appvoid commited on
Commit
9f91f48
·
verified ·
1 Parent(s): e99b627

Qwen90M step 495000; complete resumable checkpoint

Browse files
Files changed (6) hide show
  1. complete.json +5 -5
  2. coverage.json +50 -50
  3. latest.json +1 -1
  4. model.safetensors +1 -1
  5. shard_scheduler.json +275 -107
  6. trainer.pt +2 -2
complete.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "step": 490000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
- "model.safetensors": "f2108bb7a48fce23b36f33461a1e756b5caa7d1607df7ec77a36189ce14cd91a",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
- "trainer.pt": "057bc6789cc9af9e9681c23890813cd87d445ef5153b156a985027190827e3c4",
12
  "training_config.json": "c8bdf052510c2d9092e5a6824b119dd6584dc7697f8c4a8e9b7ecf3057c5bbb6",
13
  "performance.json": "234744063e094dfd7b525da4e181ac493fac400c830f871c7e8c59127ed10dc9",
14
- "shard_scheduler.json": "b69378129bd07cf55de152eb8802dc876a8d2b68eceb701d39f8a9e31bcbbbb5",
15
- "coverage.json": "719ac7183a62a1284cd10c856bf1014ea4d9dbf7c92da34175071b83066b7d81",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063"
 
1
  {
2
+ "step": 495000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
+ "model.safetensors": "04a9c9ab7806b229ff0c1f81bf062a0cb0f4ff40115f9dc54fcd35653143dba5",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
+ "trainer.pt": "b09148f8bdb03fa1cdc8767b6fe281fcc58f0012b4c4869a74f3162a6d4c4a19",
12
  "training_config.json": "c8bdf052510c2d9092e5a6824b119dd6584dc7697f8c4a8e9b7ecf3057c5bbb6",
13
  "performance.json": "234744063e094dfd7b525da4e181ac493fac400c830f871c7e8c59127ed10dc9",
14
+ "shard_scheduler.json": "5090f346773104837959c2e40fec42df5511a51ba2bc53922efca54c043ec515",
15
+ "coverage.json": "79dfd44233cf538b7f2dfa46bc0fdc879f8607366262089a9f9c1e5fcd1d4480",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063"
coverage.json CHANGED
@@ -1,107 +1,107 @@
1
  {
2
- "step": 490000,
3
  "stats": {
4
- "climbmix/blocks": 17640000,
5
- "climbmix/targets": 29485571146,
6
- "climbmix/input_tokens": 29487439993,
7
- "ultra_qa/blocks": 6615000,
8
- "ultra_qa/targets": 10791309156,
9
- "ultra_qa/input_tokens": 10791329945,
10
- "cortex/blocks": 6615000,
11
- "cortex/targets": 251600465,
12
- "cortex/input_tokens": 10515435320,
13
- "ultra_style/blocks": 6615000,
14
- "ultra_style/targets": 10462953915,
15
- "ultra_style/input_tokens": 10463502402,
16
- "openmath/blocks": 4410000,
17
- "openmath/targets": 6814294301,
18
- "openmath/input_tokens": 6816564814,
19
- "rewrite6/blocks": 2205000,
20
- "rewrite6/targets": 3614791751,
21
- "rewrite6/input_tokens": 3618755339
22
  },
23
  "sources": {
24
  "climbmix": {
25
- "page": 2416,
26
- "row": 2276,
27
  "start": {
28
- "file": 116,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
- "row_group": 75,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
- "offset": 0,
37
- "documents": 9898212
38
  },
39
  "rewrite6": {
40
- "page": 24,
41
- "row": 2371,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
- "row": 98304
47
  }
48
  },
49
  "epoch": 27,
50
  "offset": 0,
51
- "documents": 5646583
52
  },
53
  "openmath": {
54
- "page": 1509,
55
- "row": 3634,
56
  "start": {
57
  "file": 14,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
- "row_group": 67,
61
- "row": 764
62
  }
63
  },
64
  "epoch": 0,
65
- "offset": 0,
66
- "documents": 6184498
67
  },
68
  "ultra_qa": {
69
- "page": 652,
70
- "row": 3429,
71
  "start": {
72
  "file": 5,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
  "row_group": 1,
76
- "row": 8004
77
  }
78
  },
79
  "epoch": 0,
80
- "offset": 2048,
81
- "documents": 2674021
82
  },
83
  "ultra_style": {
84
- "page": 996,
85
- "row": 1168,
86
  "start": {
87
- "file": 5,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
- "row_group": 7,
91
- "row": 59261
92
  }
93
  },
94
  "epoch": 0,
95
- "offset": 2048,
96
- "documents": 4080784
97
  },
98
  "cortex": {
99
- "page": 128297,
100
- "row": 17,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
- "documents": 16422033
105
  }
106
  }
107
  }
 
1
  {
2
+ "step": 495000,
3
  "stats": {
4
+ "climbmix/blocks": 17820000,
5
+ "climbmix/targets": 29786461296,
6
+ "climbmix/input_tokens": 29788349282,
7
+ "ultra_qa/blocks": 6682500,
8
+ "ultra_qa/targets": 10901546468,
9
+ "ultra_qa/input_tokens": 10901567474,
10
+ "cortex/blocks": 6682500,
11
+ "cortex/targets": 254162427,
12
+ "cortex/input_tokens": 10622791425,
13
+ "ultra_style/blocks": 6682500,
14
+ "ultra_style/targets": 10570465550,
15
+ "ultra_style/input_tokens": 10571019587,
16
+ "openmath/blocks": 4455000,
17
+ "openmath/targets": 6883812928,
18
+ "openmath/input_tokens": 6886106549,
19
+ "rewrite6/blocks": 2227500,
20
+ "rewrite6/targets": 3651663877,
21
+ "rewrite6/input_tokens": 3655667526
22
  },
23
  "sources": {
24
  "climbmix": {
25
+ "page": 2441,
26
+ "row": 677,
27
  "start": {
28
+ "file": 118,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
+ "row_group": 9,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
+ "offset": 2048,
37
+ "documents": 9999013
38
  },
39
  "rewrite6": {
40
+ "page": 38,
41
+ "row": 2269,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
+ "row": 155648
47
  }
48
  },
49
  "epoch": 27,
50
  "offset": 0,
51
+ "documents": 5703825
52
  },
53
  "openmath": {
54
+ "page": 1525,
55
+ "row": 1171,
56
  "start": {
57
  "file": 14,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
+ "row_group": 133,
61
+ "row": 300
62
  }
63
  },
64
  "epoch": 0,
65
+ "offset": 2048,
66
+ "documents": 6247571
67
  },
68
  "ultra_qa": {
69
+ "page": 659,
70
+ "row": 2052,
71
  "start": {
72
  "file": 5,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
  "row_group": 1,
76
+ "row": 36676
77
  }
78
  },
79
  "epoch": 0,
80
+ "offset": 4096,
81
+ "documents": 2701316
82
  },
83
  "ultra_style": {
84
+ "page": 1006,
85
+ "row": 1193,
86
  "start": {
87
+ "file": 6,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
+ "row_group": 0,
91
+ "row": 10924
92
  }
93
  },
94
  "epoch": 0,
95
+ "offset": 0,
96
+ "documents": 4121769
97
  },
98
  "cortex": {
99
+ "page": 129606,
100
+ "row": 60,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
+ "documents": 16589628
105
  }
106
  }
107
  }
latest.json CHANGED
@@ -1 +1 @@
1
- {"step": 490000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "f2108bb7a48fce23b36f33461a1e756b5caa7d1607df7ec77a36189ce14cd91a", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "057bc6789cc9af9e9681c23890813cd87d445ef5153b156a985027190827e3c4", "training_config.json": "c8bdf052510c2d9092e5a6824b119dd6584dc7697f8c4a8e9b7ecf3057c5bbb6", "performance.json": "234744063e094dfd7b525da4e181ac493fac400c830f871c7e8c59127ed10dc9", "shard_scheduler.json": "b69378129bd07cf55de152eb8802dc876a8d2b68eceb701d39f8a9e31bcbbbb5", "coverage.json": "719ac7183a62a1284cd10c856bf1014ea4d9dbf7c92da34175071b83066b7d81", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063", "folder": "step-000000490000"}
 
1
+ {"step": 495000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "04a9c9ab7806b229ff0c1f81bf062a0cb0f4ff40115f9dc54fcd35653143dba5", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "b09148f8bdb03fa1cdc8767b6fe281fcc58f0012b4c4869a74f3162a6d4c4a19", "training_config.json": "c8bdf052510c2d9092e5a6824b119dd6584dc7697f8c4a8e9b7ecf3057c5bbb6", "performance.json": "234744063e094dfd7b525da4e181ac493fac400c830f871c7e8c59127ed10dc9", "shard_scheduler.json": "5090f346773104837959c2e40fec42df5511a51ba2bc53922efca54c043ec515", "coverage.json": "79dfd44233cf538b7f2dfa46bc0fdc879f8607366262089a9f9c1e5fcd1d4480", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063", "folder": "step-000000495000"}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:f2108bb7a48fce23b36f33461a1e756b5caa7d1607df7ec77a36189ce14cd91a
3
  size 360622744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04a9c9ab7806b229ff0c1f81bf062a0cb0f4ff40115f9dc54fcd35653143dba5
3
  size 360622744
shard_scheduler.json CHANGED
@@ -22,8 +22,8 @@
22
  "tokens_per_second": 598315.063954628,
23
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
24
  "estimated_pages_for_buffer": 1,
25
- "required_token_positions_per_second": 76295.78030360074,
26
- "required_pages_per_second_upper_estimate": 0.00527831911200346
27
  },
28
  {
29
  "source": "rewrite6",
@@ -36,8 +36,8 @@
36
  "tokens_per_second": 8533.957213384612,
37
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
38
  "estimated_pages_for_buffer": 1,
39
- "required_token_positions_per_second": 9536.972537950092,
40
- "required_pages_per_second_upper_estimate": 0.003992931247006483
41
  },
42
  {
43
  "source": "openmath",
@@ -50,8 +50,8 @@
50
  "tokens_per_second": 380144.35761751566,
51
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
52
  "estimated_pages_for_buffer": 1,
53
- "required_token_positions_per_second": 19073.945075900185,
54
- "required_pages_per_second_upper_estimate": 0.003972599745552147
55
  },
56
  {
57
  "source": "ultra_qa",
@@ -64,8 +64,8 @@
64
  "tokens_per_second": 43742.16556775435,
65
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
66
  "estimated_pages_for_buffer": 1,
67
- "required_token_positions_per_second": 28610.917613850273,
68
- "required_pages_per_second_upper_estimate": 0.0017151808246086487
69
  },
70
  {
71
  "source": "ultra_style",
@@ -78,8 +78,8 @@
78
  "tokens_per_second": 22018.86276523721,
79
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
80
  "estimated_pages_for_buffer": 1,
81
- "required_token_positions_per_second": 28610.917613850273,
82
- "required_pages_per_second_upper_estimate": 0.002224372998350104
83
  },
84
  {
85
  "source": "cortex",
@@ -92,8 +92,8 @@
92
  "tokens_per_second": 910315.184299418,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
  "estimated_pages_for_buffer": 23,
95
- "required_token_positions_per_second": 28610.917613850273,
96
- "required_pages_per_second_upper_estimate": 0.36699013114057377
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
@@ -158,111 +158,27 @@
158
  "prefetch_batches": 63,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 704566,
161
- "producer_batch_seconds": 0.38075650120313753,
162
- "producer_p95_seconds": 0.17078465200029314,
163
- "gpu_update_seconds": 0.9663443994755297,
164
- "producer_batches_per_second": 2.6263504282662,
165
- "gpu_batches_per_second": 1.034827749343543,
166
- "producer_headroom": 2.537959027415201,
167
  "sustainable": true,
168
- "buffer_seconds": 60.87969716695837,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
- "upload_pages_per_commit": 84,
172
  "max_pending_pages": 4096,
173
- "generation_bytes_per_second": 391819.6533964042,
174
- "upload_bytes_per_second": 5833319.346352694,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
- "estimated_full_shards_per_hour": 84.07537652415365,
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": 1.0425156310002421,
184
- "bytes": 44934,
185
- "pages": 1,
186
- "shards": 1
187
- },
188
- {
189
- "seconds": 1.7656013979999443,
190
- "bytes": 11283965,
191
- "pages": 34,
192
- "shards": 4
193
- },
194
- {
195
- "seconds": 1.4391079690003608,
196
- "bytes": 12352688,
197
- "pages": 39,
198
- "shards": 3
199
- },
200
- {
201
- "seconds": 0.9268819079998138,
202
- "bytes": 1009442,
203
- "pages": 28,
204
- "shards": 1
205
- },
206
- {
207
- "seconds": 1.2650270520002778,
208
- "bytes": 7586230,
209
- "pages": 36,
210
- "shards": 3
211
- },
212
- {
213
- "seconds": 1.2436492840001847,
214
- "bytes": 10846371,
215
- "pages": 34,
216
- "shards": 3
217
- },
218
- {
219
- "seconds": 0.9011205990000235,
220
- "bytes": 2637170,
221
- "pages": 37,
222
- "shards": 2
223
- },
224
- {
225
- "seconds": 0.9096160310000414,
226
- "bytes": 6003921,
227
- "pages": 36,
228
- "shards": 2
229
- },
230
- {
231
- "seconds": 1.0415576699997473,
232
- "bytes": 6858425,
233
- "pages": 32,
234
- "shards": 2
235
- },
236
- {
237
- "seconds": 1.2630921189993387,
238
- "bytes": 11470878,
239
- "pages": 37,
240
- "shards": 4
241
- },
242
- {
243
- "seconds": 1.1407325299996955,
244
- "bytes": 6174556,
245
- "pages": 37,
246
- "shards": 2
247
- },
248
- {
249
- "seconds": 0.9178262779996658,
250
- "bytes": 2279800,
251
- "pages": 27,
252
- "shards": 2
253
- },
254
- {
255
- "seconds": 1.0028891179999846,
256
- "bytes": 5661960,
257
- "pages": 32,
258
- "shards": 2
259
- },
260
- {
261
- "seconds": 1.3030792019999353,
262
- "bytes": 10375056,
263
- "pages": 39,
264
- "shards": 3
265
- },
266
  {
267
  "seconds": 1.011828892000267,
268
  "bytes": 8292186,
@@ -394,6 +310,258 @@
394
  "bytes": 1187716,
395
  "pages": 35,
396
  "shards": 1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
397
  }
398
  ],
399
  "cache_key": {
 
22
  "tokens_per_second": 598315.063954628,
23
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
24
  "estimated_pages_for_buffer": 1,
25
+ "required_token_positions_per_second": 77246.17314450147,
26
+ "required_pages_per_second_upper_estimate": 0.005344069493952184
27
  },
28
  {
29
  "source": "rewrite6",
 
36
  "tokens_per_second": 8533.957213384612,
37
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
38
  "estimated_pages_for_buffer": 1,
39
+ "required_token_positions_per_second": 9655.771643062684,
40
+ "required_pages_per_second_upper_estimate": 0.004042669951509708
41
  },
42
  {
43
  "source": "openmath",
 
50
  "tokens_per_second": 380144.35761751566,
51
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
52
  "estimated_pages_for_buffer": 1,
53
+ "required_token_positions_per_second": 19311.54328612537,
54
+ "required_pages_per_second_upper_estimate": 0.0040220851868558866
55
  },
56
  {
57
  "source": "ultra_qa",
 
64
  "tokens_per_second": 43742.16556775435,
65
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
66
  "estimated_pages_for_buffer": 1,
67
+ "required_token_positions_per_second": 28967.31492918805,
68
+ "required_pages_per_second_upper_estimate": 0.0017365462994759575
69
  },
70
  {
71
  "source": "ultra_style",
 
78
  "tokens_per_second": 22018.86276523721,
79
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
80
  "estimated_pages_for_buffer": 1,
81
+ "required_token_positions_per_second": 28967.31492918805,
82
+ "required_pages_per_second_upper_estimate": 0.002252081321991498
83
  },
84
  {
85
  "source": "cortex",
 
92
  "tokens_per_second": 910315.184299418,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
  "estimated_pages_for_buffer": 23,
95
+ "required_token_positions_per_second": 28967.31492918805,
96
+ "required_pages_per_second_upper_estimate": 0.3715616132321039
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
 
158
  "prefetch_batches": 63,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 704566,
161
+ "producer_batch_seconds": 0.18283066840626816,
162
+ "producer_p95_seconds": 0.22040481899966835,
163
+ "gpu_update_seconds": 0.9544550493405005,
164
+ "producer_batches_per_second": 5.469541892052264,
165
+ "gpu_batches_per_second": 1.0477182772420448,
166
+ "producer_headroom": 5.220431876448678,
167
  "sustainable": true,
168
+ "buffer_seconds": 60.13066810845153,
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": 476173.2566712523,
174
+ "upload_bytes_per_second": 5837253.21659938,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
+ "estimated_full_shards_per_hour": 102.17569613555123,
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": 1.011828892000267,
184
  "bytes": 8292186,
 
310
  "bytes": 1187716,
311
  "pages": 35,
312
  "shards": 1
313
+ },
314
+ {
315
+ "seconds": 1.3033395869997548,
316
+ "bytes": 10049773,
317
+ "pages": 30,
318
+ "shards": 3
319
+ },
320
+ {
321
+ "seconds": 0.8627702979993046,
322
+ "bytes": 2455057,
323
+ "pages": 28,
324
+ "shards": 2
325
+ },
326
+ {
327
+ "seconds": 1.2129529189996902,
328
+ "bytes": 11955081,
329
+ "pages": 34,
330
+ "shards": 3
331
+ },
332
+ {
333
+ "seconds": 0.95688545899975,
334
+ "bytes": 1097352,
335
+ "pages": 30,
336
+ "shards": 1
337
+ },
338
+ {
339
+ "seconds": 1.40065100099946,
340
+ "bytes": 12237786,
341
+ "pages": 38,
342
+ "shards": 4
343
+ },
344
+ {
345
+ "seconds": 1.0810250989998167,
346
+ "bytes": 5909598,
347
+ "pages": 28,
348
+ "shards": 2
349
+ },
350
+ {
351
+ "seconds": 0.9096889889988233,
352
+ "bytes": 1202765,
353
+ "pages": 34,
354
+ "shards": 1
355
+ },
356
+ {
357
+ "seconds": 1.2456794589998026,
358
+ "bytes": 7629808,
359
+ "pages": 38,
360
+ "shards": 3
361
+ },
362
+ {
363
+ "seconds": 0.9217846150004334,
364
+ "bytes": 6311419,
365
+ "pages": 21,
366
+ "shards": 2
367
+ },
368
+ {
369
+ "seconds": 1.1030007570007,
370
+ "bytes": 11547889,
371
+ "pages": 39,
372
+ "shards": 4
373
+ },
374
+ {
375
+ "seconds": 1.05019463699864,
376
+ "bytes": 6120696,
377
+ "pages": 37,
378
+ "shards": 2
379
+ },
380
+ {
381
+ "seconds": 1.070229210999969,
382
+ "bytes": 1151046,
383
+ "pages": 34,
384
+ "shards": 1
385
+ },
386
+ {
387
+ "seconds": 1.0475356750011997,
388
+ "bytes": 11473932,
389
+ "pages": 33,
390
+ "shards": 4
391
+ },
392
+ {
393
+ "seconds": 1.2914608989995031,
394
+ "bytes": 6033641,
395
+ "pages": 36,
396
+ "shards": 2
397
+ },
398
+ {
399
+ "seconds": 1.115209474999574,
400
+ "bytes": 812365,
401
+ "pages": 23,
402
+ "shards": 1
403
+ },
404
+ {
405
+ "seconds": 1.1693048819997784,
406
+ "bytes": 7948420,
407
+ "pages": 20,
408
+ "shards": 3
409
+ },
410
+ {
411
+ "seconds": 1.2017747059999238,
412
+ "bytes": 5823273,
413
+ "pages": 30,
414
+ "shards": 2
415
+ },
416
+ {
417
+ "seconds": 1.1922360469998239,
418
+ "bytes": 10997251,
419
+ "pages": 39,
420
+ "shards": 3
421
+ },
422
+ {
423
+ "seconds": 1.0726386229998752,
424
+ "bytes": 2419938,
425
+ "pages": 28,
426
+ "shards": 2
427
+ },
428
+ {
429
+ "seconds": 1.240178467000078,
430
+ "bytes": 6103593,
431
+ "pages": 36,
432
+ "shards": 2
433
+ },
434
+ {
435
+ "seconds": 0.8772162379991642,
436
+ "bytes": 2610307,
437
+ "pages": 37,
438
+ "shards": 2
439
+ },
440
+ {
441
+ "seconds": 1.183507477000603,
442
+ "bytes": 15672429,
443
+ "pages": 34,
444
+ "shards": 4
445
+ },
446
+ {
447
+ "seconds": 1.1264393319997907,
448
+ "bytes": 6331952,
449
+ "pages": 36,
450
+ "shards": 2
451
+ },
452
+ {
453
+ "seconds": 0.856032101000892,
454
+ "bytes": 2311911,
455
+ "pages": 29,
456
+ "shards": 2
457
+ },
458
+ {
459
+ "seconds": 1.5313736190000782,
460
+ "bytes": 6009990,
461
+ "pages": 37,
462
+ "shards": 3
463
+ },
464
+ {
465
+ "seconds": 1.0109308320006676,
466
+ "bytes": 5084502,
467
+ "pages": 28,
468
+ "shards": 2
469
+ },
470
+ {
471
+ "seconds": 1.1290311129996553,
472
+ "bytes": 7603249,
473
+ "pages": 38,
474
+ "shards": 3
475
+ },
476
+ {
477
+ "seconds": 1.4725222399993072,
478
+ "bytes": 12055069,
479
+ "pages": 36,
480
+ "shards": 3
481
+ },
482
+ {
483
+ "seconds": 1.2178526319985394,
484
+ "bytes": 7463726,
485
+ "pages": 32,
486
+ "shards": 3
487
+ },
488
+ {
489
+ "seconds": 1.3964880730000004,
490
+ "bytes": 6495644,
491
+ "pages": 37,
492
+ "shards": 2
493
+ },
494
+ {
495
+ "seconds": 1.2645150809985353,
496
+ "bytes": 6000960,
497
+ "pages": 35,
498
+ "shards": 2
499
+ },
500
+ {
501
+ "seconds": 0.9461197429991444,
502
+ "bytes": 2470605,
503
+ "pages": 29,
504
+ "shards": 2
505
+ },
506
+ {
507
+ "seconds": 1.2941537009992317,
508
+ "bytes": 6225550,
509
+ "pages": 37,
510
+ "shards": 2
511
+ },
512
+ {
513
+ "seconds": 1.2335469089994149,
514
+ "bytes": 11703737,
515
+ "pages": 35,
516
+ "shards": 3
517
+ },
518
+ {
519
+ "seconds": 1.1459508749994711,
520
+ "bytes": 7841640,
521
+ "pages": 34,
522
+ "shards": 3
523
+ },
524
+ {
525
+ "seconds": 1.0024716899988562,
526
+ "bytes": 5917383,
527
+ "pages": 36,
528
+ "shards": 2
529
+ },
530
+ {
531
+ "seconds": 1.1789036200007104,
532
+ "bytes": 2313669,
533
+ "pages": 29,
534
+ "shards": 2
535
+ },
536
+ {
537
+ "seconds": 1.5398991070014745,
538
+ "bytes": 10300581,
539
+ "pages": 40,
540
+ "shards": 3
541
+ },
542
+ {
543
+ "seconds": 0.9979469910012995,
544
+ "bytes": 6506963,
545
+ "pages": 36,
546
+ "shards": 2
547
+ },
548
+ {
549
+ "seconds": 1.1795985310000106,
550
+ "bytes": 8204299,
551
+ "pages": 31,
552
+ "shards": 3
553
+ },
554
+ {
555
+ "seconds": 0.9598508790004416,
556
+ "bytes": 6041026,
557
+ "pages": 37,
558
+ "shards": 2
559
+ },
560
+ {
561
+ "seconds": 0.9930732779994287,
562
+ "bytes": 6062259,
563
+ "pages": 32,
564
+ "shards": 3
565
  }
566
  ],
567
  "cache_key": {
trainer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:057bc6789cc9af9e9681c23890813cd87d445ef5153b156a985027190827e3c4
3
- size 721770431
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b09148f8bdb03fa1cdc8767b6fe281fcc58f0012b4c4869a74f3162a6d4c4a19
3
+ size 721760383