Run 4. Outer Step 0. Inner Step 2.

config.json

@@ -268,18 +268,27 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5609120,
-    5609146,
-    5609150,
-    5609155,
-    5609160
+    5609166,
+    5609170,
+    5609175,
+    5609179,
+    5609184,
+    5609188,
+    5609193,
+    5609197,
+    5609202,
+    5609206,
+    5609211,
+    5609215,
+    5609219,
+    5609223
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 0,
+  "inner_step": 2,
   "inner_steps": 0,
   "last_allreduce_block": 5470731,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:50852309c11bba1c68b1fc599de303950b01245d3f8e70c6ec061ebf1229b0c3
-size 2944
+oid sha256:5c26c5891d268e84c07f659a8ff1b7392c90189b255580ebaa0ef440de24ed1c
+size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:365a8be76326501ea811276a569f0cc28ad07a1bacda5111d7247e4a758ad97f
+oid sha256:e68a816e593fdae4f7dd816dae1353dc7d7a30448966f973289744657c8eb3bb
 size 4040701744