Training in progress, step 20, checkpoint

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd56b63f9d959f90d0f4fafc9d4ceee3e07955e87dfcfa64cb2b46ccf1025090
+oid sha256:596f84af86cbd8e1a1793fb22598bcccae463ff78bb47b5547427beebb47ec20
 size 330143034

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8717e6db6dad3bdb0d92afa70b8fbe0c4677a2360eab813a627812a62061fd4
-size 15984
+oid sha256:69a04a1208f7a0d6f51f37a136b5c2e55bf3f53b3d0fd57164c5b83ca47a2645
+size 15024

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6c7dfa9863eadd74540b95922e8cee67d3b691b168a951cb8b3349087d7a629
-size 15984
+oid sha256:080a7e72d6be938a9418e60003db90412af8a61e6434f9e9f1b598cca861dbcd
+size 15024

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e7596c043bb61ec6ab94f537047f856762fd577107f970ea392238fff0ea6ef1
-size 15984
+oid sha256:c3d114a75d37be476b865187eb2b3d29d9343b131614a08f42be0014f110ce6f
+size 15024

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:712dd66678ab2f0cf679625d414555daee92b448e39ad2db26b98722c1b424fc
-size 15984
+oid sha256:4fc5a0f78838743362c5d5378dff81ea2f7d0039da53a423f1759e861bc6b233
+size 15024

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:416b2ea21de930aef02b3dd806b882963de9b98a0a8d9ff11b933cd6bf4116bc
+oid sha256:84f22729b765b6841bd185712dd3ab8bf338866cd8396b5dce62f9950913691a
 size 1064

last-checkpoint/trainer_state.json

@@ -1,159 +1,49 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.14201183431952663,
+  "epoch": 2.857142857142857,
   "eval_steps": 20,
-  "global_step": 120,
+  "global_step": 20,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.001183431952662722,
-      "eval_loss": 3.7497785091400146,
-      "eval_runtime": 2.5725,
-      "eval_samples_per_second": 583.871,
-      "eval_steps_per_second": 36.541,
+      "epoch": 0.14285714285714285,
+      "eval_loss": 3.573837995529175,
+      "eval_runtime": 2.7173,
+      "eval_samples_per_second": 552.387,
+      "eval_steps_per_second": 4.784,
       "step": 1
     },
     {
-      "epoch": 0.011834319526627219,
+      "epoch": 1.4285714285714286,
       "grad_norm": Infinity,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 3.4365,
+      "learning_rate": 0.00019863613034027224,
+      "loss": 3.0145,
       "step": 10
     },
     {
-      "epoch": 0.023668639053254437,
+      "epoch": 2.857142857142857,
       "grad_norm": Infinity,
-      "learning_rate": 3.2000000000000005e-05,
-      "loss": 3.552,
+      "learning_rate": 0.0001879473751206489,
+      "loss": 3.0123,
       "step": 20
     },
     {
-      "epoch": 0.023668639053254437,
-      "eval_loss": 3.752192735671997,
-      "eval_runtime": 2.475,
-      "eval_samples_per_second": 606.872,
-      "eval_steps_per_second": 37.98,
+      "epoch": 2.857142857142857,
+      "eval_loss": 3.611345052719116,
+      "eval_runtime": 2.6681,
+      "eval_samples_per_second": 562.578,
+      "eval_steps_per_second": 4.872,
       "step": 20
-    },
-    {
-      "epoch": 0.03550295857988166,
-      "grad_norm": Infinity,
-      "learning_rate": 4.8e-05,
-      "loss": 3.6338,
-      "step": 30
-    },
-    {
-      "epoch": 0.047337278106508875,
-      "grad_norm": Infinity,
-      "learning_rate": 6.400000000000001e-05,
-      "loss": 3.7366,
-      "step": 40
-    },
-    {
-      "epoch": 0.047337278106508875,
-      "eval_loss": 3.750141143798828,
-      "eval_runtime": 2.4397,
-      "eval_samples_per_second": 615.651,
-      "eval_steps_per_second": 38.529,
-      "step": 40
-    },
-    {
-      "epoch": 0.05917159763313609,
-      "grad_norm": Infinity,
-      "learning_rate": 8e-05,
-      "loss": 3.7361,
-      "step": 50
-    },
-    {
-      "epoch": 0.07100591715976332,
-      "grad_norm": Infinity,
-      "learning_rate": 9.6e-05,
-      "loss": 3.5694,
-      "step": 60
-    },
-    {
-      "epoch": 0.07100591715976332,
-      "eval_loss": 3.730466842651367,
-      "eval_runtime": 2.5074,
-      "eval_samples_per_second": 599.033,
-      "eval_steps_per_second": 37.489,
-      "step": 60
-    },
-    {
-      "epoch": 0.08284023668639054,
-      "grad_norm": Infinity,
-      "learning_rate": 0.00011200000000000001,
-      "loss": 3.9619,
-      "step": 70
-    },
-    {
-      "epoch": 0.09467455621301775,
-      "grad_norm": Infinity,
-      "learning_rate": 0.00012800000000000002,
-      "loss": 3.5414,
-      "step": 80
-    },
-    {
-      "epoch": 0.09467455621301775,
-      "eval_loss": 3.7425360679626465,
-      "eval_runtime": 2.4554,
-      "eval_samples_per_second": 611.71,
-      "eval_steps_per_second": 38.283,
-      "step": 80
-    },
-    {
-      "epoch": 0.10650887573964497,
-      "grad_norm": Infinity,
-      "learning_rate": 0.000144,
-      "loss": 3.446,
-      "step": 90
-    },
-    {
-      "epoch": 0.11834319526627218,
-      "grad_norm": Infinity,
-      "learning_rate": 0.00016,
-      "loss": 3.9408,
-      "step": 100
-    },
-    {
-      "epoch": 0.11834319526627218,
-      "eval_loss": 3.747370719909668,
-      "eval_runtime": 2.528,
-      "eval_samples_per_second": 594.151,
-      "eval_steps_per_second": 37.184,
-      "step": 100
-    },
-    {
-      "epoch": 0.1301775147928994,
-      "grad_norm": Infinity,
-      "learning_rate": 0.00017600000000000002,
-      "loss": 3.4239,
-      "step": 110
-    },
-    {
-      "epoch": 0.14201183431952663,
-      "grad_norm": Infinity,
-      "learning_rate": 0.000192,
-      "loss": 3.7869,
-      "step": 120
-    },
-    {
-      "epoch": 0.14201183431952663,
-      "eval_loss": 3.740018606185913,
-      "eval_runtime": 2.5424,
-      "eval_samples_per_second": 590.789,
-      "eval_steps_per_second": 36.973,
-      "step": 120
     }
   ],
   "logging_steps": 10,
-  "max_steps": 2500,
+  "max_steps": 100,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 3,
-  "save_steps": 40,
+  "num_train_epochs": 15,
+  "save_steps": 20,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -166,8 +56,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 3255474413633536.0,
-  "train_batch_size": 2,
+  "total_flos": 4065108430946304.0,
+  "train_batch_size": 30,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:930e523ba424e07364b1eb3e1b298d7c7f59473e3b137fdf485803d46fe2b072
-size 6712
+oid sha256:2acaca78115072d9639fc12703771de8b7909ac6a6095b046cd694fed5704d2a
+size 6840