Run 4. Outer Step 40. Inner Step 46.

config.json

@@ -268,18 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5556470,
-    5556471,
-    5556474,
-    5556477,
-    5556481
+    5556483,
+    5556484,
+    5556487,
+    5556488,
+    5556490,
+    5556491,
+    5556493,
+    5556495,
+    5556496,
+    5556498
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 45,
+  "inner_step": 46,
   "inner_steps": 0,
   "last_allreduce_block": 5555115,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b91195e11147e9871f946efacf0dbe0f9ff60c36ab7768a590c66ada19ec97d7
+oid sha256:ab92f9d6e4b5a45e8ef328e7da3df0ab7aabce6114fd5544b13520618f5d2bf1
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:397ddd2e99dc03fa70f7464de6f956dc68ce3e21a38c0f6baefb3cea2db92412
+oid sha256:a4ece6ca3e4486c0628963c0918a994d879e43b28b3c779089100cc1c819e26a
 size 4040701744