Training in progress, step 120

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "total_flos": 1.337851081728e+17,
-    "train_loss": 1.9536447048187255,
-    "train_runtime": 67.9851,
-    "train_samples_per_second": 4.119,
-    "train_steps_per_second": 1.177
+    "epoch": 30.0,
+    "total_flos": 2.006776622592e+17,
+    "train_loss": 4.5579022089640295,
+    "train_runtime": 124.8123,
+    "train_samples_per_second": 3.365,
+    "train_steps_per_second": 0.961
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c482402c9c654933c0ff6f7fe6b58f8bb7b94180d5745e052edb8ea5e6eb051a
+oid sha256:2510e3b4d354c14f689771f130c169e0d1f7915c38b97ef7d3309890f451ac84
 size 166494824

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 20.0,
-    "total_flos": 1.337851081728e+17,
-    "train_loss": 1.9536447048187255,
-    "train_runtime": 67.9851,
-    "train_samples_per_second": 4.119,
-    "train_steps_per_second": 1.177
+    "epoch": 30.0,
+    "total_flos": 2.006776622592e+17,
+    "train_loss": 4.5579022089640295,
+    "train_runtime": 124.8123,
+    "train_samples_per_second": 3.365,
+    "train_steps_per_second": 0.961
 }

trainer_state.json

@@ -2,34 +2,41 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 20.0,
+  "epoch": 30.0,
   "eval_steps": 500,
-  "global_step": 80,
+  "global_step": 120,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 12.5,
-      "grad_norm": 210.8382110595703,
-      "learning_rate": 2.25e-05,
-      "loss": 2.4839,
+      "grad_norm": 288.3207702636719,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 4.9309,
       "step": 50
     },
     {
-      "epoch": 20.0,
-      "step": 80,
-      "total_flos": 1.337851081728e+17,
-      "train_loss": 1.9536447048187255,
-      "train_runtime": 67.9851,
-      "train_samples_per_second": 4.119,
-      "train_steps_per_second": 1.177
+      "epoch": 25.0,
+      "grad_norm": 51.23756408691406,
+      "learning_rate": 1.1666666666666668e-06,
+      "loss": 4.4404,
+      "step": 100
+    },
+    {
+      "epoch": 30.0,
+      "step": 120,
+      "total_flos": 2.006776622592e+17,
+      "train_loss": 4.5579022089640295,
+      "train_runtime": 124.8123,
+      "train_samples_per_second": 3.365,
+      "train_steps_per_second": 0.961
     }
   ],
   "logging_steps": 50,
-  "max_steps": 80,
+  "max_steps": 120,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 20,
+  "num_train_epochs": 30,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -43,7 +50,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.337851081728e+17,
+  "total_flos": 2.006776622592e+17,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05398c6b0d063ce41284fcd4f97224abe39cc4e784222b880209c55073892b1e
+oid sha256:7d7abd29495988426f34f1c7ec4f52b77b3753a4b217dc36f1e379bfb88dc74b
 size 5240