Training in progress, step 1000, checkpoint

checkpoint-1000/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ca03659b03a1532f811ae70c2603d003eb9f560cf9b68a9f5ecdf891f5439ebc
+oid sha256:06a7d7672c3a6bf984e7a9e4dbb49b56afb5f2515be6b87e99347eb956e544b5
 size 19690328

checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2544c332d987c9fcfc5c28cede911d573ca1ab713eca3f40306f4ada1602260e
+oid sha256:972cd5d7b13b46e435d937909060452e8f19c0e4798c3ff805c3367e1f808279
 size 38087162

checkpoint-1000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7769225d15b94262d795b404ade91d62b4b1a301871a308f6f8ee0af1de4cfee
+oid sha256:49480a4aeb6a555455d77377993190b79182d6aac36ad392ddce6669e836a2b2
 size 14244

checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ba37f82f1dc802fe5c0477cccd8ccc7dad1677bc00b8b4c523ff01f577155184
+oid sha256:9603a81ebc9be2a1d1c8d7bd738cc0f4675a43fdd998e5ab79e9e3c198b8e726
 size 1064

checkpoint-1000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.7192752361297607,
-  "best_model_checkpoint": "output/checkpoint-400",
-  "epoch": 5.105296745373325,
+  "best_metric": 0.45803794264793396,
+  "best_model_checkpoint": "output/checkpoint-150",
+  "epoch": 0.8021658477890303,
   "eval_steps": 50,
   "global_step": 1000,
   "is_hyper_param_search": false,
@@ -9,317 +9,317 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.01,
-      "learning_rate": 2.777777777777778e-07,
-      "loss": 9.5569,
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.6453,
       "step": 1
     },
     {
-      "epoch": 0.26,
-      "learning_rate": 1.25e-05,
-      "loss": 6.958,
+      "epoch": 0.04,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 1.3216,
       "step": 50
     },
     {
-      "epoch": 0.26,
-      "eval_accuracy": 0.42857142857142855,
-      "eval_loss": 6.581231117248535,
-      "eval_runtime": 28.2966,
-      "eval_samples_per_second": 6.184,
-      "eval_steps_per_second": 1.555,
+      "epoch": 0.04,
+      "eval_accuracy": 0.8057142857142857,
+      "eval_loss": 0.6264824271202087,
+      "eval_runtime": 28.8159,
+      "eval_samples_per_second": 6.073,
+      "eval_steps_per_second": 1.527,
       "step": 50
     },
     {
-      "epoch": 0.51,
-      "learning_rate": 2.6388888888888892e-05,
-      "loss": 1.6141,
+      "epoch": 0.08,
+      "learning_rate": 2.088888888888889e-05,
+      "loss": 0.918,
       "step": 100
     },
     {
-      "epoch": 0.51,
-      "eval_accuracy": 0.5028571428571429,
-      "eval_loss": 1.6640697717666626,
-      "eval_runtime": 28.1139,
-      "eval_samples_per_second": 6.225,
-      "eval_steps_per_second": 1.565,
+      "epoch": 0.08,
+      "eval_accuracy": 0.7885714285714286,
+      "eval_loss": 0.49068325757980347,
+      "eval_runtime": 26.1642,
+      "eval_samples_per_second": 6.689,
+      "eval_steps_per_second": 1.682,
       "step": 100
     },
     {
-      "epoch": 0.77,
-      "learning_rate": 4.027777777777778e-05,
-      "loss": 0.8533,
+      "epoch": 0.12,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.7219,
       "step": 150
     },
     {
-      "epoch": 0.77,
-      "eval_accuracy": 0.6342857142857142,
-      "eval_loss": 1.2631328105926514,
-      "eval_runtime": 26.205,
-      "eval_samples_per_second": 6.678,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.12,
+      "eval_accuracy": 0.7942857142857143,
+      "eval_loss": 0.45803794264793396,
+      "eval_runtime": 26.1308,
+      "eval_samples_per_second": 6.697,
+      "eval_steps_per_second": 1.684,
       "step": 150
     },
     {
-      "epoch": 1.02,
-      "learning_rate": 4.908536585365854e-05,
-      "loss": 0.4474,
+      "epoch": 0.16,
+      "learning_rate": 4.311111111111111e-05,
+      "loss": 0.6619,
       "step": 200
     },
     {
-      "epoch": 1.02,
-      "eval_accuracy": 0.7142857142857143,
-      "eval_loss": 0.9036799073219299,
-      "eval_runtime": 26.208,
-      "eval_samples_per_second": 6.677,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.16,
+      "eval_accuracy": 0.8285714285714286,
+      "eval_loss": 0.4676758646965027,
+      "eval_runtime": 26.1149,
+      "eval_samples_per_second": 6.701,
+      "eval_steps_per_second": 1.685,
       "step": 200
     },
     {
-      "epoch": 1.28,
-      "learning_rate": 4.603658536585366e-05,
-      "loss": 0.2996,
+      "epoch": 0.2,
+      "learning_rate": 4.906953966699315e-05,
+      "loss": 0.633,
       "step": 250
     },
     {
-      "epoch": 1.28,
-      "eval_accuracy": 0.6742857142857143,
-      "eval_loss": 1.1182247400283813,
-      "eval_runtime": 26.2071,
-      "eval_samples_per_second": 6.678,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.2,
+      "eval_accuracy": 0.8114285714285714,
+      "eval_loss": 0.4800652265548706,
+      "eval_runtime": 26.1161,
+      "eval_samples_per_second": 6.701,
+      "eval_steps_per_second": 1.685,
       "step": 250
     },
     {
-      "epoch": 1.53,
-      "learning_rate": 4.298780487804878e-05,
-      "loss": 0.2403,
+      "epoch": 0.24,
+      "learning_rate": 4.66209598432909e-05,
+      "loss": 0.6147,
       "step": 300
     },
     {
-      "epoch": 1.53,
-      "eval_accuracy": 0.72,
-      "eval_loss": 1.076162338256836,
-      "eval_runtime": 26.2142,
-      "eval_samples_per_second": 6.676,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.24,
+      "eval_accuracy": 0.7657142857142857,
+      "eval_loss": 0.5112562775611877,
+      "eval_runtime": 26.1127,
+      "eval_samples_per_second": 6.702,
+      "eval_steps_per_second": 1.685,
       "step": 300
     },
     {
-      "epoch": 1.79,
-      "learning_rate": 3.9939024390243905e-05,
-      "loss": 0.1982,
+      "epoch": 0.28,
+      "learning_rate": 4.4172380019588636e-05,
+      "loss": 0.6355,
       "step": 350
     },
     {
-      "epoch": 1.79,
-      "eval_accuracy": 0.6571428571428571,
-      "eval_loss": 1.4400347471237183,
-      "eval_runtime": 26.2187,
-      "eval_samples_per_second": 6.675,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.28,
+      "eval_accuracy": 0.7828571428571428,
+      "eval_loss": 0.5210281610488892,
+      "eval_runtime": 26.1048,
+      "eval_samples_per_second": 6.704,
+      "eval_steps_per_second": 1.686,
       "step": 350
     },
     {
-      "epoch": 2.04,
-      "learning_rate": 3.6890243902439025e-05,
-      "loss": 0.1926,
+      "epoch": 0.32,
+      "learning_rate": 4.1723800195886386e-05,
+      "loss": 0.593,
       "step": 400
     },
     {
-      "epoch": 2.04,
-      "eval_accuracy": 0.7942857142857143,
-      "eval_loss": 0.7192752361297607,
-      "eval_runtime": 26.2201,
-      "eval_samples_per_second": 6.674,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.32,
+      "eval_accuracy": 0.7885714285714286,
+      "eval_loss": 0.5098910927772522,
+      "eval_runtime": 26.1072,
+      "eval_samples_per_second": 6.703,
+      "eval_steps_per_second": 1.685,
       "step": 400
     },
     {
-      "epoch": 2.3,
-      "learning_rate": 3.384146341463415e-05,
-      "loss": 0.1258,
+      "epoch": 0.36,
+      "learning_rate": 3.927522037218414e-05,
+      "loss": 0.6047,
       "step": 450
     },
     {
-      "epoch": 2.3,
-      "eval_accuracy": 0.7657142857142857,
-      "eval_loss": 1.0054305791854858,
-      "eval_runtime": 26.2093,
-      "eval_samples_per_second": 6.677,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.36,
+      "eval_accuracy": 0.7314285714285714,
+      "eval_loss": 0.5602952837944031,
+      "eval_runtime": 26.1116,
+      "eval_samples_per_second": 6.702,
+      "eval_steps_per_second": 1.685,
       "step": 450
     },
     {
-      "epoch": 2.55,
-      "learning_rate": 3.079268292682927e-05,
-      "loss": 0.0857,
+      "epoch": 0.4,
+      "learning_rate": 3.682664054848188e-05,
+      "loss": 0.5661,
       "step": 500
     },
     {
-      "epoch": 2.55,
-      "eval_accuracy": 0.7828571428571428,
-      "eval_loss": 1.1228740215301514,
-      "eval_runtime": 26.1929,
-      "eval_samples_per_second": 6.681,
-      "eval_steps_per_second": 1.68,
+      "epoch": 0.4,
+      "eval_accuracy": 0.7542857142857143,
+      "eval_loss": 0.5670996904373169,
+      "eval_runtime": 26.1104,
+      "eval_samples_per_second": 6.702,
+      "eval_steps_per_second": 1.685,
       "step": 500
     },
     {
-      "epoch": 2.81,
-      "learning_rate": 2.7743902439024393e-05,
-      "loss": 0.1061,
+      "epoch": 0.44,
+      "learning_rate": 3.437806072477963e-05,
+      "loss": 0.6111,
       "step": 550
     },
     {
-      "epoch": 2.81,
-      "eval_accuracy": 0.7657142857142857,
-      "eval_loss": 1.2351529598236084,
-      "eval_runtime": 26.2161,
-      "eval_samples_per_second": 6.675,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.44,
+      "eval_accuracy": 0.7714285714285715,
+      "eval_loss": 0.5074692368507385,
+      "eval_runtime": 26.1037,
+      "eval_samples_per_second": 6.704,
+      "eval_steps_per_second": 1.686,
       "step": 550
     },
     {
-      "epoch": 3.06,
-      "learning_rate": 2.4695121951219512e-05,
-      "loss": 0.103,
+      "epoch": 0.48,
+      "learning_rate": 3.1929480901077375e-05,
+      "loss": 0.5788,
       "step": 600
     },
     {
-      "epoch": 3.06,
-      "eval_accuracy": 0.7542857142857143,
-      "eval_loss": 1.413483738899231,
-      "eval_runtime": 26.2198,
-      "eval_samples_per_second": 6.674,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.48,
+      "eval_accuracy": 0.7485714285714286,
+      "eval_loss": 0.6720603108406067,
+      "eval_runtime": 26.1036,
+      "eval_samples_per_second": 6.704,
+      "eval_steps_per_second": 1.686,
       "step": 600
     },
     {
-      "epoch": 3.32,
-      "learning_rate": 2.1646341463414634e-05,
-      "loss": 0.0588,
+      "epoch": 0.52,
+      "learning_rate": 2.9480901077375122e-05,
+      "loss": 0.562,
       "step": 650
     },
     {
-      "epoch": 3.32,
-      "eval_accuracy": 0.7542857142857143,
-      "eval_loss": 1.685744285583496,
-      "eval_runtime": 26.2164,
-      "eval_samples_per_second": 6.675,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.52,
+      "eval_accuracy": 0.7942857142857143,
+      "eval_loss": 0.5063876509666443,
+      "eval_runtime": 26.1086,
+      "eval_samples_per_second": 6.703,
+      "eval_steps_per_second": 1.685,
       "step": 650
     },
     {
-      "epoch": 3.57,
-      "learning_rate": 1.8597560975609757e-05,
-      "loss": 0.0871,
+      "epoch": 0.56,
+      "learning_rate": 2.7032321253672872e-05,
+      "loss": 0.5472,
       "step": 700
     },
     {
-      "epoch": 3.57,
-      "eval_accuracy": 0.7885714285714286,
-      "eval_loss": 1.5099748373031616,
-      "eval_runtime": 26.2038,
-      "eval_samples_per_second": 6.678,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.56,
+      "eval_accuracy": 0.7371428571428571,
+      "eval_loss": 0.5650284886360168,
+      "eval_runtime": 26.1065,
+      "eval_samples_per_second": 6.703,
+      "eval_steps_per_second": 1.685,
       "step": 700
     },
     {
-      "epoch": 3.83,
-      "learning_rate": 1.554878048780488e-05,
-      "loss": 0.0539,
+      "epoch": 0.6,
+      "learning_rate": 2.458374142997062e-05,
+      "loss": 0.551,
       "step": 750
     },
     {
-      "epoch": 3.83,
-      "eval_accuracy": 0.7714285714285715,
-      "eval_loss": 1.7970468997955322,
-      "eval_runtime": 26.2034,
-      "eval_samples_per_second": 6.679,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.6,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.5052895545959473,
+      "eval_runtime": 26.1094,
+      "eval_samples_per_second": 6.703,
+      "eval_steps_per_second": 1.685,
       "step": 750
     },
     {
-      "epoch": 4.08,
-      "learning_rate": 1.25e-05,
-      "loss": 0.0339,
+      "epoch": 0.64,
+      "learning_rate": 2.2135161606268366e-05,
+      "loss": 0.5259,
       "step": 800
     },
     {
-      "epoch": 4.08,
-      "eval_accuracy": 0.7714285714285715,
-      "eval_loss": 1.7582976818084717,
-      "eval_runtime": 26.2084,
-      "eval_samples_per_second": 6.677,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.64,
+      "eval_accuracy": 0.7485714285714286,
+      "eval_loss": 0.5384859442710876,
+      "eval_runtime": 26.0891,
+      "eval_samples_per_second": 6.708,
+      "eval_steps_per_second": 1.687,
       "step": 800
     },
     {
-      "epoch": 4.34,
-      "learning_rate": 9.451219512195123e-06,
-      "loss": 0.0279,
+      "epoch": 0.68,
+      "learning_rate": 1.9686581782566114e-05,
+      "loss": 0.5231,
       "step": 850
     },
     {
-      "epoch": 4.34,
-      "eval_accuracy": 0.7714285714285715,
-      "eval_loss": 1.7328156232833862,
-      "eval_runtime": 26.2197,
-      "eval_samples_per_second": 6.674,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.68,
+      "eval_accuracy": 0.7371428571428571,
+      "eval_loss": 0.5821840763092041,
+      "eval_runtime": 26.0996,
+      "eval_samples_per_second": 6.705,
+      "eval_steps_per_second": 1.686,
       "step": 850
     },
     {
-      "epoch": 4.59,
-      "learning_rate": 6.402439024390244e-06,
-      "loss": 0.0271,
+      "epoch": 0.72,
+      "learning_rate": 1.7238001958863857e-05,
+      "loss": 0.4987,
       "step": 900
     },
     {
-      "epoch": 4.59,
-      "eval_accuracy": 0.76,
-      "eval_loss": 2.002990484237671,
-      "eval_runtime": 26.219,
-      "eval_samples_per_second": 6.675,
-      "eval_steps_per_second": 1.678,
+      "epoch": 0.72,
+      "eval_accuracy": 0.7142857142857143,
+      "eval_loss": 0.5697786211967468,
+      "eval_runtime": 26.0884,
+      "eval_samples_per_second": 6.708,
+      "eval_steps_per_second": 1.687,
       "step": 900
     },
     {
-      "epoch": 4.85,
-      "learning_rate": 3.3536585365853664e-06,
-      "loss": 0.0324,
+      "epoch": 0.76,
+      "learning_rate": 1.4789422135161608e-05,
+      "loss": 0.5019,
       "step": 950
     },
     {
-      "epoch": 4.85,
-      "eval_accuracy": 0.7485714285714286,
-      "eval_loss": 2.0242903232574463,
-      "eval_runtime": 26.2002,
-      "eval_samples_per_second": 6.679,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.76,
+      "eval_accuracy": 0.7657142857142857,
+      "eval_loss": 0.5529425740242004,
+      "eval_runtime": 26.1154,
+      "eval_samples_per_second": 6.701,
+      "eval_steps_per_second": 1.685,
       "step": 950
     },
     {
-      "epoch": 5.11,
-      "learning_rate": 3.0487804878048784e-07,
-      "loss": 0.0372,
+      "epoch": 0.8,
+      "learning_rate": 1.2340842311459353e-05,
+      "loss": 0.515,
       "step": 1000
     },
     {
-      "epoch": 5.11,
-      "eval_accuracy": 0.76,
-      "eval_loss": 1.926530361175537,
-      "eval_runtime": 26.2122,
-      "eval_samples_per_second": 6.676,
-      "eval_steps_per_second": 1.679,
+      "epoch": 0.8,
+      "eval_accuracy": 0.7314285714285714,
+      "eval_loss": 0.571620762348175,
+      "eval_runtime": 26.0898,
+      "eval_samples_per_second": 6.708,
+      "eval_steps_per_second": 1.686,
       "step": 1000
     }
   ],
   "logging_steps": 50,
-  "max_steps": 1000,
-  "num_train_epochs": 6,
+  "max_steps": 1246,
+  "num_train_epochs": 1,
   "save_steps": 50,
-  "total_flos": 3.7295030211446784e+17,
+  "total_flos": 3.7102353838669824e+17,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a0ef36d7c31265f68a3edc8ae53ce9272dcc90a79f261e29a2a46e040c77a33
+oid sha256:2b7c1677f66b49d74d914a320393e2fa29530a9670d5824a30beac9590468d16
 size 4600