Run 4. Outer Step 28. Inner Step 21.

config.json

@@ -268,19 +268,19 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5537631,
-    5537635,
-    5537639,
-    5537643,
-    5537648,
-    5537652
+    5537657,
+    5537661,
+    5537665,
+    5537669,
+    5537674,
+    5537678
   ],
   "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": 5531651,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:41f92d2e3485881a457e428b72695b024e38136c49cee3046e08783a0e50d2bb
+oid sha256:29f20695b5d19439db82f54dca01a2247722de945a31493423d9b2714dd62398
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a19f6e81998cf7dd07cf2053c278b83041c2cc326d1dc0208f2f79da26cd72ea
+oid sha256:c6b15051d60e0adf6938b932d3b8fe659e7f162284b9a7d668b8864f2a48aea9
 size 4040701744