Outer Step 1. Inner Step 361. Batch Size 392

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5146660,
-    5146664,
-    5146669,
-    5146674,
-    5146679,
-    5146684,
-    5146689,
-    5146693,
-    5146698,
-    5146703
+    5146737,
+    5146742,
+    5146746,
+    5146751,
+    5146756,
+    5146761,
+    5146766,
+    5146770,
+    5146775,
+    5146780
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 352,
+  "inner_step": 361,
   "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 sha256:870a760538fbe0bc874bb740ea4b95b5fa5989ddd9cc706f8cf79ddc4187d062
+oid sha256:5cf92371a5b620b0f608d19c7b243bf00ef4dcb791ebdfea9f15a93286679b0f
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:82bf49be860f05c87ddb195df0f7987ea634349039a8af2891f1fe35c39ff2fd
+oid sha256:d739464d476a48a8527e1e35b70da5204a29dcb7ae595fdb8b455eb7ec881b96
 size 4040701744