Run 4. Outer Step 20. Inner Step 17.

config.json

@@ -268,26 +268,27 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5525155,
-    5525157,
-    5525159,
-    5525161,
-    5525162,
-    5525164,
-    5525166,
-    5525168,
-    5525170,
-    5525172,
-    5525173,
-    5525175,
-    5525177
+    5525180,
+    5525182,
+    5525183,
+    5525184,
+    5525186,
+    5525188,
+    5525189,
+    5525190,
+    5525192,
+    5525193,
+    5525194,
+    5525196,
+    5525198,
+    5525200
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 15,
+  "inner_step": 17,
   "inner_steps": 0,
   "last_allreduce_block": 5523943,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2a63054de502f2a4746ed1080c001df31630f1d9fe5fdd34a166cc5469b418ba
+oid sha256:cf7cfdaed7389ba468324b6dd85ae3c9c970b187afc1cb7caa0c404dd96589f7
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fe34c20e15d51b632443f22d1aa6aae66adab17937542eac0d431429cbdede3d
+oid sha256:bea61a6d18e76d13e992ddbd8cc5a9f1e186281bd327c17b4162f008f4e89011
 size 4040701744