Outer Step 0. Inner Step 86. Batch Size 160

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5155463,
-    5155469,
-    5155473,
-    5155478,
-    5155481,
-    5155487,
-    5155491,
-    5155495,
-    5155499,
-    5155503
+    5155534,
+    5155538,
+    5155542,
+    5155546,
+    5155550,
+    5155554,
+    5155558,
+    5155563,
+    5155568,
+    5155572
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 77,
+  "inner_step": 86,
   "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:de434915dc2cb8051a843d7f34f020e42e32c869f8acd85433084c48559612cd
+oid sha256:b0c9b6c787d81a5389967f07ae6c692539a551f355c494f5877d759f9a13971e
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42ee875f77c91f8ec0903582951cf65de828a4ede0d70fa81fe987af66d60a6f
+oid sha256:c34eeb388e8970ea61ba9ec2a33a8e3498411f3ad0d1170116cad9ba3e14c91b
 size 4040701744