Upload folder using huggingface_hub

adapter_config.json

@@ -12,21 +12,21 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 256,
+  "lora_alpha": 128,
   "lora_bias": false,
   "lora_dropout": 0.0,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 128,
+  "r": 64,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
     "qkv_proj",
+    "down_proj",
     "gate_up_proj",
-    "o_proj",
-    "down_proj"
+    "o_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:a304d75910dd0421e95eb1bd8b9470584dd6a41faaf2f2456b1247f80dddf925
-size 738232680
+oid sha256:99749dcc2cef2127a0028d6344c47cef7b7723bc9c66781cf98dba1ba1bf0002
+size 369133600

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3c04d67455664030c431864810ecb7f420e978f4745cfd77fed357a7d337cbf3
-size 1476611275
+oid sha256:9cd500f8476fd23fb4fd29c2d2d77e5e2c236753446fc12c3ad5abf9fcead93a
+size 738413771

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ac8091a108f9f7cd19dbe7d91d51c39045d769dcf6a14a778cf5cfeee0ee761e
+oid sha256:d2a5702fbdbf4947d84cc9a1ce591ebf6dc34ab54764e599b52b99cae31f4532
 size 14645

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f05b53c6c8b5fae09e3d46adb0f5cf2ced6d37c2dbb609562af4961873c13040
+oid sha256:e8083cb28f9178670969e8954135f0609f0eac2102b31887abccd667da0fb9cf
 size 1465

trainer_state.json

@@ -1,350 +1,170 @@
 {
-  "best_global_step": 330,
-  "best_metric": 0.0009259060025215149,
-  "best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-330",
-  "epoch": 22.0,
-  "eval_steps": 500,
-  "global_step": 330,
+  "best_global_step": 1000,
+  "best_metric": 3.814979936578311e-05,
+  "best_model_checkpoint": "/content/drive/MyDrive/lora_model/outputs/task15_microsoft/Phi-4-mini-instruct/checkpoint-1000",
+  "epoch": 250.0,
+  "eval_steps": 100,
+  "global_step": 1000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.0,
-      "grad_norm": 2.040301561355591,
-      "learning_rate": 4.666666666666667e-05,
-      "loss": 2.4299,
-      "step": 15
+      "epoch": 25.0,
+      "grad_norm": 1.5695021152496338,
+      "learning_rate": 4.9108140147495156e-05,
+      "loss": 0.8164,
+      "step": 100
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 1.4176747798919678,
-      "eval_runtime": 3.3557,
-      "eval_samples_per_second": 8.94,
-      "eval_steps_per_second": 1.192,
-      "step": 15
+      "epoch": 25.0,
+      "eval_loss": 0.008121663704514503,
+      "eval_runtime": 3.347,
+      "eval_samples_per_second": 8.963,
+      "eval_steps_per_second": 1.195,
+      "step": 100
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 1.0678095817565918,
-      "learning_rate": 4.989727297688797e-05,
-      "loss": 1.173,
-      "step": 30
+      "epoch": 50.0,
+      "grad_norm": 0.004555859602987766,
+      "learning_rate": 4.5817145782018976e-05,
+      "loss": 0.0022,
+      "step": 200
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 0.9685609340667725,
-      "eval_runtime": 3.4367,
-      "eval_samples_per_second": 8.729,
-      "eval_steps_per_second": 1.164,
-      "step": 30
+      "epoch": 50.0,
+      "eval_loss": 0.00019265948503743857,
+      "eval_runtime": 3.3476,
+      "eval_samples_per_second": 8.962,
+      "eval_steps_per_second": 1.195,
+      "step": 200
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 1.092650055885315,
-      "learning_rate": 4.956021052727731e-05,
-      "loss": 0.9354,
-      "step": 45
-    },
-    {
-      "epoch": 3.0,
-      "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": 1.079960823059082,
-      "learning_rate": 4.899146773943374e-05,
-      "loss": 0.8161,
-      "step": 60
-    },
-    {
-      "epoch": 4.0,
-      "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": 1.0952789783477783,
-      "learning_rate": 4.819640965374681e-05,
-      "loss": 0.7146,
-      "step": 75
-    },
-    {
-      "epoch": 5.0,
-      "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": 1.183589220046997,
-      "learning_rate": 4.718253617849306e-05,
-      "loss": 0.6222,
-      "step": 90
-    },
-    {
-      "epoch": 6.0,
-      "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": 1.408022165298462,
-      "learning_rate": 4.595941134201871e-05,
-      "loss": 0.5375,
-      "step": 105
-    },
-    {
-      "epoch": 7.0,
-      "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": 1.2591876983642578,
-      "learning_rate": 4.4538573073752365e-05,
-      "loss": 0.4583,
-      "step": 120
-    },
-    {
-      "epoch": 8.0,
-      "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": 1.3704065084457397,
-      "learning_rate": 4.2933424365097564e-05,
-      "loss": 0.3893,
-      "step": 135
-    },
-    {
-      "epoch": 9.0,
-      "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": 1.5167052745819092,
-      "learning_rate": 4.1159106836901674e-05,
-      "loss": 0.3179,
-      "step": 150
-    },
-    {
-      "epoch": 10.0,
-      "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": 1.8745532035827637,
-      "learning_rate": 3.923235790615907e-05,
-      "loss": 0.2582,
-      "step": 165
-    },
-    {
-      "epoch": 11.0,
-      "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": 2.4447438716888428,
-      "learning_rate": 3.717135289931774e-05,
-      "loss": 0.1923,
-      "step": 180
-    },
-    {
-      "epoch": 12.0,
-      "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": 2.0414676666259766,
-      "learning_rate": 3.4995533601559226e-05,
-      "loss": 0.1367,
-      "step": 195
-    },
-    {
-      "epoch": 13.0,
-      "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": 1.4828755855560303,
-      "learning_rate": 3.272542485937369e-05,
-      "loss": 0.0861,
-      "step": 210
-    },
-    {
-      "epoch": 14.0,
-      "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": 1.4082187414169312,
-      "learning_rate": 3.0382440966446875e-05,
-      "loss": 0.0486,
-      "step": 225
+      "epoch": 75.0,
+      "grad_norm": 0.0020954415667802095,
+      "learning_rate": 4.042641976420454e-05,
+      "loss": 0.0001,
+      "step": 300
     },
     {
-      "epoch": 15.0,
-      "eval_loss": 0.02555031143128872,
-      "eval_runtime": 3.408,
-      "eval_samples_per_second": 8.803,
-      "eval_steps_per_second": 1.174,
-      "step": 225
+      "epoch": 75.0,
+      "eval_loss": 0.00010092668526340276,
+      "eval_runtime": 3.3484,
+      "eval_samples_per_second": 8.96,
+      "eval_steps_per_second": 1.195,
+      "step": 300
     },
     {
-      "epoch": 16.0,
-      "grad_norm": 0.8797101974487305,
-      "learning_rate": 2.7988683659251474e-05,
-      "loss": 0.0264,
-      "step": 240
+      "epoch": 100.0,
+      "grad_norm": 0.0013948463601991534,
+      "learning_rate": 3.347970032633464e-05,
+      "loss": 0.0001,
+      "step": 400
     },
     {
-      "epoch": 16.0,
-      "eval_loss": 0.014887908473610878,
-      "eval_runtime": 3.4117,
-      "eval_samples_per_second": 8.793,
-      "eval_steps_per_second": 1.172,
-      "step": 240
+      "epoch": 100.0,
+      "eval_loss": 6.931247480679303e-05,
+      "eval_runtime": 3.3476,
+      "eval_samples_per_second": 8.962,
+      "eval_steps_per_second": 1.195,
+      "step": 400
     },
     {
-      "epoch": 17.0,
-      "grad_norm": 0.8584854602813721,
-      "learning_rate": 2.556673362788225e-05,
-      "loss": 0.0187,
-      "step": 255
+      "epoch": 125.0,
+      "grad_norm": 0.0010975338518619537,
+      "learning_rate": 2.567767174262942e-05,
+      "loss": 0.0001,
+      "step": 500
     },
     {
-      "epoch": 17.0,
-      "eval_loss": 0.008547956123948097,
-      "eval_runtime": 3.402,
-      "eval_samples_per_second": 8.818,
-      "eval_steps_per_second": 1.176,
-      "step": 255
+      "epoch": 125.0,
+      "eval_loss": 5.398171560955234e-05,
+      "eval_runtime": 3.3487,
+      "eval_samples_per_second": 8.959,
+      "eval_steps_per_second": 1.194,
+      "step": 500
     },
     {
-      "epoch": 18.0,
-      "grad_norm": 0.604076623916626,
-      "learning_rate": 2.3139437508846155e-05,
-      "loss": 0.0091,
-      "step": 270
+      "epoch": 150.0,
+      "grad_norm": 0.0008582558366470039,
+      "learning_rate": 1.7807289465156852e-05,
+      "loss": 0.0,
+      "step": 600
     },
     {
-      "epoch": 18.0,
-      "eval_loss": 0.004486155230551958,
-      "eval_runtime": 3.4077,
-      "eval_samples_per_second": 8.804,
-      "eval_steps_per_second": 1.174,
-      "step": 270
+      "epoch": 150.0,
+      "eval_loss": 4.567861833493225e-05,
+      "eval_runtime": 3.3716,
+      "eval_samples_per_second": 8.898,
+      "eval_steps_per_second": 1.186,
+      "step": 600
     },
     {
-      "epoch": 19.0,
-      "grad_norm": 0.48527106642723083,
-      "learning_rate": 2.072969236913799e-05,
-      "loss": 0.0053,
-      "step": 285
+      "epoch": 175.0,
+      "grad_norm": 0.0008014155901037157,
+      "learning_rate": 1.0662403474940072e-05,
+      "loss": 0.0,
+      "step": 700
     },
     {
-      "epoch": 19.0,
-      "eval_loss": 0.00216322997584939,
-      "eval_runtime": 3.4076,
-      "eval_samples_per_second": 8.804,
-      "eval_steps_per_second": 1.174,
-      "step": 285
+      "epoch": 175.0,
+      "eval_loss": 4.1445640817983076e-05,
+      "eval_runtime": 3.3696,
+      "eval_samples_per_second": 8.903,
+      "eval_steps_per_second": 1.187,
+      "step": 700
     },
     {
-      "epoch": 20.0,
-      "grad_norm": 0.1501542329788208,
-      "learning_rate": 1.836022971459737e-05,
-      "loss": 0.002,
-      "step": 300
+      "epoch": 200.0,
+      "grad_norm": 0.000805235467851162,
+      "learning_rate": 4.963686213009133e-06,
+      "loss": 0.0,
+      "step": 800
     },
     {
-      "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": 200.0,
+      "eval_loss": 3.887030470650643e-05,
+      "eval_runtime": 3.368,
+      "eval_samples_per_second": 8.907,
+      "eval_steps_per_second": 1.188,
+      "step": 800
     },
     {
-      "epoch": 21.0,
-      "grad_norm": 0.03647935017943382,
-      "learning_rate": 1.60534010600301e-05,
-      "loss": 0.0013,
-      "step": 315
+      "epoch": 225.0,
+      "grad_norm": 0.0007790795643813908,
+      "learning_rate": 1.2859416000311293e-06,
+      "loss": 0.0,
+      "step": 900
     },
     {
-      "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": 225.0,
+      "eval_loss": 3.820070196525194e-05,
+      "eval_runtime": 3.3697,
+      "eval_samples_per_second": 8.903,
+      "eval_steps_per_second": 1.187,
+      "step": 900
     },
     {
-      "epoch": 22.0,
-      "grad_norm": 0.026030737906694412,
-      "learning_rate": 1.3830967083844942e-05,
-      "loss": 0.001,
-      "step": 330
+      "epoch": 250.0,
+      "grad_norm": 0.0007654105429537594,
+      "learning_rate": 1.2715602095703104e-10,
+      "loss": 0.0,
+      "step": 1000
     },
     {
-      "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": 250.0,
+      "eval_loss": 3.814979936578311e-05,
+      "eval_runtime": 3.3687,
+      "eval_samples_per_second": 8.905,
+      "eval_steps_per_second": 1.187,
+      "step": 1000
     }
   ],
-  "logging_steps": 10,
-  "max_steps": 500,
+  "logging_steps": 100,
+  "max_steps": 1000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 34,
-  "save_steps": 500,
+  "num_train_epochs": 250,
+  "save_steps": 100,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -352,13 +172,13 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }
   },
-  "total_flos": 1.5057254808465408e+16,
-  "train_batch_size": 2,
+  "total_flos": 3.728409374976e+16,
+  "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:63ac9574a16bf682e538ba77e88c6fab9115c89e658abece48293457ececb1e0
+oid sha256:c6809084fc7322e88dbbc60a5ef23da7c8428195cb539247da17111c2913aee7
 size 6033