Outer Step 4. Inner Step 34. Batch Size 4

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5149159,
-    5149163,
-    5149168,
-    5149172,
-    5149176,
-    5149180,
-    5149184,
-    5149188,
-    5149193,
-    5149197
+    5149226,
+    5149231,
+    5149235,
+    5149239,
+    5149244,
+    5149248,
+    5149252,
+    5149256,
+    5149261,
+    5149265
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 24,
+  "inner_step": 34,
   "inner_steps": 0,
   "last_allreduce_block": 5148895,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e9494ae4cf944f393c44fbf2df5e6b3ee0e78f0b302f51d5d574ef84c9288060
+oid sha256:e21ea25c1ff42d5f5898c8c6fe96895c4a1e48d81221d8e5dc235492087e6ef8
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6c766d338eafae6766d35f5063495beda05a0a90861099fcd03b355809880612
+oid sha256:1632e311698fd0db9f807f3c0aae67d87b8db3acbbd43630db3044a328de55d9
 size 4040701744