Outer Step 2. Inner Step 156. Batch Size 360

config.json

@@ -268,23 +268,23 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5147977,
-    5147981,
-    5147985,
-    5147989,
-    5147993,
-    5147996,
-    5148002,
-    5148006,
-    5148010,
-    5148014
+    5148044,
+    5148048,
+    5148053,
+    5148057,
+    5148061,
+    5148065,
+    5148069,
+    5148073,
+    5148077,
+    5148081
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 147,
+  "inner_step": 156,
   "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 sha256:90bf9485ba91196acd7689989f9ae1b4a91cf5beaf9fd1ed9fd44419db31aaa2
+oid sha256:47f2d94353a6fe41e902b15b4394169126a4e563e72d40f87afe7737613ddd8c
 size 8081781770

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:825feb284d0f668b17be474b92bb654155a6c62c158be04cbdbffdc7014eac5c
+oid sha256:f031c603c34a6c3ba0fe61a8ee88758e6ff2ddaed21d7842b55ba5360e1f5db7
 size 4040701744