Training in progress, step 1000, checkpoint

checkpoint-1000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e968dbb3380de7a3cb81dafc561acc0238e3df7f4a2d3593f54687f9b501d6cf
+oid sha256:3af484d2981c1d8d51caba3ea0b20da0d73e3958223f010171f35184084a130c
 size 435544704

checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:88ebb55308c87e3e9097aa3b8979c4cd29b29eeb02dfb4f603a07a47ce4c1d5f
+oid sha256:02f0f0a5fe76c042386505973b8005fe81c1768a92f6d6341ae31e07c213c82a
 size 871183627

checkpoint-1000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f46683c066aee705f7b93bdbd4b829ca423162b1fb577c522b9b1c1625a9c1c1
+oid sha256:39586f70e92c891dbcaab37909f4cf86c0501394c80c92f12d33b7c2ac0ad3c7
 size 14709

checkpoint-1000/scaler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:14ae2a2128444abab378aa06c09a61a84665f758fcc19fc46f5789b0bc1b5665
+size 1383

checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3381cb33b25da3196bf5781a9390f345898978fd83e9e4bca1d36cf8c7145e8a
+oid sha256:ab1794c1f570008e45f746af78e6132c4682c43814cf01f7e4ba820420220475
 size 1465

checkpoint-1000/trainer_state.json

@@ -1,33 +1,39 @@
 {
-  "best_global_step": 982,
-  "best_metric": 3.760134696960449,
+  "best_global_step": 501,
+  "best_metric": 3.908651113510132,
   "best_model_checkpoint": null,
-  "epoch": 2.0366598778004072,
+  "epoch": 1.996007984031936,
   "eval_steps": 500,
   "global_step": 1000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
+    {
+      "epoch": 0.001996007984031936,
+      "grad_norm": 30.08072853088379,
+      "learning_rate": 0.0,
+      "loss": 10.3965,
+      "step": 1
+    },
     {
       "epoch": 1.0,
-      "eval_loss": 4.097745895385742,
-      "eval_runtime": 66.4624,
-      "eval_samples_per_second": 504.782,
-      "eval_steps_per_second": 1.986,
-      "step": 491
+      "grad_norm": 1.243941068649292,
+      "learning_rate": 0.0001,
+      "loss": 4.8834,
+      "step": 501
     },
     {
-      "epoch": 2.0,
-      "eval_loss": 3.760134696960449,
-      "eval_runtime": 66.4623,
-      "eval_samples_per_second": 504.783,
-      "eval_steps_per_second": 1.986,
-      "step": 982
+      "epoch": 1.0,
+      "eval_loss": 3.908651113510132,
+      "eval_runtime": 19.9819,
+      "eval_samples_per_second": 1586.036,
+      "eval_steps_per_second": 6.206,
+      "step": 501
     }
   ],
-  "logging_steps": 4000,
-  "max_steps": 2455,
+  "logging_steps": 500,
+  "max_steps": 2505,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 5,
   "save_steps": 500,
@@ -43,7 +49,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.6717202915328e+16,
+  "total_flos": 1.6720991649792e+16,
   "train_batch_size": 256,
   "trial_name": null,
   "trial_params": null

checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9a57dba9b9c8c3226400ebcad0a2060a021a094dd027a7ac1acf9378f5c6cc27
+oid sha256:cb7a8afa9c6745812d26be318b1c7d9348962a9c4c0d9cbb9442934f28d74eb6
 size 5969