Outer Step 1. Inner Step 400. Batch Size 224

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5146671,
-    5146675,
-    5146679,
-    5146683,
-    5146687,
-    5146691,
-    5146695,
-    5146698,
-    5146703,
-    5146707
+    5146740,
+    5146743,
+    5146747,
+    5146751,
+    5146755,
+    5146760,
+    5146764,
+    5146768,
+    5146772,
+    5146776
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 390,
+  "inner_step": 400,
   "inner_steps": 0,
   "last_allreduce_block": 5143202,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a316e358f22db16503474f9b59485daa8b50f1eb8990cf8e0ad91d6d7c5b11fb
+oid sha256:97064e31452d86289044a24d4589791eff6984e713c32dc04b5fedf45df8f853
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad9873b70289735f467d6c563f40575bdfde34c6424344d1048ddc5cc13e031b
+oid sha256:6058ef6a9c86d2d2c3717f25789a71d7e473c5c0115f53cc61406274fe6d6074
 size 4040701744