appvoid commited on
Commit
a2613f2
·
verified ·
1 Parent(s): 243d8f8

Qwen90M step 850000; complete resumable checkpoint

Browse files
Files changed (7) hide show
  1. complete.json +6 -6
  2. coverage.json +47 -47
  3. latest.json +1 -1
  4. model.safetensors +1 -1
  5. shard_scheduler.json +265 -25
  6. trainer.pt +2 -2
  7. training_config.json +1 -1
complete.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "step": 845000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
- "model.safetensors": "28d525df26bddcdad1f7e0ee21d00c4e2f6f2750838939feccfbef122663248b",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
- "trainer.pt": "87fc13206a6115414dc5698ba185a42d91483c871e4ad1d988dd31693f1ea24e",
12
- "training_config.json": "0ab64376a29947d9a9f670c0ece06c1900ab6f9ec82482957af5fc2c70beec58",
13
  "performance.json": "172134497605e50805257548c05018674e515323c63930c6237cd46d43786367",
14
- "shard_scheduler.json": "57f3f5db4104672d1623ef748776a5e35f680600431f1b4c0d4f28fc3f966af9",
15
- "coverage.json": "e2b9a86150908d61f021a350e36374564851cf14a57027b0b71fa60173ea1a90",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "052aedaef69b550457da29474e17df28fa635c99594a7b9c0881f3ac3827000e"
 
1
  {
2
+ "step": 850000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
+ "model.safetensors": "f63bca940961856416cd144272392c1c67906248d33fb6e3738aff0532ce52c9",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
+ "trainer.pt": "4916aa1216dc03a48e4cb8239cd7e2932ad783cf6e9cbdb87bd24f8b6be09e37",
12
+ "training_config.json": "4e8f8b2fc7193aabdbca12082e7c7cc3a0ed080bd5ff16e060cc02e10ffc53dd",
13
  "performance.json": "172134497605e50805257548c05018674e515323c63930c6237cd46d43786367",
14
+ "shard_scheduler.json": "f42d57a08c557b635b650e19e67a2af084db229b91e924fb49859e2aafade793",
15
+ "coverage.json": "3c7e63e40360305cb0d89f8e2d539d74a3ba66d8c1dc5a856f8c62962a636831",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "052aedaef69b550457da29474e17df28fa635c99594a7b9c0881f3ac3827000e"
coverage.json CHANGED
@@ -1,107 +1,107 @@
1
  {
2
- "step": 845000,
3
  "stats": {
4
- "climbmix/blocks": 30420000,
5
- "climbmix/targets": 50852257328,
6
- "climbmix/input_tokens": 50855479577,
7
- "ultra_qa/blocks": 11407500,
8
- "ultra_qa/targets": 18610655674,
9
- "ultra_qa/input_tokens": 18610691463,
10
- "cortex/blocks": 11407500,
11
- "cortex/targets": 433787145,
12
- "cortex/input_tokens": 18133094420,
13
- "ultra_style/blocks": 11407500,
14
- "ultra_style/targets": 18043082009,
15
- "ultra_style/input_tokens": 18044028468,
16
- "openmath/blocks": 7605000,
17
- "openmath/targets": 11751095221,
18
- "openmath/input_tokens": 11755011037,
19
- "rewrite6/blocks": 3802500,
20
- "rewrite6/targets": 6233650055,
21
- "rewrite6/input_tokens": 6240484559
22
  },
23
  "sources": {
24
  "climbmix": {
25
- "page": 4166,
26
- "row": 476,
27
  "start": {
28
- "file": 201,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
- "row_group": 54,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
- "offset": 0,
37
- "documents": 17064412
38
  },
39
  "rewrite6": {
40
- "page": 20,
41
- "row": 812,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
- "row": 81920
47
  }
48
  },
49
  "epoch": 47,
50
- "offset": 2048,
51
- "documents": 9736720
52
  },
53
  "openmath": {
54
- "page": 2603,
55
- "row": 3612,
56
  "start": {
57
  "file": 24,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
- "row_group": 182,
61
- "row": 289
62
  }
63
  },
64
  "epoch": 0,
65
  "offset": 2048,
66
- "documents": 10665500
67
  },
68
  "ultra_qa": {
69
- "page": 1125,
70
- "row": 3194,
71
  "start": {
72
  "file": 8,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
- "row_group": 6,
76
- "row": 63755
77
  }
78
  },
79
  "epoch": 0,
80
  "offset": 2048,
81
- "documents": 4611194
82
  },
83
  "ultra_style": {
84
- "page": 1718,
85
- "row": 1338,
86
  "start": {
87
  "file": 10,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 2,
91
- "row": 17547
92
  }
93
  },
94
  "epoch": 0,
95
  "offset": 4096,
96
- "documents": 7038266
97
  },
98
  "cortex": {
99
- "page": 221233,
100
- "row": 96,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
- "documents": 28317920
105
  }
106
  }
107
  }
 
1
  {
2
+ "step": 850000,
3
  "stats": {
4
+ "climbmix/blocks": 30600000,
5
+ "climbmix/targets": 51152990791,
6
+ "climbmix/input_tokens": 51156232218,
7
+ "ultra_qa/blocks": 11475000,
8
+ "ultra_qa/targets": 18720819448,
9
+ "ultra_qa/input_tokens": 18720855442,
10
+ "cortex/blocks": 11475000,
11
+ "cortex/targets": 436347195,
12
+ "cortex/input_tokens": 18240375745,
13
+ "ultra_style/blocks": 11475000,
14
+ "ultra_style/targets": 18150549959,
15
+ "ultra_style/input_tokens": 18151501590,
16
+ "openmath/blocks": 7650000,
17
+ "openmath/targets": 11820568087,
18
+ "openmath/input_tokens": 11824507080,
19
+ "rewrite6/blocks": 3825000,
20
+ "rewrite6/targets": 6270523226,
21
+ "rewrite6/input_tokens": 6277398213
22
  },
23
  "sources": {
24
  "climbmix": {
25
+ "page": 4190,
26
+ "row": 3595,
27
  "start": {
28
+ "file": 202,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
+ "row_group": 68,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
+ "offset": 12288,
37
+ "documents": 17165835
38
  },
39
  "rewrite6": {
40
+ "page": 34,
41
+ "row": 1198,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
+ "row": 139264
47
  }
48
  },
49
  "epoch": 47,
50
+ "offset": 0,
51
+ "documents": 9794450
52
  },
53
  "openmath": {
54
+ "page": 2619,
55
+ "row": 1184,
56
  "start": {
57
  "file": 24,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
+ "row_group": 247,
61
+ "row": 825
62
  }
63
  },
64
  "epoch": 0,
65
  "offset": 2048,
66
+ "documents": 10728608
67
  },
68
  "ultra_qa": {
69
+ "page": 1132,
70
+ "row": 1781,
71
  "start": {
72
  "file": 8,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
+ "row_group": 7,
76
+ "row": 27950
77
  }
78
  },
79
  "epoch": 0,
80
  "offset": 2048,
81
+ "documents": 4638453
82
  },
83
  "ultra_style": {
84
+ "page": 1728,
85
+ "row": 889,
86
  "start": {
87
  "file": 10,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 2,
91
+ "row": 58507
92
  }
93
  },
94
  "epoch": 0,
95
  "offset": 4096,
96
+ "documents": 7078777
97
  },
98
  "cortex": {
99
+ "page": 222542,
100
+ "row": 73,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
+ "documents": 28485449
105
  }
106
  }
107
  }
latest.json CHANGED
@@ -1 +1 @@
1
- {"step": 845000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "28d525df26bddcdad1f7e0ee21d00c4e2f6f2750838939feccfbef122663248b", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "87fc13206a6115414dc5698ba185a42d91483c871e4ad1d988dd31693f1ea24e", "training_config.json": "0ab64376a29947d9a9f670c0ece06c1900ab6f9ec82482957af5fc2c70beec58", "performance.json": "172134497605e50805257548c05018674e515323c63930c6237cd46d43786367", "shard_scheduler.json": "57f3f5db4104672d1623ef748776a5e35f680600431f1b4c0d4f28fc3f966af9", "coverage.json": "e2b9a86150908d61f021a350e36374564851cf14a57027b0b71fa60173ea1a90", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "052aedaef69b550457da29474e17df28fa635c99594a7b9c0881f3ac3827000e", "folder": "step-000000845000"}
 
1
+ {"step": 850000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "f63bca940961856416cd144272392c1c67906248d33fb6e3738aff0532ce52c9", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "4916aa1216dc03a48e4cb8239cd7e2932ad783cf6e9cbdb87bd24f8b6be09e37", "training_config.json": "4e8f8b2fc7193aabdbca12082e7c7cc3a0ed080bd5ff16e060cc02e10ffc53dd", "performance.json": "172134497605e50805257548c05018674e515323c63930c6237cd46d43786367", "shard_scheduler.json": "f42d57a08c557b635b650e19e67a2af084db229b91e924fb49859e2aafade793", "coverage.json": "3c7e63e40360305cb0d89f8e2d539d74a3ba66d8c1dc5a856f8c62962a636831", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "052aedaef69b550457da29474e17df28fa635c99594a7b9c0881f3ac3827000e", "folder": "step-000000850000"}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:28d525df26bddcdad1f7e0ee21d00c4e2f6f2750838939feccfbef122663248b
3
  size 360622744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f63bca940961856416cd144272392c1c67906248d33fb6e3738aff0532ce52c9
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": 76039.94462746628,
26
- "required_pages_per_second_upper_estimate": 0.005260619806307925
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": 9504.993078433285,
40
- "required_pages_per_second_upper_estimate": 0.003979542115113849
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": 19009.98615686657,
54
- "required_pages_per_second_upper_estimate": 0.003959278789427566
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": 28514.979235299852,
68
- "required_pages_per_second_upper_estimate": 0.001709429465303973
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": 28514.979235299852,
82
- "required_pages_per_second_upper_estimate": 0.002216914211406139
83
  },
84
  {
85
  "source": "cortex",
@@ -91,9 +91,9 @@
91
  "read_seconds": 0.04061180899998362,
92
  "tokens_per_second": 910315.184299418,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
- "estimated_pages_for_buffer": 22,
95
- "required_token_positions_per_second": 28514.979235299852,
96
- "required_pages_per_second_upper_estimate": 0.3657595366311342
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
@@ -155,27 +155,27 @@
155
  "shard_benchmark_scope": "Local packing/read measurements; upload capacity measured only on real data commits. Targets larger than available pages are sample-limited.",
156
  "estimated_fresh_generation_batch_seconds": 2.558590814491007,
157
  "selected": {
158
- "prefetch_batches": 62,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 720750,
161
- "producer_batch_seconds": 0.05139413005469606,
162
- "producer_p95_seconds": 0.15724994399988645,
163
- "gpu_update_seconds": 0.969595656088482,
164
- "producer_batches_per_second": 19.457474986652226,
165
- "gpu_batches_per_second": 1.0313577559063893,
166
- "producer_headroom": 18.86588322550829,
167
  "sustainable": true,
168
- "buffer_seconds": 60.11493067748588,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
- "upload_pages_per_commit": 83,
172
  "max_pending_pages": 4096,
173
- "generation_bytes_per_second": 403027.47604566225,
174
- "upload_bytes_per_second": 3951706.6780117266,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
- "estimated_full_shards_per_hour": 86.48031435992623,
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": [
@@ -304,6 +304,246 @@
304
  "bytes": 10661870,
305
  "pages": 35,
306
  "shards": 3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
307
  }
308
  ],
309
  "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": 76577.90463803377,
26
+ "required_pages_per_second_upper_estimate": 0.005297837128078182
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": 9572.238079754221,
40
+ "required_pages_per_second_upper_estimate": 0.004007696192931617
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": 19144.476159508442,
54
+ "required_pages_per_second_upper_estimate": 0.0039872895102379905
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": 28716.714239262663,
68
+ "required_pages_per_second_upper_estimate": 0.001721523170760028
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": 28716.714239262663,
82
+ "required_pages_per_second_upper_estimate": 0.0022325982206257418
83
  },
84
  {
85
  "source": "cortex",
 
91
  "read_seconds": 0.04061180899998362,
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": 28716.714239262663,
96
+ "required_pages_per_second_upper_estimate": 0.3683471766557979
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
 
155
  "shard_benchmark_scope": "Local packing/read measurements; upload capacity measured only on real data commits. Targets larger than available pages are sample-limited.",
156
  "estimated_fresh_generation_batch_seconds": 2.558590814491007,
157
  "selected": {
158
+ "prefetch_batches": 63,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 720750,
161
+ "producer_batch_seconds": 0.40197195971875743,
162
+ "producer_p95_seconds": 0.2049835260004329,
163
+ "gpu_update_seconds": 0.962784243686157,
164
+ "producer_batches_per_second": 2.4877357134553795,
165
+ "gpu_batches_per_second": 1.0386543055288868,
166
+ "producer_headroom": 2.3951527473701795,
167
  "sustainable": true,
168
+ "buffer_seconds": 60.65540735222789,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
+ "upload_pages_per_commit": 77,
172
  "max_pending_pages": 4096,
173
+ "generation_bytes_per_second": 335950.1157080666,
174
+ "upload_bytes_per_second": 4550117.712570053,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
+ "estimated_full_shards_per_hour": 72.08707431251048,
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": [
 
304
  "bytes": 10661870,
305
  "pages": 35,
306
  "shards": 3
307
+ },
308
+ {
309
+ "seconds": 1.9814531229999375,
310
+ "bytes": 6026004,
311
+ "pages": 34,
312
+ "shards": 2
313
+ },
314
+ {
315
+ "seconds": 1.2945359899999858,
316
+ "bytes": 2043786,
317
+ "pages": 19,
318
+ "shards": 2
319
+ },
320
+ {
321
+ "seconds": 0.9116111450002791,
322
+ "bytes": 853572,
323
+ "pages": 25,
324
+ "shards": 1
325
+ },
326
+ {
327
+ "seconds": 1.2689381730001514,
328
+ "bytes": 12316311,
329
+ "pages": 40,
330
+ "shards": 3
331
+ },
332
+ {
333
+ "seconds": 1.3365711859996736,
334
+ "bytes": 11197506,
335
+ "pages": 36,
336
+ "shards": 4
337
+ },
338
+ {
339
+ "seconds": 1.029820503999872,
340
+ "bytes": 1262307,
341
+ "pages": 35,
342
+ "shards": 1
343
+ },
344
+ {
345
+ "seconds": 1.888025431999722,
346
+ "bytes": 5850227,
347
+ "pages": 28,
348
+ "shards": 2
349
+ },
350
+ {
351
+ "seconds": 1.452950692000286,
352
+ "bytes": 7702767,
353
+ "pages": 37,
354
+ "shards": 3
355
+ },
356
+ {
357
+ "seconds": 1.831108772999869,
358
+ "bytes": 11417123,
359
+ "pages": 37,
360
+ "shards": 3
361
+ },
362
+ {
363
+ "seconds": 1.3576498049997099,
364
+ "bytes": 7486146,
365
+ "pages": 35,
366
+ "shards": 3
367
+ },
368
+ {
369
+ "seconds": 0.8907806349998282,
370
+ "bytes": 954545,
371
+ "pages": 28,
372
+ "shards": 1
373
+ },
374
+ {
375
+ "seconds": 1.3382855260006181,
376
+ "bytes": 6196201,
377
+ "pages": 35,
378
+ "shards": 2
379
+ },
380
+ {
381
+ "seconds": 1.8011739470002794,
382
+ "bytes": 11923984,
383
+ "pages": 37,
384
+ "shards": 4
385
+ },
386
+ {
387
+ "seconds": 0.9351237119999496,
388
+ "bytes": 1136147,
389
+ "pages": 34,
390
+ "shards": 1
391
+ },
392
+ {
393
+ "seconds": 1.611343528999896,
394
+ "bytes": 7401953,
395
+ "pages": 31,
396
+ "shards": 3
397
+ },
398
+ {
399
+ "seconds": 1.7755565249999563,
400
+ "bytes": 12020074,
401
+ "pages": 40,
402
+ "shards": 3
403
+ },
404
+ {
405
+ "seconds": 1.294827431000158,
406
+ "bytes": 6143505,
407
+ "pages": 34,
408
+ "shards": 2
409
+ },
410
+ {
411
+ "seconds": 1.2396700680001231,
412
+ "bytes": 7470065,
413
+ "pages": 39,
414
+ "shards": 3
415
+ },
416
+ {
417
+ "seconds": 1.6093791190005504,
418
+ "bytes": 874866,
419
+ "pages": 25,
420
+ "shards": 1
421
+ },
422
+ {
423
+ "seconds": 1.1941245199996047,
424
+ "bytes": 7420221,
425
+ "pages": 39,
426
+ "shards": 3
427
+ },
428
+ {
429
+ "seconds": 1.6006398999998055,
430
+ "bytes": 9959535,
431
+ "pages": 29,
432
+ "shards": 3
433
+ },
434
+ {
435
+ "seconds": 1.2196721769996657,
436
+ "bytes": 7030636,
437
+ "pages": 39,
438
+ "shards": 2
439
+ },
440
+ {
441
+ "seconds": 1.0609669329996905,
442
+ "bytes": 7430334,
443
+ "pages": 38,
444
+ "shards": 3
445
+ },
446
+ {
447
+ "seconds": 1.0616652130001967,
448
+ "bytes": 6074717,
449
+ "pages": 35,
450
+ "shards": 2
451
+ },
452
+ {
453
+ "seconds": 1.3232113710000704,
454
+ "bytes": 5040655,
455
+ "pages": 31,
456
+ "shards": 3
457
+ },
458
+ {
459
+ "seconds": 1.5191957560000446,
460
+ "bytes": 7567101,
461
+ "pages": 34,
462
+ "shards": 3
463
+ },
464
+ {
465
+ "seconds": 1.065770677000728,
466
+ "bytes": 897206,
467
+ "pages": 26,
468
+ "shards": 1
469
+ },
470
+ {
471
+ "seconds": 1.459628982999675,
472
+ "bytes": 13449017,
473
+ "pages": 40,
474
+ "shards": 4
475
+ },
476
+ {
477
+ "seconds": 1.560673630000565,
478
+ "bytes": 10527927,
479
+ "pages": 42,
480
+ "shards": 3
481
+ },
482
+ {
483
+ "seconds": 1.384273941000174,
484
+ "bytes": 1166498,
485
+ "pages": 35,
486
+ "shards": 1
487
+ },
488
+ {
489
+ "seconds": 1.245850141000119,
490
+ "bytes": 7339507,
491
+ "pages": 30,
492
+ "shards": 3
493
+ },
494
+ {
495
+ "seconds": 1.0799733499998183,
496
+ "bytes": 6017108,
497
+ "pages": 34,
498
+ "shards": 2
499
+ },
500
+ {
501
+ "seconds": 1.2428365799996755,
502
+ "bytes": 12889604,
503
+ "pages": 40,
504
+ "shards": 4
505
+ },
506
+ {
507
+ "seconds": 1.21177975899991,
508
+ "bytes": 6176322,
509
+ "pages": 36,
510
+ "shards": 2
511
+ },
512
+ {
513
+ "seconds": 1.4487905369996952,
514
+ "bytes": 925527,
515
+ "pages": 25,
516
+ "shards": 1
517
+ },
518
+ {
519
+ "seconds": 1.6813211230000888,
520
+ "bytes": 7642707,
521
+ "pages": 39,
522
+ "shards": 3
523
+ },
524
+ {
525
+ "seconds": 1.8502533870005209,
526
+ "bytes": 10875382,
527
+ "pages": 36,
528
+ "shards": 3
529
+ },
530
+ {
531
+ "seconds": 1.247339760999239,
532
+ "bytes": 856828,
533
+ "pages": 26,
534
+ "shards": 1
535
+ },
536
+ {
537
+ "seconds": 1.4432682090000526,
538
+ "bytes": 13357057,
539
+ "pages": 39,
540
+ "shards": 4
541
+ },
542
+ {
543
+ "seconds": 1.271395830000074,
544
+ "bytes": 6214888,
545
+ "pages": 36,
546
+ "shards": 2
547
  }
548
  ],
549
  "cache_key": {
trainer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:87fc13206a6115414dc5698ba185a42d91483c871e4ad1d988dd31693f1ea24e
3
- size 721767423
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4916aa1216dc03a48e4cb8239cd7e2932ad783cf6e9cbdb87bd24f8b6be09e37
3
+ size 721765631
training_config.json CHANGED
@@ -53,7 +53,7 @@
53
  "cpu_threads": 4,
54
  "page_rows": 4096,
55
  "cortex_page_rows": 128,
56
- "prefetch_batches": 62,
57
  "prefetch_low_water": 31,
58
  "memory_pages": 12,
59
  "cache_gb": 20,
 
53
  "cpu_threads": 4,
54
  "page_rows": 4096,
55
  "cortex_page_rows": 128,
56
+ "prefetch_batches": 63,
57
  "prefetch_low_water": 31,
58
  "memory_pages": 12,
59
  "cache_gb": 20,