Training in progress, step 50, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:217d3946b7803902022eec7cfd8fe9488050ab79dfd3fbe7b1875bbd6d1ae713
+oid sha256:0baf8aebd0d315fca0a75c5249758710a0ae88f24e69941b6c0dfa70d70a065b
 size 847599616

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d3bd819f5f239a35fa547eccf498f447b392f7befceb13f93b74e1b0076ad315
+oid sha256:c6f89651ce6488a74e905c7535a0e9e4be29248063c5ce092a76f4964cc72d9e
 size 1386414411

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f0c3a7585079a88f7e60217509a0b7358c2fc795f19cc94faeed7068f7d0ce34
+oid sha256:3aa39656b4a39e592ac10990c9c5819213db01eda0c44a0561e566e469c0a6e7
 size 14917

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:692fc434e90b39dd1c760af3bb9bd206fa31fb36c8e62d4b9a61187caf62dd58
+oid sha256:16b53619cb66bd690ba5e205b33c7d37deaa9bf03570e020bba1d0093af3aa44
 size 14917

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:055103685036cd172833dfba9d67c8c5638fe5ea929e5909dfadcb533f6d9014
+oid sha256:47f0e4a0950d20841a2c1d04b40779122c778aedec691c8d056f034b68e3aedb
 size 1465

last-checkpoint/trainer_state.json

@@ -2,278 +2,134 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.14,
-  "eval_steps": 5,
-  "global_step": 140,
+  "epoch": 0.025,
+  "eval_steps": 10,
+  "global_step": 50,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.005,
-      "eval_accuracy": 0.05346666666666667,
-      "eval_loss": 20.794218063354492,
-      "eval_runtime": 35.3413,
-      "eval_samples_per_second": 14.148,
-      "eval_steps_per_second": 7.074,
+      "epoch": 0.0025,
+      "grad_norm": 7.581125259399414,
+      "learning_rate": 1.6e-07,
+      "loss": 86.44205322265626,
       "step": 5
     },
     {
-      "epoch": 0.01,
-      "eval_accuracy": 0.07933333333333334,
-      "eval_loss": 19.725038528442383,
-      "eval_runtime": 35.9943,
-      "eval_samples_per_second": 13.891,
-      "eval_steps_per_second": 6.946,
+      "epoch": 0.005,
+      "grad_norm": 870.7899169921875,
+      "learning_rate": 3.6e-07,
+      "loss": 305.9718505859375,
       "step": 10
     },
     {
-      "epoch": 0.015,
-      "eval_accuracy": 0.087,
-      "eval_loss": 18.806058883666992,
-      "eval_runtime": 40.1906,
-      "eval_samples_per_second": 12.441,
-      "eval_steps_per_second": 6.22,
-      "step": 15
-    },
-    {
-      "epoch": 0.02,
-      "eval_accuracy": 0.08553333333333334,
-      "eval_loss": 18.3583984375,
-      "eval_runtime": 36.0213,
-      "eval_samples_per_second": 13.881,
-      "eval_steps_per_second": 6.94,
-      "step": 20
-    },
-    {
-      "epoch": 0.025,
-      "eval_accuracy": 0.08153333333333333,
-      "eval_loss": 18.101879119873047,
-      "eval_runtime": 41.6916,
-      "eval_samples_per_second": 11.993,
-      "eval_steps_per_second": 5.996,
-      "step": 25
-    },
-    {
-      "epoch": 0.03,
-      "eval_accuracy": 0.07606666666666667,
-      "eval_loss": 17.941076278686523,
-      "eval_runtime": 36.7785,
-      "eval_samples_per_second": 13.595,
-      "eval_steps_per_second": 6.797,
-      "step": 30
-    },
-    {
-      "epoch": 0.035,
-      "eval_accuracy": 0.08273333333333334,
-      "eval_loss": 17.723947525024414,
-      "eval_runtime": 40.5439,
-      "eval_samples_per_second": 12.332,
-      "eval_steps_per_second": 6.166,
-      "step": 35
-    },
-    {
-      "epoch": 0.04,
-      "eval_accuracy": 0.08126666666666667,
-      "eval_loss": 17.56175994873047,
-      "eval_runtime": 35.5502,
-      "eval_samples_per_second": 14.065,
-      "eval_steps_per_second": 7.032,
-      "step": 40
-    },
-    {
-      "epoch": 0.045,
-      "eval_accuracy": 0.08,
-      "eval_loss": 17.587928771972656,
-      "eval_runtime": 40.4006,
-      "eval_samples_per_second": 12.376,
-      "eval_steps_per_second": 6.188,
-      "step": 45
-    },
-    {
-      "epoch": 0.05,
-      "eval_accuracy": 0.07426666666666666,
-      "eval_loss": 17.396371841430664,
-      "eval_runtime": 35.5138,
-      "eval_samples_per_second": 14.079,
-      "eval_steps_per_second": 7.04,
-      "step": 50
-    },
-    {
-      "epoch": 0.055,
-      "eval_accuracy": 0.08266666666666667,
-      "eval_loss": 17.291400909423828,
-      "eval_runtime": 40.6248,
-      "eval_samples_per_second": 12.308,
-      "eval_steps_per_second": 6.154,
-      "step": 55
-    },
-    {
-      "epoch": 0.06,
-      "eval_accuracy": 0.07613333333333333,
-      "eval_loss": 17.195777893066406,
-      "eval_runtime": 36.5259,
-      "eval_samples_per_second": 13.689,
-      "eval_steps_per_second": 6.844,
-      "step": 60
-    },
-    {
-      "epoch": 0.065,
-      "eval_accuracy": 0.07053333333333334,
-      "eval_loss": 17.156831741333008,
-      "eval_runtime": 40.9838,
-      "eval_samples_per_second": 12.2,
-      "eval_steps_per_second": 6.1,
-      "step": 65
-    },
-    {
-      "epoch": 0.07,
-      "eval_accuracy": 0.07253333333333334,
-      "eval_loss": 17.09288215637207,
-      "eval_runtime": 35.8052,
-      "eval_samples_per_second": 13.964,
-      "eval_steps_per_second": 6.982,
-      "step": 70
-    },
-    {
-      "epoch": 0.075,
-      "eval_accuracy": 0.07566666666666666,
-      "eval_loss": 17.033140182495117,
-      "eval_runtime": 39.2943,
-      "eval_samples_per_second": 12.725,
-      "eval_steps_per_second": 6.362,
-      "step": 75
-    },
-    {
-      "epoch": 0.08,
-      "eval_accuracy": 0.07093333333333333,
-      "eval_loss": 16.976137161254883,
-      "eval_runtime": 35.688,
-      "eval_samples_per_second": 14.01,
-      "eval_steps_per_second": 7.005,
-      "step": 80
-    },
-    {
-      "epoch": 0.085,
-      "eval_accuracy": 0.0764,
-      "eval_loss": 16.9167537689209,
-      "eval_runtime": 40.3553,
-      "eval_samples_per_second": 12.39,
-      "eval_steps_per_second": 6.195,
-      "step": 85
+      "epoch": 0.005,
+      "eval_accuracy": 0.029019607843137254,
+      "eval_loss": 68.04701232910156,
+      "eval_runtime": 35.4394,
+      "eval_samples_per_second": 14.109,
+      "eval_steps_per_second": 1.778,
+      "step": 10
     },
     {
-      "epoch": 0.09,
-      "eval_accuracy": 0.0748,
-      "eval_loss": 16.922040939331055,
-      "eval_runtime": 36.0637,
-      "eval_samples_per_second": 13.864,
-      "eval_steps_per_second": 6.932,
-      "step": 90
+      "epoch": 0.0075,
+      "grad_norm": 225.20162963867188,
+      "learning_rate": 5.6e-07,
+      "loss": 179.27216796875,
+      "step": 15
     },
     {
-      "epoch": 0.095,
-      "eval_accuracy": 0.06873333333333333,
-      "eval_loss": 16.8882999420166,
-      "eval_runtime": 40.7406,
-      "eval_samples_per_second": 12.273,
-      "eval_steps_per_second": 6.136,
-      "step": 95
+      "epoch": 0.01,
+      "grad_norm": 79.13408660888672,
+      "learning_rate": 7.599999999999999e-07,
+      "loss": 130.82646484375,
+      "step": 20
     },
     {
-      "epoch": 0.1,
-      "grad_norm": 3.5192229747772217,
-      "learning_rate": 3.973800426880847e-06,
-      "loss": 144.29556640625,
-      "step": 100
+      "epoch": 0.01,
+      "eval_accuracy": 0.02961176470588235,
+      "eval_loss": 31.59725570678711,
+      "eval_runtime": 36.9264,
+      "eval_samples_per_second": 13.54,
+      "eval_steps_per_second": 1.706,
+      "step": 20
     },
     {
-      "epoch": 0.1,
-      "eval_accuracy": 0.07646666666666667,
-      "eval_loss": 16.831998825073242,
-      "eval_runtime": 35.9838,
-      "eval_samples_per_second": 13.895,
-      "eval_steps_per_second": 6.948,
-      "step": 100
+      "epoch": 0.0125,
+      "grad_norm": 72.2133560180664,
+      "learning_rate": 9.6e-07,
+      "loss": 124.6850830078125,
+      "step": 25
     },
     {
-      "epoch": 0.105,
-      "eval_accuracy": 0.075,
-      "eval_loss": 16.776647567749023,
-      "eval_runtime": 40.0532,
-      "eval_samples_per_second": 12.483,
-      "eval_steps_per_second": 6.242,
-      "step": 105
+      "epoch": 0.015,
+      "grad_norm": 94.8714828491211,
+      "learning_rate": 1.16e-06,
+      "loss": 123.9353515625,
+      "step": 30
     },
     {
-      "epoch": 0.11,
-      "eval_accuracy": 0.07446666666666667,
-      "eval_loss": 16.77096176147461,
-      "eval_runtime": 35.8224,
-      "eval_samples_per_second": 13.958,
-      "eval_steps_per_second": 6.979,
-      "step": 110
+      "epoch": 0.015,
+      "eval_accuracy": 0.04030196078431372,
+      "eval_loss": 30.649425506591797,
+      "eval_runtime": 38.2598,
+      "eval_samples_per_second": 13.069,
+      "eval_steps_per_second": 1.647,
+      "step": 30
     },
     {
-      "epoch": 0.115,
-      "eval_accuracy": 0.07686666666666667,
-      "eval_loss": 16.71495246887207,
-      "eval_runtime": 40.5454,
-      "eval_samples_per_second": 12.332,
-      "eval_steps_per_second": 6.166,
-      "step": 115
+      "epoch": 0.0175,
+      "grad_norm": 54.928829193115234,
+      "learning_rate": 1.3600000000000001e-06,
+      "loss": 125.992138671875,
+      "step": 35
     },
     {
-      "epoch": 0.12,
-      "eval_accuracy": 0.07413333333333333,
-      "eval_loss": 16.717641830444336,
-      "eval_runtime": 34.8477,
-      "eval_samples_per_second": 14.348,
-      "eval_steps_per_second": 7.174,
-      "step": 120
+      "epoch": 0.02,
+      "grad_norm": 70.58757019042969,
+      "learning_rate": 1.5599999999999999e-06,
+      "loss": 120.77716064453125,
+      "step": 40
     },
     {
-      "epoch": 0.125,
-      "eval_accuracy": 0.0676,
-      "eval_loss": 16.717758178710938,
-      "eval_runtime": 39.1101,
-      "eval_samples_per_second": 12.784,
-      "eval_steps_per_second": 6.392,
-      "step": 125
+      "epoch": 0.02,
+      "eval_accuracy": 0.04015686274509804,
+      "eval_loss": 29.560930252075195,
+      "eval_runtime": 37.9793,
+      "eval_samples_per_second": 13.165,
+      "eval_steps_per_second": 1.659,
+      "step": 40
     },
     {
-      "epoch": 0.13,
-      "eval_accuracy": 0.07173333333333333,
-      "eval_loss": 16.712247848510742,
-      "eval_runtime": 35.0018,
-      "eval_samples_per_second": 14.285,
-      "eval_steps_per_second": 7.142,
-      "step": 130
+      "epoch": 0.0225,
+      "grad_norm": 81.62490844726562,
+      "learning_rate": 1.7599999999999999e-06,
+      "loss": 121.851220703125,
+      "step": 45
     },
     {
-      "epoch": 0.135,
-      "eval_accuracy": 0.0726,
-      "eval_loss": 16.65378189086914,
-      "eval_runtime": 39.4554,
-      "eval_samples_per_second": 12.673,
-      "eval_steps_per_second": 6.336,
-      "step": 135
+      "epoch": 0.025,
+      "grad_norm": 65.93287658691406,
+      "learning_rate": 1.96e-06,
+      "loss": 120.95302734375,
+      "step": 50
     },
     {
-      "epoch": 0.14,
-      "eval_accuracy": 0.0776,
-      "eval_loss": 16.60840606689453,
-      "eval_runtime": 35.1512,
-      "eval_samples_per_second": 14.224,
-      "eval_steps_per_second": 7.112,
-      "step": 140
+      "epoch": 0.025,
+      "eval_accuracy": 0.04131764705882353,
+      "eval_loss": 28.05109977722168,
+      "eval_runtime": 38.147,
+      "eval_samples_per_second": 13.107,
+      "eval_steps_per_second": 1.652,
+      "step": 50
     }
   ],
-  "logging_steps": 100,
-  "max_steps": 1000,
+  "logging_steps": 5,
+  "max_steps": 2000,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
-  "save_steps": 10,
+  "save_steps": 50,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -287,7 +143,7 @@
     }
   },
   "total_flos": 0.0,
-  "train_batch_size": 1,
+  "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9bdaf07830b33937d6f32edfcc276b67b7d1e1e7aade843e68aded14ec567bf9
+oid sha256:57ed8b4eb49e711b8df54cadbcace7d2d15a05286403e857af5f3e540c8ca416
 size 5265