appvoid commited on
Commit
6b42dd4
·
verified ·
1 Parent(s): 847f7ff

Qwen90M step 280248; complete resumable checkpoint

Browse files
Files changed (6) hide show
  1. complete.json +5 -5
  2. coverage.json +40 -40
  3. latest.json +1 -1
  4. model.safetensors +1 -1
  5. shard_scheduler.json +43 -25
  6. trainer.pt +2 -2
complete.json CHANGED
@@ -1,18 +1,18 @@
1
  {
2
- "step": 280000,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
- "model.safetensors": "b202beb2e692c902003e4624a58ed61b492689af5528e336e78027d172cdc6a5",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
- "trainer.pt": "086c04c0e606c24ef7dab8710d39cde6eff9e0a7ec67fc90f9930de401842100",
12
  "training_config.json": "95ba9e002790d11959d1191b36ed994bfa10a6da51c22c772e143bec517a9783",
13
  "performance.json": "7ab839f3b8798329051e29b39e08bd08f8678f49ed79d09e3b70b9de78a4765c",
14
- "shard_scheduler.json": "bb41f80e95e1faea1ad1243a8e7d9ee991b7eb64bf0e6b9b463722b029efa994",
15
- "coverage.json": "c7677efe249bc40dea701da173ea31fe609c61d934d175645935757dc9bffe1c",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063"
 
1
  {
2
+ "step": 280248,
3
  "files": {
4
  "config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7",
5
  "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4",
6
+ "model.safetensors": "ab3d534f71afc536f11eacb5cac49b41552e56c521fecc7b301c5b0f3ba211e7",
7
  "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03",
8
  "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300",
9
  "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263",
10
  "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6",
11
+ "trainer.pt": "93080e3d08484f8e03c44bc9ba9acf6621eb9dc2cb2663dd5f9becbb62b890ee",
12
  "training_config.json": "95ba9e002790d11959d1191b36ed994bfa10a6da51c22c772e143bec517a9783",
13
  "performance.json": "7ab839f3b8798329051e29b39e08bd08f8678f49ed79d09e3b70b9de78a4765c",
14
+ "shard_scheduler.json": "1a27a42e8cdc3d5af3957968177c27079c751995153e489c49bae7322e8344e1",
15
+ "coverage.json": "c8401a996db8071c49afdc637c6460b684b49b8ba1e5678dd9298a5b147168b1",
16
  "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"
17
  },
18
  "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063"
coverage.json CHANGED
@@ -1,58 +1,58 @@
1
  {
2
- "step": 280000,
3
  "stats": {
4
- "climbmix/blocks": 10080000,
5
- "climbmix/targets": 16850702549,
6
- "climbmix/input_tokens": 16851770299,
7
- "ultra_qa/blocks": 3780000,
8
- "ultra_qa/targets": 6166762492,
9
- "ultra_qa/input_tokens": 6166774323,
10
- "cortex/blocks": 3780000,
11
- "cortex/targets": 143784713,
12
- "cortex/input_tokens": 6008633757,
13
- "ultra_style/blocks": 3780000,
14
- "ultra_style/targets": 5978866890,
15
- "ultra_style/input_tokens": 5979181245,
16
- "openmath/blocks": 2520000,
17
- "openmath/targets": 3893894891,
18
- "openmath/input_tokens": 3895193089,
19
- "rewrite6/blocks": 1260000,
20
- "rewrite6/targets": 2065551785,
21
- "rewrite6/input_tokens": 2067816586
22
  },
23
  "sources": {
24
  "climbmix": {
25
- "page": 1380,
26
- "row": 1360,
27
  "start": {
28
  "file": 66,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
- "row_group": 66,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
  "offset": 2048,
37
- "documents": 5653840
38
  },
39
  "rewrite6": {
40
- "page": 35,
41
- "row": 2118,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
- "row": 143360
47
  }
48
  },
49
  "epoch": 15,
50
  "offset": 0,
51
- "documents": 3226538
52
  },
53
  "openmath": {
54
  "page": 863,
55
- "row": 65,
56
  "start": {
57
  "file": 8,
58
  "reader": {
@@ -62,12 +62,12 @@
62
  }
63
  },
64
  "epoch": 0,
65
- "offset": 0,
66
- "documents": 3534913
67
  },
68
  "ultra_qa": {
69
  "page": 373,
70
- "row": 122,
71
  "start": {
72
  "file": 2,
73
  "reader": {
@@ -77,31 +77,31 @@
77
  }
78
  },
79
  "epoch": 0,
80
- "offset": 4096,
81
- "documents": 1527930
82
  },
83
  "ultra_style": {
84
- "page": 569,
85
- "row": 3492,
86
  "start": {
87
  "file": 3,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 3,
91
- "row": 18427
92
  }
93
  },
94
  "epoch": 0,
95
  "offset": 2048,
96
- "documents": 2334116
97
  },
98
  "cortex": {
99
- "page": 73309,
100
- "row": 69,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
- "documents": 9383621
105
  }
106
  }
107
  }
 
1
  {
2
+ "step": 280248,
3
  "stats": {
4
+ "climbmix/blocks": 10088928,
5
+ "climbmix/targets": 16865599592,
6
+ "climbmix/input_tokens": 16866668333,
7
+ "ultra_qa/blocks": 3783348,
8
+ "ultra_qa/targets": 6172241217,
9
+ "ultra_qa/input_tokens": 6172253060,
10
+ "cortex/blocks": 3783348,
11
+ "cortex/targets": 143908822,
12
+ "cortex/input_tokens": 6013958320,
13
+ "ultra_style/blocks": 3783348,
14
+ "ultra_style/targets": 5984022238,
15
+ "ultra_style/input_tokens": 5984336893,
16
+ "openmath/blocks": 2522232,
17
+ "openmath/targets": 3897349251,
18
+ "openmath/input_tokens": 3898648602,
19
+ "rewrite6/blocks": 1261116,
20
+ "rewrite6/targets": 2067382557,
21
+ "rewrite6/input_tokens": 2069649291
22
  },
23
  "sources": {
24
  "climbmix": {
25
+ "page": 1381,
26
+ "row": 2390,
27
  "start": {
28
  "file": 66,
29
  "reader": {
30
  "format": "parquet-row-v1",
31
+ "row_group": 70,
32
  "row": 1024
33
  }
34
  },
35
  "epoch": 0,
36
  "offset": 2048,
37
+ "documents": 5658966
38
  },
39
  "rewrite6": {
40
+ "page": 36,
41
+ "row": 797,
42
  "start": {
43
  "file": 0,
44
  "reader": {
45
  "format": "jsonl-row-v1",
46
+ "row": 147456
47
  }
48
  },
49
  "epoch": 15,
50
  "offset": 0,
51
+ "documents": 3229313
52
  },
53
  "openmath": {
54
  "page": 863,
55
+ "row": 3211,
56
  "start": {
57
  "file": 8,
58
  "reader": {
 
62
  }
63
  },
64
  "epoch": 0,
65
+ "offset": 2048,
66
+ "documents": 3538059
67
  },
68
  "ultra_qa": {
69
  "page": 373,
70
+ "row": 1465,
71
  "start": {
72
  "file": 2,
73
  "reader": {
 
77
  }
78
  },
79
  "epoch": 0,
80
+ "offset": 2048,
81
+ "documents": 1529273
82
  },
83
  "ultra_style": {
84
+ "page": 570,
85
+ "row": 1617,
86
  "start": {
87
  "file": 3,
88
  "reader": {
89
  "format": "parquet-row-v1",
90
  "row_group": 3,
91
+ "row": 22523
92
  }
93
  },
94
  "epoch": 0,
95
  "offset": 2048,
96
+ "documents": 2336337
97
  },
98
  "cortex": {
99
+ "page": 73374,
100
+ "row": 54,
101
  "start": null,
102
  "epoch": 0,
103
  "offset": 0,
104
+ "documents": 9391926
105
  }
106
  }
107
  }
latest.json CHANGED
@@ -1 +1 @@
1
- {"step": 280000, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "b202beb2e692c902003e4624a58ed61b492689af5528e336e78027d172cdc6a5", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "086c04c0e606c24ef7dab8710d39cde6eff9e0a7ec67fc90f9930de401842100", "training_config.json": "95ba9e002790d11959d1191b36ed994bfa10a6da51c22c772e143bec517a9783", "performance.json": "7ab839f3b8798329051e29b39e08bd08f8678f49ed79d09e3b70b9de78a4765c", "shard_scheduler.json": "bb41f80e95e1faea1ad1243a8e7d9ee991b7eb64bf0e6b9b463722b029efa994", "coverage.json": "c7677efe249bc40dea701da173ea31fe609c61d934d175645935757dc9bffe1c", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063", "folder": "step-000000280000"}
 
1
+ {"step": 280248, "files": {"config.json": "cbc72137da3dae65085a2a8c39ad7634d16bf92d785ca00d283ee82fe32b37e7", "generation_config.json": "54ed0d19f6226888c91157964ede8a892bd5ae175bfad85336152b3d9b6cd7c4", "model.safetensors": "ab3d534f71afc536f11eacb5cac49b41552e56c521fecc7b301c5b0f3ba211e7", "tokenization_byte.py": "279b9990f77b1eaedc7c1184fe13d60273a713ead856200f1fb9ae323c7b1e03", "tokenizer_config.json": "0ecbd4fd39e35125800d806af981ba0cacd8a7ba5f02c9e729c4ecc54af6d300", "special_tokens_map.json": "20d4b94d0e657c3f44d9123522c106e7315b6c1a1b6d259478f17bd2918b1263", "byte_vocab.json": "57a1cd2e70dd5b7cc6fd2eaa12ebfcd173c1620fa5d45c2c6282511c29b6a5f6", "trainer.pt": "93080e3d08484f8e03c44bc9ba9acf6621eb9dc2cb2663dd5f9becbb62b890ee", "training_config.json": "95ba9e002790d11959d1191b36ed994bfa10a6da51c22c772e143bec517a9783", "performance.json": "7ab839f3b8798329051e29b39e08bd08f8678f49ed79d09e3b70b9de78a4765c", "shard_scheduler.json": "1a27a42e8cdc3d5af3957968177c27079c751995153e489c49bae7322e8344e1", "coverage.json": "c8401a996db8071c49afdc637c6460b684b49b8ba1e5678dd9298a5b147168b1", "README.md": "9ce0f9efacbc46177a7e818b6aed274d84e7145ff944b5c7d6ecd5d33b9d4419"}, "signature": "9fee47918905b0750bee87876aa424f021ccc4507ee8db22a11d01fbc6109063", "folder": "step-000000280248"}
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b202beb2e692c902003e4624a58ed61b492689af5528e336e78027d172cdc6a5
3
  size 360622744
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab3d534f71afc536f11eacb5cac49b41552e56c521fecc7b301c5b0f3ba211e7
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": 75842.57462698415,
26
- "required_pages_per_second_upper_estimate": 0.006573842147418966
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": 9480.321828373018,
40
- "required_pages_per_second_upper_estimate": 0.0037818421207806837
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": 18960.643656746037,
54
- "required_pages_per_second_upper_estimate": 0.0043862112140569684
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": 28440.965485119053,
68
- "required_pages_per_second_upper_estimate": 0.0017319055798300411
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": 28440.965485119053,
82
- "required_pages_per_second_upper_estimate": 0.002189464252675848
83
  },
84
  {
85
  "source": "cortex",
@@ -92,8 +92,8 @@
92
  "tokens_per_second": 1015177.8395898147,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
  "estimated_pages_for_buffer": 22,
95
- "required_token_positions_per_second": 28440.965485119053,
96
- "required_pages_per_second_upper_estimate": 0.3531416055368222
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
@@ -158,24 +158,24 @@
158
  "prefetch_batches": 62,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 690670,
161
- "producer_batch_seconds": 0.080577971249987,
162
- "producer_p95_seconds": 0.18257642799926543,
163
- "gpu_update_seconds": 0.9721188971051651,
164
- "producer_batches_per_second": 12.410339755236285,
165
- "gpu_batches_per_second": 1.028680753946725,
166
- "producer_headroom": 12.06432579556068,
167
  "sustainable": true,
168
- "buffer_seconds": 60.271371620520235,
169
  "buffer_drain_seconds": null,
170
- "shard_target_mib": 32,
171
- "upload_pages_per_commit": 160,
172
  "max_pending_pages": 4096,
173
- "generation_bytes_per_second": 905156.4916824569,
174
- "upload_bytes_per_second": 607793.4295288165,
175
- "upload_sustainable": false,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
- "estimated_full_shards_per_hour": 97.11275607516899,
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": [
@@ -190,6 +190,24 @@
190
  "bytes": 935968,
191
  "pages": 27,
192
  "shards": 1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
193
  }
194
  ],
195
  "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": 75609.08177184862,
26
+ "required_pages_per_second_upper_estimate": 0.006553603578518049
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": 9451.135221481078,
40
+ "required_pages_per_second_upper_estimate": 0.0037701991469128284
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": 18902.270442962155,
54
+ "required_pages_per_second_upper_estimate": 0.004372707598381542
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": 28353.405664443235,
68
+ "required_pages_per_second_upper_estimate": 0.0017265736461417594
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": 28353.405664443235,
82
+ "required_pages_per_second_upper_estimate": 0.0021827236552991267
83
  },
84
  {
85
  "source": "cortex",
 
92
  "tokens_per_second": 1015177.8395898147,
93
  "note": "Bounded fresh sample; full-page latency is an extrapolation.",
94
  "estimated_pages_for_buffer": 22,
95
+ "required_token_positions_per_second": 28353.405664443235,
96
+ "required_pages_per_second_upper_estimate": 0.3520544056078974
97
  }
98
  ],
99
  "gpu_measurement": "Selected forward/backward estimate; replaced by live successful update timings.",
 
158
  "prefetch_batches": 62,
159
  "prefetch_capacity_limit": 512,
160
  "batch_memory_estimate_bytes": 690670,
161
+ "producer_batch_seconds": 0.07287284228125657,
162
+ "producer_p95_seconds": 0.1761082670000178,
163
+ "gpu_update_seconds": 0.9751209546820736,
164
+ "producer_batches_per_second": 13.722533233168638,
165
+ "gpu_batches_per_second": 1.0255138044141794,
166
+ "producer_headroom": 13.381129706983886,
167
  "sustainable": true,
168
+ "buffer_seconds": 60.457499190288566,
169
  "buffer_drain_seconds": null,
170
+ "shard_target_mib": 16,
171
+ "upload_pages_per_commit": 79,
172
  "max_pending_pages": 4096,
173
+ "generation_bytes_per_second": 382834.16564839584,
174
+ "upload_bytes_per_second": 4501863.534696111,
175
+ "upload_sustainable": true,
176
  "logical_page_rows": 4096,
177
  "logical_cortex_page_rows": 128,
178
+ "estimated_full_shards_per_hour": 82.14730002488047,
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": [
 
190
  "bytes": 935968,
191
  "pages": 27,
192
  "shards": 1
193
+ },
194
+ {
195
+ "seconds": 1.5188891550005792,
196
+ "bytes": 5974366,
197
+ "pages": 29,
198
+ "shards": 2
199
+ },
200
+ {
201
+ "seconds": 1.1771950970005491,
202
+ "bytes": 7732243,
203
+ "pages": 17,
204
+ "shards": 3
205
+ },
206
+ {
207
+ "seconds": 1.3709215139997468,
208
+ "bytes": 10777643,
209
+ "pages": 39,
210
+ "shards": 3
211
  }
212
  ],
213
  "cache_key": {
trainer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:086c04c0e606c24ef7dab8710d39cde6eff9e0a7ec67fc90f9930de401842100
3
- size 721763967
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93080e3d08484f8e03c44bc9ba9acf6621eb9dc2cb2663dd5f9becbb62b890ee
3
+ size 721763455