "eval_runtime": 3.4013,
+      "eval_samples_per_second": 8.82,
+      "eval_steps_per_second": 1.176,
       "step": 45
     },
     {
       "epoch": 4.0,
-      "grad_norm": 1.2298760414123535,
+      "grad_norm": 1.079960823059082,
       "learning_rate": 4.899146773943374e-05,
-      "loss": 0.8145,
+      "loss": 0.8161,
       "step": 60
     },
     {
       "epoch": 4.0,
-      "eval_loss": 0.7362357974052429,
-      "eval_runtime": 3.5012,
-      "eval_samples_per_second": 8.569,
-      "eval_steps_per_second": 1.142,
+      "eval_loss": 0.7339344024658203,
+      "eval_runtime": 3.4059,
+      "eval_samples_per_second": 8.808,
+      "eval_steps_per_second": 1.174,
       "step": 60
     },
     {
       "epoch": 5.0,
-      "grad_norm": 1.222394347190857,
+      "grad_norm": 1.0952789783477783,
       "learning_rate": 4.819640965374681e-05,
-      "loss": 0.7171,
+      "loss": 0.7146,
       "step": 75
     },
     {
       "epoch": 5.0,
-      "eval_loss": 0.6475105881690979,
-      "eval_runtime": 3.5192,
-      "eval_samples_per_second": 8.525,
-      "eval_steps_per_second": 1.137,
+      "eval_loss": 0.6446787714958191,
+      "eval_runtime": 3.4087,
+      "eval_samples_per_second": 8.801,
+      "eval_steps_per_second": 1.173,
       "step": 75
     },
     {
       "epoch": 6.0,
-      "grad_norm": 1.3320592641830444,
+      "grad_norm": 1.183589220046997,
       "learning_rate": 4.718253617849306e-05,
-      "loss": 0.6252,
+      "loss": 0.6222,
       "step": 90
     },
     {
       "epoch": 6.0,
-      "eval_loss": 0.5572862029075623,
-      "eval_runtime": 3.5119,
-      "eval_samples_per_second": 8.542,
-      "eval_steps_per_second": 1.139,
+      "eval_loss": 0.5590938925743103,
+      "eval_runtime": 3.4083,
+      "eval_samples_per_second": 8.802,
+      "eval_steps_per_second": 1.174,
       "step": 90
     },
     {
       "epoch": 7.0,
-      "grad_norm": 1.3688997030258179,
+      "grad_norm": 1.408022165298462,
       "learning_rate": 4.595941134201871e-05,
-      "loss": 0.5356,
+      "loss": 0.5375,
       "step": 105
     },
     {
       "epoch": 7.0,
-      "eval_loss": 0.47074607014656067,
-      "eval_runtime": 3.4988,
-      "eval_samples_per_second": 8.574,
-      "eval_steps_per_second": 1.143,
+      "eval_loss": 0.48502209782600403,
+      "eval_runtime": 3.4091,
+      "eval_samples_per_second": 8.8,
+      "eval_steps_per_second": 1.173,
       "step": 105
     },
     {
       "epoch": 8.0,
-      "grad_norm": 1.474031925201416,
+      "grad_norm": 1.2591876983642578,
       "learning_rate": 4.4538573073752365e-05,
-      "loss": 0.4563,
+      "loss": 0.4583,
       "step": 120
     },
     {
       "epoch": 8.0,
-      "eval_loss": 0.4020730257034302,
-      "eval_runtime": 3.5099,
-      "eval_samples_per_second": 8.547,
-      "eval_steps_per_second": 1.14,
+      "eval_loss": 0.4147873818874359,
+      "eval_runtime": 3.4094,
+      "eval_samples_per_second": 8.799,
+      "eval_steps_per_second": 1.173,
       "step": 120
     },
     {
       "epoch": 9.0,
-      "grad_norm": 1.3377634286880493,
+      "grad_norm": 1.3704065084457397,
       "learning_rate": 4.2933424365097564e-05,
-      "loss": 0.3795,
+      "loss": 0.3893,
       "step": 135
     },
     {
       "epoch": 9.0,
-      "eval_loss": 0.3274089992046356,
-      "eval_runtime": 3.5065,
-      "eval_samples_per_second": 8.556,
-      "eval_steps_per_second": 1.141,
+      "eval_loss": 0.3290213942527771,
+      "eval_runtime": 3.4058,
+      "eval_samples_per_second": 8.809,
+      "eval_steps_per_second": 1.174,
       "step": 135
     },
     {
       "epoch": 10.0,
-      "grad_norm": 1.619517207145691,
+      "grad_norm": 1.5167052745819092,
       "learning_rate": 4.1159106836901674e-05,
-      "loss": 0.3099,
+      "loss": 0.3179,
       "step": 150
     },
     {
       "epoch": 10.0,
-      "eval_loss": 0.27303346991539,
-      "eval_runtime": 3.5078,
-      "eval_samples_per_second": 8.552,
-      "eval_steps_per_second": 1.14,
+      "eval_loss": 0.27090534567832947,
+      "eval_runtime": 3.4112,
+      "eval_samples_per_second": 8.795,
+      "eval_steps_per_second": 1.173,
       "step": 150
     },
     {
       "epoch": 11.0,
-      "grad_norm": 1.7652794122695923,
+      "grad_norm": 1.8745532035827637,
       "learning_rate": 3.923235790615907e-05,
-      "loss": 0.2477,
+      "loss": 0.2582,
       "step": 165
     },
     {
       "epoch": 11.0,
-      "eval_loss": 0.19887271523475647,
-      "eval_runtime": 3.5121,
-      "eval_samples_per_second": 8.542,
-      "eval_steps_per_second": 1.139,
+      "eval_loss": 0.21189619600772858,
+      "eval_runtime": 3.4123,
+      "eval_samples_per_second": 8.792,
+      "eval_steps_per_second": 1.172,
       "step": 165
     },
     {
       "epoch": 12.0,
-      "grad_norm": 2.3318214416503906,
+      "grad_norm": 2.4447438716888428,
       "learning_rate": 3.717135289931774e-05,
-      "loss": 0.1757,
+      "loss": 0.1923,
       "step": 180
     },
     {
       "epoch": 12.0,
-      "eval_loss": 0.14593930542469025,
-      "eval_runtime": 3.5028,
-      "eval_samples_per_second": 8.565,
-      "eval_steps_per_second": 1.142,
+      "eval_loss": 0.15350165963172913,
+      "eval_runtime": 3.4052,
+      "eval_samples_per_second": 8.81,
+      "eval_steps_per_second": 1.175,
       "step": 180
     },
     {
       "epoch": 13.0,
-      "grad_norm": 1.7895382642745972,
+      "grad_norm": 2.0414676666259766,
       "learning_rate": 3.4995533601559226e-05,
-      "loss": 0.1216,
+      "loss": 0.1367,
       "step": 195
     },
     {
       "epoch": 13.0,
-      "eval_loss": 0.09015119075775146,
-      "eval_runtime": 3.4988,
-      "eval_samples_per_second": 8.574,
-      "eval_steps_per_second": 1.143,
+      "eval_loss": 0.10060039162635803,
+      "eval_runtime": 3.409,
+      "eval_samples_per_second": 8.8,
+      "eval_steps_per_second": 1.173,
       "step": 195
     },
     {
       "epoch": 14.0,
-      "grad_norm": 1.532551646232605,
+      "grad_norm": 1.4828755855560303,
       "learning_rate": 3.272542485937369e-05,
-      "loss": 0.0751,
+      "loss": 0.0861,
       "step": 210
     },
     {
       "epoch": 14.0,
-      "eval_loss": 0.05144398286938667,
-      "eval_runtime": 3.5058,
-      "eval_samples_per_second": 8.557,
-      "eval_steps_per_second": 1.141,
+      "eval_loss": 0.06639143079519272,
+      "eval_runtime": 3.4048,
+      "eval_samples_per_second": 8.811,
+      "eval_steps_per_second": 1.175,
       "step": 210
     },
     {
       "epoch": 15.0,
-      "grad_norm": 1.421689748764038,
+      "grad_norm": 1.4082187414169312,
       "learning_rate": 3.0382440966446875e-05,
-      "loss": 0.042,
+      "loss": 0.0486,
       "step": 225
     },
     {
       "epoch": 15.0,
-      "eval_loss": 0.022208373993635178,
-      "eval_runtime": 3.4998,
-      "eval_samples_per_second": 8.572,
-      "eval_steps_per_second": 1.143,
+      "eval_loss": 0.02555031143128872,
+      "eval_runtime": 3.408,
+      "eval_samples_per_second": 8.803,
+      "eval_steps_per_second": 1.174,
       "step": 225
     },
     {
       "epoch": 16.0,
-      "grad_norm": 0.6442980766296387,
+      "grad_norm": 0.8797101974487305,
       "learning_rate": 2.7988683659251474e-05,
-      "loss": 0.0212,
+      "loss": 0.0264,
       "step": 240
     },
     {
       "epoch": 16.0,
-      "eval_loss": 0.008795789442956448,
-      "eval_runtime": 3.5099,
-      "eval_samples_per_second": 8.547,
-      "eval_steps_per_second": 1.14,
+      "eval_loss": 0.014887908473610878,
+      "eval_runtime": 3.4117,
+      "eval_samples_per_second": 8.793,
+      "eval_steps_per_second": 1.172,
       "step": 240
     },
     {
       "epoch": 17.0,
-      "grad_norm": 0.6399418711662292,
+      "grad_norm": 0.8584854602813721,
       "learning_rate": 2.556673362788225e-05,
-      "loss": 0.0116,
+      "loss": 0.0187,
       "step": 255
     },
     {
       "epoch": 17.0,
-      "eval_loss": 0.007612744811922312,
-      "eval_runtime": 3.5148,
-      "eval_samples_per_second": 8.535,
-      "eval_steps_per_second": 1.138,
+      "eval_loss": 0.008547956123948097,
+      "eval_runtime": 3.402,
+      "eval_samples_per_second": 8.818,
+      "eval_steps_per_second": 1.176,
       "step": 255
     },
     {
       "epoch": 18.0,
-      "grad_norm": 0.6358581781387329,
+      "grad_norm": 0.604076623916626,
       "learning_rate": 2.3139437508846155e-05,
-      "loss": 0.0069,
+      "loss": 0.0091,
       "step": 270
     },
     {
       "epoch": 18.0,
-      "eval_loss": 0.00314411916770041,
-      "eval_runtime": 3.5022,
-      "eval_samples_per_second": 8.566,
-      "eval_steps_per_second": 1.142,
+      "eval_loss": 0.004486155230551958,
+      "eval_runtime": 3.4077,
+      "eval_samples_per_second": 8.804,
+      "eval_steps_per_second": 1.174,
       "step": 270
     },
     {
       "epoch": 19.0,
-      "grad_norm": 0.09405054897069931,
+      "grad_norm": 0.48527106642723083,
       "learning_rate": 2.072969236913799e-05,
-      "loss": 0.0033,
+      "loss": 0.0053,
       "step": 285
     },
     {
       "epoch": 19.0,
-      "eval_loss": 0.002059183781966567,
-      "eval_runtime": 3.5115,
-      "eval_samples_per_second": 8.543,
-      "eval_steps_per_second": 1.139,
+      "eval_loss": 0.00216322997584939,
+      "eval_runtime": 3.4076,
+      "eval_samples_per_second": 8.804,
+      "eval_steps_per_second": 1.174,
       "step": 285
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 0.1501542329788208,
+      "learning_rate": 1.836022971459737e-05,
+      "loss": 0.002,
+      "step": 300
+    },
+    {
+      "epoch": 20.0,
+      "eval_loss": 0.0013952606823295355,
+      "eval_runtime": 3.4086,
+      "eval_samples_per_second": 8.801,
+      "eval_steps_per_second": 1.174,
+      "step": 300
+    },
+    {
+      "epoch": 21.0,
+      "grad_norm": 0.03647935017943382,
+      "learning_rate": 1.60534010600301e-05,
+      "loss": 0.0013,
+      "step": 315
+    },
+    {
+      "epoch": 21.0,
+      "eval_loss": 0.0011086437152698636,
+      "eval_runtime": 3.4025,
+      "eval_samples_per_second": 8.817,
+      "eval_steps_per_second": 1.176,
+      "step": 315
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 0.026030737906694412,
+      "learning_rate": 1.3830967083844942e-05,
+      "loss": 0.001,
+      "step": 330
+    },
+    {
+      "epoch": 22.0,
+      "eval_loss": 0.0009259060025215149,
+      "eval_runtime": 3.4043,
+      "eval_samples_per_second": 8.812,
+      "eval_steps_per_second": 1.175,
+      "step": 330
+    },
+    {
+      "epoch": 23.0,
+      "grad_norm": 0.021359996870160103,
+      "learning_rate": 1.1713892356143239e-05,
+      "loss": 0.0009,
+      "step": 345
+    },
+    {
+      "epoch": 23.0,
+      "eval_loss": 0.0008289433317258954,
+      "eval_runtime": 3.4102,
+      "eval_samples_per_second": 8.797,
+      "eval_steps_per_second": 1.173,
+      "step": 345
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 0.01979278400540352,
+      "learning_rate": 9.722147576623743e-06,
+      "loss": 0.0008,
+      "step": 360
+    },
+    {
+      "epoch": 24.0,
+      "eval_loss": 0.0007673827931284904,
+      "eval_runtime": 3.4084,
+      "eval_samples_per_second": 8.802,
+      "eval_steps_per_second": 1.174,
+      "step": 360
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 0.015766171738505363,
+      "learning_rate": 7.87452118782363e-06,
+      "loss": 0.0007,
+      "step": 375
+    },
+    {
+      "epoch": 25.0,
+      "eval_loss": 0.0007245048182085156,
+      "eval_runtime": 3.413,
+      "eval_samples_per_second": 8.79,
+      "eval_steps_per_second": 1.172,
+      "step": 375
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 0.01694435626268387,
+      "learning_rate": 6.188442140777742e-06,
+      "loss": 0.0007,
+      "step": 390
+    },
+    {
+      "epoch": 26.0,
+      "eval_loss": 0.0006994114373810589,
+      "eval_runtime": 3.4088,
+      "eval_samples_per_second": 8.801,
+      "eval_steps_per_second": 1.173,
+      "step": 390
+    },
+    {
+      "epoch": 27.0,
+      "grad_norm": 0.014102422632277012,
+      "learning_rate": 4.679815484975505e-06,
+      "loss": 0.0007,
+      "step": 405
+    },
+    {
+      "epoch": 27.0,
+      "eval_loss": 0.0006800881819799542,
+      "eval_runtime": 3.4112,
+      "eval_samples_per_second": 8.795,
+      "eval_steps_per_second": 1.173,
+      "step": 405
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 0.015260938555002213,
+      "learning_rate": 3.3628723335213885e-06,
+      "loss": 0.0007,
+      "step": 420
+    },
+    {
+      "epoch": 28.0,
+      "eval_loss": 0.0006676352932117879,
+      "eval_runtime": 3.4094,
+      "eval_samples_per_second": 8.799,
+      "eval_steps_per_second": 1.173,
+      "step": 420
+    },
+    {
+      "epoch": 29.0,
+      "grad_norm": 0.01491440087556839,
+      "learning_rate": 2.250035618801241e-06,
+      "loss": 0.0007,
+      "step": 435
+    },
+    {
+      "epoch": 29.0,
+      "eval_loss": 0.0006499449955299497,
+      "eval_runtime": 3.409,
+      "eval_samples_per_second": 8.8,
+      "eval_steps_per_second": 1.173,
+      "step": 435
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 0.014446167275309563,
+      "learning_rate": 1.351802905002386e-06,
+      "loss": 0.0006,
+      "step": 450
+    },
+    {
+      "epoch": 30.0,
+      "eval_loss": 0.0006467067869380116,
+      "eval_runtime": 3.4096,
+      "eval_samples_per_second": 8.799,
+      "eval_steps_per_second": 1.173,
+      "step": 450
+    },
+    {
+      "epoch": 31.0,
+      "grad_norm": 0.013305812142789364,
+      "learning_rate": 6.766473629355452e-07,
+      "loss": 0.0006,
+      "step": 465
+    },
+    {
+      "epoch": 31.0,
+      "eval_loss": 0.0006454469985328615,
+      "eval_runtime": 3.4011,
+      "eval_samples_per_second": 8.821,
+      "eval_steps_per_second": 1.176,
+      "step": 465
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 0.015049861744046211,
+      "learning_rate": 2.3093784127863062e-07,
+      "loss": 0.0006,
+      "step": 480
+    },
+    {
+      "epoch": 32.0,
+      "eval_loss": 0.0006429204368032515,
+      "eval_runtime": 3.404,
+      "eval_samples_per_second": 8.813,
+      "eval_steps_per_second": 1.175,
+      "step": 480
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 0.014723357744514942,
+      "learning_rate": 1.8878788223009036e-08,
+      "loss": 0.0006,
+      "step": 495
+    },
+    {
+      "epoch": 33.0,
+      "eval_loss": 0.0006475438713096082,
+      "eval_runtime": 3.4074,
+      "eval_samples_per_second": 8.804,
+      "eval_steps_per_second": 1.174,
+      "step": 495
+    },
+    {
+      "epoch": 33.333333333333336,
+      "grad_norm": 0.014316793531179428,
+      "learning_rate": 5.244749650301639e-10,
+      "loss": 0.0007,
+      "step": 500
+    },
+    {
+      "epoch": 33.333333333333336,
+      "eval_loss": 0.0006410721107386053,
+      "eval_runtime": 3.4172,
+      "eval_samples_per_second": 8.779,
+      "eval_steps_per_second": 1.171,
+      "step": 500
     }
   ],
   "logging_steps": 10,
@@ -307,12 +532,12 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.2993765576523776e+16,
+  "total_flos": 2.279189598799872e+16,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76427348473653c4b6e8520475a690655dcc9b305f7beb72a808945fcbdf04fb
+oid sha256:63ac9574a16bf682e538ba77e88c6fab9115c89e658abece48293457ececb1e0
 size 6033