Add new SentenceTransformer model
Browse files- 1_Pooling/config.json +3 -3
- README.md +99 -98
- config.json +4 -4
- config_sentence_transformers.json +2 -2
- model.safetensors +2 -2
- tokenizer_config.json +1 -1
1_Pooling/config.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
-
"word_embedding_dimension":
|
| 3 |
-
"pooling_mode_cls_token":
|
| 4 |
-
"pooling_mode_mean_tokens":
|
| 5 |
"pooling_mode_max_tokens": false,
|
| 6 |
"pooling_mode_mean_sqrt_len_tokens": false,
|
| 7 |
"pooling_mode_weightedmean_tokens": false,
|
|
|
|
| 1 |
{
|
| 2 |
+
"word_embedding_dimension": 1024,
|
| 3 |
+
"pooling_mode_cls_token": false,
|
| 4 |
+
"pooling_mode_mean_tokens": true,
|
| 5 |
"pooling_mode_max_tokens": false,
|
| 6 |
"pooling_mode_mean_sqrt_len_tokens": false,
|
| 7 |
"pooling_mode_weightedmean_tokens": false,
|
README.md
CHANGED
|
@@ -12,54 +12,51 @@ tags:
|
|
| 12 |
- retrieval
|
| 13 |
- reranking
|
| 14 |
- generated_from_trainer
|
| 15 |
-
- dataset_size:
|
| 16 |
-
- loss:
|
| 17 |
-
base_model:
|
| 18 |
widget:
|
| 19 |
-
- source_sentence:
|
|
|
|
| 20 |
sentences:
|
| 21 |
-
-
|
| 22 |
-
|
| 23 |
-
-
|
| 24 |
-
|
| 25 |
-
-
|
| 26 |
-
|
| 27 |
-
|
|
|
|
| 28 |
sentences:
|
| 29 |
-
-
|
| 30 |
-
|
| 31 |
-
-
|
| 32 |
-
|
| 33 |
-
-
|
| 34 |
-
|
| 35 |
-
- source_sentence:
|
| 36 |
-
|
| 37 |
sentences:
|
| 38 |
-
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
- It was formed in December 2014 from elements of the dissolved 51st Mechanized
|
| 44 |
-
Brigade and newly mobilized units .
|
| 45 |
-
- source_sentence: The band pursued `` signals `` in January 2012 in three weeks ,
|
| 46 |
-
and drums were recorded in a day and a half .
|
| 47 |
sentences:
|
| 48 |
-
-
|
| 49 |
-
|
| 50 |
-
-
|
| 51 |
-
|
| 52 |
-
-
|
| 53 |
-
|
| 54 |
-
- source_sentence:
|
| 55 |
-
|
| 56 |
sentences:
|
| 57 |
-
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
with red lines and text .
|
| 63 |
datasets:
|
| 64 |
- redis/langcache-sentencepairs-v2
|
| 65 |
pipeline_tag: sentence-similarity
|
|
@@ -84,42 +81,42 @@ model-index:
|
|
| 84 |
type: test
|
| 85 |
metrics:
|
| 86 |
- type: cosine_accuracy@1
|
| 87 |
-
value: 0.
|
| 88 |
name: Cosine Accuracy@1
|
| 89 |
- type: cosine_precision@1
|
| 90 |
-
value: 0.
|
| 91 |
name: Cosine Precision@1
|
| 92 |
- type: cosine_recall@1
|
| 93 |
-
value: 0.
|
| 94 |
name: Cosine Recall@1
|
| 95 |
- type: cosine_ndcg@10
|
| 96 |
-
value: 0.
|
| 97 |
name: Cosine Ndcg@10
|
| 98 |
- type: cosine_mrr@1
|
| 99 |
-
value: 0.
|
| 100 |
name: Cosine Mrr@1
|
| 101 |
- type: cosine_map@100
|
| 102 |
-
value: 0.
|
| 103 |
name: Cosine Map@100
|
| 104 |
- type: cosine_auc_precision_cache_hit_ratio
|
| 105 |
-
value: 0.
|
| 106 |
name: Cosine Auc Precision Cache Hit Ratio
|
| 107 |
- type: cosine_auc_similarity_distribution
|
| 108 |
-
value: 0.
|
| 109 |
name: Cosine Auc Similarity Distribution
|
| 110 |
---
|
| 111 |
|
| 112 |
# Redis fine-tuned BiEncoder model for semantic caching on LangCache
|
| 113 |
|
| 114 |
-
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [
|
| 115 |
|
| 116 |
## Model Details
|
| 117 |
|
| 118 |
### Model Description
|
| 119 |
- **Model Type:** Sentence Transformer
|
| 120 |
-
- **Base model:** [
|
| 121 |
- **Maximum Sequence Length:** 100 tokens
|
| 122 |
-
- **Output Dimensionality:**
|
| 123 |
- **Similarity Function:** Cosine Similarity
|
| 124 |
- **Training Dataset:**
|
| 125 |
- [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2)
|
|
@@ -137,7 +134,7 @@ This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [A
|
|
| 137 |
```
|
| 138 |
SentenceTransformer(
|
| 139 |
(0): Transformer({'max_seq_length': 100, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
|
| 140 |
-
(1): Pooling({'word_embedding_dimension':
|
| 141 |
)
|
| 142 |
```
|
| 143 |
|
|
@@ -159,20 +156,20 @@ from sentence_transformers import SentenceTransformer
|
|
| 159 |
model = SentenceTransformer("redis/langcache-embed-experimental")
|
| 160 |
# Run inference
|
| 161 |
sentences = [
|
| 162 |
-
'
|
| 163 |
-
'
|
| 164 |
-
'
|
| 165 |
]
|
| 166 |
embeddings = model.encode(sentences)
|
| 167 |
print(embeddings.shape)
|
| 168 |
-
# [3,
|
| 169 |
|
| 170 |
# Get the similarity scores for the embeddings
|
| 171 |
similarities = model.similarity(embeddings, embeddings)
|
| 172 |
print(similarities)
|
| 173 |
-
# tensor([[1.0000, 0.9844, 0.
|
| 174 |
-
# [0.9844, 0.
|
| 175 |
-
# [0.
|
| 176 |
```
|
| 177 |
|
| 178 |
<!--
|
|
@@ -208,16 +205,16 @@ You can finetune this model on your own dataset.
|
|
| 208 |
* Dataset: `test`
|
| 209 |
* Evaluated with <code>ir_evaluator.CustomInformationRetrievalEvaluator</code>
|
| 210 |
|
| 211 |
-
| Metric | Value
|
| 212 |
-
|
| 213 |
-
| cosine_accuracy@1 | 0.
|
| 214 |
-
| cosine_precision@1 | 0.
|
| 215 |
-
| cosine_recall@1 | 0.
|
| 216 |
-
| **cosine_ndcg@10** | **0.
|
| 217 |
-
| cosine_mrr@1 | 0.
|
| 218 |
-
| cosine_map@100 | 0.
|
| 219 |
-
| cosine_auc_precision_cache_hit_ratio | 0.
|
| 220 |
-
| cosine_auc_similarity_distribution | 0.
|
| 221 |
|
| 222 |
<!--
|
| 223 |
## Bias, Risks and Limitations
|
|
@@ -238,23 +235,25 @@ You can finetune this model on your own dataset.
|
|
| 238 |
#### LangCache Sentence Pairs (all)
|
| 239 |
|
| 240 |
* Dataset: [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2)
|
| 241 |
-
* Size:
|
| 242 |
-
* Columns: <code>
|
| 243 |
* Approximate statistics based on the first 1000 samples:
|
| 244 |
-
| |
|
| 245 |
-
|
| 246 |
-
| type | string | string |
|
| 247 |
-
| details | <ul><li>min: 8 tokens</li><li>mean: 27.
|
| 248 |
* Samples:
|
| 249 |
-
|
|
| 250 |
-
|
| 251 |
-
| <code>The newer Punts are still very much in existence today and race in the same fleets as the older boats .</code> | <code>The newer punts are still very much in existence today and run in the same fleets as the older boats .</code> | <code>
|
| 252 |
-
| <code>
|
| 253 |
-
| <code>
|
| 254 |
-
* Loss:
|
| 255 |
```json
|
| 256 |
{
|
| 257 |
-
"
|
|
|
|
|
|
|
| 258 |
}
|
| 259 |
```
|
| 260 |
|
|
@@ -263,30 +262,32 @@ You can finetune this model on your own dataset.
|
|
| 263 |
#### LangCache Sentence Pairs (all)
|
| 264 |
|
| 265 |
* Dataset: [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2)
|
| 266 |
-
* Size:
|
| 267 |
-
* Columns: <code>
|
| 268 |
* Approximate statistics based on the first 1000 samples:
|
| 269 |
-
| |
|
| 270 |
-
|
| 271 |
-
| type | string | string |
|
| 272 |
-
| details | <ul><li>min: 8 tokens</li><li>mean: 27.
|
| 273 |
* Samples:
|
| 274 |
-
|
|
| 275 |
-
|
| 276 |
-
| <code>The newer Punts are still very much in existence today and race in the same fleets as the older boats .</code> | <code>The newer punts are still very much in existence today and run in the same fleets as the older boats .</code> | <code>
|
| 277 |
-
| <code>
|
| 278 |
-
| <code>
|
| 279 |
-
* Loss:
|
| 280 |
```json
|
| 281 |
{
|
| 282 |
-
"
|
|
|
|
|
|
|
| 283 |
}
|
| 284 |
```
|
| 285 |
|
| 286 |
### Training Logs
|
| 287 |
| Epoch | Step | test_cosine_ndcg@10 |
|
| 288 |
|:-----:|:----:|:-------------------:|
|
| 289 |
-
| -1 | -1 | 0.
|
| 290 |
|
| 291 |
|
| 292 |
### Framework Versions
|
|
|
|
| 12 |
- retrieval
|
| 13 |
- reranking
|
| 14 |
- generated_from_trainer
|
| 15 |
+
- dataset_size:9233417
|
| 16 |
+
- loss:ArcFaceInBatchLoss
|
| 17 |
+
base_model: answerdotai/ModernBERT-large
|
| 18 |
widget:
|
| 19 |
+
- source_sentence: Hayley Vaughan portrayed Ripa on the ABC daytime soap opera , ``
|
| 20 |
+
All My Children `` , between 1990 and 2002 .
|
| 21 |
sentences:
|
| 22 |
+
- Traxxpad is a music application for Sony 's PlayStation Portable published by
|
| 23 |
+
Definitive Studios and developed by Eidos Interactive .
|
| 24 |
+
- Between 1990 and 2002 , Hayley Vaughan Ripa portrayed in the ABC soap opera ``
|
| 25 |
+
All My Children `` .
|
| 26 |
+
- Between 1990 and 2002 , Ripa Hayley portrayed Vaughan in the ABC soap opera ``
|
| 27 |
+
All My Children `` .
|
| 28 |
+
- source_sentence: Olivella monilifera is a species of dwarf sea snail , small gastropod
|
| 29 |
+
mollusk in the family Olivellidae , the marine olives .
|
| 30 |
sentences:
|
| 31 |
+
- Olivella monilifera is a species of the dwarf - sea snail , small gastropod mollusk
|
| 32 |
+
in the Olivellidae family , the marine olives .
|
| 33 |
+
- He was cut by the Browns after being signed by the Bills in 2013 . He was later
|
| 34 |
+
released .
|
| 35 |
+
- Olivella monilifera is a kind of sea snail , marine gastropod mollusk in the Olivellidae
|
| 36 |
+
family , the dwarf olives .
|
| 37 |
+
- source_sentence: Hayashi said that Mackey `` is a sort of `` of the original model
|
| 38 |
+
for Tenchi .
|
| 39 |
sentences:
|
| 40 |
+
- In the summer of 2009 , Ellick shot a documentary about Malala Yousafzai .
|
| 41 |
+
- Hayashi said that Mackey is `` sort of `` the original model for Tenchi .
|
| 42 |
+
- Mackey said that Hayashi is `` sort of `` the original model for Tenchi .
|
| 43 |
+
- source_sentence: Much of the film was shot on location in Los Angeles and in nearby
|
| 44 |
+
Burbank and Glendale .
|
|
|
|
|
|
|
|
|
|
|
|
|
| 45 |
sentences:
|
| 46 |
+
- Much of the film was shot on location in Los Angeles and in nearby Burbank and
|
| 47 |
+
Glendale .
|
| 48 |
+
- Much of the film was shot on site in Burbank and Glendale and in the nearby Los
|
| 49 |
+
Angeles .
|
| 50 |
+
- Traxxpad is a music application for the Sony PlayStation Portable developed by
|
| 51 |
+
the Definitive Studios and published by Eidos Interactive .
|
| 52 |
+
- source_sentence: According to him , the earth is the carrier of his artistic work
|
| 53 |
+
, which is only integrated into the creative process by minimal changes .
|
| 54 |
sentences:
|
| 55 |
+
- National players are Bold players .
|
| 56 |
+
- According to him , earth is the carrier of his artistic work being integrated
|
| 57 |
+
into the creative process only by minimal changes .
|
| 58 |
+
- According to him , earth is the carrier of his creative work being integrated
|
| 59 |
+
into the artistic process only by minimal changes .
|
|
|
|
| 60 |
datasets:
|
| 61 |
- redis/langcache-sentencepairs-v2
|
| 62 |
pipeline_tag: sentence-similarity
|
|
|
|
| 81 |
type: test
|
| 82 |
metrics:
|
| 83 |
- type: cosine_accuracy@1
|
| 84 |
+
value: 0.44081091729646477
|
| 85 |
name: Cosine Accuracy@1
|
| 86 |
- type: cosine_precision@1
|
| 87 |
+
value: 0.44081091729646477
|
| 88 |
name: Cosine Precision@1
|
| 89 |
- type: cosine_recall@1
|
| 90 |
+
value: 0.42663486382682986
|
| 91 |
name: Cosine Recall@1
|
| 92 |
- type: cosine_ndcg@10
|
| 93 |
+
value: 0.6274011007244752
|
| 94 |
name: Cosine Ndcg@10
|
| 95 |
- type: cosine_mrr@1
|
| 96 |
+
value: 0.44081091729646477
|
| 97 |
name: Cosine Mrr@1
|
| 98 |
- type: cosine_map@100
|
| 99 |
+
value: 0.5749605963252064
|
| 100 |
name: Cosine Map@100
|
| 101 |
- type: cosine_auc_precision_cache_hit_ratio
|
| 102 |
+
value: 0.27130175854619276
|
| 103 |
name: Cosine Auc Precision Cache Hit Ratio
|
| 104 |
- type: cosine_auc_similarity_distribution
|
| 105 |
+
value: 0.40770905754259995
|
| 106 |
name: Cosine Auc Similarity Distribution
|
| 107 |
---
|
| 108 |
|
| 109 |
# Redis fine-tuned BiEncoder model for semantic caching on LangCache
|
| 110 |
|
| 111 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [answerdotai/ModernBERT-large](https://huggingface.co/answerdotai/ModernBERT-large) on the [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2) dataset. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for sentence pair similarity.
|
| 112 |
|
| 113 |
## Model Details
|
| 114 |
|
| 115 |
### Model Description
|
| 116 |
- **Model Type:** Sentence Transformer
|
| 117 |
+
- **Base model:** [answerdotai/ModernBERT-large](https://huggingface.co/answerdotai/ModernBERT-large) <!-- at revision 45bb4654a4d5aaff24dd11d4781fa46d39bf8c13 -->
|
| 118 |
- **Maximum Sequence Length:** 100 tokens
|
| 119 |
+
- **Output Dimensionality:** 1024 dimensions
|
| 120 |
- **Similarity Function:** Cosine Similarity
|
| 121 |
- **Training Dataset:**
|
| 122 |
- [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2)
|
|
|
|
| 134 |
```
|
| 135 |
SentenceTransformer(
|
| 136 |
(0): Transformer({'max_seq_length': 100, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
|
| 137 |
+
(1): Pooling({'word_embedding_dimension': 1024, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
|
| 138 |
)
|
| 139 |
```
|
| 140 |
|
|
|
|
| 156 |
model = SentenceTransformer("redis/langcache-embed-experimental")
|
| 157 |
# Run inference
|
| 158 |
sentences = [
|
| 159 |
+
'According to him , the earth is the carrier of his artistic work , which is only integrated into the creative process by minimal changes .',
|
| 160 |
+
'According to him , earth is the carrier of his artistic work being integrated into the creative process only by minimal changes .',
|
| 161 |
+
'According to him , earth is the carrier of his creative work being integrated into the artistic process only by minimal changes .',
|
| 162 |
]
|
| 163 |
embeddings = model.encode(sentences)
|
| 164 |
print(embeddings.shape)
|
| 165 |
+
# [3, 1024]
|
| 166 |
|
| 167 |
# Get the similarity scores for the embeddings
|
| 168 |
similarities = model.similarity(embeddings, embeddings)
|
| 169 |
print(similarities)
|
| 170 |
+
# tensor([[1.0000, 0.9844, 0.9844],
|
| 171 |
+
# [0.9844, 0.9961, 0.9922],
|
| 172 |
+
# [0.9844, 0.9922, 0.9961]], dtype=torch.bfloat16)
|
| 173 |
```
|
| 174 |
|
| 175 |
<!--
|
|
|
|
| 205 |
* Dataset: `test`
|
| 206 |
* Evaluated with <code>ir_evaluator.CustomInformationRetrievalEvaluator</code>
|
| 207 |
|
| 208 |
+
| Metric | Value |
|
| 209 |
+
|:-------------------------------------|:-----------|
|
| 210 |
+
| cosine_accuracy@1 | 0.4408 |
|
| 211 |
+
| cosine_precision@1 | 0.4408 |
|
| 212 |
+
| cosine_recall@1 | 0.4266 |
|
| 213 |
+
| **cosine_ndcg@10** | **0.6274** |
|
| 214 |
+
| cosine_mrr@1 | 0.4408 |
|
| 215 |
+
| cosine_map@100 | 0.575 |
|
| 216 |
+
| cosine_auc_precision_cache_hit_ratio | 0.2713 |
|
| 217 |
+
| cosine_auc_similarity_distribution | 0.4077 |
|
| 218 |
|
| 219 |
<!--
|
| 220 |
## Bias, Risks and Limitations
|
|
|
|
| 235 |
#### LangCache Sentence Pairs (all)
|
| 236 |
|
| 237 |
* Dataset: [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2)
|
| 238 |
+
* Size: 126,938 training samples
|
| 239 |
+
* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
|
| 240 |
* Approximate statistics based on the first 1000 samples:
|
| 241 |
+
| | anchor | positive | negative |
|
| 242 |
+
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
|
| 243 |
+
| type | string | string | string |
|
| 244 |
+
| details | <ul><li>min: 8 tokens</li><li>mean: 27.27 tokens</li><li>max: 49 tokens</li></ul> | <ul><li>min: 8 tokens</li><li>mean: 27.27 tokens</li><li>max: 48 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 26.54 tokens</li><li>max: 61 tokens</li></ul> |
|
| 245 |
* Samples:
|
| 246 |
+
| anchor | positive | negative |
|
| 247 |
+
|:--------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------|
|
| 248 |
+
| <code>The newer Punts are still very much in existence today and race in the same fleets as the older boats .</code> | <code>The newer punts are still very much in existence today and run in the same fleets as the older boats .</code> | <code>how can I get financial freedom as soon as possible?</code> |
|
| 249 |
+
| <code>The newer punts are still very much in existence today and run in the same fleets as the older boats .</code> | <code>The newer Punts are still very much in existence today and race in the same fleets as the older boats .</code> | <code>The older Punts are still very much in existence today and race in the same fleets as the newer boats .</code> |
|
| 250 |
+
| <code>Turner Valley , was at the Turner Valley Bar N Ranch Airport , southwest of the Turner Valley Bar N Ranch , Alberta , Canada .</code> | <code>Turner Valley , , was located at Turner Valley Bar N Ranch Airport , southwest of Turner Valley Bar N Ranch , Alberta , Canada .</code> | <code>Turner Valley Bar N Ranch Airport , , was located at Turner Valley Bar N Ranch , southwest of Turner Valley , Alberta , Canada .</code> |
|
| 251 |
+
* Loss: <code>losses.ArcFaceInBatchLoss</code> with these parameters:
|
| 252 |
```json
|
| 253 |
{
|
| 254 |
+
"scale": 20.0,
|
| 255 |
+
"similarity_fct": "cos_sim",
|
| 256 |
+
"gather_across_devices": false
|
| 257 |
}
|
| 258 |
```
|
| 259 |
|
|
|
|
| 262 |
#### LangCache Sentence Pairs (all)
|
| 263 |
|
| 264 |
* Dataset: [LangCache Sentence Pairs (all)](https://huggingface.co/datasets/redis/langcache-sentencepairs-v2)
|
| 265 |
+
* Size: 126,938 evaluation samples
|
| 266 |
+
* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
|
| 267 |
* Approximate statistics based on the first 1000 samples:
|
| 268 |
+
| | anchor | positive | negative |
|
| 269 |
+
|:--------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
|
| 270 |
+
| type | string | string | string |
|
| 271 |
+
| details | <ul><li>min: 8 tokens</li><li>mean: 27.27 tokens</li><li>max: 49 tokens</li></ul> | <ul><li>min: 8 tokens</li><li>mean: 27.27 tokens</li><li>max: 48 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 26.54 tokens</li><li>max: 61 tokens</li></ul> |
|
| 272 |
* Samples:
|
| 273 |
+
| anchor | positive | negative |
|
| 274 |
+
|:--------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------|
|
| 275 |
+
| <code>The newer Punts are still very much in existence today and race in the same fleets as the older boats .</code> | <code>The newer punts are still very much in existence today and run in the same fleets as the older boats .</code> | <code>how can I get financial freedom as soon as possible?</code> |
|
| 276 |
+
| <code>The newer punts are still very much in existence today and run in the same fleets as the older boats .</code> | <code>The newer Punts are still very much in existence today and race in the same fleets as the older boats .</code> | <code>The older Punts are still very much in existence today and race in the same fleets as the newer boats .</code> |
|
| 277 |
+
| <code>Turner Valley , was at the Turner Valley Bar N Ranch Airport , southwest of the Turner Valley Bar N Ranch , Alberta , Canada .</code> | <code>Turner Valley , , was located at Turner Valley Bar N Ranch Airport , southwest of Turner Valley Bar N Ranch , Alberta , Canada .</code> | <code>Turner Valley Bar N Ranch Airport , , was located at Turner Valley Bar N Ranch , southwest of Turner Valley , Alberta , Canada .</code> |
|
| 278 |
+
* Loss: <code>losses.ArcFaceInBatchLoss</code> with these parameters:
|
| 279 |
```json
|
| 280 |
{
|
| 281 |
+
"scale": 20.0,
|
| 282 |
+
"similarity_fct": "cos_sim",
|
| 283 |
+
"gather_across_devices": false
|
| 284 |
}
|
| 285 |
```
|
| 286 |
|
| 287 |
### Training Logs
|
| 288 |
| Epoch | Step | test_cosine_ndcg@10 |
|
| 289 |
|:-----:|:----:|:-------------------:|
|
| 290 |
+
| -1 | -1 | 0.6274 |
|
| 291 |
|
| 292 |
|
| 293 |
### Framework Versions
|
config.json
CHANGED
|
@@ -19,10 +19,10 @@
|
|
| 19 |
"global_rope_theta": 160000.0,
|
| 20 |
"gradient_checkpointing": false,
|
| 21 |
"hidden_activation": "gelu",
|
| 22 |
-
"hidden_size":
|
| 23 |
"initializer_cutoff_factor": 2.0,
|
| 24 |
"initializer_range": 0.02,
|
| 25 |
-
"intermediate_size":
|
| 26 |
"layer_norm_eps": 1e-05,
|
| 27 |
"local_attention": 128,
|
| 28 |
"local_rope_theta": 10000.0,
|
|
@@ -32,8 +32,8 @@
|
|
| 32 |
"model_type": "modernbert",
|
| 33 |
"norm_bias": false,
|
| 34 |
"norm_eps": 1e-05,
|
| 35 |
-
"num_attention_heads":
|
| 36 |
-
"num_hidden_layers":
|
| 37 |
"pad_token_id": 50283,
|
| 38 |
"position_embedding_type": "absolute",
|
| 39 |
"repad_logits_with_grad": false,
|
|
|
|
| 19 |
"global_rope_theta": 160000.0,
|
| 20 |
"gradient_checkpointing": false,
|
| 21 |
"hidden_activation": "gelu",
|
| 22 |
+
"hidden_size": 1024,
|
| 23 |
"initializer_cutoff_factor": 2.0,
|
| 24 |
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 2624,
|
| 26 |
"layer_norm_eps": 1e-05,
|
| 27 |
"local_attention": 128,
|
| 28 |
"local_rope_theta": 10000.0,
|
|
|
|
| 32 |
"model_type": "modernbert",
|
| 33 |
"norm_bias": false,
|
| 34 |
"norm_eps": 1e-05,
|
| 35 |
+
"num_attention_heads": 16,
|
| 36 |
+
"num_hidden_layers": 28,
|
| 37 |
"pad_token_id": 50283,
|
| 38 |
"position_embedding_type": "absolute",
|
| 39 |
"repad_logits_with_grad": false,
|
config_sentence_transformers.json
CHANGED
|
@@ -1,4 +1,5 @@
|
|
| 1 |
{
|
|
|
|
| 2 |
"__version__": {
|
| 3 |
"sentence_transformers": "5.1.0",
|
| 4 |
"transformers": "4.56.0",
|
|
@@ -9,6 +10,5 @@
|
|
| 9 |
"document": ""
|
| 10 |
},
|
| 11 |
"default_prompt_name": null,
|
| 12 |
-
"similarity_fn_name": "cosine"
|
| 13 |
-
"model_type": "SentenceTransformer"
|
| 14 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"model_type": "SentenceTransformer",
|
| 3 |
"__version__": {
|
| 4 |
"sentence_transformers": "5.1.0",
|
| 5 |
"transformers": "4.56.0",
|
|
|
|
| 10 |
"document": ""
|
| 11 |
},
|
| 12 |
"default_prompt_name": null,
|
| 13 |
+
"similarity_fn_name": "cosine"
|
|
|
|
| 14 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f84ab85561ef1687cbe92711a697e05cba61d13ec58fc3ec279bd64147715b8
|
| 3 |
+
size 789580328
|
tokenizer_config.json
CHANGED
|
@@ -938,7 +938,7 @@
|
|
| 938 |
"input_ids",
|
| 939 |
"attention_mask"
|
| 940 |
],
|
| 941 |
-
"model_max_length":
|
| 942 |
"pad_token": "[PAD]",
|
| 943 |
"sep_token": "[SEP]",
|
| 944 |
"tokenizer_class": "PreTrainedTokenizerFast",
|
|
|
|
| 938 |
"input_ids",
|
| 939 |
"attention_mask"
|
| 940 |
],
|
| 941 |
+
"model_max_length": 8192,
|
| 942 |
"pad_token": "[PAD]",
|
| 943 |
"sep_token": "[SEP]",
|
| 944 |
"tokenizer_class": "PreTrainedTokenizerFast",
|