Outer Step 0. Inner Step 225. Batch Size 404

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5132752,
-    5132757,
-    5132763,
-    5132768,
-    5132774,
-    5132784,
-    5132790,
-    5132797,
-    5132802,
-    5132808
+    5132840,
+    5132852,
+    5132858,
+    5132864,
+    5132871,
+    5132877,
+    5132883,
+    5132888,
+    5132896,
+    5132902
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 217,
+  "inner_step": 225,
   "inner_steps": 0,
   "last_allreduce_block": 5072405,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0fca57a8f6acfd5e7a6f43e587b859d1cb1dab46c40ee464f4609dc19ce06b57
+oid sha256:168e53dfa269ee22d18ffc96e089154f4ca458d35960375086dd453b216aa6ce
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:494a51793759423c7c22c8f5c9c68252ce2d47d11174f0e000344fcce0c7e10e
+oid sha256:b6faebfe3957cc11fcab141db50aa66abfab2222662e82b0fa5239929a2cb3fb
 size 4040701744