Run 4. Outer Step 4. Inner Step 9.

config.json

@@ -268,21 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5486650,
-    5486652,
-    5486655,
-    5486657,
-    5486660,
-    5486663,
-    5486665,
-    5486667
+    5486670,
+    5486674,
+    5486677,
+    5486679,
+    5486682
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 8,
+  "inner_step": 9,
   "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 sha256:4b1cc19c344da07c57a85744c82c9474123a6a8d88826118391b48a5d7b89c93
+oid sha256:73532076f095037232d71e042ef576e4a43ab6db076a079eeb3060439cb1fe5c
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aeba61a4884234f04f616ca34c80be6a87f0870e6bddc958eb7affb1c57dfebf
+oid sha256:12f9996ddb4d5fc39b50630c59397028237f79728754085c139415fe2726052b
 size 4040701744