Training in progress, step 500, checkpoint

checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:620d6e7c6fe5fec79cc452d40a6769e258718089ea22c8dc6e144c3bd0f670de
+oid sha256:d8329d2cc307e739195088026e2fd5111bd83ae59a2339444451ffc2bcf323d9
 size 577789320

checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:300b153fd04b5f1d0d0c5ca840ccf3fbc28d427d2e7dd2638bed18bda640626b
+oid sha256:641bfba23f9e06c960e3886fb85293cd5654fcc712b47f9a4efad7249e452f14
 size 1155772233

checkpoint-500/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bc1c71206c3ebb75270a918c997498b42af9ae88c8c69b43ce735a0883a88364
+oid sha256:42f8436ff41efaeac4ed5cc91663e67b8bb07c4e8be6d4eeaa077d213146c7e8
 size 14244

checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:92e482854e4ff3c178dda878d03bf3633330934a775fc818e1843bd607d65d07
+oid sha256:59c5da1b5b4f7c36f93ab88fdfad2f08c02a4309458a2395e3f5b5f0b79b879d
 size 1064

checkpoint-500/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 0.5375720262527466,
+  "best_metric": 0.48496538400650024,
   "best_model_checkpoint": "mikhail-panzo/fil_b32_le5_s4000/checkpoint-500",
-  "epoch": 11.11111111111111,
+  "epoch": 44.44444444444444,
   "eval_steps": 500,
   "global_step": 500,
   "is_hyper_param_search": false,
@@ -9,88 +9,88 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 1.1111111111111112,
-      "grad_norm": 4.398683071136475,
-      "learning_rate": 2.4500000000000004e-07,
-      "loss": 0.8108,
+      "epoch": 4.444444444444445,
+      "grad_norm": 4.288165092468262,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.7989,
       "step": 50
     },
     {
-      "epoch": 2.2222222222222223,
-      "grad_norm": 3.726436138153076,
-      "learning_rate": 4.95e-07,
-      "loss": 0.792,
+      "epoch": 8.88888888888889,
+      "grad_norm": 2.4493463039398193,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.7701,
       "step": 100
     },
     {
-      "epoch": 3.3333333333333335,
-      "grad_norm": 2.0379676818847656,
-      "learning_rate": 7.450000000000001e-07,
-      "loss": 0.7413,
+      "epoch": 13.333333333333334,
+      "grad_norm": 1.3798907995224,
+      "learning_rate": 7.5e-07,
+      "loss": 0.7374,
       "step": 150
     },
     {
-      "epoch": 4.444444444444445,
-      "grad_norm": 2.0460364818573,
-      "learning_rate": 9.950000000000002e-07,
-      "loss": 0.7342,
+      "epoch": 17.77777777777778,
+      "grad_norm": 2.6173877716064453,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.7101,
       "step": 200
     },
     {
-      "epoch": 5.555555555555555,
-      "grad_norm": 2.803539276123047,
-      "learning_rate": 1.2450000000000002e-06,
-      "loss": 0.7147,
+      "epoch": 22.22222222222222,
+      "grad_norm": 1.3756701946258545,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6887,
       "step": 250
     },
     {
-      "epoch": 6.666666666666667,
-      "grad_norm": 1.9818241596221924,
-      "learning_rate": 1.495e-06,
-      "loss": 0.7072,
+      "epoch": 26.666666666666668,
+      "grad_norm": 1.0456843376159668,
+      "learning_rate": 1.5e-06,
+      "loss": 0.672,
       "step": 300
     },
     {
-      "epoch": 7.777777777777778,
-      "grad_norm": 3.081399917602539,
-      "learning_rate": 1.745e-06,
-      "loss": 0.6843,
+      "epoch": 31.11111111111111,
+      "grad_norm": 0.9518398642539978,
+      "learning_rate": 1.75e-06,
+      "loss": 0.6583,
       "step": 350
     },
     {
-      "epoch": 8.88888888888889,
-      "grad_norm": 2.4465785026550293,
-      "learning_rate": 1.9950000000000004e-06,
-      "loss": 0.657,
+      "epoch": 35.55555555555556,
+      "grad_norm": 1.305978775024414,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.6258,
       "step": 400
     },
     {
-      "epoch": 10.0,
-      "grad_norm": 2.929323434829712,
-      "learning_rate": 2.245e-06,
-      "loss": 0.6597,
+      "epoch": 40.0,
+      "grad_norm": 0.8857623338699341,
+      "learning_rate": 2.25e-06,
+      "loss": 0.5835,
       "step": 450
     },
     {
-      "epoch": 11.11111111111111,
-      "grad_norm": 1.419311285018921,
-      "learning_rate": 2.4950000000000003e-06,
-      "loss": 0.6246,
+      "epoch": 44.44444444444444,
+      "grad_norm": 1.052182912826538,
+      "learning_rate": 2.5e-06,
+      "loss": 0.5539,
       "step": 500
     },
     {
-      "epoch": 11.11111111111111,
-      "eval_loss": 0.5375720262527466,
-      "eval_runtime": 9.2568,
-      "eval_samples_per_second": 17.177,
-      "eval_steps_per_second": 2.161,
+      "epoch": 44.44444444444444,
+      "eval_loss": 0.48496538400650024,
+      "eval_runtime": 9.4574,
+      "eval_samples_per_second": 16.812,
+      "eval_steps_per_second": 2.115,
       "step": 500
     }
   ],
   "logging_steps": 50,
   "max_steps": 4000,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 89,
+  "num_train_epochs": 364,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -104,7 +104,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 3441129941684160.0,
+  "total_flos": 1.3753846024576416e+16,
   "train_batch_size": 16,
   "trial_name": null,
   "trial_params": null

checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:05b7ecdded947540b9670cedf7cfe87afc7199b237efb642b915ffbb505d51dc
+oid sha256:2de69a241d012253ce4b1526ecd0f9fd0f7dfbedfea74bfff08144d42d9f523a
 size 5304