checkpoint 9000

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5425a36f1e325bee6d82b1fcc07102ff534ecb18f72844c64a5049b9705ee1a2
+oid sha256:0a61e6e5352efdd5b7f9da781d485d41785a65d97707d8a50f8ed3703500e382
 size 1262032932

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:40c8da2258f27071c4927075f92382af808544f48a8473848bacbf9ee6bd88e1
+oid sha256:d61d798a6ebe2df9f0c91788f23fc918b91a4b2bf0166521c19127eccff65502
 size 2490610241

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ee893b4ad54887290f9b4069066f0404d323d7fb2e1360ff8f1c42e1bb7aca50
+oid sha256:70d3644a4cf26f14e27a5301570cd0ec97d359dfc1ab88070196c7de228b37c2
 size 14639

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8133a804599afbc4c397ae559484c26a48033b26cd1cb76daab66148cf4c67eb
+oid sha256:0fb799b5054cd707634c1757db8b87f65d5adfd51f621867293ee2d7170c2937
 size 627

trainer_state.json

@@ -1,9 +1,9 @@
 {
-  "best_metric": 0.3648131787776947,
-  "best_model_checkpoint": "./working/checkpoint-8000",
-  "epoch": 95.23809523809524,
+  "best_metric": 0.3642527461051941,
+  "best_model_checkpoint": "./working/checkpoint-9000",
+  "epoch": 107.14285714285714,
   "eval_steps": 1000,
-  "global_step": 8000,
+  "global_step": 9000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -647,6 +647,86 @@
       "eval_steps_per_second": 4.029,
       "eval_wer": 0.6170212765957447,
       "step": 8000
+    },
+    {
+      "epoch": 96.43,
+      "grad_norm": 2.8603055477142334,
+      "learning_rate": 7.14e-06,
+      "loss": 0.9576,
+      "step": 8100
+    },
+    {
+      "epoch": 97.62,
+      "grad_norm": 2.931117296218872,
+      "learning_rate": 6.7650000000000005e-06,
+      "loss": 0.9579,
+      "step": 8200
+    },
+    {
+      "epoch": 98.81,
+      "grad_norm": 3.449780225753784,
+      "learning_rate": 6.39e-06,
+      "loss": 0.9535,
+      "step": 8300
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 4.3435468673706055,
+      "learning_rate": 6.015000000000001e-06,
+      "loss": 0.9463,
+      "step": 8400
+    },
+    {
+      "epoch": 101.19,
+      "grad_norm": 2.2839837074279785,
+      "learning_rate": 5.64e-06,
+      "loss": 0.9413,
+      "step": 8500
+    },
+    {
+      "epoch": 102.38,
+      "grad_norm": 3.1021485328674316,
+      "learning_rate": 5.2649999999999996e-06,
+      "loss": 0.9436,
+      "step": 8600
+    },
+    {
+      "epoch": 103.57,
+      "grad_norm": 2.9421229362487793,
+      "learning_rate": 4.890000000000001e-06,
+      "loss": 0.939,
+      "step": 8700
+    },
+    {
+      "epoch": 104.76,
+      "grad_norm": 2.0578436851501465,
+      "learning_rate": 4.515e-06,
+      "loss": 0.9338,
+      "step": 8800
+    },
+    {
+      "epoch": 105.95,
+      "grad_norm": 3.5860297679901123,
+      "learning_rate": 4.14e-06,
+      "loss": 0.9315,
+      "step": 8900
+    },
+    {
+      "epoch": 107.14,
+      "grad_norm": 2.1002159118652344,
+      "learning_rate": 3.765e-06,
+      "loss": 0.9279,
+      "step": 9000
+    },
+    {
+      "epoch": 107.14,
+      "eval_cer": 0.13245100020206102,
+      "eval_loss": 0.3642527461051941,
+      "eval_runtime": 10.5077,
+      "eval_samples_per_second": 31.976,
+      "eval_steps_per_second": 3.997,
+      "eval_wer": 0.6248053969901401,
+      "step": 9000
     }
   ],
   "logging_steps": 100,
@@ -654,7 +734,7 @@
   "num_input_tokens_seen": 0,
   "num_train_epochs": 120,
   "save_steps": 1000,
-  "total_flos": 2.4098927133551096e+19,
+  "total_flos": 2.7115316842745717e+19,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null