Training in progress, step 700, checkpoint

checkpoint-700/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db7fdeb0323d53de315aa90b913ce901a91eda8cd9f0e67d5bc47a170e020655
+oid sha256:bf77a7c1fd44dd3b150147cd2fffe878b7fd689deb8bb07f62f9fb4062a131cf
 size 9873829

checkpoint-700/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fcb139eacb9d2507bb7cc82320a30294de3c631243324674152bd026bdc54572
+oid sha256:6c020262f5f9f88db32caf56fa334dbb6625730ab2d8d9a994e491853626359c
 size 42788

checkpoint-700/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b5bd959545174877cf1f2846fdb832b0eeabc4f2c3ed0bb189d289d673c20fdd
+oid sha256:7e8e0d13424de41f2d3dd75843c59f2acf4f5c61cdef60b073f4aee97951614f
 size 14244

checkpoint-700/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:41b6c424d1ec058567267ff5869c79e515d0afabd27f2a21fce1ee8663b75bd3
+oid sha256:00f0a0254b7e61177c287e87b876b738deea627e195cf94456de79f1af81a13d
 size 1064

checkpoint-700/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.4556220769882202,
-  "best_model_checkpoint": "output/checkpoint-600",
-  "epoch": 0.7061790668348046,
+  "best_metric": 0.6392325758934021,
+  "best_model_checkpoint": "output/checkpoint-700",
+  "epoch": 0.5615160934523212,
   "eval_steps": 50,
   "global_step": 700,
   "is_hyper_param_search": false,
@@ -10,226 +10,226 @@
   "log_history": [
     {
       "epoch": 0.0,
-      "learning_rate": 1.6666666666666667e-06,
-      "loss": 0.5465,
+      "learning_rate": 1.3333333333333332e-06,
+      "loss": 0.785,
       "step": 1
     },
     {
-      "epoch": 0.05,
-      "learning_rate": 7.5e-05,
-      "loss": 0.7053,
+      "epoch": 0.04,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 0.6547,
       "step": 50
     },
     {
-      "epoch": 0.05,
-      "eval_accuracy": 0.6561679790026247,
-      "eval_loss": 0.6764907836914062,
-      "eval_runtime": 59.7766,
-      "eval_samples_per_second": 6.374,
-      "eval_steps_per_second": 1.606,
+      "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,
       "step": 50
     },
     {
-      "epoch": 0.1,
-      "learning_rate": 0.00015666666666666666,
-      "loss": 0.6576,
+      "epoch": 0.08,
+      "learning_rate": 0.00012666666666666666,
+      "loss": 0.6654,
       "step": 100
     },
     {
-      "epoch": 0.1,
-      "eval_accuracy": 0.7217847769028871,
-      "eval_loss": 0.574826717376709,
-      "eval_runtime": 59.8708,
-      "eval_samples_per_second": 6.364,
-      "eval_steps_per_second": 1.603,
+      "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,
       "step": 100
     },
     {
-      "epoch": 0.15,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 0.5874,
+      "epoch": 0.12,
+      "learning_rate": 0.00019333333333333333,
+      "loss": 0.6491,
       "step": 150
     },
     {
-      "epoch": 0.15,
-      "eval_accuracy": 0.7480314960629921,
-      "eval_loss": 0.5501317381858826,
-      "eval_runtime": 55.7487,
-      "eval_samples_per_second": 6.834,
-      "eval_steps_per_second": 1.722,
+      "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,
       "step": 150
     },
     {
-      "epoch": 0.2,
-      "learning_rate": 0.0002948780487804878,
-      "loss": 0.5775,
+      "epoch": 0.16,
+      "learning_rate": 0.00026,
+      "loss": 0.641,
       "step": 200
     },
     {
-      "epoch": 0.2,
-      "eval_accuracy": 0.7401574803149606,
-      "eval_loss": 0.5592061877250671,
-      "eval_runtime": 54.7265,
-      "eval_samples_per_second": 6.962,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 200
     },
     {
-      "epoch": 0.25,
-      "learning_rate": 0.00027658536585365855,
-      "loss": 0.5821,
+      "epoch": 0.2,
+      "learning_rate": 0.0002941234084231146,
+      "loss": 0.6521,
       "step": 250
     },
     {
-      "epoch": 0.25,
-      "eval_accuracy": 0.7664041994750657,
-      "eval_loss": 0.5005777478218079,
-      "eval_runtime": 54.7418,
-      "eval_samples_per_second": 6.96,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 250
     },
     {
-      "epoch": 0.3,
-      "learning_rate": 0.00025829268292682926,
-      "loss": 0.5699,
+      "epoch": 0.24,
+      "learning_rate": 0.00027943192948090106,
+      "loss": 0.638,
       "step": 300
     },
     {
-      "epoch": 0.3,
-      "eval_accuracy": 0.6745406824146981,
-      "eval_loss": 0.7000961303710938,
-      "eval_runtime": 54.7375,
-      "eval_samples_per_second": 6.96,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 300
     },
     {
-      "epoch": 0.35,
-      "learning_rate": 0.00023999999999999998,
-      "loss": 0.5079,
+      "epoch": 0.28,
+      "learning_rate": 0.00026474045053868754,
+      "loss": 0.6857,
       "step": 350
     },
     {
-      "epoch": 0.35,
-      "eval_accuracy": 0.7322834645669292,
-      "eval_loss": 0.5262107253074646,
-      "eval_runtime": 54.7143,
-      "eval_samples_per_second": 6.963,
-      "eval_steps_per_second": 1.755,
+      "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,
       "step": 350
     },
     {
-      "epoch": 0.4,
-      "learning_rate": 0.00022170731707317072,
-      "loss": 0.5826,
+      "epoch": 0.32,
+      "learning_rate": 0.000250048971596474,
+      "loss": 0.6469,
       "step": 400
     },
     {
-      "epoch": 0.4,
-      "eval_accuracy": 0.7296587926509186,
-      "eval_loss": 0.5798280835151672,
-      "eval_runtime": 54.7163,
-      "eval_samples_per_second": 6.963,
-      "eval_steps_per_second": 1.755,
+      "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,
       "step": 400
     },
     {
-      "epoch": 0.45,
-      "learning_rate": 0.00020341463414634146,
-      "loss": 0.5009,
+      "epoch": 0.36,
+      "learning_rate": 0.00023535749265426052,
+      "loss": 0.6813,
       "step": 450
     },
     {
-      "epoch": 0.45,
-      "eval_accuracy": 0.7506561679790026,
-      "eval_loss": 0.5315915942192078,
-      "eval_runtime": 54.7235,
-      "eval_samples_per_second": 6.962,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 450
     },
     {
-      "epoch": 0.5,
-      "learning_rate": 0.00018512195121951218,
-      "loss": 0.4756,
+      "epoch": 0.4,
+      "learning_rate": 0.000220666013712047,
+      "loss": 0.645,
       "step": 500
     },
     {
-      "epoch": 0.5,
-      "eval_accuracy": 0.7349081364829396,
-      "eval_loss": 0.5232318639755249,
-      "eval_runtime": 54.7185,
-      "eval_samples_per_second": 6.963,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 500
     },
     {
-      "epoch": 0.55,
-      "learning_rate": 0.00016682926829268292,
-      "loss": 0.4791,
+      "epoch": 0.44,
+      "learning_rate": 0.0002059745347698335,
+      "loss": 0.7044,
       "step": 550
     },
     {
-      "epoch": 0.55,
-      "eval_accuracy": 0.7847769028871391,
-      "eval_loss": 0.447153776884079,
-      "eval_runtime": 54.7183,
-      "eval_samples_per_second": 6.963,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 550
     },
     {
-      "epoch": 0.61,
-      "learning_rate": 0.00014853658536585364,
-      "loss": 0.4415,
+      "epoch": 0.48,
+      "learning_rate": 0.00019128305582761997,
+      "loss": 0.6399,
       "step": 600
     },
     {
-      "epoch": 0.61,
-      "eval_accuracy": 0.7847769028871391,
-      "eval_loss": 0.4556220769882202,
-      "eval_runtime": 54.7334,
-      "eval_samples_per_second": 6.961,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 600
     },
     {
-      "epoch": 0.66,
-      "learning_rate": 0.00013024390243902438,
-      "loss": 0.4528,
+      "epoch": 0.52,
+      "learning_rate": 0.00017659157688540645,
+      "loss": 0.6039,
       "step": 650
     },
     {
-      "epoch": 0.66,
-      "eval_accuracy": 0.7690288713910761,
-      "eval_loss": 0.47037920355796814,
-      "eval_runtime": 54.7397,
-      "eval_samples_per_second": 6.96,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 650
     },
     {
-      "epoch": 0.71,
-      "learning_rate": 0.00011195121951219511,
-      "loss": 0.5177,
+      "epoch": 0.56,
+      "learning_rate": 0.00016190009794319292,
+      "loss": 0.6321,
       "step": 700
     },
     {
-      "epoch": 0.71,
-      "eval_accuracy": 0.7742782152230971,
-      "eval_loss": 0.4634643793106079,
-      "eval_runtime": 54.7337,
-      "eval_samples_per_second": 6.961,
-      "eval_steps_per_second": 1.754,
+      "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,
       "step": 700
     }
   ],
   "logging_steps": 50,
-  "max_steps": 1000,
-  "num_train_epochs": 2,
+  "max_steps": 1246,
+  "num_train_epochs": 1,
   "save_steps": 100,
-  "total_flos": 1.2097024614875136e+17,
+  "total_flos": 2.6209416656252928e+17,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-700/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:303b684398cbe6b39e16627081218289c5098690bdfd2926fac0c4c4b61ed03d
+oid sha256:e32e120c6d65576e6fe76fdf63d907877e27469586cc746d8cd6a71d07635ea9
 size 4472