Training in progress, step 1000, checkpoint

checkpoint-1000/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f828f36ba945f1507c45a05f1a37f545095d2612ef9fba0b9631668b79085e82
+oid sha256:5131ccd1b8b62e01483e1c1e4039bf31f4b7015ac0cf46af540e95e2d50d7b4d
 size 9873829

checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:132bf7c54f35b9948e3ef98a12b6351e47f4e66da211c0a1b4706af057765569
+oid sha256:7d2922ddb0fb6c301b058d43a35e2217be71a93f5e9313e04904162f151513f9
 size 42788

checkpoint-1000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ad48fe3d5e8f6602a346a0df641401c9bcfe846ed4c7781c79e319b8ec0cf6a
+oid sha256:428d36d5acbdf9f3d66c093fcb7f0ae60e56d217f73ffb0684c29ad50bc36a53
 size 14244

checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:344ce23fea7895dd5e72f1ce513e8d0929df797342da489b0d29a933753167cf
+oid sha256:92114783b5f8ad9adf7b9b236fff840926827a113216ecc0b2c7c5ba1cdc1d87
 size 1064

checkpoint-1000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.6392325758934021,
-  "best_model_checkpoint": "output/checkpoint-700",
-  "epoch": 0.8021658477890303,
+  "best_metric": 0.6053406596183777,
+  "best_model_checkpoint": "output/checkpoint-500",
+  "epoch": 0.8018442417560389,
   "eval_steps": 50,
   "global_step": 1000,
   "is_hyper_param_search": false,
@@ -11,315 +11,315 @@
     {
       "epoch": 0.0,
       "learning_rate": 1.3333333333333332e-06,
-      "loss": 0.785,
+      "loss": 0.6133,
       "step": 1
     },
     {
       "epoch": 0.04,
       "learning_rate": 5.9999999999999995e-05,
-      "loss": 0.6547,
+      "loss": 0.61,
       "step": 50
     },
     {
       "epoch": 0.04,
-      "eval_accuracy": 0.6626865671641791,
-      "eval_loss": 0.8300915956497192,
-      "eval_runtime": 61.5689,
-      "eval_samples_per_second": 5.441,
-      "eval_steps_per_second": 1.364,
+      "eval_accuracy": 0.6895522388059702,
+      "eval_loss": 0.6718239784240723,
+      "eval_runtime": 61.4305,
+      "eval_samples_per_second": 5.453,
+      "eval_steps_per_second": 1.367,
       "step": 50
     },
     {
       "epoch": 0.08,
       "learning_rate": 0.00012666666666666666,
-      "loss": 0.6654,
+      "loss": 0.6233,
       "step": 100
     },
     {
       "epoch": 0.08,
-      "eval_accuracy": 0.6895522388059702,
-      "eval_loss": 0.6663276553153992,
-      "eval_runtime": 56.4633,
-      "eval_samples_per_second": 5.933,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6835820895522388,
+      "eval_loss": 0.695776641368866,
+      "eval_runtime": 61.5801,
+      "eval_samples_per_second": 5.44,
+      "eval_steps_per_second": 1.364,
       "step": 100
     },
     {
       "epoch": 0.12,
       "learning_rate": 0.00019333333333333333,
-      "loss": 0.6491,
+      "loss": 0.6187,
       "step": 150
     },
     {
       "epoch": 0.12,
-      "eval_accuracy": 0.6537313432835821,
-      "eval_loss": 0.9172552824020386,
-      "eval_runtime": 56.4705,
-      "eval_samples_per_second": 5.932,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6895522388059702,
+      "eval_loss": 0.6547008156776428,
+      "eval_runtime": 61.4876,
+      "eval_samples_per_second": 5.448,
+      "eval_steps_per_second": 1.366,
       "step": 150
     },
     {
       "epoch": 0.16,
       "learning_rate": 0.00026,
-      "loss": 0.641,
+      "loss": 0.667,
       "step": 200
     },
     {
       "epoch": 0.16,
-      "eval_accuracy": 0.6865671641791045,
-      "eval_loss": 0.6528732180595398,
-      "eval_runtime": 56.4485,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6686567164179105,
+      "eval_loss": 0.7339211702346802,
+      "eval_runtime": 61.0301,
+      "eval_samples_per_second": 5.489,
+      "eval_steps_per_second": 1.376,
       "step": 200
     },
     {
       "epoch": 0.2,
-      "learning_rate": 0.0002941234084231146,
-      "loss": 0.6521,
+      "learning_rate": 0.0002941291585127201,
+      "loss": 0.6896,
       "step": 250
     },
     {
       "epoch": 0.2,
-      "eval_accuracy": 0.6835820895522388,
-      "eval_loss": 0.8197052478790283,
-      "eval_runtime": 56.4577,
-      "eval_samples_per_second": 5.934,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6388059701492538,
+      "eval_loss": 0.6455020904541016,
+      "eval_runtime": 61.0356,
+      "eval_samples_per_second": 5.489,
+      "eval_steps_per_second": 1.376,
       "step": 250
     },
     {
       "epoch": 0.24,
-      "learning_rate": 0.00027943192948090106,
-      "loss": 0.638,
+      "learning_rate": 0.0002794520547945205,
+      "loss": 0.6732,
       "step": 300
     },
     {
       "epoch": 0.24,
-      "eval_accuracy": 0.6835820895522388,
-      "eval_loss": 0.6961421370506287,
-      "eval_runtime": 56.4506,
-      "eval_samples_per_second": 5.934,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6388059701492538,
+      "eval_loss": 0.8866483569145203,
+      "eval_runtime": 61.5006,
+      "eval_samples_per_second": 5.447,
+      "eval_steps_per_second": 1.366,
       "step": 300
     },
     {
       "epoch": 0.28,
-      "learning_rate": 0.00026474045053868754,
-      "loss": 0.6857,
+      "learning_rate": 0.0002647749510763209,
+      "loss": 0.6255,
       "step": 350
     },
     {
       "epoch": 0.28,
-      "eval_accuracy": 0.6835820895522388,
-      "eval_loss": 0.7876304388046265,
-      "eval_runtime": 56.4651,
-      "eval_samples_per_second": 5.933,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6746268656716418,
+      "eval_loss": 0.6250954866409302,
+      "eval_runtime": 61.4499,
+      "eval_samples_per_second": 5.452,
+      "eval_steps_per_second": 1.367,
       "step": 350
     },
     {
       "epoch": 0.32,
-      "learning_rate": 0.000250048971596474,
-      "loss": 0.6469,
+      "learning_rate": 0.0002500978473581213,
+      "loss": 0.6522,
       "step": 400
     },
     {
       "epoch": 0.32,
-      "eval_accuracy": 0.6626865671641791,
-      "eval_loss": 0.7487305402755737,
-      "eval_runtime": 56.4489,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6925373134328359,
+      "eval_loss": 0.7217190861701965,
+      "eval_runtime": 61.1569,
+      "eval_samples_per_second": 5.478,
+      "eval_steps_per_second": 1.374,
       "step": 400
     },
     {
       "epoch": 0.36,
-      "learning_rate": 0.00023535749265426052,
-      "loss": 0.6813,
+      "learning_rate": 0.00023542074363992171,
+      "loss": 0.6592,
       "step": 450
     },
     {
       "epoch": 0.36,
-      "eval_accuracy": 0.6626865671641791,
-      "eval_loss": 0.8447557091712952,
-      "eval_runtime": 56.4428,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6417910447761194,
+      "eval_loss": 0.636791467666626,
+      "eval_runtime": 61.431,
+      "eval_samples_per_second": 5.453,
+      "eval_steps_per_second": 1.367,
       "step": 450
     },
     {
       "epoch": 0.4,
-      "learning_rate": 0.000220666013712047,
-      "loss": 0.645,
+      "learning_rate": 0.00022074363992172208,
+      "loss": 0.624,
       "step": 500
     },
     {
       "epoch": 0.4,
-      "eval_accuracy": 0.6835820895522388,
-      "eval_loss": 0.6624641418457031,
-      "eval_runtime": 56.4531,
-      "eval_samples_per_second": 5.934,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.7044776119402985,
+      "eval_loss": 0.6053406596183777,
+      "eval_runtime": 61.4573,
+      "eval_samples_per_second": 5.451,
+      "eval_steps_per_second": 1.367,
       "step": 500
     },
     {
       "epoch": 0.44,
-      "learning_rate": 0.0002059745347698335,
-      "loss": 0.7044,
+      "learning_rate": 0.0002060665362035225,
+      "loss": 0.6098,
       "step": 550
     },
     {
       "epoch": 0.44,
-      "eval_accuracy": 0.7014925373134329,
-      "eval_loss": 0.63779217004776,
-      "eval_runtime": 56.4382,
-      "eval_samples_per_second": 5.936,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6477611940298508,
+      "eval_loss": 0.8857873678207397,
+      "eval_runtime": 61.4645,
+      "eval_samples_per_second": 5.45,
+      "eval_steps_per_second": 1.367,
       "step": 550
     },
     {
       "epoch": 0.48,
-      "learning_rate": 0.00019128305582761997,
-      "loss": 0.6399,
+      "learning_rate": 0.00019138943248532287,
+      "loss": 0.6357,
       "step": 600
     },
     {
       "epoch": 0.48,
-      "eval_accuracy": 0.6716417910447762,
-      "eval_loss": 0.7784742712974548,
-      "eval_runtime": 56.4356,
-      "eval_samples_per_second": 5.936,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.7074626865671642,
+      "eval_loss": 0.607388436794281,
+      "eval_runtime": 61.4941,
+      "eval_samples_per_second": 5.448,
+      "eval_steps_per_second": 1.366,
       "step": 600
     },
     {
       "epoch": 0.52,
-      "learning_rate": 0.00017659157688540645,
-      "loss": 0.6039,
+      "learning_rate": 0.0001767123287671233,
+      "loss": 0.6103,
       "step": 650
     },
     {
       "epoch": 0.52,
-      "eval_accuracy": 0.6985074626865672,
-      "eval_loss": 0.665508508682251,
-      "eval_runtime": 56.4424,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6746268656716418,
+      "eval_loss": 0.6098934412002563,
+      "eval_runtime": 56.4929,
+      "eval_samples_per_second": 5.93,
+      "eval_steps_per_second": 1.487,
       "step": 650
     },
     {
       "epoch": 0.56,
-      "learning_rate": 0.00016190009794319292,
-      "loss": 0.6321,
+      "learning_rate": 0.00016203522504892366,
+      "loss": 0.6025,
       "step": 700
     },
     {
       "epoch": 0.56,
-      "eval_accuracy": 0.6716417910447762,
-      "eval_loss": 0.6392325758934021,
-      "eval_runtime": 56.437,
-      "eval_samples_per_second": 5.936,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6895522388059702,
+      "eval_loss": 0.6117912530899048,
+      "eval_runtime": 61.1994,
+      "eval_samples_per_second": 5.474,
+      "eval_steps_per_second": 1.373,
       "step": 700
     },
     {
       "epoch": 0.6,
-      "learning_rate": 0.00014720861900097942,
-      "loss": 0.6448,
+      "learning_rate": 0.00014735812133072405,
+      "loss": 0.5837,
       "step": 750
     },
     {
       "epoch": 0.6,
-      "eval_accuracy": 0.6686567164179105,
-      "eval_loss": 0.7872520089149475,
-      "eval_runtime": 56.4423,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6925373134328359,
+      "eval_loss": 0.6470033526420593,
+      "eval_runtime": 61.3513,
+      "eval_samples_per_second": 5.46,
+      "eval_steps_per_second": 1.369,
       "step": 750
     },
     {
       "epoch": 0.64,
-      "learning_rate": 0.0001325171400587659,
-      "loss": 0.6033,
+      "learning_rate": 0.00013268101761252445,
+      "loss": 0.623,
       "step": 800
     },
     {
       "epoch": 0.64,
-      "eval_accuracy": 0.6895522388059702,
-      "eval_loss": 0.723906934261322,
-      "eval_runtime": 56.4576,
-      "eval_samples_per_second": 5.934,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6776119402985075,
+      "eval_loss": 0.7178429365158081,
+      "eval_runtime": 56.5122,
+      "eval_samples_per_second": 5.928,
+      "eval_steps_per_second": 1.486,
       "step": 800
     },
     {
       "epoch": 0.68,
-      "learning_rate": 0.00011782566111655239,
-      "loss": 0.6174,
+      "learning_rate": 0.00011800391389432484,
+      "loss": 0.6285,
       "step": 850
     },
     {
       "epoch": 0.68,
-      "eval_accuracy": 0.6925373134328359,
-      "eval_loss": 0.6893543601036072,
-      "eval_runtime": 56.4403,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6805970149253732,
+      "eval_loss": 0.7012776732444763,
+      "eval_runtime": 61.0309,
+      "eval_samples_per_second": 5.489,
+      "eval_steps_per_second": 1.376,
       "step": 850
     },
     {
       "epoch": 0.72,
-      "learning_rate": 0.00010313418217433888,
-      "loss": 0.5859,
+      "learning_rate": 0.00010332681017612522,
+      "loss": 0.609,
       "step": 900
     },
     {
       "epoch": 0.72,
-      "eval_accuracy": 0.6746268656716418,
-      "eval_loss": 0.6730663776397705,
-      "eval_runtime": 56.4424,
-      "eval_samples_per_second": 5.935,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6656716417910448,
+      "eval_loss": 0.6416519284248352,
+      "eval_runtime": 56.5079,
+      "eval_samples_per_second": 5.928,
+      "eval_steps_per_second": 1.487,
       "step": 900
     },
     {
       "epoch": 0.76,
-      "learning_rate": 8.844270323212536e-05,
+      "learning_rate": 8.864970645792562e-05,
       "loss": 0.6,
       "step": 950
     },
     {
       "epoch": 0.76,
-      "eval_accuracy": 0.6686567164179105,
-      "eval_loss": 0.6526519060134888,
-      "eval_runtime": 56.4284,
-      "eval_samples_per_second": 5.937,
-      "eval_steps_per_second": 1.489,
+      "eval_accuracy": 0.6835820895522388,
+      "eval_loss": 0.6628562808036804,
+      "eval_runtime": 61.1667,
+      "eval_samples_per_second": 5.477,
+      "eval_steps_per_second": 1.373,
       "step": 950
     },
     {
       "epoch": 0.8,
-      "learning_rate": 7.375122428991184e-05,
-      "loss": 0.6127,
+      "learning_rate": 7.397260273972601e-05,
+      "loss": 0.6142,
       "step": 1000
     },
     {
       "epoch": 0.8,
-      "eval_accuracy": 0.7014925373134329,
-      "eval_loss": 0.6604557037353516,
-      "eval_runtime": 56.4364,
-      "eval_samples_per_second": 5.936,
-      "eval_steps_per_second": 1.488,
+      "eval_accuracy": 0.6805970149253732,
+      "eval_loss": 0.635610044002533,
+      "eval_runtime": 61.6579,
+      "eval_samples_per_second": 5.433,
+      "eval_steps_per_second": 1.362,
       "step": 1000
     }
   ],
   "logging_steps": 50,
-  "max_steps": 1246,
+  "max_steps": 1247,
   "num_train_epochs": 1,
   "save_steps": 100,
-  "total_flos": 3.754567767823565e+17,
+  "total_flos": 3.745803772140749e+17,
   "trial_name": null,
   "trial_params": null
 }