Model save

{checkpoint-4200 → checkpoint-4400}/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6ddef0f5049947aa8210403bee49ef1fd880b41459272b4c05c3eb32308f2f44
+oid sha256:2ed693bd3b91cdbbe1430c3f7f43d32b0b9b099aa5180fb31ece0f93c73b26ee
 size 2490337809

{checkpoint-4300 → checkpoint-4400}/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ca11c993941e9cd903a98a8d6a851a925b5bc173fcb10c435092a4ff2bbe4c56
+oid sha256:f3c90a3d0a519652bea358ec5bc82e2a0b669425ce58bf0947c048a13d83b74d
 size 1262063089

{checkpoint-4200 → checkpoint-4400}/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4e03d76f10a67412455e92aefe077a6da7230abf78e55d42dec22dfbaea90dbd
+oid sha256:0bb367ac83047020ca661382c5402cfaed86ec3e8ea3ce8b448d44d9e43e2965
 size 14567

{checkpoint-4300 → checkpoint-4400}/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6924042b31741a307a9c2d3e5d4148105227679127f306862679a760336ca11c
+oid sha256:cb287cb35474e15fb80cdda16160c24d664008c453ecb31933c1042ff81dd3e9
 size 559

{checkpoint-4200 → checkpoint-4400}/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7fb4edd93c6eef5711ff572cfd2e93da18eb0ae8a7738f853f88ca1ecfb5e8b9
+oid sha256:379bb05b55352db422771e61095d894e4be45df2c25710a995432fffb6cdd028
 size 623

{checkpoint-4300 → checkpoint-4400}/trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 47.25136612021858,
-  "global_step": 4300,
+  "epoch": 48.349726775956285,
+  "global_step": 4400,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1683,11 +1683,50 @@
       "eval_steps_per_second": 0.802,
       "eval_wer": 0.12693726937269373,
       "step": 4300
+    },
+    {
+      "epoch": 47.47,
+      "learning_rate": 2.284313725490196e-05,
+      "loss": 0.7326,
+      "step": 4320
+    },
+    {
+      "epoch": 47.69,
+      "learning_rate": 2.088235294117647e-05,
+      "loss": 0.724,
+      "step": 4340
+    },
+    {
+      "epoch": 47.91,
+      "learning_rate": 1.892156862745098e-05,
+      "loss": 0.7316,
+      "step": 4360
+    },
+    {
+      "epoch": 48.13,
+      "learning_rate": 1.696078431372549e-05,
+      "loss": 0.7566,
+      "step": 4380
+    },
+    {
+      "epoch": 48.35,
+      "learning_rate": 1.5e-05,
+      "loss": 0.7479,
+      "step": 4400
+    },
+    {
+      "epoch": 48.35,
+      "eval_loss": 0.1490115076303482,
+      "eval_runtime": 188.3681,
+      "eval_samples_per_second": 25.71,
+      "eval_steps_per_second": 0.807,
+      "eval_wer": 0.12761850695430033,
+      "step": 4400
     }
   ],
   "max_steps": 4550,
   "num_train_epochs": 50,
-  "total_flos": 6.648181588185401e+19,
+  "total_flos": 6.805564586486029e+19,
   "trial_name": null,
   "trial_params": null
 }

{checkpoint-4300 → checkpoint-4500}/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:58cce60fe69424e2a89d307caa661955fd8777a8855b339f2af2ad4573bb989c
+oid sha256:6ddcef79ae9c9cbc49da385ba75dde3728f6718eeb4a1ea9f6af1e9bb92ea574
 size 2490337809

{checkpoint-4200 → checkpoint-4500}/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:588e62e516d26a20bbfce337c36c9f1a83451a67d19e583d1cdceb96aff5859f
+oid sha256:3eae58c0bb9c908100e64bb4cfc44af074bab8941dffb754f675e676ebd7972b
 size 1262063089

{checkpoint-4300 → checkpoint-4500}/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:412f958acca27a8108970c9e2ebea1c88954db1eb284dd8fe51687e899621b23
-size 14503
+oid sha256:7ae13169f4754f47cf02dcd08d7a9d3852557179a00533dd7f00f4265d8a8fba
+size 14567

{checkpoint-4200 → checkpoint-4500}/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7018904ee53324745014ee00f0192d0f3344069918f4c8dd54467134a03ae355
+oid sha256:a4c027f185a945dda35351dc282299cb9a8d301ee07589b42eb812ad50ff4977
 size 559

{checkpoint-4300 → checkpoint-4500}/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d9bfd4ddf4228b39539babfcc9c868a929c00cb0946eb4b8ecfaa5fa12431293
+oid sha256:c583f6b4d18dff47ed1b5d6c7fe67a5bf74f8906e7b32b87ae71e701db2a021c
 size 623

{checkpoint-4200 → checkpoint-4500}/trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 46.15300546448088,
-  "global_step": 4200,
+  "epoch": 49.44808743169399,
+  "global_step": 4500,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -1644,11 +1644,128 @@
       "eval_steps_per_second": 0.798,
       "eval_wer": 0.1271927334657962,
       "step": 4200
+    },
+    {
+      "epoch": 46.37,
+      "learning_rate": 3.264705882352941e-05,
+      "loss": 0.7283,
+      "step": 4220
+    },
+    {
+      "epoch": 46.59,
+      "learning_rate": 3.0686274509803923e-05,
+      "loss": 0.7218,
+      "step": 4240
+    },
+    {
+      "epoch": 46.81,
+      "learning_rate": 2.872549019607843e-05,
+      "loss": 0.7341,
+      "step": 4260
+    },
+    {
+      "epoch": 47.03,
+      "learning_rate": 2.6764705882352942e-05,
+      "loss": 0.7293,
+      "step": 4280
+    },
+    {
+      "epoch": 47.25,
+      "learning_rate": 2.480392156862745e-05,
+      "loss": 0.7194,
+      "step": 4300
+    },
+    {
+      "epoch": 47.25,
+      "eval_loss": 0.14928147196769714,
+      "eval_runtime": 189.625,
+      "eval_samples_per_second": 25.54,
+      "eval_steps_per_second": 0.802,
+      "eval_wer": 0.12693726937269373,
+      "step": 4300
+    },
+    {
+      "epoch": 47.47,
+      "learning_rate": 2.284313725490196e-05,
+      "loss": 0.7326,
+      "step": 4320
+    },
+    {
+      "epoch": 47.69,
+      "learning_rate": 2.088235294117647e-05,
+      "loss": 0.724,
+      "step": 4340
+    },
+    {
+      "epoch": 47.91,
+      "learning_rate": 1.892156862745098e-05,
+      "loss": 0.7316,
+      "step": 4360
+    },
+    {
+      "epoch": 48.13,
+      "learning_rate": 1.696078431372549e-05,
+      "loss": 0.7566,
+      "step": 4380
+    },
+    {
+      "epoch": 48.35,
+      "learning_rate": 1.5e-05,
+      "loss": 0.7479,
+      "step": 4400
+    },
+    {
+      "epoch": 48.35,
+      "eval_loss": 0.1490115076303482,
+      "eval_runtime": 188.3681,
+      "eval_samples_per_second": 25.71,
+      "eval_steps_per_second": 0.807,
+      "eval_wer": 0.12761850695430033,
+      "step": 4400
+    },
+    {
+      "epoch": 48.57,
+      "learning_rate": 1.303921568627451e-05,
+      "loss": 0.7139,
+      "step": 4420
+    },
+    {
+      "epoch": 48.79,
+      "learning_rate": 1.107843137254902e-05,
+      "loss": 0.7345,
+      "step": 4440
+    },
+    {
+      "epoch": 49.01,
+      "learning_rate": 9.117647058823529e-06,
+      "loss": 0.756,
+      "step": 4460
+    },
+    {
+      "epoch": 49.23,
+      "learning_rate": 7.15686274509804e-06,
+      "loss": 0.7196,
+      "step": 4480
+    },
+    {
+      "epoch": 49.45,
+      "learning_rate": 5.196078431372549e-06,
+      "loss": 0.7132,
+      "step": 4500
+    },
+    {
+      "epoch": 49.45,
+      "eval_loss": 0.1501067876815796,
+      "eval_runtime": 190.1989,
+      "eval_samples_per_second": 25.463,
+      "eval_steps_per_second": 0.799,
+      "eval_wer": 0.1265398807834232,
+      "step": 4500
     }
   ],
   "max_steps": 4550,
   "num_train_epochs": 50,
-  "total_flos": 6.496086001622114e+19,
+  "total_flos": 6.9574010203448205e+19,
   "trial_name": null,
   "trial_params": null
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1a459f72ab853af3d30be6993d7c2796ebd50d28c9cae60928258ccc702d6e46
+oid sha256:c284a4869295d3f95db2001f0fef26d374358f6f33b31cf64b9994c12e4cfd91
 size 1262063089