Outer Step 8. Inner Step 89. Batch Size 396

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5291853,
-    5291857,
-    5291861,
-    5291864,
-    5291868,
-    5291872,
-    5291875,
-    5291879,
-    5291882,
-    5291886
+    5291930,
+    5291934,
+    5291937,
+    5291941,
+    5291944,
+    5291947,
+    5291951,
+    5291954,
+    5291958,
+    5291962
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 78,
+  "inner_step": 89,
   "inner_steps": 0,
   "last_allreduce_block": 5283755,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:392b8ab21d45a589a0357894a1e14824a22505de279ae5a1e8b60d9b900f3786
+oid sha256:8bbd76c7bbf48ceecb3700eee5c3f4d974afd4f5fdf9d2226ba4c4f58a678800
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:af12813d0c9e25e4d5d400a257d1e7f540de01eaf796f428b11686b5590e3bd6
+oid sha256:293ce37f0bc8c3f900961e151e7d139a91c11a8219a3c01e7efd1da2e88750d8
 size 4040701744