Outer Step 0. Inner Step 157. Batch Size 452

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5141645,
-    5141663,
-    5141666,
-    5141671,
-    5141679,
-    5141684,
-    5141689,
-    5141696,
-    5141706,
-    5141712
+    5141740,
+    5141751,
+    5141760,
+    5141768,
+    5141775,
+    5141781,
+    5141789,
+    5141796,
+    5141802,
+    5141812
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 150,
+  "inner_step": 157,
   "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 sha256:3ecba5c9ccdc9b6047c0fc9ac24eab19c64d1550f996591db000014a7cfacc50
+oid sha256:2ef6fefca58c0cf9cbed11b17995e192de529f09a27607bf242b44d1f8b3f073
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b6a26a6c52be47d2c1808740899ad022b8ca3f01706200d729eab6d72e67cadb
+oid sha256:59134d3ef25da687a7b726221aab67ca6d9f63f39d96b2f742c312c9c386b5f8
 size 4040701744