appvoid commited on
Commit
34a288a
·
verified ·
1 Parent(s): dc9c017

Qwen90M step 655000; complete resumable checkpoint

Browse files
Files changed (7) hide show
  1. complete.json +6 -6
  2. coverage.json +45 -45
  3. latest.json +1 -1
  4. model.safetensors +1 -1
  5. shard_scheduler.json +264 -24
  6. trainer.pt +2 -2
  7. training_config.json +2 -2
complete.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "step": 650000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
- "model.safetensors": "da11afecaf8f801d2a5947a448520d9952df92bc0f9b587308168087a5a10740",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
- "trainer.pt": "ac22b9531603fa7bf3fc68ecd8b4c5213bb5f505ba92d8ef622eca0950b1cf9c",
12
- "training_config.json": "6c3c3ccff89dabb5098a22bc207c2f3abaee8b3bffec3f6cf6ad361a63b1a7fb",
13
  "performance.json": "5949473856ba60c6ea6e09958b880f6c16bf1db77e55259c9de9f88123633edf",
14
- "shard_scheduler.json": "55d15968e7c444b02f214615e89058b62eb3f772fd3481670f0251a16bf0c4b8",
15
- "coverage.json": "14205d54ff1f945944acaf4f3ef998712fa0fda76b995cac46c73b375d4ed339",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "33a459803c8cffa8929a90def998b4ad2c7c52fd1245365a9ee946b5552d845a"
 
1
  {
2
+ "step": 655000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
+ "model.safetensors": "f16c0f02e7591de5e4117964ecbe2cf4bfaf21e66738848488a66d81a023a4af",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
+ "trainer.pt": "2417fe4e5ccee0dce50398268fdd14b4c593283295e3ff26549369405615ffe6",
12
+ "training_config.json": "7234b7fffb86a474d2daeb3e14d8a78dc8439f4687eabb77722a8c9363322e29",
13
  "performance.json": "5949473856ba60c6ea6e09958b880f6c16bf1db77e55259c9de9f88123633edf",
14
+ "shard_scheduler.json": "ddb75d8dad5fb16ded850552a66cff01427ff6316465608b9827b1c59db198d3",
15
+ "coverage.json": "a0e4dda5a02b5917719ba5e1e749b733a37d59cd87fb879d106b7851fab0223f",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "33a459803c8cffa8929a90def998b4ad2c7c52fd1245365a9ee946b5552d845a"
coverage.json CHANGED
@@ -1,107 +1,107 @@
1
  {
2
- "step": 650000,
3
  "stats": {
4
- "climbmix/blocks": 23400000,
5
- "climbmix/targets": 39118037740,
6
- "climbmix/input_tokens": 39120516591,
7
- "ultra_qa/blocks": 8775000,
8
- "ultra_qa/targets": 14316037642,
9
- "ultra_qa/input_tokens": 14316065300,
10
- "cortex/blocks": 8775000,
11
- "cortex/targets": 333666382,
12
- "cortex/input_tokens": 13948402548,
13
- "ultra_style/blocks": 8775000,
14
- "ultra_style/targets": 13879953673,
15
- "ultra_style/input_tokens": 13880681377,
16
- "openmath/blocks": 5850000,
17
- "openmath/targets": 9039281192,
18
- "openmath/input_tokens": 9042292425,
19
- "rewrite6/blocks": 2925000,
20
- "rewrite6/targets": 4795141737,
21
- "rewrite6/input_tokens": 4800399327
22
  },
23
  "sources": {
24
  "climbmix": {
25
- "page": 3204,
26
- "row": 2005,
27
  "start": {
28
- "file": 155,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
- "row_group": 8,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
  "offset": 2048,
37
- "documents": 13125589
38
  },
39
  "rewrite6": {
40
- "page": 23,
41
- "row": 1350,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
- "row": 94208
47
  }
48
  },
49
  "epoch": 36,
50
  "offset": 0,
51
- "documents": 7490102
52
  },
53
  "openmath": {
54
- "page": 2002,
55
- "row": 2865,
56
  "start": {
57
  "file": 18,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
- "row_group": 340,
61
- "row": 492
62
  }
63
  },
64
  "epoch": 0,
65
  "offset": 0,
66
- "documents": 8203057
67
  },
68
  "ultra_qa": {
69
- "page": 866,
70
- "row": 133,
71
  "start": {
72
  "file": 6,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
- "row_group": 6,
76
- "row": 42065
77
  }
78
  },
79
  "epoch": 0,
80
  "offset": 2048,
81
- "documents": 3547269
82
  },
83
  "ultra_style": {
84
- "page": 1321,
85
- "row": 2238,
86
  "start": {
87
  "file": 7,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 7,
91
- "row": 20750
92
  }
93
  },
94
  "epoch": 0,
95
  "offset": 0,
96
- "documents": 5413054
97
  },
98
  "cortex": {
99
- "page": 170183,
100
- "row": 27,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
- "documents": 21783451
105
  }
106
  }
107
  }
 
1
  {
2
+ "step": 655000,
3
  "stats": {
4
+ "climbmix/blocks": 23580000,
5
+ "climbmix/targets": 39418845444,
6
+ "climbmix/input_tokens": 39421343499,
7
+ "ultra_qa/blocks": 8842500,
8
+ "ultra_qa/targets": 14426061014,
9
+ "ultra_qa/input_tokens": 14426088863,
10
+ "cortex/blocks": 8842500,
11
+ "cortex/targets": 336226856,
12
+ "cortex/input_tokens": 14055659885,
13
+ "ultra_style/blocks": 8842500,
14
+ "ultra_style/targets": 13986714370,
15
+ "ultra_style/input_tokens": 13987447583,
16
+ "openmath/blocks": 5895000,
17
+ "openmath/targets": 9108714338,
18
+ "openmath/input_tokens": 9111748654,
19
+ "rewrite6/blocks": 2947500,
20
+ "rewrite6/targets": 4831985405,
21
+ "rewrite6/input_tokens": 4837283018
22
  },
23
  "sources": {
24
  "climbmix": {
25
+ "page": 3229,
26
+ "row": 903,
27
  "start": {
28
+ "file": 156,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
+ "row_group": 25,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
  "offset": 2048,
37
+ "documents": 13226887
38
  },
39
  "rewrite6": {
40
+ "page": 37,
41
+ "row": 1218,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
+ "row": 151552
47
  }
48
  },
49
  "epoch": 36,
50
  "offset": 0,
51
+ "documents": 7547314
52
  },
53
  "openmath": {
54
+ "page": 2018,
55
+ "row": 374,
56
  "start": {
57
  "file": 18,
58
  "reader": {
59
  "format": "parquet-row-v1",
60
+ "row_group": 406,
61
+ "row": 28
62
  }
63
  },
64
  "epoch": 0,
65
  "offset": 0,
66
+ "documents": 8266102
67
  },
68
  "ultra_qa": {
69
+ "page": 872,
70
+ "row": 2816,
71
  "start": {
72
  "file": 6,
73
  "reader": {
74
  "format": "parquet-row-v1",
75
+ "row_group": 7,
76
+ "row": 1931
77
  }
78
  },
79
  "epoch": 0,
80
  "offset": 2048,
81
+ "documents": 3574528
82
  },
83
  "ultra_style": {
84
+ "page": 1331,
85
+ "row": 2732,
86
  "start": {
87
  "file": 7,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 7,
91
+ "row": 61710
92
  }
93
  },
94
  "epoch": 0,
95
  "offset": 0,
96
+ "documents": 5454508
97
  },
98
  "cortex": {
99
+ "page": 171491,
100
+ "row": 113,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
+ "documents": 21950961
105
  }
106
  }
107
  }
latest.json CHANGED
@@ -1 +1 @@
1
- {"step": 650000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "da11afecaf8f801d2a5947a448520d9952df92bc0f9b587308168087a5a10740", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "ac22b9531603fa7bf3fc68ecd8b4c5213bb5f505ba92d8ef622eca0950b1cf9c", "training_config.json": "6c3c3ccff89dabb5098a22bc207c2f3abaee8b3bffec3f6cf6ad361a63b1a7fb", "performance.json": "5949473856ba60c6ea6e09958b880f6c16bf1db77e55259c9de9f88123633edf", "shard_scheduler.json": "55d15968e7c444b02f214615e89058b62eb3f772fd3481670f0251a16bf0c4b8", "coverage.json": "14205d54ff1f945944acaf4f3ef998712fa0fda76b995cac46c73b375d4ed339", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "33a459803c8cffa8929a90def998b4ad2c7c52fd1245365a9ee946b5552d845a", "folder": "step-000000650000"}
 
1
+ {"step": 655000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "f16c0f02e7591de5e4117964ecbe2cf4bfaf21e66738848488a66d81a023a4af", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "2417fe4e5ccee0dce50398268fdd14b4c593283295e3ff26549369405615ffe6", "training_config.json": "7234b7fffb86a474d2daeb3e14d8a78dc8439f4687eabb77722a8c9363322e29", "performance.json": "5949473856ba60c6ea6e09958b880f6c16bf1db77e55259c9de9f88123633edf", "shard_scheduler.json": "ddb75d8dad5fb16ded850552a66cff01427ff6316465608b9827b1c59db198d3", "coverage.json": "a0e4dda5a02b5917719ba5e1e749b733a37d59cd87fb879d106b7851fab0223f", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "33a459803c8cffa8929a90def998b4ad2c7c52fd1245365a9ee946b5552d845a", "folder": "step-000000655000"}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:da11afecaf8f801d2a5947a448520d9952df92bc0f9b587308168087a5a10740
3
  size 360622744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f16c0f02e7591de5e4117964ecbe2cf4bfaf21e66738848488a66d81a023a4af
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": 76930.97608111532,
26
- "required_pages_per_second_upper_estimate": 0.005322263429749181
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": 9616.372010139416,
40
- "required_pages_per_second_upper_estimate": 0.00402617414796263
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": 19232.74402027883,
54
- "required_pages_per_second_upper_estimate": 0.0040056733778564375
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": 28849.116030418245,
68
- "required_pages_per_second_upper_estimate": 0.001729460455973976
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": 28849.116030418245,
82
- "required_pages_per_second_upper_estimate": 0.0022428918775141563
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": 28849.116030418245,
96
- "required_pages_per_second_upper_estimate": 0.3700454846707744
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": 63,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 720750,
161
- "producer_batch_seconds": 0.07012926607813341,
162
- "producer_p95_seconds": 0.16191080499993404,
163
- "gpu_update_seconds": 0.9583655863440738,
164
- "producer_batches_per_second": 14.259382080027272,
165
- "gpu_batches_per_second": 1.043443143461308,
166
- "producer_headroom": 13.665701068029515,
167
  "sustainable": true,
168
- "buffer_seconds": 60.37703193967665,
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": 466217.07800612436,
174
- "upload_bytes_per_second": 1588135.0581166148,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
- "estimated_full_shards_per_hour": 100.03933196199225,
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": [
@@ -256,6 +256,246 @@
256
  "bytes": 7353219,
257
  "pages": 37,
258
  "shards": 3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
259
  }
260
  ],
261
  "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": 77686.07267619061,
26
+ "required_pages_per_second_upper_estimate": 0.005374502764261977
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": 9710.759084523826,
40
+ "required_pages_per_second_upper_estimate": 0.004065692044981137
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": 19421.518169047653,
54
+ "required_pages_per_second_upper_estimate": 0.004044990054735903
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": 29132.27725357148,
68
+ "required_pages_per_second_upper_estimate": 0.0017464355389398593
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": 29132.27725357148,
82
+ "required_pages_per_second_upper_estimate": 0.0022649064171197278
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": 29132.27725357148,
96
+ "required_pages_per_second_upper_estimate": 0.37367757280655045
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": 64,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 720750,
161
+ "producer_batch_seconds": 0.05681737146881005,
162
+ "producer_p95_seconds": 0.18071310099912807,
163
+ "gpu_update_seconds": 0.9490504212680623,
164
+ "producer_batches_per_second": 17.600251017401447,
165
+ "gpu_batches_per_second": 1.053684796497811,
166
+ "producer_headroom": 16.703525642488483,
167
  "sustainable": true,
168
+ "buffer_seconds": 60.739226961155985,
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": 354606.1178936668,
174
+ "upload_bytes_per_second": 3409288.2780103944,
175
  "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
+ "estimated_full_shards_per_hour": 76.09021809203627,
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": [
 
256
  "bytes": 7353219,
257
  "pages": 37,
258
  "shards": 3
259
+ },
260
+ {
261
+ "seconds": 1.3394989830003396,
262
+ "bytes": 6035169,
263
+ "pages": 35,
264
+ "shards": 2
265
+ },
266
+ {
267
+ "seconds": 1.5462701479991665,
268
+ "bytes": 6440094,
269
+ "pages": 21,
270
+ "shards": 2
271
+ },
272
+ {
273
+ "seconds": 1.5451833110000734,
274
+ "bytes": 7247709,
275
+ "pages": 30,
276
+ "shards": 3
277
+ },
278
+ {
279
+ "seconds": 1.4297130429995377,
280
+ "bytes": 4940477,
281
+ "pages": 36,
282
+ "shards": 2
283
+ },
284
+ {
285
+ "seconds": 1.1281410550000146,
286
+ "bytes": 6022620,
287
+ "pages": 32,
288
+ "shards": 2
289
+ },
290
+ {
291
+ "seconds": 1.545225169999867,
292
+ "bytes": 7780089,
293
+ "pages": 38,
294
+ "shards": 3
295
+ },
296
+ {
297
+ "seconds": 1.7348476629995275,
298
+ "bytes": 1091907,
299
+ "pages": 35,
300
+ "shards": 1
301
+ },
302
+ {
303
+ "seconds": 1.6013349519998883,
304
+ "bytes": 17762321,
305
+ "pages": 31,
306
+ "shards": 5
307
+ },
308
+ {
309
+ "seconds": 1.5430965340001421,
310
+ "bytes": 6138495,
311
+ "pages": 36,
312
+ "shards": 2
313
+ },
314
+ {
315
+ "seconds": 1.151366422999672,
316
+ "bytes": 1194300,
317
+ "pages": 34,
318
+ "shards": 1
319
+ },
320
+ {
321
+ "seconds": 1.6349133659996369,
322
+ "bytes": 7762216,
323
+ "pages": 38,
324
+ "shards": 3
325
+ },
326
+ {
327
+ "seconds": 1.4393719609997788,
328
+ "bytes": 9407988,
329
+ "pages": 35,
330
+ "shards": 3
331
+ },
332
+ {
333
+ "seconds": 1.460241586999473,
334
+ "bytes": 2517815,
335
+ "pages": 34,
336
+ "shards": 2
337
+ },
338
+ {
339
+ "seconds": 1.2018508789997213,
340
+ "bytes": 11749429,
341
+ "pages": 34,
342
+ "shards": 3
343
+ },
344
+ {
345
+ "seconds": 1.1237053380000361,
346
+ "bytes": 5628004,
347
+ "pages": 34,
348
+ "shards": 2
349
+ },
350
+ {
351
+ "seconds": 1.1295762649997414,
352
+ "bytes": 7499226,
353
+ "pages": 38,
354
+ "shards": 3
355
+ },
356
+ {
357
+ "seconds": 1.2592496409997693,
358
+ "bytes": 5842307,
359
+ "pages": 27,
360
+ "shards": 2
361
+ },
362
+ {
363
+ "seconds": 1.3798525270003665,
364
+ "bytes": 2740249,
365
+ "pages": 37,
366
+ "shards": 2
367
+ },
368
+ {
369
+ "seconds": 1.102272197000275,
370
+ "bytes": 5978754,
371
+ "pages": 29,
372
+ "shards": 2
373
+ },
374
+ {
375
+ "seconds": 1.5570587419997537,
376
+ "bytes": 16077242,
377
+ "pages": 41,
378
+ "shards": 4
379
+ },
380
+ {
381
+ "seconds": 1.4845605279997471,
382
+ "bytes": 2728370,
383
+ "pages": 37,
384
+ "shards": 2
385
+ },
386
+ {
387
+ "seconds": 0.968915202000062,
388
+ "bytes": 6356080,
389
+ "pages": 36,
390
+ "shards": 2
391
+ },
392
+ {
393
+ "seconds": 1.188576070999261,
394
+ "bytes": 5752447,
395
+ "pages": 32,
396
+ "shards": 2
397
+ },
398
+ {
399
+ "seconds": 1.4639023600002474,
400
+ "bytes": 7690441,
401
+ "pages": 33,
402
+ "shards": 3
403
+ },
404
+ {
405
+ "seconds": 0.9298561730001893,
406
+ "bytes": 6431657,
407
+ "pages": 36,
408
+ "shards": 2
409
+ },
410
+ {
411
+ "seconds": 1.255959656999039,
412
+ "bytes": 8183161,
413
+ "pages": 33,
414
+ "shards": 3
415
+ },
416
+ {
417
+ "seconds": 1.190589429001193,
418
+ "bytes": 6346216,
419
+ "pages": 35,
420
+ "shards": 3
421
+ },
422
+ {
423
+ "seconds": 1.0324392470010935,
424
+ "bytes": 4954714,
425
+ "pages": 37,
426
+ "shards": 2
427
+ },
428
+ {
429
+ "seconds": 1.8339393939986621,
430
+ "bytes": 7623870,
431
+ "pages": 30,
432
+ "shards": 3
433
+ },
434
+ {
435
+ "seconds": 1.1467835090006702,
436
+ "bytes": 6065504,
437
+ "pages": 37,
438
+ "shards": 2
439
+ },
440
+ {
441
+ "seconds": 1.5038798329987912,
442
+ "bytes": 7468080,
443
+ "pages": 38,
444
+ "shards": 3
445
+ },
446
+ {
447
+ "seconds": 1.3404178419987147,
448
+ "bytes": 11740523,
449
+ "pages": 31,
450
+ "shards": 3
451
+ },
452
+ {
453
+ "seconds": 1.198196088000259,
454
+ "bytes": 6150655,
455
+ "pages": 36,
456
+ "shards": 2
457
+ },
458
+ {
459
+ "seconds": 1.157564569999522,
460
+ "bytes": 2379174,
461
+ "pages": 28,
462
+ "shards": 2
463
+ },
464
+ {
465
+ "seconds": 1.1466768219997903,
466
+ "bytes": 6100233,
467
+ "pages": 35,
468
+ "shards": 2
469
+ },
470
+ {
471
+ "seconds": 1.4511324699997203,
472
+ "bytes": 6333083,
473
+ "pages": 39,
474
+ "shards": 3
475
+ },
476
+ {
477
+ "seconds": 1.0026024300004792,
478
+ "bytes": 6038075,
479
+ "pages": 36,
480
+ "shards": 2
481
+ },
482
+ {
483
+ "seconds": 1.4379317339989939,
484
+ "bytes": 11505212,
485
+ "pages": 31,
486
+ "shards": 3
487
+ },
488
+ {
489
+ "seconds": 1.3395487400011916,
490
+ "bytes": 6734300,
491
+ "pages": 31,
492
+ "shards": 3
493
+ },
494
+ {
495
+ "seconds": 0.9863095519995113,
496
+ "bytes": 6175312,
497
+ "pages": 36,
498
+ "shards": 2
499
  }
500
  ],
501
  "cache_key": {
trainer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ac22b9531603fa7bf3fc68ecd8b4c5213bb5f505ba92d8ef622eca0950b1cf9c
3
- size 721778047
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2417fe4e5ccee0dce50398268fdd14b4c593283295e3ff26549369405615ffe6
3
+ size 721760959
training_config.json CHANGED
@@ -53,8 +53,8 @@
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,
60
  "force_rebenchmark": false,
 
53
  "cpu_threads": 4,
54
  "page_rows": 4096,
55
  "cortex_page_rows": 128,
56
+ "prefetch_batches": 64,
57
+ "prefetch_low_water": 32,
58
  "memory_pages": 12,
59
  "cache_gb": 20,
60
  "force_rebenchmark": false,