Run 4. Outer Step 4. Inner Step 6.

config.json

@@ -268,20 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5486536,
-    5486538,
-    5486540,
-    5486542,
-    5486545,
-    5486548,
-    5486551
+    5486554,
+    5486557,
+    5486561,
+    5486563,
+    5486564
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 5,
+  "inner_step": 6,
   "inner_steps": 0,
   "last_allreduce_block": 5485540,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76be166f84ca6e0d4bf6af48331e0f0bb28fbefba7081ec9d84c70aeaf318cc5
+oid sha256:1d4abc6cbb1c5c53473e1a9bff64d950e5ce2d7e81d6bbb69d07ee6f6aa35dd5
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:25f0acf09e0ca60196e8df90c029cf14c9d3f5d151c17aaf1e0686da4ff11a7a
+oid sha256:ae51398dd1a5bb716d556064d39efc320324fd3f5c3a95ee8dc6b1fb2ab27ae3
 size 4040701744