Run 4. Outer Step 26. Inner Step 37.

config.json

@@ -268,25 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5535773,
-    5535774,
-    5535777,
-    5535780,
-    5535781,
-    5535783,
-    5535785,
-    5535787,
-    5535788,
-    5535789,
-    5535790,
-    5535792
+    5535794,
+    5535797,
+    5535799,
+    5535802,
+    5535805
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 36,
+  "inner_step": 37,
   "inner_steps": 0,
   "last_allreduce_block": 5534554,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c4b0425fc5e649aedaa58aa5d92ebd30444046ee42eeb2dd559b59aaa0a4737
+oid sha256:434ec5470d1191522bc02f0bdc056b2983b5914945d0f261316c1967c8824758
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:000a2d8f679b7cd799897c5f35eb5a148ff8de63024b56c69c903272b56d7ce8
+oid sha256:c8376c2ebcadee945739701dfd0bea4592d31433d364787bc13ea0eb9929cc02
 size 4040701744