Run 4. Outer Step 28. Inner Step 38.

config.json

@@ -268,18 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5538205,
-    5538208,
-    5538211,
-    5538214,
-    5538216
+    5538218,
+    5538221,
+    5538223,
+    5538224,
+    5538226,
+    5538228,
+    5538230,
+    5538233,
+    5538235,
+    5538238
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 37,
+  "inner_step": 38,
   "inner_steps": 0,
   "last_allreduce_block": 5537093,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a92f4cf833a687dd14202bd54e0b41a990c178232db764a119815f4f3bd15b87
+oid sha256:fc7f06c963b88d503d910d5e51eeee7403bfa93f2f9d1363ca95c8eee30edea1
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:25d6b47aa13dc7cb8bf07199b00adcadbc1f7700fe500e6ddb110ac720b510d9
+oid sha256:3af7e8ffeea0b65bbdf762dc92f326b1879c803271a8f35a3e5a39769d630028
 size 4040701744