Hugging Face's logo

dotlabs
/
void.1

Text Generation
Transformers
TensorBoard
Safetensors
English
qwen3
byte-level
pretraining
symbolic
text-generation-inference
Model card Files
xet
 Metrics Community
3
appvoid commited on
Commit
905147d
·
verified ·
1 Parent(s): 8cc8a7a

Update TensorBoard training metrics

Browse files
Files changed (1) hide show
  1. benchmark_cache/shard_scheduler.json +29 -29
benchmark_cache/shard_scheduler.json CHANGED
@@ -22,8 +22,8 @@
22
  "tokens_per_second": 1062967.6168812835,
23
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
24
  "estimated_pages_for_buffer": 1,
25
- "required_token_positions_per_second": 74477.16912629639,
26
- "required_pages_per_second_upper_estimate": 0.006455492259207954
27
  },
28
  {
29
  "source": "rewrite6",
@@ -36,8 +36,8 @@
36
  "tokens_per_second": 1206672.263377472,
37
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
38
  "estimated_pages_for_buffer": 1,
39
- "required_token_positions_per_second": 9309.646140787048,
40
- "required_pages_per_second_upper_estimate": 0.0037137570371737066
41
  },
42
  {
43
  "source": "openmath",
@@ -50,8 +50,8 @@
50
  "tokens_per_second": 410042.7581134458,
51
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
52
  "estimated_pages_for_buffer": 1,
53
- "required_token_positions_per_second": 18619.292281574097,
54
- "required_pages_per_second_upper_estimate": 0.004307245580989959
55
  },
56
  {
57
  "source": "ultra_qa",
@@ -64,8 +64,8 @@
64
  "tokens_per_second": 2163486.5624068812,
65
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
66
  "estimated_pages_for_buffer": 1,
67
- "required_token_positions_per_second": 27928.938422361145,
68
- "required_pages_per_second_upper_estimate": 0.001700725818106467
69
  },
70
  {
71
  "source": "ultra_style",
@@ -78,8 +78,8 @@
78
  "tokens_per_second": 1546694.2737633225,
79
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
80
  "estimated_pages_for_buffer": 1,
81
- "required_token_positions_per_second": 27928.938422361145,
82
- "required_pages_per_second_upper_estimate": 0.002150046992003118
83
  },
84
  {
85
  "source": "cortex",
@@ -92,8 +92,8 @@
92
  "tokens_per_second": 1015177.8395898147,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
  "estimated_pages_for_buffer": 21,
95
- "required_token_positions_per_second": 27928.938422361145,
96
- "required_pages_per_second_upper_estimate": 0.34678394306171256
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
@@ -158,33 +158,27 @@
158
  "prefetch_batches": 61,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 688946,
161
- "producer_batch_seconds": 0.5124305633830204,
162
- "producer_p95_seconds": 0.44285217400465626,
163
- "gpu_update_seconds": 0.989940955932066,
164
- "producer_batches_per_second": 1.9514839111041506,
165
- "gpu_batches_per_second": 1.010161256595817,
166
- "producer_headroom": 1.9318538484444898,
167
  "sustainable": true,
168
- "buffer_seconds": 60.38639831185603,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
- "upload_pages_per_commit": 79,
172
  "max_pending_pages": 4096,
173
- "generation_bytes_per_second": 245545.11968123144,
174
- "upload_bytes_per_second": 6253193.739659444,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
- "estimated_full_shards_per_hour": 52.68826668575008,
179
  "selection_policy": "Smallest shard within 90% of measured local peak, enlarged to amortize 30 seconds of measured page production; bounded to 256 MiB."
180
  },
181
  "upload_measurements": [
182
- {
183
- "seconds": 0.9625181489973329,
184
- "bytes": 5689945,
185
- "pages": 27,
186
- "shards": 2
187
- },
188
  {
189
  "seconds": 0.9489649809984257,
190
  "bytes": 5862208,
@@ -562,6 +556,12 @@
562
  "bytes": 11596177,
563
  "pages": 39,
564
  "shards": 4
 
 
 
 
 
 
565
  }
566
  ],
567
  "cache_key": {
 
22
  "tokens_per_second": 1062967.6168812835,
23
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
24
  "estimated_pages_for_buffer": 1,
25
+ "required_token_positions_per_second": 74488.08882560692,
26
+ "required_pages_per_second_upper_estimate": 0.006456438751068466
27
  },
28
  {
29
  "source": "rewrite6",
 
36
  "tokens_per_second": 1206672.263377472,
37
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
38
  "estimated_pages_for_buffer": 1,
39
+ "required_token_positions_per_second": 9311.011103200864,
40
+ "required_pages_per_second_upper_estimate": 0.0037143015410885847
41
  },
42
  {
43
  "source": "openmath",
 
50
  "tokens_per_second": 410042.7581134458,
51
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
52
  "estimated_pages_for_buffer": 1,
53
+ "required_token_positions_per_second": 18622.02220640173,
54
+ "required_pages_per_second_upper_estimate": 0.004307877101053795
55
  },
56
  {
57
  "source": "ultra_qa",
 
64
  "tokens_per_second": 2163486.5624068812,
65
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
66
  "estimated_pages_for_buffer": 1,
67
+ "required_token_positions_per_second": 27933.03330960259,
68
+ "required_pages_per_second_upper_estimate": 0.001700975175255258
69
  },
70
  {
71
  "source": "ultra_style",
 
78
  "tokens_per_second": 1546694.2737633225,
79
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
80
  "estimated_pages_for_buffer": 1,
81
+ "required_token_positions_per_second": 27933.03330960259,
82
+ "required_pages_per_second_upper_estimate": 0.002150362227758338
83
  },
84
  {
85
  "source": "cortex",
 
92
  "tokens_per_second": 1015177.8395898147,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
  "estimated_pages_for_buffer": 21,
95
+ "required_token_positions_per_second": 27933.03330960259,
96
+ "required_pages_per_second_upper_estimate": 0.3468347878565453
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
 
158
  "prefetch_batches": 61,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 688946,
161
+ "producer_batch_seconds": 0.144511708648281,
162
+ "producer_p95_seconds": 0.5217355390050216,
163
+ "gpu_update_seconds": 0.9897958339703621,
164
+ "producer_batches_per_second": 6.919854518043547,
165
+ "gpu_batches_per_second": 1.0103093644966215,
166
+ "producer_headroom": 6.849243173640491,
167
  "sustainable": true,
168
+ "buffer_seconds": 60.37754587219209,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
+ "upload_pages_per_commit": 78,
172
  "max_pending_pages": 4096,
173
+ "generation_bytes_per_second": 246534.60347499957,
174
+ "upload_bytes_per_second": 6254719.971833577,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
+ "estimated_full_shards_per_hour": 52.90058687388888,
179
  "selection_policy": "Smallest shard within 90% of measured local peak, enlarged to amortize 30 seconds of measured page production; bounded to 256 MiB."
180
  },
181
  "upload_measurements": [
 
 
 
 
 
 
182
  {
183
  "seconds": 0.9489649809984257,
184
  "bytes": 5862208,
 
556
  "bytes": 11596177,
557
  "pages": 39,
558
  "shards": 4
559
+ },
560
+ {
561
+ "seconds": 1.1032275320030749,
562
+ "bytes": 6673144,
563
+ "pages": 28,
564
+ "shards": 2
565
  }
566
  ],
567
  "cache_key": {