Run 4. Outer Step 53. Inner Step 0.

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "markalan324/minor12",
+  "_name_or_path": "distributed/optimized-gpt2-1b",
   "activation_function": "gelu_new",
   "all_reduce_scores": {
     "0": "NON_PARTICIPATING",
@@ -11,16 +11,16 @@
     "103": "NON_PARTICIPATING",
     "104": "NON_PARTICIPATING",
     "105": "NON_PARTICIPATING",
-    "106": "SUCCESS",
-    "107": "NON_PARTICIPATING",
+    "106": "NON_PARTICIPATING",
+    "107": "SUCCESS",
     "108": "NON_PARTICIPATING",
-    "109": "NON_PARTICIPATING",
-    "11": "NON_PARTICIPATING",
-    "110": "SUCCESS",
+    "109": "SUCCESS",
+    "11": "SUCCESS",
+    "110": "NON_PARTICIPATING",
     "111": "NON_PARTICIPATING",
     "112": "NON_PARTICIPATING",
-    "113": "NON_PARTICIPATING",
-    "114": "SUCCESS",
+    "113": "SUCCESS",
+    "114": "NON_PARTICIPATING",
     "115": "NON_PARTICIPATING",
     "116": "NON_PARTICIPATING",
     "117": "NON_PARTICIPATING",
@@ -30,21 +30,21 @@
     "120": "NON_PARTICIPATING",
     "121": "NON_PARTICIPATING",
     "122": "NON_PARTICIPATING",
-    "123": "SUCCESS",
-    "124": "SUCCESS",
+    "123": "NON_PARTICIPATING",
+    "124": "NON_PARTICIPATING",
     "125": "NON_PARTICIPATING",
-    "126": "NON_PARTICIPATING",
+    "126": "SUCCESS",
     "127": "NON_PARTICIPATING",
     "128": "NON_PARTICIPATING",
     "129": "NON_PARTICIPATING",
     "13": "SUCCESS",
-    "130": "SUCCESS",
+    "130": "NON_PARTICIPATING",
     "131": "NON_PARTICIPATING",
     "132": "NON_PARTICIPATING",
     "133": "NON_PARTICIPATING",
     "134": "NON_PARTICIPATING",
     "135": "NON_PARTICIPATING",
-    "136": "SUCCESS",
+    "136": "NON_PARTICIPATING",
     "137": "NON_PARTICIPATING",
     "138": "SUCCESS",
     "139": "NON_PARTICIPATING",
@@ -52,15 +52,15 @@
     "140": "NON_PARTICIPATING",
     "141": "NON_PARTICIPATING",
     "142": "NON_PARTICIPATING",
-    "143": "SUCCESS",
+    "143": "NON_PARTICIPATING",
     "144": "NON_PARTICIPATING",
     "145": "NON_PARTICIPATING",
     "146": "NON_PARTICIPATING",
     "147": "NON_PARTICIPATING",
     "148": "SUCCESS",
-    "149": "SUCCESS",
-    "15": "SUCCESS",
-    "150": "SUCCESS",
+    "149": "NON_PARTICIPATING",
+    "15": "NON_PARTICIPATING",
+    "150": "NON_PARTICIPATING",
     "151": "NON_PARTICIPATING",
     "152": "NON_PARTICIPATING",
     "153": "NON_PARTICIPATING",
@@ -69,43 +69,43 @@
     "156": "NON_PARTICIPATING",
     "157": "NON_PARTICIPATING",
     "158": "NON_PARTICIPATING",
-    "159": "NON_PARTICIPATING",
+    "159": "SUCCESS",
     "16": "SUCCESS",
     "160": "NON_PARTICIPATING",
     "161": "NON_PARTICIPATING",
-    "162": "NON_PARTICIPATING",
+    "162": "SUCCESS",
     "163": "NON_PARTICIPATING",
-    "164": "NON_PARTICIPATING",
-    "165": "NON_PARTICIPATING",
+    "164": "SUCCESS",
+    "165": "SUCCESS",
     "166": "NON_PARTICIPATING",
     "167": "NON_PARTICIPATING",
     "168": "NON_PARTICIPATING",
-    "169": "NON_PARTICIPATING",
+    "169": "SUCCESS",
     "17": "SUCCESS",
     "170": "NON_PARTICIPATING",
     "171": "NON_PARTICIPATING",
-    "172": "SUCCESS",
-    "173": "SUCCESS",
+    "172": "NON_PARTICIPATING",
+    "173": "NON_PARTICIPATING",
     "174": "NON_PARTICIPATING",
-    "175": "NON_PARTICIPATING",
+    "175": "SUCCESS",
     "176": "NON_PARTICIPATING",
     "177": "NON_PARTICIPATING",
     "178": "NON_PARTICIPATING",
     "179": "NON_PARTICIPATING",
     "18": "SUCCESS",
     "180": "NON_PARTICIPATING",
-    "181": "NON_PARTICIPATING",
+    "181": "SUCCESS",
     "182": "NON_PARTICIPATING",
     "183": "NON_PARTICIPATING",
     "184": "NON_PARTICIPATING",
-    "185": "NON_PARTICIPATING",
-    "186": "SUCCESS",
+    "185": "SUCCESS",
+    "186": "NON_PARTICIPATING",
     "187": "NON_PARTICIPATING",
     "188": "NON_PARTICIPATING",
-    "189": "SUCCESS",
-    "19": "NON_PARTICIPATING",
+    "189": "NON_PARTICIPATING",
+    "19": "SUCCESS",
     "190": "NON_PARTICIPATING",
-    "191": "NON_PARTICIPATING",
+    "191": "SUCCESS",
     "192": "NON_PARTICIPATING",
     "193": "NON_PARTICIPATING",
     "194": "NON_PARTICIPATING",
@@ -124,7 +124,7 @@
     "205": "NON_PARTICIPATING",
     "206": "NON_PARTICIPATING",
     "207": "NON_PARTICIPATING",
-    "208": "NON_PARTICIPATING",
+    "208": "SUCCESS",
     "209": "NON_PARTICIPATING",
     "21": "SUCCESS",
     "210": "NON_PARTICIPATING",
@@ -132,7 +132,7 @@
     "212": "NON_PARTICIPATING",
     "213": "NON_PARTICIPATING",
     "214": "NON_PARTICIPATING",
-    "215": "SUCCESS",
+    "215": "NON_PARTICIPATING",
     "216": "NON_PARTICIPATING",
     "217": "NON_PARTICIPATING",
     "218": "NON_PARTICIPATING",
@@ -143,13 +143,13 @@
     "222": "NON_PARTICIPATING",
     "223": "NON_PARTICIPATING",
     "224": "NON_PARTICIPATING",
-    "225": "SUCCESS",
+    "225": "NON_PARTICIPATING",
     "226": "NON_PARTICIPATING",
     "227": "NON_PARTICIPATING",
     "228": "NON_PARTICIPATING",
-    "229": "NON_PARTICIPATING",
+    "229": "SUCCESS",
     "23": "SUCCESS",
-    "230": "NON_PARTICIPATING",
+    "230": "SUCCESS",
     "231": "NON_PARTICIPATING",
     "232": "SUCCESS",
     "233": "NON_PARTICIPATING",
@@ -157,15 +157,15 @@
     "235": "NON_PARTICIPATING",
     "236": "NON_PARTICIPATING",
     "237": "NON_PARTICIPATING",
-    "238": "NON_PARTICIPATING",
+    "238": "SUCCESS",
     "239": "SUCCESS",
     "24": "SUCCESS",
     "240": "NON_PARTICIPATING",
     "241": "NON_PARTICIPATING",
-    "242": "SUCCESS",
+    "242": "NON_PARTICIPATING",
     "243": "NON_PARTICIPATING",
     "244": "NON_PARTICIPATING",
-    "245": "NON_PARTICIPATING",
+    "245": "SUCCESS",
     "246": "NON_PARTICIPATING",
     "247": "NON_PARTICIPATING",
     "248": "NON_PARTICIPATING",
@@ -176,40 +176,40 @@
     "252": "NON_PARTICIPATING",
     "253": "NON_PARTICIPATING",
     "254": "NON_PARTICIPATING",
-    "255": "SUCCESS",
+    "255": "NON_PARTICIPATING",
     "26": "SUCCESS",
     "27": "SUCCESS",
     "28": "SUCCESS",
     "29": "SUCCESS",
     "3": "SUCCESS",
-    "30": "SUCCESS",
-    "31": "SUCCESS",
+    "30": "NON_PARTICIPATING",
+    "31": "NON_PARTICIPATING",
     "32": "SUCCESS",
     "33": "NON_PARTICIPATING",
-    "34": "NON_PARTICIPATING",
-    "35": "SUCCESS",
-    "36": "NON_PARTICIPATING",
+    "34": "SUCCESS",
+    "35": "NON_PARTICIPATING",
+    "36": "SUCCESS",
     "37": "SUCCESS",
     "38": "SUCCESS",
     "39": "SUCCESS",
     "4": "SUCCESS",
     "40": "SUCCESS",
     "41": "SUCCESS",
-    "42": "SUCCESS",
+    "42": "NON_PARTICIPATING",
     "43": "SUCCESS",
     "44": "SUCCESS",
     "45": "SUCCESS",
     "46": "NON_PARTICIPATING",
-    "47": "NON_PARTICIPATING",
+    "47": "SUCCESS",
     "48": "NON_PARTICIPATING",
     "49": "SUCCESS",
-    "5": "NON_PARTICIPATING",
+    "5": "SUCCESS",
     "50": "SUCCESS",
     "51": "SUCCESS",
     "52": "SUCCESS",
-    "53": "NON_PARTICIPATING",
+    "53": "SUCCESS",
     "54": "SUCCESS",
-    "55": "NON_PARTICIPATING",
+    "55": "SUCCESS",
     "56": "SUCCESS",
     "57": "SUCCESS",
     "58": "NON_PARTICIPATING",
@@ -223,9 +223,9 @@
     "65": "SUCCESS",
     "66": "SUCCESS",
     "67": "SUCCESS",
-    "68": "NON_PARTICIPATING",
-    "69": "NON_PARTICIPATING",
-    "7": "SUCCESS",
+    "68": "SUCCESS",
+    "69": "SUCCESS",
+    "7": "NON_PARTICIPATING",
     "70": "NON_PARTICIPATING",
     "71": "NON_PARTICIPATING",
     "72": "SUCCESS",
@@ -242,13 +242,13 @@
     "82": "SUCCESS",
     "83": "NON_PARTICIPATING",
     "84": "SUCCESS",
-    "85": "SUCCESS",
-    "86": "SUCCESS",
+    "85": "NON_PARTICIPATING",
+    "86": "NON_PARTICIPATING",
     "87": "SUCCESS",
     "88": "SUCCESS",
     "89": "SUCCESS",
     "9": "NON_PARTICIPATING",
-    "90": "SUCCESS",
+    "90": "NON_PARTICIPATING",
     "91": "SUCCESS",
     "92": "SUCCESS",
     "93": "NON_PARTICIPATING",
@@ -268,20 +268,20 @@
     "AutoModelForCausalLM": "distributed/optimized-gpt2-500m--modeling_gpt_optimized.GPTOptim"
   },
   "block_list": [
-    5576127,
-    5576131,
-    5576134,
-    5576138,
-    5576142
+    5576211,
+    5576268,
+    5576272,
+    5576276,
+    5576279
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 45,
+  "inner_step": 0,
   "inner_steps": 0,
-  "last_allreduce_block": 5546853,
+  "last_allreduce_block": 5576204,
   "layer_norm_epsilon": 1e-05,
   "model_type": "gpt_optimized",
   "n_embd": 1280,

inner_optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c2a2afaedb40e97e19f0b64e19483f16a597842c4a60fb134b5f1d2751c4adaf
+oid sha256:a2aa65e675ec439925db3631bae1c7cf62fa09a82b0632ef8da4c90d7b80021a
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:71347963f9fd4a13767c8e7e473c05463327a06316a93e33147f800baa9f4a0e
+oid sha256:9ba35fe72ed12b54c4e91ff5e916450689614454f76f3fc50436b55c099335a8
 size 4040701744