Training checkpoint - Epoch 2, Step 2

checkpoint-2/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a1e26f482fdd70d763d84e01a2995298706db54404e59c5203e0703fc8086f39
+oid sha256:ff04924ce62b9fc764833152a08658641d4b6b50a484515d0e9b4deae068963d
 size 14244

checkpoint-2/trainer_state.json

@@ -3,7 +3,7 @@
   "best_metric": null,
   "best_model_checkpoint": null,
   "epoch": 2.0,
-  "eval_steps": 4000,
+  "eval_steps": 500,
   "global_step": 2,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
@@ -15,6 +15,24 @@
       "learning_rate": 0.0,
       "loss": 2.4073,
       "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "eval_cosine_accuracy": 0.8333333134651184,
+      "eval_loss": 1.1724432706832886,
+      "eval_runtime": 2.8395,
+      "eval_samples_per_second": 2.113,
+      "eval_steps_per_second": 0.352,
+      "step": 1
+    },
+    {
+      "epoch": 2.0,
+      "eval_cosine_accuracy": 0.8333333134651184,
+      "eval_loss": 1.1766316890716553,
+      "eval_runtime": 0.2923,
+      "eval_samples_per_second": 20.526,
+      "eval_steps_per_second": 3.421,
+      "step": 2
     }
   ],
   "logging_steps": 4000,

checkpoint-2/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6951bb839d239af6b75461c6692130490816c65513ca3277e113a0116269ead9
+oid sha256:d15f3f8d8ef7d48fb84d1cbac4916e98a1684a23c018308e301ab1bed527a935
 size 5624

checkpoint-2/training_metrics.json

@@ -1,5 +1,18 @@
 {
-  "metrics_history": [],
+  "metrics_history": [
+    {
+      "epoch": 1.0,
+      "global_step": 1,
+      "eval_loss": 1.1724432706832886,
+      "eval_cosine_accuracy": 0.8333333134651184
+    },
+    {
+      "epoch": 2.0,
+      "global_step": 2,
+      "eval_loss": 1.1766316890716553,
+      "eval_cosine_accuracy": 0.8333333134651184
+    }
+  ],
   "current_epoch": 2,
   "global_step": 2
 }