--- datasets: - sentence-transformers/gooaq language: - en library_name: sentence-transformers license: apache-2.0 metrics: - cosine_accuracy@1 - cosine_accuracy@3 - cosine_accuracy@5 - cosine_accuracy@10 - cosine_precision@1 - cosine_precision@3 - cosine_precision@5 - cosine_precision@10 - cosine_recall@1 - cosine_recall@3 - cosine_recall@5 - cosine_recall@10 - cosine_ndcg@10 - cosine_mrr@10 - cosine_map@100 pipeline_tag: sentence-similarity tags: - sentence-transformers - sentence-similarity - feature-extraction - generated_from_trainer - dataset_size:3012496 - loss:MatryoshkaLoss - loss:MultipleNegativesRankingLoss widget: - source_sentence: how to sign legal documents as power of attorney? sentences: - 'After the principal''s name, write “by” and then sign your own name. Under or after the signature line, indicate your status as POA by including any of the following identifiers: as POA, as Agent, as Attorney in Fact or as Power of Attorney.' - '[''From the Home screen, swipe left to Apps.'', ''Tap Transfer my Data.'', ''Tap Menu (...).'', ''Tap Export to SD card.'']' - Ginger Dank Nugs (Grape) - 350mg. Feast your eyes on these unique and striking gourmet chocolates; Coco Nugs created by Ginger Dank. Crafted to resemble perfect nugs of cannabis, each of the 10 buds contains 35mg of THC. ... This is a perfect product for both cannabis and chocolate lovers, who appreciate a little twist. - source_sentence: how to delete vdom in fortigate? sentences: - Go to System -> VDOM -> VDOM2 and select 'Delete'. This VDOM is now successfully removed from the configuration. - 'Both combination birth control pills and progestin-only pills may cause headaches as a side effect. Additional side effects of birth control pills may include: breast tenderness. nausea.' - White cheese tends to show imperfections more readily and as consumers got more used to yellow-orange cheese, it became an expected option. Today, many cheddars are yellow. While most cheesemakers use annatto, some use an artificial coloring agent instead, according to Sachs. - source_sentence: where are earthquakes most likely to occur on earth? sentences: - Zelle in the Bank of the America app is a fast, safe, and easy way to send and receive money with family and friends who have a bank account in the U.S., all with no fees. Money moves in minutes directly between accounts that are already enrolled with Zelle. - It takes about 3 days for a spacecraft to reach the Moon. During that time a spacecraft travels at least 240,000 miles (386,400 kilometers) which is the distance between Earth and the Moon. - Most earthquakes occur along the edge of the oceanic and continental plates. The earth's crust (the outer layer of the planet) is made up of several pieces, called plates. The plates under the oceans are called oceanic plates and the rest are continental plates. - source_sentence: fix iphone is disabled connect to itunes without itunes? sentences: - To fix a disabled iPhone or iPad without iTunes, you have to erase your device. Click on the "Erase iPhone" option and confirm your selection. Wait for a while as the "Find My iPhone" feature will remotely erase your iOS device. Needless to say, it will also disable its lock. - How Māui brought fire to the world. One evening, after eating a hearty meal, Māui lay beside his fire staring into the flames. ... In the middle of the night, while everyone was sleeping, Māui went from village to village and extinguished all the fires until not a single fire burned in the world. - Angry Orchard makes a variety of year-round craft cider styles, including Angry Orchard Crisp Apple, a fruit-forward hard cider that balances the sweetness of culinary apples with dryness and bright acidity of bittersweet apples for a complex, refreshing taste. - source_sentence: how to reverse a video on tiktok that's not yours? sentences: - '[''Tap "Effects" at the bottom of your screen — it\''s an icon that looks like a clock. Open the Effects menu. ... '', ''At the end of the new list that appears, tap "Time." Select "Time" at the end. ... '', ''Select "Reverse" — you\''ll then see a preview of your new, reversed video appear on the screen.'']' - Franchise Facts Poke Bar has a franchise fee of up to $30,000, with a total initial investment range of $157,800 to $438,000. The initial cost of a franchise includes several fees -- Unlock this franchise to better understand the costs such as training and territory fees. - Relative age is the age of a rock layer (or the fossils it contains) compared to other layers. It can be determined by looking at the position of rock layers. Absolute age is the numeric age of a layer of rocks or fossils. Absolute age can be determined by using radiometric dating. co2_eq_emissions: emissions: 6.448001991119035 energy_consumed: 0.0165885485310573 source: codecarbon training_type: fine-tuning on_cloud: false cpu_model: 13th Gen Intel(R) Core(TM) i7-13700K ram_total_size: 31.777088165283203 hours_used: 0.109 hardware_used: 1 x NVIDIA GeForce RTX 3090 model-index: - name: Static Embeddings with BERT uncased tokenizer finetuned on GooAQ pairs results: - task: type: information-retrieval name: Information Retrieval dataset: name: gooaq 1024 dev type: gooaq-1024-dev metrics: - type: cosine_accuracy@1 value: 0.6309 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.8409 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.8986 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.9444 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.6309 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.28029999999999994 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.17972000000000002 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.09444000000000002 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.6309 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.8409 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.8986 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.9444 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.7932643237589305 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.7440336111111036 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.7465739001132767 name: Cosine Map@100 - task: type: information-retrieval name: Information Retrieval dataset: name: gooaq 512 dev type: gooaq-512-dev metrics: - type: cosine_accuracy@1 value: 0.6271 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.8366 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.8946 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.9431 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.6271 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.27886666666666665 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.17892000000000002 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.09431000000000002 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.6271 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.8366 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.8946 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.9431 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.7904860196985286 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.7408453174603101 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.7434337897783787 name: Cosine Map@100 - task: type: information-retrieval name: Information Retrieval dataset: name: gooaq 256 dev type: gooaq-256-dev metrics: - type: cosine_accuracy@1 value: 0.6192 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.8235 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.8866 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.9364 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.6192 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.27449999999999997 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.17732000000000003 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.09364000000000001 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.6192 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.8235 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.8866 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.9364 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.7821476540310974 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.7321259126984055 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.7348893313013708 name: Cosine Map@100 - task: type: information-retrieval name: Information Retrieval dataset: name: gooaq 128 dev type: gooaq-128-dev metrics: - type: cosine_accuracy@1 value: 0.5942 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.804 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.8721 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.9249 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.5942 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.268 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.17442000000000002 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.09249 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.5942 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.804 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.8721 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.9249 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.7627845665665897 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.7103426587301529 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.7133975871277517 name: Cosine Map@100 - task: type: information-retrieval name: Information Retrieval dataset: name: gooaq 64 dev type: gooaq-64-dev metrics: - type: cosine_accuracy@1 value: 0.556 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.7553 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.8267 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.8945 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.556 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.25176666666666664 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.16534000000000001 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.08945 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.556 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.7553 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.8267 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.8945 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.7246435400765202 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.6701957142857087 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.6743443703166442 name: Cosine Map@100 - task: type: information-retrieval name: Information Retrieval dataset: name: gooaq 32 dev type: gooaq-32-dev metrics: - type: cosine_accuracy@1 value: 0.4628 name: Cosine Accuracy@1 - type: cosine_accuracy@3 value: 0.6619 name: Cosine Accuracy@3 - type: cosine_accuracy@5 value: 0.7415 name: Cosine Accuracy@5 - type: cosine_accuracy@10 value: 0.8241 name: Cosine Accuracy@10 - type: cosine_precision@1 value: 0.4628 name: Cosine Precision@1 - type: cosine_precision@3 value: 0.2206333333333333 name: Cosine Precision@3 - type: cosine_precision@5 value: 0.1483 name: Cosine Precision@5 - type: cosine_precision@10 value: 0.08241 name: Cosine Precision@10 - type: cosine_recall@1 value: 0.4628 name: Cosine Recall@1 - type: cosine_recall@3 value: 0.6619 name: Cosine Recall@3 - type: cosine_recall@5 value: 0.7415 name: Cosine Recall@5 - type: cosine_recall@10 value: 0.8241 name: Cosine Recall@10 - type: cosine_ndcg@10 value: 0.6387155548290799 name: Cosine Ndcg@10 - type: cosine_mrr@10 value: 0.5797731349206319 name: Cosine Mrr@10 - type: cosine_map@100 value: 0.5857231820662888 name: Cosine Map@100 --- # Static Embeddings with BERT uncased tokenizer finetuned on GooAQ pairs This is a [sentence-transformers](https://www.SBERT.net) model trained on the [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) dataset. It maps sentences & paragraphs to a 1024-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more. This model was trained using the [train_script.py](train_script.py) code. ## Model Details ### Model Description - **Model Type:** Sentence Transformer - **Maximum Sequence Length:** inf tokens - **Output Dimensionality:** 1024 tokens - **Similarity Function:** Cosine Similarity - **Training Dataset:** - [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) - **Language:** en - **License:** apache-2.0 ### Model Sources - **Documentation:** [Sentence Transformers Documentation](https://sbert.net) - **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers) - **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers) ### Full Model Architecture ``` SentenceTransformer( (0): StaticEmbedding( (embedding): EmbeddingBag(30522, 1024, mode='mean') ) ) ``` ## Usage ### Direct Usage (Sentence Transformers) First install the Sentence Transformers library: ```bash pip install -U sentence-transformers ``` Then you can load this model and run inference. ```python from sentence_transformers import SentenceTransformer # Download from the 🤗 Hub model = SentenceTransformer("tomaarsen/static-bert-uncased-gooaq") # Run inference sentences = [ "how to reverse a video on tiktok that's not yours?", '[\'Tap "Effects" at the bottom of your screen — it\\\'s an icon that looks like a clock. Open the Effects menu. ... \', \'At the end of the new list that appears, tap "Time." Select "Time" at the end. ... \', \'Select "Reverse" — you\\\'ll then see a preview of your new, reversed video appear on the screen.\']', 'Relative age is the age of a rock layer (or the fossils it contains) compared to other layers. It can be determined by looking at the position of rock layers. Absolute age is the numeric age of a layer of rocks or fossils. Absolute age can be determined by using radiometric dating.', ] embeddings = model.encode(sentences) print(embeddings.shape) # [3, 1024] # Get the similarity scores for the embeddings similarities = model.similarity(embeddings, embeddings) print(similarities.shape) # [3, 3] ``` ## Evaluation ### Metrics #### Information Retrieval * Dataset: `gooaq-1024-dev` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.6309 | | cosine_accuracy@3 | 0.8409 | | cosine_accuracy@5 | 0.8986 | | cosine_accuracy@10 | 0.9444 | | cosine_precision@1 | 0.6309 | | cosine_precision@3 | 0.2803 | | cosine_precision@5 | 0.1797 | | cosine_precision@10 | 0.0944 | | cosine_recall@1 | 0.6309 | | cosine_recall@3 | 0.8409 | | cosine_recall@5 | 0.8986 | | cosine_recall@10 | 0.9444 | | cosine_ndcg@10 | 0.7933 | | cosine_mrr@10 | 0.744 | | **cosine_map@100** | **0.7466** | #### Information Retrieval * Dataset: `gooaq-512-dev` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.6271 | | cosine_accuracy@3 | 0.8366 | | cosine_accuracy@5 | 0.8946 | | cosine_accuracy@10 | 0.9431 | | cosine_precision@1 | 0.6271 | | cosine_precision@3 | 0.2789 | | cosine_precision@5 | 0.1789 | | cosine_precision@10 | 0.0943 | | cosine_recall@1 | 0.6271 | | cosine_recall@3 | 0.8366 | | cosine_recall@5 | 0.8946 | | cosine_recall@10 | 0.9431 | | cosine_ndcg@10 | 0.7905 | | cosine_mrr@10 | 0.7408 | | **cosine_map@100** | **0.7434** | #### Information Retrieval * Dataset: `gooaq-256-dev` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.6192 | | cosine_accuracy@3 | 0.8235 | | cosine_accuracy@5 | 0.8866 | | cosine_accuracy@10 | 0.9364 | | cosine_precision@1 | 0.6192 | | cosine_precision@3 | 0.2745 | | cosine_precision@5 | 0.1773 | | cosine_precision@10 | 0.0936 | | cosine_recall@1 | 0.6192 | | cosine_recall@3 | 0.8235 | | cosine_recall@5 | 0.8866 | | cosine_recall@10 | 0.9364 | | cosine_ndcg@10 | 0.7821 | | cosine_mrr@10 | 0.7321 | | **cosine_map@100** | **0.7349** | #### Information Retrieval * Dataset: `gooaq-128-dev` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.5942 | | cosine_accuracy@3 | 0.804 | | cosine_accuracy@5 | 0.8721 | | cosine_accuracy@10 | 0.9249 | | cosine_precision@1 | 0.5942 | | cosine_precision@3 | 0.268 | | cosine_precision@5 | 0.1744 | | cosine_precision@10 | 0.0925 | | cosine_recall@1 | 0.5942 | | cosine_recall@3 | 0.804 | | cosine_recall@5 | 0.8721 | | cosine_recall@10 | 0.9249 | | cosine_ndcg@10 | 0.7628 | | cosine_mrr@10 | 0.7103 | | **cosine_map@100** | **0.7134** | #### Information Retrieval * Dataset: `gooaq-64-dev` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.556 | | cosine_accuracy@3 | 0.7553 | | cosine_accuracy@5 | 0.8267 | | cosine_accuracy@10 | 0.8945 | | cosine_precision@1 | 0.556 | | cosine_precision@3 | 0.2518 | | cosine_precision@5 | 0.1653 | | cosine_precision@10 | 0.0895 | | cosine_recall@1 | 0.556 | | cosine_recall@3 | 0.7553 | | cosine_recall@5 | 0.8267 | | cosine_recall@10 | 0.8945 | | cosine_ndcg@10 | 0.7246 | | cosine_mrr@10 | 0.6702 | | **cosine_map@100** | **0.6743** | #### Information Retrieval * Dataset: `gooaq-32-dev` * Evaluated with [InformationRetrievalEvaluator](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.InformationRetrievalEvaluator) | Metric | Value | |:--------------------|:-----------| | cosine_accuracy@1 | 0.4628 | | cosine_accuracy@3 | 0.6619 | | cosine_accuracy@5 | 0.7415 | | cosine_accuracy@10 | 0.8241 | | cosine_precision@1 | 0.4628 | | cosine_precision@3 | 0.2206 | | cosine_precision@5 | 0.1483 | | cosine_precision@10 | 0.0824 | | cosine_recall@1 | 0.4628 | | cosine_recall@3 | 0.6619 | | cosine_recall@5 | 0.7415 | | cosine_recall@10 | 0.8241 | | cosine_ndcg@10 | 0.6387 | | cosine_mrr@10 | 0.5798 | | **cosine_map@100** | **0.5857** | ## Training Details ### Training Dataset #### gooaq * Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c) * Size: 3,012,496 training samples * Columns: question and answer * Approximate statistics based on the first 1000 samples: | | question | answer | |:--------|:-----------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | question | answer | |:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | what is the difference between broilers and layers? | An egg laying poultry is called egger or layer whereas broilers are reared for obtaining meat. So a layer should be able to produce more number of large sized eggs, without growing too much. On the other hand, a broiler should yield more meat and hence should be able to grow well. | | what is the difference between chronological order and spatial order? | As a writer, you should always remember that unlike chronological order and the other organizational methods for data, spatial order does not take into account the time. Spatial order is primarily focused on the location. All it does is take into account the location of objects and not the time. | | is kamagra same as viagra? | Kamagra is thought to contain the same active ingredient as Viagra, sildenafil citrate. In theory, it should work in much the same way as Viagra, taking about 45 minutes to take effect, and lasting for around 4-6 hours. However, this will vary from person to person. | * Loss: [MatryoshkaLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters: ```json { "loss": "MultipleNegativesRankingLoss", "matryoshka_dims": [ 1024, 512, 256, 128, 64, 32 ], "matryoshka_weights": [ 1, 1, 1, 1, 1, 1 ], "n_dims_per_step": -1 } ``` ### Evaluation Dataset #### gooaq * Dataset: [gooaq](https://huggingface.co/datasets/sentence-transformers/gooaq) at [b089f72](https://huggingface.co/datasets/sentence-transformers/gooaq/tree/b089f728748a068b7bc5234e5bcf5b25e3c8279c) * Size: 3,012,496 evaluation samples * Columns: question and answer * Approximate statistics based on the first 1000 samples: | | question | answer | |:--------|:-----------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------| | type | string | string | | details | | | * Samples: | question | answer | |:-----------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | how do i program my directv remote with my tv? | ['Press MENU on your remote.', 'Select Settings & Help > Settings > Remote Control > Program Remote.', 'Choose the device (TV, audio, DVD) you wish to program. ... ', 'Follow the on-screen prompts to complete programming.'] | | are rodrigues fruit bats nocturnal? | Before its numbers were threatened by habitat destruction, storms, and hunting, some of those groups could number 500 or more members. Sunrise, sunset. Rodrigues fruit bats are most active at dawn, at dusk, and at night. | | why does your heart rate increase during exercise bbc bitesize? | During exercise there is an increase in physical activity and muscle cells respire more than they do when the body is at rest. The heart rate increases during exercise. The rate and depth of breathing increases - this makes sure that more oxygen is absorbed into the blood, and more carbon dioxide is removed from it. | * Loss: [MatryoshkaLoss](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#matryoshkaloss) with these parameters: ```json { "loss": "MultipleNegativesRankingLoss", "matryoshka_dims": [ 1024, 512, 256, 128, 64, 32 ], "matryoshka_weights": [ 1, 1, 1, 1, 1, 1 ], "n_dims_per_step": -1 } ``` ### Training Hyperparameters #### Non-Default Hyperparameters - `eval_strategy`: steps - `per_device_train_batch_size`: 2048 - `per_device_eval_batch_size`: 2048 - `learning_rate`: 0.2 - `num_train_epochs`: 1 - `warmup_ratio`: 0.1 - `bf16`: True - `batch_sampler`: no_duplicates #### All Hyperparameters
Click to expand - `overwrite_output_dir`: False - `do_predict`: False - `eval_strategy`: steps - `prediction_loss_only`: True - `per_device_train_batch_size`: 2048 - `per_device_eval_batch_size`: 2048 - `per_gpu_train_batch_size`: None - `per_gpu_eval_batch_size`: None - `gradient_accumulation_steps`: 1 - `eval_accumulation_steps`: None - `torch_empty_cache_steps`: None - `learning_rate`: 0.2 - `weight_decay`: 0.0 - `adam_beta1`: 0.9 - `adam_beta2`: 0.999 - `adam_epsilon`: 1e-08 - `max_grad_norm`: 1.0 - `num_train_epochs`: 1 - `max_steps`: -1 - `lr_scheduler_type`: linear - `lr_scheduler_kwargs`: {} - `warmup_ratio`: 0.1 - `warmup_steps`: 0 - `log_level`: passive - `log_level_replica`: warning - `log_on_each_node`: True - `logging_nan_inf_filter`: True - `save_safetensors`: True - `save_on_each_node`: False - `save_only_model`: False - `restore_callback_states_from_checkpoint`: False - `no_cuda`: False - `use_cpu`: False - `use_mps_device`: False - `seed`: 42 - `data_seed`: None - `jit_mode_eval`: False - `use_ipex`: False - `bf16`: True - `fp16`: False - `fp16_opt_level`: O1 - `half_precision_backend`: auto - `bf16_full_eval`: False - `fp16_full_eval`: False - `tf32`: None - `local_rank`: 0 - `ddp_backend`: None - `tpu_num_cores`: None - `tpu_metrics_debug`: False - `debug`: [] - `dataloader_drop_last`: False - `dataloader_num_workers`: 0 - `dataloader_prefetch_factor`: None - `past_index`: -1 - `disable_tqdm`: False - `remove_unused_columns`: True - `label_names`: None - `load_best_model_at_end`: False - `ignore_data_skip`: False - `fsdp`: [] - `fsdp_min_num_params`: 0 - `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False} - `fsdp_transformer_layer_cls_to_wrap`: None - `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None} - `deepspeed`: None - `label_smoothing_factor`: 0.0 - `optim`: adamw_torch - `optim_args`: None - `adafactor`: False - `group_by_length`: False - `length_column_name`: length - `ddp_find_unused_parameters`: None - `ddp_bucket_cap_mb`: None - `ddp_broadcast_buffers`: False - `dataloader_pin_memory`: True - `dataloader_persistent_workers`: False - `skip_memory_metrics`: True - `use_legacy_prediction_loop`: False - `push_to_hub`: False - `resume_from_checkpoint`: None - `hub_model_id`: None - `hub_strategy`: every_save - `hub_private_repo`: False - `hub_always_push`: False - `gradient_checkpointing`: False - `gradient_checkpointing_kwargs`: None - `include_inputs_for_metrics`: False - `eval_do_concat_batches`: True - `fp16_backend`: auto - `push_to_hub_model_id`: None - `push_to_hub_organization`: None - `mp_parameters`: - `auto_find_batch_size`: False - `full_determinism`: False - `torchdynamo`: None - `ray_scope`: last - `ddp_timeout`: 1800 - `torch_compile`: False - `torch_compile_backend`: None - `torch_compile_mode`: None - `dispatch_batches`: None - `split_batches`: None - `include_tokens_per_second`: False - `include_num_input_tokens_seen`: False - `neftune_noise_alpha`: None - `optim_target_modules`: None - `batch_eval_metrics`: False - `eval_on_start`: False - `eval_use_gather_object`: False - `batch_sampler`: no_duplicates - `multi_dataset_batch_sampler`: proportional
### Training Logs | Epoch | Step | Training Loss | Validation Loss | gooaq-1024-dev_cosine_map@100 | gooaq-512-dev_cosine_map@100 | gooaq-256-dev_cosine_map@100 | gooaq-128-dev_cosine_map@100 | gooaq-64-dev_cosine_map@100 | gooaq-32-dev_cosine_map@100 | |:------:|:----:|:-------------:|:---------------:|:-----------------------------:|:----------------------------:|:----------------------------:|:----------------------------:|:---------------------------:|:---------------------------:| | 0 | 0 | - | - | 0.2095 | 0.2010 | 0.1735 | 0.1381 | 0.0750 | 0.0331 | | 0.0007 | 1 | 34.953 | - | - | - | - | - | - | - | | 0.0682 | 100 | 16.2504 | - | - | - | - | - | - | - | | 0.1363 | 200 | 5.9502 | - | - | - | - | - | - | - | | 0.1704 | 250 | - | 1.6781 | 0.6791 | 0.6729 | 0.6619 | 0.6409 | 0.5904 | 0.4934 | | 0.2045 | 300 | 4.8411 | - | - | - | - | - | - | - | | 0.2727 | 400 | 4.336 | - | - | - | - | - | - | - | | 0.3408 | 500 | 4.0484 | 1.3935 | 0.7104 | 0.7055 | 0.6968 | 0.6756 | 0.6322 | 0.5358 | | 0.4090 | 600 | 3.8378 | - | - | - | - | - | - | - | | 0.4772 | 700 | 3.6765 | - | - | - | - | - | - | - | | 0.5112 | 750 | - | 1.2549 | 0.7246 | 0.7216 | 0.7133 | 0.6943 | 0.6482 | 0.5582 | | 0.5453 | 800 | 3.5439 | - | - | - | - | - | - | - | | 0.6135 | 900 | 3.4284 | - | - | - | - | - | - | - | | 0.6817 | 1000 | 3.3576 | 1.1656 | 0.7359 | 0.7338 | 0.7252 | 0.7040 | 0.6604 | 0.5715 | | 0.7498 | 1100 | 3.2456 | - | - | - | - | - | - | - | | 0.8180 | 1200 | 3.2014 | - | - | - | - | - | - | - | | 0.8521 | 1250 | - | 1.1133 | 0.7438 | 0.7398 | 0.7310 | 0.7099 | 0.6704 | 0.5796 | | 0.8862 | 1300 | 3.1536 | - | - | - | - | - | - | - | | 0.9543 | 1400 | 3.0696 | - | - | - | - | - | - | - | | 1.0 | 1467 | - | - | 0.7466 | 0.7434 | 0.7349 | 0.7134 | 0.6743 | 0.5857 | ### Environmental Impact Carbon emissions were measured using [CodeCarbon](https://github.com/mlco2/codecarbon). - **Energy Consumed**: 0.017 kWh - **Carbon Emitted**: 0.006 kg of CO2 - **Hours Used**: 0.109 hours ### Training Hardware - **On Cloud**: No - **GPU Model**: 1 x NVIDIA GeForce RTX 3090 - **CPU Model**: 13th Gen Intel(R) Core(TM) i7-13700K - **RAM Size**: 31.78 GB ### Framework Versions - Python: 3.11.6 - Sentence Transformers: 3.2.0.dev0 - Transformers: 4.43.4 - PyTorch: 2.5.0.dev20240807+cu121 - Accelerate: 0.31.0 - Datasets: 2.20.0 - Tokenizers: 0.19.1 ## Citation ### BibTeX #### Sentence Transformers ```bibtex @inproceedings{reimers-2019-sentence-bert, title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks", author = "Reimers, Nils and Gurevych, Iryna", booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing", month = "11", year = "2019", publisher = "Association for Computational Linguistics", url = "https://arxiv.org/abs/1908.10084", } ``` #### MatryoshkaLoss ```bibtex @misc{kusupati2024matryoshka, title={Matryoshka Representation Learning}, author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi}, year={2024}, eprint={2205.13147}, archivePrefix={arXiv}, primaryClass={cs.LG} } ``` #### MultipleNegativesRankingLoss ```bibtex @misc{henderson2017efficient, title={Efficient Natural Language Response Suggestion for Smart Reply}, author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil}, year={2017}, eprint={1705.00652}, archivePrefix={arXiv}, primaryClass={cs.CL} } ```