Run 4. Outer Step 49. Inner Step 63.

config.json

@@ -268,28 +268,40 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5571416,
-    5571418,
-    5571420,
-    5571421,
-    5571422,
-    5571424,
-    5571426,
-    5571427,
-    5571427,
-    5571428,
-    5571429,
-    5571431,
-    5571433,
-    5571435,
-    5571437
+    5571439,
+    5571441,
+    5571443,
+    5571444,
+    5571445,
+    5571447,
+    5571448,
+    5571449,
+    5571450,
+    5571451,
+    5571452,
+    5571453,
+    5571454,
+    5571456,
+    5571458,
+    5571459,
+    5571460,
+    5571461,
+    5571463,
+    5571465,
+    5571466,
+    5571468,
+    5571469,
+    5571470,
+    5571471,
+    5571472,
+    5571473
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 61,
+  "inner_step": 63,
   "inner_steps": 0,
   "last_allreduce_block": 5566279,
   "layer_norm_epsilon": 1e-05,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:641c6b4306952e92f789cc8ea6af2850e61c123ee468047b87843ee5e9f70b6d
+oid sha256:cf0a03120c9bb7c5f845844a51e2e5ed0a23c66a00df52c0ee067f52146616c9
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d34bf33c10c416fca29812b1f1e499b6fbffcc878846d769f3e5c05e83ab8012
+oid sha256:6e7b709a8825dd08467510bfd85c142896756ecad56a84198dcc7bb7d48285fd
 size 4040701744