Outer Step 0. Inner Step 112. Batch Size 336

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5140205,
-    5140208,
-    5140212,
-    5140215,
-    5140219,
-    5140222,
-    5140226,
-    5140229,
-    5140233,
-    5140236
+    5140372,
+    5140412,
+    5140455,
+    5140525,
+    5140566,
+    5140665,
+    5140702,
+    5140722,
+    5140766,
+    5140799
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 104,
+  "inner_step": 112,
   "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:bffda1e9633bb600ba31098672c32e3aa11034e85f05992755aa690f6f5404c1
+oid sha256:76f3c3d028d99eeb57af329a6b0207bc68d4f1aa523ab2474467d6352124d0c5
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1fe82a7ea4316050b496bcd668ca4540c332edba560080e94a6645c60f0ed24a
+oid sha256:41faa70560f53fd26efd73421fb84a48c7da227f042461865a9da6976a22947e
 size 4040701744