Run 4. Outer Step 0. Inner Step 3.

config.json

@@ -268,18 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5592022,
-    5592026,
-    5592030,
-    5592035,
-    5592039
+    5592108,
+    5592112,
+    5592117,
+    5592121,
+    5592125
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 0,
+  "inner_step": 3,
   "inner_steps": 0,
   "last_allreduce_block": 5470731,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:06872b06f3bac1643a90c4514e3ef63d6c6141dc2e5328adfc37d46d148b03df
-size 3357
+oid sha256:68ea47d4e8c0a0099e782a7889727d7c67823a8499b01df01f77353356ed7cdb
+size 8081782503

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:365a8be76326501ea811276a569f0cc28ad07a1bacda5111d7247e4a758ad97f
+oid sha256:b67d9761458ebeab1ada8f2f3a35931b65aa7fc7ce30ede7ed9433f260567c75
 size 4040701744