--- language: en license: apache-2.0 library_name: setfit tags: - setfit - sentence-transformers - text-classification - generated_from_setfit_trainer base_model: sentence-transformers/paraphrase-mpnet-base-v2 metrics: - accuracy - precision - recall - f1 widget: - text: 'in the shadows of arrakis lie one secret... dune 2 is a masterpiece the film starts exactly at the moment the first part ended. although according to villeneuve it is designed to be understood without having seen the first part, i highly recommend watching dune: part one before.the first half of the film is poetic. great shots of the desert. developing new characters (especially fremen) that we had seen briefly in the first part and, above all, we can see an evolution of some characters that we already knew. i highlight here the character of stilgar (javier bardem) and chani (zendaya), who undoubtedly gets one of the great performances of the film. bravo to villeneuve for having developed chani''s character (wasted in herbert''s novel), here chani shines until the end. bardem, is also a fundamental piece that gives freshness to the serious and majestic plot, adding dynamism and even funny moments.the second half of the film goes to paul atreides. i can''t say too much here, only that we see one of the greatest evolutions of a character that i''ve been seen in cinema. the last 45 minutes of the film are simply some of the best i have ever witnessed on a screen. everything that was harvested in dune: part one now makes complete sense. the characters of florence pugh and austin butler stand out, the first one, in the background but being a totally key character, especially for everything that is going to come (messiah); and the second one for his excellent performance as a villain (far above the baron, a villain who personally never convinced me).all of this is great, however, an element of equal importance to everything mentioned was: the music of hans zimmer. spectacular songs and sounds that make you simply feel vibrate (literally... if you see it in imax) every moment inside you. it is no coincidence that nolan''s best films have zimmer''s seal. his music creates art. his music creates blockbusters. not because it''s commercial, but because it creates emotions within you. every note, sound and song are placed at the right time and moment in the film. hans zimmer''s music is undoubtedly the specie, the water of life... of dune: part two.with all this we are left with what is not only the best science fiction film ever created, but what is (be careful what i am going to say) the best film of this 21st century (at least to date), with permission from the dark knight and interstellar.villeneuve gives us a great masterpiece. here, however, the messiah is not paul atreides, nor villeneuve himself... the messiah is the spectator. you. this film will push you to be a better you, it will make you dream big, it will remind you that, like paul, you can decide when to lead your life and inspire others. you are the chosen one to, within a simple movie, find the gold that will make you better in your daily life.this valuable lesson is not achieved by ordinary pieces of art, but only by a few works... which some of us dare to call... masterpieces.because... in the shadows of arrakis lie one secret... dune 2 is a masterpiece.' - text: does justice to the books awesome! stunning! the film follows through the spirit of the books with the ever-present internal turmoil of paul. it has the feel of mysticism of the books. the cinematography is spectacular and so is hans zimmer's music. the freemen and harkons are depicted really well and so is the planet dune. at the same time, there are many details and happenings left behind, so for people who have not read the books it's a bit difficult to follow the narrative. the movie doesn't follow thoroughly the destiny of the other of the closest friends and servents of the duke, who survived the massacre in part 1. 5 out of 8 found this helpful. was this review helpful? sign in to vote. permalink - text: epic but empty i'll precursor this by saying 1) i have not read the books 2) i am a huge denis villeneuve fan 3) i found part one underwhelming.while i find there to be many issues with this film, perhaps its biggest weakness imo is that i hold no emotional connection with any of the characters, despite the film's lengthy runtime. the protagonist's mother is written as incredibly annoying, the villains are underused, florence pugh's character has 0 likability and acts simply as a story device. bardem's character is ok and funny at times and brolin doesn't have a standout moment. zendaya's character is the sole exception to this, who is the beating heart of the film.at times, i also felt that chalamet, despite being an amazing actor, is perhaps miscast, though having not read the material, i cannot comment further on this.the pacing felt off, as it stays too long in places it shouldn't, while sacrificing moments that could linger longer. this coming from a person that enjoys long epics and slow burn films. the movie looks outstanding though feels closer to a connection of rushed story beats. the direction it takes with the main character also does not feel earned.perhaps the second biggest issue is that i do not totally buy into the world that has been created. it feels large but empty and lacking character. almost too clean and strange for the sake of it. style over substance. we stay for most of the time in one or two places which doesn't help. there were also some weird choice of cuts and jumps in scenes that hampered the flow of the film.with that said, there are a couple of epic moments that stood out, which are 1) the whole zendaya bazooka scene and 2) the giant worm riding.overall, the film is enjoyable. however without engaging characters, a believable world, and pacing issues, the film falls short of greatness, perhaps way short. - text: pure respect for the original book dune has got to be one of the hardest books to translate into a film. so much of the writing comes from within each characters head.watching denis villeneuve take every important detail and bring it to life was nothing short of pure wonder. you can really tell that dune was one of his favourite stories. his visual direction, the breathtaking audio, the beautiful sets, and the top-tier acting are all big elements in the overall production... but sometimes there's there's something else beyond that. the passion that really shines through.one of the best movies ever made. period. kudos to everyone involved in every aspect of the production. and a huge thank you to the studio for allowing these people to perform their craft to the best of their abilities. 2 out of 3 found this helpful. was this review helpful? sign in to vote. permalink - text: 'movie-making excellence!!! denis villeneuve''s dune: part two is one of the best sci-fi fantasy movies to be ever produced.the adapted screenplay is almost true to its source material. the "spice" of the book has been captured faithfully and even there are few deviations from the book, i choose to believe the same would have taken to adapt the movie in best possible manner.the soundtrack of the movie by the han zimmer elevates the mood and make each scene feel meaningful and epic. the sound design complements the visual prowess perfectly.dennis villeneue''s dune trilogy will be his magnus opus. each shot has been carefully crafted to almost perfection leaving no waste behind. we are truly blessed to witness his work.with an ensemble cast of industry veterans and hollywood''s future biggest and brightest stars, the casting of new characters can be considered as great. each actor/actress have done justice to their roles with standout performances by rebecca ferguson, zendaya, timothee and austin butler.production design, costumes and make-up is excellenta must watch for every movie-goer. experience it in imax if possible.' pipeline_tag: text-classification inference: true model-index: - name: SetFit with sentence-transformers/paraphrase-mpnet-base-v2 on data/raw/15239678.jsonl results: - task: type: text-classification name: Text Classification dataset: name: data/raw/15239678.jsonl type: unknown split: test metrics: - type: accuracy value: 0.8571428571428571 name: Accuracy - type: precision value: 0.9959514170040485 name: Precision - type: recall value: 0.8512110726643599 name: Recall - type: f1 value: 0.917910447761194 name: F1 --- # SetFit with sentence-transformers/paraphrase-mpnet-base-v2 on data/raw/15239678.jsonl This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance is used for classification. The model has been trained using an efficient few-shot learning technique that involves: 1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning. 2. Training a classification head with features from the fine-tuned Sentence Transformer. ## Model Details ### Model Description - **Model Type:** SetFit - **Sentence Transformer body:** [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2) - **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance - **Maximum Sequence Length:** 512 tokens - **Number of Classes:** 2 classes - **Language:** en - **License:** apache-2.0 ### Model Sources - **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit) - **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055) - **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit) ### Model Labels | Label | Examples | |:---------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | negative | | | positive | | ## Evaluation ### Metrics | Label | Accuracy | Precision | Recall | F1 | |:--------|:---------|:----------|:-------|:-------| | **all** | 0.8571 | 0.9960 | 0.8512 | 0.9179 | ## Uses ### Direct Use for Inference First install the SetFit library: ```bash pip install setfit ``` Then you can load this model and run inference. ```python from setfit import SetFitModel # Download from the 🤗 Hub model = SetFitModel.from_pretrained("carlesoctav/SentimentClassifierDune64shot") # Run inference preds = model("does justice to the books awesome! stunning! the film follows through the spirit of the books with the ever-present internal turmoil of paul. it has the feel of mysticism of the books. the cinematography is spectacular and so is hans zimmer's music. the freemen and harkons are depicted really well and so is the planet dune. at the same time, there are many details and happenings left behind, so for people who have not read the books it's a bit difficult to follow the narrative. the movie doesn't follow thoroughly the destiny of the other of the closest friends and servents of the duke, who survived the massacre in part 1. 5 out of 8 found this helpful. was this review helpful? sign in to vote. permalink") ``` ## Training Details ### Training Set Metrics | Training set | Min | Median | Max | |:-------------|:----|:---------|:-----| | Word count | 108 | 228.9219 | 1595 | | Label | Training Sample Count | |:---------|:----------------------| | negative | 64 | | positive | 64 | ### Training Hyperparameters - batch_size: (16, 16) - num_epochs: (1, 1) - max_steps: -1 - sampling_strategy: oversampling - body_learning_rate: (2e-05, 1e-05) - head_learning_rate: 0.01 - loss: CosineSimilarityLoss - distance_metric: cosine_distance - margin: 0.25 - end_to_end: False - use_amp: False - warmup_proportion: 0.1 - seed: 42 - eval_max_steps: -1 - load_best_model_at_end: True ### Training Results | Epoch | Step | Training Loss | Validation Loss | |:-------:|:-------:|:-------------:|:---------------:| | 0.0019 | 1 | 0.2744 | - | | 0.0962 | 50 | 0.0513 | - | | 0.1923 | 100 | 0.0022 | - | | 0.2885 | 150 | 0.0003 | - | | 0.3846 | 200 | 0.0001 | - | | 0.4808 | 250 | 0.0001 | - | | 0.5769 | 300 | 0.0001 | - | | 0.6731 | 350 | 0.0001 | - | | 0.7692 | 400 | 0.0001 | - | | 0.8654 | 450 | 0.0001 | - | | 0.9615 | 500 | 0.0001 | - | | **1.0** | **520** | **-** | **0.1845** | * The bold row denotes the saved checkpoint. ### Framework Versions - Python: 3.10.11 - SetFit: 1.0.3 - Sentence Transformers: 2.5.1 - Transformers: 4.38.2 - PyTorch: 2.0.1 - Datasets: 2.18.0 - Tokenizers: 0.15.2 ## Citation ### BibTeX ```bibtex @article{https://doi.org/10.48550/arxiv.2209.11055, doi = {10.48550/ARXIV.2209.11055}, url = {https://arxiv.org/abs/2209.11055}, author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren}, keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences}, title = {Efficient Few-Shot Learning Without Prompts}, publisher = {arXiv}, year = {2022}, copyright = {Creative Commons Attribution 4.0 International} } ```