Outer Step 2. Inner Step 119. Batch Size 108

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5147652,
-    5147656,
-    5147661,
-    5147665,
-    5147669,
-    5147673,
-    5147677,
-    5147681,
-    5147685,
-    5147689
+    5147722,
+    5147725,
+    5147729,
+    5147733,
+    5147738,
+    5147741,
+    5147745,
+    5147749,
+    5147753,
+    5147757
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 109,
+  "inner_step": 119,
   "inner_steps": 0,
   "last_allreduce_block": 5146835,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:00fc655ac6f0c1b1c13a26115c9bf61de0af59e92c6f74e9f5adf17174d5099d
+oid sha256:84b50b54e3c63a5f8f5aa45a9973a859f14d1f2194517c10ee5be568565582ec
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa061d1d6c0854a74f45b408f7721ae8a194d0ff3e3fa46aca94eafadb9d7506
+oid sha256:29ffb4bb00c2d73ce16e616e1dc7e70f906e395795046670b010b292928ef3a8
 size 4040701744