Run 4. Outer Step 0. Inner Step 3.

config.json

@@ -268,27 +268,17 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5609166,
-    5609170,
-    5609175,
-    5609179,
-    5609184,
-    5609188,
-    5609193,
-    5609197,
-    5609202,
-    5609206,
-    5609211,
-    5609215,
-    5609219,
-    5609223
+    5609229,
+    5609234,
+    5609239,
+    5609244
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 2,
+  "inner_step": 3,
   "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 sha256:5c26c5891d268e84c07f659a8ff1b7392c90189b255580ebaa0ef440de24ed1c
+oid sha256:1706debd5143351e7a9f71d01cd9b28f8696ccdae2c24c7f0aae36bd8349ab26
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e68a816e593fdae4f7dd816dae1353dc7d7a30448966f973289744657c8eb3bb
+oid sha256:0b1a4d4c0617e3878feb8c3e738c232965424044bf9fefd795bfa073e7e2eaf9
 size 4040701744