Run 4. Outer Step 4. Inner Step 3.

config.json

@@ -268,20 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5486448,
-    5486450,
-    5486452,
-    5486454,
-    5486458,
-    5486461,
-    5486464
+    5486468,
+    5486472,
+    5486475,
+    5486477,
+    5486480
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 2,
+  "inner_step": 3,
   "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 sha256:454a56f33c7413ac5f2592e2585ed9e6a9dbb7ad513d96a59af49f79ed668990
+oid sha256:4a5c543a8c0d5da23d62770777c0a7fb8466bf1c49fc23d85cc9c7235a225218
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:396c8bbd2930bf443f9b26b8f137b581026f40edf48953be78999ed40d2fea95
+oid sha256:7d2c6b0eda6556b503e7807071c99ed2f5584b0a338a09410f9b126ab0c135bd
 size 4040701744