Run 4. Outer Step 12. Inner Step 3.

config.json

@@ -268,18 +268,20 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5508517,
-    5508521,
-    5508525,
-    5508529,
-    5508532
+    5508536,
+    5508540,
+    5508543,
+    5508547,
+    5508551,
+    5508554,
+    5508558
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 2,
+  "inner_step": 3,
   "inner_steps": 0,
   "last_allreduce_block": 5506953,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1d140bef646993ba3bd38a9d6a1bc6afd52e8ea312acfa0363d6e94c8a4c600c
+oid sha256:8caed6a3e67f412ef3382898d36fc2ce84b8821c1845861cdef4f91287ec8293
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:839b770e8fe3c7ca2a41212f487a4611b3f2707a4c9eec4c725dae41f51e34b9
+oid sha256:20eca89e5845f9117efb9a779eda2bd1423a6764202b6d7cea99a2af7a28d295
 size 4040701744