Upload food embeddings model

README.md

@@ -4,35 +4,35 @@ tags:
 - sentence-similarity
 - feature-extraction
 - generated_from_trainer
-- dataset_size:4256
+- dataset_size:35819
 - loss:ContrastiveLoss
 base_model: sentence-transformers/all-mpnet-base-v2
 widget:
-- source_sentence: So Delicious Key Lime Yogurt
+- source_sentence: Broccoli, stalks raw
   sentences:
-  - Squash, yellow raw
-  - Babyfood, mixed fruit yogurt
-  - Beef, rib eye steak/roast bone-in lip-on raw
-- source_sentence: Cocoa Bumpers Cereal, Quaker Mother's
+  - Carrots, canned no salt
+  - Squash, Indian raw
+  - Biscuit, Popeyes
+- source_sentence: Cereal, General Mills Cheerios
   sentences:
-  - Lovebird Cereal Honey Box
-  - Ham, canned roasted
-  - Chicken, light meat with skin, cooked stewed
-- source_sentence: Broadbeans, raw immature seeds
+  - Chocolate pudding, ready-to-eat
+  - Mackerel, Atlantic cooked
+  - Cereal, Malt-O-Meal Berry Colossal Crunch
+- source_sentence: Beef Tenderloin, lean cooked broiled
   sentences:
-  - Peas, canned rinsed
-  - Promin Minestrone Soup
-  - Rice, brown long-grain cooked
-- source_sentence: Chicken, dark meat thigh meat and skin, added solution cooked braised
+  - Elk, tenderloin lean cooked broiled
+  - Chicken, capons giblets cooked simmered
+  - Barley, pearled cooked
+- source_sentence: Beef, New Zealand eye round slow roasted
   sentences:
-  - Moose, raw
-  - Chickpeas, cooked with salt
-  - Sausage, pork turkey and beef reduced sodium
-- source_sentence: Shortening, soy and cottonseed for pastries
+  - Sorghum flour, white pearled raw
+  - Beef, Denver cut steak, grilled
+  - Pudding, chocolate instant with 2% milk
+- source_sentence: Beef, shoulder steak boneless grilled
   sentences:
-  - Soup, chicken noodle reduced sodium
-  - Sea lion kidney, Steller (Alaska Native)
-  - Salad, McDonald's side
+  - Pork, bacon, cooked pan-fried
+  - Oyster, eastern breaded fried
+  - Beef, top blade steak, grilled select
 pipeline_tag: sentence-similarity
 library_name: sentence-transformers
 metrics:
@@ -49,10 +49,10 @@ model-index:
       type: validation
     metrics:
     - type: pearson_cosine
-      value: 0.8269809784218102
+      value: 0.8767870213264454
       name: Pearson Cosine
     - type: spearman_cosine
-      value: 0.845955787172452
+      value: 0.8665397416848721
       name: Spearman Cosine
 ---
 
@@ -106,9 +106,9 @@ from sentence_transformers import SentenceTransformer
 model = SentenceTransformer("jonny9f/food_embeddings4")
 # Run inference
 sentences = [
-    'Shortening, soy and cottonseed for pastries',
-    'Sea lion kidney, Steller (Alaska Native)',
-    'Soup, chicken noodle reduced sodium',
+    'Beef, shoulder steak boneless grilled',
+    'Beef, top blade steak, grilled select',
+    'Pork, bacon, cooked pan-fried',
 ]
 embeddings = model.encode(sentences)
 print(embeddings.shape)
@@ -153,10 +153,10 @@ You can finetune this model on your own dataset.
 * Dataset: `validation`
 * Evaluated with [<code>EmbeddingSimilarityEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)
 
-| Metric              | Value     |
-|:--------------------|:----------|
-| pearson_cosine      | 0.827     |
-| **spearman_cosine** | **0.846** |
+| Metric              | Value      |
+|:--------------------|:-----------|
+| pearson_cosine      | 0.8768     |
+| **spearman_cosine** | **0.8665** |
 
 <!--
 ## Bias, Risks and Limitations
@@ -177,19 +177,19 @@ You can finetune this model on your own dataset.
 #### Unnamed Dataset
 
 
-* Size: 4,256 training samples
+* Size: 35,819 training samples
 * Columns: <code>sentence_0</code>, <code>sentence_1</code>, and <code>label</code>
 * Approximate statistics based on the first 1000 samples:
-  |         | sentence_0                                                                       | sentence_1                                                                       | label                                                           |
-  |:--------|:---------------------------------------------------------------------------------|:---------------------------------------------------------------------------------|:----------------------------------------------------------------|
-  | type    | string                                                                           | string                                                                           | float                                                           |
-  | details | <ul><li>min: 3 tokens</li><li>mean: 9.91 tokens</li><li>max: 27 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 9.96 tokens</li><li>max: 24 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.39</li><li>max: 0.85</li></ul> |
+  |         | sentence_0                                                                        | sentence_1                                                                       | label                                                           |
+  |:--------|:----------------------------------------------------------------------------------|:---------------------------------------------------------------------------------|:----------------------------------------------------------------|
+  | type    | string                                                                            | string                                                                           | float                                                           |
+  | details | <ul><li>min: 3 tokens</li><li>mean: 10.09 tokens</li><li>max: 25 tokens</li></ul> | <ul><li>min: 3 tokens</li><li>mean: 9.88 tokens</li><li>max: 24 tokens</li></ul> | <ul><li>min: 0.0</li><li>mean: 0.33</li><li>max: 0.85</li></ul> |
 * Samples:
-  | sentence_0                                   | sentence_1                                      | label                           |
-  |:---------------------------------------------|:------------------------------------------------|:--------------------------------|
-  | <code>Fava Beans, cooked without salt</code> | <code>Red Kidney Beans, cooked with salt</code> | <code>0.85</code>               |
-  | <code>Spaghetti squash, raw</code>           | <code>Mushrooms, white cooked</code>            | <code>0.5719985961914062</code> |
-  | <code>Chicken, back with skin roasted</code> | <code>Beef rib, roasted</code>                  | <code>0.0</code>                |
+  | sentence_0                                                     | sentence_1                                             | label                           |
+  |:---------------------------------------------------------------|:-------------------------------------------------------|:--------------------------------|
+  | <code>Instant Oats, maple and brown sugar fortified dry</code> | <code>Chocolate frosting, creamy dry mix</code>        | <code>0.0</code>                |
+  | <code>Fried Chicken Breast, meat only extra crispy KFC</code>  | <code>Brothers Natural Fruit Crisps Strawberry</code>  | <code>0.0</code>                |
+  | <code>Sesame seed dressing, regular</code>                     | <code>Italian dressing, fat-free salad dressing</code> | <code>0.7745922088623046</code> |
 * Loss: [<code>ContrastiveLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#contrastiveloss) with these parameters:
   ```json
   {
@@ -202,8 +202,8 @@ You can finetune this model on your own dataset.
 ### Training Hyperparameters
 #### Non-Default Hyperparameters
 
-- `per_device_train_batch_size`: 32
-- `per_device_eval_batch_size`: 32
+- `per_device_train_batch_size`: 128
+- `per_device_eval_batch_size`: 128
 - `num_train_epochs`: 1
 - `multi_dataset_batch_sampler`: round_robin
 
@@ -214,8 +214,8 @@ You can finetune this model on your own dataset.
 - `do_predict`: False
 - `eval_strategy`: no
 - `prediction_loss_only`: True
-- `per_device_train_batch_size`: 32
-- `per_device_eval_batch_size`: 32
+- `per_device_train_batch_size`: 128
+- `per_device_eval_batch_size`: 128
 - `per_gpu_train_batch_size`: None
 - `per_gpu_eval_batch_size`: None
 - `gradient_accumulation_steps`: 1
@@ -331,7 +331,7 @@ You can finetune this model on your own dataset.
 ### Training Logs
 | Epoch | Step | validation_spearman_cosine |
 |:-----:|:----:|:--------------------------:|
-| 1.0   | 133  | 0.8460                     |
+| 1.0   | 280  | 0.8665                     |
 
 
 ### Framework Versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:44512fb0fe9566fddc194c4ecd20617070852653233eccc58ae88f6fc48e2c73
+oid sha256:1d1cb1586ee53cc57a2af6f1c1b763aa44287d52636dcd9d832e724a9f9fec6f
 size 437967672