Run 4. Outer Step 18. Inner Step 21.

config.json

@@ -268,19 +268,19 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5521723,
-    5521727,
-    5521730,
-    5521734,
-    5521738,
-    5521741
+    5521746,
+    5521750,
+    5521754,
+    5521757,
+    5521760,
+    5521764
   ],
   "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": 5509976,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ec96b07a149e50f49059a3b6895ac10e00a5b21948b0b60200cb8cfe47e2caf9
+oid sha256:c95da08550a7a75511b097bcb19bba4d266aea85c4f010a890a57977e88ecbbf
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4e0552843dd29dc2bc72fee569ed27aedea73f79e6d884834842d5b7d9440010
+oid sha256:28e56f173cabe6f66875e17f49112ef6de823e5d5b5f143a5e5597aa279af757
 size 4040701744