Outer Step 3. Inner Step 5. Batch Size 448

config.json

@@ -268,18 +268,18 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5076021,
-    5076024,
-    5076028,
-    5076044,
-    5076048
+    5076062,
+    5076066,
+    5076069,
+    5076072,
+    5076076
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 1,
+  "inner_step": 5,
   "inner_steps": 0,
   "last_allreduce_block": 5063053,
   "layer_norm_epsilon": 1e-05,

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e750574176d02ab0f0ca58dba6a62712f57bbe526c28f6610d8d1c9f3cc14737
+oid sha256:909b23598bb80ef40cd3f8942b2f51d9945424aa965b60b568993130bc3920ea
 size 4040701744