Training in progress, step 2000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8a6231b9a43188b3030be892c139d7894a62c3fea99c8c6c7eb17b00365a57d7
+oid sha256:5284a72c66df303e073df72ce9f1ae1e29ea633668e40d9822a8ccaa64add555
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:61e1d1dea84b6c812c4ddd82f2f7f4ea581df5126382ca1bf3a48d5758288836
+oid sha256:e8358e2ab5d7859161edb7560d72769c1dccae73dcd94f8bb280a6f4c6a418e5
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:923a3b0ad55df1dc931abd14ea79f0c2ae7566e46df76993edc14d667615816a
+oid sha256:b7c9308b4bec49af406c284201ce7a6649c508431fa62b7fd440c998b6a016ca
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42de46b2e4c5179833d46fe28b80f469c3d40255a80ec3a104ef7ad1fd3d8210
-size 14180
+oid sha256:2512ba1bd332fd1303d75224b66504a865cf77c5c151bdcdfcda52c23dbe2db6
+size 14244

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3adce9cc9139db0a326056a6714b81b8c58d21b77098b37dae23cf33f822452c
+oid sha256:b71627c700396d7041b5a8f195ae9cbb8512fe53c30c6da53a207b7194343ba8
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f03648c25873beacb3afe9bdbd6315225ba0d09c0be3adf22a9ad7dc42cafe47
+oid sha256:2d83848f1b1bfd2394c14189ddfec57565343ac6855bbb1685fedf0549ec3643
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 1250,
-  "best_metric": 4.76948356628418,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-1000",
-  "epoch": 0.1153757403276671,
+  "best_global_step": 2000,
+  "best_metric": 4.666446208953857,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-2000",
+  "epoch": 0.15383432043688947,
   "eval_steps": 250,
-  "global_step": 1500,
+  "global_step": 2000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -266,6 +266,92 @@
       "eval_samples_per_second": 53.979,
       "eval_steps_per_second": 13.495,
       "step": 1500
+    },
+    {
+      "epoch": 0.11922159833858934,
+      "grad_norm": 3.392902135848999,
+      "learning_rate": 0.0001945822403449082,
+      "loss": 4.859,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12306745634951158,
+      "grad_norm": 4.783381938934326,
+      "learning_rate": 0.00019432252032309172,
+      "loss": 4.7791,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12691331436043382,
+      "grad_norm": 3.96708345413208,
+      "learning_rate": 0.00019406280030127523,
+      "loss": 4.7743,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13075917237135604,
+      "grad_norm": 5.381892681121826,
+      "learning_rate": 0.00019380308027945874,
+      "loss": 4.8915,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13460503038227828,
+      "grad_norm": 3.269899606704712,
+      "learning_rate": 0.00019354336025764228,
+      "loss": 4.8823,
+      "step": 1750
+    },
+    {
+      "epoch": 0.13460503038227828,
+      "eval_loss": 4.689135551452637,
+      "eval_runtime": 18.4665,
+      "eval_samples_per_second": 54.152,
+      "eval_steps_per_second": 13.538,
+      "step": 1750
+    },
+    {
+      "epoch": 0.13845088839320052,
+      "grad_norm": 2.5163233280181885,
+      "learning_rate": 0.0001932836402358258,
+      "loss": 4.8513,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14229674640412276,
+      "grad_norm": 4.244402885437012,
+      "learning_rate": 0.0001930239202140093,
+      "loss": 4.6855,
+      "step": 1850
+    },
+    {
+      "epoch": 0.146142604415045,
+      "grad_norm": 6.551025867462158,
+      "learning_rate": 0.00019276420019219285,
+      "loss": 4.8173,
+      "step": 1900
+    },
+    {
+      "epoch": 0.14998846242596722,
+      "grad_norm": 2.791435956954956,
+      "learning_rate": 0.00019250448017037633,
+      "loss": 4.6983,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15383432043688947,
+      "grad_norm": 3.4012036323547363,
+      "learning_rate": 0.00019224476014855984,
+      "loss": 4.8178,
+      "step": 2000
+    },
+    {
+      "epoch": 0.15383432043688947,
+      "eval_loss": 4.666446208953857,
+      "eval_runtime": 18.3775,
+      "eval_samples_per_second": 54.414,
+      "eval_steps_per_second": 13.604,
+      "step": 2000
     }
   ],
   "logging_steps": 50,