appvoid commited on
Commit
613e582
·
verified ·
1 Parent(s): ee0bfe9

Qwen90M step 205000; 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 +277 -25
  6. trainer.pt +2 -2
  7. training_config.json +2 -2
complete.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "step": 200000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
- "model.safetensors": "9e1098c2f4131285db26132369fd315a230c2a1ed6433490bf04e1d55302bd5c",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
- "trainer.pt": "121ebbd424d06e2eca68e9473ccb7613c34366b9b444edaa47833ef5a5344919",
12
- "training_config.json": "95ba9e002790d11959d1191b36ed994bfa10a6da51c22c772e143bec517a9783",
13
  "performance.json": "7f72f0a0e868a4baec06f238f922d0c5d7e397a51386c11ac32acc776d4d3522",
14
- "shard_scheduler.json": "d840cc690a9276bb980e6c1c2e254aea3ef8a4071771396fd2e073516734a949",
15
- "coverage.json": "33255e32c3d706b5219a7812b83fcf6ed8998528c0f3efa9d1b474d0202a80c0",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063"
 
1
  {
2
+ "step": 205000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
+ "model.safetensors": "81479c438b659ae76d14a43cced464395f2a4a922f5a7d702440d60c97a8d05d",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
+ "trainer.pt": "7e9a9266ee128ed17a9d65fb5caa62965e42f185628b8d60fc5b3a810de04245",
12
+ "training_config.json": "2122ad6f8105987f9b87dd55d13b632084853908566b081c829ffe801259e2bf",
13
  "performance.json": "7f72f0a0e868a4baec06f238f922d0c5d7e397a51386c11ac32acc776d4d3522",
14
+ "shard_scheduler.json": "6b78eed1b3798fed0667fc8eba1b21df9d8e0ade64a2f028654262590a5a3502",
15
+ "coverage.json": "cca4a03276b249b32504e7328f2319c84bec1a8153f03dd9d2c1e39fa1f9e3a6",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063"
coverage.json CHANGED
@@ -1,107 +1,107 @@
1
  {
2
- "step": 200000,
3
  "stats": {
4
- "climbmix/blocks": 7200000,
5
- "climbmix/targets": 12033775305,
6
- "climbmix/input_tokens": 12034537799,
7
- "ultra_qa/blocks": 2700000,
8
- "ultra_qa/targets": 4405129990,
9
- "ultra_qa/input_tokens": 4405138485,
10
- "cortex/blocks": 2700000,
11
- "cortex/targets": 102703673,
12
- "cortex/input_tokens": 4291696006,
13
- "ultra_style/blocks": 2700000,
14
- "ultra_style/targets": 4270419549,
15
- "ultra_style/input_tokens": 4270644269,
16
- "openmath/blocks": 1800000,
17
- "openmath/targets": 2781111203,
18
- "openmath/input_tokens": 2782038425,
19
- "rewrite6/blocks": 900000,
20
- "rewrite6/targets": 1475418140,
21
- "rewrite6/input_tokens": 1477035705
22
  },
23
  "sources": {
24
  "climbmix": {
25
- "page": 986,
26
- "row": 1904,
27
  "start": {
28
- "file": 47,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
- "row_group": 56,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
- "offset": 4096,
37
- "documents": 4040560
38
  },
39
  "rewrite6": {
40
- "page": 10,
41
- "row": 4037,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
- "row": 40960
47
  }
48
  },
49
  "epoch": 11,
50
  "offset": 0,
51
- "documents": 2304441
52
  },
53
  "openmath": {
54
- "page": 616,
55
- "row": 2034,
56
  "start": {
57
  "file": 5,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
- "row_group": 339,
61
- "row": 886
62
  }
63
  },
64
  "epoch": 0,
65
  "offset": 0,
66
- "documents": 2525170
67
  },
68
  "ultra_qa": {
69
- "page": 266,
70
- "row": 1939,
71
  "start": {
72
  "file": 2,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
- "row_group": 0,
76
- "row": 50285
77
  }
78
  },
79
  "epoch": 0,
80
  "offset": 2048,
81
- "documents": 1091475
82
  },
83
  "ultra_style": {
84
- "page": 407,
85
- "row": 311,
86
  "start": {
87
  "file": 2,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 3,
91
- "row": 41668
92
  }
93
  },
94
  "epoch": 0,
95
- "offset": 2048,
96
- "documents": 1667383
97
  },
98
  "cortex": {
99
- "page": 52363,
100
- "row": 94,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
- "documents": 6702558
105
  }
106
  }
107
  }
 
1
  {
2
+ "step": 205000,
3
  "stats": {
4
+ "climbmix/blocks": 7380000,
5
+ "climbmix/targets": 12335030804,
6
+ "climbmix/input_tokens": 12335812509,
7
+ "ultra_qa/blocks": 2767500,
8
+ "ultra_qa/targets": 4515170109,
9
+ "ultra_qa/input_tokens": 4515178820,
10
+ "cortex/blocks": 2767500,
11
+ "cortex/targets": 105260709,
12
+ "cortex/input_tokens": 4399023648,
13
+ "ultra_style/blocks": 2767500,
14
+ "ultra_style/targets": 4377672891,
15
+ "ultra_style/input_tokens": 4377902456,
16
+ "openmath/blocks": 1845000,
17
+ "openmath/targets": 2850633588,
18
+ "openmath/input_tokens": 2851583832,
19
+ "rewrite6/blocks": 922500,
20
+ "rewrite6/targets": 1512304899,
21
+ "rewrite6/input_tokens": 1513963391
22
  },
23
  "sources": {
24
  "climbmix": {
25
+ "page": 1011,
26
+ "row": 415,
27
  "start": {
28
+ "file": 48,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
+ "row_group": 74,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
+ "offset": 2048,
37
+ "documents": 4141471
38
  },
39
  "rewrite6": {
40
+ "page": 25,
41
+ "row": 802,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
+ "row": 102400
47
  }
48
  },
49
  "epoch": 11,
50
  "offset": 0,
51
+ "documents": 2362646
52
  },
53
  "openmath": {
54
+ "page": 631,
55
+ "row": 3535,
56
  "start": {
57
  "file": 5,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
+ "row_group": 401,
61
+ "row": 326
62
  }
63
  },
64
  "epoch": 0,
65
  "offset": 0,
66
+ "documents": 2588111
67
  },
68
  "ultra_qa": {
69
+ "page": 273,
70
+ "row": 561,
71
  "start": {
72
  "file": 2,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
+ "row_group": 1,
76
+ "row": 14394
77
  }
78
  },
79
  "epoch": 0,
80
  "offset": 2048,
81
+ "documents": 1118769
82
  },
83
  "ultra_style": {
84
+ "page": 416,
85
+ "row": 3531,
86
  "start": {
87
  "file": 2,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 3,
91
+ "row": 78532
92
  }
93
  },
94
  "epoch": 0,
95
+ "offset": 0,
96
+ "documents": 1707467
97
  },
98
  "cortex": {
99
+ "page": 53673,
100
+ "row": 65,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
+ "documents": 6870209
105
  }
106
  }
107
  }
latest.json CHANGED
@@ -1 +1 @@
1
- {"step": 200000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "9e1098c2f4131285db26132369fd315a230c2a1ed6433490bf04e1d55302bd5c", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "121ebbd424d06e2eca68e9473ccb7613c34366b9b444edaa47833ef5a5344919", "training_config.json": "95ba9e002790d11959d1191b36ed994bfa10a6da51c22c772e143bec517a9783", "performance.json": "7f72f0a0e868a4baec06f238f922d0c5d7e397a51386c11ac32acc776d4d3522", "shard_scheduler.json": "d840cc690a9276bb980e6c1c2e254aea3ef8a4071771396fd2e073516734a949", "coverage.json": "33255e32c3d706b5219a7812b83fcf6ed8998528c0f3efa9d1b474d0202a80c0", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063", "folder": "step-000000200000"}
 
1
+ {"step": 205000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "81479c438b659ae76d14a43cced464395f2a4a922f5a7d702440d60c97a8d05d", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "7e9a9266ee128ed17a9d65fb5caa62965e42f185628b8d60fc5b3a810de04245", "training_config.json": "2122ad6f8105987f9b87dd55d13b632084853908566b081c829ffe801259e2bf", "performance.json": "7f72f0a0e868a4baec06f238f922d0c5d7e397a51386c11ac32acc776d4d3522", "shard_scheduler.json": "6b78eed1b3798fed0667fc8eba1b21df9d8e0ade64a2f028654262590a5a3502", "coverage.json": "cca4a03276b249b32504e7328f2319c84bec1a8153f03dd9d2c1e39fa1f9e3a6", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063", "folder": "step-000000205000"}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:9e1098c2f4131285db26132369fd315a230c2a1ed6433490bf04e1d55302bd5c
3
  size 360622744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81479c438b659ae76d14a43cced464395f2a4a922f5a7d702440d60c97a8d05d
3
  size 360622744
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": 75216.18075821365,
26
- "required_pages_per_second_upper_estimate": 0.006519547914454685
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": 9402.022594776707,
40
- "required_pages_per_second_upper_estimate": 0.003750607385821249
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": 18804.045189553413,
54
- "required_pages_per_second_upper_estimate": 0.0043499849147108464
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": 28206.06778433012,
68
- "required_pages_per_second_upper_estimate": 0.0017176015422649853
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": 28206.06778433012,
82
- "required_pages_per_second_upper_estimate": 0.002171381177430663
83
  },
84
  {
85
  "source": "cortex",
@@ -91,9 +91,9 @@
91
  "read_seconds": 0.042780299000014566,
92
  "tokens_per_second": 1015177.8395898147,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
- "estimated_pages_for_buffer": 22,
95
- "required_token_positions_per_second": 28206.06778433012,
96
- "required_pages_per_second_upper_estimate": 0.35022496224505656
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": 0.1434574083060457,
157
  "selected": {
158
- "prefetch_batches": 62,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 688946,
161
- "producer_batch_seconds": 0.10476856211718655,
162
- "producer_p95_seconds": 0.18885634199978085,
163
- "gpu_update_seconds": 0.980214619471341,
164
- "producer_batches_per_second": 9.544847994396184,
165
- "gpu_batches_per_second": 1.0201847433568476,
166
- "producer_headroom": 9.355999544738848,
167
  "sustainable": true,
168
- "buffer_seconds": 60.773306407223146,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
- "upload_pages_per_commit": 81,
172
  "max_pending_pages": 4096,
173
- "generation_bytes_per_second": 364056.88282280555,
174
- "upload_bytes_per_second": 5423901.113376007,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
- "estimated_full_shards_per_hour": 78.11813224328161,
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": [
@@ -274,6 +274,258 @@
274
  "bytes": 2653813,
275
  "pages": 36,
276
  "shards": 2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
277
  }
278
  ],
279
  "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": 74801.05554614915,
26
+ "required_pages_per_second_upper_estimate": 0.006483565913198165
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": 9350.131943268643,
40
+ "required_pages_per_second_upper_estimate": 0.0037299074291003045
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": 18700.263886537286,
54
+ "required_pages_per_second_upper_estimate": 0.0043259769367466165
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": 28050.39582980593,
68
+ "required_pages_per_second_upper_estimate": 0.0017081219369820858
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": 28050.39582980593,
82
+ "required_pages_per_second_upper_estimate": 0.00215939711944384
83
  },
84
  {
85
  "source": "cortex",
 
91
  "read_seconds": 0.042780299000014566,
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": 28050.39582980593,
96
+ "required_pages_per_second_upper_estimate": 0.3482920375703829
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": 0.1434574083060457,
157
  "selected": {
158
+ "prefetch_batches": 61,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 688946,
161
+ "producer_batch_seconds": 0.18545458216408406,
162
+ "producer_p95_seconds": 0.21071651100101008,
163
+ "gpu_update_seconds": 0.9856545400554259,
164
+ "producer_batches_per_second": 5.392155795402419,
165
+ "gpu_batches_per_second": 1.0145542473164761,
166
+ "producer_headroom": 5.31480284042457,
167
  "sustainable": true,
168
+ "buffer_seconds": 60.12492694338098,
169
  "buffer_drain_seconds": null,
170
  "shard_target_mib": 16,
171
+ "upload_pages_per_commit": 82,
172
  "max_pending_pages": 4096,
173
+ "generation_bytes_per_second": 359053.68233167636,
174
+ "upload_bytes_per_second": 5324605.411232466,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
+ "estimated_full_shards_per_hour": 77.04456188643186,
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": [
 
274
  "bytes": 2653813,
275
  "pages": 36,
276
  "shards": 2
277
+ },
278
+ {
279
+ "seconds": 1.7666170649999913,
280
+ "bytes": 10799529,
281
+ "pages": 36,
282
+ "shards": 3
283
+ },
284
+ {
285
+ "seconds": 0.9170148619996326,
286
+ "bytes": 257577,
287
+ "pages": 8,
288
+ "shards": 1
289
+ },
290
+ {
291
+ "seconds": 1.4457078639998144,
292
+ "bytes": 7654224,
293
+ "pages": 37,
294
+ "shards": 3
295
+ },
296
+ {
297
+ "seconds": 0.9410484550007823,
298
+ "bytes": 1134403,
299
+ "pages": 34,
300
+ "shards": 1
301
+ },
302
+ {
303
+ "seconds": 1.2444776140000613,
304
+ "bytes": 6243354,
305
+ "pages": 26,
306
+ "shards": 2
307
+ },
308
+ {
309
+ "seconds": 1.5528372400003718,
310
+ "bytes": 12775308,
311
+ "pages": 38,
312
+ "shards": 4
313
+ },
314
+ {
315
+ "seconds": 1.527598386000136,
316
+ "bytes": 5658593,
317
+ "pages": 26,
318
+ "shards": 2
319
+ },
320
+ {
321
+ "seconds": 1.7142034060007063,
322
+ "bytes": 6133193,
323
+ "pages": 36,
324
+ "shards": 2
325
+ },
326
+ {
327
+ "seconds": 1.0039940289998412,
328
+ "bytes": 2541534,
329
+ "pages": 36,
330
+ "shards": 2
331
+ },
332
+ {
333
+ "seconds": 1.7366310759998669,
334
+ "bytes": 10198556,
335
+ "pages": 30,
336
+ "shards": 3
337
+ },
338
+ {
339
+ "seconds": 0.7939092509996044,
340
+ "bytes": 1225371,
341
+ "pages": 32,
342
+ "shards": 1
343
+ },
344
+ {
345
+ "seconds": 1.2473491540004034,
346
+ "bytes": 13161042,
347
+ "pages": 35,
348
+ "shards": 4
349
+ },
350
+ {
351
+ "seconds": 1.0770104249995711,
352
+ "bytes": 6141587,
353
+ "pages": 36,
354
+ "shards": 2
355
+ },
356
+ {
357
+ "seconds": 1.1525904230002197,
358
+ "bytes": 7408784,
359
+ "pages": 38,
360
+ "shards": 3
361
+ },
362
+ {
363
+ "seconds": 1.0340140519992929,
364
+ "bytes": 5739884,
365
+ "pages": 26,
366
+ "shards": 2
367
+ },
368
+ {
369
+ "seconds": 0.9941408290005711,
370
+ "bytes": 1118169,
371
+ "pages": 34,
372
+ "shards": 1
373
+ },
374
+ {
375
+ "seconds": 1.4089016780008023,
376
+ "bytes": 7539453,
377
+ "pages": 38,
378
+ "shards": 3
379
+ },
380
+ {
381
+ "seconds": 1.256968259000132,
382
+ "bytes": 11567319,
383
+ "pages": 33,
384
+ "shards": 3
385
+ },
386
+ {
387
+ "seconds": 1.1466406099998494,
388
+ "bytes": 5899615,
389
+ "pages": 35,
390
+ "shards": 2
391
+ },
392
+ {
393
+ "seconds": 0.9532060609999462,
394
+ "bytes": 7422959,
395
+ "pages": 29,
396
+ "shards": 3
397
+ },
398
+ {
399
+ "seconds": 1.2087593340002059,
400
+ "bytes": 6148456,
401
+ "pages": 35,
402
+ "shards": 2
403
+ },
404
+ {
405
+ "seconds": 1.2602108099999896,
406
+ "bytes": 2561145,
407
+ "pages": 35,
408
+ "shards": 3
409
+ },
410
+ {
411
+ "seconds": 1.5218778089993066,
412
+ "bytes": 10001472,
413
+ "pages": 31,
414
+ "shards": 3
415
+ },
416
+ {
417
+ "seconds": 1.655338631999257,
418
+ "bytes": 6858339,
419
+ "pages": 37,
420
+ "shards": 2
421
+ },
422
+ {
423
+ "seconds": 1.3612198569990142,
424
+ "bytes": 7511604,
425
+ "pages": 29,
426
+ "shards": 3
427
+ },
428
+ {
429
+ "seconds": 1.111224463000326,
430
+ "bytes": 5878363,
431
+ "pages": 35,
432
+ "shards": 2
433
+ },
434
+ {
435
+ "seconds": 1.1176849410003342,
436
+ "bytes": 5784032,
437
+ "pages": 33,
438
+ "shards": 2
439
+ },
440
+ {
441
+ "seconds": 1.2254727769995952,
442
+ "bytes": 7561832,
443
+ "pages": 34,
444
+ "shards": 3
445
+ },
446
+ {
447
+ "seconds": 1.1802483909996226,
448
+ "bytes": 1055971,
449
+ "pages": 33,
450
+ "shards": 1
451
+ },
452
+ {
453
+ "seconds": 1.59867495799881,
454
+ "bytes": 13042114,
455
+ "pages": 35,
456
+ "shards": 4
457
+ },
458
+ {
459
+ "seconds": 1.1432860509994498,
460
+ "bytes": 5906480,
461
+ "pages": 31,
462
+ "shards": 2
463
+ },
464
+ {
465
+ "seconds": 1.2417195279995212,
466
+ "bytes": 5623378,
467
+ "pages": 33,
468
+ "shards": 2
469
+ },
470
+ {
471
+ "seconds": 0.9634915199985699,
472
+ "bytes": 7597611,
473
+ "pages": 38,
474
+ "shards": 3
475
+ },
476
+ {
477
+ "seconds": 0.761399758999687,
478
+ "bytes": 1061264,
479
+ "pages": 32,
480
+ "shards": 1
481
+ },
482
+ {
483
+ "seconds": 1.0512442660001398,
484
+ "bytes": 5862459,
485
+ "pages": 28,
486
+ "shards": 2
487
+ },
488
+ {
489
+ "seconds": 1.2080343199995696,
490
+ "bytes": 17359027,
491
+ "pages": 39,
492
+ "shards": 5
493
+ },
494
+ {
495
+ "seconds": 1.0087294929999189,
496
+ "bytes": 1073260,
497
+ "pages": 33,
498
+ "shards": 1
499
+ },
500
+ {
501
+ "seconds": 1.4433996169991588,
502
+ "bytes": 7239065,
503
+ "pages": 29,
504
+ "shards": 3
505
+ },
506
+ {
507
+ "seconds": 0.894821537000098,
508
+ "bytes": 1079304,
509
+ "pages": 31,
510
+ "shards": 1
511
+ },
512
+ {
513
+ "seconds": 1.1567702649990679,
514
+ "bytes": 10579644,
515
+ "pages": 35,
516
+ "shards": 3
517
+ },
518
+ {
519
+ "seconds": 1.6480453410003975,
520
+ "bytes": 7697533,
521
+ "pages": 38,
522
+ "shards": 3
523
+ },
524
+ {
525
+ "seconds": 1.2853153289997863,
526
+ "bytes": 6668992,
527
+ "pages": 31,
528
+ "shards": 2
529
  }
530
  ],
531
  "cache_key": {
trainer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:121ebbd424d06e2eca68e9473ccb7613c34366b9b444edaa47833ef5a5344919
3
- size 721772159
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7e9a9266ee128ed17a9d65fb5caa62965e42f185628b8d60fc5b3a810de04245
3
+ size 721767807
training_config.json CHANGED
@@ -46,8 +46,8 @@
46
  "cpu_threads": 4,
47
  "page_rows": 4096,
48
  "cortex_page_rows": 128,
49
- "prefetch_batches": 62,
50
- "prefetch_low_water": 31,
51
  "memory_pages": 12,
52
  "cache_gb": 20,
53
  "force_rebenchmark": false,
 
46
  "cpu_threads": 4,
47
  "page_rows": 4096,
48
  "cortex_page_rows": 128,
49
+ "prefetch_batches": 61,
50
+ "prefetch_low_water": 30,
51
  "memory_pages": 12,
52
  "cache_gb": 20,
53
  "force_rebenchmark": false,