Dataset Preview
Duplicate
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed because of a cast error
Error code:   DatasetGenerationCastError
Exception:    DatasetGenerationCastError
Message:      An error occurred while generating the dataset

All the data files must have the same columns, but at some point there are 4 new columns ({'has_base_model', 'has_model_card', 'id', 'downloads'}) and 19 missing columns ({'base_id', 'timestamp', 'aratio', 'ztimestamp', 'tratio', 'zbytes_out', 'pred_ratio', 'ratio', 'tbytes_out', 'bytes_out', 'shape', 'atimestamp', 'bytes_in', 'zratio', 'dist', 'ttimestamp', 'abytes_out', 'param_name', 'target_id'}).

This happened while the csv dataset builder was generating data using

hf://datasets/tensordex/tensordex-ae-cache/model_hub_crawl/model_snapshot_merged_last_month.csv (at revision fcefa9e4025810518875dc7b3607c25012e923ca), ['hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/compression_data/real_compression_all_models.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/model_hub_crawl/model_snapshot_merged_last_month.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/model_level_reduction/trace_global_flexsplit.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/model_level_reduction/trace_global_zipllm.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/sample_pairs.tsv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/tests/output/algo_benchmark/logs/bench_cached_facility_parsed.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/tests/output/compare_methods/zipllm_all_models.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/tests/output/openzl_benchmark.csv']

Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1837, in _prepare_split_single
                  writer.write_table(table)
                  ~~~~~~~~~~~~~~~~~~^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
                  self._write_table(pa_table, writer_batch_size=writer_batch_size)
                  ~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
                  pa_table = table_cast(pa_table, self._schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              id: string
              has_model_card: string
              has_base_model: string
              downloads: double
              -- schema metadata --
              pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 760
              to
              {'target_id': Value('string'), 'base_id': Value('string'), 'param_name': Value('string'), 'shape': Value('string'), 'bytes_in': Value('int64'), 'dist': Value('float64'), 'pred_ratio': Value('float64'), 'bytes_out': Value('float64'), 'ratio': Value('float64'), 'timestamp': Value('string'), 'zbytes_out': Value('float64'), 'zratio': Value('float64'), 'ztimestamp': Value('string'), 'abytes_out': Value('int64'), 'aratio': Value('float64'), 'atimestamp': Value('string'), 'tbytes_out': Value('float64'), 'tratio': Value('float64'), 'ttimestamp': Value('string')}
              because column names don't match
              
              During handling of the above exception, another exception occurred:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
                  parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
                                                                        ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      builder, max_dataset_size_bytes=max_dataset_size_bytes
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
                  builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
                  ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1683, in _prepare_split
                  for job_id, done, content in self._prepare_split_single(
                                               ~~~~~~~~~~~~~~~~~~~~~~~~~~^
                      gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  ):
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1839, in _prepare_split_single
                  raise DatasetGenerationCastError.from_cast_error(
                  ...<4 lines>...
                  )
              datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
              
              All the data files must have the same columns, but at some point there are 4 new columns ({'has_base_model', 'has_model_card', 'id', 'downloads'}) and 19 missing columns ({'base_id', 'timestamp', 'aratio', 'ztimestamp', 'tratio', 'zbytes_out', 'pred_ratio', 'ratio', 'tbytes_out', 'bytes_out', 'shape', 'atimestamp', 'bytes_in', 'zratio', 'dist', 'ttimestamp', 'abytes_out', 'param_name', 'target_id'}).
              
              This happened while the csv dataset builder was generating data using
              
              hf://datasets/tensordex/tensordex-ae-cache/model_hub_crawl/model_snapshot_merged_last_month.csv (at revision fcefa9e4025810518875dc7b3607c25012e923ca), ['hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/compression_data/real_compression_all_models.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/model_hub_crawl/model_snapshot_merged_last_month.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/model_level_reduction/trace_global_flexsplit.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/model_level_reduction/trace_global_zipllm.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/sample_pairs.tsv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/tests/output/algo_benchmark/logs/bench_cached_facility_parsed.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/tests/output/compare_methods/zipllm_all_models.csv', 'hf://datasets/tensordex/tensordex-ae-cache@fcefa9e4025810518875dc7b3607c25012e923ca/tests/output/openzl_benchmark.csv']
              
              Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

target_id
string
base_id
string
param_name
string
shape
string
bytes_in
int64
dist
float64
pred_ratio
float64
bytes_out
float64
ratio
float64
timestamp
string
zbytes_out
float64
zratio
float64
ztimestamp
string
abytes_out
int64
aratio
float64
atimestamp
string
tbytes_out
float64
tratio
float64
ttimestamp
string
3248a0f161d6937f8c3f07db208b24de
194960234f8ebfb7fc7ae25eb31ee1bb
model.embed_tokens.weight
32000x4096
262,144,000
0.004926
0.650411
211,927,731
0.80844
2025-12-04T21:05:03.483697
195,096,504
0.744234
2026-02-19T16:56:07.689881
186,294,406
0.710657
2026-03-04T05:28:48.850238
185,553,846
0.707832
2026-03-28T08:07:53.354419
5ea4a1ab8b0a1aaae65c1ca2a85497f6
c8779cf20be6da057db20875ab9c6774
model.layers.16.post_attention_layernorm.weight
(4096,)
8,192
null
0.679419
null
null
null
2,496
0.304688
2026-03-02T02:12:54.814017
3,424
0.417969
2026-03-04T05:29:01.595867
4,234
0.516846
2026-03-28T08:08:09.619902
443a2d256a8a7f3645dc647951a850fe
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.346427
4,286
0.523193
2026-03-03T19:35:38.161811
4,162
0.508057
2026-03-02T02:12:39.548724
4,334
0.529053
2026-03-04T05:28:48.850238
4,123
0.503296
2026-03-28T08:07:53.354419
4a6e08467dffa729ee9428429cc35a8e
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.379075
4,683
0.571655
2026-03-03T19:35:38.161811
4,437
0.541626
2026-03-02T02:12:39.548724
4,722
0.576416
2026-03-04T05:28:48.850238
4,560
0.556641
2026-03-28T08:07:53.354419
35aae388080d850874eb51657eee4371
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.296857
3,957
0.483032
2026-03-03T19:35:38.161811
3,775
0.460815
2026-03-02T02:12:39.548724
3,945
0.481567
2026-03-04T05:28:48.850238
3,784
0.461914
2026-03-28T08:07:53.354419
6ddf3bcc186c47904ea606eb28c17ee1
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.356536
4,446
0.542725
2026-03-03T19:35:38.161811
4,345
0.530396
2026-03-02T02:12:39.548724
4,553
0.555786
2026-03-04T05:28:48.850238
4,336
0.529297
2026-03-28T08:07:53.354419
362ed56a7e25b4435f4de07b2c8694f4
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.461052
5,285
0.645142
2026-03-03T19:35:38.161811
5,126
0.625732
2026-03-02T02:12:39.548724
5,505
0.671997
2026-03-04T05:28:48.850238
5,188
0.633301
2026-03-28T08:07:53.354419
47e06499bb3c4cf330fd602ad7a63f08
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.381408
4,455
0.543823
2026-03-03T19:35:38.161811
4,487
0.547729
2026-03-02T02:12:39.548724
4,595
0.560913
2026-03-04T05:28:48.850238
4,308
0.525879
2026-03-28T08:07:53.354419
e6baa62b488d6439818ec4307f67185e
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.347534
4,432
0.541016
2026-03-03T19:35:38.161811
4,446
0.542725
2026-03-02T02:12:39.548724
4,572
0.558105
2026-03-04T05:28:48.850238
4,285
0.523071
2026-03-28T08:07:53.354419
f99dc1fc55018b117e26c080bf182048
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.418577
4,753
0.5802
2026-03-03T19:35:38.161811
4,568
0.557617
2026-03-02T02:12:39.548724
4,873
0.594849
2026-03-04T05:28:48.850238
4,673
0.570435
2026-03-28T08:07:53.354419
b18b26dd32e5a90bec78d7ae4bae60be
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.347073
4,368
0.533203
2026-03-03T19:35:38.161811
4,367
0.533081
2026-03-02T02:12:39.548724
4,491
0.548218
2026-03-04T05:28:48.850238
4,252
0.519043
2026-03-28T08:07:53.354419
814f4ecb2c3d12deb5530c9895dde408
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.323864
4,206
0.513428
2026-03-03T19:35:38.161811
4,121
0.503052
2026-03-02T02:12:39.548724
4,278
0.522217
2026-03-04T05:28:48.850238
4,046
0.493896
2026-03-28T08:07:53.354419
d74cbedcfabccc8c6ba2a3009acc8603
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.329859
4,124
0.503418
2026-03-03T19:35:38.161811
3,994
0.487549
2026-03-02T02:12:39.548724
4,147
0.506226
2026-03-04T05:28:48.850238
3,958
0.483154
2026-03-28T08:07:53.354419
a0b1c2d53a5c70d2f24718fce2f55694
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.371927
4,419
0.539429
2026-03-03T19:35:38.161811
4,333
0.528931
2026-03-02T02:12:39.548724
4,496
0.548828
2026-03-04T05:28:48.850238
4,288
0.523438
2026-03-28T08:07:53.354419
fe6328b965b858e9744512626fd651eb
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.466475
4,915
0.599976
2026-03-03T19:35:38.161811
4,709
0.574829
2026-03-02T02:12:39.548724
4,958
0.605225
2026-03-04T05:28:48.850238
4,797
0.585571
2026-03-28T08:07:53.354419
f398f52a2be2012565188aff1b9ea9f5
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.377053
4,721
0.576294
2026-03-03T19:35:38.161811
4,719
0.57605
2026-03-02T02:12:39.548724
4,844
0.591309
2026-03-04T05:28:48.850238
4,586
0.559814
2026-03-28T08:07:53.354419
5a4d383f82e8f49d46332b885759d3ce
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.368266
4,658
0.568604
2026-03-03T19:35:38.161811
4,633
0.565552
2026-03-02T02:12:39.548724
4,766
0.581787
2026-03-04T05:28:48.850238
4,547
0.555054
2026-03-28T08:07:53.354419
8ba52f2ed5c69e1f6fac645a6298cd96
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.364453
4,649
0.567505
2026-03-03T19:35:38.161811
4,675
0.570679
2026-03-02T02:12:39.548724
4,769
0.582153
2026-03-04T05:28:48.850238
4,478
0.546631
2026-03-28T08:07:53.354419
5b66f354f7976a8baacdea5b8deeefc6
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.37289
4,648
0.567383
2026-03-03T19:35:38.161811
4,611
0.562866
2026-03-02T02:12:39.548724
4,729
0.577271
2026-03-04T05:28:48.850238
4,452
0.543457
2026-03-28T08:07:53.354419
d8ca594960eabd5ce130450fc6c485b8
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.34932
4,478
0.546631
2026-03-03T19:35:38.161811
4,400
0.537109
2026-03-02T02:12:39.548724
4,505
0.549927
2026-03-04T05:28:48.850238
4,291
0.523804
2026-03-28T08:07:53.354419
e453bed5cfea4a67f9883b63c6666797
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.472567
5,027
0.613647
2026-03-03T19:35:38.161811
4,811
0.58728
2026-03-02T02:12:39.548724
5,111
0.623901
2026-03-04T05:28:48.850238
4,950
0.604248
2026-03-28T08:07:53.354419
7fd560b6ba10967195f6e95e00e69d0e
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.493496
5,047
0.616089
2026-03-03T19:35:38.161811
4,805
0.586548
2026-03-02T02:12:39.548724
5,097
0.622192
2026-03-04T05:28:48.850238
4,949
0.604126
2026-03-28T08:07:53.354419
73db0cff9009184e236a56576c8e29f4
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.476671
5,377
0.656372
2026-03-03T19:35:38.161811
5,199
0.634644
2026-03-02T02:12:39.548724
5,565
0.679321
2026-03-04T05:28:48.850238
5,277
0.644165
2026-03-28T08:07:53.354419
dd3788cb43343ac87e5a585411d213e3
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.546266
5,794
0.707275
2026-03-03T19:35:38.161811
5,827
0.711304
2026-03-02T02:12:39.548724
6,133
0.748657
2026-03-04T05:28:48.850238
5,802
0.708252
2026-03-28T08:07:53.354419
ac19f75592df40f9a2a8a4e61e61ff03
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.483473
5,576
0.680664
2026-03-03T19:35:38.161811
5,491
0.670288
2026-03-02T02:12:39.548724
5,878
0.717529
2026-03-04T05:28:48.850238
5,584
0.681641
2026-03-28T08:07:53.354419
dc65ce228465fa7c4780456fc9cdc6b9
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.537036
5,663
0.691284
2026-03-03T19:35:38.161811
5,640
0.688477
2026-03-02T02:12:39.548724
5,995
0.731812
2026-03-04T05:28:48.850238
5,672
0.692383
2026-03-28T08:07:53.354419
cd636c33a2a15f1360ce310243ac3f1d
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.553781
5,723
0.698608
2026-03-03T19:35:38.161811
5,725
0.698853
2026-03-02T02:12:39.548724
6,053
0.738892
2026-03-04T05:28:48.850238
5,733
0.699829
2026-03-28T08:07:53.354419
6bbcd86fae5943448cfbce53e01e5b4f
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.447907
4,961
0.605591
2026-03-03T19:35:38.161811
4,828
0.589355
2026-03-02T02:12:39.548724
5,164
0.630371
2026-03-04T05:28:48.850238
4,923
0.600952
2026-03-28T08:07:53.354419
400a1334374483b647ecfac9cabb75fd
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.360615
4,439
0.54187
2026-03-03T19:35:38.161811
4,330
0.528564
2026-03-02T02:12:39.548724
4,550
0.55542
2026-03-04T05:28:48.850238
4,331
0.528687
2026-03-28T08:07:53.354419
14c570028ee9dd81c63d200650a4274a
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.374381
4,468
0.54541
2026-03-03T19:35:38.161811
4,425
0.540161
2026-03-02T02:12:39.548724
4,631
0.565308
2026-03-04T05:28:48.850238
4,389
0.535767
2026-03-28T08:07:53.354419
cc2b23dc1ad9c1186aba4e4c5d978c76
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.385986
4,511
0.550659
2026-03-03T19:35:38.161811
4,435
0.541382
2026-03-02T02:12:39.548724
4,652
0.567871
2026-03-04T05:28:48.850238
4,409
0.538208
2026-03-28T08:07:53.354419
96ceb78ee2f72921232b3c8bd3a9be08
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.359255
4,241
0.5177
2026-03-03T19:35:38.161811
4,113
0.502075
2026-03-02T02:12:39.548724
4,280
0.522461
2026-03-04T05:28:48.850238
4,064
0.496094
2026-03-28T08:07:53.354419
ab801770ee153643bb05076f2b05eea0
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.36306
4,392
0.536133
2026-03-03T19:35:38.161811
4,277
0.522095
2026-03-02T02:12:39.548724
4,480
0.546875
2026-03-04T05:28:48.850238
4,296
0.524414
2026-03-28T08:07:53.354419
dc5451caf4088fd90e8703e116c80435
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.362441
4,493
0.548462
2026-03-03T19:35:38.161811
4,402
0.537354
2026-03-02T02:12:39.548724
4,602
0.561768
2026-03-04T05:28:48.850238
4,346
0.530518
2026-03-28T08:07:53.354419
7e136a5b9106ce92607bac4cb4e6bf31
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.343231
4,387
0.535522
2026-03-03T19:35:38.161811
4,291
0.523804
2026-03-02T02:12:39.548724
4,487
0.547729
2026-03-04T05:28:48.850238
4,239
0.517456
2026-03-28T08:07:53.354419
349b3f4612a816b9fd117c58d5ef025d
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.380942
4,519
0.551636
2026-03-03T19:35:38.161811
4,421
0.539673
2026-03-02T02:12:39.548724
4,641
0.566528
2026-03-04T05:28:48.850238
4,400
0.537109
2026-03-28T08:07:53.354419
f02ab927e46471dfaff648f18d3b92aa
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.372362
4,454
0.543701
2026-03-03T19:35:38.161811
4,392
0.536133
2026-03-02T02:12:39.548724
4,591
0.560425
2026-03-04T05:28:48.850238
4,325
0.527954
2026-03-28T08:07:53.354419
6ebe196e6bbf175a94b132cf189b5bd8
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.376494
4,489
0.547974
2026-03-03T19:35:38.161811
4,395
0.536499
2026-03-02T02:12:39.548724
4,586
0.559814
2026-03-04T05:28:48.850238
4,350
0.531006
2026-03-28T08:07:53.354419
19e73399b9824fef66a7cd8af900796c
014f5ccb342104fe99a81a5c6906ff55
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.551807
5,677
0.692993
2026-03-03T19:35:38.161811
5,519
0.673706
2026-03-02T02:12:39.548724
5,860
0.715332
2026-03-04T05:28:48.850238
5,722
0.698486
2026-03-28T08:07:53.354419
8d117f20f15a2881c2f053c8cfa4245d
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.360677
4,518
0.551514
2026-03-03T19:35:38.161811
4,400
0.537109
2026-03-02T02:12:39.548724
4,642
0.56665
2026-03-04T05:28:48.850238
4,423
0.539917
2026-03-28T08:07:53.354419
573e8976b90e24e8841cdb95863fc84d
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.381284
4,695
0.57312
2026-03-03T19:35:38.161811
4,579
0.55896
2026-03-02T02:12:39.548724
4,847
0.591675
2026-03-04T05:28:48.850238
4,609
0.562622
2026-03-28T08:07:53.354419
fb1fbd2e223bd48ee8cb867d1fdc6f90
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.343631
4,287
0.523315
2026-03-03T19:35:38.161811
4,159
0.50769
2026-03-02T02:12:39.548724
4,340
0.529785
2026-03-04T05:28:48.850238
4,122
0.503174
2026-03-28T08:07:53.354419
3895ea74d82d09c0d61da0764e809b09
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.29489
4,048
0.494141
2026-03-03T19:35:38.161811
3,838
0.468506
2026-03-02T02:12:39.548724
4,042
0.493408
2026-03-04T05:28:48.850238
3,861
0.471313
2026-03-28T08:07:53.354419
eae9cd66a856368592519f1f5267b6fd
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.352863
4,363
0.532593
2026-03-03T19:35:38.161811
4,242
0.517822
2026-03-02T02:12:39.548724
4,454
0.543701
2026-03-04T05:28:48.850238
4,261
0.520142
2026-03-28T08:07:53.354419
4107cd8787e3864b4b6de60d8fb5ead2
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.367087
4,525
0.552368
2026-03-03T19:35:38.161811
4,420
0.539551
2026-03-02T02:12:39.548724
4,629
0.565063
2026-03-04T05:28:48.850238
4,438
0.541748
2026-03-28T08:07:53.354419
22705dcb33d25df60e8bc6bb0bee01b9
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.378888
4,513
0.550903
2026-03-03T19:35:38.161811
4,392
0.536133
2026-03-02T02:12:39.548724
4,630
0.565186
2026-03-04T05:28:48.850238
4,407
0.537964
2026-03-28T08:07:53.354419
e153e536e4cd49bd1764b7f5c2c79cea
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.374598
4,527
0.552612
2026-03-03T19:35:38.161811
4,388
0.535645
2026-03-02T02:12:39.548724
4,615
0.563354
2026-03-04T05:28:48.850238
4,415
0.53894
2026-03-28T08:07:53.354419
684e5a7a12d45dcbc89deecc926a67fa
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.361482
4,505
0.549927
2026-03-03T19:35:38.161811
4,354
0.531494
2026-03-02T02:12:39.548724
4,602
0.561768
2026-03-04T05:28:48.850238
4,395
0.536499
2026-03-28T08:07:53.354419
cb80c34730e98b7eda93b536c587f191
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.374101
4,484
0.547363
2026-03-03T19:35:38.161811
4,363
0.532593
2026-03-02T02:12:39.548724
4,596
0.561035
2026-03-04T05:28:48.850238
4,377
0.534302
2026-03-28T08:07:53.354419
feb16d9be4a1cbbe2c1e483212567833
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.365197
4,513
0.550903
2026-03-03T19:35:38.161811
4,407
0.537964
2026-03-02T02:12:39.548724
4,629
0.565063
2026-03-04T05:28:48.850238
4,398
0.536865
2026-03-28T08:07:53.354419
16d20d2031bff2661ca6c2b217cf9026
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.362286
4,545
0.55481
2026-03-03T19:35:38.161811
4,416
0.539063
2026-03-02T02:12:39.548724
4,655
0.568237
2026-03-04T05:28:48.850238
4,442
0.542236
2026-03-28T08:07:53.354419
5d2c410622b9389e07f26b6e0477db43
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.376307
4,560
0.556641
2026-03-03T19:35:38.161811
4,458
0.544189
2026-03-02T02:12:39.548724
4,701
0.573853
2026-03-04T05:28:48.850238
4,471
0.545776
2026-03-28T08:07:53.354419
0e445ee4c53e59c4039bb6e4d3ff749e
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.373387
4,518
0.551514
2026-03-03T19:35:38.161811
4,384
0.535156
2026-03-02T02:12:39.548724
4,628
0.564941
2026-03-04T05:28:48.850238
4,404
0.537598
2026-03-28T08:07:53.354419
5d4db94b5e67198bc17ee6384d039c1a
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.366034
4,538
0.553955
2026-03-03T19:35:38.161811
4,388
0.535645
2026-03-02T02:12:39.548724
4,607
0.562378
2026-03-04T05:28:48.850238
4,413
0.538696
2026-03-28T08:07:53.354419
81a216c1192dd30b3f086ae03d1e7e88
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.368452
4,518
0.551514
2026-03-03T19:35:38.161811
4,436
0.541504
2026-03-02T02:12:39.548724
4,648
0.567383
2026-03-04T05:28:48.850238
4,416
0.539063
2026-03-28T08:07:53.354419
fd7d78cec56a1dc21db73043a1812051
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.367397
4,511
0.550659
2026-03-03T19:35:38.161811
4,381
0.53479
2026-03-02T02:12:39.548724
4,612
0.562988
2026-03-04T05:28:48.850238
4,422
0.539795
2026-03-28T08:07:53.354419
f33c36eca9845b06253eb1a6ee45201b
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.358884
4,501
0.549438
2026-03-03T19:35:38.161811
4,377
0.534302
2026-03-02T02:12:39.548724
4,606
0.562256
2026-03-04T05:28:48.850238
4,403
0.537476
2026-03-28T08:07:53.354419
e307d1156e17bda7a958bd37128c4eec
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.373324
4,507
0.550171
2026-03-03T19:35:38.161811
4,474
0.546143
2026-03-02T02:12:39.548724
4,633
0.565552
2026-03-04T05:28:48.850238
4,356
0.531738
2026-03-28T08:07:53.354419
d8b05e516ee7e66e054a98202a2f0486
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.370623
4,532
0.553223
2026-03-03T19:35:38.161811
4,409
0.538208
2026-03-02T02:12:39.548724
4,631
0.565308
2026-03-04T05:28:48.850238
4,420
0.539551
2026-03-28T08:07:53.354419
b80907a36c0f4642f80f1b7e21d0abcd
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.319828
4,084
0.498535
2026-03-03T19:35:38.161811
3,891
0.474976
2026-03-02T02:12:39.548724
4,112
0.501953
2026-03-04T05:28:48.850238
3,935
0.480347
2026-03-28T08:07:53.354419
dc97499d4f7d78dce95bb6eff53bb5fd
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.331842
4,343
0.530151
2026-03-03T19:35:38.161811
4,194
0.511963
2026-03-02T02:12:39.548724
4,414
0.538818
2026-03-04T05:28:48.850238
4,233
0.516724
2026-03-28T08:07:53.354419
2fd7defeb5d7f9f66a98fdd621f44a86
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.279216
3,874
0.4729
2026-03-03T19:35:38.161811
3,631
0.443237
2026-03-02T02:12:39.548724
3,873
0.472778
2026-03-04T05:28:48.850238
3,706
0.452393
2026-03-28T08:07:53.354419
f68de2f77ce5a8da9e6b28712b6ccfaa
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.231196
3,372
0.411621
2026-03-03T19:35:38.161811
3,118
0.380615
2026-03-02T02:12:39.548724
3,395
0.414429
2026-03-04T05:28:48.850238
3,204
0.391113
2026-03-28T08:07:53.354419
a65a886e0a8ed51f8bc3057089b41bf5
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.253076
3,598
0.439209
2026-03-03T19:35:38.161811
3,360
0.410156
2026-03-02T02:12:39.548724
3,620
0.441895
2026-03-04T05:28:48.850238
3,462
0.422607
2026-03-28T08:07:53.354419
afd9a2731e9396ac85bcf7367543bf78
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.292838
3,986
0.486572
2026-03-03T19:35:38.161811
3,771
0.460327
2026-03-02T02:12:39.548724
3,986
0.486572
2026-03-04T05:28:48.850238
3,821
0.466431
2026-03-28T08:07:53.354419
b737650e230906c1057c954f207f7abc
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.340501
4,353
0.531372
2026-03-03T19:35:38.161811
4,179
0.510132
2026-03-02T02:12:39.548724
4,397
0.536743
2026-03-04T05:28:48.850238
4,219
0.515015
2026-03-28T08:07:53.354419
f3d76f03fb7135304a04590368e21fe5
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.374536
4,497
0.54895
2026-03-03T19:35:38.161811
4,383
0.535034
2026-03-02T02:12:39.548724
4,615
0.563354
2026-03-04T05:28:48.850238
4,420
0.539551
2026-03-28T08:07:53.354419
b88da55b73ff4d892580c4d7f36b812a
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.279099
3,818
0.466064
2026-03-03T19:35:38.161811
3,599
0.439331
2026-03-02T02:12:39.548724
3,819
0.466187
2026-03-04T05:28:48.850238
3,636
0.443848
2026-03-28T08:07:53.354419
4f85baf61210bbad0ecfffe313f39f48
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.271868
3,610
0.440674
2026-03-03T19:35:38.161811
3,365
0.410767
2026-03-02T02:12:39.548724
3,604
0.439941
2026-03-04T05:28:48.850238
3,428
0.418457
2026-03-28T08:07:53.354419
0e69274675e07ca4a3f46eb169335dae
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.306492
4,101
0.50061
2026-03-03T19:35:38.161811
3,887
0.474487
2026-03-02T02:12:39.548724
4,098
0.500244
2026-03-04T05:28:48.850238
3,929
0.479614
2026-03-28T08:07:53.354419
b3c012fdd308ee580770ea4bd091eab4
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.257698
3,662
0.447021
2026-03-03T19:35:38.161811
3,381
0.41272
2026-03-02T02:12:39.548724
3,636
0.443848
2026-03-04T05:28:48.850238
3,500
0.427246
2026-03-28T08:07:53.354419
3000ee3e53331079fdcde6bd081bf02b
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.287737
3,932
0.47998
2026-03-03T19:35:38.161811
3,720
0.454102
2026-03-02T02:12:39.548724
3,939
0.480835
2026-03-04T05:28:48.850238
3,775
0.460815
2026-03-28T08:07:53.354419
febaaaa990a02e69597b2da0da0f6ae2
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.315833
4,163
0.508179
2026-03-03T19:35:38.161811
3,971
0.484741
2026-03-02T02:12:39.548724
4,181
0.510376
2026-03-04T05:28:48.850238
3,990
0.487061
2026-03-28T08:07:53.354419
1601fc167774c31bb71a7a9a493c807c
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.334164
4,366
0.532959
2026-03-03T19:35:38.161811
4,206
0.513428
2026-03-02T02:12:39.548724
4,426
0.540283
2026-03-04T05:28:48.850238
4,223
0.515503
2026-03-28T08:07:53.354419
84e6b68148c76358afccdb28194b8395
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.325414
4,139
0.505249
2026-03-03T19:35:38.161811
3,970
0.484619
2026-03-02T02:12:39.548724
4,180
0.510254
2026-03-04T05:28:48.850238
3,984
0.486328
2026-03-28T08:07:53.354419
6d066fab3ece447df6ee12f3cf924a62
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.308957
4,033
0.49231
2026-03-03T19:35:38.161811
3,847
0.469604
2026-03-02T02:12:39.548724
4,072
0.49707
2026-03-04T05:28:48.850238
3,887
0.474487
2026-03-28T08:07:53.354419
44881c185a60f9ac8321f15cdcb79a05
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.302564
4,042
0.493408
2026-03-03T19:35:38.161811
3,863
0.471558
2026-03-02T02:12:39.548724
4,072
0.49707
2026-03-04T05:28:48.850238
3,885
0.474243
2026-03-28T08:07:53.354419
e01e2c6913f5736e80d08bad8794fbb7
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.293225
3,933
0.480103
2026-03-03T19:35:38.161811
3,705
0.452271
2026-03-02T02:12:39.548724
3,914
0.477783
2026-03-04T05:28:48.850238
3,763
0.459351
2026-03-28T08:07:53.354419
56bef9b8da2fad779721ad0353657afd
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.364701
4,503
0.549683
2026-03-03T19:35:38.161811
4,376
0.53418
2026-03-02T02:12:39.548724
4,614
0.563232
2026-03-04T05:28:48.850238
4,384
0.535156
2026-03-28T08:07:53.354419
b0b7c4ee4c9c75476ce24662fa45bb39
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.345197
4,320
0.527344
2026-03-03T19:35:38.161811
4,128
0.503906
2026-03-02T02:12:39.548724
4,361
0.532349
2026-03-04T05:28:48.850238
4,208
0.513672
2026-03-28T08:07:53.354419
d7c5a1e19bc13c2e8adf77a1c65288dc
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.340562
4,339
0.529663
2026-03-03T19:35:38.161811
4,155
0.507202
2026-03-02T02:12:39.548724
4,400
0.537109
2026-03-04T05:28:48.850238
4,218
0.514893
2026-03-28T08:07:53.354419
31d61ee97990a71013ff660834949364
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.258993
3,602
0.439697
2026-03-03T19:35:38.161811
3,372
0.411621
2026-03-02T02:12:39.548724
3,599
0.439331
2026-03-04T05:28:48.850238
3,442
0.420166
2026-03-28T08:07:53.354419
df60fd8ce3fb1e435160d95bb0de27a9
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.376525
4,544
0.554688
2026-03-03T19:35:38.161811
4,397
0.536743
2026-03-02T02:12:39.548724
4,631
0.565308
2026-03-04T05:28:48.850238
4,429
0.540649
2026-03-28T08:07:53.354419
269529281d90f790379421cb5c0ce000
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.387357
4,630
0.565186
2026-03-03T19:35:38.161811
4,528
0.552734
2026-03-02T02:12:39.548724
4,782
0.58374
2026-03-04T05:28:48.850238
4,529
0.552856
2026-03-28T08:07:53.354419
91a3572dc58822383210ffc729b35f17
014f5ccb342104fe99a81a5c6906ff55
model.layers.0.input_layernorm.weight
4096
8,192
null
0.578688
5,955
0.726929
2025-12-05T16:52:04.321338
5,539
0.676147
2026-02-19T16:56:07.689881
5,883
0.71814
2026-03-04T05:28:48.850238
5,799
0.707886
2026-03-28T08:07:53.354419
0aa24edc9370a27884bd5d52d1e0f5fc
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.340807
4,362
0.532471
2026-03-03T19:35:38.161811
4,206
0.513428
2026-03-02T02:12:39.548724
4,422
0.539795
2026-03-04T05:28:48.850238
4,229
0.516235
2026-03-28T08:07:53.354419
dedbcd260ec171d034d5491d1e724df8
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.354313
4,491
0.548218
2026-03-03T19:35:38.161811
4,375
0.534058
2026-03-02T02:12:39.548724
4,595
0.560913
2026-03-04T05:28:48.850238
4,400
0.537109
2026-03-28T08:07:53.354419
ab77915b293abeebce7e48c522822e5e
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.37022
4,515
0.551147
2026-03-03T19:35:38.161811
4,401
0.537231
2026-03-02T02:12:39.548724
4,637
0.56604
2026-03-04T05:28:48.850238
4,440
0.541992
2026-03-28T08:07:53.354419
0e5f93618f1ad10a3cf97a0421353489
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.382529
4,508
0.550293
2026-03-03T19:35:38.161811
4,378
0.534424
2026-03-02T02:12:39.548724
4,620
0.563965
2026-03-04T05:28:48.850238
4,402
0.537354
2026-03-28T08:07:53.354419
87c6f205b38b231f2da465624de5af14
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.365693
4,614
0.563232
2026-03-03T19:35:38.161811
4,624
0.564453
2026-03-02T02:12:39.548724
4,732
0.577637
2026-03-04T05:28:48.850238
4,472
0.545898
2026-03-28T08:07:53.354419
e05e941626a652cd5199c668c3efd629
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.394279
4,625
0.564575
2026-03-03T19:35:38.161811
4,458
0.544189
2026-03-02T02:12:39.548724
4,741
0.578735
2026-03-04T05:28:48.850238
4,490
0.548096
2026-03-28T08:07:53.354419
7d87b0153415b60c280efb748441359d
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.373076
4,461
0.544556
2026-03-03T19:35:38.161811
4,372
0.533691
2026-03-02T02:12:39.548724
4,579
0.55896
2026-03-04T05:28:48.850238
4,352
0.53125
2026-03-28T08:07:53.354419
a8a52529b8c151bc63d3ea08db883d6b
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.360894
4,464
0.544922
2026-03-03T19:35:38.161811
4,401
0.537231
2026-03-02T02:12:39.548724
4,561
0.556763
2026-03-04T05:28:48.850238
4,329
0.528442
2026-03-28T08:07:53.354419
1fb7d82aee1dff9c5b5838d0d33620dd
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.435783
4,980
0.60791
2026-03-03T19:35:38.161811
4,835
0.59021
2026-03-02T02:12:39.548724
5,164
0.630371
2026-03-04T05:28:48.850238
4,876
0.595215
2026-03-28T08:07:53.354419
223febc4bf18d5f7bb38d33d225f5c46
8bf0998a3c8a072007996de4d1efaf2a
model.layers.0.input_layernorm.weight
(4096,)
8,192
null
0.343569
4,283
0.522827
2026-03-03T19:35:38.161811
4,148
0.506348
2026-03-02T02:12:39.548724
4,325
0.527954
2026-03-04T05:28:48.850238
4,105
0.501099
2026-03-28T08:07:53.354419
d5f963c672096bce971fd8d443661b1e
aecd8e5a3fa91f6d4c7cb69cd5a095e7
model.layers.1.input_layernorm.weight
4096
8,192
null
0.058975
704
0.085938
2025-12-05T16:39:37.174549
691
0.084351
2026-02-20T01:02:20.149594
1,390
0.169678
2026-03-07T00:51:34.501108114+00:00
718
0.087646
2026-03-28T19:35:55
5d1af597300558ef6bc2befb18a15a7d
aecd8e5a3fa91f6d4c7cb69cd5a095e7
model.layers.1.input_layernorm.weight
4096
8,192
null
0.038494
224
0.027344
2025-12-05T16:39:37.174549
277
0.033813
2026-02-20T01:02:20.149594
1,212
0.147949
2026-03-06T19:24:02.563615300+00:00
268
0.032715
2026-03-28T19:35:55
46570a0c070462c9953c09e2a11b3efc
aecd8e5a3fa91f6d4c7cb69cd5a095e7
model.layers.1.input_layernorm.weight
4096
8,192
null
0.224926
3,082
0.376221
2025-12-05T16:39:37.174549
2,106
0.25708
2026-03-03T08:51:33.913828
2,647
0.32312
2026-03-06T07:22:21.601430802+00:00
2,281
0.278442
2026-03-29T23:08:58
b659ea13609451fe41be6aeb49a0f189
acb4eef1decdbd970ac8906680154a72
model.layers.30.post_attention_layernorm.weight
4096
8,192
0.006332
0.044878
502
0.061279
2025-12-05T16:39:37.174549
439
0.053589
2026-02-13T18:24:37.653069
1,280
0.15625
2026-03-06T02:30:56.842390818+00:00
null
null
null
86cb5b8443ae547da1e5841d53fdca5c
aecd8e5a3fa91f6d4c7cb69cd5a095e7
model.layers.1.input_layernorm.weight
4096
8,192
null
0.196689
2,697
0.329224
2025-12-05T16:39:37.174549
1,921
0.234497
2026-02-20T01:02:20.149594
2,483
0.303101
2026-03-06T04:10:16.794041082+00:00
2,067
0.252319
2026-03-28T19:35:55
End of preview.

TensorDex — Artifact Evaluation Cache

Data companion for the TensorDex SOSP artifact (code repoae/). It holds the pre-computed results and the raw tensors needed to reproduce the paper's figures and to verify them from scratch, so reviewers don't re-run the full 40 TB pipeline.

The code repo's ae/download_cache.py pulls this into ae/cache/.

Contents

Path Size Role
results.db 5.2 GB the 11.4 M-pair compression cache (every figure's numbers)
sample_blobs/<xx>/<yy>/<id>.safetensors ~2 GB raw tensor blobs for the Tier-1 sample (content-addressed)
sample_pairs.tsv the (target, base) pairs the sample covers
data/tensordb_s3/metadata.db 567 MB slim: tensor sizes + model→tensor map (for reduction charts)
model_hub_crawl/ ~127 MB Hugging Face crawl stats (Fig 2, Fig 4)
tests/output/, compression_data/, model_level_reduction/ ~500 MB baseline plans & traces the chart modules read

results.db is a self-contained SQLite file (WAL folded in), slimmed to exactly the columns the AE charts and verification scripts read (original row ids are preserved, so subsampled charts reproduce the paper's figures byte-for-byte). Blobs are keyed by XXH3-128 of the raw tensor bytes — id == hash(bytes); that is what Tier-1 verify_sample.py re-checks.

results.db column guide

One row per compressed (target, base) tensor pair of the trace:

Column(s) Meaning
target_id, base_id XXH3-128 content hashes of the two tensors (= blob ids)
param_name, shape, bytes_in tensor identity and raw size
bcs_dist TensorSketch (BCS) distance between the pair — TensorPred's input
tratio / tbytes_out TensorX delta codec (TensorDex-TX, 65.1 %) — re-derived bit-exact by make verify
fratio / fbytes_out FM++ delta codec (the 70.5 % headline) — re-derived bit-exact after make ae-fmpp
ratio / bytes_out BitX (ZipLLM's delta codec) baseline
aratio / abytes_out, zratio dense per-pair delta ratios computed during development — the fitting cache TensorPred is trained and evaluated on (Fig 13; 5.77 M pairs vs ~1 M for FM++); zratio is its zstd fallback for tiny tensors, also used by the entropy/CDF analyses
pred_ratio TensorPred's stored prediction of aratiomake verify-predict re-fits the model from scratch and recovers this column to ~1e-4
timestamp, ttimestamp row provenance (trace ordering; legacy-row reporting in verify)

Use

# in the code repo
python ae/download_cache.py --repo <this-dataset-id>
make figures      # Tier 0 — re-plot every figure
make verify       # Tier 1 — re-derive a random sample from these blobs

Provenance

Generated by the code repo's authoring tools: ae/stage_data.py (chart inputs

  • slim metadata.db), ae/build_sample_bundle.py (the blob sample). Regenerate a larger sample with --budget-gb N.
Downloads last month
102