Update TensorBoard training metrics

benchmark_cache/shard_scheduler.json

@@ -22,8 +22,8 @@
       "tokens_per_second": 1062967.6168812835,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 75718.71545211473,
-      "required_pages_per_second_upper_estimate": 0.0065631063480594935
+      "required_token_positions_per_second": 75429.76421096348,
+      "required_pages_per_second_upper_estimate": 0.006538060786816728
     },
     {
       "source": "rewrite6",
@@ -36,8 +36,8 @@
       "tokens_per_second": 1206672.263377472,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 9464.839431514341,
-      "required_pages_per_second_upper_estimate": 0.0037756659611912964
+      "required_token_positions_per_second": 9428.720526370435,
+      "required_pages_per_second_upper_estimate": 0.003761257589903636
     },
     {
       "source": "openmath",
@@ -50,8 +50,8 @@
       "tokens_per_second": 410042.7581134458,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 18929.678863028683,
-      "required_pages_per_second_upper_estimate": 0.004379048053992215
+      "required_token_positions_per_second": 18857.44105274087,
+      "required_pages_per_second_upper_estimate": 0.004362337107924169
     },
     {
       "source": "ultra_qa",
@@ -64,8 +64,8 @@
       "tokens_per_second": 2163486.5624068812,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 28394.518294543024,
-      "required_pages_per_second_upper_estimate": 0.0017290771896135365
+      "required_token_positions_per_second": 28286.161579111304,
+      "required_pages_per_second_upper_estimate": 0.0017224788341474943
     },
     {
       "source": "ultra_style",
@@ -78,8 +78,8 @@
       "tokens_per_second": 1546694.2737633225,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 1,
-      "required_token_positions_per_second": 28394.518294543024,
-      "required_pages_per_second_upper_estimate": 0.0021858886193712526
+      "required_token_positions_per_second": 28286.161579111304,
+      "required_pages_per_second_upper_estimate": 0.002177547019466733
     },
     {
       "source": "cortex",
@@ -92,8 +92,8 @@
       "tokens_per_second": 1015177.8395898147,
       "note": "Bounded fresh sample; full-page latency is an extrapolation.",
       "estimated_pages_for_buffer": 22,
-      "required_token_positions_per_second": 28394.518294543024,
-      "required_pages_per_second_upper_estimate": 0.3525648868786151
+      "required_token_positions_per_second": 28286.161579111304,
+      "required_pages_per_second_upper_estimate": 0.3512194591195513
     }
   ],
   "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
@@ -158,24 +158,24 @@
     "prefetch_batches": 62,
     "prefetch_capacity_limit": 512,
     "batch_memory_estimate_bytes": 688946,
-    "producer_batch_seconds": 0.10471276342164515,
-    "producer_p95_seconds": 0.3463279750030779,
-    "gpu_update_seconds": 0.9737090699409225,
-    "producer_batches_per_second": 9.549934194490852,
-    "gpu_batches_per_second": 1.027000806370914,
-    "producer_headroom": 9.298857542514702,
+    "producer_batch_seconds": 0.1195543993672743,
+    "producer_p95_seconds": 0.35193734799759113,
+    "gpu_update_seconds": 0.9774390888163995,
+    "producer_batches_per_second": 8.36439315736072,
+    "gpu_batches_per_second": 1.0230816543370698,
+    "producer_headroom": 8.17568482623279,
     "sustainable": true,
-    "buffer_seconds": 60.369962336337196,
+    "buffer_seconds": 60.60122350661677,
     "buffer_drain_seconds": null,
     "shard_target_mib": 16,
-    "upload_pages_per_commit": 80,
+    "upload_pages_per_commit": 82,
     "max_pending_pages": 4096,
-    "generation_bytes_per_second": 254681.48879160965,
+    "generation_bytes_per_second": 270500.14458436874,
     "upload_bytes_per_second": 3732263.5224966407,
     "upload_sustainable": true,
     "logical_page_rows": 4096,
     "logical_cortex_page_rows": 128,
-    "estimated_full_shards_per_hour": 54.64871881304948,
+    "estimated_full_shards_per_hour": 58.04303410671517,
     "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": [

metrics/void-byte-v1/1790434090210070261-a05fc3fb/train.jsonl

@@ -305,3 +305,4 @@
 {"step": 70500, "loss": 0.6558354012668133, "lr": 0.0003, "grad_norm": 0.13466285050920596, "recovery_retries": 0, "loss_scale": 1.0, "microbatch": 86, "input_bytes_and_specials": 10317975409, "tokens_per_second": 149755.84670505646, "data_wait_s": 1.0271000064676628e-05, "gpu_peak_gb": 85.46427917480469}
 {"step": 70600, "loss": 0.5950672663748264, "lr": 0.0003, "grad_norm": 0.13279087819561, "recovery_retries": 0, "loss_scale": 1.0, "microbatch": 86, "input_bytes_and_specials": 10332681970, "tokens_per_second": 128305.54160269313, "data_wait_s": 1.6959998902166262e-05, "gpu_peak_gb": 85.46427917480469}
 {"step": 70700, "loss": 0.6360750216990709, "lr": 0.0003, "grad_norm": 0.13459834338447882, "recovery_retries": 0, "loss_scale": 1.0, "microbatch": 86, "input_bytes_and_specials": 10347387295, "tokens_per_second": 150647.55868196956, "data_wait_s": 1.1276999430265278e-05, "gpu_peak_gb": 85.46427917480469}
+{"step": 70800, "loss": 0.6045117974281311, "lr": 0.0003, "grad_norm": 0.12535527682268818, "recovery_retries": 0, "loss_scale": 1.0, "microbatch": 86, "input_bytes_and_specials": 10362013685, "tokens_per_second": 149354.95903517498, "data_wait_s": 9.227998816641048e-06, "gpu_peak_gb": 85.46427917480469}

runs/void-byte-v1/events.out.tfevents.1790434090.372bd512-57cc-4928-b44a-3b375a0fa071-5spg5.451.0.1790434090210070261-a05fc3fb

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:415d2e3e3a469facc78c34cadba6116cbb494d1294c2f132542ca3bc98087ce4
-size 655089
+oid sha256:545cbf14601a3346429144c73b7efb54b12b2a45da4ff6ce5dbde0b0d0254fab
+size 657207