Run 4. Outer Step 20. Inner Step 18.

config.json

@@ -268,23 +268,27 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5525332,
-    5525334,
-    5525337,
-    5525339,
-    5525342,
-    5525345,
-    5525348,
-    5525349,
-    5525352,
-    5525355
+    5525358,
+    5525361,
+    5525363,
+    5525364,
+    5525366,
+    5525368,
+    5525369,
+    5525371,
+    5525372,
+    5525374,
+    5525376,
+    5525378,
+    5525380,
+    5525382
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 17,
+  "inner_step": 18,
   "inner_steps": 0,
   "last_allreduce_block": 5523943,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42aac8df9ca8604893eaf6eab3526f82183a936e8cb41c898b094fa637386642
+oid sha256:45260bfc899649ca42fc2083fc2a238851df85565acd94f76f02ee21f1cbbce8
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e145bf0f05e4a216794f99c908cd55362cb3a8f823a3776082d60d9fdbc7b7e3
+oid sha256:1654ef41c1aedd4635a0f04bb7474c81f568eb840640a3483f45269aff32767d
 size 4040701744