Run 4. Outer Step 28. Inner Step 21.

config.json

@@ -268,18 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5537637,
-    5537641,
-    5537645,
-    5537649,
-    5537653
+    5537660,
+    5537664,
+    5537668,
+    5537673,
+    5537677
   ],
   "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": 5535854,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b9f28be9ad773dc8a6d968643209975590ec5ba8bf1a31ef73920d73c0dc3847
+oid sha256:cad7b175bc418cfda3d7ea5795996a5cc3b8df7d01a19fe89bf2a8b0dce6ad15
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f7e72b061fd031a0bc64a0318bbe5c33d8482ed4ce129dcd4671c97c96b41d6c
+oid sha256:461f22492050f1a46cbd6af6557c77a66b3c0c9681c607fec66d8df810366177
 size 4040701744