Hugging Face's logo

widebluesky
/
wbs-embedding-demo

Sentence Similarity
sentence-transformers
Safetensors
bert
advai-notebooks
embedding
demo
synthetic-data
random-initialization
text-embeddings-inference
Model card Files
xet
 Community
widebluesky commited on
Commit
b44712d
·
verified ·
1 Parent(s): e9079e1

Associate verified demo dataset and pin training-data provenance

Browse files
Files changed (2) hide show
  1. README.md +18 -12
  2. publish_manifest.json +2 -2
README.md CHANGED
@@ -8,7 +8,7 @@ tags:
8
  - random-initialization
9
  pipeline_tag: sentence-similarity
10
  datasets:
11
- - widebluesky/wbs-retrieval-demo
12
  ---
13
 
14
  # widebluesky/wbs-embedding-demo
@@ -92,27 +92,33 @@ resuming training and local machine paths are not part of this publication.
92
 
93
  ## Training and evaluation dataset
94
 
95
- Dataset: [widebluesky/wbs-retrieval-demo](https://huggingface.co/datasets/widebluesky/wbs-retrieval-demo).
96
- Verified dataset revision: [`c75ca4e872c63c4683b266a4924a493968a1111f`](https://huggingface.co/datasets/widebluesky/wbs-retrieval-demo/tree/c75ca4e872c63c4683b266a4924a493968a1111f) (`demo-v1`).
97
 
98
- This is the synthetic source dataset used for this trained demo release.
99
- Its record fingerprint matches the saved training-run provenance; the published
100
- dataset was reloaded and checked record-by-record before associating it with this model.
 
 
101
 
102
- Training uses `embedding_training` (24 query-positive samples). Evaluation uses
103
- `queries` (12 train / 3 validation / 3 test), the shared `corpus` (12 documents),
104
- and `qrels` (216 judgments). The Ranker demo shares this benchmark but owns different training pairs.
105
 
106
  ```python
107
  from datasets import load_dataset
108
 
109
  data = load_dataset(
110
- "widebluesky/wbs-retrieval-demo", "embedding_training",
111
- revision="c75ca4e872c63c4683b266a4924a493968a1111f", token=True,
112
  )
113
  ```
114
 
115
  Access uses the existing Hugging Face login; both model and dataset repositories are private.
116
  Original project-native data files are available under `source/` in the dataset repository.
117
- Canonical dataset fingerprint: `f96589e88e79ce684f3c8733ae1a3c98a8f028459f6e82e3477efeb2867b2a19`.
118
  Training sample fingerprint: `cebc151f0e9c04eae8c92a1d79a8c3ae6771b1a81fb7484e7a13a34555ee9173`.
 
 
 
 
 
8
  - random-initialization
9
  pipeline_tag: sentence-similarity
10
  datasets:
11
+ - widebluesky/wbs-embedding-demo
12
  ---
13
 
14
  # widebluesky/wbs-embedding-demo
 
92
 
93
  ## Training and evaluation dataset
94
 
95
+ Dataset: [widebluesky/wbs-embedding-demo](https://huggingface.co/datasets/widebluesky/wbs-embedding-demo).
96
+ Verified dataset revision: [`d5d422450670f840377c9f8bb3b11b7919cdadfa`](https://huggingface.co/datasets/widebluesky/wbs-embedding-demo/tree/d5d422450670f840377c9f8bb3b11b7919cdadfa) (`demo-v1`).
97
 
98
+ This independently versioned task dataset contains the exact synthetic records used
99
+ for this trained demo release. It includes its own training samples and a complete
100
+ local copy of the evaluation benchmark. No other dataset repository is required.
101
+ All records were reloaded from the pinned Hub revision and checked against the
102
+ original training provenance before associating this dataset with the model.
103
 
104
+ Training uses `embedding_training` (24 query-positive contrastive samples).
105
+ Evaluation uses `queries` (12 train / 3 validation / 3 test), `corpus` (12 documents)
106
+ and `qrels` (216 judgments). Ranker training pairs and candidates are not included.
107
 
108
  ```python
109
  from datasets import load_dataset
110
 
111
  data = load_dataset(
112
+ "widebluesky/wbs-embedding-demo", "embedding_training",
113
+ revision="d5d422450670f840377c9f8bb3b11b7919cdadfa", token=True,
114
  )
115
  ```
116
 
117
  Access uses the existing Hugging Face login; both model and dataset repositories are private.
118
  Original project-native data files are available under `source/` in the dataset repository.
119
+ Canonical benchmark fingerprint: `f96589e88e79ce684f3c8733ae1a3c98a8f028459f6e82e3477efeb2867b2a19`.
120
  Training sample fingerprint: `cebc151f0e9c04eae8c92a1d79a8c3ae6771b1a81fb7484e7a13a34555ee9173`.
121
+
122
+ This association supersedes the former shared retrieval dataset association.
123
+ The original `widebluesky/wbs-retrieval-demo` dataset and its pinned `demo-v1` revision
124
+ remain available for historical reproducibility.
publish_manifest.json CHANGED
@@ -9,8 +9,8 @@
9
  },
10
  {
11
  "path": "README.md",
12
- "bytes": 3895,
13
- "sha256": "ca6e4e41957a9b88f6ca07871813eae4b7d6ac851a68a4177d014364b131f06e"
14
  },
15
  {
16
  "path": "advai_metadata.json",
 
9
  },
10
  {
11
  "path": "README.md",
12
+ "bytes": 4256,
13
+ "sha256": "c4ced8a8837245aa74142ccb0929e7b2c36d81864339ca303b9cad4e58097cff"
14
  },
15
  {
16
  "path": "advai_metadata.json",