Outer Step 0. Inner Step 104. Batch Size 20

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5140150,
-    5140154,
-    5140158,
-    5140161,
-    5140165,
-    5140168,
-    5140171,
-    5140174,
-    5140178,
-    5140181
+    5140205,
+    5140208,
+    5140212,
+    5140215,
+    5140219,
+    5140222,
+    5140226,
+    5140229,
+    5140233,
+    5140236
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 94,
+  "inner_step": 104,
   "inner_steps": 0,
   "last_allreduce_block": 5072405,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6af3dd51a28ba6f05997bfca4725072f684ca84fcbda3ab6ab744d67a83be0f8
+oid sha256:bffda1e9633bb600ba31098672c32e3aa11034e85f05992755aa690f6f5404c1
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:485651178d4004c3ad575ee2320837162a0cc608f8177cd04a073563ee257fc5
+oid sha256:1fe82a7ea4316050b496bcd668ca4540c332edba560080e94a6645c60f0ed24a
 size 4040701744