Training in progress, step 3000, checkpoint

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f2b4cbe8b1b80bab80643ae95440812326a8a6e7bccc79e0f31fd0b001f92483
+oid sha256:e598bf8b3c9e467b65b1bb82c8cbdd7f6d9c1972d320fbed970e890963dcdc5d
 size 36730224

last-checkpoint/ar_diffusion_info.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:683cb6e78851437b6d551b60a7733fd0cf8a96582a781e0a27157158ccd47347
+oid sha256:7d2d6f677b30f619e4a135ea8b093f0ab4667e8aba8c81cd7afaa73bd22e2ff8
 size 1544

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:142ce534c6cf070debd44d21ea1fe5c477f0438570d559dbda8928fe80687677
+oid sha256:4dc13a71e65c3e33eb37c20ebdf6182925491cef76344bbf33cf66c0cd5d4873
 size 73588346

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:74818f2f3daacf3c0fca3dde38a211872e713dacdefaa18d7435559b30dc197d
-size 14244
+oid sha256:0cdad770e6415de25f2503b80a04125ce36c191e7ebbe40cc70935c15b3079b5
+size 14180

last-checkpoint/scaler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21c1bf6c47692563bb8660483ef35d0f3e0d75e9fd318c04f77453b710e51e5c
+oid sha256:1fd6bb9f68e7b13e6013a02cf99adaeff381c90e8c38480bfbcc7ea752fc628b
 size 988

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:eeeb94e0675525b5bb02698eb79efe748a90c2efeae2eda028133bdd4c5927c0
+oid sha256:1e2f4f94b8df8c39a3dbd1dfca81bfb52c1274fbec2120e8d28a3f0e8b8a87c3
 size 1064

last-checkpoint/trainer_state.json

@@ -1,10 +1,10 @@
 {
-  "best_global_step": 2500,
-  "best_metric": 4.597048282623291,
-  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-2500",
-  "epoch": 0.19229290054611184,
+  "best_global_step": 3000,
+  "best_metric": 4.579595565795898,
+  "best_model_checkpoint": "./ar-diffusion-checkpoints-fixed/checkpoint-3000",
+  "epoch": 0.2307514806553342,
   "eval_steps": 250,
-  "global_step": 2500,
+  "global_step": 3000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -438,6 +438,92 @@
       "eval_samples_per_second": 54.025,
       "eval_steps_per_second": 13.506,
       "step": 2500
+    },
+    {
+      "epoch": 0.19613875855703408,
+      "grad_norm": 2.59525465965271,
+      "learning_rate": 0.00018938783990857857,
+      "loss": 4.4364,
+      "step": 2550
+    },
+    {
+      "epoch": 0.19998461656795632,
+      "grad_norm": 3.844686985015869,
+      "learning_rate": 0.00018912811988676208,
+      "loss": 4.6437,
+      "step": 2600
+    },
+    {
+      "epoch": 0.20383047457887854,
+      "grad_norm": 3.4633946418762207,
+      "learning_rate": 0.0001888683998649456,
+      "loss": 4.5016,
+      "step": 2650
+    },
+    {
+      "epoch": 0.20767633258980078,
+      "grad_norm": 3.7852296829223633,
+      "learning_rate": 0.0001886086798431291,
+      "loss": 4.5845,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21152219060072303,
+      "grad_norm": 3.8716065883636475,
+      "learning_rate": 0.00018834895982131265,
+      "loss": 4.3669,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21152219060072303,
+      "eval_loss": 4.602295875549316,
+      "eval_runtime": 18.4747,
+      "eval_samples_per_second": 54.128,
+      "eval_steps_per_second": 13.532,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21536804861164527,
+      "grad_norm": 3.9932167530059814,
+      "learning_rate": 0.00018808923979949616,
+      "loss": 4.4545,
+      "step": 2800
+    },
+    {
+      "epoch": 0.21921390662256748,
+      "grad_norm": 4.182176113128662,
+      "learning_rate": 0.00018782951977767967,
+      "loss": 4.5355,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22305976463348973,
+      "grad_norm": 3.776895523071289,
+      "learning_rate": 0.00018756979975586318,
+      "loss": 4.6859,
+      "step": 2900
+    },
+    {
+      "epoch": 0.22690562264441197,
+      "grad_norm": 3.9219324588775635,
+      "learning_rate": 0.0001873100797340467,
+      "loss": 4.5186,
+      "step": 2950
+    },
+    {
+      "epoch": 0.2307514806553342,
+      "grad_norm": 3.7879323959350586,
+      "learning_rate": 0.0001870503597122302,
+      "loss": 4.612,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2307514806553342,
+      "eval_loss": 4.579595565795898,
+      "eval_runtime": 18.476,
+      "eval_samples_per_second": 54.124,
+      "eval_steps_per_second": 13.531,
+      "step": 3000
     }
   ],
   "logging_steps": 50,