Update to 3rd epoch (#3)

- Update to 3rd epoch (d0952bd6981f1ea64fbf5c45bfac4813048441c2)

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f60ab1968e3383e1d0d99a3068cc74957a84c043bf5eda03d22efd0f8ea9f8f9
+oid sha256:e74d42214f527d418c0910ec128f6d3b5f1523570670c171baf66d4a0b95cefb
 size 134263757

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:81e5d1b0fcd6e6a2841ac0a59753960210a56af903a223d6b0a46462ec2c9c96
+oid sha256:25a3db7ff130b07d4e4b23a83e562d5a4ef3cf845f13e971f940342e4037d75b
 size 268514565

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:757b82c50c74e12673c2fa57a7a74b1902252dcab25b574e433c6ead4b7d6d80
+oid sha256:846c8f223a95d3e869043e3bb0dfb8a2e257ae81486c479822c9e12ad0bacd7c
 size 14575

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:48d79510a393105d03ace4b25923bb16766b9918be883f120389bf31e783c069
+oid sha256:cb57538b0fddb988f32d5e33311bcf25efee1aa4001ec3e33ef4a2dd884d77d3
 size 627

trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 7.230657989877079,
-  "global_step": 5000,
+  "epoch": 4.3383947939262475,
+  "global_step": 3000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1806,1211 +1806,11 @@
       "learning_rate": 0.0002,
       "loss": 0.8549,
       "step": 3000
-    },
-    {
-      "epoch": 4.35,
-      "learning_rate": 0.0002,
-      "loss": 0.8248,
-      "step": 3010
-    },
-    {
-      "epoch": 4.37,
-      "learning_rate": 0.0002,
-      "loss": 0.8262,
-      "step": 3020
-    },
-    {
-      "epoch": 4.38,
-      "learning_rate": 0.0002,
-      "loss": 0.9076,
-      "step": 3030
-    },
-    {
-      "epoch": 4.4,
-      "learning_rate": 0.0002,
-      "loss": 0.8706,
-      "step": 3040
-    },
-    {
-      "epoch": 4.41,
-      "learning_rate": 0.0002,
-      "loss": 0.8503,
-      "step": 3050
-    },
-    {
-      "epoch": 4.43,
-      "learning_rate": 0.0002,
-      "loss": 0.8351,
-      "step": 3060
-    },
-    {
-      "epoch": 4.44,
-      "learning_rate": 0.0002,
-      "loss": 0.8267,
-      "step": 3070
-    },
-    {
-      "epoch": 4.45,
-      "learning_rate": 0.0002,
-      "loss": 0.9008,
-      "step": 3080
-    },
-    {
-      "epoch": 4.47,
-      "learning_rate": 0.0002,
-      "loss": 0.8624,
-      "step": 3090
-    },
-    {
-      "epoch": 4.48,
-      "learning_rate": 0.0002,
-      "loss": 0.8703,
-      "step": 3100
-    },
-    {
-      "epoch": 4.5,
-      "learning_rate": 0.0002,
-      "loss": 0.8284,
-      "step": 3110
-    },
-    {
-      "epoch": 4.51,
-      "learning_rate": 0.0002,
-      "loss": 0.7899,
-      "step": 3120
-    },
-    {
-      "epoch": 4.53,
-      "learning_rate": 0.0002,
-      "loss": 0.9195,
-      "step": 3130
-    },
-    {
-      "epoch": 4.54,
-      "learning_rate": 0.0002,
-      "loss": 0.881,
-      "step": 3140
-    },
-    {
-      "epoch": 4.56,
-      "learning_rate": 0.0002,
-      "loss": 0.8647,
-      "step": 3150
-    },
-    {
-      "epoch": 4.57,
-      "learning_rate": 0.0002,
-      "loss": 0.8115,
-      "step": 3160
-    },
-    {
-      "epoch": 4.58,
-      "learning_rate": 0.0002,
-      "loss": 0.8282,
-      "step": 3170
-    },
-    {
-      "epoch": 4.6,
-      "learning_rate": 0.0002,
-      "loss": 0.9087,
-      "step": 3180
-    },
-    {
-      "epoch": 4.61,
-      "learning_rate": 0.0002,
-      "loss": 0.8789,
-      "step": 3190
-    },
-    {
-      "epoch": 4.63,
-      "learning_rate": 0.0002,
-      "loss": 0.8359,
-      "step": 3200
-    },
-    {
-      "epoch": 4.64,
-      "learning_rate": 0.0002,
-      "loss": 0.8223,
-      "step": 3210
-    },
-    {
-      "epoch": 4.66,
-      "learning_rate": 0.0002,
-      "loss": 0.8178,
-      "step": 3220
-    },
-    {
-      "epoch": 4.67,
-      "learning_rate": 0.0002,
-      "loss": 0.9243,
-      "step": 3230
-    },
-    {
-      "epoch": 4.69,
-      "learning_rate": 0.0002,
-      "loss": 0.8732,
-      "step": 3240
-    },
-    {
-      "epoch": 4.7,
-      "learning_rate": 0.0002,
-      "loss": 0.8427,
-      "step": 3250
-    },
-    {
-      "epoch": 4.71,
-      "learning_rate": 0.0002,
-      "loss": 0.8358,
-      "step": 3260
-    },
-    {
-      "epoch": 4.73,
-      "learning_rate": 0.0002,
-      "loss": 0.8308,
-      "step": 3270
-    },
-    {
-      "epoch": 4.74,
-      "learning_rate": 0.0002,
-      "loss": 0.9175,
-      "step": 3280
-    },
-    {
-      "epoch": 4.76,
-      "learning_rate": 0.0002,
-      "loss": 0.8806,
-      "step": 3290
-    },
-    {
-      "epoch": 4.77,
-      "learning_rate": 0.0002,
-      "loss": 0.848,
-      "step": 3300
-    },
-    {
-      "epoch": 4.79,
-      "learning_rate": 0.0002,
-      "loss": 0.8159,
-      "step": 3310
-    },
-    {
-      "epoch": 4.8,
-      "learning_rate": 0.0002,
-      "loss": 0.8202,
-      "step": 3320
-    },
-    {
-      "epoch": 4.82,
-      "learning_rate": 0.0002,
-      "loss": 0.9166,
-      "step": 3330
-    },
-    {
-      "epoch": 4.83,
-      "learning_rate": 0.0002,
-      "loss": 0.8756,
-      "step": 3340
-    },
-    {
-      "epoch": 4.84,
-      "learning_rate": 0.0002,
-      "loss": 0.842,
-      "step": 3350
-    },
-    {
-      "epoch": 4.86,
-      "learning_rate": 0.0002,
-      "loss": 0.831,
-      "step": 3360
-    },
-    {
-      "epoch": 4.87,
-      "learning_rate": 0.0002,
-      "loss": 0.8228,
-      "step": 3370
-    },
-    {
-      "epoch": 4.89,
-      "learning_rate": 0.0002,
-      "loss": 0.9109,
-      "step": 3380
-    },
-    {
-      "epoch": 4.9,
-      "learning_rate": 0.0002,
-      "loss": 0.8746,
-      "step": 3390
-    },
-    {
-      "epoch": 4.92,
-      "learning_rate": 0.0002,
-      "loss": 0.8605,
-      "step": 3400
-    },
-    {
-      "epoch": 4.93,
-      "learning_rate": 0.0002,
-      "loss": 0.8263,
-      "step": 3410
-    },
-    {
-      "epoch": 4.95,
-      "learning_rate": 0.0002,
-      "loss": 0.8009,
-      "step": 3420
-    },
-    {
-      "epoch": 4.96,
-      "learning_rate": 0.0002,
-      "loss": 0.9056,
-      "step": 3430
-    },
-    {
-      "epoch": 4.97,
-      "learning_rate": 0.0002,
-      "loss": 0.8647,
-      "step": 3440
-    },
-    {
-      "epoch": 4.99,
-      "learning_rate": 0.0002,
-      "loss": 0.8481,
-      "step": 3450
-    },
-    {
-      "epoch": 5.0,
-      "learning_rate": 0.0002,
-      "loss": 0.7768,
-      "step": 3460
-    },
-    {
-      "epoch": 5.02,
-      "learning_rate": 0.0002,
-      "loss": 0.8872,
-      "step": 3470
-    },
-    {
-      "epoch": 5.03,
-      "learning_rate": 0.0002,
-      "loss": 0.8517,
-      "step": 3480
-    },
-    {
-      "epoch": 5.05,
-      "learning_rate": 0.0002,
-      "loss": 0.8096,
-      "step": 3490
-    },
-    {
-      "epoch": 5.06,
-      "learning_rate": 0.0002,
-      "loss": 0.7853,
-      "step": 3500
-    },
-    {
-      "epoch": 5.08,
-      "learning_rate": 0.0002,
-      "loss": 0.7469,
-      "step": 3510
-    },
-    {
-      "epoch": 5.09,
-      "learning_rate": 0.0002,
-      "loss": 0.9045,
-      "step": 3520
-    },
-    {
-      "epoch": 5.1,
-      "learning_rate": 0.0002,
-      "loss": 0.8387,
-      "step": 3530
-    },
-    {
-      "epoch": 5.12,
-      "learning_rate": 0.0002,
-      "loss": 0.8039,
-      "step": 3540
-    },
-    {
-      "epoch": 5.13,
-      "learning_rate": 0.0002,
-      "loss": 0.7629,
-      "step": 3550
-    },
-    {
-      "epoch": 5.15,
-      "learning_rate": 0.0002,
-      "loss": 0.7336,
-      "step": 3560
-    },
-    {
-      "epoch": 5.16,
-      "learning_rate": 0.0002,
-      "loss": 0.9018,
-      "step": 3570
-    },
-    {
-      "epoch": 5.18,
-      "learning_rate": 0.0002,
-      "loss": 0.8511,
-      "step": 3580
-    },
-    {
-      "epoch": 5.19,
-      "learning_rate": 0.0002,
-      "loss": 0.8043,
-      "step": 3590
-    },
-    {
-      "epoch": 5.21,
-      "learning_rate": 0.0002,
-      "loss": 0.7877,
-      "step": 3600
-    },
-    {
-      "epoch": 5.22,
-      "learning_rate": 0.0002,
-      "loss": 0.7438,
-      "step": 3610
-    },
-    {
-      "epoch": 5.23,
-      "learning_rate": 0.0002,
-      "loss": 0.903,
-      "step": 3620
-    },
-    {
-      "epoch": 5.25,
-      "learning_rate": 0.0002,
-      "loss": 0.8562,
-      "step": 3630
-    },
-    {
-      "epoch": 5.26,
-      "learning_rate": 0.0002,
-      "loss": 0.8198,
-      "step": 3640
-    },
-    {
-      "epoch": 5.28,
-      "learning_rate": 0.0002,
-      "loss": 0.7953,
-      "step": 3650
-    },
-    {
-      "epoch": 5.29,
-      "learning_rate": 0.0002,
-      "loss": 0.7606,
-      "step": 3660
-    },
-    {
-      "epoch": 5.31,
-      "learning_rate": 0.0002,
-      "loss": 0.8996,
-      "step": 3670
-    },
-    {
-      "epoch": 5.32,
-      "learning_rate": 0.0002,
-      "loss": 0.8519,
-      "step": 3680
-    },
-    {
-      "epoch": 5.34,
-      "learning_rate": 0.0002,
-      "loss": 0.8307,
-      "step": 3690
-    },
-    {
-      "epoch": 5.35,
-      "learning_rate": 0.0002,
-      "loss": 0.7875,
-      "step": 3700
-    },
-    {
-      "epoch": 5.37,
-      "learning_rate": 0.0002,
-      "loss": 0.7508,
-      "step": 3710
-    },
-    {
-      "epoch": 5.38,
-      "learning_rate": 0.0002,
-      "loss": 0.9087,
-      "step": 3720
-    },
-    {
-      "epoch": 5.39,
-      "learning_rate": 0.0002,
-      "loss": 0.8564,
-      "step": 3730
-    },
-    {
-      "epoch": 5.41,
-      "learning_rate": 0.0002,
-      "loss": 0.8265,
-      "step": 3740
-    },
-    {
-      "epoch": 5.42,
-      "learning_rate": 0.0002,
-      "loss": 0.794,
-      "step": 3750
-    },
-    {
-      "epoch": 5.44,
-      "learning_rate": 0.0002,
-      "loss": 0.7502,
-      "step": 3760
-    },
-    {
-      "epoch": 5.45,
-      "learning_rate": 0.0002,
-      "loss": 0.8966,
-      "step": 3770
-    },
-    {
-      "epoch": 5.47,
-      "learning_rate": 0.0002,
-      "loss": 0.8521,
-      "step": 3780
-    },
-    {
-      "epoch": 5.48,
-      "learning_rate": 0.0002,
-      "loss": 0.8195,
-      "step": 3790
-    },
-    {
-      "epoch": 5.5,
-      "learning_rate": 0.0002,
-      "loss": 0.7985,
-      "step": 3800
-    },
-    {
-      "epoch": 5.51,
-      "learning_rate": 0.0002,
-      "loss": 0.7676,
-      "step": 3810
-    },
-    {
-      "epoch": 5.52,
-      "learning_rate": 0.0002,
-      "loss": 0.9075,
-      "step": 3820
-    },
-    {
-      "epoch": 5.54,
-      "learning_rate": 0.0002,
-      "loss": 0.852,
-      "step": 3830
-    },
-    {
-      "epoch": 5.55,
-      "learning_rate": 0.0002,
-      "loss": 0.8306,
-      "step": 3840
-    },
-    {
-      "epoch": 5.57,
-      "learning_rate": 0.0002,
-      "loss": 0.7964,
-      "step": 3850
-    },
-    {
-      "epoch": 5.58,
-      "learning_rate": 0.0002,
-      "loss": 0.7599,
-      "step": 3860
-    },
-    {
-      "epoch": 5.6,
-      "learning_rate": 0.0002,
-      "loss": 0.8959,
-      "step": 3870
-    },
-    {
-      "epoch": 5.61,
-      "learning_rate": 0.0002,
-      "loss": 0.8489,
-      "step": 3880
-    },
-    {
-      "epoch": 5.63,
-      "learning_rate": 0.0002,
-      "loss": 0.8095,
-      "step": 3890
-    },
-    {
-      "epoch": 5.64,
-      "learning_rate": 0.0002,
-      "loss": 0.7912,
-      "step": 3900
-    },
-    {
-      "epoch": 5.65,
-      "learning_rate": 0.0002,
-      "loss": 0.7555,
-      "step": 3910
-    },
-    {
-      "epoch": 5.67,
-      "learning_rate": 0.0002,
-      "loss": 0.8884,
-      "step": 3920
-    },
-    {
-      "epoch": 5.68,
-      "learning_rate": 0.0002,
-      "loss": 0.8375,
-      "step": 3930
-    },
-    {
-      "epoch": 5.7,
-      "learning_rate": 0.0002,
-      "loss": 0.8313,
-      "step": 3940
-    },
-    {
-      "epoch": 5.71,
-      "learning_rate": 0.0002,
-      "loss": 0.7983,
-      "step": 3950
-    },
-    {
-      "epoch": 5.73,
-      "learning_rate": 0.0002,
-      "loss": 0.7691,
-      "step": 3960
-    },
-    {
-      "epoch": 5.74,
-      "learning_rate": 0.0002,
-      "loss": 0.8946,
-      "step": 3970
-    },
-    {
-      "epoch": 5.76,
-      "learning_rate": 0.0002,
-      "loss": 0.8528,
-      "step": 3980
-    },
-    {
-      "epoch": 5.77,
-      "learning_rate": 0.0002,
-      "loss": 0.8293,
-      "step": 3990
-    },
-    {
-      "epoch": 5.78,
-      "learning_rate": 0.0002,
-      "loss": 0.7929,
-      "step": 4000
-    },
-    {
-      "epoch": 5.8,
-      "learning_rate": 0.0002,
-      "loss": 0.773,
-      "step": 4010
-    },
-    {
-      "epoch": 5.81,
-      "learning_rate": 0.0002,
-      "loss": 0.8988,
-      "step": 4020
-    },
-    {
-      "epoch": 5.83,
-      "learning_rate": 0.0002,
-      "loss": 0.8639,
-      "step": 4030
-    },
-    {
-      "epoch": 5.84,
-      "learning_rate": 0.0002,
-      "loss": 0.8219,
-      "step": 4040
-    },
-    {
-      "epoch": 5.86,
-      "learning_rate": 0.0002,
-      "loss": 0.8106,
-      "step": 4050
-    },
-    {
-      "epoch": 5.87,
-      "learning_rate": 0.0002,
-      "loss": 0.7746,
-      "step": 4060
-    },
-    {
-      "epoch": 5.89,
-      "learning_rate": 0.0002,
-      "loss": 0.8874,
-      "step": 4070
-    },
-    {
-      "epoch": 5.9,
-      "learning_rate": 0.0002,
-      "loss": 0.8497,
-      "step": 4080
-    },
-    {
-      "epoch": 5.91,
-      "learning_rate": 0.0002,
-      "loss": 0.818,
-      "step": 4090
-    },
-    {
-      "epoch": 5.93,
-      "learning_rate": 0.0002,
-      "loss": 0.7891,
-      "step": 4100
-    },
-    {
-      "epoch": 5.94,
-      "learning_rate": 0.0002,
-      "loss": 0.7508,
-      "step": 4110
-    },
-    {
-      "epoch": 5.96,
-      "learning_rate": 0.0002,
-      "loss": 0.8985,
-      "step": 4120
-    },
-    {
-      "epoch": 5.97,
-      "learning_rate": 0.0002,
-      "loss": 0.8496,
-      "step": 4130
-    },
-    {
-      "epoch": 5.99,
-      "learning_rate": 0.0002,
-      "loss": 0.8176,
-      "step": 4140
-    },
-    {
-      "epoch": 6.0,
-      "learning_rate": 0.0002,
-      "loss": 0.7464,
-      "step": 4150
-    },
-    {
-      "epoch": 6.02,
-      "learning_rate": 0.0002,
-      "loss": 0.8725,
-      "step": 4160
-    },
-    {
-      "epoch": 6.03,
-      "learning_rate": 0.0002,
-      "loss": 0.8238,
-      "step": 4170
-    },
-    {
-      "epoch": 6.04,
-      "learning_rate": 0.0002,
-      "loss": 0.7955,
-      "step": 4180
-    },
-    {
-      "epoch": 6.06,
-      "learning_rate": 0.0002,
-      "loss": 0.7304,
-      "step": 4190
-    },
-    {
-      "epoch": 6.07,
-      "learning_rate": 0.0002,
-      "loss": 0.6796,
-      "step": 4200
-    },
-    {
-      "epoch": 6.09,
-      "learning_rate": 0.0002,
-      "loss": 0.887,
-      "step": 4210
-    },
-    {
-      "epoch": 6.1,
-      "learning_rate": 0.0002,
-      "loss": 0.831,
-      "step": 4220
-    },
-    {
-      "epoch": 6.12,
-      "learning_rate": 0.0002,
-      "loss": 0.7878,
-      "step": 4230
-    },
-    {
-      "epoch": 6.13,
-      "learning_rate": 0.0002,
-      "loss": 0.7574,
-      "step": 4240
-    },
-    {
-      "epoch": 6.15,
-      "learning_rate": 0.0002,
-      "loss": 0.6826,
-      "step": 4250
-    },
-    {
-      "epoch": 6.16,
-      "learning_rate": 0.0002,
-      "loss": 0.881,
-      "step": 4260
-    },
-    {
-      "epoch": 6.17,
-      "learning_rate": 0.0002,
-      "loss": 0.8245,
-      "step": 4270
-    },
-    {
-      "epoch": 6.19,
-      "learning_rate": 0.0002,
-      "loss": 0.7827,
-      "step": 4280
-    },
-    {
-      "epoch": 6.2,
-      "learning_rate": 0.0002,
-      "loss": 0.7556,
-      "step": 4290
-    },
-    {
-      "epoch": 6.22,
-      "learning_rate": 0.0002,
-      "loss": 0.6812,
-      "step": 4300
-    },
-    {
-      "epoch": 6.23,
-      "learning_rate": 0.0002,
-      "loss": 0.8812,
-      "step": 4310
-    },
-    {
-      "epoch": 6.25,
-      "learning_rate": 0.0002,
-      "loss": 0.8373,
-      "step": 4320
-    },
-    {
-      "epoch": 6.26,
-      "learning_rate": 0.0002,
-      "loss": 0.7915,
-      "step": 4330
-    },
-    {
-      "epoch": 6.28,
-      "learning_rate": 0.0002,
-      "loss": 0.7575,
-      "step": 4340
-    },
-    {
-      "epoch": 6.29,
-      "learning_rate": 0.0002,
-      "loss": 0.6715,
-      "step": 4350
-    },
-    {
-      "epoch": 6.31,
-      "learning_rate": 0.0002,
-      "loss": 0.8731,
-      "step": 4360
-    },
-    {
-      "epoch": 6.32,
-      "learning_rate": 0.0002,
-      "loss": 0.8362,
-      "step": 4370
-    },
-    {
-      "epoch": 6.33,
-      "learning_rate": 0.0002,
-      "loss": 0.7917,
-      "step": 4380
-    },
-    {
-      "epoch": 6.35,
-      "learning_rate": 0.0002,
-      "loss": 0.7803,
-      "step": 4390
-    },
-    {
-      "epoch": 6.36,
-      "learning_rate": 0.0002,
-      "loss": 0.6743,
-      "step": 4400
-    },
-    {
-      "epoch": 6.38,
-      "learning_rate": 0.0002,
-      "loss": 0.8894,
-      "step": 4410
-    },
-    {
-      "epoch": 6.39,
-      "learning_rate": 0.0002,
-      "loss": 0.8375,
-      "step": 4420
-    },
-    {
-      "epoch": 6.41,
-      "learning_rate": 0.0002,
-      "loss": 0.793,
-      "step": 4430
-    },
-    {
-      "epoch": 6.42,
-      "learning_rate": 0.0002,
-      "loss": 0.7616,
-      "step": 4440
-    },
-    {
-      "epoch": 6.44,
-      "learning_rate": 0.0002,
-      "loss": 0.6821,
-      "step": 4450
-    },
-    {
-      "epoch": 6.45,
-      "learning_rate": 0.0002,
-      "loss": 0.884,
-      "step": 4460
-    },
-    {
-      "epoch": 6.46,
-      "learning_rate": 0.0002,
-      "loss": 0.8264,
-      "step": 4470
-    },
-    {
-      "epoch": 6.48,
-      "learning_rate": 0.0002,
-      "loss": 0.7921,
-      "step": 4480
-    },
-    {
-      "epoch": 6.49,
-      "learning_rate": 0.0002,
-      "loss": 0.7699,
-      "step": 4490
-    },
-    {
-      "epoch": 6.51,
-      "learning_rate": 0.0002,
-      "loss": 0.6964,
-      "step": 4500
-    },
-    {
-      "epoch": 6.52,
-      "learning_rate": 0.0002,
-      "loss": 0.8871,
-      "step": 4510
-    },
-    {
-      "epoch": 6.54,
-      "learning_rate": 0.0002,
-      "loss": 0.8286,
-      "step": 4520
-    },
-    {
-      "epoch": 6.55,
-      "learning_rate": 0.0002,
-      "loss": 0.792,
-      "step": 4530
-    },
-    {
-      "epoch": 6.57,
-      "learning_rate": 0.0002,
-      "loss": 0.7692,
-      "step": 4540
-    },
-    {
-      "epoch": 6.58,
-      "learning_rate": 0.0002,
-      "loss": 0.6903,
-      "step": 4550
-    },
-    {
-      "epoch": 6.59,
-      "learning_rate": 0.0002,
-      "loss": 0.8976,
-      "step": 4560
-    },
-    {
-      "epoch": 6.61,
-      "learning_rate": 0.0002,
-      "loss": 0.8361,
-      "step": 4570
-    },
-    {
-      "epoch": 6.62,
-      "learning_rate": 0.0002,
-      "loss": 0.7947,
-      "step": 4580
-    },
-    {
-      "epoch": 6.64,
-      "learning_rate": 0.0002,
-      "loss": 0.7715,
-      "step": 4590
-    },
-    {
-      "epoch": 6.65,
-      "learning_rate": 0.0002,
-      "loss": 0.7046,
-      "step": 4600
-    },
-    {
-      "epoch": 6.67,
-      "learning_rate": 0.0002,
-      "loss": 0.8952,
-      "step": 4610
-    },
-    {
-      "epoch": 6.68,
-      "learning_rate": 0.0002,
-      "loss": 0.848,
-      "step": 4620
-    },
-    {
-      "epoch": 6.7,
-      "learning_rate": 0.0002,
-      "loss": 0.7885,
-      "step": 4630
-    },
-    {
-      "epoch": 6.71,
-      "learning_rate": 0.0002,
-      "loss": 0.7753,
-      "step": 4640
-    },
-    {
-      "epoch": 6.72,
-      "learning_rate": 0.0002,
-      "loss": 0.7011,
-      "step": 4650
-    },
-    {
-      "epoch": 6.74,
-      "learning_rate": 0.0002,
-      "loss": 0.8948,
-      "step": 4660
-    },
-    {
-      "epoch": 6.75,
-      "learning_rate": 0.0002,
-      "loss": 0.8186,
-      "step": 4670
-    },
-    {
-      "epoch": 6.77,
-      "learning_rate": 0.0002,
-      "loss": 0.7991,
-      "step": 4680
-    },
-    {
-      "epoch": 6.78,
-      "learning_rate": 0.0002,
-      "loss": 0.756,
-      "step": 4690
-    },
-    {
-      "epoch": 6.8,
-      "learning_rate": 0.0002,
-      "loss": 0.6999,
-      "step": 4700
-    },
-    {
-      "epoch": 6.81,
-      "learning_rate": 0.0002,
-      "loss": 0.8991,
-      "step": 4710
-    },
-    {
-      "epoch": 6.83,
-      "learning_rate": 0.0002,
-      "loss": 0.8334,
-      "step": 4720
-    },
-    {
-      "epoch": 6.84,
-      "learning_rate": 0.0002,
-      "loss": 0.7978,
-      "step": 4730
-    },
-    {
-      "epoch": 6.85,
-      "learning_rate": 0.0002,
-      "loss": 0.7514,
-      "step": 4740
-    },
-    {
-      "epoch": 6.87,
-      "learning_rate": 0.0002,
-      "loss": 0.6968,
-      "step": 4750
-    },
-    {
-      "epoch": 6.88,
-      "learning_rate": 0.0002,
-      "loss": 0.8783,
-      "step": 4760
-    },
-    {
-      "epoch": 6.9,
-      "learning_rate": 0.0002,
-      "loss": 0.8291,
-      "step": 4770
-    },
-    {
-      "epoch": 6.91,
-      "learning_rate": 0.0002,
-      "loss": 0.8059,
-      "step": 4780
-    },
-    {
-      "epoch": 6.93,
-      "learning_rate": 0.0002,
-      "loss": 0.7764,
-      "step": 4790
-    },
-    {
-      "epoch": 6.94,
-      "learning_rate": 0.0002,
-      "loss": 0.6978,
-      "step": 4800
-    },
-    {
-      "epoch": 6.96,
-      "learning_rate": 0.0002,
-      "loss": 0.8875,
-      "step": 4810
-    },
-    {
-      "epoch": 6.97,
-      "learning_rate": 0.0002,
-      "loss": 0.8238,
-      "step": 4820
-    },
-    {
-      "epoch": 6.98,
-      "learning_rate": 0.0002,
-      "loss": 0.7776,
-      "step": 4830
-    },
-    {
-      "epoch": 7.0,
-      "learning_rate": 0.0002,
-      "loss": 0.7001,
-      "step": 4840
-    },
-    {
-      "epoch": 7.01,
-      "learning_rate": 0.0002,
-      "loss": 0.8558,
-      "step": 4850
-    },
-    {
-      "epoch": 7.03,
-      "learning_rate": 0.0002,
-      "loss": 0.8014,
-      "step": 4860
-    },
-    {
-      "epoch": 7.04,
-      "learning_rate": 0.0002,
-      "loss": 0.7671,
-      "step": 4870
-    },
-    {
-      "epoch": 7.06,
-      "learning_rate": 0.0002,
-      "loss": 0.7089,
-      "step": 4880
-    },
-    {
-      "epoch": 7.07,
-      "learning_rate": 0.0002,
-      "loss": 0.6218,
-      "step": 4890
-    },
-    {
-      "epoch": 7.09,
-      "learning_rate": 0.0002,
-      "loss": 0.8637,
-      "step": 4900
-    },
-    {
-      "epoch": 7.1,
-      "learning_rate": 0.0002,
-      "loss": 0.8098,
-      "step": 4910
-    },
-    {
-      "epoch": 7.11,
-      "learning_rate": 0.0002,
-      "loss": 0.7572,
-      "step": 4920
-    },
-    {
-      "epoch": 7.13,
-      "learning_rate": 0.0002,
-      "loss": 0.7368,
-      "step": 4930
-    },
-    {
-      "epoch": 7.14,
-      "learning_rate": 0.0002,
-      "loss": 0.6251,
-      "step": 4940
-    },
-    {
-      "epoch": 7.16,
-      "learning_rate": 0.0002,
-      "loss": 0.8433,
-      "step": 4950
-    },
-    {
-      "epoch": 7.17,
-      "learning_rate": 0.0002,
-      "loss": 0.8094,
-      "step": 4960
-    },
-    {
-      "epoch": 7.19,
-      "learning_rate": 0.0002,
-      "loss": 0.7691,
-      "step": 4970
-    },
-    {
-      "epoch": 7.2,
-      "learning_rate": 0.0002,
-      "loss": 0.7159,
-      "step": 4980
-    },
-    {
-      "epoch": 7.22,
-      "learning_rate": 0.0002,
-      "loss": 0.6349,
-      "step": 4990
-    },
-    {
-      "epoch": 7.23,
-      "learning_rate": 0.0002,
-      "loss": 0.854,
-      "step": 5000
     }
   ],
   "max_steps": 5000,
   "num_train_epochs": 8,
-  "total_flos": 1.0738578956363366e+18,
+  "total_flos": 6.44896481399808e+17,
   "trial_name": null,
   "trial_params": null
 }