Run 4. Outer Step 23. Inner Step 21.

config.json

@@ -268,22 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5531513,
-    5531516,
-    5531519,
-    5531522,
-    5531524,
-    5531527,
-    5531529,
-    5531531,
-    5531533
+    5531536,
+    5531538,
+    5531540,
+    5531541,
+    5531542
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 20,
+  "inner_step": 21,
   "inner_steps": 0,
   "last_allreduce_block": 5529444,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4729fa6fa36debb84b2f53e0aeb7208c39a1c3031c55e43ee05177bdea25a580
+oid sha256:709921ba056365e9cf11db7a0a024c4aef962928f30039be88b3d108cb1b2411
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa68b700c6be272584ce983de9c87a7b2d45afb2f874e5db95bdd6618e7fd51e
+oid sha256:3a38d5e2dca911f49fae065c19c469636695e6c1047124bcfa635f1e35f56568
 size 4040701744