Qwen90M step 765000; complete resumable checkpoint

complete.json

@@ -1,19 +1,19 @@
 {
-  "step": 760000,
+  "step": 765000,
   "files": {
     "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
     "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
-    "model.safetensors": "a860e83f899b4e90ca04decae15751893c788a0a68e4bf4ef98f0325ea54d165",
+    "model.safetensors": "6b4ad24005ad10d830f5b4dd0b826962f001a14e87dcbc136addc786b07770c8",
     "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
     "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
     "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
     "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
-    "trainer.pt": "1b1f259fdc0e4e7c034af4b926b3e23d2a45e4a91f9e75f8c00269d04f243d83",
-    "training_config.json": "75248af0a321738b405c3d51d1d55c7040dbc0785759d95f8dbdf6cd071a6687",
-    "performance.json": "48e3aa7938b07d5e08810d60bc208cc2cb9133c4c39a24151758b79306675379",
-    "shard_scheduler.json": "8546a039abf7332253379c1f06f6d9b77cec29d6925ca8a503635e01f245c443",
-    "coverage.json": "e5698dd1fdd8d06e82449a48c972062af4c83576c8da391fd360e5358f8f05e2",
+    "trainer.pt": "07dc267a3de4e00552f0a06fb17ec8e0dcd6e1e709ac7d235e0e71960df8e5ea",
+    "training_config.json": "4e8f8b2fc7193aabdbca12082e7c7cc3a0ed080bd5ff16e060cc02e10ffc53dd",
+    "performance.json": "d91455eeac02a2960a60e3a2edc7d99796576b62edb422642a8beb3b824ff02d",
+    "shard_scheduler.json": "4c919b423fdadb08dc04e9ee2624261806d3f9f3f73d63e1a0959bb729e52ff3",
+    "coverage.json": "b421dc96b600cc122bd4c40c0c3b6e29b4666eec433d9dfb0299a93825a0d58c",
     "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
   },
-  "signature": "13510bfc04d8fed598fd0d09c1f778fbaebf19cefdd81f23f5ef3b909ef94a63"
+  "signature": "052aedaef69b550457da29474e17df28fa635c99594a7b9c0881f3ac3827000e"
 }

coverage.json

@@ -1,107 +1,107 @@
 {
-  "step": 760000,
+  "step": 765000,
   "stats": {
-    "climbmix/blocks": 27360000,
-    "climbmix/targets": 45737235053,
-    "climbmix/input_tokens": 45740133081,
-    "ultra_qa/blocks": 10260000,
-    "ultra_qa/targets": 16737918320,
-    "ultra_qa/input_tokens": 16737950563,
-    "cortex/blocks": 10260000,
-    "cortex/targets": 390103024,
-    "cortex/input_tokens": 16308780977,
-    "ultra_style/blocks": 10260000,
-    "ultra_style/targets": 16228024880,
-    "ultra_style/input_tokens": 16228875947,
-    "openmath/blocks": 6840000,
-    "openmath/targets": 10568964438,
-    "openmath/input_tokens": 10572485614,
-    "rewrite6/blocks": 3420000,
-    "rewrite6/targets": 5606566408,
-    "rewrite6/input_tokens": 5612713435
+    "climbmix/blocks": 27540000,
+    "climbmix/targets": 46037548584,
+    "climbmix/input_tokens": 46040465939,
+    "ultra_qa/blocks": 10327500,
+    "ultra_qa/targets": 16848102853,
+    "ultra_qa/input_tokens": 16848135318,
+    "cortex/blocks": 10327500,
+    "cortex/targets": 392690448,
+    "cortex/input_tokens": 16416010630,
+    "ultra_style/blocks": 10327500,
+    "ultra_style/targets": 16334987427,
+    "ultra_style/input_tokens": 16335844116,
+    "openmath/blocks": 6885000,
+    "openmath/targets": 10638497464,
+    "openmath/input_tokens": 10642041928,
+    "rewrite6/blocks": 3442500,
+    "rewrite6/targets": 5643572276,
+    "rewrite6/input_tokens": 5649759720
   },
   "sources": {
     "climbmix": {
-      "page": 3746,
-      "row": 3857,
+      "page": 3771,
+      "row": 3176,
       "start": {
-        "file": 181,
+        "file": 182,
         "reader": {
           "format": "parquet-row-v1",
-          "row_group": 28,
+          "row_group": 44,
           "row": 1024
         }
       },
       "epoch": 0,
-      "offset": 2048,
-      "documents": 15347473
+      "offset": 14336,
+      "documents": 15449192
     },
     "rewrite6": {
-      "page": 31,
-      "row": 3430,
+      "page": 45,
+      "row": 3598,
       "start": {
         "file": 0,
         "reader": {
           "format": "jsonl-row-v1",
-          "row": 126976
+          "row": 184320
         }
       },
       "epoch": 42,
       "offset": 0,
-      "documents": 8757374
+      "documents": 8814886
     },
     "openmath": {
-      "page": 2341,
-      "row": 3059,
+      "page": 2357,
+      "row": 768,
       "start": {
-        "file": 21,
+        "file": 22,
         "reader": {
           "format": "parquet-row-v1",
-          "row_group": 419,
-          "row": 86
+          "row_group": 47,
+          "row": 972
         }
       },
       "epoch": 0,
       "offset": 0,
-      "documents": 9591795
+      "documents": 9655040
     },
     "ultra_qa": {
-      "page": 1012,
-      "row": 2311,
+      "page": 1019,
+      "row": 952,
       "start": {
-        "file": 7,
+        "file": 8,
         "reader": {
           "format": "parquet-row-v1",
-          "row_group": 7,
-          "row": 55616
+          "row_group": 0,
+          "row": 16813
         }
       },
       "epoch": 0,
       "offset": 2048,
-      "documents": 4147463
+      "documents": 4174776
     },
     "ultra_style": {
-      "page": 1545,
-      "row": 1981,
+      "page": 1555,
+      "row": 2608,
       "start": {
         "file": 9,
         "reader": {
           "format": "parquet-row-v1",
-          "row_group": 1,
-          "row": 78482
+          "row_group": 2,
+          "row": 34550
         }
       },
       "epoch": 0,
       "offset": 2048,
-      "documents": 6330301
+      "documents": 6371888
     },
     "cortex": {
-      "page": 198979,
-      "row": 69,
+      "page": 200288,
+      "row": 88,
       "start": null,
       "epoch": 0,
       "offset": 0,
-      "documents": 25469381
+      "documents": 25636952
     }
   }
 }

latest.json

@@ -1 +1 @@
-{"step": 760000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "a860e83f899b4e90ca04decae15751893c788a0a68e4bf4ef98f0325ea54d165", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "1b1f259fdc0e4e7c034af4b926b3e23d2a45e4a91f9e75f8c00269d04f243d83", "training_config.json": "75248af0a321738b405c3d51d1d55c7040dbc0785759d95f8dbdf6cd071a6687", "performance.json": "48e3aa7938b07d5e08810d60bc208cc2cb9133c4c39a24151758b79306675379", "shard_scheduler.json": "8546a039abf7332253379c1f06f6d9b77cec29d6925ca8a503635e01f245c443", "coverage.json": "e5698dd1fdd8d06e82449a48c972062af4c83576c8da391fd360e5358f8f05e2", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "13510bfc04d8fed598fd0d09c1f778fbaebf19cefdd81f23f5ef3b909ef94a63", "folder": "step-000000760000"}
+{"step": 765000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "6b4ad24005ad10d830f5b4dd0b826962f001a14e87dcbc136addc786b07770c8", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "07dc267a3de4e00552f0a06fb17ec8e0dcd6e1e709ac7d235e0e71960df8e5ea", "training_config.json": "4e8f8b2fc7193aabdbca12082e7c7cc3a0ed080bd5ff16e060cc02e10ffc53dd", "performance.json": "d91455eeac02a2960a60e3a2edc7d99796576b62edb422642a8beb3b824ff02d", "shard_scheduler.json": "4c919b423fdadb08dc04e9ee2624261806d3f9f3f73d63e1a0959bb729e52ff3", "coverage.json": "b421dc96b600cc122bd4c40c0c3b6e29b4666eec433d9dfb0299a93825a0d58c", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "052aedaef69b550457da29474e17df28fa635c99594a7b9c0881f3ac3827000e", "folder": "step-000000765000"}

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a860e83f899b4e90ca04decae15751893c788a0a68e4bf4ef98f0325ea54d165
+oid sha256:6b4ad24005ad10d830f5b4dd0b826962f001a14e87dcbc136addc786b07770c8
 size 360622744

performance.json

@@ -49,11 +49,11 @@
   "selected": {
     "microbatch": 90,
     "mode": "compiled",
-    "microbatch_seconds": 0.9371568949999869,
+    "microbatch_seconds": 0.9481044159999783,
     "peak_gb": 89.37790203094482,
     "fits_budget": true,
-    "estimated_update_seconds": 0.9371568949999869,
-    "input_tokens_per_second": 196679.98067709096
+    "estimated_update_seconds": 0.9481044159999783,
+    "input_tokens_per_second": 194408.96687059014
   },
   "measurements": [
     {

shard_scheduler.json

@@ -22,8 +22,8 @@
       "tokens_per_second": 598315.063954628,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 76663.29501248694,
-      "required_pages_per_second_upper_estimate": 0.005303744632315819
+      "required_token_positions_per_second": 77160.6514700967,
+      "required_pages_per_second_upper_estimate": 0.0053381529060792365
     },
     {
       "source": "rewrite6",
@@ -36,8 +36,8 @@
       "tokens_per_second": 8533.957213384612,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 9582.911876560867,
-      "required_pages_per_second_upper_estimate": 0.004012165088760336
+      "required_token_positions_per_second": 9645.081433762087,
+      "required_pages_per_second_upper_estimate": 0.004038194184112503
     },
     {
       "source": "openmath",
@@ -50,8 +50,8 @@
       "tokens_per_second": 380144.35761751566,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 19165.823753121735,
-      "required_pages_per_second_upper_estimate": 0.0039917356510137374
+      "required_token_positions_per_second": 19290.162867524174,
+      "required_pages_per_second_upper_estimate": 0.004017632209500813
     },
     {
       "source": "ultra_qa",
@@ -64,8 +64,8 @@
       "tokens_per_second": 43742.16556775435,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 28748.735629682604,
-      "required_pages_per_second_upper_estimate": 0.0017234428042218715
+      "required_token_positions_per_second": 28935.244301286257,
+      "required_pages_per_second_upper_estimate": 0.0017346237143022584
     },
     {
       "source": "ultra_style",
@@ -78,8 +78,8 @@
       "tokens_per_second": 22018.86276523721,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 28748.735629682604,
-      "required_pages_per_second_upper_estimate": 0.002235087742883681
+      "required_token_positions_per_second": 28935.244301286257,
+      "required_pages_per_second_upper_estimate": 0.0022495879717359174
     },
     {
       "source": "cortex",
@@ -92,8 +92,8 @@
       "tokens_per_second": 910315.184299418,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 23,
-      "required_token_positions_per_second": 28748.735629682604,
-      "required_pages_per_second_upper_estimate": 0.3687579126702146
+      "required_token_positions_per_second": 28935.244301286257,
+      "required_pages_per_second_upper_estimate": 0.37115024565213706
     }
   ],
   "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
@@ -158,409 +158,91 @@
     "prefetch_batches": 63,
     "prefetch_capacity_limit": 512,
     "batch_memory_estimate_bytes": 720750,
-    "producer_batch_seconds": 0.09382273250031403,
-    "producer_p95_seconds": 0.3345595279970439,
-    "gpu_update_seconds": 0.9617118594757916,
-    "producer_batches_per_second": 10.65839773955265,
-    "gpu_batches_per_second": 1.039812486605997,
-    "producer_headroom": 10.250307509137754,
+    "producer_batch_seconds": 0.22211623922648727,
+    "producer_p95_seconds": 0.16718530399975862,
+    "gpu_update_seconds": 0.9555129278369688,
+    "producer_batches_per_second": 4.5021471797040515,
+    "gpu_batches_per_second": 1.0465583152953652,
+    "producer_headroom": 4.30185983323197,
     "sustainable": true,
-    "buffer_seconds": 60.58784714697487,
+    "buffer_seconds": 60.197314453729035,
     "buffer_drain_seconds": null,
     "shard_target_mib": 16,
-    "upload_pages_per_commit": 83,
+    "upload_pages_per_commit": 79,
     "max_pending_pages": 4096,
-    "generation_bytes_per_second": 256556.76260388948,
-    "upload_bytes_per_second": 4861160.122505256,
+    "generation_bytes_per_second": 471718.0651086896,
+    "upload_bytes_per_second": 5138876.858680128,
     "upload_sustainable": true,
     "logical_page_rows": 4096,
     "logical_cortex_page_rows": 128,
-    "estimated_full_shards_per_hour": 55.05110891902459,
+    "estimated_full_shards_per_hour": 101.21971573777691,
     "selection_policy": "Smallest shard within 90% of measured local peak, enlarged to amortize 30 seconds of measured page production; bounded to 256 MiB."
   },
   "upload_measurements": [
     {
-      "seconds": 0.9614488920014992,
-      "bytes": 6537093,
-      "pages": 33,
-      "shards": 3
-    },
-    {
-      "seconds": 1.2357967999996617,
-      "bytes": 11997860,
-      "pages": 35,
-      "shards": 3
-    },
-    {
-      "seconds": 1.7381740189994161,
-      "bytes": 7938212,
-      "pages": 40,
-      "shards": 3
-    },
-    {
-      "seconds": 1.2239695549978933,
-      "bytes": 5214335,
-      "pages": 32,
-      "shards": 2
-    },
-    {
-      "seconds": 1.2570640209996782,
-      "bytes": 6323177,
-      "pages": 35,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1513970990017697,
-      "bytes": 2421895,
-      "pages": 31,
-      "shards": 2
-    },
-    {
-      "seconds": 0.9733352420007577,
-      "bytes": 6126927,
-      "pages": 37,
-      "shards": 2
-    },
-    {
-      "seconds": 1.3508422129998507,
-      "bytes": 16469825,
-      "pages": 33,
-      "shards": 4
-    },
-    {
-      "seconds": 1.220427556996583,
-      "bytes": 2558602,
-      "pages": 34,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1153945629994269,
-      "bytes": 6138686,
-      "pages": 33,
-      "shards": 2
-    },
-    {
-      "seconds": 1.4000027559959562,
-      "bytes": 7623395,
-      "pages": 39,
-      "shards": 3
-    },
-    {
-      "seconds": 1.0437422140021226,
-      "bytes": 887331,
-      "pages": 27,
+      "seconds": 0.9713157099999989,
+      "bytes": 23380,
+      "pages": 1,
       "shards": 1
     },
     {
-      "seconds": 1.3929757290025009,
-      "bytes": 11170736,
-      "pages": 35,
-      "shards": 3
-    },
-    {
-      "seconds": 1.5524120170011884,
-      "bytes": 8373554,
-      "pages": 38,
-      "shards": 3
-    },
-    {
-      "seconds": 1.806914691995189,
-      "bytes": 5833325,
-      "pages": 30,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1171102930020425,
-      "bytes": 1254155,
+      "seconds": 1.497052147999966,
+      "bytes": 11817101,
       "pages": 34,
-      "shards": 1
-    },
-    {
-      "seconds": 1.3684268549986882,
-      "bytes": 11656046,
-      "pages": 37,
       "shards": 4
     },
     {
-      "seconds": 1.130367224002839,
-      "bytes": 6468621,
-      "pages": 37,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1125211710022995,
-      "bytes": 2380554,
-      "pages": 28,
-      "shards": 2
-    },
-    {
-      "seconds": 1.3700068609978189,
-      "bytes": 12013278,
-      "pages": 37,
-      "shards": 3
-    },
-    {
-      "seconds": 1.232138457999099,
-      "bytes": 9991874,
+      "seconds": 1.0985135669998272,
+      "bytes": 6249223,
       "pages": 36,
-      "shards": 3
-    },
-    {
-      "seconds": 1.27796818799834,
-      "bytes": 2573475,
-      "pages": 32,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1166499749961076,
-      "bytes": 6085922,
-      "pages": 33,
       "shards": 2
     },
     {
-      "seconds": 0.8856082540005445,
-      "bytes": 2405776,
-      "pages": 29,
-      "shards": 2
-    },
-    {
-      "seconds": 1.0989994560004561,
-      "bytes": 10674348,
+      "seconds": 1.3139188940003805,
+      "bytes": 2690437,
       "pages": 38,
       "shards": 3
     },
     {
-      "seconds": 1.3233692020003218,
-      "bytes": 11880619,
-      "pages": 39,
-      "shards": 3
-    },
-    {
-      "seconds": 0.9542889080039458,
-      "bytes": 2355430,
-      "pages": 29,
-      "shards": 2
-    },
-    {
-      "seconds": 1.3224051930010319,
-      "bytes": 6398826,
-      "pages": 37,
-      "shards": 2
-    },
-    {
-      "seconds": 1.916404732997762,
-      "bytes": 5130348,
-      "pages": 29,
-      "shards": 2
-    },
-    {
-      "seconds": 1.503245963001973,
-      "bytes": 7636802,
-      "pages": 37,
-      "shards": 3
-    },
-    {
-      "seconds": 1.082158730001538,
-      "bytes": 6031838,
-      "pages": 38,
-      "shards": 2
-    },
-    {
-      "seconds": 1.433015736998641,
-      "bytes": 8073393,
-      "pages": 27,
-      "shards": 3
-    },
-    {
-      "seconds": 1.7437503740002285,
-      "bytes": 10517926,
-      "pages": 37,
-      "shards": 3
-    },
-    {
-      "seconds": 1.3904311470032553,
-      "bytes": 899587,
-      "pages": 31,
-      "shards": 1
-    },
-    {
-      "seconds": 1.836203224993369,
-      "bytes": 7625671,
-      "pages": 39,
-      "shards": 4
-    },
-    {
-      "seconds": 1.115198325001984,
-      "bytes": 5971301,
-      "pages": 29,
-      "shards": 2
-    },
-    {
-      "seconds": 1.9192805860002409,
-      "bytes": 6532250,
-      "pages": 41,
-      "shards": 3
-    },
-    {
-      "seconds": 1.5230046290016617,
-      "bytes": 11668778,
+      "seconds": 1.4740382569998474,
+      "bytes": 11539917,
       "pages": 32,
       "shards": 3
     },
     {
-      "seconds": 2.0409051120004733,
-      "bytes": 6281809,
-      "pages": 36,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1541477030041278,
-      "bytes": 6906912,
+      "seconds": 1.530777683999986,
+      "bytes": 10014604,
       "pages": 34,
       "shards": 3
     },
     {
-      "seconds": 1.7220873259939253,
-      "bytes": 6147366,
-      "pages": 28,
-      "shards": 2
-    },
-    {
-      "seconds": 3.3549275529949227,
-      "bytes": 2682623,
-      "pages": 38,
-      "shards": 2
-    },
-    {
-      "seconds": 1.4113562120037386,
-      "bytes": 6081250,
+      "seconds": 1.4347790020001412,
+      "bytes": 2652487,
       "pages": 37,
       "shards": 2
     },
     {
-      "seconds": 1.8620396420010366,
-      "bytes": 10625025,
-      "pages": 23,
-      "shards": 3
-    },
-    {
-      "seconds": 1.826674156000081,
-      "bytes": 7716672,
-      "pages": 41,
-      "shards": 3
-    },
-    {
-      "seconds": 1.3658487110005808,
-      "bytes": 5996229,
-      "pages": 31,
-      "shards": 2
-    },
-    {
-      "seconds": 1.6789640829956625,
-      "bytes": 1167744,
-      "pages": 34,
-      "shards": 1
-    },
-    {
-      "seconds": 1.3386350959990523,
-      "bytes": 11890624,
-      "pages": 35,
-      "shards": 4
-    },
-    {
-      "seconds": 1.2723760490043787,
-      "bytes": 6238095,
+      "seconds": 1.0014089890000832,
+      "bytes": 6005067,
       "pages": 36,
       "shards": 2
     },
     {
-      "seconds": 1.1314651699940441,
-      "bytes": 8436407,
-      "pages": 37,
-      "shards": 3
-    },
-    {
-      "seconds": 1.675168465000752,
-      "bytes": 5970136,
+      "seconds": 0.9762548340004287,
+      "bytes": 2290276,
       "pages": 28,
       "shards": 2
     },
     {
-      "seconds": 1.2014365330032888,
-      "bytes": 5687823,
+      "seconds": 1.4867014489991561,
+      "bytes": 11031744,
       "pages": 35,
-      "shards": 2
-    },
-    {
-      "seconds": 1.1749995130012394,
-      "bytes": 7282459,
-      "pages": 38,
       "shards": 3
     },
     {
-      "seconds": 1.3900761400000192,
-      "bytes": 5687350,
-      "pages": 29,
-      "shards": 2
-    },
-    {
-      "seconds": 1.0886611489986535,
-      "bytes": 2750948,
-      "pages": 38,
-      "shards": 2
-    },
-    {
-      "seconds": 2.3056074749983964,
-      "bytes": 15312112,
-      "pages": 34,
-      "shards": 4
-    },
-    {
-      "seconds": 1.2607366750016809,
-      "bytes": 5997714,
-      "pages": 33,
-      "shards": 2
-    },
-    {
-      "seconds": 1.251247890999366,
-      "bytes": 2744476,
-      "pages": 38,
-      "shards": 2
-    },
-    {
-      "seconds": 1.2621945279970532,
-      "bytes": 5911926,
-      "pages": 29,
-      "shards": 2
-    },
-    {
-      "seconds": 1.4384337969968328,
-      "bytes": 10703677,
-      "pages": 35,
-      "shards": 3
-    },
-    {
-      "seconds": 1.1966409989981912,
-      "bytes": 2669509,
-      "pages": 38,
-      "shards": 2
-    },
-    {
-      "seconds": 1.7882141709997086,
-      "bytes": 11758267,
-      "pages": 28,
-      "shards": 3
-    },
-    {
-      "seconds": 1.0542783380005858,
-      "bytes": 2768181,
-      "pages": 35,
-      "shards": 2
-    },
-    {
-      "seconds": 1.5631635209938395,
-      "bytes": 10280055,
-      "pages": 36,
+      "seconds": 2.084590124000897,
+      "bytes": 12097526,
+      "pages": 37,
       "shards": 3
     }
   ],
@@ -606,7 +288,7 @@
   },
   "cache_reused": true,
   "resume_probe_seconds": [
-    3.2689711630000033,
-    0.017895548000012695
+    3.0803219919999947,
+    0.10207374499998423
   ]
 }

trainer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1b1f259fdc0e4e7c034af4b926b3e23d2a45e4a91f9e75f8c00269d04f243d83
-size 721774335
+oid sha256:07dc267a3de4e00552f0a06fb17ec8e0dcd6e1e709ac7d235e0e71960df8e5ea
+size 721777663

training_config.json

@@ -32,7 +32,7 @@
     "openmath": 10,
     "rewrite6": 5
   },
-  "lr": 5e-05,
+  "lr": 3e-05,
   "weight_decay": 0.1,
   "lr_schedule": "auto",
   "warmup_steps": 1000,