Outer Step 0. Inner Step 104. Batch Size 132

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5155604,
-    5155608,
-    5155613,
-    5155618,
-    5155623,
-    5155628,
-    5155633,
-    5155637,
-    5155642,
-    5155647
+    5155684,
+    5155689,
+    5155694,
+    5155699,
+    5155704,
+    5155708,
+    5155712,
+    5155716,
+    5155719,
+    5155725
   ],
   "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": 4016775,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:23957ca11f66f4b6a190961e83747bfb715eb9c57131cc8a29426eacb4b8fbe7
+oid sha256:363a226383b390fadd207e77b3aece1689cf6578af29ce45077f2b53d9fcd446
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:88e93a28dd11e7613ef678c37b9c8169530df15240487ad5deb31dca9580eea3
+oid sha256:ca090293f5b3a464adc9b72ff8d819c81b7f071df28a75aeb1a951db31ee6125
 size 4040701744