Run 3. Outer Step 10. Inner Step 5.

config.json

@@ -268,18 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5422062,
-    5422066,
-    5422070,
-    5422073,
-    5422076
+    5422106,
+    5422109,
+    5422113,
+    5422116,
+    5422121
   ],
   "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": 5398844,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4e0b7037e10aea41ec6e934d630aaeb9c6224ff0ba2a1935127a104328e81bc0
+oid sha256:168491da68772eb9182472394e07486f99d3f24ec87380dc437950be2dfb2ddb
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:27749ac120853e7d3a7dc72370a5975fed0931cc3e0d3742cea5549a0b0ca566
+oid sha256:a38f5b6a06642bec9c65570e6ade17c107b4065db32d7a128a3b0bb4a2bf3b20
 size 4040701744