Training in progress, step 20000

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec52caad045b3f34632491a3e60266fca49fb4a8996857ded69858dac82d39b4
+oid sha256:6e887bb747da025cd1ec575f2a274df97b695997849c7b157e69b159e3552f28
 size 202193937

last-checkpoint/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c4934ec353ee8c0b201d8896e01193b73d8d525ed3d2c04d990e1d6af1e4f1ff
+oid sha256:c840f1ee65fbec232eb9372c840a41795b59843a5f5fe59c38c627af00e5f47a
 size 102501541

last-checkpoint/rng_state_0.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_1.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_2.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_3.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_4.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_5.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_6.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/rng_state_7.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:161364992086d421e640cb46d16fb2f26cefdd719c2e2fd276cf5231abbcc82c
+oid sha256:31462fb462555c2db423119a4c7fdb710f8856a89d18c7416080095d8e371b58
 size 14503

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4931c97be633a6e3808391ac03b792837805ded64ff2a578babfa79e60dc22a7
+oid sha256:c84272d9e8a15cd78c706d16b77833d53dbfbd182e8ad79e3cd658ef6c3eaaf6
 size 623

last-checkpoint/trainer_state.json

@@ -1,8 +1,8 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.22305496074232692,
-  "global_step": 10000,
+  "epoch": 0.44610992148465384,
+  "global_step": 20000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -206,11 +206,211 @@
       "eval_samples_per_second": 1186.109,
       "eval_steps_per_second": 18.589,
       "step": 10000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 0.6235,
+      "step": 10500
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 0.6146,
+      "step": 11000
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.6032013893127441,
+      "eval_runtime": 1.9441,
+      "eval_samples_per_second": 1181.504,
+      "eval_steps_per_second": 18.517,
+      "step": 11000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000138,
+      "loss": 0.604,
+      "step": 11500
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 0.5936,
+      "step": 12000
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5801298022270203,
+      "eval_runtime": 1.9051,
+      "eval_samples_per_second": 1205.71,
+      "eval_steps_per_second": 18.897,
+      "step": 12000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00015,
+      "loss": 0.583,
+      "step": 12500
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.000156,
+      "loss": 0.573,
+      "step": 13000
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.5577847361564636,
+      "eval_runtime": 1.8487,
+      "eval_samples_per_second": 1242.481,
+      "eval_steps_per_second": 19.473,
+      "step": 13000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000162,
+      "loss": 0.5636,
+      "step": 13500
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000168,
+      "loss": 0.5544,
+      "step": 14000
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.5371876955032349,
+      "eval_runtime": 1.956,
+      "eval_samples_per_second": 1174.317,
+      "eval_steps_per_second": 18.405,
+      "step": 14000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 0.546,
+      "step": 14500
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.5382,
+      "step": 15000
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5219064354896545,
+      "eval_runtime": 1.9007,
+      "eval_samples_per_second": 1208.524,
+      "eval_steps_per_second": 18.941,
+      "step": 15000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000186,
+      "loss": 0.5291,
+      "step": 15500
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 0.5214,
+      "step": 16000
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5052428841590881,
+      "eval_runtime": 1.9463,
+      "eval_samples_per_second": 1180.202,
+      "eval_steps_per_second": 18.497,
+      "step": 16000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.000198,
+      "loss": 0.5141,
+      "step": 16500
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.000204,
+      "loss": 0.5074,
+      "step": 17000
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.4891608655452728,
+      "eval_runtime": 1.9578,
+      "eval_samples_per_second": 1173.284,
+      "eval_steps_per_second": 18.388,
+      "step": 17000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 0.5006,
+      "step": 17500
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 0.4943,
+      "step": 18000
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.4745289087295532,
+      "eval_runtime": 1.9643,
+      "eval_samples_per_second": 1169.354,
+      "eval_steps_per_second": 18.327,
+      "step": 18000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 0.4886,
+      "step": 18500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 0.4834,
+      "step": 19000
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.4653691351413727,
+      "eval_runtime": 1.8946,
+      "eval_samples_per_second": 1212.382,
+      "eval_steps_per_second": 19.001,
+      "step": 19000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.000234,
+      "loss": 0.4781,
+      "step": 19500
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 0.4727,
+      "step": 20000
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.4524119198322296,
+      "eval_runtime": 1.8731,
+      "eval_samples_per_second": 1226.293,
+      "eval_steps_per_second": 19.219,
+      "step": 20000
     }
   ],
   "max_steps": 500000,
   "num_train_epochs": 12,
-  "total_flos": 3.194871387745e+20,
+  "total_flos": 6.38974277549e+20,
   "trial_name": null,
   "trial_params": null
 }

pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c4934ec353ee8c0b201d8896e01193b73d8d525ed3d2c04d990e1d6af1e4f1ff
+oid sha256:c840f1ee65fbec232eb9372c840a41795b59843a5f5fe59c38c627af00e5f47a
 size 102501541