Run 3. Outer Step 11. Inner Step 5.

config.json

@@ -268,18 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5424744,
-    5424748,
-    5424752,
-    5424755,
-    5424759
+    5424795,
+    5424799,
+    5424803,
+    5424807,
+    5424811
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 3,
+  "inner_step": 5,
   "inner_steps": 0,
   "last_allreduce_block": 5422706,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a21de9db33c7ba24e2f5d5f220208b7531eec6d951cea4a2ab9d56a0a06af5bd
+oid sha256:fd2cf50031f98c0001d2b36207bbad22e76c7239921ef83e3e820e73b7e17f6d
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:78e6a28b9b192d25cc44dbf6496981713934099ecbea746277df3176a0238e56
+oid sha256:a418b4a5bd98fe468d814248730f278e8383b592bf3925c56b3a3e812d4e6b75
 size 4040701744