Training in progress, step 4100, checkpoint

last-checkpoint/adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6f654fa6534ae48416fd3b50f90a80eafdeed68b33a32a6aace4caab3df83a1c
+oid sha256:4ed2742c10327a32eb422aaa133c3032dc538b6a336d8b357616a430bb5a9eac
 size 50349441

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0e3ac252efb905580c340792dde23762e62e7c716fc2902b717985412f24587c
+oid sha256:25cf6bc5917a6db676a0dec291a2cc7176611b0fbcc8fc2a6b9308a5f99d2a7a
 size 100691721

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5840f7374f51c709f33701383022a2fde86c49fec768033add729bc923ee95d3
+oid sha256:8e0ce6d2dc6f370c842f935b375eb168ab79f78ba12ff7121f2ed0954b0deab8
 size 14575

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:593f43b0c83adb27a2db37a6418c2ef12a213bbc2a02f2dc881de6846a69a931
+oid sha256:f55ebe991e4838deda9ff3044672335e03697b65b3fe0753cca61998e974ee51
 size 627

last-checkpoint/trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 1.2198114395141602,
-  "best_model_checkpoint": "./outputs/checkpoint-4000",
-  "epoch": 2.914754098360656,
+  "best_metric": 1.207747220993042,
+  "best_model_checkpoint": "./outputs/checkpoint-4100",
+  "epoch": 2.9876138433515482,
   "eval_steps": 100,
-  "global_step": 4000,
+  "global_step": 4100,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -567,13 +567,27 @@
       "eval_samples_per_second": 46.997,
       "eval_steps_per_second": 5.88,
       "step": 4000
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 0.0002,
+      "loss": 1.1977,
+      "step": 4100
+    },
+    {
+      "epoch": 2.99,
+      "eval_loss": 1.207747220993042,
+      "eval_runtime": 133.474,
+      "eval_samples_per_second": 47.005,
+      "eval_steps_per_second": 5.881,
+      "step": 4100
     }
   ],
   "logging_steps": 100,
   "max_steps": 4116,
   "num_train_epochs": 3,
   "save_steps": 100,
-  "total_flos": 2.384126359199662e+17,
+  "total_flos": 2.4438930355335168e+17,
   "trial_name": null,
   "trial_params": null
 }