Training in progress, step 2500, checkpoint

last-checkpoint/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fafc67897a73ac0e70a597348c3155bc381e60dd90c67d5d6ed7fc84ba105613
+oid sha256:57c4831d4e47f3c33f5f8969a02983374946ab3e5f383e907aca585a86e5b94d
 size 6425529112

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b0f91ec07067e47aed7410c0ad51ca6a2cf72d126a46306b9c6488df8bde784a
+oid sha256:4159a3207a568baa044ce95dd31d1026fc5b9d5c7ec0e1a1965827fde31d1cab
 size 12851224679

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7949a82b4956d6dc1683588213cba80de08632b440866c7377566d6b5ddef0f6
+oid sha256:8ab5f3bc9f12c23cdfc964a4b7a0357b0f5ab599c9285206de2dfd5600d45e1d
 size 1465

last-checkpoint/trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.5625,
+  "epoch": 1.953125,
   "eval_steps": 500,
-  "global_step": 2000,
+  "global_step": 2500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -408,6 +408,106 @@
       "mean_token_accuracy": 0.9926387491822243,
       "num_tokens": 10201426.0,
       "step": 2000
+    },
+    {
+      "entropy": 0.030316293751820923,
+      "epoch": 1.6015625,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.435763888888889e-06,
+      "loss": 0.021486189365386963,
+      "mean_token_accuracy": 0.9927674040198327,
+      "num_tokens": 10454276.0,
+      "step": 2050
+    },
+    {
+      "entropy": 0.02887956439051777,
+      "epoch": 1.640625,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.001736111111112e-06,
+      "loss": 0.020056800842285158,
+      "mean_token_accuracy": 0.9931298586726188,
+      "num_tokens": 10707497.0,
+      "step": 2100
+    },
+    {
+      "entropy": 0.028787780185230077,
+      "epoch": 1.6796875,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.5677083333333335e-06,
+      "loss": 0.0200783896446228,
+      "mean_token_accuracy": 0.9929500755667686,
+      "num_tokens": 10962955.0,
+      "step": 2150
+    },
+    {
+      "entropy": 0.028631422137841582,
+      "epoch": 1.71875,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.1336805555555562e-06,
+      "loss": 0.02008913516998291,
+      "mean_token_accuracy": 0.9930686053633689,
+      "num_tokens": 11217333.0,
+      "step": 2200
+    },
+    {
+      "entropy": 0.02850784788839519,
+      "epoch": 1.7578125,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.699652777777778e-06,
+      "loss": 0.01996502876281738,
+      "mean_token_accuracy": 0.9930599120259285,
+      "num_tokens": 11472495.0,
+      "step": 2250
+    },
+    {
+      "entropy": 0.028683945639058947,
+      "epoch": 1.796875,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.265625e-06,
+      "loss": 0.019921081066131593,
+      "mean_token_accuracy": 0.9932510870695114,
+      "num_tokens": 11728662.0,
+      "step": 2300
+    },
+    {
+      "entropy": 0.02975102465134114,
+      "epoch": 1.8359375,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.8315972222222223e-06,
+      "loss": 0.020963990688323976,
+      "mean_token_accuracy": 0.9929134699702263,
+      "num_tokens": 11983347.0,
+      "step": 2350
+    },
+    {
+      "entropy": 0.02991143790073693,
+      "epoch": 1.875,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3975694444444446e-06,
+      "loss": 0.020808370113372804,
+      "mean_token_accuracy": 0.9929909712076187,
+      "num_tokens": 12236526.0,
+      "step": 2400
+    },
+    {
+      "entropy": 0.030044674500823022,
+      "epoch": 1.9140625,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.635416666666667e-07,
+      "loss": 0.021207802295684815,
+      "mean_token_accuracy": 0.9927184066176414,
+      "num_tokens": 12489562.0,
+      "step": 2450
+    },
+    {
+      "entropy": 0.02905242417007685,
+      "epoch": 1.953125,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.295138888888889e-07,
+      "loss": 0.020578203201293947,
+      "mean_token_accuracy": 0.9929935920238495,
+      "num_tokens": 12748444.0,
+      "step": 2500
     }
   ],
   "logging_steps": 50,
@@ -427,7 +527,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.936166915593175e+17,
+  "total_flos": 2.4192422651355955e+17,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null