Run 4. Outer Step 53. Inner Step 37.

config.json

@@ -1,9 +1,9 @@
 {
-  "_name_or_path": "distributed/optimized-gpt2-1b",
+  "_name_or_path": "jockibaba/jockibaba4",
   "activation_function": "gelu_new",
   "all_reduce_scores": {
     "0": "NON_PARTICIPATING",
-    "1": "FAIL",
+    "1": "SUCCESS",
     "10": "NON_PARTICIPATING",
     "100": "NON_PARTICIPATING",
     "101": "NON_PARTICIPATING",
@@ -12,14 +12,14 @@
     "104": "NON_PARTICIPATING",
     "105": "NON_PARTICIPATING",
     "106": "NON_PARTICIPATING",
-    "107": "NON_PARTICIPATING",
+    "107": "SUCCESS",
     "108": "NON_PARTICIPATING",
-    "109": "NON_PARTICIPATING",
+    "109": "SUCCESS",
     "11": "SUCCESS",
-    "110": "SUCCESS",
-    "111": "SUCCESS",
+    "110": "NON_PARTICIPATING",
+    "111": "NON_PARTICIPATING",
     "112": "NON_PARTICIPATING",
-    "113": "NON_PARTICIPATING",
+    "113": "SUCCESS",
     "114": "NON_PARTICIPATING",
     "115": "NON_PARTICIPATING",
     "116": "NON_PARTICIPATING",
@@ -30,21 +30,21 @@
     "120": "NON_PARTICIPATING",
     "121": "NON_PARTICIPATING",
     "122": "NON_PARTICIPATING",
-    "123": "SUCCESS",
+    "123": "NON_PARTICIPATING",
     "124": "NON_PARTICIPATING",
     "125": "NON_PARTICIPATING",
-    "126": "NON_PARTICIPATING",
+    "126": "SUCCESS",
     "127": "NON_PARTICIPATING",
-    "128": "SUCCESS",
+    "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",
@@ -56,31 +56,31 @@
     "144": "NON_PARTICIPATING",
     "145": "NON_PARTICIPATING",
     "146": "NON_PARTICIPATING",
-    "147": "SUCCESS",
-    "148": "NON_PARTICIPATING",
+    "147": "NON_PARTICIPATING",
+    "148": "SUCCESS",
     "149": "NON_PARTICIPATING",
-    "15": "SUCCESS",
+    "15": "NON_PARTICIPATING",
     "150": "NON_PARTICIPATING",
     "151": "NON_PARTICIPATING",
     "152": "NON_PARTICIPATING",
     "153": "NON_PARTICIPATING",
-    "154": "SUCCESS",
+    "154": "NON_PARTICIPATING",
     "155": "NON_PARTICIPATING",
     "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": "SUCCESS",
-    "165": "NON_PARTICIPATING",
+    "165": "SUCCESS",
     "166": "NON_PARTICIPATING",
     "167": "NON_PARTICIPATING",
     "168": "NON_PARTICIPATING",
-    "169": "NON_PARTICIPATING",
+    "169": "SUCCESS",
     "17": "SUCCESS",
     "170": "NON_PARTICIPATING",
     "171": "NON_PARTICIPATING",
@@ -92,20 +92,20 @@
     "177": "NON_PARTICIPATING",
     "178": "NON_PARTICIPATING",
     "179": "NON_PARTICIPATING",
-    "18": "NON_PARTICIPATING",
+    "18": "SUCCESS",
     "180": "NON_PARTICIPATING",
-    "181": "NON_PARTICIPATING",
+    "181": "SUCCESS",
     "182": "NON_PARTICIPATING",
     "183": "NON_PARTICIPATING",
     "184": "NON_PARTICIPATING",
-    "185": "NON_PARTICIPATING",
+    "185": "SUCCESS",
     "186": "NON_PARTICIPATING",
     "187": "NON_PARTICIPATING",
     "188": "NON_PARTICIPATING",
     "189": "NON_PARTICIPATING",
     "19": "SUCCESS",
     "190": "NON_PARTICIPATING",
-    "191": "NON_PARTICIPATING",
+    "191": "SUCCESS",
     "192": "NON_PARTICIPATING",
     "193": "NON_PARTICIPATING",
     "194": "NON_PARTICIPATING",
@@ -115,35 +115,35 @@
     "198": "NON_PARTICIPATING",
     "199": "NON_PARTICIPATING",
     "2": "SUCCESS",
-    "20": "FAIL",
+    "20": "SUCCESS",
     "200": "NON_PARTICIPATING",
     "201": "NON_PARTICIPATING",
-    "202": "SUCCESS",
+    "202": "NON_PARTICIPATING",
     "203": "NON_PARTICIPATING",
     "204": "NON_PARTICIPATING",
     "205": "NON_PARTICIPATING",
     "206": "NON_PARTICIPATING",
     "207": "NON_PARTICIPATING",
-    "208": "NON_PARTICIPATING",
+    "208": "SUCCESS",
     "209": "NON_PARTICIPATING",
     "21": "SUCCESS",
     "210": "NON_PARTICIPATING",
     "211": "NON_PARTICIPATING",
     "212": "NON_PARTICIPATING",
     "213": "NON_PARTICIPATING",
-    "214": "SUCCESS",
+    "214": "NON_PARTICIPATING",
     "215": "NON_PARTICIPATING",
     "216": "NON_PARTICIPATING",
     "217": "NON_PARTICIPATING",
     "218": "NON_PARTICIPATING",
     "219": "NON_PARTICIPATING",
     "22": "SUCCESS",
-    "220": "SUCCESS",
+    "220": "NON_PARTICIPATING",
     "221": "NON_PARTICIPATING",
     "222": "NON_PARTICIPATING",
     "223": "NON_PARTICIPATING",
     "224": "NON_PARTICIPATING",
-    "225": "SUCCESS",
+    "225": "NON_PARTICIPATING",
     "226": "NON_PARTICIPATING",
     "227": "NON_PARTICIPATING",
     "228": "NON_PARTICIPATING",
@@ -151,25 +151,25 @@
     "23": "SUCCESS",
     "230": "SUCCESS",
     "231": "NON_PARTICIPATING",
-    "232": "NON_PARTICIPATING",
+    "232": "SUCCESS",
     "233": "NON_PARTICIPATING",
     "234": "NON_PARTICIPATING",
     "235": "NON_PARTICIPATING",
     "236": "NON_PARTICIPATING",
     "237": "NON_PARTICIPATING",
-    "238": "NON_PARTICIPATING",
-    "239": "NON_PARTICIPATING",
+    "238": "SUCCESS",
+    "239": "SUCCESS",
     "24": "SUCCESS",
     "240": "NON_PARTICIPATING",
     "241": "NON_PARTICIPATING",
     "242": "NON_PARTICIPATING",
     "243": "NON_PARTICIPATING",
     "244": "NON_PARTICIPATING",
-    "245": "NON_PARTICIPATING",
+    "245": "SUCCESS",
     "246": "NON_PARTICIPATING",
     "247": "NON_PARTICIPATING",
     "248": "NON_PARTICIPATING",
-    "249": "SUCCESS",
+    "249": "NON_PARTICIPATING",
     "25": "SUCCESS",
     "250": "SUCCESS",
     "251": "NON_PARTICIPATING",
@@ -182,33 +182,33 @@
     "28": "SUCCESS",
     "29": "SUCCESS",
     "3": "SUCCESS",
-    "30": "SUCCESS",
-    "31": "SUCCESS",
+    "30": "NON_PARTICIPATING",
+    "31": "NON_PARTICIPATING",
     "32": "SUCCESS",
     "33": "NON_PARTICIPATING",
     "34": "SUCCESS",
-    "35": "SUCCESS",
+    "35": "NON_PARTICIPATING",
     "36": "SUCCESS",
     "37": "SUCCESS",
     "38": "SUCCESS",
     "39": "SUCCESS",
-    "4": "FAIL",
+    "4": "SUCCESS",
     "40": "SUCCESS",
     "41": "SUCCESS",
     "42": "NON_PARTICIPATING",
     "43": "SUCCESS",
-    "44": "FAIL",
+    "44": "SUCCESS",
     "45": "SUCCESS",
     "46": "NON_PARTICIPATING",
-    "47": "NON_PARTICIPATING",
+    "47": "SUCCESS",
     "48": "NON_PARTICIPATING",
     "49": "SUCCESS",
-    "5": "FAIL",
+    "5": "SUCCESS",
     "50": "SUCCESS",
     "51": "SUCCESS",
     "52": "SUCCESS",
     "53": "SUCCESS",
-    "54": "FAIL",
+    "54": "SUCCESS",
     "55": "SUCCESS",
     "56": "SUCCESS",
     "57": "SUCCESS",
@@ -220,18 +220,18 @@
     "62": "SUCCESS",
     "63": "SUCCESS",
     "64": "NON_PARTICIPATING",
-    "65": "FAIL",
+    "65": "SUCCESS",
     "66": "SUCCESS",
     "67": "SUCCESS",
     "68": "SUCCESS",
     "69": "SUCCESS",
-    "7": "SUCCESS",
+    "7": "NON_PARTICIPATING",
     "70": "NON_PARTICIPATING",
     "71": "NON_PARTICIPATING",
     "72": "SUCCESS",
     "73": "NON_PARTICIPATING",
-    "74": "FAIL",
-    "75": "FAIL",
+    "74": "SUCCESS",
+    "75": "SUCCESS",
     "76": "SUCCESS",
     "77": "SUCCESS",
     "78": "SUCCESS",
@@ -247,8 +247,8 @@
     "87": "SUCCESS",
     "88": "SUCCESS",
     "89": "SUCCESS",
-    "9": "SUCCESS",
-    "90": "SUCCESS",
+    "9": "NON_PARTICIPATING",
+    "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": [
-    5586982,
-    5586986,
-    5586990,
-    5586994,
-    5586998
+    5587040,
+    5587072,
+    5587076,
+    5587080,
+    5587085
   ],
   "block_size": 1024,
   "bos_token_id": 50256,
   "embd_pdrop": 0.1,
   "eos_token_id": 50256,
   "initializer_range": 0.02,
-  "inner_step": 25,
+  "inner_step": 37,
   "inner_steps": 0,
-  "last_allreduce_block": 5586129,
+  "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:f6cb7d70b7c8f757d0889df53fbef5835db93054070f06ffa338a457c1754ea7
+oid sha256:1f808fd5f3f4abaf456a2b48a553e263d4e4f553914779f5d540cae1253a3e88
 size 8081782026

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e382a297fe6e9c50e96a5fd88455a88c1f03b2e0f2079a3bc60fb06dcf79a5c5
+oid sha256:60e55addb0593c19bf457ee433300770fae57d69e18e707e4e688ad915872b45
 size 4040701744