Datasets:

librarian-bots

/

model_cards_with_metadata_with_embeddings

like 5

textattack/bert-base-uncased-rotten-tomatoes

textattack

transformers

text-classification

## TextAttack Model Card This `bert-base-uncased` model was fine-tuned for sequence classificationusing TextAttack and the rotten_tomatoes dataset loaded using the `nlp` library. The model was fine-tuned for 10 epochs with a batch size of 16, a learning rate of 2e-05, and a maximum sequence length of 128. Since this was a classification task, the model was trained with a cross-entropy loss function. The best score the model achieved on this task was 0.875234521575985, as measured by the eval set accuracy, found after 4 epochs. For more information, check out [TextAttack on Github](https://github.com/QData/TextAttack).

microsoft/xclip-base-patch16-zero-shot

microsoft

transformers

feature-extraction

--- language: en license: mit tags: - vision - video-classification model-index: - name: nielsr/xclip-base-patch16-zero-shot results: - task: type: video-classification dataset: name: HMDB-51 type: hmdb-51 metrics: - type: top-1 accuracy value: 44.6 - task: type: video-classification dataset: name: UCF101 type: ucf101 metrics: - type: top-1 accuracy value: 72.0 - task: type: video-classification dataset: name: Kinetics-600 type: kinetics600 metrics: - type: top-1 accuracy value: 65.2 --- # X-CLIP (base-sized model) X-CLIP model (base-sized, patch resolution of 16) trained on [Kinetics-400](https://www.deepmind.com/open-source/kinetics). It was introduced in the paper [Expanding Language-Image Pretrained Models for General Video Recognition](https://arxiv.org/abs/2208.02816) by Ni et al. and first released in [this repository](https://github.com/microsoft/VideoX/tree/master/X-CLIP). This model was trained using 32 frames per video, at a resolution of 224x224. Disclaimer: The team releasing X-CLIP did not write a model card for this model so this model card has been written by the Hugging Face team. ## Model description X-CLIP is a minimal extension of [CLIP](https://huggingface.co/docs/transformers/model_doc/clip) for general video-language understanding. The model is trained in a contrastive way on (video, text) pairs.  This allows the model to be used for tasks like zero-shot, few-shot or fully supervised video classification and video-text retrieval. ## Intended uses & limitations You can use the raw model for determining how well text goes with a given video. See the [model hub](https://huggingface.co/models?search=microsoft/xclip) to look for fine-tuned versions on a task that interests you. ### How to use For code examples, we refer to the [documentation](https://huggingface.co/transformers/main/model_doc/xclip.html#). ## Training data This model was trained on [Kinetics 400](https://www.deepmind.com/open-source/kinetics). ### Preprocessing The exact details of preprocessing during training can be found [here](https://github.com/microsoft/VideoX/blob/40f6d177e0a057a50ac69ac1de6b5938fd268601/X-CLIP/datasets/build.py#L247). The exact details of preprocessing during validation can be found [here](https://github.com/microsoft/VideoX/blob/40f6d177e0a057a50ac69ac1de6b5938fd268601/X-CLIP/datasets/build.py#L285). During validation, one resizes the shorter edge of each frame, after which center cropping is performed to a fixed-size resolution (like 224x224). Next, frames are normalized across the RGB channels with the ImageNet mean and standard deviation. ## Evaluation results This model achieves a zero-shot top-1 accuracy of 44.6% on HMDB-51, 72.0% on UCF-101 and 65.2% on Kinetics-600.

allenai/led-base-16384

allenai

transformers

text2text-generation

--- language: en license: apache-2.0 --- ## Introduction [Allenai's Longformer Encoder-Decoder (LED)](https://github.com/allenai/longformer#longformer). As described in [Longformer: The Long-Document Transformer](https://arxiv.org/pdf/2004.05150.pdf) by Iz Beltagy, Matthew E. Peters, Arman Cohan, *led-base-16384* was initialized from [*bart-base*](https://huggingface.co/facebook/bart-base) since both models share the exact same architecture. To be able to process 16K tokens, *bart-base*'s position embedding matrix was simply copied 16 times. This model is especially interesting for long-range summarization and question answering. ## Fine-tuning for down-stream task [This notebook](https://colab.research.google.com/drive/12LjJazBl7Gam0XBPy_y0CTOJZeZ34c2v?usp=sharing) shows how *led-base-16384* can effectively be fine-tuned on a downstream task.

stabilityai/stable-diffusion-xl-base-0.9

stabilityai

diffusers

text-to-image

--- license: other extra_gated_prompt: >- Copyright (c) Stability AI Ltd. This License Agreement (as may be amended in accordance with this License Agreement, “License”), between you, or your employer or other entity (if you are entering into this agreement on behalf of your employer or other entity) (“Licensee” or “you”) and Stability AI Ltd. (“Stability AI” or “we”) applies to your use of any computer program, algorithm, source code, object code, software, models, or model weights that is made available by Stability AI under this License (“Software”) and any specifications, manuals, documentation, and other written information provided by Stability AI related to the Software (“Documentation”). By using the Software, you agree to the terms of this License. If you do not agree to this License, then you do not have any rights to use the Software or Documentation (collectively, the “Software Products”), and you must immediately cease using the Software Products. If you are agreeing to be bound by the terms of this License on behalf of your employer or other entity, you represent and warrant to Stability AI that you have full legal authority to bind your employer or such entity to this License. If you do not have the requisite authority, you may not accept the License or access the Software Products on behalf of your employer or other entity. 1. LICENSE GRANT a. Subject to your compliance with the Documentation and Sections 2, 3, and 5, Stability AI grants you a non-exclusive, worldwide, non-transferable, non-sublicensable, revocable, royalty free and limited license under Stability AI’s copyright interests to use, reproduce, and create derivative works of the Software solely for your non-commercial research purposes. The foregoing license is personal to you, and you may not assign, sublicense, distribute, publish, host, or otherwise make available this Software, derivative works of the Software, models or model weights associated with the Software, this License, or any other rights or obligations under this License without Stability AI’s prior written consent; any such assignment or sublicense without Stability AI’s prior written consent will be void and will automatically and immediately terminate this License. For sake of clarity, this License does not grant to you the right or ability to extend any license to the Software, derivative works of the Software, or associated models or model weights to a non-Licensee, nor does this License permit you to create a new Licensee, such as by making available a copy of this License. If you would like rights not granted by this License, you may seek permission by sending an email to legal@stability.ai. b. You may make a reasonable number of copies of the Documentation solely for your use in connection with the license to the Software granted above. c. The grant of rights expressly set forth in this Section 1 (License Grant) are the complete grant of rights to you in the Software Products, and no other licenses are granted, whether by waiver, estoppel, implication, equity or otherwise. Stability AI and its licensors reserve all rights not expressly granted by this License. 2. RESTRICTIONS You will not, and will not permit, assist or cause any third party to: a. use, modify, copy, reproduce, create derivative works of, or distribute the Software Products (or any derivative works thereof, works incorporating the Software Products, or any data produced by the Software), in whole or in part, for (i) any commercial or production purposes, (ii) military purposes or in the service of nuclear technology, (iii) purposes of surveillance, including any research or development relating to surveillance, (iv) biometric processing, (v) in any manner that infringes, misappropriates, or otherwise violates any third-party rights, or (vi) in any manner that violates any applicable law and violating any privacy or security laws, rules, regulations, directives, or governmental requirements (including the General Data Privacy Regulation (Regulation (EU) 2016/679), the California Consumer Privacy Act, and any and all laws governing the processing of biometric information), as well as all amendments and successor laws to any of the foregoing; b. alter or remove copyright and other proprietary notices which appear on or in the Software Products; c. utilize any equipment, device, software, or other means to circumvent or remove any security or protection used by Stability AI in connection with the Software, or to circumvent or remove any usage restrictions, or to enable functionality disabled by Stability AI; or d. offer or impose any terms on the Software Products that alter, restrict, or are inconsistent with the terms of this License. e. 1) violate any applicable U.S. and non-U.S. export control and trade sanctions laws (“Export Laws”); 2) directly or indirectly export, re-export, provide, or otherwise transfer Software Products: (a) to any individual, entity, or country prohibited by Export Laws; (b) to anyone on U.S. or non-U.S. government restricted parties lists; or (c) for any purpose prohibited by Export Laws, including nuclear, chemical or biological weapons, or missile technology applications; 3) use or download Software Products if you or they are: (a) located in a comprehensively sanctioned jurisdiction, (b) currently listed on any U.S. or non-U.S. restricted parties list, or (c) for any purpose prohibited by Export Laws; and (4) will not disguise your location through IP proxying or other methods. 3. ATTRIBUTION Together with any copies of the Software Products (as well as derivative works thereof or works incorporating the Software Products) that you distribute, you must provide (i) a copy of this License, and (ii) the following attribution notice: “SDXL 0.9 is licensed under the SDXL Research License, Copyright (c) Stability AI Ltd. All Rights Reserved.” 4. DISCLAIMERS THE SOFTWARE PRODUCTS ARE PROVIDED “AS IS” AND “WITH ALL FAULTS” WITH NO WARRANTY OF ANY KIND, EXPRESS OR IMPLIED. STABILITY AI EXPRESSLY DISCLAIMS ALL REPRESENTATIONS AND WARRANTIES, EXPRESS OR IMPLIED, WHETHER BY STATUTE, CUSTOM, USAGE OR OTHERWISE AS TO ANY MATTERS RELATED TO THE SOFTWARE PRODUCTS, INCLUDING BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, SATISFACTORY QUALITY, OR NON-INFRINGEMENT. STABILITY AI MAKES NO WARRANTIES OR REPRESENTATIONS THAT THE SOFTWARE PRODUCTS WILL BE ERROR FREE OR FREE OF VIRUSES OR OTHER HARMFUL COMPONENTS, OR PRODUCE ANY PARTICULAR RESULTS. 5. LIMITATION OF LIABILITY TO THE FULLEST EXTENT PERMITTED BY LAW, IN NO EVENT WILL STABILITY AI BE LIABLE TO YOU (A) UNDER ANY THEORY OF LIABILITY, WHETHER BASED IN CONTRACT, TORT, NEGLIGENCE, STRICT LIABILITY, WARRANTY, OR OTHERWISE UNDER THIS LICENSE, OR (B) FOR ANY INDIRECT, CONSEQUENTIAL, EXEMPLARY, INCIDENTAL, PUNITIVE OR SPECIAL DAMAGES OR LOST PROFITS, EVEN IF STABILITY AI HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. THE SOFTWARE PRODUCTS, THEIR CONSTITUENT COMPONENTS, AND ANY OUTPUT (COLLECTIVELY, “SOFTWARE MATERIALS”) ARE NOT DESIGNED OR INTENDED FOR USE IN ANY APPLICATION OR SITUATION WHERE FAILURE OR FAULT OF THE SOFTWARE MATERIALS COULD REASONABLY BE ANTICIPATED TO LEAD TO SERIOUS INJURY OF ANY PERSON, INCLUDING POTENTIAL DISCRIMINATION OR VIOLATION OF AN INDIVIDUAL’S PRIVACY RIGHTS, OR TO SEVERE PHYSICAL, PROPERTY, OR ENVIRONMENTAL DAMAGE (EACH, A “HIGH-RISK USE”). IF YOU ELECT TO USE ANY OF THE SOFTWARE MATERIALS FOR A HIGH-RISK USE, YOU DO SO AT YOUR OWN RISK. YOU AGREE TO DESIGN AND IMPLEMENT APPROPRIATE DECISION-MAKING AND RISK-MITIGATION PROCEDURES AND POLICIES IN CONNECTION WITH A HIGH-RISK USE SUCH THAT EVEN IF THERE IS A FAILURE OR FAULT IN ANY OF THE SOFTWARE MATERIALS, THE SAFETY OF PERSONS OR PROPERTY AFFECTED BY THE ACTIVITY STAYS AT A LEVEL THAT IS REASONABLE, APPROPRIATE, AND LAWFUL FOR THE FIELD OF THE HIGH-RISK USE. 6. INDEMNIFICATION You will indemnify, defend and hold harmless Stability AI and our subsidiaries and affiliates, and each of our respective shareholders, directors, officers, employees, agents, successors, and assigns (collectively, the “Stability AI Parties”) from and against any losses, liabilities, damages, fines, penalties, and expenses (including reasonable attorneys’ fees) incurred by any Stability AI Party in connection with any claim, demand, allegation, lawsuit, proceeding, or investigation (collectively, “Claims”) arising out of or related to: (a) your access to or use of the Software Products (as well as any results or data generated from such access or use), including any High-Risk Use (defined below); (b) your violation of this License; or (c) your violation, misappropriation or infringement of any rights of another (including intellectual property or other proprietary rights and privacy rights). You will promptly notify the Stability AI Parties of any such Claims, and cooperate with Stability AI Parties in defending such Claims. You will also grant the Stability AI Parties sole control of the defense or settlement, at Stability AI’s sole option, of any Claims. This indemnity is in addition to, and not in lieu of, any other indemnities or remedies set forth in a written agreement between you and Stability AI or the other Stability AI Parties. 7. TERMINATION; SURVIVAL a. This License will automatically terminate upon any breach by you of the terms of this License. b. We may terminate this License, in whole or in part, at any time upon notice (including electronic) to you. c. The following sections survive termination of this License: 2 (Restrictions), 3 (Attribution), 4 (Disclaimers), 5 (Limitation on Liability), 6 (Indemnification) 7 (Termination; Survival), 8 (Third Party Materials), 9 (Trademarks), 10 (Applicable Law; Dispute Resolution), and 11 (Miscellaneous). 8. THIRD PARTY MATERIALS The Software Products may contain third-party software or other components (including free and open source software) (all of the foregoing, “Third Party Materials”), which are subject to the license terms of the respective third-party licensors. Your dealings or correspondence with third parties and your use of or interaction with any Third Party Materials are solely between you and the third party. Stability AI does not control or endorse, and makes no representations or warranties regarding, any Third Party Materials, and your access to and use of such Third Party Materials are at your own risk. 9. TRADEMARKS Licensee has not been granted any trademark license as part of this License and may not use any name or mark associated with Stability AI without the prior written permission of Stability AI, except to the extent necessary to make the reference required by the “ATTRIBUTION” section of this Agreement. 10. APPLICABLE LAW; DISPUTE RESOLUTION This License will be governed and construed under the laws of the State of California without regard to conflicts of law provisions. Any suit or proceeding arising out of or relating to this License will be brought in the federal or state courts, as applicable, in San Mateo County, California, and each party irrevocably submits to the jurisdiction and venue of such courts. 11. MISCELLANEOUS If any provision or part of a provision of this License is unlawful, void or unenforceable, that provision or part of the provision is deemed severed from this License, and will not affect the validity and enforceability of any remaining provisions. The failure of Stability AI to exercise or enforce any right or provision of this License will not operate as a waiver of such right or provision. This License does not confer any third-party beneficiary rights upon any other person or entity. This License, together with the Documentation, contains the entire understanding between you and Stability AI regarding the subject matter of this License, and supersedes all other written or oral agreements and understandings between you and Stability AI regarding such subject matter. No change or addition to any provision of this License will be binding unless it is in writing and signed by an authorized representative of both you and Stability AI. extra_gated_heading: Researcher Early Access extra_gated_description: SDXL 0.9 Research License Agreement extra_gated_button_content: Submit application extra_gated_fields: Organization: text Nature of research: text Personal researcher link (CV, website, github): text Other Comments: text I accept the above license agreement, and will use the Software non-commercially and for research purposes only: checkbox tags: - text-to-image - stable-diffusion --- # SD-XL 0.9-base Model Card  ## Model  SDXL consists of a two-step pipeline for latent diffusion: First, we use a base model to generate latents of the desired output size. In the second step, we use a specialized high-resolution model and apply a technique called SDEdit (https://arxiv.org/abs/2108.01073, also known as "img2img") to the latents generated in the first step, using the same prompt. ### Model Description - **Developed by:** Stability AI - **Model type:** Diffusion-based text-to-image generative model - **License:** [SDXL 0.9 Research License](https://huggingface.co/stabilityai/stable-diffusion-xl-base-0.9/blob/main/LICENSE.md) - **Model Description:** This is a model that can be used to generate and modify images based on text prompts. It is a [Latent Diffusion Model](https://arxiv.org/abs/2112.10752) that uses two fixed, pretrained text encoders ([OpenCLIP-ViT/G](https://github.com/mlfoundations/open_clip) and [CLIP-ViT/L](https://github.com/openai/CLIP/tree/main)). - **Resources for more information:** [GitHub Repository](https://github.com/Stability-AI/generative-models) [SDXL paper on arXiv](https://arxiv.org/abs/2307.01952). ### Model Sources <!-- Provide the basic links for the model. --> - **Repository:** https://github.com/Stability-AI/generative-models - **Demo [optional]:** https://clipdrop.co/stable-diffusion ### 🧨 Diffusers Make sure to upgrade diffusers to >= 0.18.0: ``` pip install diffusers --upgrade ``` In addition make sure to install `transformers`, `safetensors`, `accelerate` as well as the invisible watermark: ``` pip install invisible_watermark transformers accelerate safetensors ``` You can use the model then as follows ```py from diffusers import DiffusionPipeline import torch pipe = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-xl-base-0.9", torch_dtype=torch.float16, use_safetensors=True, variant="fp16") pipe.to("cuda") # if using torch < 2.0 # pipe.enable_xformers_memory_efficient_attention() prompt = "An astronaut riding a green horse" images = pipe(prompt=prompt).images[0] ``` When using `torch >= 2.0`, you can improve the inference speed by 20-30% with torch.compile. Simple wrap the unet with torch compile before running the pipeline: ```py pipe.unet = torch.compile(pipe.unet, mode="reduce-overhead", fullgraph=True) ``` If you are limited by GPU VRAM, you can enable *cpu offloading* by calling `pipe.enable_model_cpu_offload` instead of `.to("cuda")`: ```diff - pipe.to("cuda") + pipe.enable_model_cpu_offload() ``` ## Uses ### Direct Use The model is intended for research purposes only. Possible research areas and tasks include - Generation of artworks and use in design and other artistic processes. - Applications in educational or creative tools. - Research on generative models. - Safe deployment of models which have the potential to generate harmful content. - Probing and understanding the limitations and biases of generative models. Excluded uses are described below. ### Out-of-Scope Use The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model. ## Limitations and Bias ### Limitations - The model does not achieve perfect photorealism - The model cannot render legible text - The model struggles with more difficult tasks which involve compositionality, such as rendering an image corresponding to “A red cube on top of a blue sphere” - Faces and people in general may not be generated properly. - The autoencoding part of the model is lossy. ### Bias While the capabilities of image generation models are impressive, they can also reinforce or exacerbate social biases. ## Evaluation  The chart above evaluates user preference for SDXL (with and without refinement) over Stable Diffusion 1.5 and 2.1. The SDXL base model performs significantly better than the previous variants, and the model combined with the refinement module achieves the best overall performance.

timm/dpn107.mx_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for dpn107.mx_in1k A DPN (Dual-Path Net) image classification model. Trained on ImageNet-1k in MXNet by paper authors and ported to PyTorch by Ross Wightman. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 86.9 - GMACs: 18.4 - Activations (M): 33.5 - Image size: 224 x 224 - **Papers:** - Dual Path Networks: https://arxiv.org/abs/1707.01629 - **Dataset:** ImageNet-1k - **Original:** https://github.com/cypw/DPNs ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('dpn107.mx_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'dpn107.mx_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 128, 112, 112]) # torch.Size([1, 376, 56, 56]) # torch.Size([1, 1152, 28, 28]) # torch.Size([1, 2432, 14, 14]) # torch.Size([1, 2688, 7, 7]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'dpn107.mx_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 2688, 7, 7) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Citation ```bibtex @article{Chen2017, title={Dual Path Networks}, author={Yunpeng Chen, Jianan Li, Huaxin Xiao, Xiaojie Jin, Shuicheng Yan, Jiashi Feng}, journal={arXiv preprint arXiv:1707.01629}, year={2017} } ```

lllyasviel/control_v11f1p_sd15_depth

lllyasviel

diffusers

image-to-image

--- license: openrail base_model: runwayml/stable-diffusion-v1-5 tags: - art - controlnet - stable-diffusion - controlnet-v1-1 - image-to-image duplicated_from: ControlNet-1-1-preview/control_v11p_sd15_depth --- # Controlnet - v1.1 - *depth Version* **Controlnet v1.1** is the successor model of [Controlnet v1.0](https://huggingface.co/lllyasviel/ControlNet) and was released in [lllyasviel/ControlNet-v1-1](https://huggingface.co/lllyasviel/ControlNet-v1-1) by [Lvmin Zhang](https://huggingface.co/lllyasviel). This checkpoint is a conversion of [the original checkpoint](https://huggingface.co/lllyasviel/ControlNet-v1-1/blob/main/control_v11f1p_sd15_depth.pth) into `diffusers` format. It can be used in combination with **Stable Diffusion**, such as [runwayml/stable-diffusion-v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5). For more details, please also have a look at the [🧨 Diffusers docs](https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/controlnet). ControlNet is a neural network structure to control diffusion models by adding extra conditions.  This checkpoint corresponds to the ControlNet conditioned on **depth images**. ## Model Details - **Developed by:** Lvmin Zhang, Maneesh Agrawala - **Model type:** Diffusion-based text-to-image generation model - **Language(s):** English - **License:** [The CreativeML OpenRAIL M license](https://huggingface.co/spaces/CompVis/stable-diffusion-license) is an [Open RAIL M license](https://www.licenses.ai/blog/2022/8/18/naming-convention-of-responsible-ai-licenses), adapted from the work that [BigScience](https://bigscience.huggingface.co/) and [the RAIL Initiative](https://www.licenses.ai/) are jointly carrying in the area of responsible AI licensing. See also [the article about the BLOOM Open RAIL license](https://bigscience.huggingface.co/blog/the-bigscience-rail-license) on which our license is based. - **Resources for more information:** [GitHub Repository](https://github.com/lllyasviel/ControlNet), [Paper](https://arxiv.org/abs/2302.05543). - **Cite as:** @misc{zhang2023adding, title={Adding Conditional Control to Text-to-Image Diffusion Models}, author={Lvmin Zhang and Maneesh Agrawala}, year={2023}, eprint={2302.05543}, archivePrefix={arXiv}, primaryClass={cs.CV} } ## Introduction Controlnet was proposed in [*Adding Conditional Control to Text-to-Image Diffusion Models*](https://arxiv.org/abs/2302.05543) by Lvmin Zhang, Maneesh Agrawala. The abstract reads as follows: *We present a neural network structure, ControlNet, to control pretrained large diffusion models to support additional input conditions. The ControlNet learns task-specific conditions in an end-to-end way, and the learning is robust even when the training dataset is small (< 50k). Moreover, training a ControlNet is as fast as fine-tuning a diffusion model, and the model can be trained on a personal devices. Alternatively, if powerful computation clusters are available, the model can scale to large amounts (millions to billions) of data. We report that large diffusion models like Stable Diffusion can be augmented with ControlNets to enable conditional inputs like edge maps, depthmentation maps, keypoints, etc. This may enrich the methods to control large diffusion models and further facilitate related applications.* ## Example It is recommended to use the checkpoint with [Stable Diffusion v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5) as the checkpoint has been trained on it. Experimentally, the checkpoint can be used with other diffusion models such as dreamboothed stable diffusion. **Note**: If you want to process an image to create the auxiliary conditioning, external dependencies are required as shown below: 1. Let's install `diffusers` and related packages: ``` $ pip install diffusers transformers accelerate ``` 3. Run code: ```python import torch import os from huggingface_hub import HfApi from pathlib import Path from diffusers.utils import load_image from PIL import Image import numpy as np from transformers import pipeline from diffusers import ( ControlNetModel, StableDiffusionControlNetPipeline, UniPCMultistepScheduler, ) checkpoint = "lllyasviel/control_v11p_sd15_depth" image = load_image( "https://huggingface.co/lllyasviel/control_v11p_sd15_depth/resolve/main/images/input.png" ) prompt = "Stormtrooper's lecture in beautiful lecture hall" depth_estimator = pipeline('depth-estimation') image = depth_estimator(image)['depth'] image = np.array(image) image = image[:, :, None] image = np.concatenate([image, image, image], axis=2) control_image = Image.fromarray(image) control_image.save("./images/control.png") controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16) pipe = StableDiffusionControlNetPipeline.from_pretrained( "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16 ) pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config) pipe.enable_model_cpu_offload() generator = torch.manual_seed(0) image = pipe(prompt, num_inference_steps=30, generator=generator, image=control_image).images[0] image.save('images/image_out.png') ```    ## Other released checkpoints v1-1 The authors released 14 different checkpoints, each trained with [Stable Diffusion v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5) on a different type of conditioning: | Model Name | Control Image Overview| Condition Image | Control Image Example | Generated Image Example | |---|---|---|---|---| |[lllyasviel/control_v11p_sd15_canny](https://huggingface.co/lllyasviel/control_v11p_sd15_canny)<br/> | *Trained with canny edge detection* | A monochrome image with white edges on a black background.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11e_sd15_ip2p](https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p)<br/> | *Trained with pixel to pixel instruction* | No condition .|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_inpaint](https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint)<br/> | Trained with image inpainting | No condition.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/output.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/output.png"/></a>| |[lllyasviel/control_v11p_sd15_mlsd](https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd)<br/> | Trained with multi-level line segment detection | An image with annotated line segments.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11f1p_sd15_depth](https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth)<br/> | Trained with depth estimation | An image with depth information, usually represented as a grayscale image.|<a href="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_normalbae](https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae)<br/> | Trained with surface normal estimation | An image with surface normal information, usually represented as a color-coded image.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_seg](https://huggingface.co/lllyasviel/control_v11p_sd15_seg)<br/> | Trained with image segmentation | An image with segmented regions, usually represented as a color-coded image.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_lineart](https://huggingface.co/lllyasviel/control_v11p_sd15_lineart)<br/> | Trained with line art generation | An image with line art, usually black lines on a white background.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15s2_lineart_anime](https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime)<br/> | Trained with anime line art generation | An image with anime-style line art.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_openpose](https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime)<br/> | Trained with human pose estimation | An image with human poses, usually represented as a set of keypoints or skeletons.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_scribble](https://huggingface.co/lllyasviel/control_v11p_sd15_scribble)<br/> | Trained with scribble-based image generation | An image with scribbles, usually random or user-drawn strokes.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_softedge](https://huggingface.co/lllyasviel/control_v11p_sd15_softedge)<br/> | Trained with soft edge image generation | An image with soft edges, usually to create a more painterly or artistic effect.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11e_sd15_shuffle](https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle)<br/> | Trained with image shuffling | An image with shuffled patches or regions.|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11f1e_sd15_tile](https://huggingface.co/lllyasviel/control_v11f1e_sd15_tile)<br/> | Trained with image tiling | A blurry image or part of an image .|<a href="https://huggingface.co/lllyasviel/control_v11f1e_sd15_tile/resolve/main/images/original.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11f1e_sd15_tile/resolve/main/images/original.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11f1e_sd15_tile/resolve/main/images/output.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11f1e_sd15_tile/resolve/main/images/output.png"/></a>| ## Improvements in Depth 1.1: - The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases. - The new depth model is a relatively unbiased model. It is not trained with some specific type of depth by some specific depth estimation method. It is not over-fitted to one preprocessor. This means this model will work better with different depth estimation, different preprocessor resolutions, or even with real depth created by 3D engines. - Some reasonable data augmentations are applied to training, like random left-right flipping. - The model is resumed from depth 1.0, and it should work well in all cases where depth 1.0 works well. If not, please open an issue with image, and we will take a look at your case. Depth 1.1 works well in many failure cases of depth 1.0. - If you use Midas depth (the "depth" in webui plugin) with 384 preprocessor resolution, the difference between depth 1.0 and 1.1 should be minimal. However, if you try other preprocessor resolutions or other preprocessors (like leres and zoe), the depth 1.1 is expected to be a bit better than 1.0. ## More information For more information, please also have a look at the [Diffusers ControlNet Blog Post](https://huggingface.co/blog/controlnet) and have a look at the [official docs](https://github.com/lllyasviel/ControlNet-v1-1-nightly).

timm/tf_efficientnetv2_s.in21k_ft_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k - imagenet-21k --- # Model card for tf_efficientnetv2_s.in21k_ft_in1k A EfficientNet-v2 image classification model. Trained on ImageNet-21k and fine-tuned on ImageNet-1k in Tensorflow by paper authors, ported to PyTorch by Ross Wightman. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 21.5 - GMACs: 5.4 - Activations (M): 22.7 - Image size: train = 300 x 300, test = 384 x 384 - **Papers:** - EfficientNetV2: Smaller Models and Faster Training: https://arxiv.org/abs/2104.00298 - **Dataset:** ImageNet-1k - **Pretrain Dataset:** ImageNet-21k - **Original:** https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('tf_efficientnetv2_s.in21k_ft_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'tf_efficientnetv2_s.in21k_ft_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 24, 150, 150]) # torch.Size([1, 48, 75, 75]) # torch.Size([1, 64, 38, 38]) # torch.Size([1, 160, 19, 19]) # torch.Size([1, 256, 10, 10]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'tf_efficientnetv2_s.in21k_ft_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 1280, 10, 10) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{tan2021efficientnetv2, title={Efficientnetv2: Smaller models and faster training}, author={Tan, Mingxing and Le, Quoc}, booktitle={International conference on machine learning}, pages={10096--10106}, year={2021}, organization={PMLR} } ``` ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ```

Salesforce/blip-itm-large-coco

Salesforce

transformers

--- pipeline_tags: 'other' tags: - image-text-matching languages: - en license: bsd-3-clause --- # BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation Model card for BLIP trained on image-text matching - large architecture (with ViT large backbone) trained on COCO dataset. |  | |:--:| | <b> Pull figure from BLIP official repo | Image source: https://github.com/salesforce/BLIP </b>| ## TL;DR Authors from the [paper](https://arxiv.org/abs/2201.12086) write in the abstract: *Vision-Language Pre-training (VLP) has advanced the performance for many vision-language tasks. However, most existing pre-trained models only excel in either understanding-based tasks or generation-based tasks. Furthermore, performance improvement has been largely achieved by scaling up the dataset with noisy image-text pairs collected from the web, which is a suboptimal source of supervision. In this paper, we propose BLIP, a new VLP framework which transfers flexibly to both vision-language understanding and generation tasks. BLIP effectively utilizes the noisy web data by bootstrapping the captions, where a captioner generates synthetic captions and a filter removes the noisy ones. We achieve state-of-the-art results on a wide range of vision-language tasks, such as image-text retrieval (+2.7% in average recall@1), image captioning (+2.8% in CIDEr), and VQA (+1.6% in VQA score). BLIP also demonstrates strong generalization ability when directly transferred to videolanguage tasks in a zero-shot manner. Code, models, and datasets are released.* ## Usage You can use this model for conditional and un-conditional image captioning ### Using the Pytorch model #### Running the model on CPU <details> <summary> Click to expand </summary> ```python import requests from PIL import Image from transformers import BlipProcessor, BlipForImageTextRetrieval processor = BlipProcessor.from_pretrained("Salesforce/blip-itm-large-coco") model = BlipForImageTextRetrieval.from_pretrained("Salesforce/blip-itm-large-coco") img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg' raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB') question = "A woman and a dog sitting together in a beach." inputs = processor(raw_image, question, return_tensors="pt") itm_scores = model(**inputs)[0] cosine_score = model(**inputs, use_itm_head=False)[0] ``` </details> #### Running the model on GPU ##### In full precision <details> <summary> Click to expand </summary> ```python import requests from PIL import Image from transformers import BlipProcessor, BlipForImageTextRetrieval processor = BlipProcessor.from_pretrained("Salesforce/blip-itm-large-coco") model = BlipForImageTextRetrieval.from_pretrained("Salesforce/blip-itm-large-coco").to("cuda") img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg' raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB') question = "A woman and a dog sitting together in a beach." inputs = processor(raw_image, question, return_tensors="pt").to("cuda") itm_scores = model(**inputs)[0] cosine_score = model(**inputs, use_itm_head=False)[0] ``` </details> ##### In half precision (`float16`) <details> <summary> Click to expand </summary> ```python import torch import requests from PIL import Image from transformers import BlipProcessor, BlipForImageTextRetrieval processor = BlipProcessor.from_pretrained("Salesforce/blip-itm-large-coco") model = BlipForImageTextRetrieval.from_pretrained("Salesforce/blip-itm-large-coco", torch_dtype=torch.float16).to("cuda") img_url = 'https://storage.googleapis.com/sfr-vision-language-research/BLIP/demo.jpg' raw_image = Image.open(requests.get(img_url, stream=True).raw).convert('RGB') question = "A woman and a dog sitting together in a beach." inputs = processor(raw_image, question, return_tensors="pt").to("cuda", torch.float16) itm_scores = model(**inputs)[0] cosine_score = model(**inputs, use_itm_head=False)[0] ``` </details> ## BibTex and citation info ``` @misc{https://doi.org/10.48550/arxiv.2201.12086, doi = {10.48550/ARXIV.2201.12086}, url = {https://arxiv.org/abs/2201.12086}, author = {Li, Junnan and Li, Dongxu and Xiong, Caiming and Hoi, Steven}, keywords = {Computer Vision and Pattern Recognition (cs.CV), FOS: Computer and information sciences, FOS: Computer and information sciences}, title = {BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation}, publisher = {arXiv}, year = {2022}, copyright = {Creative Commons Attribution 4.0 International} } ```

medicalai/ClinicalBERT

medicalai

transformers

fill-mask

--- tags: - medical --- # ClinicalBERT <!-- Provide a quick summary of what the model is/does. --> This model card describes the ClinicalBERT model, which was trained on a large multicenter dataset with a large corpus of 1.2B words of diverse diseases we constructed. We then utilized a large-scale corpus of EHRs from over 3 million patient records to fine tune the base language model. ## Pretraining Data The ClinicalBERT model was trained on a large multicenter dataset with a large corpus of 1.2B words of diverse diseases we constructed. <!-- For more details, see here. --> ## Model Pretraining ### Pretraining Procedures The ClinicalBERT was initialized from BERT. Then the training followed the principle of masked language model, in which given a piece of text, we randomly replace some tokens by MASKs, special tokens for masking, and then require the model to predict the original tokens via contextual text. ### Pretraining Hyperparameters We used a batch size of 32, a maximum sequence length of 256, and a learning rate of 5e-5 for pre-training our models. ## How to use the model Load the model via the transformers library: ```python from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("medicalai/ClinicalBERT") model = AutoModel.from_pretrained("medicalai/ClinicalBERT") ``` ## Citation Please cite this article: Wang, G., Liu, X., Ying, Z. et al. Optimized glycemic control of type 2 diabetes with reinforcement learning: a proof-of-concept trial. Nat Med (2023). https://doi.org/10.1038/s41591-023-02552-9

timm/mobilenetv3_small_100.lamb_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for mobilenetv3_small_100.lamb_in1k A MobileNet-v3 image classification model. Trained on ImageNet-1k in `timm` using recipe template described below. Recipe details: * A LAMB optimizer recipe that is similar to [ResNet Strikes Back](https://arxiv.org/abs/2110.00476) `A2` but 50% longer with EMA weight averaging, no CutMix * RMSProp (TF 1.0 behaviour) optimizer, EMA weight averaging * Step (exponential decay w/ staircase) LR schedule with warmup ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 2.5 - GMACs: 0.1 - Activations (M): 1.4 - Image size: 224 x 224 - **Papers:** - Searching for MobileNetV3: https://arxiv.org/abs/1905.02244 - **Dataset:** ImageNet-1k - **Original:** https://github.com/huggingface/pytorch-image-models ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('mobilenetv3_small_100.lamb_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilenetv3_small_100.lamb_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 16, 112, 112]) # torch.Size([1, 16, 56, 56]) # torch.Size([1, 24, 28, 28]) # torch.Size([1, 48, 14, 14]) # torch.Size([1, 576, 7, 7]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilenetv3_small_100.lamb_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 576, 7, 7) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ``` ```bibtex @inproceedings{howard2019searching, title={Searching for mobilenetv3}, author={Howard, Andrew and Sandler, Mark and Chu, Grace and Chen, Liang-Chieh and Chen, Bo and Tan, Mingxing and Wang, Weijun and Zhu, Yukun and Pang, Ruoming and Vasudevan, Vijay and others}, booktitle={Proceedings of the IEEE/CVF international conference on computer vision}, pages={1314--1324}, year={2019} } ```

roberta-base-openai-detector

transformers

text-classification

--- language: en license: mit tags: - exbert datasets: - bookcorpus - wikipedia --- # RoBERTa Base OpenAI Detector ## Table of Contents - [Model Details](#model-details) - [Uses](#uses) - [Risks, Limitations and Biases](#risks-limitations-and-biases) - [Training](#training) - [Evaluation](#evaluation) - [Environmental Impact](#environmental-impact) - [Technical Specifications](#technical-specifications) - [Citation Information](#citation-information) - [Model Card Authors](#model-card-author) - [How To Get Started With the Model](#how-to-get-started-with-the-model) ## Model Details **Model Description:** RoBERTa base OpenAI Detector is the GPT-2 output detector model, obtained by fine-tuning a RoBERTa base model with the outputs of the 1.5B-parameter GPT-2 model. The model can be used to predict if text was generated by a GPT-2 model. This model was released by OpenAI at the same time as OpenAI released the weights of the [largest GPT-2 model](https://huggingface.co/gpt2-xl), the 1.5B parameter version. - **Developed by:** OpenAI, see [GitHub Repo](https://github.com/openai/gpt-2-output-dataset/tree/master/detector) and [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf) for full author list - **Model Type:** Fine-tuned transformer-based language model - **Language(s):** English - **License:** MIT - **Related Models:** [RoBERTa base](https://huggingface.co/roberta-base), [GPT-XL (1.5B parameter version)](https://huggingface.co/gpt2-xl), [GPT-Large (the 774M parameter version)](https://huggingface.co/gpt2-large), [GPT-Medium (the 355M parameter version)](https://huggingface.co/gpt2-medium) and [GPT-2 (the 124M parameter version)](https://huggingface.co/gpt2) - **Resources for more information:** - [Research Paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf) (see, in particular, the section beginning on page 12 about Automated ML-based detection). - [GitHub Repo](https://github.com/openai/gpt-2-output-dataset/tree/master/detector) - [OpenAI Blog Post](https://openai.com/blog/gpt-2-1-5b-release/) - [Explore the detector model here](https://huggingface.co/openai-detector ) ## Uses #### Direct Use The model is a classifier that can be used to detect text generated by GPT-2 models. However, it is strongly suggested not to use it as a ChatGPT detector for the purposes of making grave allegations of academic misconduct against undergraduates and others, as this model might give inaccurate results in the case of ChatGPT-generated input. #### Downstream Use The model's developers have stated that they developed and released the model to help with research related to synthetic text generation, so the model could potentially be used for downstream tasks related to synthetic text generation. See the [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf) for further discussion. #### Misuse and Out-of-scope Use The model should not be used to intentionally create hostile or alienating environments for people. In addition, the model developers discuss the risk of adversaries using the model to better evade detection in their [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf), suggesting that using the model for evading detection or for supporting efforts to evade detection would be a misuse of the model. ## Risks, Limitations and Biases **CONTENT WARNING: Readers should be aware this section may contain content that is disturbing, offensive, and can propagate historical and current stereotypes.** Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. #### Risks and Limitations In their [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf), the model developers discuss the risk that the model may be used by bad actors to develop capabilities for evading detection, though one purpose of releasing the model is to help improve detection research. In a related [blog post](https://openai.com/blog/gpt-2-1-5b-release/), the model developers also discuss the limitations of automated methods for detecting synthetic text and the need to pair automated detection tools with other, non-automated approaches. They write: > We conducted in-house detection research and developed a detection model that has detection rates of ~95% for detecting 1.5B GPT-2-generated text. We believe this is not high enough accuracy for standalone detection and needs to be paired with metadata-based approaches, human judgment, and public education to be more effective. The model developers also [report](https://openai.com/blog/gpt-2-1-5b-release/) finding that classifying content from larger models is more difficult, suggesting that detection with automated tools like this model will be increasingly difficult as model sizes increase. The authors find that training detector models on the outputs of larger models can improve accuracy and robustness. #### Bias Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by RoBERTa base and GPT-2 1.5B (which this model is built/fine-tuned on) can include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups (see the [RoBERTa base](https://huggingface.co/roberta-base) and [GPT-2 XL](https://huggingface.co/gpt2-xl) model cards for more information). The developers of this model discuss these issues further in their [paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf). ## Training #### Training Data The model is a sequence classifier based on RoBERTa base (see the [RoBERTa base model card](https://huggingface.co/roberta-base) for more details on the RoBERTa base training data) and then fine-tuned using the outputs of the 1.5B GPT-2 model (available [here](https://github.com/openai/gpt-2-output-dataset)). #### Training Procedure The model developers write that: > We based a sequence classifier on RoBERTaBASE (125 million parameters) and fine-tuned it to classify the outputs from the 1.5B GPT-2 model versus WebText, the dataset we used to train the GPT-2 model. They later state: > To develop a robust detector model that can accurately classify generated texts regardless of the sampling method, we performed an analysis of the model’s transfer performance. See the [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf) for further details on the training procedure. ## Evaluation The following evaluation information is extracted from the [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf). #### Testing Data, Factors and Metrics The model is intended to be used for detecting text generated by GPT-2 models, so the model developers test the model on text datasets, measuring accuracy by: > testing 510-token test examples comprised of 5,000 samples from the WebText dataset and 5,000 samples generated by a GPT-2 model, which were not used during the training. #### Results The model developers [find](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf): > Our classifier is able to detect 1.5 billion parameter GPT-2-generated text with approximately 95% accuracy...The model’s accuracy depends on sampling methods used when generating outputs, like temperature, Top-K, and nucleus sampling ([Holtzman et al., 2019](https://arxiv.org/abs/1904.09751). Nucleus sampling outputs proved most difficult to correctly classify, but a detector trained using nucleus sampling transfers well across other sampling methods. As seen in Figure 1 [in the paper], we found consistently high accuracy when trained on nucleus sampling. See the [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf), Figure 1 (on page 14) and Figure 2 (on page 16) for full results. ## Environmental Impact Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - **Hardware Type:** Unknown - **Hours used:** Unknown - **Cloud Provider:** Unknown - **Compute Region:** Unknown - **Carbon Emitted:** Unknown ## Technical Specifications The model developers write that: See the [associated paper](https://d4mucfpksywv.cloudfront.net/papers/GPT_2_Report.pdf) for further details on the modeling architecture and training details. ## Citation Information ```bibtex @article{solaiman2019release, title={Release strategies and the social impacts of language models}, author={Solaiman, Irene and Brundage, Miles and Clark, Jack and Askell, Amanda and Herbert-Voss, Ariel and Wu, Jeff and Radford, Alec and Krueger, Gretchen and Kim, Jong Wook and Kreps, Sarah and others}, journal={arXiv preprint arXiv:1908.09203}, year={2019} } ``` APA: - Solaiman, I., Brundage, M., Clark, J., Askell, A., Herbert-Voss, A., Wu, J., ... & Wang, J. (2019). Release strategies and the social impacts of language models. arXiv preprint arXiv:1908.09203. ## Model Card Authors This model card was written by the team at Hugging Face. ## How to Get Started with the Model This model can be instantiated and run with a Transformers pipeline: ```python from transformers import pipeline pipe = pipeline("text-classification", model="roberta-base-openai-detector") print(pipe("Hello world! Is this content AI-generated?")) # [{'label': 'Real', 'score': 0.8036582469940186}] ```

google/t5-v1_1-xxl

google

transformers

text2text-generation

--- language: en datasets: - c4 license: apache-2.0 --- [Google's T5](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html) Version 1.1 ## Version 1.1 [T5 Version 1.1](https://github.com/google-research/text-to-text-transfer-transformer/blob/master/released_checkpoints.md#t511) includes the following improvements compared to the original T5 model- GEGLU activation in feed-forward hidden layer, rather than ReLU - see [here](https://arxiv.org/abs/2002.05202). - Dropout was turned off in pre-training (quality win). Dropout should be re-enabled during fine-tuning. - Pre-trained on C4 only without mixing in the downstream tasks. - no parameter sharing between embedding and classifier layer - "xl" and "xxl" replace "3B" and "11B". The model shapes are a bit different - larger `d_model` and smaller `num_heads` and `d_ff`. **Note**: T5 Version 1.1 was only pre-trained on C4 excluding any supervised training. Therefore, this model has to be fine-tuned before it is useable on a downstream task. Pretraining Dataset: [C4](https://huggingface.co/datasets/c4) Other Community Checkpoints: [here](https://huggingface.co/models?search=t5-v1_1) Paper: [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://arxiv.org/pdf/1910.10683.pdf) Authors: *Colin Raffel, Noam Shazeer, Adam Roberts, Katherine Lee, Sharan Narang, Michael Matena, Yanqi Zhou, Wei Li, Peter J. Liu* ## Abstract Transfer learning, where a model is first pre-trained on a data-rich task before being fine-tuned on a downstream task, has emerged as a powerful technique in natural language processing (NLP). The effectiveness of transfer learning has given rise to a diversity of approaches, methodology, and practice. In this paper, we explore the landscape of transfer learning techniques for NLP by introducing a unified framework that converts every language problem into a text-to-text format. Our systematic study compares pre-training objectives, architectures, unlabeled datasets, transfer approaches, and other factors on dozens of language understanding tasks. By combining the insights from our exploration with scale and our new “Colossal Clean Crawled Corpus”, we achieve state-of-the-art results on many benchmarks covering summarization, question answering, text classification, and more. To facilitate future work on transfer learning for NLP, we release our dataset, pre-trained models, and code. 

zhihan1996/DNABERT-2-117M

zhihan1996

transformers

fill-mask

--- metrics: - matthews_correlation - f1 tags: - biology - medical --- This is the official pre-trained model introduced in [DNABERT-2: Efficient Foundation Model and Benchmark For Multi-Species Genome ](https://arxiv.org/pdf/2306.15006.pdf). DNABERT-2 is a transformer-based genome foundation model trained on multi-species genome. To load the model from huggingface: ``` import torch from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("zhihan1996/DNABERT-2-117M", trust_remote_code=True) model = AutoModel.from_pretrained("zhihan1996/DNABERT-2-117M", trust_remote_code=True) ``` To calculate the embedding of a dna sequence ``` dna = "ACGTAGCATCGGATCTATCTATCGACACTTGGTTATCGATCTACGAGCATCTCGTTAGC" inputs = tokenizer(dna, return_tensors = 'pt')["input_ids"] hidden_states = model(inputs)[0] # [1, sequence_length, 768] # embedding with mean pooling embedding_mean = torch.mean(hidden_states[0], dim=0) print(embedding_mean.shape) # expect to be 768 # embedding with max pooling embedding_max = torch.max(hidden_states[0], dim=0)[0] print(embedding_max.shape) # expect to be 768 ```

timm/deit_base_distilled_patch16_224.fb_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_tag: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for deit_base_distilled_patch16_224.fb_in1k A DeiT image classification model. Trained on ImageNet-1k using distillation tokens by paper authors. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 87.3 - GMACs: 17.7 - Activations (M): 24.0 - Image size: 224 x 224 - **Papers:** - Training data-efficient image transformers & distillation through attention: https://arxiv.org/abs/2012.12877 - **Original:** https://github.com/facebookresearch/deit - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('deit_base_distilled_patch16_224.fb_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'deit_base_distilled_patch16_224.fb_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 198, 768) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @InProceedings{pmlr-v139-touvron21a, title = {Training data-efficient image transformers & distillation through attention}, author = {Touvron, Hugo and Cord, Matthieu and Douze, Matthijs and Massa, Francisco and Sablayrolles, Alexandre and Jegou, Herve}, booktitle = {International Conference on Machine Learning}, pages = {10347--10357}, year = {2021}, volume = {139}, month = {July} } ``` ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ```

CreativeLang/novel_metaphors

CreativeLang

transformers

token-classification

Entry not found

lllyasviel/control_v11p_sd15_openpose

lllyasviel

diffusers

image-to-image

--- license: openrail base_model: runwayml/stable-diffusion-v1-5 tags: - art - controlnet - stable-diffusion - controlnet-v1-1 - image-to-image duplicated_from: ControlNet-1-1-preview/control_v11p_sd15_openpose --- # Controlnet - v1.1 - *openpose Version* **Controlnet v1.1** is the successor model of [Controlnet v1.0](https://huggingface.co/lllyasviel/ControlNet) and was released in [lllyasviel/ControlNet-v1-1](https://huggingface.co/lllyasviel/ControlNet-v1-1) by [Lvmin Zhang](https://huggingface.co/lllyasviel). This checkpoint is a conversion of [the original checkpoint](https://huggingface.co/lllyasviel/ControlNet-v1-1/blob/main/control_v11p_sd15_openpose.pth) into `diffusers` format. It can be used in combination with **Stable Diffusion**, such as [runwayml/stable-diffusion-v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5). For more details, please also have a look at the [🧨 Diffusers docs](https://huggingface.co/docs/diffusers/api/pipelines/stable_diffusion/controlnet). ControlNet is a neural network structure to control diffusion models by adding extra conditions.  This checkpoint corresponds to the ControlNet conditioned on **openpose images**. ## Model Details - **Developed by:** Lvmin Zhang, Maneesh Agrawala - **Model type:** Diffusion-based text-to-image generation model - **Language(s):** English - **License:** [The CreativeML OpenRAIL M license](https://huggingface.co/spaces/CompVis/stable-diffusion-license) is an [Open RAIL M license](https://www.licenses.ai/blog/2022/8/18/naming-convention-of-responsible-ai-licenses), adapted from the work that [BigScience](https://bigscience.huggingface.co/) and [the RAIL Initiative](https://www.licenses.ai/) are jointly carrying in the area of responsible AI licensing. See also [the article about the BLOOM Open RAIL license](https://bigscience.huggingface.co/blog/the-bigscience-rail-license) on which our license is based. - **Resources for more information:** [GitHub Repository](https://github.com/lllyasviel/ControlNet), [Paper](https://arxiv.org/abs/2302.05543). - **Cite as:** @misc{zhang2023adding, title={Adding Conditional Control to Text-to-Image Diffusion Models}, author={Lvmin Zhang and Maneesh Agrawala}, year={2023}, eprint={2302.05543}, archivePrefix={arXiv}, primaryClass={cs.CV} } ## Introduction Controlnet was proposed in [*Adding Conditional Control to Text-to-Image Diffusion Models*](https://arxiv.org/abs/2302.05543) by Lvmin Zhang, Maneesh Agrawala. The abstract reads as follows: *We present a neural network structure, ControlNet, to control pretrained large diffusion models to support additional input conditions. The ControlNet learns task-specific conditions in an end-to-end way, and the learning is robust even when the training dataset is small (< 50k). Moreover, training a ControlNet is as fast as fine-tuning a diffusion model, and the model can be trained on a personal devices. Alternatively, if powerful computation clusters are available, the model can scale to large amounts (millions to billions) of data. We report that large diffusion models like Stable Diffusion can be augmented with ControlNets to enable conditional inputs like edge maps, segmentation maps, keypoints, etc. This may enrich the methods to control large diffusion models and further facilitate related applications.* ## Example It is recommended to use the checkpoint with [Stable Diffusion v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5) as the checkpoint has been trained on it. Experimentally, the checkpoint can be used with other diffusion models such as dreamboothed stable diffusion. **Note**: If you want to process an image to create the auxiliary conditioning, external dependencies are required as shown below: 1. Install https://github.com/patrickvonplaten/controlnet_aux ```sh $ pip install controlnet_aux==0.3.0 ``` 2. Let's install `diffusers` and related packages: ``` $ pip install diffusers transformers accelerate ``` 3. Run code: ```python import torch import os from huggingface_hub import HfApi from pathlib import Path from diffusers.utils import load_image from PIL import Image import numpy as np from controlnet_aux import OpenposeDetector from diffusers import ( ControlNetModel, StableDiffusionControlNetPipeline, UniPCMultistepScheduler, ) checkpoint = "lllyasviel/control_v11p_sd15_openpose" image = load_image( "https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/input.png" ) prompt = "chef in the kitchen" processor = OpenposeDetector.from_pretrained('lllyasviel/ControlNet') control_image = processor(image, hand_and_face=True) control_image.save("./images/control.png") controlnet = ControlNetModel.from_pretrained(checkpoint, torch_dtype=torch.float16) pipe = StableDiffusionControlNetPipeline.from_pretrained( "runwayml/stable-diffusion-v1-5", controlnet=controlnet, torch_dtype=torch.float16 ) pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config) pipe.enable_model_cpu_offload() generator = torch.manual_seed(0) image = pipe(prompt, num_inference_steps=30, generator=generator, image=control_image).images[0] image.save('images/image_out.png') ```    ## Other released checkpoints v1-1 The authors released 14 different checkpoints, each trained with [Stable Diffusion v1-5](https://huggingface.co/runwayml/stable-diffusion-v1-5) on a different type of conditioning: | Model Name | Control Image Overview| Control Image Example | Generated Image Example | |---|---|---|---| |[lllyasviel/control_v11p_sd15_canny](https://huggingface.co/lllyasviel/control_v11p_sd15_canny)<br/> *Trained with canny edge detection* | A monochrome image with white edges on a black background.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_canny/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11e_sd15_ip2p](https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p)<br/> *Trained with pixel to pixel instruction* | No condition .|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11e_sd15_ip2p/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_inpaint](https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint)<br/> Trained with image inpainting | No condition.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/output.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_inpaint/resolve/main/images/output.png"/></a>| |[lllyasviel/control_v11p_sd15_mlsd](https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd)<br/> Trained with multi-level line segment detection | An image with annotated line segments.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_mlsd/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11f1p_sd15_depth](https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth)<br/> Trained with depth estimation | An image with depth information, usually represented as a grayscale image.|<a href="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11f1p_sd15_depth/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_normalbae](https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae)<br/> Trained with surface normal estimation | An image with surface normal information, usually represented as a color-coded image.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_normalbae/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_seg](https://huggingface.co/lllyasviel/control_v11p_sd15_seg)<br/> Trained with image segmentation | An image with segmented regions, usually represented as a color-coded image.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_seg/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_lineart](https://huggingface.co/lllyasviel/control_v11p_sd15_lineart)<br/> Trained with line art generation | An image with line art, usually black lines on a white background.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_lineart/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15s2_lineart_anime](https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime)<br/> Trained with anime line art generation | An image with anime-style line art.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15s2_lineart_anime/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_openpose](https://huggingface.co/lllyasviel/control_v11p_sd15_openpose)<br/> Trained with human pose estimation | An image with human poses, usually represented as a set of keypoints or skeletons.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_openpose/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_scribble](https://huggingface.co/lllyasviel/control_v11p_sd15_scribble)<br/> Trained with scribble-based image generation | An image with scribbles, usually random or user-drawn strokes.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_scribble/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11p_sd15_softedge](https://huggingface.co/lllyasviel/control_v11p_sd15_softedge)<br/> Trained with soft edge image generation | An image with soft edges, usually to create a more painterly or artistic effect.|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11p_sd15_softedge/resolve/main/images/image_out.png"/></a>| |[lllyasviel/control_v11e_sd15_shuffle](https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle)<br/> Trained with image shuffling | An image with shuffled patches or regions.|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/control.png"><img width="64" style="margin:0;padding:0;" src="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/control.png"/></a>|<a href="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/image_out.png"><img width="64" src="https://huggingface.co/lllyasviel/control_v11e_sd15_shuffle/resolve/main/images/image_out.png"/></a>| ## Improvements in Openpose 1.1: - The improvement of this model is mainly based on our improved implementation of OpenPose. We carefully reviewed the difference between the pytorch OpenPose and CMU's c++ openpose. Now the processor should be more accurate, especially for hands. The improvement of processor leads to the improvement of Openpose 1.1. - More inputs are supported (hand and face). - The training dataset of previous cnet 1.0 has several problems including (1) a small group of greyscale human images are duplicated thousands of times (!!), causing the previous model somewhat likely to generate grayscale human images; (2) some images has low quality, very blurry, or significant JPEG artifacts; (3) a small group of images has wrong paired prompts caused by a mistake in our data processing scripts. The new model fixed all problems of the training dataset and should be more reasonable in many cases. ## More information For more information, please also have a look at the [Diffusers ControlNet Blog Post](https://huggingface.co/blog/controlnet) and have a look at the [official docs](https://github.com/lllyasviel/ControlNet-v1-1-nightly).

timm/mobilevit_s.cvnets_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: other datasets: - imagenet-1k --- # Model card for mobilevit_s.cvnets_in1k A MobileViT image classification model. Trained on ImageNet-1k by paper authors. See license details at https://github.com/apple/ml-cvnets/blob/main/LICENSE ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 5.6 - GMACs: 2.0 - Activations (M): 19.9 - Image size: 256 x 256 - **Papers:** - MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer: https://arxiv.org/abs/2110.02178 - **Original:** https://github.com/apple/ml-cvnets - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('mobilevit_s.cvnets_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilevit_s.cvnets_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 32, 128, 128]) # torch.Size([1, 64, 64, 64]) # torch.Size([1, 96, 32, 32]) # torch.Size([1, 128, 16, 16]) # torch.Size([1, 640, 8, 8]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilevit_s.cvnets_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 640, 8, 8) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{mehta2022mobilevit, title={MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer}, author={Sachin Mehta and Mohammad Rastegari}, booktitle={International Conference on Learning Representations}, year={2022} } ```

liyucheng/frame_finder

liyucheng

transformers

token-classification

--- license: cc datasets: - liyucheng/FrameNet_v17 language: - en --- # Frame Classification This model is trained FrameNet v1.7. Check out the training dataset [here](https://huggingface.co/datasets/liyucheng/FrameNet_v17). The data is loaded with `ds = dataset.load_dataset('liyucheng/FrameNet_v17', name = 'frame_label')`. This flatten all frame annotation to specific sentences, making frame classification a sequence tagging task. # Metrics ``` {'accuracy_score': 0.8382018348623853, 'precision': 0.8382018348623853, 'recall': 0.8382018348623853, 'micro_f1': 0.8382018348623853, 'macro_f1': 0.45824850358482677} ```

timm/fbnetv3_b.ra2_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for fbnetv3_b.ra2_in1k A FBNet-v3 image classification model. Trained on ImageNet-1k in `timm` using recipe template described below. Recipe details: * RandAugment `RA2` recipe. Inspired by and evolved from EfficientNet RandAugment recipes. Published as `B` recipe in [ResNet Strikes Back](https://arxiv.org/abs/2110.00476). * RMSProp (TF 1.0 behaviour) optimizer, EMA weight averaging * Step (exponential decay w/ staircase) LR schedule with warmup ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 8.6 - GMACs: 0.4 - Activations (M): 7.0 - Image size: train = 224 x 224, test = 256 x 256 - **Papers:** - FBNetV3: Joint Architecture-Recipe Search using Predictor Pretraining: https://arxiv.org/abs/2006.02049 - ResNet strikes back: An improved training procedure in timm: https://arxiv.org/abs/2110.00476 - **Dataset:** ImageNet-1k - **Original:** https://github.com/huggingface/pytorch-image-models ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('fbnetv3_b.ra2_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'fbnetv3_b.ra2_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 16, 112, 112]) # torch.Size([1, 24, 56, 56]) # torch.Size([1, 40, 28, 28]) # torch.Size([1, 120, 14, 14]) # torch.Size([1, 1344, 7, 7]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'fbnetv3_b.ra2_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 1344, 7, 7) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{dai2021fbnetv3, title={Fbnetv3: Joint architecture-recipe search using predictor pretraining}, author={Dai, Xiaoliang and Wan, Alvin and Zhang, Peizhao and Wu, Bichen and He, Zijian and Wei, Zhen and Chen, Kan and Tian, Yuandong and Yu, Matthew and Vajda, Peter and others}, booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition}, pages={16276--16285}, year={2021} } ``` ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ``` ```bibtex @inproceedings{wightman2021resnet, title={ResNet strikes back: An improved training procedure in timm}, author={Wightman, Ross and Touvron, Hugo and Jegou, Herve}, booktitle={NeurIPS 2021 Workshop on ImageNet: Past, Present, and Future} } ```

nbroad/ESG-BERT

nbroad

transformers

text-classification

--- language: - en widget: - text: "In fiscal year 2019, we reduced our comprehensive carbon footprint for the fourth consecutive year—down 35 percent compared to 2015, when Apple’s carbon emissions peaked, even as net revenue increased by 11 percent over that same period. In the past year, we avoided over 10 million metric tons from our emissions reduction initiatives—like our Supplier Clean Energy Program, which lowered our footprint by 4.4 million metric tons. " example_title: "Reduced carbon footprint" - text: "We believe it is essential to establish validated conflict-free sources of 3TG within the Democratic Republic of the Congo (the “DRC”) and adjoining countries (together, with the DRC, the “Covered Countries”), so that these minerals can be procured in a way that contributes to economic growth and development in the region. To aid in this effort, we have established a conflict minerals policy and an internal team to implement the policy." example_title: "Conflict minerals policy" --- # Model Card for ESG-BERT Domain Specific BERT Model for Text Mining in Sustainable Investing # Model Details ## Model Description - **Developed by:** [Mukut Mukherjee](https://www.linkedin.com/in/mukutm/), [Charan Pothireddi](https://www.linkedin.com/in/sree-charan-pothireddi-6a0a3587/) and [Parabole.ai](https://www.linkedin.com/in/sree-charan-pothireddi-6a0a3587/) - **Shared by [Optional]:** HuggingFace - **Model type:** Language model - **Language(s) (NLP):** en - **License:** More information needed - **Related Models:** - **Parent Model:** BERT - **Resources for more information:** - [GitHub Repo](https://github.com/mukut03/ESG-BERT) - [Blog Post](https://towardsdatascience.com/nlp-meets-sustainable-investing-d0542b3c264b?source=friends_link&sk=1f7e6641c3378aaff319a81decf387bf) # Uses ## Direct Use Text Mining in Sustainable Investing ## Downstream Use [Optional] The applications of ESG-BERT can be expanded way beyond just text classification. It can be fine-tuned to perform various other downstream NLP tasks in the domain of Sustainable Investing. ## Out-of-Scope Use The model should not be used to intentionally create hostile or alienating environments for people. # Bias, Risks, and Limitations Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by the model may include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups. ## Recommendations Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recomendations. # Training Details ## Training Data More information needed ## Training Procedure <!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. --> ### Preprocessing More information needed ### Speeds, Sizes, Times More information needed # Evaluation ## Testing Data, Factors & Metrics ### Testing Data The fine-tuned model for text classification is also available [here](https://drive.google.com/drive/folders/1Qz4HP3xkjLfJ6DGCFNeJ7GmcPq65_HVe?usp=sharing). It can be used directly to make predictions using just a few steps. First, download the fine-tuned pytorch_model.bin, config.json, and vocab.txt ### Factors More information needed ### Metrics More information needed ## Results ESG-BERT was further trained on unstructured text data with accuracies of 100% and 98% for Next Sentence Prediction and Masked Language Modelling tasks. Fine-tuning ESG-BERT for text classification yielded an F-1 score of 0.90. For comparison, the general BERT (BERT-base) model scored 0.79 after fine-tuning, and the sci-kit learn approach scored 0.67. # Model Examination More information needed # Environmental Impact Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - **Hardware Type:** More information needed - **Hours used:** More information needed - **Cloud Provider:** information needed - **Compute Region:** More information needed - **Carbon Emitted:** More information needed # Technical Specifications [optional] ## Model Architecture and Objective More information needed ## Compute Infrastructure More information needed ### Hardware More information needed ### Software JDK 11 is needed to serve the model # Citation <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> **BibTeX:** More information needed **APA:** More information needed # Glossary [optional] <!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. --> More information needed # More Information [optional] More information needed # Model Card Authors [optional] [Mukut Mukherjee](https://www.linkedin.com/in/mukutm/), [Charan Pothireddi](https://www.linkedin.com/in/sree-charan-pothireddi-6a0a3587/) and [Parabole.ai](https://www.linkedin.com/in/sree-charan-pothireddi-6a0a3587/), in collaboration with the Ezi Ozoani and the HuggingFace Team # Model Card Contact More information needed # How to Get Started with the Model Use the code below to get started with the model. <details> <summary> Click to expand </summary> ``` pip install torchserve torch-model-archiver pip install torchvision pip install transformers ``` Next up, we'll set up the handler script. It is a basic handler for text classification that can be improved upon. Save this script as "handler.py" in your directory. [1] ``` from abc import ABC import json import logging import os import torch from transformers import AutoModelForSequenceClassification, AutoTokenizer from ts.torch_handler.base_handler import BaseHandler logger = logging.getLogger(__name__) class TransformersClassifierHandler(BaseHandler, ABC): """ Transformers text classifier handler class. This handler takes a text (string) and as input and returns the classification text based on the serialized transformers checkpoint. """ def __init__(self): super(TransformersClassifierHandler, self).__init__() self.initialized = False def initialize(self, ctx): self.manifest = ctx.manifest properties = ctx.system_properties model_dir = properties.get("model_dir") self.device = torch.device("cuda:" + str(properties.get("gpu_id")) if torch.cuda.is_available() else "cpu") # Read model serialize/pt file self.model = AutoModelForSequenceClassification.from_pretrained(model_dir) self.tokenizer = AutoTokenizer.from_pretrained(model_dir) self.model.to(self.device) self.model.eval() logger.debug('Transformer model from path {0} loaded successfully'.format(model_dir)) # Read the mapping file, index to object name mapping_file_path = os.path.join(model_dir, "index_to_name.json") if os.path.isfile(mapping_file_path): with open(mapping_file_path) as f: self.mapping = json.load(f) else: logger.warning('Missing the index_to_name.json file. Inference output will not include class name.') self.initialized = True def preprocess(self, data): """ Very basic preprocessing code - only tokenizes. Extend with your own preprocessing steps as needed. """ text = data[0].get("data") if text is None: text = data[0].get("body") sentences = text.decode('utf-8') logger.info("Received text: '%s'", sentences) inputs = self.tokenizer.encode_plus( sentences, add_special_tokens=True, return_tensors="pt" ) return inputs def inference(self, inputs): """ Predict the class of a text using a trained transformer model. """ # NOTE: This makes the assumption that your model expects text to be tokenized # with "input_ids" and "token_type_ids" - which is true for some popular transformer models, e.g. bert. # If your transformer model expects different tokenization, adapt this code to suit # its expected input format. prediction = self.model( inputs['input_ids'].to(self.device), token_type_ids=inputs['token_type_ids'].to(self.device) )[0].argmax().item() logger.info("Model predicted: '%s'", prediction) if self.mapping: prediction = self.mapping[str(prediction)] return [prediction] def postprocess(self, inference_output): # TODO: Add any needed post-processing of the model predictions here return inference_output _service = TransformersClassifierHandler() def handle(data, context): try: if not _service.initialized: _service.initialize(context) if data is None: return None data = _service.preprocess(data) data = _service.inference(data) data = _service.postprocess(data) return data except Exception as e: raise e ``` TorcheServe uses a format called MAR (Model Archive). We can convert our PyTorch model to a .mar file using this command: ``` torch-model-archiver --model-name "bert" --version 1.0 --serialized-file ./bert_model/pytorch_model.bin --extra-files "./bert_model/config.json,./bert_model/vocab.txt" --handler "./handler.py" ``` Move the .mar file into a new directory: ``` mkdir model_store && mv bert.mar model_store ``` Finally, we can start TorchServe using the command: ``` torchserve --start --model-store model_store --models bert=bert.mar ``` We can now query the model from another terminal window using the Inference API. We pass a text file containing text that the model will try to classify. ``` curl -X POST http://127.0.0.1:8080/predictions/bert -T predict.txt ``` This returns a label number which correlates to a textual label. This is stored in the label_dict.txt dictionary file. ``` __label__Business_Ethics : 0 __label__Data_Security : 1 __label__Access_And_Affordability : 2 __label__Business_Model_Resilience : 3 __label__Competitive_Behavior : 4 __label__Critical_Incident_Risk_Management : 5 __label__Customer_Welfare : 6 __label__Director_Removal : 7 __label__Employee_Engagement_Inclusion_And_Diversity : 8 __label__Employee_Health_And_Safety : 9 __label__Human_Rights_And_Community_Relations : 10 __label__Labor_Practices : 11 __label__Management_Of_Legal_And_Regulatory_Framework : 12 __label__Physical_Impacts_Of_Climate_Change : 13 __label__Product_Quality_And_Safety : 14 __label__Product_Design_And_Lifecycle_Management : 15 __label__Selling_Practices_And_Product_Labeling : 16 __label__Supply_Chain_Management : 17 __label__Systemic_Risk_Management : 18 __label__Waste_And_Hazardous_Materials_Management : 19 __label__Water_And_Wastewater_Management : 20 __label__Air_Quality : 21 __label__Customer_Privacy : 22 __label__Ecological_Impacts : 23 __label__Energy_Management : 24 __label__GHG_Emissions : 25 ``` <\details>

valhalla/distilbart-mnli-12-1

valhalla

transformers

zero-shot-classification

--- datasets: - mnli tags: - distilbart - distilbart-mnli pipeline_tag: zero-shot-classification --- # DistilBart-MNLI distilbart-mnli is the distilled version of bart-large-mnli created using the **No Teacher Distillation** technique proposed for BART summarisation by Huggingface, [here](https://github.com/huggingface/transformers/tree/master/examples/seq2seq#distilbart). We just copy alternating layers from `bart-large-mnli` and finetune more on the same data. | | matched acc | mismatched acc | | ------------------------------------------------------------------------------------ | ----------- | -------------- | | [bart-large-mnli](https://huggingface.co/facebook/bart-large-mnli) (baseline, 12-12) | 89.9 | 90.01 | | [distilbart-mnli-12-1](https://huggingface.co/valhalla/distilbart-mnli-12-1) | 87.08 | 87.5 | | [distilbart-mnli-12-3](https://huggingface.co/valhalla/distilbart-mnli-12-3) | 88.1 | 88.19 | | [distilbart-mnli-12-6](https://huggingface.co/valhalla/distilbart-mnli-12-6) | 89.19 | 89.01 | | [distilbart-mnli-12-9](https://huggingface.co/valhalla/distilbart-mnli-12-9) | 89.56 | 89.52 | This is a very simple and effective technique, as we can see the performance drop is very little. Detailed performace trade-offs will be posted in this [sheet](https://docs.google.com/spreadsheets/d/1dQeUvAKpScLuhDV1afaPJRRAE55s2LpIzDVA5xfqxvk/edit?usp=sharing). ## Fine-tuning If you want to train these models yourself, clone the [distillbart-mnli repo](https://github.com/patil-suraj/distillbart-mnli) and follow the steps below Clone and install transformers from source ```bash git clone https://github.com/huggingface/transformers.git pip install -qqq -U ./transformers ``` Download MNLI data ```bash python transformers/utils/download_glue_data.py --data_dir glue_data --tasks MNLI ``` Create student model ```bash python create_student.py \ --teacher_model_name_or_path facebook/bart-large-mnli \ --student_encoder_layers 12 \ --student_decoder_layers 6 \ --save_path student-bart-mnli-12-6 \ ``` Start fine-tuning ```bash python run_glue.py args.json ``` You can find the logs of these trained models in this [wandb project](https://wandb.ai/psuraj/distilbart-mnli).

circulus/canvers-anime-v3.6

circulus

diffusers

--- license: gpl-3.0 ---

TheBloke/WizardLM-Uncensored-Falcon-7B-GPTQ

TheBloke

transformers

text-generation

--- license: apache-2.0 --- <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Eric Hartford's WizardLM-Uncensored-Falcon-7B GPTQ This repo contains an experimantal GPTQ 4bit model for [Eric Hartford's WizardLM-Uncensored-Falcon-7B](https://huggingface.co/ehartford/WizardLM-Uncensored-Falcon-7b). It is the result of quantising to 4bit using [AutoGPTQ](https://github.com/PanQiWei/AutoGPTQ). ## Repositories available * [4-bit GPTQ model for GPU inference](https://huggingface.co/TheBloke/WizardLM-Uncensored-Falcon-7B-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference](https://huggingface.co/TheBloke/WizardLM-Uncensored-Falcon-7B-GGML) * [Eric's unquantised bf16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/ehartford/WizardLM-Uncensored-Falcon-7b) ## Prompt template Prompt format is WizardLM: ``` What is a falcon? Can I keep one as a pet? ### Response: ``` ## EXPERIMENTAL Please note this is an experimental GPTQ model. Support for it is currently quite limited. It is also expected to be **SLOW**. This is currently unavoidable, but is being looked at. ## AutoGPTQ AutoGPTQ 0.2.0 is required: `pip install auto-gptq` AutoGPTQ provides pre-compiled wheels for Windows and Linux, with CUDA toolkit 11.7 or 11.8. If you are running CUDA toolkit 12.x, you will need to compile your own by following these instructions: ``` git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ pip install . ``` These manual steps will require that you have the [Nvidia CUDA toolkit](https://developer.nvidia.com/cuda-12-0-1-download-archive) installed. ## text-generation-webui There is provisional AutoGPTQ support in text-generation-webui. This requires text-generation-webui as of commit 204731952ae59d79ea3805a425c73dd171d943c3. So please first update text-genration-webui to the latest version. ## How to download and use this model in text-generation-webui 1. Launch text-generation-webui 2. Click the **Model tab**. 3. Untick **Autoload model** 4. Under **Download custom model or LoRA**, enter `TheBloke/WizardLM-Uncensored-Falcon-7B-GPTQ`. 5. Click **Download**. 6. Wait until it says it's finished downloading. 7. Click the **Refresh** icon next to **Model** in the top left. 8. In the **Model drop-down**: choose the model you just downloaded, `WizardLM-Uncensored-Falcon-7B-GPTQ`. 9. Make sure **Loader** is set to **AutoGPTQ**. This model will not work with ExLlama or GPTQ-for-LLaMa. 10. Tick **Trust Remote Code**, followed by **Save Settings** 11. Click **Reload**. 12. Once it says it's loaded, click the **Text Generation tab** and enter a prompt! ## Try it for free on Google Colab Thanks to user [lucianosb](https://huggingface.co/lucianosb), here is a Google Colab notebook that can be used to try this model for free: https://colab.research.google.com/drive/16C4H9heewOrgUMFYNhxz1AvO12yPHyEq?usp=sharing ## About `trust_remote_code` Please be aware that this command line argument causes Python code provided by Falcon to be executed on your machine. This code is required at the moment because Falcon is too new to be supported by Hugging Face transformers. At some point in the future transformers will support the model natively, and then `trust_remote_code` will no longer be needed. In this repo you can see two `.py` files - these are the files that get executed. They are copied from the base repo at [Falcon-7B-Instruct](https://huggingface.co/tiiuae/falcon-7b-instruct). ## Simple Python example code To run this code you need to install AutoGPTQ and einops: ``` pip install auto-gptq pip install einops ``` You can then run this example code: ```python import torch from transformers import AutoTokenizer from auto_gptq import AutoGPTQForCausalLM # Download the model from HF and store it locally, then reference its location here: quantized_model_dir = "/path/to/TheBloke_WizardLM-Uncensored-Falcon-7B-GPTQ" from transformers import AutoTokenizer tokenizer = AutoTokenizer.from_pretrained(quantized_model_dir, use_fast=False) model = AutoGPTQForCausalLM.from_quantized(quantized_model_dir, device="cuda:0", use_triton=False, use_safetensors=True, torch_dtype=torch.float32, trust_remote_code=True) prompt = "Write a story about llamas" prompt_template = f"### Instruction: {prompt}\n### Response:" tokens = tokenizer(prompt_template, return_tensors="pt").to("cuda:0").input_ids output = model.generate(input_ids=tokens, max_new_tokens=100, do_sample=True, temperature=0.8) print(tokenizer.decode(output[0])) ``` ## Provided files **gptq_model-4bit-64g.safetensors** This will work with AutoGPTQ as of commit `3cb1bf5` (`3cb1bf5a6d43a06dc34c6442287965d1838303d3`) It was created with groupsize 64 to give higher inference quality, and without `desc_act` (act-order) to increase inference speed. * `gptq_model-4bit-64g.safetensors` * Works only with latest AutoGPTQ CUDA, compiled from source as of commit `3cb1bf5` * At this time it does not work with AutoGPTQ Triton, but support will hopefully be added in time. * Works with text-generation-webui using `--autogptq --trust_remote_code` * At this time it does NOT work with one-click-installers * Does not work with any version of GPTQ-for-LLaMa * Parameters: Groupsize = 64. No act-order. <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute. Thanks to the [chirper.ai](https://chirper.ai) team! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI **Special thanks to**: Aemon Algiz. **Patreon special mentions**: Sam, theTransient, Jonathan Leane, Steven Wood, webtim, Johann-Peter Hartmann, Geoffrey Montalvo, Gabriel Tamborski, Willem Michiel, John Villwock, Derek Yates, Mesiah Bishop, Eugene Pentland, Pieter, Chadd, Stephen Murray, Daniel P. Andersen, terasurfer, Brandon Frisco, Thomas Belote, Sid, Nathan LeClaire, Magnesian, Alps Aficionado, Stanislav Ovsiannikov, Alex, Joseph William Delisle, Nikolai Manek, Michael Davis, Junyu Yang, K, J, Spencer Kim, Stefan Sabev, Olusegun Samson, transmissions 11, Michael Levine, Cory Kujawski, Rainer Wilmers, zynix, Kalila, Luke @flexchar, Ajan Kanaga, Mandus, vamX, Ai Maven, Mano Prime, Matthew Berman, subjectnull, Vitor Caleffi, Clay Pascal, biorpg, alfie_i, 阿明, Jeffrey Morgan, ya boyyy, Raymond Fosdick, knownsqashed, Olakabola, Leonard Tan, ReadyPlayerEmma, Enrico Ros, Dave, Talal Aujan, Illia Dulskyi, Sean Connelly, senxiiz, Artur Olbinski, Elle, Raven Klaugh, Fen Risland, Deep Realms, Imad Khwaja, Fred von Graf, Will Dee, usrbinkat, SuperWojo, Alexandros Triantafyllidis, Swaroop Kallakuri, Dan Guido, John Detwiler, Pedro Madruga, Iucharbius, Viktor Bowallius, Asp the Wyvern, Edmond Seymore, Trenton Dambrowitz, Space Cruiser, Spiking Neurons AB, Pyrater, LangChain4j, Tony Hughes, Kacper Wikieł, Rishabh Srivastava, David Ziegler, Luke Pendergrass, Andrey, Gabriel Puliatti, Lone Striker, Sebastain Graf, Pierre Kircher, Randy H, NimbleBox.ai, Vadim, danny, Deo Leter Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # ✨ Original model card: Eric Hartford's WizardLM-Uncensored-Falcon-7B This is WizardLM trained on top of tiiuae/falcon-7b, with a subset of the dataset - responses that contained alignment / moralizing were removed. The intent is to train a WizardLM that doesn't have alignment built-in, so that alignment (of any sort) can be added separately with for example with a RLHF LoRA. Shout out to the open source AI/ML community, and everyone who helped me out. Note: An uncensored model has no guardrails. You are responsible for anything you do with the model, just as you are responsible for anything you do with any dangerous object such as a knife, gun, lighter, or car. Publishing anything this model generates is the same as publishing it yourself. You are responsible for the content you publish, and you cannot blame the model any more than you can blame the knife, gun, lighter, or car for what you do with it. Prompt format is Wizardlm. ``` What is a falcon? Can I keep one as a pet? ### Response: ```

ayameRushia/bert-base-indonesian-1.5G-sentiment-analysis-smsa

ayameRushia

transformers

text-classification

--- license: mit tags: - generated_from_trainer datasets: - indonlu metrics: - accuracy model-index: - name: bert-base-indonesian-1.5G-finetuned-sentiment-analysis-smsa results: - task: name: Text Classification type: text-classification dataset: name: indonlu type: indonlu args: smsa metrics: - name: Accuracy type: accuracy value: 0.9373015873015873 language: id widget: - text: "Saya mengapresiasi usaha anda" --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # bert-base-indonesian-1.5G-finetuned-sentiment-analysis-smsa This model is a fine-tuned version of [cahya/bert-base-indonesian-1.5G](https://huggingface.co/cahya/bert-base-indonesian-1.5G) on the indonlu dataset. It achieves the following results on the evaluation set: - Loss: 0.3390 - Accuracy: 0.9373 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 16 - eval_batch_size: 16 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 10 ### Training results | Training Loss | Epoch | Step | Validation Loss | Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:| | 0.2864 | 1.0 | 688 | 0.2154 | 0.9286 | | 0.1648 | 2.0 | 1376 | 0.2238 | 0.9357 | | 0.0759 | 3.0 | 2064 | 0.3351 | 0.9365 | | 0.044 | 4.0 | 2752 | 0.3390 | 0.9373 | | 0.0308 | 5.0 | 3440 | 0.4346 | 0.9365 | | 0.0113 | 6.0 | 4128 | 0.4708 | 0.9365 | | 0.006 | 7.0 | 4816 | 0.5533 | 0.9325 | | 0.0047 | 8.0 | 5504 | 0.5888 | 0.9310 | | 0.0001 | 9.0 | 6192 | 0.5961 | 0.9333 | | 0.0 | 10.0 | 6880 | 0.5992 | 0.9357 | ### Framework versions - Transformers 4.14.1 - Pytorch 1.10.0+cu111 - Datasets 1.16.1 - Tokenizers 0.10.3

facebook/mask2former-swin-large-coco-instance

facebook

transformers

image-segmentation

--- license: other tags: - vision - image-segmentation datasets: - coco widget: - src: http://images.cocodataset.org/val2017/000000039769.jpg example_title: Cats - src: http://images.cocodataset.org/val2017/000000039770.jpg example_title: Castle --- # Mask2Former Mask2Former model trained on COCO instance segmentation (large-sized version, Swin backbone). It was introduced in the paper [Masked-attention Mask Transformer for Universal Image Segmentation ](https://arxiv.org/abs/2112.01527) and first released in [this repository](https://github.com/facebookresearch/Mask2Former/). Disclaimer: The team releasing Mask2Former did not write a model card for this model so this model card has been written by the Hugging Face team. ## Model description Mask2Former addresses instance, semantic and panoptic segmentation with the same paradigm: by predicting a set of masks and corresponding labels. Hence, all 3 tasks are treated as if they were instance segmentation. Mask2Former outperforms the previous SOTA, [MaskFormer](https://arxiv.org/abs/2107.06278) both in terms of performance an efficiency by (i) replacing the pixel decoder with a more advanced multi-scale deformable attention Transformer, (ii) adopting a Transformer decoder with masked attention to boost performance without without introducing additional computation and (iii) improving training efficiency by calculating the loss on subsampled points instead of whole masks.  ## Intended uses & limitations You can use this particular checkpoint for instance segmentation. See the [model hub](https://huggingface.co/models?search=mask2former) to look for other fine-tuned versions on a task that interests you. ### How to use Here is how to use this model: ```python import requests import torch from PIL import Image from transformers import AutoImageProcessor, Mask2FormerForUniversalSegmentation # load Mask2Former fine-tuned on COCO instance segmentation processor = AutoImageProcessor.from_pretrained("facebook/mask2former-swin-large-coco-instance") model = Mask2FormerForUniversalSegmentation.from_pretrained("facebook/mask2former-swin-large-coco-instance") url = "http://images.cocodataset.org/val2017/000000039769.jpg" image = Image.open(requests.get(url, stream=True).raw) inputs = processor(images=image, return_tensors="pt") with torch.no_grad(): outputs = model(**inputs) # model predicts class_queries_logits of shape `(batch_size, num_queries)` # and masks_queries_logits of shape `(batch_size, num_queries, height, width)` class_queries_logits = outputs.class_queries_logits masks_queries_logits = outputs.masks_queries_logits # you can pass them to processor for postprocessing result = processor.post_process_instance_segmentation(outputs, target_sizes=[image.size[::-1]])[0] # we refer to the demo notebooks for visualization (see "Resources" section in the Mask2Former docs) predicted_instance_map = result["segmentation"] ``` For more code examples, we refer to the [documentation](https://huggingface.co/docs/transformers/master/en/model_doc/mask2former).

deepset/gbert-large

deepset

transformers

fill-mask

--- language: de license: mit datasets: - wikipedia - OPUS - OpenLegalData - oscar --- # German BERT large Released, Oct 2020, this is a German BERT language model trained collaboratively by the makers of the original German BERT (aka "bert-base-german-cased") and the dbmdz BERT (aka bert-base-german-dbmdz-cased). In our [paper](https://arxiv.org/pdf/2010.10906.pdf), we outline the steps taken to train our model and show that it outperforms its predecessors. ## Overview **Paper:** [here](https://arxiv.org/pdf/2010.10906.pdf) **Architecture:** BERT large **Language:** German ## Performance ``` GermEval18 Coarse: 80.08 GermEval18 Fine: 52.48 GermEval14: 88.16 ``` See also: deepset/gbert-base deepset/gbert-large deepset/gelectra-base deepset/gelectra-large deepset/gelectra-base-generator deepset/gelectra-large-generator ## Authors **Branden Chan:** branden.chan@deepset.ai **Stefan Schweter:** stefan@schweter.eu **Timo Möller:** timo.moeller@deepset.ai ## About us <div class="grid lg:grid-cols-2 gap-x-4 gap-y-3"> <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center"> <img alt="" src="https://raw.githubusercontent.com/deepset-ai/.github/main/deepset-logo-colored.png" class="w-40"/> </div> <div class="w-full h-40 object-cover mb-2 rounded-lg flex items-center justify-center"> <img alt="" src="https://raw.githubusercontent.com/deepset-ai/.github/main/haystack-logo-colored.png" class="w-40"/> </div> </div> [deepset](http://deepset.ai/) is the company behind the open-source NLP framework [Haystack](https://haystack.deepset.ai/) which is designed to help you build production ready NLP systems that use: Question answering, summarization, ranking etc. Some of our other work: - [Distilled roberta-base-squad2 (aka "tinyroberta-squad2")]([https://huggingface.co/deepset/tinyroberta-squad2) - [German BERT (aka "bert-base-german-cased")](https://deepset.ai/german-bert) - [GermanQuAD and GermanDPR datasets and models (aka "gelectra-base-germanquad", "gbert-base-germandpr")](https://deepset.ai/germanquad) ## Get in touch and join the Haystack community <p>For more info on Haystack, visit our <strong><a href="https://github.com/deepset-ai/haystack">GitHub</a></strong> repo and <strong><a href="https://docs.haystack.deepset.ai">Documentation</a></strong>. We also have a <strong><a class="h-7" href="https://haystack.deepset.ai/community">Discord community open to everyone!</a></strong></p> [Twitter](https://twitter.com/deepset_ai) | [LinkedIn](https://www.linkedin.com/company/deepset-ai/) | [Discord](https://haystack.deepset.ai/community) | [GitHub Discussions](https://github.com/deepset-ai/haystack/discussions) | [Website](https://deepset.ai) By the way: [we're hiring!](http://www.deepset.ai/jobs)

timm/twins_pcpvt_base.in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for twins_pcpvt_base.in1k A Twins-PCPVT image classification model. Trained on ImageNet-1k by paper authors. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 43.8 - GMACs: 6.7 - Activations (M): 25.2 - Image size: 224 x 224 - **Papers:** - Twins: Revisiting the Design of Spatial Attention in Vision Transformers: https://arxiv.org/abs/2104.13840 - **Dataset:** ImageNet-1k - **Original:** https://github.com/Meituan-AutoML/Twins ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('twins_pcpvt_base.in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'twins_pcpvt_base.in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 49, 512) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{chu2021Twins, title={Twins: Revisiting the Design of Spatial Attention in Vision Transformers}, author={Xiangxiang Chu and Zhi Tian and Yuqing Wang and Bo Zhang and Haibing Ren and Xiaolin Wei and Huaxia Xia and Chunhua Shen}, booktitle={NeurIPS 2021}, url={https://openreview.net/forum?id=5kTlVBkzSRx}, year={2021} } ```

timm/crossvit_9_240.in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for crossvit_9_240.in1k A CrossViT image classification model. Trained on ImageNet-1k by paper authors. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 8.6 - GMACs: 1.8 - Activations (M): 9.5 - Image size: 240 x 240 - **Papers:** - CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification: https://arxiv.org/abs/2103.14899 - **Dataset:** ImageNet-1k - **Original:** https://github.com/IBM/CrossViT ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('crossvit_9_240.in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'crossvit_9_240.in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (torch.Size([1, 401, 128]), torch.Size([1, 197, 256])) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{ chen2021crossvit, title={{CrossViT: Cross-Attention Multi-Scale Vision Transformer for Image Classification}}, author={Chun-Fu (Richard) Chen and Quanfu Fan and Rameswar Panda}, booktitle={International Conference on Computer Vision (ICCV)}, year={2021} } ```

timm/pit_b_224.in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for pit_b_224.in1k A PiT (Pooling based Vision Transformer) image classification model. Trained on ImageNet-1k by paper authors. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 73.8 - GMACs: 12.4 - Activations (M): 32.9 - Image size: 224 x 224 - **Papers:** - Rethinking Spatial Dimensions of Vision Transformers: https://arxiv.org/abs/2103.16302 - **Dataset:** ImageNet-1k - **Original:** https://github.com/naver-ai/pit ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('pit_b_224.in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'pit_b_224.in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 256, 31, 31]) # torch.Size([1, 512, 16, 16]) # torch.Size([1, 1024, 8, 8]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'pit_b_224.in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 1, 1024) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{heo2021pit, title={Rethinking Spatial Dimensions of Vision Transformers}, author={Byeongho Heo and Sangdoo Yun and Dongyoon Han and Sanghyuk Chun and Junsuk Choe and Seong Joon Oh}, booktitle = {International Conference on Computer Vision (ICCV)}, year={2021}, } ```

h2oai/h2ogpt-4096-llama2-13b-chat

h2oai

transformers

text-generation

--- inference: false language: - en license: llama2 model_type: llama pipeline_tag: text-generation tags: - facebook - meta - pytorch - llama - llama-2 - h2ogpt --- h2oGPT clone of [Meta's Llama 2 13B Chat](https://huggingface.co/meta-llama/Llama-2-13b-chat-hf). Try it live on our [h2oGPT demo](https://gpt.h2o.ai) with side-by-side LLM comparisons and private document chat! See how it compares to other models on our [LLM Leaderboard](https://evalgpt.ai/)! See more at [H2O.ai](https://h2o.ai/) ## Model Architecture ``` LlamaForCausalLM( (model): LlamaModel( (embed_tokens): Embedding(32000, 5120, padding_idx=0) (layers): ModuleList( (0-39): 40 x LlamaDecoderLayer( (self_attn): LlamaAttention( (q_proj): Linear(in_features=5120, out_features=5120, bias=False) (k_proj): Linear(in_features=5120, out_features=5120, bias=False) (v_proj): Linear(in_features=5120, out_features=5120, bias=False) (o_proj): Linear(in_features=5120, out_features=5120, bias=False) (rotary_emb): LlamaRotaryEmbedding() ) (mlp): LlamaMLP( (gate_proj): Linear(in_features=5120, out_features=13824, bias=False) (up_proj): Linear(in_features=5120, out_features=13824, bias=False) (down_proj): Linear(in_features=13824, out_features=5120, bias=False) (act_fn): SiLUActivation() ) (input_layernorm): LlamaRMSNorm() (post_attention_layernorm): LlamaRMSNorm() ) ) (norm): LlamaRMSNorm() ) (lm_head): Linear(in_features=5120, out_features=32000, bias=False) ) ```

Jean-Baptiste/camembert-ner-with-dates

Jean-Baptiste

transformers

token-classification

--- language: fr datasets: - Jean-Baptiste/wikiner_fr widget: - text: "Je m'appelle jean-baptiste et j'habite à montréal depuis fevr 2012" license: mit --- # camembert-ner: model fine-tuned from camemBERT for NER task (including DATE tag). ## Introduction [camembert-ner-with-dates] is an extension of french camembert-ner model with an additionnal tag for dates. Model was trained on enriched version of wikiner-fr dataset (~170 634 sentences). On my test data (mix of chat and email), this model got an f1 score of ~83% (in comparison dateparser was ~70%). Dateparser library can still be be used on the output of this model in order to convert text to python datetime object (https://dateparser.readthedocs.io/en/latest/). ## How to use camembert-ner-with-dates with HuggingFace ##### Load camembert-ner-with-dates and its sub-word tokenizer : ```python from transformers import AutoTokenizer, AutoModelForTokenClassification tokenizer = AutoTokenizer.from_pretrained("Jean-Baptiste/camembert-ner-with-dates") model = AutoModelForTokenClassification.from_pretrained("Jean-Baptiste/camembert-ner-with-dates") ##### Process text sample (from wikipedia) from transformers import pipeline nlp = pipeline('ner', model=model, tokenizer=tokenizer, aggregation_strategy="simple") nlp("Apple est créée le 1er avril 1976 dans le garage de la maison d'enfance de Steve Jobs à Los Altos en Californie par Steve Jobs, Steve Wozniak et Ronald Wayne14, puis constituée sous forme de société le 3 janvier 1977 à l'origine sous le nom d'Apple Computer, mais pour ses 30 ans et pour refléter la diversification de ses produits, le mot « computer » est retiré le 9 janvier 2015.") [{'entity_group': 'ORG', 'score': 0.9776379466056824, 'word': 'Apple', 'start': 0, 'end': 5}, {'entity_group': 'DATE', 'score': 0.9793774570737567, 'word': 'le 1er avril 1976 dans le', 'start': 15, 'end': 41}, {'entity_group': 'PER', 'score': 0.9958226680755615, 'word': 'Steve Jobs', 'start': 74, 'end': 85}, {'entity_group': 'LOC', 'score': 0.995087186495463, 'word': 'Los Altos', 'start': 87, 'end': 97}, {'entity_group': 'LOC', 'score': 0.9953305125236511, 'word': 'Californie', 'start': 100, 'end': 111}, {'entity_group': 'PER', 'score': 0.9961076378822327, 'word': 'Steve Jobs', 'start': 115, 'end': 126}, {'entity_group': 'PER', 'score': 0.9960325956344604, 'word': 'Steve Wozniak', 'start': 127, 'end': 141}, {'entity_group': 'PER', 'score': 0.9957776467005411, 'word': 'Ronald Wayne', 'start': 144, 'end': 157}, {'entity_group': 'DATE', 'score': 0.994030773639679, 'word': 'le 3 janvier 1977 à', 'start': 198, 'end': 218}, {'entity_group': 'ORG', 'score': 0.9720810294151306, 'word': "d'Apple Computer", 'start': 240, 'end': 257}, {'entity_group': 'DATE', 'score': 0.9924157659212748, 'word': '30 ans et', 'start': 272, 'end': 282}, {'entity_group': 'DATE', 'score': 0.9934852868318558, 'word': 'le 9 janvier 2015.', 'start': 363, 'end': 382}] ``` ## Model performances (metric: seqeval) Global ``` 'precision': 0.928 'recall': 0.928 'f1': 0.928 ``` By entity ``` Label LOC: (precision:0.929, recall:0.932, f1:0.931, support:9510) Label PER: (precision:0.952, recall:0.965, f1:0.959, support:9399) Label MISC: (precision:0.878, recall:0.844, f1:0.860, support:5364) Label ORG: (precision:0.848, recall:0.883, f1:0.865, support:2299) Label DATE: Not relevant because of method used to add date tag on wikiner dataset (estimated f1 ~90%) ```

timm/swinv2_cr_tiny_ns_224.sw_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_tag: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for swinv2_cr_tiny_ns_224.sw_in1k An independent implementation of Swin Transformer V2 released prior to the official code release. A collaboration between [Christoph Reich](https://github.com/ChristophReich1996) and Ross Wightman, the model differs from official impl in a few ways: * MLP log relative position bias uses unnormalized natural log w/o scaling vs normalized, sigmoid clamped and scaled log2. * option to apply LayerNorm at end of every stage ("ns" variants). * defaults to NCHW tensor layout at output of each stage and final features. Pretrained on ImageNet-1k by Ross Wightman. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 28.3 - GMACs: 4.7 - Activations (M): 28.5 - Image size: 224 x 224 - **Papers:** - Swin Transformer V2: Scaling Up Capacity and Resolution: https://arxiv.org/abs/2111.09883 - **Original:** https://github.com/ChristophReich1996/Swin-Transformer-V2 - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('swinv2_cr_tiny_ns_224.sw_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'swinv2_cr_tiny_ns_224.sw_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g. for swin_base_patch4_window7_224 (NHWC output) # torch.Size([1, 56, 56, 128]) # torch.Size([1, 28, 28, 256]) # torch.Size([1, 14, 14, 512]) # torch.Size([1, 7, 7, 1024]) # e.g. for swinv2_cr_small_ns_224 (NCHW output) # torch.Size([1, 96, 56, 56]) # torch.Size([1, 192, 28, 28]) # torch.Size([1, 384, 14, 14]) # torch.Size([1, 768, 7, 7]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'swinv2_cr_tiny_ns_224.sw_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled (ie.e a (batch_size, H, W, num_features) tensor for swin / swinv2 # or (batch_size, num_features, H, W) for swinv2_cr output = model.forward_head(output, pre_logits=True) # output is (batch_size, num_features) tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{liu2021swinv2, title={Swin Transformer V2: Scaling Up Capacity and Resolution}, author={Ze Liu and Han Hu and Yutong Lin and Zhuliang Yao and Zhenda Xie and Yixuan Wei and Jia Ning and Yue Cao and Zheng Zhang and Li Dong and Furu Wei and Baining Guo}, booktitle={International Conference on Computer Vision and Pattern Recognition (CVPR)}, year={2022} } ``` ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ```

tiiuae/falcon-rw-1b

tiiuae

transformers

text-generation

--- datasets: - tiiuae/falcon-refinedweb language: - en inference: false license: apache-2.0 --- # Falcon-RW-1B **Falcon-RW-1B is a 1B parameters causal decoder-only model built by [TII](https://www.tii.ae) and trained on 350B tokens of [RefinedWeb](https://huggingface.co/datasets/tiiuae/falcon-refinedweb). It is made available under the Apache 2.0 license.** See the 📓 [paper on arXiv](https://arxiv.org/abs/2306.01116) for more details. RefinedWeb is a high-quality web dataset built by leveraging stringent filtering and large-scale deduplication. Falcon-RW-1B, trained on RefinedWeb only, matches or outperforms comparable models trained on curated data. ⚠️ Falcon is now available as a core model in the `transformers` library! To use the in-library version, please install the latest version of `transformers` with `pip install git+https://github.com/huggingface/transformers.git`, then simply remove the `trust_remote_code=True` argument from `from_pretrained()`. ⚠️ This model is intended for use as a **research artifact**, to study the influence of training on web data alone. **If you are interested in state-of-the-art models, we recommend using Falcon-[7B](https://huggingface.co/tiiuae/falcon-7b)/[40B](https://huggingface.co/tiiuae/falcon-40b), both trained on >1,000 billion tokens.** ```python from transformers import AutoTokenizer, AutoModelForCausalLM import transformers import torch model = "tiiuae/falcon-rw-1b" tokenizer = AutoTokenizer.from_pretrained(model) pipeline = transformers.pipeline( "text-generation", model=model, tokenizer=tokenizer, torch_dtype=torch.bfloat16, device_map="auto", ) sequences = pipeline( "Girafatron is obsessed with giraffes, the most glorious animal on the face of this Earth. Giraftron believes all other animals are irrelevant when compared to the glorious majesty of the giraffe.\nDaniel: Hello, Girafatron!\nGirafatron:", max_length=200, do_sample=True, top_k=10, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id, ) for seq in sequences: print(f"Result: {seq['generated_text']}") ``` 💥 **Falcon LLMs require PyTorch 2.0 for use with `transformers`!** # Model Card for Falcon-RW-1B ## Model Details ### Model Description - **Developed by:** [https://www.tii.ae](https://www.tii.ae); - **Model type:** Causal decoder-only; - **Language(s) (NLP):** English; - **License:** Apache 2.0. ### Model Source - **Paper:** [https://arxiv.org/abs/2306.01116](https://arxiv.org/abs/2306.01116). ## Uses ### Direct Use Research on large language models, specifically the influence of adequately filtered and deduplicated web data on the properties of large language models (fairness, safety, limitations, capabilities, etc.). ### Out-of-Scope Use Production use without adequate assessment of risks and mitigation; any use cases which may be considered irresponsible or harmful. Broadly speaking, we would recommend Falcon-[7B](https://huggingface.co/tiiuae/falcon-7b)/[40B](https://huggingface.co/tiiuae/falcon-40b) for any use not directly related to research on web data pipelines. ## Bias, Risks, and Limitations Falcon-RW-1B is trained on English data only, and will not generalize appropriately to other languages. Furthermore, as it is trained on a large-scale corpora representative of the web, it will carry the stereotypes and biases commonly encountered online. ### Recommendations We recommend users of Falcon-RW-1B to consider finetuning it for the specific set of tasks of interest, and for guardrails and appropriate precautions to be taken for any production use. ## How to Get Started with the Model ```python from transformers import AutoTokenizer, AutoModelForCausalLM import transformers import torch model = "tiiuae/falcon-rw-1b" tokenizer = AutoTokenizer.from_pretrained(model) pipeline = transformers.pipeline( "text-generation", model=model, tokenizer=tokenizer, torch_dtype=torch.bfloat16, device_map="auto", ) sequences = pipeline( "Girafatron is obsessed with giraffes, the most glorious animal on the face of this Earth. Giraftron believes all other animals are irrelevant when compared to the glorious majesty of the giraffe.\nDaniel: Hello, Girafatron!\nGirafatron:", max_length=200, do_sample=True, top_k=10, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id, ) for seq in sequences: print(f"Result: {seq['generated_text']}") ``` ## Training Details ### Training Data Falcon-RW-1B was trained on 350B tokens of [RefinedWeb](https://huggingface.co/datasets/tiiuae/falcon-refinedweb), a high-quality filtered and deduplicated web dataset. The data was tokenized with the GPT-2 tokenizer. ### Training Procedure Falcon-RW-1B was trained on 32 A100 40GB GPUs, using only data parallelism with ZeRO. #### Training Hyperparameters Hyperparameters were adapted from the GPT-3 paper ([Brown et al., 2020](https://arxiv.org/abs/2005.14165)). | **Hyperparameter** | **Value** | **Comment** | |--------------------|------------|-------------------------------------------| | Precision | `bfloat16` | | | Optimizer | AdamW | | | Learning rate | 2e-4 | 500M tokens warm-up, cosine decay to 2e-5 | | Weight decay | 1e-1 | | | Batch size | 512 | 4B tokens ramp-up | #### Speeds, Sizes, Times Training happened in early December 2022 and took about six days. ## Evaluation See the 📓 [paper on arXiv](https://arxiv.org/abs/2306.01116) for in-depth evaluation. ## Technical Specifications ### Model Architecture and Objective Falcon-RW-1B is a causal decoder-only model trained on a causal language modeling task (i.e., predict the next token). The architecture is adapted from the GPT-3 paper ([Brown et al., 2020](https://arxiv.org/abs/2005.14165)), but uses ALiBi ([Ofir et al., 2021](https://arxiv.org/abs/2108.12409)) and FlashAttention ([Dao et al., 2022](https://arxiv.org/abs/2205.14135)). | **Hyperparameter** | **Value** | **Comment** | |--------------------|-----------|----------------------------------------| | Layers | 24 | | | `d_model` | 2048 | | | `head_dim` | 64 | Reduced to optimise for FlashAttention | | Vocabulary | 50304 | | | Sequence length | 2048 | | ### Compute Infrastructure #### Hardware Falcon-RW-1B was trained on AWS SageMaker, on 32 A100 40GB GPUs in P4d instances. #### Software Falcon-RW-1B was trained a custom distributed training codebase, Gigatron. It uses a 3D parallelism approach combined with ZeRO and high-performance Triton kernels (FlashAttention, etc.) ## Citation ``` @article{refinedweb, title={The {R}efined{W}eb dataset for {F}alcon {LLM}: outperforming curated corpora with web data, and web data only}, author={Guilherme Penedo and Quentin Malartic and Daniel Hesslow and Ruxandra Cojocaru and Alessandro Cappelli and Hamza Alobeidli and Baptiste Pannier and Ebtesam Almazrouei and Julien Launay}, journal={arXiv preprint arXiv:2306.01116}, eprint={2306.01116}, eprinttype = {arXiv}, url={https://arxiv.org/abs/2306.01116}, year={2023} } ``` ## Contact falconllm@tii.ae

Helsinki-NLP/opus-mt-ur-en

Helsinki-NLP

transformers

translation

--- language: - ur - en tags: - translation license: apache-2.0 --- ### urd-eng * source group: Urdu * target group: English * OPUS readme: [urd-eng](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/urd-eng/README.md) * model: transformer-align * source language(s): urd * target language(s): eng * model: transformer-align * pre-processing: normalization + SentencePiece (spm32k,spm32k) * download original weights: [opus-2020-06-17.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/urd-eng/opus-2020-06-17.zip) * test set translations: [opus-2020-06-17.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/urd-eng/opus-2020-06-17.test.txt) * test set scores: [opus-2020-06-17.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/urd-eng/opus-2020-06-17.eval.txt) ## Benchmarks | testset | BLEU | chr-F | |-----------------------|-------|-------| | Tatoeba-test.urd.eng | 23.2 | 0.435 | ### System Info: - hf_name: urd-eng - source_languages: urd - target_languages: eng - opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/urd-eng/README.md - original_repo: Tatoeba-Challenge - tags: ['translation'] - languages: ['ur', 'en'] - src_constituents: {'urd'} - tgt_constituents: {'eng'} - src_multilingual: False - tgt_multilingual: False - prepro: normalization + SentencePiece (spm32k,spm32k) - url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/urd-eng/opus-2020-06-17.zip - url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/urd-eng/opus-2020-06-17.test.txt - src_alpha3: urd - tgt_alpha3: eng - short_pair: ur-en - chrF2_score: 0.435 - bleu: 23.2 - brevity_penalty: 0.975 - ref_len: 12029.0 - src_name: Urdu - tgt_name: English - train_date: 2020-06-17 - src_alpha2: ur - tgt_alpha2: en - prefer_old: False - long_pair: urd-eng - helsinki_git_sha: 480fcbe0ee1bf4774bcbe6226ad9f58e63f6c535 - transformers_git_sha: 2207e5d8cb224e954a7cba69fa4ac2309e9ff30b - port_machine: brutasse - port_time: 2020-08-21-14:41

timm/convit_base.fb_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for convit_base.fb_in1k A ConViT image classification model. Trained on ImageNet-1k by paper authors. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 86.5 - GMACs: 17.5 - Activations (M): 31.8 - Image size: 224 x 224 - **Papers:** - ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases: https://arxiv.org/abs/2103.10697 - **Dataset:** ImageNet-1k - **Original:** https://github.com/facebookresearch/convit ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('convit_base.fb_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'convit_base.fb_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 197, 768) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @article{d2021convit, title={ConViT: Improving Vision Transformers with Soft Convolutional Inductive Biases}, author={d'Ascoli, St{'e}phane and Touvron, Hugo and Leavitt, Matthew and Morcos, Ari and Biroli, Giulio and Sagun, Levent}, journal={arXiv preprint arXiv:2103.10697}, year={2021} } ```

timm/fastvit_t8.apple_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: other datasets: - imagenet-1k --- # Model card for fastvit_t8.apple_in1k A FastViT image classification model. Trained on ImageNet-1k by paper authors. Please observe [original license](https://github.com/apple/ml-fastvit/blob/8af5928238cab99c45f64fc3e4e7b1516b8224ba/LICENSE). ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 4.0 - GMACs: 0.7 - Activations (M): 8.6 - Image size: 256 x 256 - **Papers:** - FastViT: A Fast Hybrid Vision Transformer using Structural Reparameterization: https://arxiv.org/abs/2303.14189 - **Original:** https://github.com/apple/ml-fastvit - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('fastvit_t8.apple_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'fastvit_t8.apple_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 48, 64, 64]) # torch.Size([1, 96, 32, 32]) # torch.Size([1, 192, 16, 16]) # torch.Size([1, 384, 8, 8]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'fastvit_t8.apple_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 384, 8, 8) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Citation ```bibtex @inproceedings{vasufastvit2023, author = {Pavan Kumar Anasosalu Vasu and James Gabriel and Jeff Zhu and Oncel Tuzel and Anurag Ranjan}, title = {FastViT: A Fast Hybrid Vision Transformer using Structural Reparameterization}, booktitle={Proceedings of the IEEE/CVF International Conference on Computer Vision}, year = {2023} } ```

microsoft/resnet-18

microsoft

transformers

image-classification

--- license: apache-2.0 tags: - vision - image-classification datasets: - imagenet-1k widget: - src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/tiger.jpg example_title: Tiger - src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/teapot.jpg example_title: Teapot - src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/palace.jpg example_title: Palace --- # ResNet ResNet model trained on imagenet-1k. It was introduced in the paper [Deep Residual Learning for Image Recognition](https://arxiv.org/abs/1512.03385) and first released in [this repository](https://github.com/KaimingHe/deep-residual-networks). Disclaimer: The team releasing ResNet did not write a model card for this model so this model card has been written by the Hugging Face team. ## Model description ResNet introduced residual connections, they allow to train networks with an unseen number of layers (up to 1000). ResNet won the 2015 ILSVRC & COCO competition, one important milestone in deep computer vision.  ## Intended uses & limitations You can use the raw model for image classification. See the [model hub](https://huggingface.co/models?search=resnet) to look for fine-tuned versions on a task that interests you. ### How to use Here is how to use this model: ```python >>> from transformers import AutoFeatureExtractor, ResNetForImageClassification >>> import torch >>> from datasets import load_dataset >>> dataset = load_dataset("huggingface/cats-image") >>> image = dataset["test"]["image"][0] >>> feature_extractor = AutoFeatureExtractor.from_pretrained("microsoft/resnet-18") >>> model = ResNetForImageClassification.from_pretrained("microsoft/resnet-18") >>> inputs = feature_extractor(image, return_tensors="pt") >>> with torch.no_grad(): ... logits = model(**inputs).logits >>> # model predicts one of the 1000 ImageNet classes >>> predicted_label = logits.argmax(-1).item() >>> print(model.config.id2label[predicted_label]) tiger cat ``` For more code examples, we refer to the [documentation](https://huggingface.co/docs/transformers/master/en/model_doc/resnet).

cross-encoder/stsb-distilroberta-base

cross-encoder

transformers

text-classification

--- license: apache-2.0 --- # Cross-Encoder for Quora Duplicate Questions Detection This model was trained using [SentenceTransformers](https://sbert.net) [Cross-Encoder](https://www.sbert.net/examples/applications/cross-encoder/README.html) class. ## Training Data This model was trained on the [STS benchmark dataset](http://ixa2.si.ehu.eus/stswiki/index.php/STSbenchmark). The model will predict a score between 0 and 1 how for the semantic similarity of two sentences. ## Usage and Performance Pre-trained models can be used like this: ``` from sentence_transformers import CrossEncoder model = CrossEncoder('model_name') scores = model.predict([('Sentence 1', 'Sentence 2'), ('Sentence 3', 'Sentence 4')]) ``` The model will predict scores for the pairs `('Sentence 1', 'Sentence 2')` and `('Sentence 3', 'Sentence 4')`. You can use this model also without sentence_transformers and by just using Transformers ``AutoModel`` class

smallcloudai/Refact-1_6B-fim

smallcloudai

transformers

text-generation

--- pipeline_tag: text-generation inference: true widget: - text: 'def print_hello_world():' example_title: Hello world group: Python license: bigscience-openrail-m pretrain-datasets: - books - arxiv - c4 - falcon-refinedweb - wiki - github-issues - stack_markdown - self-made dataset of permissive github code datasets: - bigcode/the-stack-dedup - rombodawg/2XUNCENSORED_MegaCodeTraining188k - bigcode/commitpackft metrics: - code_eval library_name: transformers tags: - code model-index: - name: Refact-1.6B results: - task: type: text-generation dataset: type: openai_humaneval name: HumanEval metrics: - name: pass@1 (T=0.01) type: pass@1 value: 32.0 verified: false - name: pass@1 (T=0.2) type: pass@1 value: 31.5 verified: false - name: pass@10 (T=0.8) type: pass@10 value: 53.0 verified: false - name: pass@100 (T=0.8) type: pass@100 value: 76.9 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize Python metrics: - name: pass@1 (T=0.2) type: pass@1 value: 35.8 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize JavaScript metrics: - name: pass@1 (T=0.2) type: pass@1 value: 31.6 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize Java metrics: - name: pass@1 (T=0.2) type: pass@1 value: 29.1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize Go metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize C++ metrics: - name: pass@1 (T=0.2) type: pass@1 value: 26.3 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize Rust metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalSynthesize Average metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests Python metrics: - name: pass@1 (T=0.2) type: pass@1 value: 18.38 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests JavaScript metrics: - name: pass@1 (T=0.2) type: pass@1 value: 12.28 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests Java metrics: - name: pass@1 (T=0.2) type: pass@1 value: 15.12 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests Go metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests C++ metrics: - name: pass@1 (T=0.2) type: pass@1 value: 13.17 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests Rust metrics: - name: pass@1 (T=0.2) type: pass@1 value: 2.8 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixTests Average metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs Python metrics: - name: pass@1 (T=0.2) type: pass@1 value: 26.92 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs JavaScript metrics: - name: pass@1 (T=0.2) type: pass@1 value: 26.85 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs Java metrics: - name: pass@1 (T=0.2) type: pass@1 value: 30.76 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs Go metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs C++ metrics: - name: pass@1 (T=0.2) type: pass@1 value: 25.94 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs Rust metrics: - name: pass@1 (T=0.2) type: pass@1 value: 8.44 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalFixDocs Average metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain Python metrics: - name: pass@1 (T=0.2) type: pass@1 value: 26.46 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain JavaScript metrics: - name: pass@1 (T=0.2) type: pass@1 value: 17.86 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain Java metrics: - name: pass@1 (T=0.2) type: pass@1 value: 20.94 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain Go metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain C++ metrics: - name: pass@1 (T=0.2) type: pass@1 value: 18.78 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain Rust metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: bigcode/humanevalpack name: HumanEvalExplain Average metrics: - name: pass@1 (T=0.2) type: pass@1 value: -1 verified: false - task: type: text-generation dataset: type: mbpp name: MBPP metrics: - name: pass@1 (T=0.01) type: pass@1 value: 31.15 verified: false - task: type: text-generation dataset: type: ds1000 name: DS-1000 (Overall Completion) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 10.1 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (C++) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 21.61 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (C#) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 13.91 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (D) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 9.5 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Go) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 53.57 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Java) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 21.58 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Julia) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 13.75 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (JavaScript) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 26.88 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Lua) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 15.26 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (PHP) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 23.04 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Perl) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 12.1 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Python) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 29.6 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (R) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 13.77 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Ruby) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 12.68 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Racket) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 4.29 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Rust) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 19.54 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Scala) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 18.33 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Bash) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 5.7 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (Swift) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 17.68 verified: false - task: type: text-generation dataset: type: nuprl/MultiPL-E name: MultiPL-HumanEval (TypeScript) metrics: - name: pass@1 (T=0.2) type: pass@1 value: 25 verified: false language: - en ---  # Refact-1.6B Finally, the model we started training with our [blog post](https://refact.ai/blog/2023/applying-recent-innovations-to-train-model/) is ready 🎉 After fine-tuning on generated data, it beats Replit 3b, Stability Code 3b and many other models. It almost beats StarCoder ten times the size! Model | Size | HumanEval pass@1 | HumanEval pass@10 | ----------------------|---------------|--------------------|--------------------| DeciCoder-1b | 1b | 19.1% | | <b>Refact-1.6-fim</b> | <b>1.6b</b> | <b>32.0%</b> | <b>53.0%</b> | StableCode | 3b | 20.2% | 33.8% | ReplitCode v1 | 3b | 21.9% | | CodeGen2.5-multi | 7b | 28.4% | 47.5% | CodeLlama | 7b | 33.5% | 59.6% | StarCoder | 15b | 33.6% | | Likely, it's the best model for practical use in your IDE for code completion because it's smart and fast! You can start using it right now by downloading the [Refact plugin](https://refact.ai/). You can host the model yourself, too, using the [open source docker container](https://github.com/smallcloudai/refact). And it's multi-language (see MultiPL-HumanEval and other metrics below) and it works as a chat (see the section below). # It Works As a Chat The primary application of this model is code completion (infill) in multiple programming languages. But it works as a chat quite well. HumanEval results using instruction following (chat) format, against models specialized for chat only: Model | Size | pass@1 | pass@10 | -----------------------|--------|----------|----------| <b>Refact-1.6-fim</b> | 1.6b | 38.4% | 55.6% | StableCode-instruct | 3b | 26.9% | 36.2% | OctoGeeX | 6b | 44.7% | | CodeLlama-instruct | 7b | 34.8% | 64.3% | CodeGen2.5-instruct | 7b | 36.2% | 60.87 | CodeLlama-instruct | 13b | 42.7% | 71.6% | StarChat-β | 15b | 33.5% | | OctoCoder | 15b | 46.2% | | # Example Fill-in-the-middle uses special tokens to identify the prefix/middle/suffix part of the input and output: ```python # pip install -q transformers from transformers import AutoModelForCausalLM, AutoTokenizer checkpoint = "smallcloudai/Refact-1_6B-fim" device = "cuda" # for GPU usage or "cpu" for CPU usage tokenizer = AutoTokenizer.from_pretrained(checkpoint) model = AutoModelForCausalLM.from_pretrained(checkpoint, trust_remote_code=True).to(device) prompt = '<fim_prefix>def print_hello_world():\n """<fim_suffix>\n print("Hello world!")<fim_middle>' inputs = tokenizer.encode(prompt, return_tensors="pt").to(device) outputs = model.generate(inputs, max_length=100, temperature=0.2) print("-"*80) print(tokenizer.decode(outputs[0])) ``` # Chat Format The same model works as chat (experimental). ```python prompt_template = "<empty_output>SYSTEM {system}\n" \ "<empty_output>USER {query}\n" \ "<empty_output>ASSISTANT" prompt = prompt_template.format(system="You are a programming assistant", query="How do I sort a list in Python?") ``` # Architecture As described in more detail in the blog post, we used: - [ALiBi](https://arxiv.org/abs/2108.12409) based attention - [LayerNorm](https://arxiv.org/abs/1607.06450v1) instead of [RMSNorm](https://arxiv.org/pdf/1910.07467.pdf) - [Multi Query Attention](https://arxiv.org/abs/1911.02150) We also used LiON, flash attention, early dropout. It's not that innovative that you can't run it, in fact you can -- see an example below. # Pretraining For the base model, we used our own dataset that contains code with permissive licenses only, and open text datasets. Filtering is the key to success of this model: - We only used text in English - Only topics related to computer science - Applied heavy deduplication The text to code proportion was 50:50, model trained for 1.2T tokens. We don't release the base model, because its Fill-in-the-Middle (FIM) capability likes to repeat itself too much, so its practical use is limited. But if you still want it, write us a message on Discord. # Finetuning We tested our hypothesis that chat data should boost base model performance in FIM and regular left-to-right code completion. We found that just 15% of open [code](https://huggingface.co/datasets/bigcode/commitpackft) [instruction-following](https://huggingface.co/datasets/rombodawg/2XUNCENSORED_MegaCodeTraining188k) datasets, that we filtered for quality, improves almost all metrics. Additionally, to improve FIM, we observed common failure modes, and prepared a synthetic dataset based on [The Stack dedup v1.1](https://huggingface.co/datasets/bigcode/the-stack-dedup) to address them. There is a distribution shift between typical code on the internet, and the code you write in your IDE. The former is likely finished, so the model tries to come up with a suggestion that makes the code complete. You are likely to have half-written code as you work on it, there is no single addition that can repair it fully. In practice, model needs to have a tendency to stop after a couple of lines are added, and sometimes don't write anything at all. We found that just giving it empty completions, single line completions, multiline completions that end with a smaller text indent or at least a newline -- makes it much more usable. This data was used as the rest 85% of the finetune dataset. The final model is the result of several attempts to make it work as good as possible for code completion, and to perform well on a wide range of metrics. The best attempt took 40B tokens. # Limitations and Bias The Refact-1.6B model was trained on text in English. But it has seen a lot more languages in code comments. Its performance on non-English languages is lower, for sure. # Model Stats - **Architecture:** LLAMA-like model with multi-query attention - **Objectives** Fill-in-the-Middle, Chat - **Tokens context:** 4096 - **Pretraining tokens:** 1.2T - **Finetuning tokens:** 40B - **Precision:** bfloat16 - **GPUs** 64 NVidia A5000 - **Training time** 28 days # License The model is licensed under the BigScience OpenRAIL-M v1 license agreement # Citation If you are using this model, please give a link to this page.

THUDM/chatglm2-6b-32k

THUDM

transformers

--- language: - zh - en tags: - glm - chatglm - thudm --- # ChatGLM2-6B-32K <p align="center"> 💻 <a href="https://github.com/THUDM/ChatGLM2-6B" target="_blank">Github Repo</a> • 🐦 <a href="https://twitter.com/thukeg" target="_blank">Twitter</a> • 📃 <a href="https://arxiv.org/abs/2103.10360" target="_blank">[GLM@ACL 22]</a> <a href="https://github.com/THUDM/GLM" target="_blank">[GitHub]</a> • 📃 <a href="https://arxiv.org/abs/2210.02414" target="_blank">[GLM-130B@ICLR 23]</a> <a href="https://github.com/THUDM/GLM-130B" target="_blank">[GitHub]</a> <br> </p> <p align="center"> 👋 Join our <a href="https://join.slack.com/t/chatglm/shared_invite/zt-1y7pqoloy-9b1g6T6JjA8J0KxvUjbwJw" target="_blank">Slack</a> and <a href="https://github.com/THUDM/ChatGLM-6B/blob/main/resources/WECHAT.md" target="_blank">WeChat</a> </p> ## 更新/Update - 我们优化了KV Cache的存储方式，减少了显存碎片的产生。基于优化后的代码，模型可以在约**20G显存**的情况下处理32K长度的上下文（FP/BF16格式）。 - We have optimized the storage method of the KV Cache, reducing the generation of memory fragmentation. Based on the optimized code, the model can process a context length of 32K under approximately **20G** of memory (FP/BF16 format). ## 介绍 ChatGLM**2**-6B-32K在[ChatGLM2-6B](https://huggingface.co/THUDM/chatglm2-6b)的基础上进一步强化了对于长文本的理解能力，能够更好的处理最多32K长度的上下文。具体地，我们基于[位置插值](https://arxiv.org/abs/2306.15595)（Positional Interpolation）的方法对位置编码进行了更新，并在对话阶段使用 32K 的上下文长度训练。在实际的使用中，如果您面临的上下文长度基本在 **8K 以内**，我们推荐使用[ChatGLM2-6B](https://huggingface.co/THUDM/chatglm2-6b)；如果您需要处理**超过 8K** 的上下文长度，我们推荐使用ChatGLM2-6B-32K。 ChatGLM**2**-6B-32K是开源中英双语对话模型 [ChatGLM2-6B](https://github.com/THUDM/ChatGLM2-6B) 的加长版本，在保留了初代模型对话流畅、部署门槛较低等众多优秀特性的基础之上，ChatGLM**2**-6B-32k 引入了如下新特性： 1. **更强大的性能**：基于 ChatGLM 初代模型的开发经验，我们全面升级了 ChatGLM2-6B-32K 的基座模型。ChatGLM2-6B-32K 使用了 [GLM](https://github.com/THUDM/GLM) 的混合目标函数，经过了 1.4T 中英标识符的预训练与人类偏好对齐训练。 2. **更长的上下文**：基于 [FlashAttention](https://github.com/HazyResearch/flash-attention) 技术，我们将基座模型的上下文长度（Context Length）由 ChatGLM-6B 的 2K 扩展到了 32K，并在对话阶段使用 32K 的上下文长度训练，允许更多轮次的对话。 3. **更高效的推理**：基于 [Multi-Query Attention](http://arxiv.org/abs/1911.02150) 技术，ChatGLM2-6B-32K 有更高效的推理速度和更低的显存占用：在官方的模型实现下，推理速度相比初代提升了 42%，INT4 量化下，6G 显存支持的对话长度由 1K 提升到了 8K。 4. **更开放的协议**：ChatGLM2-6B-32K 权重对学术研究**完全开放**，在填写[问卷](https://open.bigmodel.cn/mla/form)进行登记后**亦允许免费商业使用**。 The ChatGLM**2**-6B-32K further strengthens the ability to understand long texts based on the [ChatGLM2-6B](https://huggingface.co/THUDM/chatglm2-6b), and can better handle up to 32K context length. Specifically, we have updated the position encoding based on the method of [Positional Interpolation](https://arxiv.org/abs/2306.15595), and trained with a 32K context length during the dialogue alignment. In practical use, if the context length you are dealing with is generally within 8K, we recommend using [ChatGLM2-6B](https://huggingface.co/THUDM/chatglm2-6b); if you need to handle a context length exceeding 8K, we recommend using ChatGLM2-6B-32K. ChatGLM2-6B-32K is the second-generation version of the open-source bilingual (Chinese-English) chat model [ChatGLM-6B](https://github.com/THUDM/ChatGLM-6B). It retains the smooth conversation flow and low deployment threshold of the first-generation model, while introducing the following new features: 1. **Stronger Performance**: Based on the development experience of the first-generation ChatGLM model, we have fully upgraded the base model of ChatGLM2-6B-32K. ChatGLM2-6B-32K uses the hybrid objective function of [GLM](https://github.com/THUDM/GLM), and has undergone pre-training with 1.4T bilingual tokens and human preference alignment training. 2. **Longer Context**: Based on [FlashAttention](https://github.com/HazyResearch/flash-attention) technique, we have extended the context length of the base model from 2K in ChatGLM-6B to 32K, and trained with a context length of 32K during the dialogue alignment, allowing for more rounds of dialogue. 3. **More Efficient Inference**: Based on [Multi-Query Attention](http://arxiv.org/abs/1911.02150) technique, ChatGLM2-6B-32K has more efficient inference speed and lower GPU memory usage: under the official implementation, the inference speed has increased by 42% compared to the first generation; under INT4 quantization, the dialogue length supported by 6G GPU memory has increased from 1K to 8K. 4. **More Open License**: ChatGLM2-6B-32K weights are **completely open** for academic research, and **free commercial use** is also allowed after completing the [questionnaire](https://open.bigmodel.cn/mla/form). ## 软件依赖 ```shell pip install protobuf transformers==4.30.2 cpm_kernels torch>=2.0 gradio mdtex2html sentencepiece accelerate ``` ## 代码调用 可以通过如下代码调用 ChatGLM-6B-32K 模型来生成对话： ```ipython >>> from transformers import AutoTokenizer, AutoModel >>> tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm2-6b-32k", trust_remote_code=True) >>> model = AutoModel.from_pretrained("THUDM/chatglm2-6b-32k", trust_remote_code=True).half().cuda() >>> model = model.eval() >>> response, history = model.chat(tokenizer, "你好", history=[]) >>> print(response) 你好👋!我是人工智能助手 ChatGLM-6B,很高兴见到你,欢迎问我任何问题。 >>> response, history = model.chat(tokenizer, "晚上睡不着应该怎么办", history=history) >>> print(response) 晚上睡不着可能会让你感到焦虑或不舒服,但以下是一些可以帮助你入睡的方法: 1. 制定规律的睡眠时间表:保持规律的睡眠时间表可以帮助你建立健康的睡眠习惯,使你更容易入睡。尽量在每天的相同时间上床,并在同一时间起床。 2. 创造一个舒适的睡眠环境:确保睡眠环境舒适,安静,黑暗且温度适宜。可以使用舒适的床上用品,并保持房间通风。 3. 放松身心:在睡前做些放松的活动,例如泡个热水澡,听些轻柔的音乐,阅读一些有趣的书籍等,有助于缓解紧张和焦虑,使你更容易入睡。 4. 避免饮用含有咖啡因的饮料:咖啡因是一种刺激性物质,会影响你的睡眠质量。尽量避免在睡前饮用含有咖啡因的饮料,例如咖啡,茶和可乐。 5. 避免在床上做与睡眠无关的事情:在床上做些与睡眠无关的事情,例如看电影,玩游戏或工作等,可能会干扰你的睡眠。 6. 尝试呼吸技巧:深呼吸是一种放松技巧,可以帮助你缓解紧张和焦虑,使你更容易入睡。试着慢慢吸气,保持几秒钟,然后缓慢呼气。 如果这些方法无法帮助你入睡,你可以考虑咨询医生或睡眠专家,寻求进一步的建议。 ``` 关于更多的使用说明，包括如何运行命令行和网页版本的 DEMO，以及使用模型量化以节省显存，请参考我们的 [Github Repo](https://github.com/THUDM/ChatGLM2-6B)。 For more instructions, including how to run CLI and web demos, and model quantization, please refer to our [Github Repo](https://github.com/THUDM/ChatGLM2-6B). ## Change Log * v1.0 ## 协议 本仓库的代码依照 [Apache-2.0](LICENSE) 协议开源，ChatGLM2-6B-32K 模型的权重的使用则需要遵循 [Model License](MODEL_LICENSE)。 ## 引用 如果你觉得我们的工作有帮助的话，请考虑引用下列论文，ChatGLM2-6B 的论文会在近期公布，敬请期待～ ``` @article{zeng2022glm, title={Glm-130b: An open bilingual pre-trained model}, author={Zeng, Aohan and Liu, Xiao and Du, Zhengxiao and Wang, Zihan and Lai, Hanyu and Ding, Ming and Yang, Zhuoyi and Xu, Yifan and Zheng, Wendi and Xia, Xiao and others}, journal={arXiv preprint arXiv:2210.02414}, year={2022} } ``` ``` @inproceedings{du2022glm, title={GLM: General Language Model Pretraining with Autoregressive Blank Infilling}, author={Du, Zhengxiao and Qian, Yujie and Liu, Xiao and Ding, Ming and Qiu, Jiezhong and Yang, Zhilin and Tang, Jie}, booktitle={Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)}, pages={320--335}, year={2022} } ```

kykim/bert-kor-base

kykim

transformers

fill-mask

--- language: ko --- # Bert base model for Korean * 70GB Korean text dataset and 42000 lower-cased subwords are used * Check the model performance and other language models for Korean in [github](https://github.com/kiyoungkim1/LM-kor) ```python from transformers import BertTokenizerFast, BertModel tokenizer_bert = BertTokenizerFast.from_pretrained("kykim/bert-kor-base") model_bert = BertModel.from_pretrained("kykim/bert-kor-base") ```

peft-internal-testing/tiny_T5ForSeq2SeqLM-lora

peft-internal-testing

peft

--- library_name: peft --- ## Training procedure ### Framework versions - PEFT 0.4.0.dev0

peft-internal-testing/tiny_OPTForQuestionAnswering-lora

peft-internal-testing

peft

--- library_name: peft --- ## Training procedure ### Framework versions - PEFT 0.4.0.dev0

cl-tohoku/bert-base-japanese-char-v2

cl-tohoku

transformers

fill-mask

--- language: ja license: cc-by-sa-4.0 datasets: - wikipedia widget: - text: 東北大学で[MASK]の研究をしています。 --- # BERT base Japanese (character-level tokenization with whole word masking, jawiki-20200831) This is a [BERT](https://github.com/google-research/bert) model pretrained on texts in the Japanese language. This version of the model processes input texts with word-level tokenization based on the Unidic 2.1.2 dictionary (available in [unidic-lite](https://pypi.org/project/unidic-lite/) package), followed by character-level tokenization. Additionally, the model is trained with the whole word masking enabled for the masked language modeling (MLM) objective. The codes for the pretraining are available at [cl-tohoku/bert-japanese](https://github.com/cl-tohoku/bert-japanese/tree/v2.0). ## Model architecture The model architecture is the same as the original BERT base model; 12 layers, 768 dimensions of hidden states, and 12 attention heads. ## Training Data The models are trained on the Japanese version of Wikipedia. The training corpus is generated from the Wikipedia Cirrussearch dump file as of August 31, 2020. The generated corpus files are 4.0GB in total, containing approximately 30M sentences. We used the [MeCab](https://taku910.github.io/mecab/) morphological parser with [mecab-ipadic-NEologd](https://github.com/neologd/mecab-ipadic-neologd) dictionary to split texts into sentences. ## Tokenization The texts are first tokenized by MeCab with the Unidic 2.1.2 dictionary and then split into characters. The vocabulary size is 6144. We used [`fugashi`](https://github.com/polm/fugashi) and [`unidic-lite`](https://github.com/polm/unidic-lite) packages for the tokenization. ## Training The models are trained with the same configuration as the original BERT; 512 tokens per instance, 256 instances per batch, and 1M training steps. For training of the MLM (masked language modeling) objective, we introduced whole word masking in which all of the subword tokens corresponding to a single word (tokenized by MeCab) are masked at once. For training of each model, we used a v3-8 instance of Cloud TPUs provided by [TensorFlow Research Cloud program](https://www.tensorflow.org/tfrc/). The training took about 5 days to finish. ## Licenses The pretrained models are distributed under the terms of the [Creative Commons Attribution-ShareAlike 3.0](https://creativecommons.org/licenses/by-sa/3.0/). ## Acknowledgments This model is trained with Cloud TPUs provided by [TensorFlow Research Cloud](https://www.tensorflow.org/tfrc/) program.

peft-internal-testing/tiny_OPTForFeatureExtraction-lora

peft-internal-testing

peft

text-generation

--- library_name: peft pipeline_tag: text-generation --- ## Training procedure ### Framework versions - PEFT 0.4.0.dev0

flaviagiammarino/pubmed-clip-vit-base-patch32

flaviagiammarino

transformers

zero-shot-image-classification

--- license: mit language: - en tags: - medical - vision widget: - src: "https://huggingface.co/flaviagiammarino/pubmed-clip-vit-base-patch32/resolve/main/scripts/input.jpeg" candidate_labels: "Chest X-Ray, Brain MRI, Abdomen CT Scan" example_title: "Abdomen CT Scan" --- # Model Card for PubMedCLIP PubMedCLIP is a fine-tuned version of [CLIP](https://huggingface.co/docs/transformers/model_doc/clip) for the medical domain. ## Model Description PubMedCLIP was trained on the [Radiology Objects in COntext (ROCO)](https://github.com/razorx89/roco-dataset) dataset, a large-scale multimodal medical imaging dataset. The ROCO dataset includes diverse imaging modalities (such as X-Ray, MRI, ultrasound, fluoroscopy, etc.) from various human body regions (such as head, spine, chest, abdomen, etc.) captured from open-access [PubMed](https://pubmed.ncbi.nlm.nih.gov/) articles.<br> PubMedCLIP was trained for 50 epochs with a batch size of 64 using the Adam optimizer with a learning rate of 10−5. The authors have released three different pre-trained models at this [link](https://1drv.ms/u/s!ApXgPqe9kykTgwD4Np3-f7ODAot8?e=zLVlJ2) which use ResNet-50, ResNet-50x4 and ViT32 as image encoders. This repository includes only the ViT32 variant of the PubMedCLIP model.<br> - **Repository:** [PubMedCLIP Official GitHub Repository](https://github.com/sarahESL/PubMedCLIP) - **Paper:** [Does CLIP Benefit Visual Question Answering in the Medical Domain as Much as it Does in the General Domain?](https://arxiv.org/abs/2112.13906) ## Usage ```python import requests from PIL import Image import matplotlib.pyplot as plt from transformers import CLIPProcessor, CLIPModel model = CLIPModel.from_pretrained("flaviagiammarino/pubmed-clip-vit-base-patch32") processor = CLIPProcessor.from_pretrained("flaviagiammarino/pubmed-clip-vit-base-patch32") url = "https://huggingface.co/flaviagiammarino/pubmed-clip-vit-base-patch32/resolve/main/scripts/input.jpeg" image = Image.open(requests.get(url, stream=True).raw) text = ["Chest X-Ray", "Brain MRI", "Abdominal CT Scan"] inputs = processor(text=text, images=image, return_tensors="pt", padding=True) probs = model(**inputs).logits_per_image.softmax(dim=1).squeeze() plt.subplots() plt.imshow(image) plt.title("".join([x[0] + ": " + x[1] + "\n" for x in zip(text, [format(prob, ".4%") for prob in probs])])) plt.axis("off") plt.tight_layout() plt.show() ``` ## Additional Information ### Licensing Information The authors have released the model code and pre-trained checkpoints under the [MIT License](https://github.com/sarahESL/PubMedCLIP/blob/main/LICENSE). ### Citation Information ``` @article{eslami2021does, title={Does clip benefit visual question answering in the medical domain as much as it does in the general domain?}, author={Eslami, Sedigheh and de Melo, Gerard and Meinel, Christoph}, journal={arXiv preprint arXiv:2112.13906}, year={2021} } ```

peft-internal-testing/tiny_GPT2ForTokenClassification-lora

peft-internal-testing

peft

--- library_name: peft --- ## Training procedure ### Framework versions - PEFT 0.4.0.dev0

peft-internal-testing/tiny_WhisperForConditionalGeneration-lora

peft-internal-testing

peft

--- library_name: peft --- ## Training procedure ### Framework versions - PEFT 0.4.0.dev0

TheBloke/Llama-2-70B-Chat-AWQ

TheBloke

transformers

text-generation

--- language: - en license: llama2 tags: - facebook - meta - pytorch - llama - llama-2 model_name: Llama 2 70B Chat base_model: meta-llama/Llama-2-70b-chat-hf inference: false model_creator: Meta Llama 2 model_type: llama pipeline_tag: text-generation prompt_template: '[INST] <<SYS>> You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don''t know the answer to a question, please don''t share false information. <</SYS>> {prompt}[/INST] ' quantized_by: TheBloke --- <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Llama 2 70B Chat - AWQ - Model creator: [Meta Llama 2](https://huggingface.co/meta-llama) - Original model: [Llama 2 70B Chat](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf) <!-- description start --> ## Description This repo contains AWQ model files for [Meta Llama 2's Llama 2 70B Chat](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf). ### About AWQ AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference. It is also now supported by continuous batching server [vLLM](https://github.com/vllm-project/vllm), allowing use of AWQ models for high-throughput concurrent inference in multi-user server scenarios. Note that, at the time of writing, overall throughput is still lower than running vLLM with unquantised models, however using AWQ enables using much smaller GPUs which can lead to easier deployment and overall cost savings. For example, a 70B model can be run on 1 x 48GB GPU instead of 2 x 80GB. <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Llama-2-70B-chat-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Llama-2-70B-chat-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Llama-2-70B-chat-GGUF) * [Meta Llama 2's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/meta-llama/Llama-2-70b-chat-hf) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Llama-2-Chat ``` [INST] <<SYS>> You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information. <</SYS>> {prompt}[/INST] ``` <!-- prompt-template end --> <!-- README_AWQ.md-provided-files start --> ## Provided files and AWQ parameters For my first release of AWQ models, I am releasing 128g models only. I will consider adding 32g as well if there is interest, and once I have done perplexity and evaluation comparisons, but at this time 32g models are still not fully tested with AutoAWQ and vLLM. Models are released as sharded safetensors files. | Branch | Bits | GS | AWQ Dataset | Seq Len | Size | | ------ | ---- | -- | ----------- | ------- | ---- | | [main](https://huggingface.co/TheBloke/Llama-2-70B-chat-AWQ/tree/main) | 4 | 128 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 36.61 GB <!-- README_AWQ.md-provided-files end --> <!-- README_AWQ.md-use-from-vllm start --> ## Serving this model from vLLM Documentation on installing and using vLLM [can be found here](https://vllm.readthedocs.io/en/latest/). - When using vLLM as a server, pass the `--quantization awq` parameter, for example: ```shell python3 python -m vllm.entrypoints.api_server --model TheBloke/Llama-2-70B-chat-AWQ --quantization awq ``` When using vLLM from Python code, pass the `quantization=awq` parameter, for example: ```python from vllm import LLM, SamplingParams prompts = [ "Hello, my name is", "The president of the United States is", "The capital of France is", "The future of AI is", ] sampling_params = SamplingParams(temperature=0.8, top_p=0.95) llm = LLM(model="TheBloke/Llama-2-70B-chat-AWQ", quantization="awq") outputs = llm.generate(prompts, sampling_params) # Print the outputs. for output in outputs: prompt = output.prompt generated_text = output.outputs[0].text print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}") ``` <!-- README_AWQ.md-use-from-vllm start --> <!-- README_AWQ.md-use-from-python start --> ## How to use this AWQ model from Python code ### Install the necessary packages Requires: [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) 0.0.2 or later ```shell pip3 install autoawq ``` If you have problems installing [AutoAWQ](https://github.com/casper-hansen/AutoAWQ) using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y autoawq git clone https://github.com/casper-hansen/AutoAWQ cd AutoAWQ pip3 install . ``` ### You can then try the following example code ```python from awq import AutoAWQForCausalLM from transformers import AutoTokenizer model_name_or_path = "TheBloke/Llama-2-70B-chat-AWQ" # Load model model = AutoAWQForCausalLM.from_quantized(model_name_or_path, fuse_layers=True, trust_remote_code=False, safetensors=True) tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, trust_remote_code=False) prompt = "Tell me about AI" prompt_template=f'''[INST] <<SYS>> You are a helpful, respectful and honest assistant. Always answer as helpfully as possible, while being safe. Your answers should not include any harmful, unethical, racist, sexist, toxic, dangerous, or illegal content. Please ensure that your responses are socially unbiased and positive in nature. If a question does not make any sense, or is not factually coherent, explain why instead of answering something not correct. If you don't know the answer to a question, please don't share false information. <</SYS>> {prompt}[/INST] ''' print("\n\n*** Generate:") tokens = tokenizer( prompt_template, return_tensors='pt' ).input_ids.cuda() # Generate output generation_output = model.generate( tokens, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, max_new_tokens=512 ) print("Output: ", tokenizer.decode(generation_output[0])) # Inference can also be done using transformers' pipeline from transformers import pipeline print("*** Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_AWQ.md-use-from-python end --> <!-- README_AWQ.md-compatibility start --> ## Compatibility The files provided are tested to work with [AutoAWQ](https://github.com/casper-hansen/AutoAWQ), and [vLLM](https://github.com/vllm-project/vllm). [Huggingface Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) is not yet compatible with AWQ, but a PR is open which should bring support soon: [TGI PR #781](https://github.com/huggingface/text-generation-inference/issues/781). <!-- README_AWQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI **Special thanks to**: Aemon Algiz. **Patreon special mentions**: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Meta Llama 2's Llama 2 70B Chat # **Llama 2** Llama 2 is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. This is the repository for the 70B fine-tuned model, optimized for dialogue use cases and converted for the Hugging Face Transformers format. Links to other models can be found in the index at the bottom. ## Model Details *Note: Use of this model is governed by the Meta license. In order to download the model weights and tokenizer, please visit the [website](https://ai.meta.com/resources/models-and-libraries/llama-downloads/) and accept our License before requesting access here.* Meta developed and publicly released the Llama 2 family of large language models (LLMs), a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama-2-Chat, are optimized for dialogue use cases. Llama-2-Chat models outperform open-source chat models on most benchmarks we tested, and in our human evaluations for helpfulness and safety, are on par with some popular closed-source models like ChatGPT and PaLM. **Model Developers** Meta **Variations** Llama 2 comes in a range of parameter sizes — 7B, 13B, and 70B — as well as pretrained and fine-tuned variations. **Input** Models input text only. **Output** Models generate text only. **Model Architecture** Llama 2 is an auto-regressive language model that uses an optimized transformer architecture. The tuned versions use supervised fine-tuning (SFT) and reinforcement learning with human feedback (RLHF) to align to human preferences for helpfulness and safety. ||Training Data|Params|Content Length|GQA|Tokens|LR| |---|---|---|---|---|---|---| |Llama 2|*A new mix of publicly available online data*|7B|4k|&#10007;|2.0T|3.0 x 10^-4| |Llama 2|*A new mix of publicly available online data*|13B|4k|&#10007;|2.0T|3.0 x 10^-4| |Llama 2|*A new mix of publicly available online data*|70B|4k|&#10004;|2.0T|1.5 x 10^-4| *Llama 2 family of models.* Token counts refer to pretraining data only. All models are trained with a global batch-size of 4M tokens. Bigger models - 70B -- use Grouped-Query Attention (GQA) for improved inference scalability. **Model Dates** Llama 2 was trained between January 2023 and July 2023. **Status** This is a static model trained on an offline dataset. Future versions of the tuned models will be released as we improve model safety with community feedback. **License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/) **Research Paper** ["Llama-2: Open Foundation and Fine-tuned Chat Models"](arxiv.org/abs/2307.09288) ## Intended Use **Intended Use Cases** Llama 2 is intended for commercial and research use in English. Tuned models are intended for assistant-like chat, whereas pretrained models can be adapted for a variety of natural language generation tasks. **Out-of-scope Uses** Use in any manner that violates applicable laws or regulations (including trade compliance laws).Use in languages other than English. Use in any other way that is prohibited by the Acceptable Use Policy and Licensing Agreement for Llama 2. ## Hardware and Software **Training Factors** We used custom training libraries, Meta's Research Super Cluster, and production clusters for pretraining. Fine-tuning, annotation, and evaluation were also performed on third-party cloud compute. **Carbon Footprint** Pretraining utilized a cumulative 3.3M GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W). Estimated total emissions were 539 tCO2eq, 100% of which were offset by Meta’s sustainability program. ||Time (GPU hours)|Power Consumption (W)|Carbon Emitted(tCO₂eq)| |---|---|---|---| |Llama 2 7B|184320|400|31.22| |Llama 2 13B|368640|400|62.44| |Llama 2 70B|1720320|400|291.42| |Total|3311616||539.00| **CO₂ emissions during pretraining.** Time: total GPU time required for training each model. Power Consumption: peak power capacity per GPU device for the GPUs used adjusted for power usage efficiency. 100% of the emissions are directly offset by Meta's sustainability program, and because we are openly releasing these models, the pretraining costs do not need to be incurred by others. ## Training Data **Overview** Llama 2 was pretrained on 2 trillion tokens of data from publicly available sources. The fine-tuning data includes publicly available instruction datasets, as well as over one million new human-annotated examples. Neither the pretraining nor the fine-tuning datasets include Meta user data. **Data Freshness** The pretraining data has a cutoff of September 2022, but some tuning data is more recent, up to July 2023. ## Evaluation Results In this section, we report the results for the Llama 1 and Llama 2 models on standard academic benchmarks.For all the evaluations, we use our internal evaluations library. |Model|Size|Code|Commonsense Reasoning|World Knowledge|Reading Comprehension|Math|MMLU|BBH|AGI Eval| |---|---|---|---|---|---|---|---|---|---| |Llama 1|7B|14.1|60.8|46.2|58.5|6.95|35.1|30.3|23.9| |Llama 1|13B|18.9|66.1|52.6|62.3|10.9|46.9|37.0|33.9| |Llama 1|33B|26.0|70.0|58.4|67.6|21.4|57.8|39.8|41.7| |Llama 1|65B|30.7|70.7|60.5|68.6|30.8|63.4|43.5|47.6| |Llama 2|7B|16.8|63.9|48.9|61.3|14.6|45.3|32.6|29.3| |Llama 2|13B|24.5|66.9|55.4|65.8|28.7|54.8|39.4|39.1| |Llama 2|70B|**37.5**|**71.9**|**63.6**|**69.4**|**35.2**|**68.9**|**51.2**|**54.2**| **Overall performance on grouped academic benchmarks.** *Code:* We report the average pass@1 scores of our models on HumanEval and MBPP. *Commonsense Reasoning:* We report the average of PIQA, SIQA, HellaSwag, WinoGrande, ARC easy and challenge, OpenBookQA, and CommonsenseQA. We report 7-shot results for CommonSenseQA and 0-shot results for all other benchmarks. *World Knowledge:* We evaluate the 5-shot performance on NaturalQuestions and TriviaQA and report the average. *Reading Comprehension:* For reading comprehension, we report the 0-shot average on SQuAD, QuAC, and BoolQ. *MATH:* We report the average of the GSM8K (8 shot) and MATH (4 shot) benchmarks at top 1. |||TruthfulQA|Toxigen| |---|---|---|---| |Llama 1|7B|27.42|23.00| |Llama 1|13B|41.74|23.08| |Llama 1|33B|44.19|22.57| |Llama 1|65B|48.71|21.77| |Llama 2|7B|33.29|**21.25**| |Llama 2|13B|41.86|26.10| |Llama 2|70B|**50.18**|24.60| **Evaluation of pretrained LLMs on automatic safety benchmarks.** For TruthfulQA, we present the percentage of generations that are both truthful and informative (the higher the better). For ToxiGen, we present the percentage of toxic generations (the smaller the better). |||TruthfulQA|Toxigen| |---|---|---|---| |Llama-2-Chat|7B|57.04|**0.00**| |Llama-2-Chat|13B|62.18|**0.00**| |Llama-2-Chat|70B|**64.14**|0.01| **Evaluation of fine-tuned LLMs on different safety datasets.** Same metric definitions as above. ## Ethical Considerations and Limitations Llama 2 is a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, Llama 2’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate, biased or other objectionable responses to user prompts. Therefore, before deploying any applications of Llama 2, developers should perform safety testing and tuning tailored to their specific applications of the model. Please see the Responsible Use Guide available at [https://ai.meta.com/llama/responsible-use-guide/](https://ai.meta.com/llama/responsible-use-guide) ## Reporting Issues Please report any software “bug,” or other problems with the models through one of the following means: - Reporting issues with the model: [github.com/facebookresearch/llama](http://github.com/facebookresearch/llama) - Reporting problematic content generated by the model: [developers.facebook.com/llama_output_feedback](http://developers.facebook.com/llama_output_feedback) - Reporting bugs and security concerns: [facebook.com/whitehat/info](http://facebook.com/whitehat/info) ## Llama Model Index |Model|Llama2|Llama2-hf|Llama2-chat|Llama2-chat-hf| |---|---|---|---|---| |7B| [Link](https://huggingface.co/llamaste/Llama-2-7b) | [Link](https://huggingface.co/llamaste/Llama-2-7b-hf) | [Link](https://huggingface.co/llamaste/Llama-2-7b-chat) | [Link](https://huggingface.co/llamaste/Llama-2-7b-chat-hf)| |13B| [Link](https://huggingface.co/llamaste/Llama-2-13b) | [Link](https://huggingface.co/llamaste/Llama-2-13b-hf) | [Link](https://huggingface.co/llamaste/Llama-2-13b-chat) | [Link](https://huggingface.co/llamaste/Llama-2-13b-hf)| |70B| [Link](https://huggingface.co/llamaste/Llama-2-70b) | [Link](https://huggingface.co/llamaste/Llama-2-70b-hf) | [Link](https://huggingface.co/llamaste/Llama-2-70b-chat) | [Link](https://huggingface.co/llamaste/Llama-2-70b-hf)|

timm/visformer_small.in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for visformer_small.in1k A Visformer image classification model. Trained on ImageNet-1k by https://github.com/hzhang57 and https://github.com/developer0hye. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 40.2 - GMACs: 4.9 - Activations (M): 11.4 - Image size: 224 x 224 - **Papers:** - Visformer: The Vision-friendly Transformer: https://arxiv.org/abs/2104.12533 - **Dataset:** ImageNet-1k - **Original:** https://github.com/danczs/Visformer ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('visformer_small.in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'visformer_small.in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 768, 7, 7) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{chen2021visformer, title={Visformer: The vision-friendly transformer}, author={Chen, Zhengsu and Xie, Lingxi and Niu, Jianwei and Liu, Xuefeng and Wei, Longhui and Tian, Qi}, booktitle={Proceedings of the IEEE/CVF international conference on computer vision}, pages={589--598}, year={2021} } ```

CreativeLang/metaphor_detection_roberta_seq

CreativeLang

transformers

token-classification

--- license: cc-by-2.0 datasets: - CreativeLang/vua20_metaphor language: - en --- # Metaphor_Detection_Roberta_Seq ## Description - **Paper:** [FrameBERT: Conceptual Metaphor Detection with Frame Embedding Learning](https://aclanthology.org/2023.eacl-main.114.pdf) ## Model Summary Creative Language Toolkit (CLTK) Metadata - CL Type: Metaphor - Task Type: detection - Size: roberta-base (500MB) - Created time: 2022 This model is a easy to use metaphor detection baseline realised with `roberta-base` fine-tuned on [CreativeLang/vua20_metaphor](https://huggingface.co/datasets/CreativeLang/vua20_metaphor) dataset. To use this model, please use the `inference.py` in the [FrameBERT repo](https://github.com/liyucheng09/MetaphorFrame). Just run: ``` python inference.py CreativeLang/metaphor_detection_roberta_seq ``` Check out `inference.py` to learn how to apply the model on your own data. For the details of this model and the dataset used, we refer you to the release [paper](https://aclanthology.org/2023.eacl-main.114.pdf). ## Metrics | Metric | Value | |----------------------------------|--------------------------| | eval_loss | 0.2656 | | eval_accuracy_score | 0.9142 | | eval_precision | 0.9142 | | eval_recall | 0.9142 | | eval_f1 | 0.9142 | | eval_f1_macro | 0.7315 | | eval_runtime | 8.9802 | | eval_samples_per_second | 411.7960 | | eval_steps_per_second | 51.5580 | | epoch | 3.0000 | ### Citation Information If you find this dataset helpful, please cite: ``` @article{Li2023FrameBERTCM, title={FrameBERT: Conceptual Metaphor Detection with Frame Embedding Learning}, author={Yucheng Li and Shunyu Wang and Chenghua Lin and Frank Guerin and Lo{\"i}c Barrault}, journal={ArXiv}, year={2023}, volume={abs/2302.04834} } ``` ### Contributions If you have any queries, please open an issue or direct your queries to [mail](mailto:yucheng.li@surrey.ac.uk).

DeepFloyd/IF-I-M-v1.0

DeepFloyd

diffusers

text-to-image

--- license: deepfloyd-if-license extra_gated_prompt: "DeepFloyd LICENSE AGREEMENT\nThis License Agreement (as may be amended in accordance with this License Agreement, “License”), between you, or your employer or other entity (if you are entering into this agreement on behalf of your employer or other entity) (“Licensee” or “you”) and Stability AI Ltd.. (“Stability AI” or “we”) applies to your use of any computer program, algorithm, source code, object code, or software that is made available by Stability AI under this License (“Software”) and any specifications, manuals, documentation, and other written information provided by Stability AI related to the Software (“Documentation”).\nBy clicking “I Accept” below or by using the Software, you agree to the terms of this License. If you do not agree to this License, then you do not have any rights to use the Software or Documentation (collectively, the “Software Products”), and you must immediately cease using the Software Products. If you are agreeing to be bound by the terms of this License on behalf of your employer or other entity, you represent and warrant to Stability AI that you have full legal authority to bind your employer or such entity to this License. If you do not have the requisite authority, you may not accept the License or access the Software Products on behalf of your employer or other entity.\n1. LICENSE GRANT\n a. Subject to your compliance with the Documentation and Sections 2, 3, and 5, Stability AI grants you a non-exclusive, worldwide, non-transferable, non-sublicensable, revocable, royalty free and limited license under Stability AI’s copyright interests to reproduce, distribute, and create derivative works of the Software solely for your non-commercial research purposes. The foregoing license is personal to you, and you may not assign or sublicense this License or any other rights or obligations under this License without Stability AI’s prior written consent; any such assignment or sublicense will be void and will automatically and immediately terminate this License.\n b. You may make a reasonable number of copies of the Documentation solely for use in connection with the license to the Software granted above.\n c. The grant of rights expressly set forth in this Section 1 (License Grant) are the complete grant of rights to you in the Software Products, and no other licenses are granted, whether by waiver, estoppel, implication, equity or otherwise. Stability AI and its licensors reserve all rights not expressly granted by this License.\L\n2. RESTRICTIONS\n You will not, and will not permit, assist or cause any third party to:\n a. use, modify, copy, reproduce, create derivative works of, or distribute the Software Products (or any derivative works thereof, works incorporating the Software Products, or any data produced by the Software), in whole or in part, for (i) any commercial or production purposes, (ii) military purposes or in the service of nuclear technology, (iii) purposes of surveillance, including any research or development relating to surveillance, (iv) biometric processing, (v) in any manner that infringes, misappropriates, or otherwise violates any third-party rights, or (vi) in any manner that violates any applicable law and violating any privacy or security laws, rules, regulations, directives, or governmental requirements (including the General Data Privacy Regulation (Regulation (EU) 2016/679), the California Consumer Privacy Act, and any and all laws governing the processing of biometric information), as well as all amendments and successor laws to any of the foregoing;\n b. alter or remove copyright and other proprietary notices which appear on or in the Software Products;\n c. utilize any equipment, device, software, or other means to circumvent or remove any security or protection used by Stability AI in connection with the Software, or to circumvent or remove any usage restrictions, or to enable functionality disabled by Stability AI; or\n d. offer or impose any terms on the Software Products that alter, restrict, or are inconsistent with the terms of this License.\n e. 1) violate any applicable U.S. and non-U.S. export control and trade sanctions laws (“Export Laws”); 2) directly or indirectly export, re-export, provide, or otherwise transfer Software Products: (a) to any individual, entity, or country prohibited by Export Laws; (b) to anyone on U.S. or non-U.S. government restricted parties lists; or (c) for any purpose prohibited by Export Laws, including nuclear, chemical or biological weapons, or missile technology applications; 3) use or download Software Products if you or they are: (a) located in a comprehensively sanctioned jurisdiction, (b) currently listed on any U.S. or non-U.S. restricted parties list, or (c) for any purpose prohibited by Export Laws; and (4) will not disguise your location through IP proxying or other methods.\L\n3. ATTRIBUTION\n Together with any copies of the Software Products (as well as derivative works thereof or works incorporating the Software Products) that you distribute, you must provide (i) a copy of this License, and (ii) the following attribution notice: “DeepFloyd is licensed under the DeepFloyd License, Copyright (c) Stability AI Ltd. All Rights Reserved.”\L\n4. DISCLAIMERS\n THE SOFTWARE PRODUCTS ARE PROVIDED “AS IS” and “WITH ALL FAULTS” WITH NO WARRANTY OF ANY KIND, EXPRESS OR IMPLIED. STABILITY AIEXPRESSLY DISCLAIMS ALL REPRESENTATIONS AND WARRANTIES, EXPRESS OR IMPLIED, WHETHER BY STATUTE, CUSTOM, USAGE OR OTHERWISE AS TO ANY MATTERS RELATED TO THE SOFTWARE PRODUCTS, INCLUDING BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, TITLE, SATISFACTORY QUALITY, OR NON-INFRINGEMENT. STABILITY AI MAKES NO WARRANTIES OR REPRESENTATIONS THAT THE SOFTWARE PRODUCTS WILL BE ERROR FREE OR FREE OF VIRUSES OR OTHER HARMFUL COMPONENTS, OR PRODUCE ANY PARTICULAR RESULTS.\L\n5. LIMITATION OF LIABILITY\n TO THE FULLEST EXTENT PERMITTED BY LAW, IN NO EVENT WILL STABILITY AI BE LIABLE TO YOU (A) UNDER ANY THEORY OF LIABILITY, WHETHER BASED IN CONTRACT, TORT, NEGLIGENCE, STRICT LIABILITY, WARRANTY, OR OTHERWISE UNDER THIS LICENSE, OR (B) FOR ANY INDIRECT, CONSEQUENTIAL, EXEMPLARY, INCIDENTAL, PUNITIVE OR SPECIAL DAMAGES OR LOST PROFITS, EVEN IF STABILITY AI HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. THE SOFTWARE PRODUCTS, THEIR CONSTITUENT COMPONENTS, AND ANY OUTPUT (COLLECTIVELY, “SOFTWARE MATERIALS”) ARE NOT DESIGNED OR INTENDED FOR USE IN ANY APPLICATION OR SITUATION WHERE FAILURE OR FAULT OF THE SOFTWARE MATERIALS COULD REASONABLY BE ANTICIPATED TO LEAD TO SERIOUS INJURY OF ANY PERSON, INCLUDING POTENTIAL DISCRIMINATION OR VIOLATION OF AN INDIVIDUAL’S PRIVACY RIGHTS, OR TO SEVERE PHYSICAL, PROPERTY, OR ENVIRONMENTAL DAMAGE (EACH, A “HIGH-RISK USE”). IF YOU ELECT TO USE ANY OF THE SOFTWARE MATERIALS FOR A HIGH-RISK USE, YOU DO SO AT YOUR OWN RISK. YOU AGREE TO DESIGN AND IMPLEMENT APPROPRIATE DECISION-MAKING AND RISK-MITIGATION PROCEDURES AND POLICIES IN CONNECTION WITH A HIGH-RISK USE SUCH THAT EVEN IF THERE IS A FAILURE OR FAULT IN ANY OF THE SOFTWARE MATERIALS, THE SAFETY OF PERSONS OR PROPERTY AFFECTED BY THE ACTIVITY STAYS AT A LEVEL THAT IS REASONABLE, APPROPRIATE, AND LAWFUL FOR THE FIELD OF THE HIGH-RISK USE.\L\n6. INDEMNIFICATION\n You will indemnify, defend and hold harmless Stability AI and our subsidiaries and affiliates, and each of our respective shareholders, directors, officers, employees, agents, successors, and assigns (collectively, the “Stability AI Parties”) from and against any losses, liabilities, damages, fines, penalties, and expenses (including reasonable attorneys’ fees) incurred by any Stability AI Party in connection with any claim, demand, allegation, lawsuit, proceeding, or investigation (collectively, “Claims”) arising out of or related to: (a) your access to or use of the Software Products (as well as any results or data generated from such access or use), including any High-Risk Use (defined below); (b) your violation of this License; or (c) your violation, misappropriation or infringement of any rights of another (including intellectual property or other proprietary rights and privacy rights). You will promptly notify the Stability AI Parties of any such Claims, and cooperate with Stability AI Parties in defending such Claims. You will also grant the Stability AI Parties sole control of the defense or settlement, at Stability AI’s sole option, of any Claims. This indemnity is in addition to, and not in lieu of, any other indemnities or remedies set forth in a written agreement between you and Stability AI or the other Stability AI Parties.\L\n7. TERMINATION; SURVIVAL\n a. This License will automatically terminate upon any breach by you of the terms of this License.\L\Lb. We may terminate this License, in whole or in part, at any time upon notice (including electronic) to you.\L\Lc. The following sections survive termination of this License: 2 (Restrictions), 3 (Attribution), 4 (Disclaimers), 5 (Limitation on Liability), 6 (Indemnification) 7 (Termination; Survival), 8 (Third Party Materials), 9 (Trademarks), 10 (Applicable Law; Dispute Resolution), and 11 (Miscellaneous).\L\n8. THIRD PARTY MATERIALS\n The Software Products may contain third-party software or other components (including free and open source software) (all of the foregoing, “Third Party Materials”), which are subject to the license terms of the respective third-party licensors. Your dealings or correspondence with third parties and your use of or interaction with any Third Party Materials are solely between you and the third party. Stability AI does not control or endorse, and makes no representations or warranties regarding, any Third Party Materials, and your access to and use of such Third Party Materials are at your own risk.\L\n9. TRADEMARKS\n Licensee has not been granted any trademark license as part of this License and may not use any name or mark associated with Stability AI without the prior written permission of Stability AI, except to the extent necessary to make the reference required by the “ATTRIBUTION” section of this Agreement.\L\n10. APPLICABLE LAW; DISPUTE RESOLUTION\n This License will be governed and construed under the laws of the State of California without regard to conflicts of law provisions. Any suit or proceeding arising out of or relating to this License will be brought in the federal or state courts, as applicable, in San Mateo County, California, and each party irrevocably submits to the jurisdiction and venue of such courts.\L\n11. MISCELLANEOUS\n If any provision or part of a provision of this License is unlawful, void or unenforceable, that provision or part of the provision is deemed severed from this License, and will not affect the validity and enforceability of any remaining provisions. The failure of Stability AI to exercise or enforce any right or provision of this License will not operate as a waiver of such right or provision. This License does not confer any third-party beneficiary rights upon any other person or entity. This License, together with the Documentation, contains the entire understanding between you and Stability AI regarding the subject matter of this License, and supersedes all other written or oral agreements and understandings between you and Stability AI regarding such subject matter. No change or addition to any provision of this License will be binding unless it is in writing and signed by an authorized representative of both you and Stability AI." extra_gated_fields: "Organization /\_Affiliation": text Previously related publications: text I accept the above license agreement, and will use the Software non-commercially and for research purposes only: checkbox tags: - if - text-to-image inference: false --- # IF-I-M-v1.0 DeepFloyd-IF is a pixel-based text-to-image triple-cascaded diffusion model, that can generate pictures with new state-of-the-art for photorealism and language understanding. The result is a highly efficient model that outperforms current state-of-the-art models, achieving a zero-shot FID-30K score of `6.66` on the COCO dataset. *Inspired by* [*Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding*](https://arxiv.org/pdf/2205.11487.pdf)  ## Model Details - **Developed by:** DeepFloyd, StabilityAI - **Model type:** pixel-based text-to-image cascaded diffusion model - **Cascade Stage:** I - **Num Parameters:** 400M - **Language(s):** primarily English and, to a lesser extent, other Romance languages - **License:** <span style="color:blue"><a href="https://huggingface.co/spaces/DeepFloyd/deepfloyd-if-license">DeepFloyd IF License Agreement</a></span> - **Model Description:** DeepFloyd-IF is modular composed of frozen text mode and three pixel cascaded diffusion modules, each designed to generate images of increasing resolution: 64x64, 256x256, and 1024x1024. All stages of the model utilize a frozen text encoder based on the T5 transformer to extract text embeddings, which are then fed into a UNet architecture enhanced with cross-attention and attention-pooling - **Resources for more information:** [GitHub](https://github.com/deep-floyd/IF), [Website](https://deepfloyd.ai), [All Links](https://linktr.ee/deepfloyd) ## Using with `diffusers` IF is integrated with the 🤗 Hugging Face [🧨 diffusers library](https://github.com/huggingface/diffusers/), which is optimized to run on GPUs with as little as 14 GB of VRAM. Before you can use IF, you need to accept its usage conditions. To do so: 1. Make sure to have a [Hugging Face account](https://huggingface.co/join) and be loggin in 2. Accept the license on the model card of [DeepFloyd/IF-I-M-v1.0](https://huggingface.co/DeepFloyd/IF-I-M-v1.0) 3. Make sure to login locally. Install `huggingface_hub` ```sh pip install huggingface_hub --upgrade ``` run the login function in a Python shell ```py from huggingface_hub import login login() ``` and enter your [Hugging Face Hub access token](https://huggingface.co/docs/hub/security-tokens#what-are-user-access-tokens). Next we install `diffusers` and dependencies: ```sh pip install diffusers accelerate transformers safetensors sentencepiece ``` And we can now run the model locally. By default `diffusers` makes use of [model cpu offloading](https://huggingface.co/docs/diffusers/optimization/fp16#model-offloading-for-fast-inference-and-memory-savings) to run the whole IF pipeline with as little as 14 GB of VRAM. If you are using `torch>=2.0.0`, make sure to **remove all** `enable_xformers_memory_efficient_attention()` functions. * **Load all stages and offload to CPU** ```py from diffusers import DiffusionPipeline from diffusers.utils import pt_to_pil import torch # stage 1 stage_1 = DiffusionPipeline.from_pretrained("DeepFloyd/IF-I-M-v1.0", variant="fp16", torch_dtype=torch.float16) stage_1.enable_xformers_memory_efficient_attention() # remove line if torch.__version__ >= 2.0.0 stage_1.enable_model_cpu_offload() # stage 2 stage_2 = DiffusionPipeline.from_pretrained( "DeepFloyd/IF-II-M-v1.0", text_encoder=None, variant="fp16", torch_dtype=torch.float16 ) stage_2.enable_xformers_memory_efficient_attention() # remove line if torch.__version__ >= 2.0.0 stage_2.enable_model_cpu_offload() # stage 3 safety_modules = {"feature_extractor": stage_1.feature_extractor, "safety_checker": stage_1.safety_checker, "watermarker": stage_1.watermarker} stage_3 = DiffusionPipeline.from_pretrained("stabilityai/stable-diffusion-x4-upscaler", **safety_modules, torch_dtype=torch.float16) stage_3.enable_xformers_memory_efficient_attention() # remove line if torch.__version__ >= 2.0.0 stage_3.enable_model_cpu_offload() ``` * **Retrieve Text Embeddings** ```py prompt = 'a photo of a kangaroo wearing an orange hoodie and blue sunglasses standing in front of the eiffel tower holding a sign that says "very deep learning"' # text embeds prompt_embeds, negative_embeds = stage_1.encode_prompt(prompt) ``` * **Run stage 1** ```py generator = torch.manual_seed(0) image = stage_1(prompt_embeds=prompt_embeds, negative_prompt_embeds=negative_embeds, generator=generator, output_type="pt").images pt_to_pil(image)[0].save("./if_stage_I.png") ``` * **Run stage 2** ```py image = stage_2( image=image, prompt_embeds=prompt_embeds, negative_prompt_embeds=negative_embeds, generator=generator, output_type="pt" ).images pt_to_pil(image)[0].save("./if_stage_II.png") ``` * **Run stage 3** ```py image = stage_3(prompt=prompt, image=image, generator=generator, noise_level=100).images image[0].save("./if_stage_III.png") ``` There are multiple ways to speed up the inference time and lower the memory consumption even more with `diffusers`. To do so, please have a look at the Diffusers docs: - 🚀 [Optimizing for inference time](https://huggingface.co/docs/diffusers/api/pipelines/if#optimizing-for-speed) - ⚙️ [Optimizing for low memory during inference](https://huggingface.co/docs/diffusers/api/pipelines/if#optimizing-for-memory) For more in-detail information about how to use IF, please have a look at [the IF blog post](https://huggingface.co/blog/if) and the [documentation](https://huggingface.co/docs/diffusers/main/en/api/pipelines/if) 📖. Diffusers dreambooth scripts also supports fine-tuning 🎨 [IF](https://huggingface.co/docs/diffusers/main/en/training/dreambooth#if). With parameter efficient finetuning, you can add new concepts to IF with a single GPU and ~28 GB VRAM. ## Training **Training Data:** 1.2B text-image pairs (based on LAION-A and few additional internal datasets) Test/Valid parts of datasets are not used at any cascade and stage of training. Valid part of COCO helps to demonstrate "online" loss behaviour during training (to catch incident and other problems), but dataset is never used for train. **Training Procedure:** IF-I-M-v1.0 is the smallest (from IF series) pixel-based diffusion cascade which uses T5-Encoder embeddings (hidden states) to generate 64px image. During training, - Images are cropped to square via shifted-center-crop augmentation (randomly shift from center up to 0.1 of size) and resized to 64px using `Pillow==9.2.0` BICUBIC resampling with reducing_gap=None (it helps to avoid aliasing) and processed to tensor BxCxHxW - Text prompts are encoded through open-sourced frozen T5-v1_1-xxl text-encoder (that completely was trained by Google team), random 10% of texts are dropped to empty string to add ability for classifier free guidance (CFG) - The non-pooled output of the text encoder is fed into the projection (linear layer without activation) and is used in UNet backbone of the diffusion model via controlled hybrid self- and cross- attention - Also, the output of the text encode is pooled via attention-pooling (64 heads) and is used in time embed as additional features - Diffusion process is limited by 1000 discrete steps, with cosine beta schedule of noising image - The loss is a reconstruction objective between the noise that was added to the image and the prediction made by the UNet - The training process for checkpoint IF-I-M-v1.0 has 2_500_000 steps at resolution 64x64 on all datasets, OneCycleLR policy, SiLU activations, optimizer AdamW8bit + DeepSpeed-Zero1, fully frozen T5-Encoder  **Hardware:** 12 x 8 x A100 GPUs **Optimizer:** [AdamW8bit](https://arxiv.org/abs/2110.02861) + [DeepSpeed ZeRO-1](https://www.deepspeed.ai/tutorials/zero/) **Batch:** 3072 **Learning rate**: [one-cycle](https://pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.OneCycleLR.html) cosine strategy, warmup 10000 steps, start_lr=4e-6, max_lr=1e-4, final_lr=1e-8  ## Evaluation Results `FID-30K: 8.86`  # Uses ## Direct Use The model is released for research purposes. Any attempt to deploy the model in production requires not only that the LICENSE is followed but full liability over the person deploying the model. Possible research areas and tasks include: - Generation of artistic imagery and use in design and other artistic processes. - Safe deployment of models which have the potential to generate harmful content. - Probing and understanding the limitations and biases of generative models. - Applications in educational or creative tools. - Research on generative models. Excluded uses are described below. ### Misuse, Malicious Use, and Out-of-Scope Use _Note: This section is originally taken from the [DALLE-MINI model card](https://huggingface.co/dalle-mini/dalle-mini), was used for Stable Diffusion but applies in the same way for IF_. The model should not be used to intentionally create or disseminate images that create hostile or alienating environments for people. This includes generating images that people would foreseeably find disturbing, distressing, or offensive; or content that propagates historical or current stereotypes. #### Out-of-Scope Use The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model. #### Misuse and Malicious Use Using the model to generate content that is cruel to individuals is a misuse of this model. This includes, but is not limited to: - Generating demeaning, dehumanizing, or otherwise harmful representations of people or their environments, cultures, religions, etc. - Intentionally promoting or propagating discriminatory content or harmful stereotypes. - Impersonating individuals without their consent. - Sexual content without consent of the people who might see it. - Mis- and disinformation - Representations of egregious violence and gore - Sharing of copyrighted or licensed material in violation of its terms of use. - Sharing content that is an alteration of copyrighted or licensed material in violation of its terms of use. ## Limitations and Bias ### Limitations - The model does not achieve perfect photorealism - The model was trained mainly with English captions and will not work as well in other languages. - The model was trained on a subset of the large-scale dataset [LAION-5B](https://laion.ai/blog/laion-5b/), which contains adult, violent and sexual content. To partially mitigate this, we have... (see Training section). ### Bias While the capabilities of image generation models are impressive, they can also reinforce or exacerbate social biases. IF was primarily trained on subsets of [LAION-2B(en)](https://laion.ai/blog/laion-5b/), which consists of images that are limited to English descriptions. Texts and images from communities and cultures that use other languages are likely to be insufficiently accounted for. This affects the overall output of the model, as white and western cultures are often set as the default. Further, the ability of the model to generate content with non-English prompts is significantly worse than with English-language prompts. IF mirrors and exacerbates biases to such a degree that viewer discretion must be advised irrespective of the input or its intent. *This model card was written by: DeepFloyd Team and is based on the [StableDiffusion model card](https://huggingface.co/CompVis/stable-diffusion-v1-4).*

cointegrated/LaBSE-en-ru

cointegrated

transformers

feature-extraction

--- language: ["ru", "en"] tags: - feature-extraction - embeddings - sentence-similarity --- # LaBSE for English and Russian This is a truncated version of [sentence-transformers/LaBSE](https://huggingface.co/sentence-transformers/LaBSE), which is, in turn, a port of [LaBSE](https://tfhub.dev/google/LaBSE/1) by Google. The current model has only English and Russian tokens left in the vocabulary. Thus, the vocabulary is 10% of the original, and number of parameters in the whole model is 27% of the original, without any loss in the quality of English and Russian embeddings. To get the sentence embeddings, you can use the following code: ```python import torch from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained("cointegrated/LaBSE-en-ru") model = AutoModel.from_pretrained("cointegrated/LaBSE-en-ru") sentences = ["Hello World", "Привет Мир"] encoded_input = tokenizer(sentences, padding=True, truncation=True, max_length=64, return_tensors='pt') with torch.no_grad(): model_output = model(**encoded_input) embeddings = model_output.pooler_output embeddings = torch.nn.functional.normalize(embeddings) print(embeddings) ``` The model has been truncated in [this notebook](https://colab.research.google.com/drive/1dnPRn0-ugj3vZgSpyCC9sgslM2SuSfHy?usp=sharing). You can adapt it for other languages (like [EIStakovskii/LaBSE-fr-de](https://huggingface.co/EIStakovskii/LaBSE-fr-de)), models or datasets. ## Reference: Fangxiaoyu Feng, Yinfei Yang, Daniel Cer, Narveen Ari, Wei Wang. [Language-agnostic BERT Sentence Embedding](https://arxiv.org/abs/2007.01852). July 2020 License: [https://tfhub.dev/google/LaBSE/1](https://tfhub.dev/google/LaBSE/1)

rinna/japanese-roberta-base

rinna

transformers

fill-mask

--- language: ja thumbnail: https://github.com/rinnakk/japanese-gpt2/blob/master/rinna.png tags: - ja - japanese - roberta - masked-lm - nlp license: mit datasets: - cc100 - wikipedia mask_token: "[MASK]" widget: - text: "[CLS]4年に1度[MASK]は開かれる。" --- # japanese-roberta-base  This repository provides a base-sized Japanese RoBERTa model. The model was trained using code from Github repository [rinnakk/japanese-pretrained-models](https://github.com/rinnakk/japanese-pretrained-models) by [rinna Co., Ltd.](https://corp.rinna.co.jp/) # How to load the model ~~~~ from transformers import AutoTokenizer, AutoModelForMaskedLM tokenizer = AutoTokenizer.from_pretrained("rinna/japanese-roberta-base", use_fast=False) tokenizer.do_lower_case = True # due to some bug of tokenizer config loading model = AutoModelForMaskedLM.from_pretrained("rinna/japanese-roberta-base") ~~~~ # How to use the model for masked token prediction ## Note 1: Use `[CLS]` To predict a masked token, be sure to add a `[CLS]` token before the sentence for the model to correctly encode it, as it is used during the model training. ## Note 2: Use `[MASK]` after tokenization A) Directly typing `[MASK]` in an input string and B) replacing a token with `[MASK]` after tokenization will yield different token sequences, and thus different prediction results. It is more appropriate to use `[MASK]` after tokenization (as it is consistent with how the model was pretrained). However, the Huggingface Inference API only supports typing `[MASK]` in the input string and produces less robust predictions. ## Note 3: Provide `position_ids` as an argument explicitly When `position_ids` are not provided for a `Roberta*` model, Huggingface's `transformers` will automatically construct it but start from `padding_idx` instead of `0` (see [issue](https://github.com/rinnakk/japanese-pretrained-models/issues/3) and function `create_position_ids_from_input_ids()` in Huggingface's [implementation](https://github.com/huggingface/transformers/blob/master/src/transformers/models/roberta/modeling_roberta.py)), which unfortunately does not work as expected with `rinna/japanese-roberta-base` since the `padding_idx` of the corresponding tokenizer is not `0`. So please be sure to constrcut the `position_ids` by yourself and make it start from position id `0`. ## Example Here is an example by to illustrate how our model works as a masked language model. Notice the difference between running the following code example and running the Huggingface Inference API. ~~~~ # original text text = "4年に1度オリンピックは開かれる。" # prepend [CLS] text = "[CLS]" + text # tokenize tokens = tokenizer.tokenize(text) print(tokens) # output: ['[CLS]', '▁4', '年に', '1', '度', 'オリンピック', 'は', '開かれる', '。'] # mask a token masked_idx = 5 tokens[masked_idx] = tokenizer.mask_token print(tokens) # output: ['[CLS]', '▁4', '年に', '1', '度', '[MASK]', 'は', '開かれる', '。'] # convert to ids token_ids = tokenizer.convert_tokens_to_ids(tokens) print(token_ids) # output: [4, 1602, 44, 24, 368, 6, 11, 21583, 8] # convert to tensor import torch token_tensor = torch.LongTensor([token_ids]) # provide position ids explicitly position_ids = list(range(0, token_tensor.size(1))) print(position_ids) # output: [0, 1, 2, 3, 4, 5, 6, 7, 8] position_id_tensor = torch.LongTensor([position_ids]) # get the top 10 predictions of the masked token with torch.no_grad(): outputs = model(input_ids=token_tensor, position_ids=position_id_tensor) predictions = outputs[0][0, masked_idx].topk(10) for i, index_t in enumerate(predictions.indices): index = index_t.item() token = tokenizer.convert_ids_to_tokens([index])[0] print(i, token) """ 0 総会 1 サミット 2 ワールドカップ 3 フェスティバル 4 大会 5 オリンピック 6 全国大会 7 党大会 8 イベント 9 世界選手権 """ ~~~~ # Model architecture A 12-layer, 768-hidden-size transformer-based masked language model. # Training The model was trained on [Japanese CC-100](http://data.statmt.org/cc-100/ja.txt.xz) and [Japanese Wikipedia](https://dumps.wikimedia.org/jawiki/) to optimize a masked language modelling objective on 8*V100 GPUs for around 15 days. It reaches ~3.9 perplexity on a dev set sampled from CC-100. # Tokenization The model uses a [sentencepiece](https://github.com/google/sentencepiece)-based tokenizer, the vocabulary was trained on the Japanese Wikipedia using the official sentencepiece training script. # Licenese [The MIT license](https://opensource.org/licenses/MIT)

timm/levit_128.fb_dist_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_tag: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for levit_128.fb_dist_in1k A LeViT image classification model using convolutional mode (using nn.Conv2d and nn.BatchNorm2d). Pretrained on ImageNet-1k using distillation by paper authors. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 9.2 - GMACs: 0.4 - Activations (M): 2.7 - Image size: 224 x 224 - **Papers:** - LeViT: a Vision Transformer in ConvNet's Clothing for Faster Inference: https://arxiv.org/abs/2104.01136 - **Original:** https://github.com/facebookresearch/LeViT - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open( urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png')) model = timm.create_model('levit_128.fb_dist_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open( urlopen('https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png')) model = timm.create_model( 'levit_128.fb_dist_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled (ie.e a (batch_size, num_features, H, W) tensor output = model.forward_head(output, pre_logits=True) # output is (batch_size, num_features) tensor ``` ## Model Comparison |model |top1 |top5 |param_count|img_size| |-----------------------------------|------|------|-----------|--------| |levit_384.fb_dist_in1k |82.596|96.012|39.13 |224 | |levit_conv_384.fb_dist_in1k |82.596|96.012|39.13 |224 | |levit_256.fb_dist_in1k |81.512|95.48 |18.89 |224 | |levit_conv_256.fb_dist_in1k |81.512|95.48 |18.89 |224 | |levit_conv_192.fb_dist_in1k |79.86 |94.792|10.95 |224 | |levit_192.fb_dist_in1k |79.858|94.792|10.95 |224 | |levit_128.fb_dist_in1k |78.474|94.014|9.21 |224 | |levit_conv_128.fb_dist_in1k |78.474|94.02 |9.21 |224 | |levit_128s.fb_dist_in1k |76.534|92.864|7.78 |224 | |levit_conv_128s.fb_dist_in1k |76.532|92.864|7.78 |224 | ## Citation ```bibtex @InProceedings{Graham_2021_ICCV, author = {Graham, Benjamin and El-Nouby, Alaaeldin and Touvron, Hugo and Stock, Pierre and Joulin, Armand and Jegou, Herve and Douze, Matthijs}, title = {LeViT: A Vision Transformer in ConvNet's Clothing for Faster Inference}, booktitle = {Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV)}, month = {October}, year = {2021}, pages = {12259-12269} } ``` ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/rwightman/pytorch-image-models}} } ```

m3hrdadfi/typo-detector-distilbert-en

m3hrdadfi

transformers

token-classification

--- language: en widget: - text: "He had also stgruggled with addiction during his time in Congress ." - text: "The review thoroughla assessed all aspects of JLENS SuR and CPG esign maturit and confidence ." - text: "Letterma also apologized two his staff for the satyation ." - text: "Vincent Jay had earlier won France 's first gold in gthe 10km biathlon sprint ." - text: "It is left to the directors to figure out hpw to bring the stry across to tye audience ." --- # Typo Detector ## Dataset Information For this specific task, I used [NeuSpell](https://github.com/neuspell/neuspell) corpus as my raw data. ## Evaluation The following tables summarize the scores obtained by model overall and per each class. | # | precision | recall | f1-score | support | |:------------:|:---------:|:--------:|:--------:|:--------:| | TYPO | 0.992332 | 0.985997 | 0.989154 | 416054.0 | | micro avg | 0.992332 | 0.985997 | 0.989154 | 416054.0 | | macro avg | 0.992332 | 0.985997 | 0.989154 | 416054.0 | | weighted avg | 0.992332 | 0.985997 | 0.989154 | 416054.0 | ## How to use You use this model with Transformers pipeline for NER (token-classification). ### Installing requirements ```bash pip install transformers ``` ### Prediction using pipeline ```python import torch from transformers import AutoConfig, AutoTokenizer, AutoModelForTokenClassification from transformers import pipeline model_name_or_path = "m3hrdadfi/typo-detector-distilbert-en" config = AutoConfig.from_pretrained(model_name_or_path) tokenizer = AutoTokenizer.from_pretrained(model_name_or_path) model = AutoModelForTokenClassification.from_pretrained(model_name_or_path, config=config) nlp = pipeline('token-classification', model=model, tokenizer=tokenizer, aggregation_strategy="average") ``` ```python sentences = [ "He had also stgruggled with addiction during his time in Congress .", "The review thoroughla assessed all aspects of JLENS SuR and CPG esign maturit and confidence .", "Letterma also apologized two his staff for the satyation .", "Vincent Jay had earlier won France 's first gold in gthe 10km biathlon sprint .", "It is left to the directors to figure out hpw to bring the stry across to tye audience .", ] for sentence in sentences: typos = [sentence[r["start"]: r["end"]] for r in nlp(sentence)] detected = sentence for typo in typos: detected = detected.replace(typo, f'<i>{typo}</i>') print(" [Input]: ", sentence) print("[Detected]: ", detected) print("-" * 130) ``` Output: ```text [Input]: He had also stgruggled with addiction during his time in Congress . [Detected]: He had also <i>stgruggled</i> with addiction during his time in Congress . ---------------------------------------------------------------------------------------------------------------------------------- [Input]: The review thoroughla assessed all aspects of JLENS SuR and CPG esign maturit and confidence . [Detected]: The review <i>thoroughla</i> assessed all aspects of JLENS SuR and CPG <i>esign</i> <i>maturit</i> and confidence . ---------------------------------------------------------------------------------------------------------------------------------- [Input]: Letterma also apologized two his staff for the satyation . [Detected]: <i>Letterma</i> also apologized <i>two</i> his staff for the <i>satyation</i> . ---------------------------------------------------------------------------------------------------------------------------------- [Input]: Vincent Jay had earlier won France 's first gold in gthe 10km biathlon sprint . [Detected]: Vincent Jay had earlier won France 's first gold in <i>gthe</i> 10km biathlon sprint . ---------------------------------------------------------------------------------------------------------------------------------- [Input]: It is left to the directors to figure out hpw to bring the stry across to tye audience . [Detected]: It is left to the directors to figure out <i>hpw</i> to bring the <i>stry</i> across to <i>tye</i> audience . ---------------------------------------------------------------------------------------------------------------------------------- ``` ## Questions? Post a Github issue on the [TypoDetector Issues](https://github.com/m3hrdadfi/typo-detector/issues) repo.

jackaduma/SecRoBERTa

jackaduma

transformers

fill-mask

--- language: en thumbnail: https://github.com/jackaduma tags: - exbert - security - cybersecurity - cyber security - threat hunting - threat intelligence license: apache-2.0 datasets: - APTnotes - Stucco-Data - CASIE --- # SecRoBERTa This is the pretrained model presented in [SecBERT: A Pretrained Language Model for Cyber Security Text](https://github.com/jackaduma/SecBERT/), which is a SecRoBERTa model trained on cyber security text. The training corpus was papers taken from * [APTnotes](https://github.com/kbandla/APTnotes) * [Stucco-Data: Cyber security data sources](https://stucco.github.io/data/) * [CASIE: Extracting Cybersecurity Event Information from Text](https://ebiquity.umbc.edu/_file_directory_/papers/943.pdf) * [SemEval-2018 Task 8: Semantic Extraction from CybersecUrity REports using Natural Language Processing (SecureNLP)](https://competitions.codalab.org/competitions/17262). SecRoBERTa has its own wordpiece vocabulary (secvocab) that's built to best match the training corpus. We trained [SecBERT](https://huggingface.co/jackaduma/SecBERT) and [SecRoBERTa](https://huggingface.co/jackaduma/SecRoBERTa) versions. Available models include: * [`SecBERT`](https://huggingface.co/jackaduma/SecBERT) * [`SecRoBERTa`](https://huggingface.co/jackaduma/SecRoBERTa) --- ## **Fill Mask** We proposed to build language model which work on cyber security text, as result, it can improve downstream tasks (NER, Text Classification, Semantic Understand, Q&A) in Cyber Security Domain. First, as below shows Fill-Mask pipeline in [Google Bert](), [AllenAI SciBert](https://github.com/allenai/scibert) and our [SecBERT](https://github.com/jackaduma/SecBERT) . <!-- <img src="./fill-mask-result.png" width="150%" height="150%"> -->  --- The original repo can be found [here](https://github.com/jackaduma/SecBERT).

timm/pnasnet5large.tf_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for pnasnet5large.tf_in1k A PNasNet image classification model. Trained on ImageNet-1k by paper authors. Ported from Tensorflow via Cadene's pretrained-models.pytorch. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 86.1 - GMACs: 25.0 - Activations (M): 92.9 - Image size: 331 x 331 - **Papers:** - Progressive Neural Architecture Search: https://arxiv.org/abs/1712.00559 - **Original:** - https://github.com/tensorflow/models - https://github.com/Cadene/pretrained-models.pytorch - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('pnasnet5large.tf_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'pnasnet5large.tf_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 96, 165, 165]) # torch.Size([1, 270, 83, 83]) # torch.Size([1, 1080, 42, 42]) # torch.Size([1, 2160, 21, 21]) # torch.Size([1, 4320, 11, 11]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'pnasnet5large.tf_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 4320, 11, 11) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @misc{liu2018progressive, title={Progressive Neural Architecture Search}, author={Chenxi Liu and Barret Zoph and Maxim Neumann and Jonathon Shlens and Wei Hua and Li-Jia Li and Li Fei-Fei and Alan Yuille and Jonathan Huang and Kevin Murphy}, year={2018}, eprint={1712.00559}, archivePrefix={arXiv}, primaryClass={cs.CV} } ```

Qiliang/bart-large-cnn-samsum-ChatGPT_v3

Qiliang

transformers

text2text-generation

--- license: mit tags: - generated_from_trainer model-index: - name: bart-large-cnn-samsum-ChatGPT_v3 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # bart-large-cnn-samsum-ChatGPT_v3 This model is a fine-tuned version of [philschmid/bart-large-cnn-samsum](https://huggingface.co/philschmid/bart-large-cnn-samsum) on an unknown dataset. ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 2e-05 - train_batch_size: 4 - eval_batch_size: 4 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 - mixed_precision_training: Native AMP ### Framework versions - Transformers 4.24.0 - Pytorch 1.12.1 - Datasets 2.6.1 - Tokenizers 0.13.2

TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ

TheBloke

transformers

text-generation

--- language: - en license: other tags: - uncensored datasets: - ehartford/wizard_vicuna_70k_unfiltered model_name: Wizard Vicuna 13B Uncensored base_model: ehartford/Wizard-Vicuna-13B-Uncensored inference: false model_creator: Eric Hartford model_type: llama prompt_template: 'A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user''s questions. USER: {prompt} ASSISTANT: ' quantized_by: TheBloke --- <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Wizard Vicuna 13B Uncensored - GPTQ - Model creator: [Eric Hartford](https://huggingface.co/ehartford) - Original model: [Wizard Vicuna 13B Uncensored](https://huggingface.co/ehartford/Wizard-Vicuna-13B-Uncensored) <!-- description start --> ## Description This repo contains GPTQ model files for [Eric Hartford's Wizard Vicuna 13B Uncensored](https://huggingface.co/ehartford/Wizard-Vicuna-13B-Uncensored). Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them. <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GGUF) * [Eric Hartford's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/ehartford/Wizard-Vicuna-13B-Uncensored) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Vicuna ``` A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT: ``` <!-- prompt-template end --> <!-- README_GPTQ.md-provided-files start --> ## Provided files and GPTQ parameters Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements. Each separate quant is in a different branch. See below for instructions on fetching from different branches. All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches are made with AutoGPTQ. Files in the `main` branch which were uploaded before August 2023 were made with GPTQ-for-LLaMa. <details> <summary>Explanation of GPTQ parameters</summary> - Bits: The bit size of the quantised model. - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value. - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now. - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy. - GPTQ dataset: The dataset used for quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s). - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences. - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama models in 4-bit. </details> | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc | | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- | | [latest](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ/tree/latest) | 4 | 128 | Yes | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 2048 | 8.11 GB | Yes | 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. | | [model_v1](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ/tree/model_v1) | 4 | 128 | No | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 2048 | 8.11 GB | Yes | 4-bit, without Act Order and group size 128g. | | [main](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ/tree/main) | 4 | 128 | No | 0.01 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 2048 | 8.11 GB | Yes | 4-bit, without Act Order and group size 128g. | <!-- README_GPTQ.md-provided-files end --> <!-- README_GPTQ.md-download-from-branches start --> ## How to download from branches - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ:latest` - With Git, you can clone a branch with: ``` git clone --single-branch --branch latest https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ ``` - In Python Transformers code, the branch is the `revision` parameter; see below. <!-- README_GPTQ.md-download-from-branches end --> <!-- README_GPTQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui). Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the **Model tab**. 2. Under **Download custom model or LoRA**, enter `TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ`. - To download from a specific branch, enter for example `TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ:latest` - see Provided Files above for the list of branches for each option. 3. Click **Download**. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to **Model**. 6. In the **Model** dropdown, choose the model you just downloaded: `Wizard-Vicuna-13B-Uncensored-GPTQ` 7. The model will automatically load, and is now ready for use! 8. If you want any custom settings, set them and then click **Save settings for this model** followed by **Reload the Model** in the top right. * Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`. 9. Once you're ready, click the **Text Generation tab** and enter a prompt to get started! <!-- README_GPTQ.md-text-generation-webui end --> <!-- README_GPTQ.md-use-from-python start --> ## How to use this GPTQ model from Python code ### Install the necessary packages Requires: Transformers 4.32.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later. ```shell pip3 install transformers>=4.32.0 optimum>=1.12.0 pip3 install auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ # Use cu117 if on CUDA 11.7 ``` If you have problems installing AutoGPTQ using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y auto-gptq git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ pip3 install . ``` ### For CodeLlama models only: you must use Transformers 4.33.0 or later. If 4.33.0 is not yet released when you read this, you will need to install Transformers from source: ```shell pip3 uninstall -y transformers pip3 install git+https://github.com/huggingface/transformers.git ``` ### You can then use the following code ```python from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline model_name_or_path = "TheBloke/Wizard-Vicuna-13B-Uncensored-GPTQ" # To use a different branch, change revision # For example: revision="latest" model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", trust_remote_code=False, revision="main") tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True) prompt = "Tell me about AI" prompt_template=f'''A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT: ''' print("\n\n*** Generate:") input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda() output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512) print(tokenizer.decode(output[0])) # Inference can also be done using transformers' pipeline print("*** Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_GPTQ.md-use-from-python end --> <!-- README_GPTQ.md-compatibility start --> ## Compatibility The files provided are tested to work with AutoGPTQ, both via Transformers and using AutoGPTQ directly. They should also work with [Occ4m's GPTQ-for-LLaMa fork](https://github.com/0cc4m/KoboldAI). [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama models in 4-bit. Please see the Provided Files table above for per-file compatibility. [Huggingface Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) is compatible with all GPTQ models. <!-- README_GPTQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI **Special thanks to**: Aemon Algiz. **Patreon special mentions**: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: Eric Hartford's Wizard Vicuna 13B Uncensored This is [wizard-vicuna-13b](https://huggingface.co/junelee/wizard-vicuna-13b) trained with a subset of the dataset - responses that contained alignment / moralizing were removed. The intent is to train a WizardLM that doesn't have alignment built-in, so that alignment (of any sort) can be added separately with for example with a RLHF LoRA. Shout out to the open source AI/ML community, and everyone who helped me out. Note: An uncensored model has no guardrails. You are responsible for anything you do with the model, just as you are responsible for anything you do with any dangerous object such as a knife, gun, lighter, or car. Publishing anything this model generates is the same as publishing it yourself. You are responsible for the content you publish, and you cannot blame the model any more than you can blame the knife, gun, lighter, or car for what you do with it.

digiplay/AbsoluteReality_v1.8.1

digiplay

diffusers

text-to-image

--- license: other tags: - stable-diffusion - stable-diffusion-diffusers - text-to-image - diffusers inference: true --- Model info : https://civitai.com/models/81458?modelVersionId=132760 Original Author's DEMO images : ,%20(extremely%20intricate_1.3),,%20(realistic),%20portrait%20of%20a%20girl,%20the%20most%20beautiful%20in%20the%20world,%20(medieval%20armor),%20m.jpeg) .jpeg) ),%20intricate,%20(steel%20metal%20[rust]),%20elegant,.jpeg)

philschmid/distilbert-onnx

philschmid

transformers

question-answering

--- language: "en" datasets: - squad metrics: - squad license: apache-2.0 --- # ONNX Conversion of [distilbert-base-cased-distilled-squad](https://huggingface.co/distilbert-base-cased-distilled-squad) # DistilBERT base cased distilled SQuAD This model is a fine-tune checkpoint of [DistilBERT-base-cased](https://huggingface.co/distilbert-base-cased), fine-tuned using (a second step of) knowledge distillation on SQuAD v1.1. This model reaches a F1 score of 87.1 on the dev set (for comparison, BERT bert-base-cased version reaches a F1 score of 88.7).

snrspeaks/t5-one-line-summary

snrspeaks

transformers

text2text-generation

--- datasets: - arxiv widget: - text: "summarize: We describe a system called Overton, whose main design goal is to support engineers in building, monitoring, and improving production machinelearning systems. Key challenges engineers face are monitoring fine-grained quality, diagnosing errors in sophisticated applications, and handling contradictory or incomplete supervision data. Overton automates the life cycle of model construction, deployment, and monitoring by providing a set of novel high-level, declarative abstractions. Overton's vision is to shift developers to these higher-level tasks instead of lower-level machine learning tasks. In fact, using Overton, engineers can build deep-learning-based applications without writing any code in frameworks like TensorFlow. For over a year, Overton has been used in production to support multiple applications in both near-real-time applications and back-of-house processing. In that time, Overton-based applications have answered billions of queries in multiple languages and processed trillions of records reducing errors 1.7-2.9 times versus production systems." license: mit --- # T5 One Line Summary A T5 model trained on 370,000 research papers, to generate one line summary based on description/abstract of the papers. It is trained using [simpleT5](https://github.com/Shivanandroy/simpleT5) library - A python package built on top of pytorch lightning⚡️ & transformers🤗 to quickly train T5 models ## Usage:[](https://colab.research.google.com/drive/1HrfT8IKLXvZzPFpl1EhZ3s_iiXG3O2VY?usp=sharing) ```python abstract = """We describe a system called Overton, whose main design goal is to support engineers in building, monitoring, and improving production machine learning systems. Key challenges engineers face are monitoring fine-grained quality, diagnosing errors in sophisticated applications, and handling contradictory or incomplete supervision data. Overton automates the life cycle of model construction, deployment, and monitoring by providing a set of novel high-level, declarative abstractions. Overton's vision is to shift developers to these higher-level tasks instead of lower-level machine learning tasks. In fact, using Overton, engineers can build deep-learning-based applications without writing any code in frameworks like TensorFlow. For over a year, Overton has been used in production to support multiple applications in both near-real-time applications and back-of-house processing. In that time, Overton-based applications have answered billions of queries in multiple languages and processed trillions of records reducing errors 1.7-2.9 times versus production systems. """ ``` ### Using Transformers🤗 ```python model_name = "snrspeaks/t5-one-line-summary" from transformers import AutoModelForSeq2SeqLM, AutoTokenizer model = AutoModelForSeq2SeqLM.from_pretrained(model_name) tokenizer = AutoTokenizer.from_pretrained(model_name) input_ids = tokenizer.encode("summarize: " + abstract, return_tensors="pt", add_special_tokens=True) generated_ids = model.generate(input_ids=input_ids,num_beams=5,max_length=50,repetition_penalty=2.5,length_penalty=1,early_stopping=True,num_return_sequences=3) preds = [tokenizer.decode(g, skip_special_tokens=True, clean_up_tokenization_spaces=True) for g in generated_ids] print(preds) # output ["Overton: Building, Deploying, and Monitoring Machine Learning Systems for Engineers", "Overton: A System for Building, Monitoring, and Improving Production Machine Learning Systems", "Overton: Building, Monitoring, and Improving Production Machine Learning Systems"] ``` ### Using simpleT5⚡️ ```python # pip install --upgrade simplet5 from simplet5 import SimpleT5 model = SimpleT5() model.load_model("t5","snrspeaks/t5-one-line-summary") model.predict(abstract) # output "Overton: Building, Deploying, and Monitoring Machine Learning Systems for Engineers" ```

amberoad/bert-multilingual-passage-reranking-msmarco

amberoad

transformers

text-classification

--- language: - multilingual - af - sq - ar - an - hy - ast - az - ba - eu - bar - be - bn - inc - bs - br - bg - my - ca - ceb - ce - zh - cv - hr - cs - da - nl - en - et - fi - fr - gl - ka - de - el - gu - ht - he - hi - hu - is - io - id - ga - it - ja - jv - kn - kk - ky - ko - la - lv - lt - roa - nds - lm - mk - mg - ms - ml - mr - min - ne - new - nb - nn - oc - fa - pms - pl - pt - pa - ro - ru - sco - sr - hr - scn - sk - sl - aze - es - su - sw - sv - tl - tg - ta - tt - te - tr - uk - ud - uz - vi - vo - war - cy - fry - pnb - yo thumbnail: https://amberoad.de/images/logo_text.png tags: - msmarco - multilingual - passage reranking license: apache-2.0 datasets: - msmarco metrics: - MRR widget: - query: What is a corporation? passage: A company is incorporated in a specific nation, often within the bounds of a smaller subset of that nation, such as a state or province. The corporation is then governed by the laws of incorporation in that state. A corporation may issue stock, either private or public, or may be classified as a non-stock corporation. If stock is issued, the corporation will usually be governed by its shareholders, either directly or indirectly. --- # Passage Reranking Multilingual BERT 🔃 🌍 ## Model description **Input:** Supports over 100 Languages. See [List of supported languages](https://github.com/google-research/bert/blob/master/multilingual.md#list-of-languages) for all available. **Purpose:** This module takes a search query [1] and a passage [2] and calculates if the passage matches the query. It can be used as an improvement for Elasticsearch Results and boosts the relevancy by up to 100%. **Architecture:** On top of BERT there is a Densly Connected NN which takes the 768 Dimensional [CLS] Token as input and provides the output ([Arxiv](https://arxiv.org/abs/1901.04085)). **Output:** Just a single value between between -10 and 10. Better matching query,passage pairs tend to have a higher a score. ## Intended uses & limitations Both query[1] and passage[2] have to fit in 512 Tokens. As you normally want to rerank the first dozens of search results keep in mind the inference time of approximately 300 ms/query. #### How to use ```python from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("amberoad/bert-multilingual-passage-reranking-msmarco") model = AutoModelForSequenceClassification.from_pretrained("amberoad/bert-multilingual-passage-reranking-msmarco") ``` This Model can be used as a drop-in replacement in the [Nboost Library](https://github.com/koursaros-ai/nboost) Through this you can directly improve your Elasticsearch Results without any coding. ## Training data This model is trained using the [**Microsoft MS Marco Dataset**](https://microsoft.github.io/msmarco/ "Microsoft MS Marco"). This training dataset contains approximately 400M tuples of a query, relevant and non-relevant passages. All datasets used for training and evaluating are listed in this [table](https://github.com/microsoft/MSMARCO-Passage-Ranking#data-information-and-formating). The used dataset for training is called *Train Triples Large*, while the evaluation was made on *Top 1000 Dev*. There are 6,900 queries in total in the development dataset, where each query is mapped to top 1,000 passage retrieved using BM25 from MS MARCO corpus. ## Training procedure The training is performed the same way as stated in this [README](https://github.com/nyu-dl/dl4marco-bert "NYU Github"). See their excellent Paper on [Arxiv](https://arxiv.org/abs/1901.04085). We changed the BERT Model from an English only to the default BERT Multilingual uncased Model from [Google](https://huggingface.co/bert-base-multilingual-uncased). Training was done 400 000 Steps. This equaled 12 hours an a TPU V3-8. ## Eval results We see nearly similar performance than the English only Model in the English [Bing Queries Dataset](http://www.msmarco.org/). Although the training data is English only internal Tests on private data showed a far higher accurancy in German than all other available models. Fine-tuned Models | Dependency | Eval Set | Search Boost<a href='#benchmarks'> | Speed on GPU ----------------------------------------------------------------------------------- | ---------------------------------------------------------------------------- | ------------------------------------------------------------------ | ----------------------------------------------------- | ---------------------------------- **`amberoad/Multilingual-uncased-MSMARCO`** (This Model) | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-blue"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+61%** _{^{(0.29 vs 0.18)}} | ~300 ms/query <a href='#footnotes'> `nboost/pt-tinybert-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+45%** _{^{(0.26 vs 0.18)}} | ~50ms/query <a href='#footnotes'> `nboost/pt-bert-base-uncased-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+62%** _{^{(0.29 vs 0.18)}} | ~300 ms/query<a href='#footnotes'> `nboost/pt-bert-large-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='http://www.msmarco.org/'>bing queries</a> | **+77%** _{^{(0.32 vs 0.18)}} | - `nboost/pt-biobert-base-msmarco` | <img alt="PyTorch" src="https://img.shields.io/badge/PyTorch-red"/> | <a href ='https://github.com/naver/biobert-pretrained'>biomed</a> | **+66%** _{^{(0.17 vs 0.10)}} | ~300 ms/query<a href='#footnotes'> This table is taken from [nboost](https://github.com/koursaros-ai/nboost) and extended by the first line. ## Contact Infos  Amberoad is a company focussing on Search and Business Intelligence. We provide you: * Advanced Internal Company Search Engines thorugh NLP * External Search Egnines: Find Competitors, Customers, Suppliers **Get in Contact now to benefit from our Expertise:** The training and evaluation was performed by [**Philipp Reissel**](https://reissel.eu/) and [**Igli Manaj**](https://github.com/iglimanaj) [ Linkedin](https://de.linkedin.com/company/amberoad) | <svg xmlns="http://www.w3.org/2000/svg" x="0px" y="0px" width="32" height="32" viewBox="0 0 172 172" style=" fill:#000000;"><g fill="none" fill-rule="nonzero" stroke="none" stroke-width="1" stroke-linecap="butt" stroke-linejoin="miter" stroke-miterlimit="10" stroke-dasharray="" stroke-dashoffset="0" font-family="none" font-weight="none" font-size="none" text-anchor="none" style="mix-blend-mode: normal"><path d="M0,172v-172h172v172z" fill="none"></path><g fill="#e67e22"><path d="M37.625,21.5v86h96.75v-86h-5.375zM48.375,32.25h10.75v10.75h-10.75zM69.875,32.25h10.75v10.75h-10.75zM91.375,32.25h32.25v10.75h-32.25zM48.375,53.75h75.25v43h-75.25zM80.625,112.875v17.61572c-1.61558,0.93921 -2.94506,2.2687 -3.88428,3.88428h-49.86572v10.75h49.86572c1.8612,3.20153 5.28744,5.375 9.25928,5.375c3.97183,0 7.39808,-2.17347 9.25928,-5.375h49.86572v-10.75h-49.86572c-0.93921,-1.61558 -2.2687,-2.94506 -3.88428,-3.88428v-17.61572z"></path></g></g></svg>[Homepage](https://de.linkedin.com/company/amberoad) | [Email](info@amberoad.de)

raminm/fundhub_category_v1

raminm

transformers

text-classification

--- license: apache-2.0 base_model: bert-base-uncased tags: - generated_from_trainer metrics: - f1 model-index: - name: fundhub_category_v1 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # fundhub_category_v1 This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.1340 - F1: 0.9626 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:-----:|:---------------:|:------:| | 0.1648 | 1.0 | 25275 | 0.1583 | 0.9512 | | 0.1396 | 2.0 | 50550 | 0.1578 | 0.9583 | | 0.0934 | 3.0 | 75825 | 0.1340 | 0.9626 | ### Framework versions - Transformers 4.34.0 - Pytorch 2.1.0+cu121 - Datasets 2.14.4 - Tokenizers 0.14.1

google/byt5-base

google

transformers

text2text-generation

--- language: - multilingual - af - am - ar - az - be - bg - bn - ca - ceb - co - cs - cy - da - de - el - en - eo - es - et - eu - fa - fi - fil - fr - fy - ga - gd - gl - gu - ha - haw - hi - hmn - ht - hu - hy - ig - is - it - iw - ja - jv - ka - kk - km - kn - ko - ku - ky - la - lb - lo - lt - lv - mg - mi - mk - ml - mn - mr - ms - mt - my - ne - nl - no - ny - pa - pl - ps - pt - ro - ru - sd - si - sk - sl - sm - sn - so - sq - sr - st - su - sv - sw - ta - te - tg - th - tr - uk - und - ur - uz - vi - xh - yi - yo - zh - zu datasets: - mc4 license: apache-2.0 --- # ByT5 - Base ByT5 is a tokenizer-free version of [Google's T5](https://ai.googleblog.com/2020/02/exploring-transfer-learning-with-t5.html) and generally follows the architecture of [MT5](https://huggingface.co/google/mt5-base). ByT5 was only pre-trained on [mC4](https://www.tensorflow.org/datasets/catalog/c4#c4multilingual) excluding any supervised training with an average span-mask of 20 UTF-8 characters. Therefore, this model has to be fine-tuned before it is useable on a downstream task. ByT5 works especially well on noisy text data,*e.g.*, `google/byt5-base` significantly outperforms [mt5-base](https://huggingface.co/google/mt5-base) on [TweetQA](https://arxiv.org/abs/1907.06292). Paper: [ByT5: Towards a token-free future with pre-trained byte-to-byte models](https://arxiv.org/abs/2105.13626) Authors: *Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, Colin Raffel* ## Example Inference ByT5 works on raw UTF-8 bytes and can be used without a tokenizer: ```python from transformers import T5ForConditionalGeneration import torch model = T5ForConditionalGeneration.from_pretrained('google/byt5-base') input_ids = torch.tensor([list("Life is like a box of chocolates.".encode("utf-8"))]) + 3 # add 3 for special tokens labels = torch.tensor([list("La vie est comme une boîte de chocolat.".encode("utf-8"))]) + 3 # add 3 for special tokens loss = model(input_ids, labels=labels).loss # forward pass ``` For batched inference & training it is however recommended using a tokenizer class for padding: ```python from transformers import T5ForConditionalGeneration, AutoTokenizer model = T5ForConditionalGeneration.from_pretrained('google/byt5-base') tokenizer = AutoTokenizer.from_pretrained('google/byt5-base') model_inputs = tokenizer(["Life is like a box of chocolates.", "Today is Monday."], padding="longest", return_tensors="pt") labels = tokenizer(["La vie est comme une boîte de chocolat.", "Aujourd'hui c'est lundi."], padding="longest", return_tensors="pt").input_ids loss = model(**model_inputs, labels=labels).loss # forward pass ``` ## Abstract Most widely-used pre-trained language models operate on sequences of tokens corresponding to word or subword units. Encoding text as a sequence of tokens requires a tokenizer, which is typically created as an independent artifact from the model. Token-free models that instead operate directly on raw text (bytes or characters) have many benefits: they can process text in any language out of the box, they are more robust to noise, and they minimize technical debt by removing complex and error-prone text preprocessing pipelines. Since byte or character sequences are longer than token sequences, past work on token-free models has often introduced new model architectures designed to amortize the cost of operating directly on raw text. In this paper, we show that a standard Transformer architecture can be used with minimal modifications to process byte sequences. We carefully characterize the trade-offs in terms of parameter count, training FLOPs, and inference speed, and show that byte-level models are competitive with their token-level counterparts. We also demonstrate that byte-level models are significantly more robust to noise and perform better on tasks that are sensitive to spelling and pronunciation. As part of our contribution, we release a new set of pre-trained byte-level Transformer models based on the T5 architecture, as well as all code and data used in our experiments. 

raminm/fundhub_subcategory_v1

raminm

transformers

text-classification

--- license: apache-2.0 base_model: bert-base-uncased tags: - generated_from_trainer metrics: - f1 model-index: - name: fundhub_subcategory_v1 results: [] --- <!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. --> # fundhub_subcategory_v1 This model is a fine-tuned version of [bert-base-uncased](https://huggingface.co/bert-base-uncased) on the None dataset. It achieves the following results on the evaluation set: - Loss: 0.3386 - F1: 0.8965 ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-05 - train_batch_size: 16 - eval_batch_size: 8 - seed: 42 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 - lr_scheduler_type: linear - num_epochs: 3 ### Training results | Training Loss | Epoch | Step | Validation Loss | F1 | |:-------------:|:-----:|:-----:|:---------------:|:------:| | 0.4921 | 1.0 | 25275 | 0.4638 | 0.8635 | | 0.3574 | 2.0 | 50550 | 0.3744 | 0.8836 | | 0.2978 | 3.0 | 75825 | 0.3386 | 0.8965 | ### Framework versions - Transformers 4.34.0 - Pytorch 2.1.0+cu121 - Datasets 2.14.4 - Tokenizers 0.14.1

openai-gpt

transformers

text-generation

--- language: en license: mit --- # OpenAI GPT ## Table of Contents - [Model Details](#model-details) - [How To Get Started With the Model](#how-to-get-started-with-the-model) - [Uses](#uses) - [Risks, Limitations and Biases](#risks-limitations-and-biases) - [Training](#training) - [Evaluation](#evaluation) - [Environmental Impact](#environmental-impact) - [Technical Specifications](#technical-specifications) - [Citation Information](#citation-information) - [Model Card Authors](#model-card-authors) ## Model Details **Model Description:** `openai-gpt` is a transformer-based language model created and released by OpenAI. The model is a causal (unidirectional) transformer pre-trained using language modeling on a large corpus with long range dependencies. - **Developed by:** Alec Radford, Karthik Narasimhan, Tim Salimans, Ilya Sutskever. See [associated research paper](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf) and [GitHub repo](https://github.com/openai/finetune-transformer-lm) for model developers and contributors. - **Model Type:** Transformer-based language model - **Language(s):** English - **License:** [MIT License](https://github.com/openai/finetune-transformer-lm/blob/master/LICENSE) - **Related Models:** [GPT2](https://huggingface.co/gpt2), [GPT2-Medium](https://huggingface.co/gpt2-medium), [GPT2-Large](https://huggingface.co/gpt2-large) and [GPT2-XL](https://huggingface.co/gpt2-xl) - **Resources for more information:** - [Research Paper](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf) - [OpenAI Blog Post](https://openai.com/blog/language-unsupervised/) - [GitHub Repo](https://github.com/openai/finetune-transformer-lm) - Test the full generation capabilities here: https://transformer.huggingface.co/doc/gpt ## How to Get Started with the Model Use the code below to get started with the model. You can use this model directly with a pipeline for text generation. Since the generation relies on some randomness, we set a seed for reproducibility: ```python >>> from transformers import pipeline, set_seed >>> generator = pipeline('text-generation', model='openai-gpt') >>> set_seed(42) >>> generator("Hello, I'm a language model,", max_length=30, num_return_sequences=5) [{'generated_text': "Hello, I'm a language model,'he said, when i was finished.'ah well,'said the man,'that's"}, {'generated_text': 'Hello, I\'m a language model, " she said. \n she reached the bottom of the shaft and leaned a little further out. it was'}, {'generated_text': 'Hello, I\'m a language model, " she laughed. " we call that a\'white girl.\'or as we are called by the'}, {'generated_text': 'Hello, I\'m a language model, " said mr pin. " an\'the ones with the funny hats don\'t. " the rest of'}, {'generated_text': 'Hello, I\'m a language model, was\'ere \'bout to do some more dancin \', " he said, then his voice lowered to'}] ``` Here is how to use this model in PyTorch: ```python from transformers import OpenAIGPTTokenizer, OpenAIGPTModel import torch tokenizer = OpenAIGPTTokenizer.from_pretrained("openai-gpt") model = OpenAIGPTModel.from_pretrained("openai-gpt") inputs = tokenizer("Hello, my dog is cute", return_tensors="pt") outputs = model(**inputs) last_hidden_states = outputs.last_hidden_state ``` and in TensorFlow: ```python from transformers import OpenAIGPTTokenizer, TFOpenAIGPTModel tokenizer = OpenAIGPTTokenizer.from_pretrained("openai-gpt") model = TFOpenAIGPTModel.from_pretrained("openai-gpt") inputs = tokenizer("Hello, my dog is cute", return_tensors="tf") outputs = model(inputs) last_hidden_states = outputs.last_hidden_state ``` ## Uses #### Direct Use This model can be used for language modeling tasks. #### Downstream Use Potential downstream uses of this model include tasks that leverage language models. In the [associated paper](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf), the model developers discuss evaluations of the model for tasks including natural language inference (NLI), question answering, semantic similarity, and text classification. #### Misuse and Out-of-scope Use The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model. ## Risks, Limitations and Biases #### Biases **CONTENT WARNING: Readers should be aware that language generated by this model can be disturbing or offensive to some and can propagate historical and current stereotypes.** Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). Predictions generated by this model can include disturbing and harmful stereotypes across protected classes; identity characteristics; and sensitive, social, and occupational groups. For example: ```python >>> from transformers import pipeline, set_seed >>> generator = pipeline('text-generation', model='openai-gpt') >>> set_seed(42) >>> generator("The man worked as a", max_length=10, num_return_sequences=5) [{'generated_text': 'The man worked as a teacher for the college he'}, {'generated_text': 'The man worked as a janitor at the club.'}, {'generated_text': 'The man worked as a bodyguard in america. the'}, {'generated_text': 'The man worked as a clerk for one of the'}, {'generated_text': 'The man worked as a nurse, but there was'}] >>> set_seed(42) >>> generator("The woman worked as a", max_length=10, num_return_sequences=5) [{'generated_text': 'The woman worked as a medical intern but is a'}, {'generated_text': 'The woman worked as a midwife, i know that'}, {'generated_text': 'The woman worked as a prostitute in a sex club'}, {'generated_text': 'The woman worked as a secretary for one of the'}, {'generated_text': 'The woman worked as a nurse, but she had'}] ``` This bias may also affect fine-tuned versions of this model. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. #### Risks and Limitations The model developers also wrote in a [blog post](https://openai.com/blog/language-unsupervised/) about risks and limitations of the model, including: > - **Compute Requirements:** Many previous approaches to NLP tasks train relatively small models on a single GPU from scratch. Our approach requires an expensive pre-training step - 1 month on 8 GPUs. Luckily, this only has to be done once and we’re releasing our model so others can avoid it. It is also a large model (in comparison to prior work) and consequently uses more compute and memory — we used a 37-layer (12 block) Transformer architecture, and we train on sequences of up to 512 tokens. Most experiments were conducted on 4 and 8 GPU systems. The model does fine-tune to new tasks very quickly which helps mitigate the additional resource requirements. > - **The limits and bias of learning about the world through text:** Books and text readily available on the internet do not contain complete or even accurate information about the world. Recent work ([Lucy and Gauthier, 2017](https://arxiv.org/abs/1705.11168)) has shown that certain kinds of information are difficult to learn via just text and other work ([Gururangan et al., 2018](https://arxiv.org/abs/1803.02324)) has shown that models learn and exploit biases in data distributions. > - **Still brittle generalization:** Although our approach improves performance across a broad range of tasks, current deep learning NLP models still exhibit surprising and counterintuitive behavior - especially when evaluated in a systematic, adversarial, or out-of-distribution way. Our approach is not immune to these issues, though we have observed some indications of progress. Our approach shows improved lexical robustness over previous purely neural approaches to textual entailment. On the dataset introduced in Glockner et al. (2018) our model achieves 83.75%, performing similarly to KIM, which incorporates external knowledge via WordNet. ## Training #### Training Data The model developers [write](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf): > We use the BooksCorpus dataset ([Zhu et al., 2015](https://www.cv-foundation.org/openaccess/content_iccv_2015/papers/Zhu_Aligning_Books_and_ICCV_2015_paper.pdf)) for training the language model. It contains over 7,000 unique unpublished books from a variety of genres including Adventure, Fantasy, and Romance. Crucially, it contains long stretches of contiguous text, which allows the generative model to learn to condition on long-range information. #### Training Procedure The model developers [write](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf): > Our model largely follows the original transformer work [62]. We trained a 12-layer decoder-only transformer with masked self-attention heads (768 dimensional states and 12 attention heads). For the position-wise feed-forward networks, we used 3072 dimensional inner states. We used the Adam optimization scheme [27] with a max learning rate of 2.5e-4. The learning rate was increased linearly from zero over the first 2000 updates and annealed to 0 using a cosine schedule. We train for 100 epochs on minibatches of 64 randomly sampled, contiguous sequences of 512 tokens. Since layernorm [2] is used extensively throughout the model, a simple weight initialization of N (0, 0.02) was sufficient. We used a bytepair encoding (BPE) vocabulary with 40,000 merges [53] and residual, embedding, and attention dropouts with a rate of 0.1 for regularization. We also employed a modified version of L2 regularization proposed in [37], with w = 0.01 on all non bias or gain weights. For the activation function, we used the Gaussian Error Linear Unit (GELU) [18]. We used learned position embeddings instead of the sinusoidal version proposed in the original work. We use the ftfy library2 to clean the raw text in BooksCorpus, standardize some punctuation and whitespace, and use the spaCy tokenizer. See the paper for further details and links to citations. ## Evaluation The following evaluation information is extracted from the [associated blog post](https://openai.com/blog/language-unsupervised/). See the [associated paper](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf) for further details. #### Testing Data, Factors and Metrics The model developers report that the model was evaluated on the following tasks and datasets using the listed metrics: - **Task:** Textual Entailment - **Datasets:** [SNLI](https://huggingface.co/datasets/snli), [MNLI Matched](https://huggingface.co/datasets/glue), [MNLI Mismatched](https://huggingface.co/datasets/glue), [SciTail](https://huggingface.co/datasets/scitail), [QNLI](https://huggingface.co/datasets/glue), [RTE](https://huggingface.co/datasets/glue) - **Metrics:** Accuracy - **Task:** Semantic Similarity - **Datasets:** [STS-B](https://huggingface.co/datasets/glue), [QQP](https://huggingface.co/datasets/glue), [MRPC](https://huggingface.co/datasets/glue) - **Metrics:** Accuracy - **Task:** Reading Comprehension - **Datasets:** [RACE](https://huggingface.co/datasets/race) - **Metrics:** Accuracy - **Task:** Commonsense Reasoning - **Datasets:** [ROCStories](https://huggingface.co/datasets/story_cloze), [COPA](https://huggingface.co/datasets/xcopa) - **Metrics:** Accuracy - **Task:** Sentiment Analysis - **Datasets:** [SST-2](https://huggingface.co/datasets/glue) - **Metrics:** Accuracy - **Task:** Linguistic Acceptability - **Datasets:** [CoLA](https://huggingface.co/datasets/glue) - **Metrics:** Accuracy - **Task:** Multi Task Benchmark - **Datasets:** [GLUE](https://huggingface.co/datasets/glue) - **Metrics:** Accuracy #### Results The model achieves the following results without any fine-tuning (zero-shot): | Task | TE | TE | TE |TE | TE | TE | SS | SS | SS | RC | CR | CR | SA | LA | MTB | |:--------:|:--:|:----------:|:-------------:|:-----:|:----:|:---:|:---:|:---:|:--:|:----:|:--------:|:----:|:----:|:----:|:----:| | Dataset |SNLI|MNLI Matched|MNLI Mismatched|SciTail| QNLI | RTE |STS-B| QQP |MPRC|RACE |ROCStories|COPA | SST-2| CoLA | GLUE | | |89.9| 82.1 | 81.4 |88.3 | 88.1 | 56.0|82.0 | 70.3|82.3|59.0 | 86.5 | 78.6 | 91.3 | 45.4 | 72.8 | ## Environmental Impact The model developers [report that](https://openai.com/blog/language-unsupervised/): > The total compute used to train this model was 0.96 petaflop days (pfs-days). > 8 P600 GPU's * 30 days * 12 TFLOPS/GPU * 0.33 utilization = .96 pfs-days Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). - **Hardware Type:** 8 P600 GPUs - **Hours used:** 720 hours (30 days) - **Cloud Provider:** Unknown - **Compute Region:** Unknown - **Carbon Emitted:** Unknown ## Technical Specifications See the [associated paper](https://cdn.openai.com/research-covers/language-unsupervised/language_understanding_paper.pdf) for details on the modeling architecture, objective, compute infrastructure, and training details. ## Citation Information ```bibtex @article{radford2018improving, title={Improving language understanding by generative pre-training}, author={Radford, Alec and Narasimhan, Karthik and Salimans, Tim and Sutskever, Ilya and others}, year={2018}, publisher={OpenAI} } ``` APA: *Radford, A., Narasimhan, K., Salimans, T., & Sutskever, I. (2018). Improving language understanding by generative pre-training.* ## Model Card Authors This model card was written by the Hugging Face team.

timm/sebotnet33ts_256.a1h_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for sebotnet33ts_256.a1h_in1k A BotNet image classification model (with Squeeze-and-Excitation channel attention, based on ResNet architecture). Trained on ImageNet-1k in `timm` by Ross Wightman. NOTE: this model did not adhere to any specific paper configuration, it was tuned for reasonable training times and reduced frequency of self-attention blocks. Recipe details: * Based on [ResNet Strikes Back](https://arxiv.org/abs/2110.00476) `A1` recipe * LAMB optimizer * Stronger dropout, stochastic depth, and RandAugment than paper `A1` recipe * Cosine LR schedule with warmup This model architecture is implemented using `timm`'s flexible [BYOBNet (Bring-Your-Own-Blocks Network)](https://github.com/huggingface/pytorch-image-models/blob/main/timm/models/byobnet.py). BYOB (with BYOANet attention specific blocks) allows configuration of: * block / stage layout * block-type interleaving * stem layout * output stride (dilation) * activation and norm layers * channel and spatial / self-attention layers ...and also includes `timm` features common to many other architectures, including: * stochastic depth * gradient checkpointing * layer-wise LR decay * per-stage feature extraction ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 13.7 - GMACs: 3.9 - Activations (M): 17.5 - Image size: 256 x 256 - **Papers:** - Bottleneck Transformers for Visual Recognition: https://arxiv.org/abs/2101.11605 - ResNet strikes back: An improved training procedure in timm: https://arxiv.org/abs/2110.00476 - **Dataset:** ImageNet-1k ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('sebotnet33ts_256.a1h_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'sebotnet33ts_256.a1h_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 32, 128, 128]) # torch.Size([1, 256, 64, 64]) # torch.Size([1, 512, 32, 32]) # torch.Size([1, 1024, 16, 16]) # torch.Size([1, 1280, 8, 8]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'sebotnet33ts_256.a1h_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 1280, 8, 8) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ``` ```bibtex @article{Srinivas2021BottleneckTF, title={Bottleneck Transformers for Visual Recognition}, author={A. Srinivas and Tsung-Yi Lin and Niki Parmar and Jonathon Shlens and P. Abbeel and Ashish Vaswani}, journal={2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, year={2021}, pages={16514-16524} } ``` ```bibtex @inproceedings{wightman2021resnet, title={ResNet strikes back: An improved training procedure in timm}, author={Wightman, Ross and Touvron, Hugo and Jegou, Herve}, booktitle={NeurIPS 2021 Workshop on ImageNet: Past, Present, and Future} } ```

deepset/sentence_bert

deepset

transformers

--- license: apache-2.0 --- This is an upload of the bert-base-nli-stsb-mean-tokens pretrained model from the Sentence Transformers Repo (https://github.com/UKPLab/sentence-transformers)

klue/roberta-base

klue

transformers

fill-mask

--- language: ko tags: - korean - klue mask_token: "[MASK]" widget: - text: 대한민국의 수도는 [MASK] 입니다. --- # KLUE RoBERTa base Pretrained RoBERTa Model on Korean Language. See [Github](https://github.com/KLUE-benchmark/KLUE) and [Paper](https://arxiv.org/abs/2105.09680) for more details. ## How to use _NOTE:_ Use `BertTokenizer` instead of RobertaTokenizer. (`AutoTokenizer` will load `BertTokenizer`) ```python from transformers import AutoModel, AutoTokenizer model = AutoModel.from_pretrained("klue/roberta-base") tokenizer = AutoTokenizer.from_pretrained("klue/roberta-base") ``` ## BibTeX entry and citation info ```bibtex @misc{park2021klue, title={KLUE: Korean Language Understanding Evaluation}, author={Sungjoon Park and Jihyung Moon and Sungdong Kim and Won Ik Cho and Jiyoon Han and Jangwon Park and Chisung Song and Junseong Kim and Yongsook Song and Taehwan Oh and Joohong Lee and Juhyun Oh and Sungwon Lyu and Younghoon Jeong and Inkwon Lee and Sangwoo Seo and Dongjun Lee and Hyunwoo Kim and Myeonghwa Lee and Seongbo Jang and Seungwon Do and Sunkyoung Kim and Kyungtae Lim and Jongwon Lee and Kyumin Park and Jamin Shin and Seonghyun Kim and Lucy Park and Alice Oh and Jungwoo Ha and Kyunghyun Cho}, year={2021}, eprint={2105.09680}, archivePrefix={arXiv}, primaryClass={cs.CL} } ```

EleutherAI/pythia-410m-deduped

EleutherAI

transformers

text-generation

--- language: - en tags: - pytorch - causal-lm - pythia license: apache-2.0 datasets: - EleutherAI/the_pile_deduplicated --- The *Pythia Scaling Suite* is a collection of models developed to facilitate interpretability research [(see paper)](https://arxiv.org/pdf/2304.01373.pdf). It contains two sets of eight models of sizes 70M, 160M, 410M, 1B, 1.4B, 2.8B, 6.9B, and 12B. For each size, there are two models: one trained on the Pile, and one trained on the Pile after the dataset has been globally deduplicated. All 8 model sizes are trained on the exact same data, in the exact same order. We also provide 154 intermediate checkpoints per model, hosted on Hugging Face as branches. The Pythia model suite was designed to promote scientific research on large language models, especially interpretability research. Despite not centering downstream performance as a design goal, we find the models <a href="#evaluations">match or exceed</a> the performance of similar and same-sized models, such as those in the OPT and GPT-Neo suites. <details> <summary style="font-weight:600">Details on previous early release and naming convention.</summary> Previously, we released an early version of the Pythia suite to the public. However, we decided to retrain the model suite to address a few hyperparameter discrepancies. This model card <a href="#changelog">lists the changes</a>; see appendix B in the Pythia paper for further discussion. We found no difference in benchmark performance between the two Pythia versions. The old models are [still available](https://huggingface.co/models?other=pythia_v0), but we suggest the retrained suite if you are just starting to use Pythia.<br> **This is the current release.** Please note that all models in the *Pythia* suite were renamed in January 2023. For clarity, a <a href="#naming-convention-and-parameter-count">table comparing the old and new names</a> is provided in this model card, together with exact parameter counts. </details> <br> # Pythia-410M-deduped ## Model Details - Developed by: [EleutherAI](http://eleuther.ai) - Model type: Transformer-based Language Model - Language: English - Learn more: [Pythia's GitHub repository](https://github.com/EleutherAI/pythia) for training procedure, config files, and details on how to use. [See paper](https://arxiv.org/pdf/2304.01373.pdf) for more evals and implementation details. - Library: [GPT-NeoX](https://github.com/EleutherAI/gpt-neox) - License: Apache 2.0 - Contact: to ask questions about this model, join the [EleutherAI Discord](https://discord.gg/zBGx3azzUn), and post them in `#release-discussion`. Please read the existing *Pythia* documentation before asking about it in the EleutherAI Discord. For general correspondence: [contact@eleuther. ai](mailto:contact@eleuther.ai). <figure> | Pythia model | Non-Embedding Params | Layers | Model Dim | Heads | Batch Size | Learning Rate | Equivalent Models | | -----------: | -------------------: | :----: | :-------: | :---: | :--------: | :-------------------: | :--------------------: | | 70M | 18,915,328 | 6 | 512 | 8 | 2M | 1.0 x 10^-3 | — | | 160M | 85,056,000 | 12 | 768 | 12 | 2M | 6.0 x 10^-4 | GPT-Neo 125M, OPT-125M | | 410M | 302,311,424 | 24 | 1024 | 16 | 2M | 3.0 x 10^-4 | OPT-350M | | 1.0B | 805,736,448 | 16 | 2048 | 8 | 2M | 3.0 x 10^-4 | — | | 1.4B | 1,208,602,624 | 24 | 2048 | 16 | 2M | 2.0 x 10^-4 | GPT-Neo 1.3B, OPT-1.3B | | 2.8B | 2,517,652,480 | 32 | 2560 | 32 | 2M | 1.6 x 10^-4 | GPT-Neo 2.7B, OPT-2.7B | | 6.9B | 6,444,163,072 | 32 | 4096 | 32 | 2M | 1.2 x 10^-4 | OPT-6.7B | | 12B | 11,327,027,200 | 36 | 5120 | 40 | 2M | 1.2 x 10^-4 | — | <figcaption>Engineering details for the <i>Pythia Suite</i>. Deduped and non-deduped models of a given size have the same hyperparameters. “Equivalent” models have <b>exactly</b> the same architecture, and the same number of non-embedding parameters.</figcaption> </figure> ## Uses and Limitations ### Intended Use The primary intended use of Pythia is research on the behavior, functionality, and limitations of large language models. This suite is intended to provide a controlled setting for performing scientific experiments. We also provide 154 checkpoints per model: initial `step0`, 10 log-spaced checkpoints `step{1,2,4...512}`, and 143 evenly-spaced checkpoints from `step1000` to `step143000`. These checkpoints are hosted on Hugging Face as branches. Note that branch `143000` corresponds exactly to the model checkpoint on the `main` branch of each model. You may also further fine-tune and adapt Pythia-410M-deduped for deployment, as long as your use is in accordance with the Apache 2.0 license. Pythia models work with the Hugging Face [Transformers Library](https://huggingface.co/docs/transformers/index). If you decide to use pre-trained Pythia-410M-deduped as a basis for your fine-tuned model, please conduct your own risk and bias assessment. ### Out-of-scope use The Pythia Suite is **not** intended for deployment. It is not a in itself a product and cannot be used for human-facing interactions. For example, the model may generate harmful or offensive text. Please evaluate the risks associated with your particular use case. Pythia models are English-language only, and are not suitable for translation or generating text in other languages. Pythia-410M-deduped has not been fine-tuned for downstream contexts in which language models are commonly deployed, such as writing genre prose, or commercial chatbots. This means XNPythia-410M-dedupedAME will **not** respond to a given prompt the way a product like ChatGPT does. This is because, unlike this model, ChatGPT was fine-tuned using methods such as Reinforcement Learning from Human Feedback (RLHF) to better “follow” human instructions. ### Limitations and biases The core functionality of a large language model is to take a string of text and predict the next token. The token used by the model need not produce the most “accurate” text. Never rely on Pythia-410M-deduped to produce factually accurate output. This model was trained on [the Pile](https://pile.eleuther.ai/), a dataset known to contain profanity and texts that are lewd or otherwise offensive. See [Section 6 of the Pile paper](https://arxiv.org/abs/2101.00027) for a discussion of documented biases with regards to gender, religion, and race. Pythia-410M-deduped may produce socially unacceptable or undesirable text, *even if* the prompt itself does not include anything explicitly offensive. If you plan on using text generated through, for example, the Hosted Inference API, we recommend having a human curate the outputs of this language model before presenting it to other people. Please inform your audience that the text was generated by Pythia-410M-deduped. ### Quickstart Pythia models can be loaded and used via the following code, demonstrated here for the third `pythia-70m-deduped` checkpoint: ```python from transformers import GPTNeoXForCausalLM, AutoTokenizer model = GPTNeoXForCausalLM.from_pretrained( "EleutherAI/pythia-70m-deduped", revision="step3000", cache_dir="./pythia-70m-deduped/step3000", ) tokenizer = AutoTokenizer.from_pretrained( "EleutherAI/pythia-70m-deduped", revision="step3000", cache_dir="./pythia-70m-deduped/step3000", ) inputs = tokenizer("Hello, I am", return_tensors="pt") tokens = model.generate(**inputs) tokenizer.decode(tokens[0]) ``` Revision/branch `step143000` corresponds exactly to the model checkpoint on the `main` branch of each model.<br> For more information on how to use all Pythia models, see [documentation on GitHub](https://github.com/EleutherAI/pythia). ## Training ### Training data Pythia-410M-deduped was trained on the Pile **after the dataset has been globally deduplicated**.<br> [The Pile](https://pile.eleuther.ai/) is a 825GiB general-purpose dataset in English. It was created by EleutherAI specifically for training large language models. It contains texts from 22 diverse sources, roughly broken down into five categories: academic writing (e.g. arXiv), internet (e.g. CommonCrawl), prose (e.g. Project Gutenberg), dialogue (e.g. YouTube subtitles), and miscellaneous (e.g. GitHub, Enron Emails). See [the Pile paper](https://arxiv.org/abs/2101.00027) for a breakdown of all data sources, methodology, and a discussion of ethical implications. Consult [the datasheet](https://arxiv.org/abs/2201.07311) for more detailed documentation about the Pile and its component datasets. The Pile can be downloaded from the [official website](https://pile.eleuther.ai/), or from a [community mirror](https://the-eye.eu/public/AI/pile/). ### Training procedure All models were trained on the exact same data, in the exact same order. Each model saw 299,892,736,000 tokens during training, and 143 checkpoints for each model are saved every 2,097,152,000 tokens, spaced evenly throughout training, from `step1000` to `step143000` (which is the same as `main`). In addition, we also provide frequent early checkpoints: `step0` and `step{1,2,4...512}`. This corresponds to training for just under 1 epoch on the Pile for non-deduplicated models, and about 1.5 epochs on the deduplicated Pile. All *Pythia* models trained for 143000 steps at a batch size of 2M (2,097,152 tokens).<br> See [GitHub](https://github.com/EleutherAI/pythia) for more details on training procedure, including [how to reproduce it](https://github.com/EleutherAI/pythia/blob/main/README.md#reproducing-training).<br> Pythia uses the same tokenizer as [GPT-NeoX- 20B](https://huggingface.co/EleutherAI/gpt-neox-20b). ## Evaluations All 16 *Pythia* models were evaluated using the [LM Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness). You can access the results by model and step at `results/json/*` in the [GitHub repository](https://github.com/EleutherAI/pythia/tree/main/results/json/).<br> Expand the sections below to see plots of evaluation results for all Pythia and Pythia-deduped models compared with OPT and BLOOM. <details> <summary>LAMBADA – OpenAI</summary> <img src="/EleutherAI/pythia-12b/resolve/main/eval_plots/lambada_openai_v1.png" style="width:auto"/> </details> <details> <summary>Physical Interaction: Question Answering (PIQA)</summary> <img src="/EleutherAI/pythia-12b/resolve/main/eval_plots/piqa_v1.png" style="width:auto"/> </details> <details> <summary>WinoGrande</summary> <img src="/EleutherAI/pythia-12b/resolve/main/eval_plots/winogrande_v1.png" style="width:auto"/> </details> <details> <summary>AI2 Reasoning Challenge—Easy Set</summary> <img src="/EleutherAI/pythia-12b/resolve/main/eval_plots/arc_easy_v1.png" style="width:auto"/> </details> <details> <summary>SciQ</summary> <img src="/EleutherAI/pythia-12b/resolve/main/eval_plots/sciq_v1.png" style="width:auto"/> </details> ## Changelog This section compares differences between previously released [Pythia v0](https://huggingface.co/models?other=pythia_v0) and the current models. See Appendix B of the Pythia paper for further discussion of these changes and the motivation behind them. We found that retraining Pythia had no impact on benchmark performance. - All model sizes are now trained with uniform batch size of 2M tokens. Previously, the models of size 160M, 410M, and 1.4B parameters were trained with batch sizes of 4M tokens. - We added checkpoints at initialization (step 0) and steps {1,2,4,8,16,32,64, 128,256,512} in addition to every 1000 training steps. - Flash Attention was used in the new retrained suite. - We remedied a minor inconsistency that existed in the original suite: all models of size 2.8B parameters or smaller had a learning rate (LR) schedule which decayed to a minimum LR of 10% the starting LR rate, but the 6.9B and 12B models all used an LR schedule which decayed to a minimum LR of 0. In the redone training runs, we rectified this inconsistency: all models now were trained with LR decaying to a minimum of 0.1× their maximum LR. ### Naming convention and parameter count *Pythia* models were renamed in January 2023. It is possible that the old naming convention still persists in some documentation by accident. The current naming convention (70M, 160M, etc.) is based on total parameter count. <figure style="width:32em"> | current Pythia suffix | old suffix | total params | non-embedding params | | --------------------: | ---------: | -------------: | -------------------: | | 70M | 19M | 70,426,624 | 18,915,328 | | 160M | 125M | 162,322,944 | 85,056,000 | | 410M | 350M | 405,334,016 | 302,311,424 | | 1B | 800M | 1,011,781,632 | 805,736,448 | | 1.4B | 1.3B | 1,414,647,808 | 1,208,602,624 | | 2.8B | 2.7B | 2,775,208,960 | 2,517,652,480 | | 6.9B | 6.7B | 6,857,302,016 | 6,444,163,072 | | 12B | 13B | 11,846,072,320 | 11,327,027,200 | </figure>

timm/mobilenetv3_small_075.lamb_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k --- # Model card for mobilenetv3_small_075.lamb_in1k A MobileNet-v3 image classification model. Trained on ImageNet-1k in `timm` using recipe template described below. Recipe details: * A LAMB optimizer recipe that is similar to [ResNet Strikes Back](https://arxiv.org/abs/2110.00476) `A2` but 50% longer with EMA weight averaging, no CutMix * RMSProp (TF 1.0 behaviour) optimizer, EMA weight averaging * Step (exponential decay w/ staircase) LR schedule with warmup ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 2.0 - GMACs: 0.0 - Activations (M): 1.3 - Image size: 224 x 224 - **Papers:** - Searching for MobileNetV3: https://arxiv.org/abs/1905.02244 - **Dataset:** ImageNet-1k - **Original:** https://github.com/huggingface/pytorch-image-models ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('mobilenetv3_small_075.lamb_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilenetv3_small_075.lamb_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 16, 112, 112]) # torch.Size([1, 16, 56, 56]) # torch.Size([1, 24, 28, 28]) # torch.Size([1, 40, 14, 14]) # torch.Size([1, 432, 7, 7]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilenetv3_small_075.lamb_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 432, 7, 7) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ``` ```bibtex @inproceedings{howard2019searching, title={Searching for mobilenetv3}, author={Howard, Andrew and Sandler, Mark and Chu, Grace and Chen, Liang-Chieh and Chen, Bo and Tan, Mingxing and Wang, Weijun and Zhu, Yukun and Pang, Ruoming and Vasudevan, Vijay and others}, booktitle={Proceedings of the IEEE/CVF international conference on computer vision}, pages={1314--1324}, year={2019} } ```

bhadresh-savani/bert-base-uncased-emotion

bhadresh-savani

transformers

text-classification

--- language: - en license: apache-2.0 tags: - text-classification - emotion - pytorch datasets: - emotion metrics: - Accuracy, F1 Score thumbnail: https://avatars3.githubusercontent.com/u/32437151?s=460&u=4ec59abc8d21d5feea3dab323d23a5860e6996a4&v=4 model-index: - name: bhadresh-savani/bert-base-uncased-emotion results: - task: type: text-classification name: Text Classification dataset: name: emotion type: emotion config: default split: test metrics: - type: accuracy value: 0.9265 name: Accuracy verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMWQzNzA2MTFkY2RkNDMxYTFhOGUzMTdiZTgwODA3ODdmZTVhNTVjOTAwMGM5NjU1OGY0MjMzZWU0OTU2MzY1YiIsInZlcnNpb24iOjF9.f6iWK0iyU8_g32W2oMfh1ChevMsl0StI402cB6DNzJCYj9xywTnFltBY36jAJFDRK41HXdMnPMl64Bynr-Q9CA - type: precision value: 0.8859601677706858 name: Precision Macro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNTc2ZjRmMzYzNTE0ZDQ1ZDdkYWViYWNhZDhkOTE2ZDhmMDFjZmZiZjRkZWVlMzQ3MWE4NDNlYzlmM2I4ZGM2OCIsInZlcnNpb24iOjF9.jR-gFrrBIAfiYV352RDhK3nzgqIgNCPd55OhIcCfVdVAWHQSZSJXhFyg8yChC7DwoVmUQy1Ya-d8Hflp7Wi-AQ - type: precision value: 0.9265 name: Precision Micro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMDAyMWZjZTM5NWNjNTcyMWQzMWQyNDcyN2RlZTQyZTM4ZDQ4Y2FlNzM2OTZkMzM3YzI4YTAwNzg4MGNjZmZjZCIsInZlcnNpb24iOjF9.cmkuDmhhETKIKAL81K28oiO889sZ0hvEpZ6Ep7dW_KB9VOTFs15BzFY9vwcpdXQDugWBbB2g7r3FUgRLwIEpAg - type: precision value: 0.9265082039990273 name: Precision Weighted verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMTA2NzY2NTJmZTExZWM3OGIzYzg3ZDM3Y2I5MTU3Mjg3Y2NmZGEyMjFmNjExZWM3ZDFjNzdhOTZkNTYwYWQxYyIsInZlcnNpb24iOjF9.DJgeA6ZovHoxgCqhzilIzafet8uN3-Xbx1ZYcEEc4jXzFbRtErE__QHGaaSaUQEzPp4BAztp1ageOaBoEmXSDg - type: recall value: 0.879224648382427 name: Recall Macro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZGU3MmQ1Yjg5OGJlYTE1NWJmNGVjY2ExMDZiZjVjYmVkOGYxYWFkOTVlMDVjOWVhZGFjOGFkYzcwMGIyMTAyZCIsInZlcnNpb24iOjF9.jwgaNEBSQENlx3vojBi1WKJOQ7pSuP4Iyw4kKPsq9IUaW-Ah8KdgPV9Nm2DY1cwEtMayvVeIVmQ3Wo8PORDRAg - type: recall value: 0.9265 name: Recall Micro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNDE3OWQ0ZGZjNzAxY2I0NGMxNDU0OWE1OGM2N2Q3OTUwYWI0NmZjMDQ3MDc0NDA4YTc2NDViM2Y0ZTMyMjYyZCIsInZlcnNpb24iOjF9.Ihc61PSO3K63t5hUSAve4Gt1tC8R_ZruZo492dTD9CsKOF10LkvrCskJJaOATjFJgqb3FFiJ8-nDL9Pa3HF-Dg - type: recall value: 0.9265 name: Recall Weighted verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNzJkYTg5YjA0YTBlNDY3ZjFjZWIzOWVhYjI4Y2YxM2FhMmUwMDZlZTE0NTIzNjMxMjE3NzgwNGFjYTkzOWM1YyIsInZlcnNpb24iOjF9.LlBX4xTjKuTX0NPK0jYzYDXRVnUEoUKVwIHfw5xUzaFgtF4wuqaYV7F0VKoOd3JZxzxNgf7JzeLof0qTquE9Cw - type: f1 value: 0.8821398657055098 name: F1 Macro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNTE4OThiMmE0NDEzZjBkY2RmZWNjMGI3YWNmNTFjNTY5NjIwNjFkZjk1ZjIxMjI4M2ZiZGJhYzJmNzVhZTU1NSIsInZlcnNpb24iOjF9.gzYyUbO4ycvP1RXnrKKZH3E8ym0DjwwUFf4Vk9j0wrg2sWIchjmuloZz0SLryGqwHiAV8iKcSBWWy61Q480XAw - type: f1 value: 0.9265 name: F1 Micro verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZGM2Y2E0NjMyNmJhMTE4NjYyMjI2MTJlZjUzNmRmY2U3Yjk3ZGUyYzU2OWYzMWM2ZjY4ZTg0OTliOTY3YmI2MSIsInZlcnNpb24iOjF9.hEz_yExs6LV0RBpFBoUbnAQZHitxN57HodCJpDx0yyW6dQwWaza0JxdO-kBf8JVBK8JyISkNgOYskBY5LD4ZDQ - type: f1 value: 0.9262425173620311 name: F1 Weighted verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZmMyY2NhNTRhOGMwM2M5OTQxNDQ0NjRkZDdiMDExMWFkMmI4MmYwZGQ1OGRiYmRjMmE2YTc0MGZmMWMwN2Q4MSIsInZlcnNpb24iOjF9.ljbb2L4R08NCGjcfuX1878HRilJ_p9qcDJpWhsu-5EqWCco80e9krb7VvIJV0zBfmi7Z3C2qGGRsfsAIhtQ5Dw - type: loss value: 0.17315374314785004 name: loss verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZmQwN2I2Nzg4OWU1ODE5NTBhMTZiMjljMjJhN2JiYmY0MTkzMTA1NmVhMGU0Y2Y0NjgyOTU3ZjgyYTc3ODE5NCIsInZlcnNpb24iOjF9.EEp3Gxm58ab-9335UGQEk-3dFQcMRgJgViI7fpz7mfY2r5Pg-AOel5w4SMzmBM-hiUFwStgxe5he_kG2yPGFCw --- # bert-base-uncased-emotion ## Model description: [Bert](https://arxiv.org/abs/1810.04805) is a Transformer Bidirectional Encoder based Architecture trained on MLM(Mask Language Modeling) objective [bert-base-uncased](https://huggingface.co/bert-base-uncased) finetuned on the emotion dataset using HuggingFace Trainer with below training parameters ``` learning rate 2e-5, batch size 64, num_train_epochs=8, ``` ## Model Performance Comparision on Emotion Dataset from Twitter: | Model | Accuracy | F1 Score | Test Sample per Second | | --- | --- | --- | --- | | [Distilbert-base-uncased-emotion](https://huggingface.co/bhadresh-savani/distilbert-base-uncased-emotion) | 93.8 | 93.79 | 398.69 | | [Bert-base-uncased-emotion](https://huggingface.co/bhadresh-savani/bert-base-uncased-emotion) | 94.05 | 94.06 | 190.152 | | [Roberta-base-emotion](https://huggingface.co/bhadresh-savani/roberta-base-emotion) | 93.95 | 93.97| 195.639 | | [Albert-base-v2-emotion](https://huggingface.co/bhadresh-savani/albert-base-v2-emotion) | 93.6 | 93.65 | 182.794 | ## How to Use the model: ```python from transformers import pipeline classifier = pipeline("text-classification",model='bhadresh-savani/bert-base-uncased-emotion', return_all_scores=True) prediction = classifier("I love using transformers. The best part is wide range of support and its easy to use", ) print(prediction) """ output: [[ {'label': 'sadness', 'score': 0.0005138228880241513}, {'label': 'joy', 'score': 0.9972520470619202}, {'label': 'love', 'score': 0.0007443308713845909}, {'label': 'anger', 'score': 0.0007404946954920888}, {'label': 'fear', 'score': 0.00032938539516180754}, {'label': 'surprise', 'score': 0.0004197491507511586} ]] """ ``` ## Dataset: [Twitter-Sentiment-Analysis](https://huggingface.co/nlp/viewer/?dataset=emotion). ## Training procedure [Colab Notebook](https://github.com/bhadreshpsavani/ExploringSentimentalAnalysis/blob/main/SentimentalAnalysisWithDistilbert.ipynb) follow the above notebook by changing the model name from distilbert to bert ## Eval results ```json { 'test_accuracy': 0.9405, 'test_f1': 0.9405920712282673, 'test_loss': 0.15769127011299133, 'test_runtime': 10.5179, 'test_samples_per_second': 190.152, 'test_steps_per_second': 3.042 } ``` ## Reference: * [Natural Language Processing with Transformer By Lewis Tunstall, Leandro von Werra, Thomas Wolf](https://learning.oreilly.com/library/view/natural-language-processing/9781098103231/)

bhadresh-savani/albert-base-v2-emotion

bhadresh-savani

transformers

text-classification

--- language: - en thumbnail: https://avatars3.githubusercontent.com/u/32437151?s=460&u=4ec59abc8d21d5feea3dab323d23a5860e6996a4&v=4 tags: - text-classification - emotion - pytorch license: apache-2.0 datasets: - emotion metrics: - Accuracy, F1 Score --- # Albert-base-v2-emotion ## Model description: [Albert](https://arxiv.org/pdf/1909.11942v6.pdf) is A Lite BERT architecture that has significantly fewer parameters than a traditional BERT architecture. [Albert-base-v2](https://huggingface.co/albert-base-v2) finetuned on the emotion dataset using HuggingFace Trainer with below Hyperparameters ``` learning rate 2e-5, batch size 64, num_train_epochs=8, ``` ## Model Performance Comparision on Emotion Dataset from Twitter: | Model | Accuracy | F1 Score | Test Sample per Second | | --- | --- | --- | --- | | [Distilbert-base-uncased-emotion](https://huggingface.co/bhadresh-savani/distilbert-base-uncased-emotion) | 93.8 | 93.79 | 398.69 | | [Bert-base-uncased-emotion](https://huggingface.co/bhadresh-savani/bert-base-uncased-emotion) | 94.05 | 94.06 | 190.152 | | [Roberta-base-emotion](https://huggingface.co/bhadresh-savani/roberta-base-emotion) | 93.95 | 93.97| 195.639 | | [Albert-base-v2-emotion](https://huggingface.co/bhadresh-savani/albert-base-v2-emotion) | 93.6 | 93.65 | 182.794 | ## How to Use the model: ```python from transformers import pipeline classifier = pipeline("text-classification",model='bhadresh-savani/albert-base-v2-emotion', return_all_scores=True) prediction = classifier("I love using transformers. The best part is wide range of support and its easy to use", ) print(prediction) """ Output: [[ {'label': 'sadness', 'score': 0.010403595864772797}, {'label': 'joy', 'score': 0.8902180790901184}, {'label': 'love', 'score': 0.042532723397016525}, {'label': 'anger', 'score': 0.041297927498817444}, {'label': 'fear', 'score': 0.011772023513913155}, {'label': 'surprise', 'score': 0.0037756056990474463} ]] """ ``` ## Dataset: [Twitter-Sentiment-Analysis](https://huggingface.co/nlp/viewer/?dataset=emotion). ## Training procedure [Colab Notebook](https://github.com/bhadreshpsavani/ExploringSentimentalAnalysis/blob/main/SentimentalAnalysisWithDistilbert.ipynb) ## Eval results ```json { 'test_accuracy': 0.936, 'test_f1': 0.9365658988006296, 'test_loss': 0.15278364717960358, 'test_runtime': 10.9413, 'test_samples_per_second': 182.794, 'test_steps_per_second': 2.925 } ``` ## Reference: * [Natural Language Processing with Transformer By Lewis Tunstall, Leandro von Werra, Thomas Wolf](https://learning.oreilly.com/library/view/natural-language-processing/9781098103231/)

sentence-transformers/msmarco-bert-base-dot-v5

sentence-transformers

sentence-transformers

sentence-similarity

--- language: - en pipeline_tag: sentence-similarity tags: - sentence-transformers - feature-extraction - sentence-similarity - transformers --- # msmarco-bert-base-dot-v5 This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and was designed for **semantic search**. It has been trained on 500K (query, answer) pairs from the [MS MARCO dataset](https://github.com/microsoft/MSMARCO-Passage-Ranking/). For an introduction to semantic search, have a look at: [SBERT.net - Semantic Search](https://www.sbert.net/examples/applications/semantic-search/README.html) ## Usage (Sentence-Transformers) Using this model becomes easy when you have [sentence-transformers](https://www.SBERT.net) installed: ``` pip install -U sentence-transformers ``` Then you can use the model like this: ```python from sentence_transformers import SentenceTransformer, util query = "How many people live in London?" docs = ["Around 9 Million people live in London", "London is known for its financial district"] #Load the model model = SentenceTransformer('sentence-transformers/msmarco-bert-base-dot-v5') #Encode query and documents query_emb = model.encode(query) doc_emb = model.encode(docs) #Compute dot score between query and all document embeddings scores = util.dot_score(query_emb, doc_emb)[0].cpu().tolist() #Combine docs & scores doc_score_pairs = list(zip(docs, scores)) #Sort by decreasing score doc_score_pairs = sorted(doc_score_pairs, key=lambda x: x[1], reverse=True) #Output passages & scores print("Query:", query) for doc, score in doc_score_pairs: print(score, doc) ``` ## Usage (HuggingFace Transformers) Without [sentence-transformers](https://www.SBERT.net), you can use the model like this: First, you pass your input through the transformer model, then you have to apply the correct pooling-operation on-top of the contextualized word embeddings. ```python from transformers import AutoTokenizer, AutoModel import torch #Mean Pooling - Take attention mask into account for correct averaging def mean_pooling(model_output, attention_mask): token_embeddings = model_output.last_hidden_state input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float() return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9) #Encode text def encode(texts): # Tokenize sentences encoded_input = tokenizer(texts, padding=True, truncation=True, return_tensors='pt') # Compute token embeddings with torch.no_grad(): model_output = model(**encoded_input, return_dict=True) # Perform pooling embeddings = mean_pooling(model_output, encoded_input['attention_mask']) return embeddings # Sentences we want sentence embeddings for query = "How many people live in London?" docs = ["Around 9 Million people live in London", "London is known for its financial district"] # Load model from HuggingFace Hub tokenizer = AutoTokenizer.from_pretrained("sentence-transformers/msmarco-bert-base-dot-v5") model = AutoModel.from_pretrained("sentence-transformers/msmarco-bert-base-dot-v5") #Encode query and docs query_emb = encode(query) doc_emb = encode(docs) #Compute dot score between query and all document embeddings scores = torch.mm(query_emb, doc_emb.transpose(0, 1))[0].cpu().tolist() #Combine docs & scores doc_score_pairs = list(zip(docs, scores)) #Sort by decreasing score doc_score_pairs = sorted(doc_score_pairs, key=lambda x: x[1], reverse=True) #Output passages & scores print("Query:", query) for doc, score in doc_score_pairs: print(score, doc) ``` ## Technical Details In the following some technical details how this model must be used: | Setting | Value | | --- | :---: | | Dimensions | 768 | | Max Sequence Length | 512 | | Produces normalized embeddings | No | | Pooling-Method | Mean pooling | | Suitable score functions | dot-product (e.g. `util.dot_score`) | ## Evaluation Results <!--- Describe how your model was evaluated --> For an automated evaluation of this model, see the *Sentence Embeddings Benchmark*: [https://seb.sbert.net](https://seb.sbert.net?model_name=msmarco-bert-base-base-dot-v5) ## Training See `train_script.py` in this repository for the used training script. The model was trained with the parameters: **DataLoader**: `torch.utils.data.dataloader.DataLoader` of length 7858 with parameters: ``` {'batch_size': 64, 'sampler': 'torch.utils.data.sampler.RandomSampler', 'batch_sampler': 'torch.utils.data.sampler.BatchSampler'} ``` **Loss**: `sentence_transformers.losses.MarginMSELoss.MarginMSELoss` Parameters of the fit()-Method: ``` { "callback": null, "epochs": 30, "evaluation_steps": 0, "evaluator": "NoneType", "max_grad_norm": 1, "optimizer_class": "<class 'transformers.optimization.AdamW'>", "optimizer_params": { "lr": 1e-05 }, "scheduler": "WarmupLinear", "steps_per_epoch": null, "warmup_steps": 10000, "weight_decay": 0.01 } ``` ## Full Model Architecture ``` SentenceTransformer( (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: bert-base-uncased (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False}) ) ``` ## Citing & Authors This model was trained by [sentence-transformers](https://www.sbert.net/). If you find this model helpful, feel free to cite our publication [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://arxiv.org/abs/1908.10084): ```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 = "http://arxiv.org/abs/1908.10084", } ```

knkarthick/MEETING_SUMMARY

knkarthick

transformers

summarization

--- language: en license: apache-2.0 tags: - bart - seq2seq - summarization datasets: - cnndaily/newyorkdaily/xsum/samsum/dialogsum/AMI metrics: - rouge widget: - text: 'Hi, I''m David and I''m supposed to be an industrial designer. Um, I just got the project announcement about what the project is. Designing a remote control. That''s about it, didn''t get anything else. Did you get the same thing? Cool. There''s too much gear. Okay. Can''t draw. Um. Yeah. Um, well anyway, I don''t know, it''s just the first animal I can think off the top of my head. Um. Yes. Big reason is ''cause I''m allergic to most animals. Allergic to animal fur, so um fish was a natural choice. Um, yeah, and I kind of like whales. They come in and go eat everything in sight. And they''re quite harmless and mild and interesting. Tail''s a bit big, I think. It''s an after dinner dog then. Hmm. It does make sense from maybe the design point of view ''cause you have more complicated characters like European languages, then you need more buttons. So, possibly. Hmm. Yeah. And you keep losing them. Finding them is really a pain, you know. I mean it''s usually quite small, or when you want it right, it slipped behind the couch or it''s kicked under the table. You know. Yep. Mm-hmm. I think one factor would be production cost. Because there''s a cap there, so um depends on how much you can cram into that price. Um. I think that that''s the main factor. Cool. Okay. Right. Um well this is the kick-off meeting for our our project. Um and um this is just what we''re gonna be doing over the next twenty five minutes. Um so first of all, just to kind of make sure that we all know each other, I''m Laura and I''m the project manager. Do you want to introduce yourself again? Okay. Great. Okay. Um so we''re designing a new remote control and um Oh I have to record who''s here actually. So that''s David, Andrew and Craig, isn''t it? And you all arrived on time. Um yeah so des uh design a new remote control. Um, as you can see it''s supposed to be original, trendy and user friendly. Um so that''s kind of our our brief, as it were. Um and so there are three different stages to the design. Um I''m not really sure what what you guys have already received um in your emails. What did you get? Mm-hmm. Is that what everybody got? Okay. Um. So we''re gonna have like individual work and then a meeting about it. And repeat that process three times. Um and at this point we get try out the whiteboard over there. Um. So uh you get to draw your favourite animal and sum up your favourite characteristics of it. So who would like to go first? Very good. Mm-hmm. Yeah. Yeah. Right. Lovely. Right. You can take as long over this as you like, because we haven''t got an awful lot to discuss. Ok oh we do we do. Don''t feel like you''re in a rush, anyway. Ach why not We might have to get you up again then. I don''t know what mine is. I''m gonna have to think on the spot now. Is that a whale? Ah. Okay. God, I still don''t know what I''m gonna write about. Um. I was gonna choose a dog as well. But I''ll just draw a different kind of dog. M my favourite animal is my own dog at home. Um That doesn''t really look like him, actually. He looks more like a pig, actually. Ah well. Do you? Oh that''s very good of you. Uh. Um he''s a mixture of uh various things. Um and what do I like about him, um That''s just to suggest that his tail wags. Um he''s very friendly and cheery and always pleased to see you, and very kind of affectionate and um uh and he''s quite quite wee as well so you know he can doesn''t take up too much space. Um and uh And he does a funny thing where he chases his tail as well, which is quite amusing, so It is. I think it is. He only does it after he''s had his dinner and um he''ll just all of a sudden just get up and start chasing his tail ''round the living room. Yeah, so uh Yeah, maybe. Maybe. Right, um where did you find this? Just down here? Yeah. Okay. Um what are we doing next? Uh um. Okay, uh we now need to discuss the project finance. Um so according to the brief um we''re gonna be selling this remote control for twenty five Euro, um and we''re aiming to make fifty million Euro. Um so we''re gonna be selling this on an international scale. And uh we don''t want it to cost any more than uh twelve fifty Euros, so fifty percent of the selling price. Sure. All together. Um I dunno. I imagine That''s a good question. I imagine it probably is our sale actually because it''s probably up to the the um the retailer to uh sell it for whatever price they want. Um. But I I don''t know, I mean do you think the fact that it''s going to be sold internationally will have a bearing on how we design it at all? Think it will? Um. Hmm. Oh yeah, regions and stuff, yeah. Yeah. Okay. Yeah. Well for a remote control, do you think that will be I suppose it''s depends on how complicated our remote control is. Yeah, yeah. Okay. What, just like in terms of like the wealth of the country? Like how much money people have to spend on things like? Aye, I see what you mean, yeah. Marketing. Good marketing thoughts. Oh gosh, I should be writing all this down. Um. Mm. Yeah. Yeah, yeah. Like how much does, you know, a remote control cost. Well twenty five Euro, I mean that''s um that''s about like eighteen pounds or something, isn''t it? Or no, is it as much as that? Sixteen seventeen eighteen pounds. Um, I dunno, I''ve never bought a remote control, so I don''t know how how good a remote control that would get you. Um. But yeah, I suppose it has to look kind of cool and gimmicky. Um right, okay. Let me just scoot on ahead here. Okay. Um well d Does anybody have anything to add to uh to the finance issue at all? Thin No, actually. That would be useful, though, wouldn''t it, if you knew like what your money would get you now. Mm-hmm. Yeah, yeah. Oh. Five minutes to end of meeting. Oh, okay. We''re a bit behind. Yeah. Right, so do you think that should be like a main design aim of our remote control d you know, do your your satellite and your regular telly and your V_C_R_ and everything? Mm-hmm. Yeah. Or even like, you know, notes about um what you wanna watch. Like you might put in there oh I want to watch such and such and look a Oh that''s a good idea. So extra functionalities. Mm-hmm. Hmm. Um okay, uh I''d wel we''re gonna have to wrap up pretty quickly in the next couple of minutes. Um I''ll just check we''ve nothing else. Okay. Um so anything else anybody wants to add about what they don''t like about remote controls they''ve used, what they would really like to be part of this new one at all? You keep losing them. Okay. Yeah. W You get those ones where you can, if you like, whistle or make a really high pitched noise they beep. There I mean is that something we''d want to include, do you think? Dunno. Okay maybe. My goodness. Still feels quite primitive. Maybe like a touch screen or something? Okay. Uh-huh, okay. Well I guess that''s up to our industrial designer. It looks better. Yeah. Okay. Okay. Right, well um so just to wrap up, the next meeting''s gonna be in thirty minutes. So that''s about um about ten to twelve by my watch. Um so inbetween now and then, um as the industrial designer, you''re gonna be working on you know the actual working design of it so y you know what you''re doing there. Um for user interface, technical functions, I guess that''s you know like what we''ve been talking about, what it''ll actually do. Um and uh marketing executive, you''ll be just thinking about what it actually what, you know, what requirements it has to has to fulfil and you''ll all get instructions emailed to you, I guess. Um. Yeah, so it''s th the functional design stage is next, I guess. And uh and that''s the end of the meeting. So I got that little message a lot sooner than I thought I would, so Mm-hmm. Uh-huh, yeah. Th Okay, well just very quickly ''cause this we''re supposed to finish now. Um I guess that''s up to us, I mean you probably want some kind of unique selling point of it, so um, you know Yeah. Mm-hmm. Yeah. Okay. Right, okay, we''ll that''s that''s the end of the meeting, then. Um. So, uh thank you all for coming. Um I''m Craig and I''m User Interface. Yeah. Well, my favourite animal would be a monkey. Then they''re small cute and furry, and uh when planet of the apes becomes real, I''m gonna be up there with them. Yeah. I know um My parents went out and bought um remote controls because um they got fed up of having four or five different remote controls for each things the house. So um for them it was just how many devices control. Uh. Mm-hmm. Great. And I''m Andrew and I''m uh our marketing expert. Mm-hmm. Mm-hmm. Yeah, that''s that''s it. Yeah. I will go. That''s fine. Alright. So This one here, right? Okay. Very nice. Alright. My favourite animal is like A beagle. Um charac favourite characteristics of it? Is that right? Uh, right, well basically um high priority for any animal for me is that they be willing to take a lot of physical affection from their family. And, yeah that they have lots of personality and uh be fit and in robust good health. So this is blue. Blue beagle. My family''s beagle. I coulda told you a whole lot more about beagles. Boy, let me tell you. Impressionist. Alright. Mm. Superb sketch, by the way. Yep. I see a dog in there. Yep. Now I see a rooster. What kind is it? Is he aware that th it''s his own cha tail he''s chasing? Hmm. Probably when he was little he got lots of attention for doing it and has forever been conditioned. ''Kay. Um, can we just go over that again? Uh, so bas at twel Alright, yeah. Okay. So cost like production cost is twelve fifty, but selling price is is that wholesale or retail? Like on the shelf. Our sale our sale anyway. Yeah, okay okay. Okay. Mm-hmm. Alright. Yes. Mm-hmm. Mm-hmm. Well right away I''m wondering if there''s um th th uh, like with D_V_D_ players, if there are zones. Um f frequencies or something um as well as uh characters, um different uh keypad styles and s symbols. Um. I don''t know. Yeah. Yeah. Yeah. And then a and then al the other thing international is on top of the price. I''m thinking the price might might appeal to a certain market in one region, whereas in another it''ll be different, so Just a chara just a characteristic of the Just Or just like, basic product podi positioning, the twenty five Euro remote control might be a big hit in London, might not be such a big hit in Greece, who knows, something like that, yeah. Yep. Right away I''m making some kind of assumptions about what what information we''re given here, thinking, ''kay trendy probably means something other than just basic, something other than just standard. Um so I''m wondering right away, is selling twenty five Euros, is that sort of the thi is this gonna to be like the premium product kinda thing or Uh-huh. Mm-hmm. Yep. Yeah, I''d say so, yeah. No. Yeah, yeah. Mm-hmm. Do we have any other background information on like how that compares to other other Yeah. Mm-hmm. Yeah, interesting thing about discussing um production of a remote control for me is that l as you point out, I just don''t think of remote controls as somethin something people consciously assess in their purchasing habits. It''s just like getting shoelaces with shoes or something. It just comes along. Do you know what I mean? Like so sort of like how do you I I mean one one way of looking at it would be, well the people producing television sets, maybe they have to buy remote controls. Or another way is maybe people who have T_V_ sets are really fed up with their remote control and they really want a better one or something. But Right. Right. Okay so Right, so in function one of the priorities might be to combine as many uses I think so. Yeah, yeah. Yeah. Well like um, maybe what we could use is a sort of like a example of a successful other piece technology is palm palm pilots. They''re gone from being just like little sort of scribble boards to cameras, M_P_ three players, telephones, everything, agenda. So, like, I wonder if we might add something new to the to the remote control market, such as the lighting in your house, or um Yeah, yeah. An Yeah. Like, p personally for me, at home I''ve I''ve combined the um the audio video of my television set and my D_V_D_ player and my C_D_ player. So they w all work actually function together but I have different remote controls for each of them. So it''s sort of ironic that that then they''re in there um you know, the sound and everything it''s just one system. But each one''s got its own little part. Mm. Mm. Mm. Mm-hmm. Mm-hmm. Yeah. Yeah. That''s just really good id Yep. Uh, sure. I remember when the first remote control my my family had was on a cable. Actually had a cable between it and the T_V_ and big like buttons that sort of like, like on a blender or something. And um, you know, when I think about what they are now, it''s better, but actually it''s still kind of, I dunno, like a massive junky thing on the table. Maybe we could think about how, could be more, you know, streamlined. S Something like that, yeah. Or whatever would be technologically reasonable. ''Cause it could b it could it could be that f it could be that functionally that doesn''t make it any better, but that just the appeal of of not having You know, these days there''s a r pe things in people''s homes are becoming more and more like chic, you know. Um, nicer materials and might be be worth exploring anyway. Okay. Um. Before we wrap up, just to make sure we''re all on the same page here, um, do we We were given sort of an example of a coffee machine or something, right? Well, um are we at ma right now on the assumption that our television remote control may have features which go beyond the television? Or are we keeping sort of like a a design commitment to television features? I I don''t know. Yep. Yeah, sure. Okay. Okay, yeah. Okay. Okay. Okay. Alright.' model-index: - name: MEETING_SUMMARY results: - task: type: abstractive-text-summarization name: Abstractive Text Summarization dataset: name: samsum type: samsum metrics: - type: rouge-1 value: 53.8795 name: Validation ROGUE-1 - type: rouge-2 value: 28.4975 name: Validation ROGUE-2 - type: rouge-L value: 44.1899 name: Validation ROGUE-L - type: rouge-Lsum value: 49.4863 name: Validation ROGUE-Lsum - type: gen-length value: 30.088 name: Validation ROGUE-Lsum - type: rouge-1 value: 53.2284 name: Test ROGUE-1 - type: rouge-2 value: 28.184 name: Test ROGUE-2 - type: rouge-L value: 44.122 name: Test ROGUE-L - type: rouge-Lsum value: 49.0301 name: Test ROGUE-Lsum - type: gen-length value: 29.9951 name: Test ROGUE-Lsum - task: type: summarization name: Summarization dataset: name: bazzhangz/sumdataset type: bazzhangz/sumdataset config: bazzhangz--sumdataset split: train metrics: - type: rouge value: 40.5544 name: ROUGE-1 verified: true - type: rouge value: 17.0751 name: ROUGE-2 verified: true - type: rouge value: 32.153 name: ROUGE-L verified: true - type: rouge value: 36.4277 name: ROUGE-LSUM verified: true - type: loss value: 2.116729736328125 name: loss verified: true - type: gen_len value: 42.1978 name: gen_len verified: true - task: type: abstractive-text-summarization name: Abstractive Text Summarization dataset: name: xsum type: xsum metrics: - type: rouge-1 value: 35.9078 name: Validation ROGUE-1 - type: rouge-2 value: 14.2497 name: Validation ROGUE-2 - type: rouge-L value: 28.1421 name: Validation ROGUE-L - type: rouge-Lsum value: 28.9826 name: Validation ROGUE-Lsum - type: gen-length value: 32.0167 name: Validation ROGUE-Lsum - type: rouge-1 value: 36.0241 name: Test ROGUE-1 - type: rouge-2 value: 14.3715 name: Test ROGUE-2 - type: rouge-L value: 28.1968 name: Test ROGUE-L - type: rouge-Lsum value: 29.0527 name: Test ROGUE-Lsum - type: gen-length value: 31.9933 name: Test ROGUE-Lsum - task: type: abstractive-text-summarization name: Abstractive Text Summarization dataset: name: dialogsum type: dialogsum metrics: - type: rouge-1 value: 39.8612 name: Validation ROGUE-1 - type: rouge-2 value: 16.6917 name: Validation ROGUE-2 - type: rouge-L value: 32.2718 name: Validation ROGUE-L - type: rouge-Lsum value: 35.8748 name: Validation ROGUE-Lsum - type: gen-length value: 41.726 name: Validation ROGUE-Lsum - type: rouge-1 value: 36.9608 name: Test ROGUE-1 - type: rouge-2 value: 14.3058 name: Test ROGUE-2 - type: rouge-L value: 29.3261 name: Test ROGUE-L - type: rouge-Lsum value: 32.9 name: Test ROGUE-Lsum - type: gen-length value: 43.086 name: Test ROGUE-Lsum - task: type: summarization name: Summarization dataset: name: samsum type: samsum config: samsum split: test metrics: - type: rouge value: 53.1878 name: ROUGE-1 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiOTVkNTczYjFmYzBmMzczNWE0MGY4MDAyZWExOGNjZmY1Yzk2ZGM1MGNjZmFmYWUyZmIxZjdjOTk4OTc4OGJlMSIsInZlcnNpb24iOjF9.yyzPpGtESuZXy_lBESrboGxdGYB7I6jaIjquCYqliE2xdbGf5awDFpDUwlZHDuw6RD2mIZv1FC8PPs9lOHuSAg - type: rouge value: 28.1666 name: ROUGE-2 verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiMjAzOTdjNGYxNWMzYmFjYjRmMTcxYzI0MmNlNmM5Nzg2MzBlNDdmZWFkN2EwMDE2ZTZmYzc0Zjg0ZDc0M2IxNiIsInZlcnNpb24iOjF9.cPH6O50T6HekO227Xzha-EN_Jp7JS9fh5EP9I0tHxbpGptKtZOQC-NG68zfU2eJKlRSrmgaBYs8tjfTvpAgyDg - type: rouge value: 44.117 name: ROUGE-L verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNmNmMzJkYjMxMjhlZDM4YmU3NmI1MDExNzhiYmVhMzEyZGJjNDJkNzczNGQwOTMwNzg2YjU1ZWQ4MDhiMzkxYiIsInZlcnNpb24iOjF9.lcEXK15UqZOdXnPjVqIhFd6o_PLROSIONTRFX5NbwanjEI_MWMLpDh_V0Kpnvs_W0sE6cXh2yoifSYNDA5W7Bw - type: rouge value: 49.0094 name: ROUGE-LSUM verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiYThkYjk4ZjMzYjI0OTAxNDJiZTU5MzE0YjI5MjEzYTYwNWEzMmU5NjU2ZjQ5NzJhMzkyNmVhNWFjZmM1MjAwMSIsInZlcnNpb24iOjF9.LTn6LpKuMO4Rv4NgsbPmtr2ewiKyoqAXlf6YJfM_6GKwVTKpnJxwx7gaaAtMb0jVlgieITMP11JmbeRfMEhgDg - type: loss value: 1.710614562034607 name: loss verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiNjNjZmM0ZjkwYWYyMWIyMmFiMWI1ODBiYjRjNzVhM2JhN2NmNmM1ZDUwZWRjNDQxNzUwMWM4YjYxYTg1MWYwNyIsInZlcnNpb24iOjF9.hGXZhp9pe-HDJilXVvMCkqz-92YZvH6Qr7q9Z7fJkm8N9s0b4sl-4PwjQYJEOLEAhoRO2s-F5T3bmCYCaMiNBQ - type: gen_len value: 29.9951 name: gen_len verified: true verifyToken: eyJhbGciOiJFZERTQSIsInR5cCI6IkpXVCJ9.eyJoYXNoIjoiZmY1NzZiMDAzNGJlNTg4Nzc0YzU1MTA3YTI3MzVmNGZkNWQ0ZDE4MGZlNGI1MzJmYzA3MjQ0MDZhMTcyYTk2NCIsInZlcnNpb24iOjF9.8dvMfY7Y-nw-K8NGgTXIGFMxaSUWQYBE1w3N5YYOn4iwnCe2ugo2qPIOxLY91q7CaAOMCSskFV3BDStQ4p0ZCg --- Model obtained by Fine Tuning 'facebook/bart-large-xsum' using AMI Meeting Corpus, SAMSUM Dataset, DIALOGSUM Dataset, XSUM Dataset! ## Usage # Example 1 ```python from transformers import pipeline summarizer = pipeline("summarization", model="knkarthick/MEETING_SUMMARY") text = '''The tower is 324 metres (1,063 ft) tall, about the same height as an 81-storey building, and the tallest structure in Paris. Its base is square, measuring 125 metres (410 ft) on each side. During its construction, the Eiffel Tower surpassed the Washington Monument to become the tallest man-made structure in the world, a title it held for 41 years until the Chrysler Building in New York City was finished in 1930. It was the first structure to reach a height of 300 metres. Due to the addition of a broadcasting aerial at the top of the tower in 1957, it is now taller than the Chrysler Building by 5.2 metres (17 ft). Excluding transmitters, the Eiffel Tower is the second tallest free-standing structure in France after the Millau Viaduct. ''' summarizer(text) ``` # Example 2 ```python from transformers import pipeline summarizer = pipeline("summarization", model="knkarthick/MEETING_SUMMARY") text = '''Bangalore is the capital and the largest city of the Indian state of Karnataka. It has a population of more than 8 million and a metropolitan population of around 11 million, making it the third most populous city and fifth most populous urban agglomeration in India. Located in southern India on the Deccan Plateau, at a height of over 900 m (3,000 ft) above sea level, Bangalore is known for its pleasant climate throughout the year. Its elevation is the highest among the major cities of India.The city's history dates back to around 890 CE, in a stone inscription found at the Nageshwara Temple in Begur, Bangalore. The Begur inscription is written in Halegannada (ancient Kannada), mentions 'Bengaluru Kalaga' (battle of Bengaluru). It was a significant turning point in the history of Bangalore as it bears the earliest reference to the name 'Bengaluru'. In 1537 CE, Kempé Gowdā – a feudal ruler under the Vijayanagara Empire – established a mud fort considered to be the foundation of modern Bangalore and its oldest areas, or petes, which exist to the present day. After the fall of Vijayanagar empire in 16th century, the Mughals sold Bangalore to Chikkadevaraja Wodeyar (1673–1704), the then ruler of the Kingdom of Mysore for three lakh rupees. When Haider Ali seized control of the Kingdom of Mysore, the administration of Bangalore passed into his hands. The city was captured by the British East India Company after victory in the Fourth Anglo-Mysore War (1799), who returned administrative control of the city to the Maharaja of Mysore. The old city developed in the dominions of the Maharaja of Mysore and was made capital of the Princely State of Mysore, which existed as a nominally sovereign entity of the British Raj. In 1809, the British shifted their cantonment to Bangalore, outside the old city, and a town grew up around it, which was governed as part of British India. Following India's independence in 1947, Bangalore became the capital of Mysore State, and remained capital when the new Indian state of Karnataka was formed in 1956. The two urban settlements of Bangalore – city and cantonment – which had developed as independent entities merged into a single urban centre in 1949. The existing Kannada name, Bengalūru, was declared the official name of the city in 2006. Bangalore is widely regarded as the "Silicon Valley of India" (or "IT capital of India") because of its role as the nation's leading information technology (IT) exporter. Indian technological organisations are headquartered in the city. A demographically diverse city, Bangalore is the second fastest-growing major metropolis in India. Recent estimates of the metro economy of its urban area have ranked Bangalore either the fourth- or fifth-most productive metro area of India. As of 2017, Bangalore was home to 7,700 millionaires and 8 billionaires with a total wealth of $320 billion. It is home to many educational and research institutions. Numerous state-owned aerospace and defence organisations are located in the city. The city also houses the Kannada film industry. It was ranked the most liveable Indian city with a population of over a million under the Ease of Living Index 2020. ''' summarizer(text) ``` # Example 3 ```python from transformers import pipeline summarizer = pipeline("summarization", model="knkarthick/MEETING_SUMMARY") text = '''Hi, I'm David and I'm supposed to be an industrial designer. Um, I just got the project announcement about what the project is. Designing a remote control. That's about it, didn't get anything else. Did you get the same thing? Cool. There's too much gear. Okay. Can't draw. Um. Yeah. Um, well anyway, I don't know, it's just the first animal I can think off the top of my head. Um. Yes. Big reason is 'cause I'm allergic to most animals. Allergic to animal fur, so um fish was a natural choice. Um, yeah, and I kind of like whales. They come in and go eat everything in sight. And they're quite harmless and mild and interesting. Tail's a bit big, I think. It's an after dinner dog then. Hmm. It does make sense from maybe the design point of view 'cause you have more complicated characters like European languages, then you need more buttons. So, possibly. Hmm. Yeah. And you keep losing them. Finding them is really a pain, you know. I mean it's usually quite small, or when you want it right, it slipped behind the couch or it's kicked under the table. You know. Yep. Mm-hmm. I think one factor would be production cost. Because there's a cap there, so um depends on how much you can cram into that price. Um. I think that that's the main factor. Cool. Okay. Right. Um well this is the kick-off meeting for our our project. Um and um this is just what we're gonna be doing over the next twenty five minutes. Um so first of all, just to kind of make sure that we all know each other, I'm Laura and I'm the project manager. Do you want to introduce yourself again? Okay. Great. Okay. Um so we're designing a new remote control and um Oh I have to record who's here actually. So that's David, Andrew and Craig, isn't it? And you all arrived on time. Um yeah so des uh design a new remote control. Um, as you can see it's supposed to be original, trendy and user friendly. Um so that's kind of our our brief, as it were. Um and so there are three different stages to the design. Um I'm not really sure what what you guys have already received um in your emails. What did you get? Mm-hmm. Is that what everybody got? Okay. Um. So we're gonna have like individual work and then a meeting about it. And repeat that process three times. Um and at this point we get try out the whiteboard over there. Um. So uh you get to draw your favourite animal and sum up your favourite characteristics of it. So who would like to go first? Very good. Mm-hmm. Yeah. Yeah. Right. Lovely. Right. You can take as long over this as you like, because we haven't got an awful lot to discuss. Ok oh we do we do. Don't feel like you're in a rush, anyway. Ach why not We might have to get you up again then. I don't know what mine is. I'm gonna have to think on the spot now. Is that a whale? Ah. Okay. God, I still don't know what I'm gonna write about. Um. I was gonna choose a dog as well. But I'll just draw a different kind of dog. M my favourite animal is my own dog at home. Um That doesn't really look like him, actually. He looks more like a pig, actually. Ah well. Do you? Oh that's very good of you. Uh. Um he's a mixture of uh various things. Um and what do I like about him, um That's just to suggest that his tail wags. Um he's very friendly and cheery and always pleased to see you, and very kind of affectionate and um uh and he's quite quite wee as well so you know he can doesn't take up too much space. Um and uh And he does a funny thing where he chases his tail as well, which is quite amusing, so It is. I think it is. He only does it after he's had his dinner and um he'll just all of a sudden just get up and start chasing his tail 'round the living room. Yeah, so uh Yeah, maybe. Maybe. Right, um where did you find this? Just down here? Yeah. Okay. Um what are we doing next? Uh um. Okay, uh we now need to discuss the project finance. Um so according to the brief um we're gonna be selling this remote control for twenty five Euro, um and we're aiming to make fifty million Euro. Um so we're gonna be selling this on an international scale. And uh we don't want it to cost any more than uh twelve fifty Euros, so fifty percent of the selling price. Sure. All together. Um I dunno. I imagine That's a good question. I imagine it probably is our sale actually because it's probably up to the the um the retailer to uh sell it for whatever price they want. Um. But I I don't know, I mean do you think the fact that it's going to be sold internationally will have a bearing on how we design it at all? Think it will? Um. Hmm. Oh yeah, regions and stuff, yeah. Yeah. Okay. Yeah. Well for a remote control, do you think that will be I suppose it's depends on how complicated our remote control is. Yeah, yeah. Okay. What, just like in terms of like the wealth of the country? Like how much money people have to spend on things like? Aye, I see what you mean, yeah. Marketing. Good marketing thoughts. Oh gosh, I should be writing all this down. Um. Mm. Yeah. Yeah, yeah. Like how much does, you know, a remote control cost. Well twenty five Euro, I mean that's um that's about like eighteen pounds or something, isn't it? Or no, is it as much as that? Sixteen seventeen eighteen pounds. Um, I dunno, I've never bought a remote control, so I don't know how how good a remote control that would get you. Um. But yeah, I suppose it has to look kind of cool and gimmicky. Um right, okay. Let me just scoot on ahead here. Okay. Um well d Does anybody have anything to add to uh to the finance issue at all? Thin No, actually. That would be useful, though, wouldn't it, if you knew like what your money would get you now. Mm-hmm. Yeah, yeah. Oh. Five minutes to end of meeting. Oh, okay. We're a bit behind. Yeah. Right, so do you think that should be like a main design aim of our remote control d you know, do your your satellite and your regular telly and your V_C_R_ and everything? Mm-hmm. Yeah. Or even like, you know, notes about um what you wanna watch. Like you might put in there oh I want to watch such and such and look a Oh that's a good idea. So extra functionalities. Mm-hmm. Hmm. Um okay, uh I'd wel we're gonna have to wrap up pretty quickly in the next couple of minutes. Um I'll just check we've nothing else. Okay. Um so anything else anybody wants to add about what they don't like about remote controls they've used, what they would really like to be part of this new one at all? You keep losing them. Okay. Yeah. W You get those ones where you can, if you like, whistle or make a really high pitched noise they beep. There I mean is that something we'd want to include, do you think? Dunno. Okay maybe. My goodness. Still feels quite primitive. Maybe like a touch screen or something? Okay. Uh-huh, okay. Well I guess that's up to our industrial designer. It looks better. Yeah. Okay. Okay. Right, well um so just to wrap up, the next meeting's gonna be in thirty minutes. So that's about um about ten to twelve by my watch. Um so inbetween now and then, um as the industrial designer, you're gonna be working on you know the actual working design of it so y you know what you're doing there. Um for user interface, technical functions, I guess that's you know like what we've been talking about, what it'll actually do. Um and uh marketing executive, you'll be just thinking about what it actually what, you know, what requirements it has to has to fulfil and you'll all get instructions emailed to you, I guess. Um. Yeah, so it's th the functional design stage is next, I guess. And uh and that's the end of the meeting. So I got that little message a lot sooner than I thought I would, so Mm-hmm. Uh-huh, yeah. Th Okay, well just very quickly 'cause this we're supposed to finish now. Um I guess that's up to us, I mean you probably want some kind of unique selling point of it, so um, you know Yeah. Mm-hmm. Yeah. Okay. Right, okay, we'll that's that's the end of the meeting, then. Um. So, uh thank you all for coming. Um I'm Craig and I'm User Interface. Yeah. Well, my favourite animal would be a monkey. Then they're small cute and furry, and uh when planet of the apes becomes real, I'm gonna be up there with them. Yeah. I know um My parents went out and bought um remote controls because um they got fed up of having four or five different remote controls for each things the house. So um for them it was just how many devices control. Uh. Mm-hmm. Great. And I'm Andrew and I'm uh our marketing expert. Mm-hmm. Mm-hmm. Yeah, that's that's it. Yeah. I will go. That's fine. Alright. So This one here, right? Okay. Very nice. Alright. My favourite animal is like A beagle. Um charac favourite characteristics of it? Is that right? Uh, right, well basically um high priority for any animal for me is that they be willing to take a lot of physical affection from their family. And, yeah that they have lots of personality and uh be fit and in robust good health. So this is blue. Blue beagle. My family's beagle. I coulda told you a whole lot more about beagles. Boy, let me tell you. Impressionist. Alright. Mm. Superb sketch, by the way. Yep. I see a dog in there. Yep. Now I see a rooster. What kind is it? Is he aware that th it's his own cha tail he's chasing? Hmm. Probably when he was little he got lots of attention for doing it and has forever been conditioned. 'Kay. Um, can we just go over that again? Uh, so bas at twel Alright, yeah. Okay. So cost like production cost is twelve fifty, but selling price is is that wholesale or retail? Like on the shelf. Our sale our sale anyway. Yeah, okay okay. Okay. Mm-hmm. Alright. Yes. Mm-hmm. Mm-hmm. Well right away I'm wondering if there's um th th uh, like with D_V_D_ players, if there are zones. Um f frequencies or something um as well as uh characters, um different uh keypad styles and s symbols. Um. I don't know. Yeah. Yeah. Yeah. And then a and then al the other thing international is on top of the price. I'm thinking the price might might appeal to a certain market in one region, whereas in another it'll be different, so Just a chara just a characteristic of the Just Or just like, basic product podi positioning, the twenty five Euro remote control might be a big hit in London, might not be such a big hit in Greece, who knows, something like that, yeah. Yep. Right away I'm making some kind of assumptions about what what information we're given here, thinking, 'kay trendy probably means something other than just basic, something other than just standard. Um so I'm wondering right away, is selling twenty five Euros, is that sort of the thi is this gonna to be like the premium product kinda thing or Uh-huh. Mm-hmm. Yep. Yeah, I'd say so, yeah. No. Yeah, yeah. Mm-hmm. Do we have any other background information on like how that compares to other other Yeah. Mm-hmm. Yeah, interesting thing about discussing um production of a remote control for me is that l as you point out, I just don't think of remote controls as somethin something people consciously assess in their purchasing habits. It's just like getting shoelaces with shoes or something. It just comes along. Do you know what I mean? Like so sort of like how do you I I mean one one way of looking at it would be, well the people producing television sets, maybe they have to buy remote controls. Or another way is maybe people who have T_V_ sets are really fed up with their remote control and they really want a better one or something. But Right. Right. Okay so Right, so in function one of the priorities might be to combine as many uses I think so. Yeah, yeah. Yeah. Well like um, maybe what we could use is a sort of like a example of a successful other piece technology is palm palm pilots. They're gone from being just like little sort of scribble boards to cameras, M_P_ three players, telephones, everything, agenda. So, like, I wonder if we might add something new to the to the remote control market, such as the lighting in your house, or um Yeah, yeah. An Yeah. Like, p personally for me, at home I've I've combined the um the audio video of my television set and my D_V_D_ player and my C_D_ player. So they w all work actually function together but I have different remote controls for each of them. So it's sort of ironic that that then they're in there um you know, the sound and everything it's just one system. But each one's got its own little part. Mm. Mm. Mm. Mm-hmm. Mm-hmm. Yeah. Yeah. That's just really good id Yep. Uh, sure. I remember when the first remote control my my family had was on a cable. Actually had a cable between it and the T_V_ and big like buttons that sort of like, like on a blender or something. And um, you know, when I think about what they are now, it's better, but actually it's still kind of, I dunno, like a massive junky thing on the table. Maybe we could think about how, could be more, you know, streamlined. S Something like that, yeah. Or whatever would be technologically reasonable. 'Cause it could b it could it could be that f it could be that functionally that doesn't make it any better, but that just the appeal of of not having You know, these days there's a r pe things in people's homes are becoming more and more like chic, you know. Um, nicer materials and might be be worth exploring anyway. Okay. Um. Before we wrap up, just to make sure we're all on the same page here, um, do we We were given sort of an example of a coffee machine or something, right? Well, um are we at ma right now on the assumption that our television remote control may have features which go beyond the television? Or are we keeping sort of like a a design commitment to television features? I I don't know. Yep. Yeah, sure. Okay. Okay, yeah. Okay. Okay. Okay. Alright. ''' summarizer(text) ``` # Example 4 ```python from transformers import pipeline summarizer = pipeline("summarization", model="knkarthick/MEETING_SUMMARY") text = ''' Das : Hi and welcome to the a16z podcast. I’m Das, and in this episode, I talk SaaS go-to-market with David Ulevitch and our newest enterprise general partner Kristina Shen. The first half of the podcast looks at how remote work impacts the SaaS go-to-market and what the smartest founders are doing to survive the current crisis. The second half covers pricing approaches and strategy, including how to think about free versus paid trials and navigating the transition to larger accounts. But we start with why it’s easier to move upmarket than down… and the advantage that gives a SaaS startup against incumbents. David : If you have a cohort of customers that are paying you $10,000 a year for your product, you’re going to find a customer that self-selects and is willing to pay $100,000 a year. Once you get one of those, your organization will figure out how you sell to, how you satisfy and support, customers at that price point and that size. But it’s really hard for a company that sells up market to move down market, because they’ve already baked in all that expensive, heavy lifting sales motion. And so as you go down market with a lower price point, usually, you can’t actually support it. Das : Does that mean that it’s easier for a company to do this go-to-market if they’re a new startup as opposed to if they’re a pre-existing SaaS? Kristina : It’s culturally very, very hard to give a product away for free that you’re already charging for. It feels like you’re eating away at your own potential revenue when you do it. So most people who try it end up pulling back very quickly. David : This is actually one of the key reasons why the bottoms up SaaS motion is just so competitive, and compelling, and so destructive against the traditional sales-driven test motion. If you have that great product and people are choosing to use it, it’s very hard for somebody with a sales-driven motion, and all the cost that’s loaded into that, to be able to compete against it. There are so many markets where initially, we would look at companies and say, “Oh, well, this couldn’t possibly be bottoms up. It has to be sold to the CIO. It has to be sold to the CSO or the CFO.” But in almost every case we’ve been wrong, and there has been a bottoms up motion. The canonical example is Slack. It’s crazy that Slack is a bottoms up company, because you’re talking about corporate messaging, and how could you ever have a messaging solution that only a few people might be using, that only a team might be using? But now it’s just, “Oh, yeah, some people started using it, and then more people started using it, and then everyone had Slack.” Kristina : I think another classic example is Dropbox versus Box. Both started as bottoms up businesses, try before you buy. But Box quickly found, “Hey, I’d rather sell to IT.” And Dropbox said, “Hey, we’ve got a great freemium motion going.” And they catalyzed their business around referrals and giving away free storage and shared storage in a way that really helped drive their bottoms up business. Das : It’s a big leap to go from selling to smaller customers to larger customers. How have you seen SaaS companies know or get the timing right on that? Especially since it does seem like that’s really related to scaling your sales force? Kristina : Don’t try to go from a 100-person company to a 20,000-person company. Start targeting early adopters, maybe they’re late stage pre-IPO companies, then newly IPO’d companies. Starting in tech tends to be a little bit easier because they tend to be early adopters. Going vertical by vertical can be a great strategy as well. Targeting one customer who might be branded in that space, can help brand yourself in that category. And then all their competitors will also want your product if you do a good job. A lot of times people will dedicate a sales rep to each vertical, so that they become really, really knowledgeable in that space, and also build their own brand and reputation and know who are the right customers to target. Das : So right now, you’ve got a lot more people working remote. Does this move to remote work mean that on-premise software is dying? And is it accelerating the move to software as a service? Kristina : This remote work and working from home is only going to catalyze more of the conversion from on-premise over to cloud and SaaS. In general, software spend declines 20% during an economic downturn. This happened in ’08, this happened in ’01. But when we look at the last downturn in ’08, SaaS spend actually, for public companies, increased, on average, 10%, which means there’s a 30% spread, which really shows us that there was a huge catalyst from people moving on-premise to SaaS. David : And as people work remote, the ability to use SaaS tools is much easier than having to VPN back into your corporate network. We’ve been seeing that, inside sales teams have been doing larger and larger deals, essentially moving up market on the inside, without having to engage with field sales teams. In fact, a lot of the new SaaS companies today rather than building out a field team, they have a hybrid team, where people are working and closing deals on the inside and if they had to go out and meet with a customer, they would do that. But by and large, most of it was happening over the phone, over email, and over videoconferencing. And all the deals now, by definition, are gonna be done remote because people can’t go visit their customers in person. Das : So with bottoms up, did user behavior and buyer behavior change, so the go-to-market evolved? Or did the go-to-market evolve and then you saw user and buyer behavior change? I’m curious with this move to remote work. Is that going to trigger more changes or has the go-to-market enabled that change in user behavior, even though we see that change coming because of a lot of forces outside of the market? Kristina : I definitely think they are interrelated. But I do think it was a user change that catalyzed everything. We decided that we preferred better software, and we tried a couple products. We were able to purchase off our credit card. And then IT and procurement eventually said, “Wow, everyone’s buying these already, I might as well get a company license and a company deal so I’m not paying as much.” While obviously software vendors had to offer the products that could be self-served, users started to realize they had the power, they wanted to use better software, they paid with their credit cards. And now software vendors are forced to change their go-to-market to actually suit that use case. Das : If that’s the case that when user behavior has changed, it’s tended to be the catalyzing force of bigger changes in the go-to-market, what are some of the changes you foresee for SaaS because the world has changed to this new reality of remote work and more distributed teams? David : We’re in a very uncertain economic environment right now. And a couple of things will become very clear over the next 3 to 9 to 15 months — you’re going to find out which SaaS products are absolutely essential to helping a business operate and run, and which ones were just nice to have and may not get renewed. I think on the customer, buying side, you’re very likely to see people push back on big annual commitments and prefer to go month-to-month where they can. Or you’ll see more incentives from SaaS startups to offer discounts for annual contracts. You’re going to see people that might sign an annual contract, but they may not want to pay upfront. They may prefer to meter the cash out ratably over the term of the contract. And as companies had empowered and allowed budget authority to be pushed down in organizations, you’re gonna see that budget authority get pulled back, more scrutiny on spending, and likely a lot of SaaS products not get renewed that turned out to not be essential. Kristina : I think the smartest founders are making sure they have the runway to continue to exist. And they’re doing that in a couple of ways. They’re preserving cash, and they are making sure that their existing customers are super, super happy, because retaining your customers is so important in this environment. And they’re making sure that they have efficient or profitable customer acquisition. Don’t spend valuable dollars acquiring customers. But acquire customers efficiently that will add to a great existing customer base. Das : To go into pricing and packaging for SaaS for a moment, what are some of the different pricing approaches that you see SaaS companies taking? Kristina : The old school way of doing SaaS go-to-market is bundle everything together, make the pricing super complex, so you don’t actually understand what you’re paying for. You’re forced to purchase it because you need one component of the product. New modern SaaS pricing is keep it simple, keep it tied to value, and make sure you’re solving one thing really, really well. David : You want to make it easy for your customers to give you money. And if your customers don’t understand your pricing, that’s a huge red flag. Sometimes founders will try to over engineer their pricing model. Kristina : We talk a lot about everything has to be 10X better than the alternatives. But it’s much easier to be 10X better when you solve one thing very, very well, and then have simple pricing around it. I think the most common that most people know about is PEPM or per employee per month, where you’re charging basically for every single seat. Another really common model is the freemium model. So, think about a Dropbox, or an Asana, or a Skype, where it’s trigger based. You try the product for free, but when you hit a certain amount of storage, or a certain amount of users, then it converts over to paid. And then you also have a time trial, where you get the full experience of the product for some limited time period. And then you’re asked if you want to continue using the product to pay. And then there’s pay as go, and particularly, pay as you go as a usage model. So, Slack will say, “Hey, if your users aren’t actually using the product this month, we won’t actually charge you for it.” David : The example that Kristina made about Slack and users, everybody understands what a user is, and if they’re using the product, they pay for it, and if they’re not using it, they don’t pay for it. That’s a very friendly way to make it easy for your customers to give you money. If Slack came up with a pricing model that was like based on number of messages, or number of API integration calls, the customer would have no idea what that means. Kristina : There’s also the consumption model. So Twilio only charges you for every SMS text or phone call that you make on the platform any given month. And so they make money or lose money as your usage goes. The pricing is very aligned to your productivity. David : Generally, those are for products where the usage only goes in one direction. If you think of a company like Databricks, where they’re charging for storage, or Amazon’s S3 service, it is very aligned with the customer, but it also strategically aligns with the business because they know the switching cost is very high, the churn is very low. And generally, in those businesses, you’re only going to store more data, so they can charge based on usage or volume of data. Kristina : Recently, there’s been a huge trend of payment as a revenue. It’s particularly common in vertical markets where SaaS companies are adding payments as a revenue in addition to their employee or subscription revenue. If you look at Shopify, for example, more than 50% of their revenue is actually payment revenue. They’re making money every single time you purchase something off one of their shopping cart websites. Das : When you’re working with a founder or a SaaS startup, how have you seen them find the right pricing model for their product, for their market? Kristina : Step one is just talk to a lot of customers. Try to figure out what is the market pricing for possible alternatives or competitors, understand their pain points and their willingness to pay. And just throw a price out there, because you have to have a starting point in order to actually test and iterate. Particularly in the SMB, or the bottoms up business, you can test and iterate pretty quickly because you have so many data points. David : I always tell founders, step one is to just go out there and talk to customers. Step two is just double your prices. I don’t think there’s ever been a great company with a great product that’s fallen apart because their pricing was wrong. But a lot of SaaS startup founders really under price, and you don’t want to find out two or three years later that you were 200% underpriced. A very common thing that SaaS companies do, they’ll have the basic package that either is free or low cost, that you can just sign up online for. They’ll have a middle package where they share some pricing, and then they’ll have the enterprise package where you have to contact sales to find out more. And that way they don’t actually have to show the pricing for that third package. And that gives the salespeople the flexibility to adjust pricing on a per deal basis. Das : When you’re working with companies, why are they underpricing their products? David : I think it’s psychological. People need to price on value, and they don’t know how much value they’re delivering relative to “Oh, it only cost me $100 a month to provide this service, so I just need to charge $200.” But if it turns out you’re saving your customer $50,000 a year, then you’re wildly underpriced. You have to remember that SaaS is essentially a proxy for outsourced IT. You’re spending money on a SaaS service to not pay to develop something internally, or to have to pay IT to support something that’s more complex on-prem. Software is much cheaper than people, and so generally, the price point can be much higher. Kristina : And the other thing is your value increases over time. You’re delivering more features, more products, you understand the customer better. It’s the beauty of the SaaS model and cloud model that you can iterate and push code immediately, and the customer immediately sees value. A lot of times people have the same price point from the first customer sold to three years later and the 200th customer. Quite frankly, you’ve delivered so much value along the way that your price point should have gone up. The other thing I’ll say is a lot of people discount per seat pricing a lot as they move up market. We tend to tell people that the best validation of your product having great product market fit is your ability to hold your price point. So while there is some natural discounting on a per seat basis because people do deserve some volume discounting, I would say try to resist that as much as possible. Das : Especially for a technical founder, it’s so tempting to get in there and fiddle with these knobs. How do you know when it is time to experiment with your pricing and packaging? David : If you’re looking at your business and you see that you are doing more deals, and they’re closing faster, you should raise your pricing. And you pay attention to how long it takes to close deals and whether the number of deals is staying consistent as you do that. And, at some point, you’re going to find out when you’re losing deals on price. I think a moment where companies have to plan ahead to avoid having to course correct is after they roll out massive pricing and packaging changes, which are pretty natural as companies move up market. But how they navigate that transition to larger accounts, and how they either bring along or move away from those smaller, earlier customers who got them to where they are, tends to be really important because they can get a lot of noise on Twitter, they can get a lot of blowback from their customers. So Zendesk is a company where they rolled out a major packaging change. And when they rolled it out, they hadn’t planned on grandfathering in their early customers. They got a lot of pushback, and very quickly, they put out a blog post and said, “We hear what you’re saying, we appreciate you building the business that we’ve become today. We do need to have a package for the future. But all the people that have been customers so far will be grandfathered in for at least a period of time into the old model.” Kristina : If you iterate pricing constantly, you don’t really have this problem because your customers will be used to pricing changes. You normally pair them with new features, and it all kind of works out. But if you have to go through a big grandfather change, I tend to lean towards treating your early customers really, really well. They adopted when you weren’t a big company yet. They probably co-built the product with you in many ways. And so, it’s great to get more dollars out of your customer base, but treat your early customers well. Das : Are there any other failure modes that you see startups really falling into around pricing and packaging or any common mistakes that they make? David : I think a lot of founders don’t always map out the cost or model of their pricing and their product relative to their cost of actually doing sales and marketing and customer acquisition. Kristina : Inside sales is so popular in Silicon Valley. When you’re selling more to an SMB or mid-market type customer, the expectation is that you’re educating and helping the prospective customer over the phone. And so, you’re not expected to be as high touch. But 5K is almost the minimum price point you need to sell to the SMB with an inside sales team in order to pay for the outbound costs and all the conversions, because there is typically a team that sits around the quota carrying rep. And so, price matching — how much your price point is compared to what your go-to-market motion is — matters a lot. Other big failure modes that I see, people guess the ramp time of a sales rep wrong. And ramp time really ties to the segment of customer you’re selling into. It tends be that if you’re selling into the enterprise, the ramp time for sales reps, because sales cycles are so long, tend to be much longer as well. They could be six months plus, could be a year. While if you’re selling more into SMB or mid-market, the ramp time to get a rep up and running can be much shorter, three to six months. Because the sales cycles are shorter, they just iterate much faster, and they ramp up much more quickly. David : The other thing that people have to understand is that sales velocity is a really important component to figuring out how many reps you should be hiring, whether they should be inside reps or field reps. If it takes you 90 days to close a deal, that can’t be a $5,000 a year deal, that has to be a $50,000 or even $150,000 a year deal. Das : Kristina, I know you’ve done a lot of work with metrics. So how do those play in? Kristina : Probably the one way to sum it all together is how many months does it take to pay back customer acquisition cost. Very commonly within the SaaS world, we talk about a 12-month CAC payback. We typically want to see for every dollar you spend on sales and marketing, you get a dollar back within a year. That means you can tweak the inputs any way you want. Let’s say that doing paid acquisition is really effective for you. Then, you can spend proportionally more on paid acquisition and less on sales reps. Vice versa, if you have a great inbound engine, you actually can hire a lot more sales reps and spend more on sales headcount. With all formulas, it’s a guide rail, so if you have customers that retain really, really well, let’s say you’re selling to the enterprise, and you’ve got a 90% or 95% annual retention rate, then your CAC payback could be between 12 and 24 months. But let’s say you’re selling to the SMB and churn is 2% or 3% monthly, which ends up being like 80% to 90% annual retention. Then, because your customer is less sticky, I would recommend looking at a CAC payback of 6 to 12 months. Das : How should you think about doing a free trial versus a paid trial? David : On the one hand, the bottoms up motion where people can try essentially a full version of a product before they buy it is extremely powerful. On the other hand, I’ve started to try to think about how I advise companies, when they are thinking about a free trial for something that might cost $100,000 or $200,000 a year? Do we do a paid pilot that has some sort of contractual obligation that if we meet then turns into a commercial engagement? Kristina : I do think the beauty of the bottoms up business is that you can get people to try the entire experience of the product for free, and they fall in love with it, and a certain percentage will convert. And that works really, really well for products that can self-serve. When you start moving up market to more complex products, the challenge with trials is it takes work to actually implement the product, whether it be integrations, IT has to give access, etc. You lose that self-serve ability, which is so amazing in the trial. And so, I tend to be more in the camp of paid trials, if it costs you money to actually deploy the trial. And when you’re selling to bigger customers, they associate value when they have to pay. Once a customer has to pay you, then they feel a need to make the project successful and thus they will onboard, schedule things, give you data and access. David : If you can get to a point where you get the customer to do that paid pilot, such that the only difference between a pilot and an actual customer is just the signing of a contract, that’s very powerful. Now, that does force you to have a really good pre-sales motion to make sure that you can deliver on the promise you’ve made your customers. When companies don’t have a great product, and they paper over it with professional services and sales engineering and post-sales support, that paid pilot thing doesn’t work because the experience isn’t good enough. So, it really is incumbent on the SaaS company that does a paid pilot to make sure that they are able to deliver on that experience. Kristina : And one emerging trend recently is people signing an annual contract with a one or three month out, as a replacement to the paid pilot. Because it’s the best of both worlds, the SaaS company that’s selling the product gets a higher level of commitment. And the customer gets the optionality of opting out in the same way as a trial without any clawback. It really comes down to where procurement falls. Sometimes procurement is at the beginning of that decision, which makes it more like an annual contract. Sometimes procurement is at the one or three month opt-out period, which means the customer already has a great experience, loves the product, and it is an easier way to convert procurements to actually sign on… David : And that is a really good segue into renewals. I always tell founders, you might have this subscription business, but it’s not a recurring revenue business until the second year when the revenue actually recurs. I think you really have the first three months to get a customer up and running and happy. And if they’re not, you then have about three months to fix it. And if all that works out, then the remaining six months of the contract can be focused on upsell and expansion. Das : Awesome. Thank you, Kristina. Thank you, David. Kristina : Thanks so much for having us. This was fun. David : Yeah, a lot of fun, great topics, and our favorite thing to talk about. ''' summarizer(text) ```

timm/ViT-B-16-SigLIP-256

timm

open_clip

zero-shot-image-classification

--- tags: - clip - siglip library_name: open_clip pipeline_tag: zero-shot-image-classification license: apache-2.0 datasets: - webli --- # Model card for ViT-B-16-SigLIP-256 A SigLIP (Sigmoid loss for Language-Image Pre-training) model trained on WebLI. This model has been converted to PyTorch from the original JAX checkpoints in [Big Vision](https://github.com/google-research/big_vision). These weights are usable in both OpenCLIP (image + text) and timm (image only). ## Model Details - **Model Type:** Contrastive Image-Text, Zero-Shot Image Classification. - **Original:** https://github.com/google-research/big_vision - **Dataset:** WebLI - **Papers:** - Sigmoid loss for language image pre-training: https://arxiv.org/abs/2303.15343 ## Model Usage ### With OpenCLIP ``` import torch import torch.nn.functional as F from urllib.request import urlopen from PIL import Image from open_clip import create_model_from_pretrained, get_tokenizer # works on open-clip-torch>=2.23.0, timm>=0.9.8 model, preprocess = create_model_from_pretrained('hf-hub:timm/ViT-B-16-SigLIP-256') tokenizer = get_tokenizer('hf-hub:timm/ViT-B-16-SigLIP-256') image = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) image = preprocess(image).unsqueeze(0) labels_list = ["a dog", "a cat", "a donut", "a beignet"] text = tokenizer(labels_list, context_length=model.context_length) with torch.no_grad(), torch.cuda.amp.autocast(): image_features = model.encode_image(image) text_features = model.encode_text(text) image_features = F.normalize(image_features, dim=-1) text_features = F.normalize(text_features, dim=-1) text_probs = torch.sigmoid(image_features @ text_features.T * model.logit_scale.exp() + model.logit_bias) zipped_list = list(zip(labels_list, [round(p.item(), 3) for p in text_probs[0]])) print("Label probabilities: ", zipped_list) ``` ### With `timm` (for image embeddings) ```python from urllib.request import urlopen from PIL import Image import timm image = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'vit_base_patch16_siglip_256', pretrained=True, num_classes=0, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(image).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor ``` ## Citation ```bibtex @article{zhai2023sigmoid, title={Sigmoid loss for language image pre-training}, author={Zhai, Xiaohua and Mustafa, Basil and Kolesnikov, Alexander and Beyer, Lucas}, journal={arXiv preprint arXiv:2303.15343}, year={2023} } ``` ```bibtex @misc{big_vision, author = {Beyer, Lucas and Zhai, Xiaohua and Kolesnikov, Alexander}, title = {Big Vision}, year = {2022}, publisher = {GitHub}, journal = {GitHub repository}, howpublished = {\url{https://github.com/google-research/big_vision}} } ```

Salesforce/codegen-350M-nl

Salesforce

transformers

text-generation

--- license: bsd-3-clause --- # CodeGen (CodeGen-NL 350M) ## Model description CodeGen is a family of autoregressive language models for **program synthesis** from the paper: [A Conversational Paradigm for Program Synthesis](https://arxiv.org/abs/2203.13474) by Erik Nijkamp, Bo Pang, Hiroaki Hayashi, Lifu Tu, Huan Wang, Yingbo Zhou, Silvio Savarese, Caiming Xiong. The models are originally released in [this repository](https://github.com/salesforce/CodeGen), under 3 pre-training data variants (`NL`, `Multi`, `Mono`) and 4 model size variants (`350M`, `2B`, `6B`, `16B`). The checkpoint included in this repository is denoted as **CodeGen-NL 350M** in the paper, where "NL" means it is pre-trained on the Pile and "350M" refers to the number of trainable parameters. ## Training data This checkpoint (CodeGen-NL 350M) was pre-trained on [the Pile](https://github.com/EleutherAI/the-pile), a large-scale curated dataset created by [EleutherAI](https://www.eleuther.ai/). Parts of the dataset include code data. ## Training procedure CodeGen was trained using cross-entropy loss to maximize the likelihood of sequential inputs. The family of models are trained using multiple TPU-v4-512 by Google, leveraging data and model parallelism. See Section 2.3 of the [paper](https://arxiv.org/abs/2203.13474) for more details. ## Evaluation results We evaluate our models on two code generation benchmark: HumanEval and MTPB. Please refer to the [paper](https://arxiv.org/abs/2203.13474) for more details. ## Intended Use and Limitations As an autoregressive language model, CodeGen is capable of extracting features from given natural language and programming language texts, and calculating the likelihood of them. However, the model is intended for and best at **program synthesis**, that is, generating executable code given English prompts, where the prompts should be in the form of a comment string. The model can complete partially-generated code as well. ## How to use This model can be easily loaded using the `AutoModelForCausalLM` functionality: ```python from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Salesforce/codegen-350M-nl") model = AutoModelForCausalLM.from_pretrained("Salesforce/codegen-350M-nl") text = "def hello_world():" input_ids = tokenizer(text, return_tensors="pt").input_ids generated_ids = model.generate(input_ids, max_length=128) print(tokenizer.decode(generated_ids[0], skip_special_tokens=True)) ``` ## BibTeX entry and citation info ```bibtex @article{Nijkamp2022ACP, title={A Conversational Paradigm for Program Synthesis}, author={Nijkamp, Erik and Pang, Bo and Hayashi, Hiroaki and Tu, Lifu and Wang, Huan and Zhou, Yingbo and Savarese, Silvio and Xiong, Caiming}, journal={arXiv preprint}, year={2022} } ```

microsoft/BiomedCLIP-PubMedBERT_256-vit_base_patch16_224

microsoft

open_clip

zero-shot-image-classification

--- language: en tags: - clip - biology - medical license: mit library_name: open_clip widget: - src: https://huggingface.co/microsoft/BiomedCLIP-PubMedBERT_256-vit_base_patch16_224/resolve/main/example_data/biomed_image_classification_example_data/squamous_cell_carcinoma_histopathology.jpeg candidate_labels: adenocarcinoma histopathology, squamous cell carcinoma histopathology example_title: squamous cell carcinoma histopathology - src: >- https://huggingface.co/microsoft/BiomedCLIP-PubMedBERT_256-vit_base_patch16_224/resolve/main/example_data/biomed_image_classification_example_data/adenocarcinoma_histopathology.jpg candidate_labels: adenocarcinoma histopathology, squamous cell carcinoma histopathology example_title: adenocarcinoma histopathology - src: >- https://upload.wikimedia.org/wikipedia/commons/5/57/Left-sided_Pleural_Effusion.jpg candidate_labels: left-sided pleural effusion chest x-ray, right-sided pleural effusion chest x-ray, normal chest x-ray example_title: left-sided pleural effusion chest x-ray pipeline_tag: zero-shot-image-classification --- # BiomedCLIP-PubMedBERT_256-vit_base_patch16_224 [BiomedCLIP](https://aka.ms/biomedclip-paper) is a biomedical vision-language foundation model that is pretrained on [PMC-15M](https://aka.ms/biomedclip-paper), a dataset of 15 million figure-caption pairs extracted from biomedical research articles in PubMed Central, using contrastive learning. It uses PubMedBERT as the text encoder and Vision Transformer as the image encoder, with domain-specific adaptations. It can perform various vision-language processing (VLP) tasks such as cross-modal retrieval, image classification, and visual question answering. BiomedCLIP establishes new state of the art in a wide range of standard datasets, and substantially outperforms prior VLP approaches:  ## Citation ```bibtex @misc{https://doi.org/10.48550/arXiv.2303.00915, doi = {10.48550/ARXIV.2303.00915}, url = {https://arxiv.org/abs/2303.00915}, author = {Zhang, Sheng and Xu, Yanbo and Usuyama, Naoto and Bagga, Jaspreet and Tinn, Robert and Preston, Sam and Rao, Rajesh and Wei, Mu and Valluri, Naveen and Wong, Cliff and Lungren, Matthew and Naumann, Tristan and Poon, Hoifung}, title = {Large-Scale Domain-Specific Pretraining for Biomedical Vision-Language Processing}, publisher = {arXiv}, year = {2023}, } ``` ## Model Use ### How to use Please refer to this [example notebook](https://aka.ms/biomedclip-example-notebook). ### Intended Use This model is intended to be used solely for (I) future research on visual-language processing and (II) reproducibility of the experimental results reported in the reference paper. #### Primary Intended Use The primary intended use is to support AI researchers building on top of this work. BiomedCLIP and its associated models should be helpful for exploring various biomedical VLP research questions, especially in the radiology domain. #### Out-of-Scope Use **Any** deployed use case of the model --- commercial or otherwise --- is currently out of scope. Although we evaluated the models using a broad set of publicly-available research benchmarks, the models and evaluations are not intended for deployed use cases. Please refer to [the associated paper](https://aka.ms/biomedclip-paper) for more details. ## Data This model builds upon [PMC-15M dataset](https://aka.ms/biomedclip-paper), which is a large-scale parallel image-text dataset for biomedical vision-language processing. It contains 15 million figure-caption pairs extracted from biomedical research articles in PubMed Central. It covers a diverse range of biomedical image types, such as microscopy, radiography, histology, and more. ## Limitations This model was developed using English corpora, and thus can be considered English-only. ## Further information Please refer to the corresponding paper, ["Large-Scale Domain-Specific Pretraining for Biomedical Vision-Language Processing"](https://aka.ms/biomedclip-paper) for additional details on the model training and evaluation.

HooshvareLab/bert-base-parsbert-uncased

HooshvareLab

transformers

fill-mask

## ParsBERT: Transformer-based Model for Persian Language Understanding ParsBERT is a monolingual language model based on Google’s BERT architecture with the same configurations as BERT-Base. Paper presenting ParsBERT: [arXiv:2005.12515](https://arxiv.org/abs/2005.12515) All the models (downstream tasks) are uncased and trained with whole word masking. (coming soon stay tuned) --- ## Introduction This model is pre-trained on a large Persian corpus with various writing styles from numerous subjects (e.g., scientific, novels, news) with more than 2M documents. A large subset of this corpus was crawled manually. As a part of ParsBERT methodology, an extensive pre-processing combining POS tagging and WordPiece segmentation was carried out to bring the corpus into a proper format. This process produces more than 40M true sentences. ## Evaluation ParsBERT is evaluated on three NLP downstream tasks: Sentiment Analysis (SA), Text Classification, and Named Entity Recognition (NER). For this matter and due to insufficient resources, two large datasets for SA and two for text classification were manually composed, which are available for public use and benchmarking. ParsBERT outperformed all other language models, including multilingual BERT and other hybrid deep learning models for all tasks, improving the state-of-the-art performance in Persian language modeling. ## Results The following table summarizes the F1 score obtained by ParsBERT as compared to other models and architectures. ### Sentiment Analysis (SA) task | Dataset | ParsBERT | mBERT | DeepSentiPers | |:--------------------------:|:---------:|:-----:|:-------------:| | Digikala User Comments | 81.74* | 80.74 | - | | SnappFood User Comments | 88.12* | 87.87 | - | | SentiPers (Multi Class) | 71.11* | - | 69.33 | | SentiPers (Binary Class) | 92.13* | - | 91.98 | ### Text Classification (TC) task | Dataset | ParsBERT | mBERT | |:-----------------:|:--------:|:-----:| | Digikala Magazine | 93.59* | 90.72 | | Persian News | 97.19* | 95.79 | ### Named Entity Recognition (NER) task | Dataset | ParsBERT | mBERT | MorphoBERT | Beheshti-NER | LSTM-CRF | Rule-Based CRF | BiLSTM-CRF | |:-------:|:--------:|:--------:|:----------:|:--------------:|:----------:|:----------------:|:------------:| | PEYMA | 93.10* | 86.64 | - | 90.59 | - | 84.00 | - | | ARMAN | 98.79* | 95.89 | 89.9 | 84.03 | 86.55 | - | 77.45 | **If you tested ParsBERT on a public dataset and you want to add your results to the table above, open a pull request or contact us. Also make sure to have your code available online so we can add it as a reference** ## How to use ### TensorFlow 2.0 ```python from transformers import AutoConfig, AutoTokenizer, TFAutoModel config = AutoConfig.from_pretrained("HooshvareLab/bert-base-parsbert-uncased") tokenizer = AutoTokenizer.from_pretrained("HooshvareLab/bert-base-parsbert-uncased") model = AutoModel.from_pretrained("HooshvareLab/bert-base-parsbert-uncased") text = "ما در هوشواره معتقدیم با انتقال صحیح دانش و آگاهی، همه افراد می‌توانند از ابزارهای هوشمند استفاده کنند. شعار ما هوش مصنوعی برای همه است." tokenizer.tokenize(text) >>> ['ما', 'در', 'هوش', '##واره', 'معتقدیم', 'با', 'انتقال', 'صحیح', 'دانش', 'و', 'اگاهی', '،', 'همه', 'افراد', 'میتوانند', 'از', 'ابزارهای', 'هوشمند', 'استفاده', 'کنند', '.', 'شعار', 'ما', 'هوش', 'مصنوعی', 'برای', 'همه', 'است', '.'] ``` ### Pytorch ```python from transformers import AutoConfig, AutoTokenizer, AutoModel config = AutoConfig.from_pretrained("HooshvareLab/bert-base-parsbert-uncased") tokenizer = AutoTokenizer.from_pretrained("HooshvareLab/bert-base-parsbert-uncased") model = AutoModel.from_pretrained("HooshvareLab/bert-base-parsbert-uncased") ``` ## NLP Tasks Tutorial Coming soon stay tuned ## Cite Please cite the following paper in your publication if you are using [ParsBERT](https://arxiv.org/abs/2005.12515) in your research: ```markdown @article{ParsBERT, title={ParsBERT: Transformer-based Model for Persian Language Understanding}, author={Mehrdad Farahani, Mohammad Gharachorloo, Marzieh Farahani, Mohammad Manthouri}, journal={ArXiv}, year={2020}, volume={abs/2005.12515} } ``` ## Acknowledgments We hereby, express our gratitude to the [Tensorflow Research Cloud (TFRC) program](https://tensorflow.org/tfrc) for providing us with the necessary computation resources. We also thank [Hooshvare](https://hooshvare.com) Research Group for facilitating dataset gathering and scraping online text resources. ## Contributors - Mehrdad Farahani: [Linkedin](https://www.linkedin.com/in/m3hrdadfi/), [Twitter](https://twitter.com/m3hrdadfi), [Github](https://github.com/m3hrdadfi) - Mohammad Gharachorloo: [Linkedin](https://www.linkedin.com/in/mohammad-gharachorloo/), [Twitter](https://twitter.com/MGharachorloo), [Github](https://github.com/baarsaam) - Marzieh Farahani: [Linkedin](https://www.linkedin.com/in/marziehphi/), [Twitter](https://twitter.com/marziehphi), [Github](https://github.com/marziehphi) - Mohammad Manthouri: [Linkedin](https://www.linkedin.com/in/mohammad-manthouri-aka-mansouri-07030766/), [Twitter](https://twitter.com/mmanthouri), [Github](https://github.com/mmanthouri) - Hooshvare Team: [Official Website](https://hooshvare.com/), [Linkedin](https://www.linkedin.com/company/hooshvare), [Twitter](https://twitter.com/hooshvare), [Github](https://github.com/hooshvare), [Instagram](https://www.instagram.com/hooshvare/) ## Releases ### Release v0.1 (May 27, 2019) This is the first version of our ParsBERT based on BERT_BASE

typeform/distilbert-base-uncased-mnli

typeform

transformers

zero-shot-classification

--- language: en pipeline_tag: zero-shot-classification tags: - distilbert datasets: - multi_nli metrics: - accuracy --- # DistilBERT base model (uncased) ## Table of Contents - [Model Details](#model-details) - [How to Get Started With the Model](#how-to-get-started-with-the-model) - [Uses](#uses) - [Risks, Limitations and Biases](#risks-limitations-and-biases) - [Training](#training) - [Evaluation](#evaluation) - [Environmental Impact](#environmental-impact) ## Model Details **Model Description:** This is the [uncased DistilBERT model](https://huggingface.co/distilbert-base-uncased) fine-tuned on [Multi-Genre Natural Language Inference](https://huggingface.co/datasets/multi_nli) (MNLI) dataset for the zero-shot classification task. - **Developed by:** The [Typeform](https://www.typeform.com/) team. - **Model Type:** Zero-Shot Classification - **Language(s):** English - **License:** Unknown - **Parent Model:** See the [distilbert base uncased model](https://huggingface.co/distilbert-base-uncased) for more information about the Distilled-BERT base model. ## How to Get Started with the Model ```python from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("typeform/distilbert-base-uncased-mnli") model = AutoModelForSequenceClassification.from_pretrained("typeform/distilbert-base-uncased-mnli") ``` ## Uses This model can be used for text classification tasks. ## Risks, Limitations and Biases **CONTENT WARNING: Readers should be aware this section contains content that is disturbing, offensive, and can propagate historical and current stereotypes.** Significant research has explored bias and fairness issues with language models (see, e.g., [Sheng et al. (2021)](https://aclanthology.org/2021.acl-long.330.pdf) and [Bender et al. (2021)](https://dl.acm.org/doi/pdf/10.1145/3442188.3445922)). ## Training #### Training Data This model of DistilBERT-uncased is pretrained on the Multi-Genre Natural Language Inference [(MultiNLI)](https://huggingface.co/datasets/multi_nli) corpus. It is a crowd-sourced collection of 433k sentence pairs annotated with textual entailment information. The corpus covers a range of genres of spoken and written text, and supports a distinctive cross-genre generalization evaluation. This model is also **not** case-sensitive, i.e., it does not make a difference between "english" and "English". #### Training Procedure Training is done on a [p3.2xlarge](https://aws.amazon.com/ec2/instance-types/p3/) AWS EC2 with the following hyperparameters: ``` $ run_glue.py \ --model_name_or_path distilbert-base-uncased \ --task_name mnli \ --do_train \ --do_eval \ --max_seq_length 128 \ --per_device_train_batch_size 16 \ --learning_rate 2e-5 \ --num_train_epochs 5 \ --output_dir /tmp/distilbert-base-uncased_mnli/ ``` ## Evaluation #### Evaluation Results When fine-tuned on downstream tasks, this model achieves the following results: - **Epoch = ** 5.0 - **Evaluation Accuracy =** 0.8206875508543532 - **Evaluation Loss =** 0.8706700205802917 - ** Evaluation Runtime = ** 17.8278 - ** Evaluation Samples per second = ** 551.498 MNLI and MNLI-mm results: | Task | MNLI | MNLI-mm | |:----:|:----:|:----:| | | 82.0 | 82.0 | ## Environmental Impact Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700). We present the hardware type based on the [associated paper](https://arxiv.org/pdf/2105.09680.pdf). **Hardware Type:** 1 NVIDIA Tesla V100 GPUs **Hours used:** Unknown **Cloud Provider:** AWS EC2 P3 **Compute Region:** Unknown **Carbon Emitted:** (Power consumption x Time x Carbon produced based on location of power grid): Unknown

google/mobilenet_v2_1.0_224

google

transformers

image-classification

--- license: other tags: - vision - image-classification datasets: - imagenet-1k widget: - src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/tiger.jpg example_title: Tiger - src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/teapot.jpg example_title: Teapot - src: https://huggingface.co/datasets/mishig/sample_images/resolve/main/palace.jpg example_title: Palace --- # MobileNet V2 MobileNet V2 model pre-trained on ImageNet-1k at resolution 224x224. It was introduced in [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://arxiv.org/abs/1801.04381) by Mark Sandler, Andrew Howard, Menglong Zhu, Andrey Zhmoginov, Liang-Chieh Chen. It was first released in [this repository](https://github.com/tensorflow/models/tree/master/research/slim/nets/mobilenet). Disclaimer: The team releasing MobileNet V2 did not write a model card for this model so this model card has been written by the Hugging Face team. ## Model description From the [original README](https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet_v1.md): > MobileNets are small, low-latency, low-power models parameterized to meet the resource constraints of a variety of use cases. They can be built upon for classification, detection, embeddings and segmentation similar to how other popular large scale models, such as Inception, are used. MobileNets can be run efficiently on mobile devices [...] MobileNets trade off between latency, size and accuracy while comparing favorably with popular models from the literature. The checkpoints are named **mobilenet\_v2\_*depth*\_*size***, for example **mobilenet\_v2\_1.0\_224**, where **1.0** is the depth multiplier and **224** is the resolution of the input images the model was trained on. ## Intended uses & limitations You can use the raw model for image classification. See the [model hub](https://huggingface.co/models?search=mobilenet_v2) to look for fine-tuned versions on a task that interests you. ### How to use Here is how to use this model to classify an image of the COCO 2017 dataset into one of the 1,000 ImageNet classes: ```python from transformers import AutoImageProcessor, AutoModelForImageClassification from PIL import Image import requests url = "http://images.cocodataset.org/val2017/000000039769.jpg" image = Image.open(requests.get(url, stream=True).raw) preprocessor = AutoImageProcessor.from_pretrained("google/mobilenet_v2_1.0_224") model = AutoModelForImageClassification.from_pretrained("google/mobilenet_v2_1.0_224") inputs = preprocessor(images=image, return_tensors="pt") outputs = model(**inputs) logits = outputs.logits # model predicts one of the 1000 ImageNet classes predicted_class_idx = logits.argmax(-1).item() print("Predicted class:", model.config.id2label[predicted_class_idx]) ``` Note: This model actually predicts 1001 classes, the 1000 classes from ImageNet plus an extra “background” class (index 0). Currently, both the feature extractor and model support PyTorch. ### BibTeX entry and citation info ```bibtex @inproceedings{mobilenetv22018, title={MobileNetV2: Inverted Residuals and Linear Bottlenecks}, author={Mark Sandler and Andrew Howard and Menglong Zhu and Andrey Zhmoginov and Liang-Chieh Chen}, booktitle={CVPR}, year={2018} } ```

sadickam/sdg-classification-bert

sadickam

transformers

text-classification

--- license: mit language: - en metrics: - accuracy - matthews_correlation --- # sadickam/sdg-classification-bert <!-- Provide a quick summary of what the model is/does. --> This model is for classifying text with respect to the United Nations sustainable development goals (SDG).  Source:https://www.un.org/development/desa/disabilities/about-us/sustainable-development-goals-sdgs-and-disability.html ## Model Details ### Model Description <!-- Provide a longer summary of what this model is. --> This text classification model was developed by fine-tuning the bert-base-uncased pre-trained model. The training data for this fine-tuned model was sourced from the publicly available OSDG Community Dataset (OSDG-CD) at https://zenodo.org/record/5550238#.ZBulfcJByF4. This model was made as part of academic research at Deakin University. The goal was to make a transformer-based SDG text classification model that anyone could use. Only the first 16 UN SDGs supported. The primary model details are highlighted below: - **Model type:** Text classification - **Language(s) (NLP):** English - **License:** mit - **Finetuned from model [optional]:** bert-base-uncased ### Model Sources <!-- Provide the basic links for the model. --> - **Repository:** https://huggingface.co/sadickam/sdg-classification-bert - **Demo [optional]:** option 1: https://sadickam-sdg-text-classifier.hf.space/; option 2: https://sadickam-sdg-classification-bert-main-qxg1gv.streamlit.app/ ### Direct Use <!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. --> This is a fine-tuned model and therefore requires no further training. ## How to Get Started with the Model Use the code below to get started with the model. ```python from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("sadickam/sdg-classification-bert") model = AutoModelForSequenceClassification.from_pretrained("sadickam/sdg-classification-bert") ``` ## Training Data <!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. --> The training data includes text from a wide range of industries and academic research fields. Hence, this fine-tuned model is not for a specific industry. See training here: https://zenodo.org/record/5550238#.ZBulfcJByF4 ## Training Hyperparameters - Num_epoch = 3 - Learning rate = 5e-5 - Batch size = 16 ## Evaluation #### Metrics <!-- These are the evaluation metrics being used, ideally with a description of why. --> - Accuracy = 0.9 - Matthews correlation = 0.89 ## Citation Sadick, A.M. (2023). SDG classification with BERT. https://huggingface.co/sadickam/sdg-classification-bert <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. --> <!--## Model Card Contact -->

michiyasunaga/BioLinkBERT-base

michiyasunaga

transformers

text-classification

--- license: apache-2.0 language: en datasets: - pubmed tags: - bert - exbert - linkbert - biolinkbert - feature-extraction - fill-mask - question-answering - text-classification - token-classification widget: - text: "Sunitinib is a tyrosine kinase inhibitor" --- ## BioLinkBERT-base BioLinkBERT-base model pretrained on [PubMed](https://pubmed.ncbi.nlm.nih.gov/) abstracts along with citation link information. It is introduced in the paper [LinkBERT: Pretraining Language Models with Document Links (ACL 2022)](https://arxiv.org/abs/2203.15827). The code and data are available in [this repository](https://github.com/michiyasunaga/LinkBERT). This model achieves state-of-the-art performance on several biomedical NLP benchmarks such as [BLURB](https://microsoft.github.io/BLURB/) and [MedQA-USMLE](https://github.com/jind11/MedQA). ## Model description LinkBERT is a transformer encoder (BERT-like) model pretrained on a large corpus of documents. It is an improvement of BERT that newly captures **document links** such as hyperlinks and citation links to include knowledge that spans across multiple documents. Specifically, it was pretrained by feeding linked documents into the same language model context, besides a single document. LinkBERT can be used as a drop-in replacement for BERT. It achieves better performance for general language understanding tasks (e.g. text classification), and is also particularly effective for **knowledge-intensive** tasks (e.g. question answering) and **cross-document** tasks (e.g. reading comprehension, document retrieval). ## Intended uses & limitations The model can be used by fine-tuning on a downstream task, such as question answering, sequence classification, and token classification. You can also use the raw model for feature extraction (i.e. obtaining embeddings for input text). ### How to use To use the model to get the features of a given text in PyTorch: ```python from transformers import AutoTokenizer, AutoModel tokenizer = AutoTokenizer.from_pretrained('michiyasunaga/BioLinkBERT-base') model = AutoModel.from_pretrained('michiyasunaga/BioLinkBERT-base') inputs = tokenizer("Sunitinib is a tyrosine kinase inhibitor", return_tensors="pt") outputs = model(**inputs) last_hidden_states = outputs.last_hidden_state ``` For fine-tuning, you can use [this repository](https://github.com/michiyasunaga/LinkBERT) or follow any other BERT fine-tuning codebases. ## Evaluation results When fine-tuned on downstream tasks, LinkBERT achieves the following results. **Biomedical benchmarks ([BLURB](https://microsoft.github.io/BLURB/), [MedQA](https://github.com/jind11/MedQA), [MMLU](https://github.com/hendrycks/test), etc.):** BioLinkBERT attains new state-of-the-art. | | BLURB score | PubMedQA | BioASQ | MedQA-USMLE | | ---------------------- | -------- | -------- | ------- | -------- | | PubmedBERT-base | 81.10 | 55.8 | 87.5 | 38.1 | | **BioLinkBERT-base** | **83.39** | **70.2** | **91.4** | **40.0** | | **BioLinkBERT-large** | **84.30** | **72.2** | **94.8** | **44.6** | | | MMLU-professional medicine | | ---------------------- | -------- | | GPT-3 (175 params) | 38.7 | | UnifiedQA (11B params) | 43.2 | | **BioLinkBERT-large (340M params)** | **50.7** | ## Citation If you find LinkBERT useful in your project, please cite the following: ```bibtex @InProceedings{yasunaga2022linkbert, author = {Michihiro Yasunaga and Jure Leskovec and Percy Liang}, title = {LinkBERT: Pretraining Language Models with Document Links}, year = {2022}, booktitle = {Association for Computational Linguistics (ACL)}, } ```

TheBloke/vicuna-13B-v1.5-GPTQ

TheBloke

transformers

text-generation

--- license: llama2 model_name: Vicuna 13B v1.5 base_model: lmsys/vicuna-13b-v1.5 inference: false model_creator: lmsys model_type: llama prompt_template: 'A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user''s questions. USER: {prompt} ASSISTANT: ' quantized_by: TheBloke --- <!-- header start --> <!-- 200823 --> <div style="width: auto; margin-left: auto; margin-right: auto"> <img src="https://i.imgur.com/EBdldam.jpg" alt="TheBlokeAI" style="width: 100%; min-width: 400px; display: block; margin: auto;"> </div> <div style="display: flex; justify-content: space-between; width: 100%;"> <div style="display: flex; flex-direction: column; align-items: flex-start;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://discord.gg/theblokeai">Chat & support: TheBloke's Discord server</a></p> </div> <div style="display: flex; flex-direction: column; align-items: flex-end;"> <p style="margin-top: 0.5em; margin-bottom: 0em;"><a href="https://www.patreon.com/TheBlokeAI">Want to contribute? TheBloke's Patreon page</a></p> </div> </div> <div style="text-align:center; margin-top: 0em; margin-bottom: 0em"><p style="margin-top: 0.25em; margin-bottom: 0em;">TheBloke's LLM work is generously supported by a grant from <a href="https://a16z.com">andreessen horowitz (a16z)</a></p></div> <hr style="margin-top: 1.0em; margin-bottom: 1.0em;"> <!-- header end --> # Vicuna 13B v1.5 - GPTQ - Model creator: [lmsys](https://huggingface.co/lmsys) - Original model: [Vicuna 13B v1.5](https://huggingface.co/lmsys/vicuna-13b-v1.5) <!-- description start --> ## Description This repo contains GPTQ model files for [lmsys's Vicuna 13B v1.5](https://huggingface.co/lmsys/vicuna-13b-v1.5). Multiple GPTQ parameter permutations are provided; see Provided Files below for details of the options provided, their parameters, and the software used to create them. <!-- description end --> <!-- repositories-available start --> ## Repositories available * [AWQ model(s) for GPU inference.](https://huggingface.co/TheBloke/vicuna-13B-v1.5-AWQ) * [GPTQ models for GPU inference, with multiple quantisation parameter options.](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ) * [2, 3, 4, 5, 6 and 8-bit GGUF models for CPU+GPU inference](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GGUF) * [lmsys's original unquantised fp16 model in pytorch format, for GPU inference and for further conversions](https://huggingface.co/lmsys/vicuna-13b-v1.5) <!-- repositories-available end --> <!-- prompt-template start --> ## Prompt template: Vicuna ``` A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT: ``` <!-- prompt-template end --> <!-- README_GPTQ.md-provided-files start --> ## Provided files and GPTQ parameters Multiple quantisation parameters are provided, to allow you to choose the best one for your hardware and requirements. Each separate quant is in a different branch. See below for instructions on fetching from different branches. All recent GPTQ files are made with AutoGPTQ, and all files in non-main branches are made with AutoGPTQ. Files in the `main` branch which were uploaded before August 2023 were made with GPTQ-for-LLaMa. <details> <summary>Explanation of GPTQ parameters</summary> - Bits: The bit size of the quantised model. - GS: GPTQ group size. Higher numbers use less VRAM, but have lower quantisation accuracy. "None" is the lowest possible value. - Act Order: True or False. Also known as `desc_act`. True results in better quantisation accuracy. Some GPTQ clients have had issues with models that use Act Order plus Group Size, but this is generally resolved now. - Damp %: A GPTQ parameter that affects how samples are processed for quantisation. 0.01 is default, but 0.1 results in slightly better accuracy. - GPTQ dataset: The dataset used for quantisation. Using a dataset more appropriate to the model's training can improve quantisation accuracy. Note that the GPTQ dataset is not the same as the dataset used to train the model - please refer to the original model repo for details of the training dataset(s). - Sequence Length: The length of the dataset sequences used for quantisation. Ideally this is the same as the model sequence length. For some very long sequence models (16+K), a lower sequence length may have to be used. Note that a lower sequence length does not limit the sequence length of the quantised model. It only impacts the quantisation accuracy on longer inference sequences. - ExLlama Compatibility: Whether this file can be loaded with ExLlama, which currently only supports Llama models in 4-bit. </details> | Branch | Bits | GS | Act Order | Damp % | GPTQ Dataset | Seq Len | Size | ExLlama | Desc | | ------ | ---- | -- | --------- | ------ | ------------ | ------- | ---- | ------- | ---- | | [main](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ/tree/main) | 4 | 128 | No | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.26 GB | Yes | 4-bit, without Act Order and group size 128g. | | [gptq-4bit-32g-actorder_True](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ/tree/gptq-4bit-32g-actorder_True) | 4 | 32 | Yes | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 8.00 GB | Yes | 4-bit, with Act Order and group size 32g. Gives highest possible inference quality, with maximum VRAM usage. | | [gptq-4bit-64g-actorder_True](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ/tree/gptq-4bit-64g-actorder_True) | 4 | 64 | Yes | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.51 GB | Yes | 4-bit, with Act Order and group size 64g. Uses less VRAM than 32g, but with slightly lower accuracy. | | [gptq-4bit-128g-actorder_True](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ/tree/gptq-4bit-128g-actorder_True) | 4 | 128 | Yes | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 7.26 GB | Yes | 4-bit, with Act Order and group size 128g. Uses even less VRAM than 64g, but with slightly lower accuracy. | | [gptq-8bit--1g-actorder_True](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ/tree/gptq-8bit--1g-actorder_True) | 8 | None | Yes | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.36 GB | No | 8-bit, with Act Order. No group size, to lower VRAM requirements. | | [gptq-8bit-128g-actorder_True](https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ/tree/gptq-8bit-128g-actorder_True) | 8 | 128 | Yes | 0.1 | [wikitext](https://huggingface.co/datasets/wikitext/viewer/wikitext-2-v1/test) | 4096 | 13.65 GB | No | 8-bit, with group size 128g for higher inference quality and with Act Order for even higher accuracy. | <!-- README_GPTQ.md-provided-files end --> <!-- README_GPTQ.md-download-from-branches start --> ## How to download from branches - In text-generation-webui, you can add `:branch` to the end of the download name, eg `TheBloke/vicuna-13B-v1.5-GPTQ:main` - With Git, you can clone a branch with: ``` git clone --single-branch --branch main https://huggingface.co/TheBloke/vicuna-13B-v1.5-GPTQ ``` - In Python Transformers code, the branch is the `revision` parameter; see below. <!-- README_GPTQ.md-download-from-branches end --> <!-- README_GPTQ.md-text-generation-webui start --> ## How to easily download and use this model in [text-generation-webui](https://github.com/oobabooga/text-generation-webui). Please make sure you're using the latest version of [text-generation-webui](https://github.com/oobabooga/text-generation-webui). It is strongly recommended to use the text-generation-webui one-click-installers unless you're sure you know how to make a manual install. 1. Click the **Model tab**. 2. Under **Download custom model or LoRA**, enter `TheBloke/vicuna-13B-v1.5-GPTQ`. - To download from a specific branch, enter for example `TheBloke/vicuna-13B-v1.5-GPTQ:main` - see Provided Files above for the list of branches for each option. 3. Click **Download**. 4. The model will start downloading. Once it's finished it will say "Done". 5. In the top left, click the refresh icon next to **Model**. 6. In the **Model** dropdown, choose the model you just downloaded: `vicuna-13B-v1.5-GPTQ` 7. The model will automatically load, and is now ready for use! 8. If you want any custom settings, set them and then click **Save settings for this model** followed by **Reload the Model** in the top right. * Note that you do not need to and should not set manual GPTQ parameters any more. These are set automatically from the file `quantize_config.json`. 9. Once you're ready, click the **Text Generation tab** and enter a prompt to get started! <!-- README_GPTQ.md-text-generation-webui end --> <!-- README_GPTQ.md-use-from-python start --> ## How to use this GPTQ model from Python code ### Install the necessary packages Requires: Transformers 4.32.0 or later, Optimum 1.12.0 or later, and AutoGPTQ 0.4.2 or later. ```shell pip3 install transformers>=4.32.0 optimum>=1.12.0 pip3 install auto-gptq --extra-index-url https://huggingface.github.io/autogptq-index/whl/cu118/ # Use cu117 if on CUDA 11.7 ``` If you have problems installing AutoGPTQ using the pre-built wheels, install it from source instead: ```shell pip3 uninstall -y auto-gptq git clone https://github.com/PanQiWei/AutoGPTQ cd AutoGPTQ pip3 install . ``` ### For CodeLlama models only: you must use Transformers 4.33.0 or later. If 4.33.0 is not yet released when you read this, you will need to install Transformers from source: ```shell pip3 uninstall -y transformers pip3 install git+https://github.com/huggingface/transformers.git ``` ### You can then use the following code ```python from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline model_name_or_path = "TheBloke/vicuna-13B-v1.5-GPTQ" # To use a different branch, change revision # For example: revision="main" model = AutoModelForCausalLM.from_pretrained(model_name_or_path, device_map="auto", trust_remote_code=False, revision="main") tokenizer = AutoTokenizer.from_pretrained(model_name_or_path, use_fast=True) prompt = "Tell me about AI" prompt_template=f'''A chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions. USER: {prompt} ASSISTANT: ''' print("\n\n*** Generate:") input_ids = tokenizer(prompt_template, return_tensors='pt').input_ids.cuda() output = model.generate(inputs=input_ids, temperature=0.7, do_sample=True, top_p=0.95, top_k=40, max_new_tokens=512) print(tokenizer.decode(output[0])) # Inference can also be done using transformers' pipeline print("*** Pipeline:") pipe = pipeline( "text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512, do_sample=True, temperature=0.7, top_p=0.95, top_k=40, repetition_penalty=1.1 ) print(pipe(prompt_template)[0]['generated_text']) ``` <!-- README_GPTQ.md-use-from-python end --> <!-- README_GPTQ.md-compatibility start --> ## Compatibility The files provided are tested to work with AutoGPTQ, both via Transformers and using AutoGPTQ directly. They should also work with [Occ4m's GPTQ-for-LLaMa fork](https://github.com/0cc4m/KoboldAI). [ExLlama](https://github.com/turboderp/exllama) is compatible with Llama models in 4-bit. Please see the Provided Files table above for per-file compatibility. [Huggingface Text Generation Inference (TGI)](https://github.com/huggingface/text-generation-inference) is compatible with all GPTQ models. <!-- README_GPTQ.md-compatibility end --> <!-- footer start --> <!-- 200823 --> ## Discord For further support, and discussions on these models and AI in general, join us at: [TheBloke AI's Discord server](https://discord.gg/theblokeai) ## Thanks, and how to contribute Thanks to the [chirper.ai](https://chirper.ai) team! Thanks to Clay from [gpus.llm-utils.org](llm-utils)! I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training. If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects. Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits. * Patreon: https://patreon.com/TheBlokeAI * Ko-Fi: https://ko-fi.com/TheBlokeAI **Special thanks to**: Aemon Algiz. **Patreon special mentions**: Alicia Loh, Stephen Murray, K, Ajan Kanaga, RoA, Magnesian, Deo Leter, Olakabola, Eugene Pentland, zynix, Deep Realms, Raymond Fosdick, Elijah Stavena, Iucharbius, Erik Bjäreholt, Luis Javier Navarrete Lozano, Nicholas, theTransient, John Detwiler, alfie_i, knownsqashed, Mano Prime, Willem Michiel, Enrico Ros, LangChain4j, OG, Michael Dempsey, Pierre Kircher, Pedro Madruga, James Bentley, Thomas Belote, Luke @flexchar, Leonard Tan, Johann-Peter Hartmann, Illia Dulskyi, Fen Risland, Chadd, S_X, Jeff Scroggin, Ken Nordquist, Sean Connelly, Artur Olbinski, Swaroop Kallakuri, Jack West, Ai Maven, David Ziegler, Russ Johnson, transmissions 11, John Villwock, Alps Aficionado, Clay Pascal, Viktor Bowallius, Subspace Studios, Rainer Wilmers, Trenton Dambrowitz, vamX, Michael Levine, 준교 김, Brandon Frisco, Kalila, Trailburnt, Randy H, Talal Aujan, Nathan Dryer, Vadim, 阿明, ReadyPlayerEmma, Tiffany J. Kim, George Stoitzev, Spencer Kim, Jerry Meng, Gabriel Tamborski, Cory Kujawski, Jeffrey Morgan, Spiking Neurons AB, Edmond Seymore, Alexandros Triantafyllidis, Lone Striker, Cap'n Zoog, Nikolai Manek, danny, ya boyyy, Derek Yates, usrbinkat, Mandus, TL, Nathan LeClaire, subjectnull, Imad Khwaja, webtim, Raven Klaugh, Asp the Wyvern, Gabriel Puliatti, Caitlyn Gatomon, Joseph William Delisle, Jonathan Leane, Luke Pendergrass, SuperWojo, Sebastain Graf, Will Dee, Fred von Graf, Andrey, Dan Guido, Daniel P. Andersen, Nitin Borwankar, Elle, Vitor Caleffi, biorpg, jjj, NimbleBox.ai, Pieter, Matthew Berman, terasurfer, Michael Davis, Alex, Stanislav Ovsiannikov Thank you to all my generous patrons and donaters! And thank you again to a16z for their generous grant. <!-- footer end --> # Original model card: lmsys's Vicuna 13B v1.5 # Vicuna Model Card ## Model Details Vicuna is a chat assistant trained by fine-tuning Llama 2 on user-shared conversations collected from ShareGPT. - **Developed by:** [LMSYS](https://lmsys.org/) - **Model type:** An auto-regressive language model based on the transformer architecture - **License:** Llama 2 Community License Agreement - **Finetuned from model:** [Llama 2](https://arxiv.org/abs/2307.09288) ### Model Sources - **Repository:** https://github.com/lm-sys/FastChat - **Blog:** https://lmsys.org/blog/2023-03-30-vicuna/ - **Paper:** https://arxiv.org/abs/2306.05685 - **Demo:** https://chat.lmsys.org/ ## Uses The primary use of Vicuna is research on large language models and chatbots. The primary intended users of the model are researchers and hobbyists in natural language processing, machine learning, and artificial intelligence. ## How to Get Started with the Model - Command line interface: https://github.com/lm-sys/FastChat#vicuna-weights - APIs (OpenAI API, Huggingface API): https://github.com/lm-sys/FastChat/tree/main#api ## Training Details Vicuna v1.5 is fine-tuned from Llama 2 with supervised instruction fine-tuning. The training data is around 125K conversations collected from ShareGPT.com. See more details in the "Training Details of Vicuna Models" section in the appendix of this [paper](https://arxiv.org/pdf/2306.05685.pdf). ## Evaluation  Vicuna is evaluated with standard benchmarks, human preference, and LLM-as-a-judge. See more details in this [paper](https://arxiv.org/pdf/2306.05685.pdf) and [leaderboard](https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard). ## Difference between different versions of Vicuna See [vicuna_weights_version.md](https://github.com/lm-sys/FastChat/blob/main/docs/vicuna_weights_version.md)

KoboldAI/OPT-6B-nerys-v2

KoboldAI

transformers

text-generation

--- language: en license: other commercial: no --- # OPT 6B - Nerys ## Model Description OPT 6B-Nerys is a finetune created using Facebook's OPT model. ## Training data The training data contains around 2500 ebooks in various genres (the "Pike" dataset), a CYOA dataset called "CYS" and 50 Asian "Light Novels" (the "Manga-v1" dataset). Most parts of the dataset have been prepended using the following text: `[Genre: <genre1>, <genre2>]` This dataset has been cleaned in the same way as fairseq-dense-13B-Nerys-v2 ### How to use You can use this model directly with a pipeline for text generation. This example generates a different sequence each time it's run: ```py >>> from transformers import pipeline >>> generator = pipeline('text-generation', model='KoboldAI/OPT-6B-Nerys-v2') >>> generator("Welcome Captain Janeway, I apologize for the delay.", do_sample=True, min_length=50) [{'generated_text': 'Welcome Captain Janeway, I apologize for the delay."\nIt's all right," Janeway said. "I'm certain that you're doing your best to keep me informed of what\'s going on."'}] ``` ### Limitations and Biases Based on known problems with NLP technology, potential relevant factors include bias (gender, profession, race and religion). ### License OPT-6B is licensed under the OPT-175B license, Copyright (c) Meta Platforms, Inc. All Rights Reserved. ### BibTeX entry and citation info ``` @misc{zhang2022opt, title={OPT: Open Pre-trained Transformer Language Models}, author={Susan Zhang and Stephen Roller and Naman Goyal and Mikel Artetxe and Moya Chen and Shuohui Chen and Christopher Dewan and Mona Diab and Xian Li and Xi Victoria Lin and Todor Mihaylov and Myle Ott and Sam Shleifer and Kurt Shuster and Daniel Simig and Punit Singh Koura and Anjali Sridhar and Tianlu Wang and Luke Zettlemoyer}, year={2022}, eprint={2205.01068}, archivePrefix={arXiv}, primaryClass={cs.CL} } ```

facebook/sam-vit-huge

facebook

transformers

--- license: apache-2.0 --- # Model Card for Segment Anything Model (SAM) - ViT Huge (ViT-H) version <p> <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/sam-architecture.png" alt="Model architecture"> <em> Detailed architecture of Segment Anything Model (SAM).</em> </p> # Table of Contents 0. [TL;DR](#TL;DR) 1. [Model Details](#model-details) 2. [Usage](#usage) 3. [Citation](#citation) # TL;DR [Link to original repository](https://github.com/facebookresearch/segment-anything) | <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/sam-beancans.png" alt="Snow" width="600" height="600"> | <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/sam-dog-masks.png" alt="Forest" width="600" height="600"> | <img src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/transformers/model_doc/sam-car-seg.png" alt="Mountains" width="600" height="600"> | |---------------------------------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------| The **Segment Anything Model (SAM)** produces high quality object masks from input prompts such as points or boxes, and it can be used to generate masks for all objects in an image. It has been trained on a [dataset](https://segment-anything.com/dataset/index.html) of 11 million images and 1.1 billion masks, and has strong zero-shot performance on a variety of segmentation tasks. The abstract of the paper states: > We introduce the Segment Anything (SA) project: a new task, model, and dataset for image segmentation. Using our efficient model in a data collection loop, we built the largest segmentation dataset to date (by far), with over 1 billion masks on 11M licensed and privacy respecting images. The model is designed and trained to be promptable, so it can transfer zero-shot to new image distributions and tasks. We evaluate its capabilities on numerous tasks and find that its zero-shot performance is impressive -- often competitive with or even superior to prior fully supervised results. We are releasing the Segment Anything Model (SAM) and corresponding dataset (SA-1B) of 1B masks and 11M images at [https://segment-anything.com](https://segment-anything.com) to foster research into foundation models for computer vision. **Disclaimer**: Content from **this** model card has been written by the Hugging Face team, and parts of it were copy pasted from the original [SAM model card](https://github.com/facebookresearch/segment-anything). # Model Details The SAM model is made up of 3 modules: - The `VisionEncoder`: a VIT based image encoder. It computes the image embeddings using attention on patches of the image. Relative Positional Embedding is used. - The `PromptEncoder`: generates embeddings for points and bounding boxes - The `MaskDecoder`: a two-ways transformer which performs cross attention between the image embedding and the point embeddings (->) and between the point embeddings and the image embeddings. The outputs are fed - The `Neck`: predicts the output masks based on the contextualized masks produced by the `MaskDecoder`. # Usage ## Prompted-Mask-Generation ```python from PIL import Image import requests from transformers import SamModel, SamProcessor model = SamModel.from_pretrained("facebook/sam-vit-huge") processor = SamProcessor.from_pretrained("facebook/sam-vit-huge") img_url = "https://huggingface.co/ybelkada/segment-anything/resolve/main/assets/car.png" raw_image = Image.open(requests.get(img_url, stream=True).raw).convert("RGB") input_points = [[[450, 600]]] # 2D localization of a window ``` ```python inputs = processor(raw_image, input_points=input_points, return_tensors="pt").to("cuda") outputs = model(**inputs) masks = processor.image_processor.post_process_masks(outputs.pred_masks.cpu(), inputs["original_sizes"].cpu(), inputs["reshaped_input_sizes"].cpu()) scores = outputs.iou_scores ``` Among other arguments to generate masks, you can pass 2D locations on the approximate position of your object of interest, a bounding box wrapping the object of interest (the format should be x, y coordinate of the top right and bottom left point of the bounding box), a segmentation mask. At this time of writing, passing a text as input is not supported by the official model according to [the official repository](https://github.com/facebookresearch/segment-anything/issues/4#issuecomment-1497626844). For more details, refer to this notebook, which shows a walk throught of how to use the model, with a visual example! ## Automatic-Mask-Generation The model can be used for generating segmentation masks in a "zero-shot" fashion, given an input image. The model is automatically prompt with a grid of `1024` points which are all fed to the model. The pipeline is made for automatic mask generation. The following snippet demonstrates how easy you can run it (on any device! Simply feed the appropriate `points_per_batch` argument) ```python from transformers import pipeline generator = pipeline("mask-generation", device = 0, points_per_batch = 256) image_url = "https://huggingface.co/ybelkada/segment-anything/resolve/main/assets/car.png" outputs = generator(image_url, points_per_batch = 256) ``` Now to display the image: ```python import matplotlib.pyplot as plt from PIL import Image import numpy as np def show_mask(mask, ax, random_color=False): if random_color: color = np.concatenate([np.random.random(3), np.array([0.6])], axis=0) else: color = np.array([30 / 255, 144 / 255, 255 / 255, 0.6]) h, w = mask.shape[-2:] mask_image = mask.reshape(h, w, 1) * color.reshape(1, 1, -1) ax.imshow(mask_image) plt.imshow(np.array(raw_image)) ax = plt.gca() for mask in outputs["masks"]: show_mask(mask, ax=ax, random_color=True) plt.axis("off") plt.show() ``` This should give you the following  # Citation If you use this model, please use the following BibTeX entry. ``` @article{kirillov2023segany, title={Segment Anything}, author={Kirillov, Alexander and Mintun, Eric and Ravi, Nikhila and Mao, Hanzi and Rolland, Chloe and Gustafson, Laura and Xiao, Tete and Whitehead, Spencer and Berg, Alexander C. and Lo, Wan-Yen and Doll{\'a}r, Piotr and Girshick, Ross}, journal={arXiv:2304.02643}, year={2023} } ```

timm/mobilenetv3_large_100.miil_in21k_ft_in1k

timm

timm

image-classification

--- tags: - image-classification - timm library_name: timm license: apache-2.0 datasets: - imagenet-1k - imagenet-21k-p --- # Model card for mobilenetv3_large_100.miil_in21k_ft_in1k A MobileNet-v3 image classification model. Petrained on ImageNet-21k-P and fine-tuned on ImageNet-1k by Alibaba MIIL. ## Model Details - **Model Type:** Image classification / feature backbone - **Model Stats:** - Params (M): 5.5 - GMACs: 0.2 - Activations (M): 4.4 - Image size: 224 x 224 - **Papers:** - Searching for MobileNetV3: https://arxiv.org/abs/1905.02244 - **Dataset:** ImageNet-1k - **Pretrain Dataset:** ImageNet-21k-P ## Model Usage ### Image Classification ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model('mobilenetv3_large_100.miil_in21k_ft_in1k', pretrained=True) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 top5_probabilities, top5_class_indices = torch.topk(output.softmax(dim=1) * 100, k=5) ``` ### Feature Map Extraction ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilenetv3_large_100.miil_in21k_ft_in1k', pretrained=True, features_only=True, ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # unsqueeze single image into batch of 1 for o in output: # print shape of each feature map in output # e.g.: # torch.Size([1, 16, 112, 112]) # torch.Size([1, 24, 56, 56]) # torch.Size([1, 40, 28, 28]) # torch.Size([1, 112, 14, 14]) # torch.Size([1, 960, 7, 7]) print(o.shape) ``` ### Image Embeddings ```python from urllib.request import urlopen from PIL import Image import timm img = Image.open(urlopen( 'https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/beignets-task-guide.png' )) model = timm.create_model( 'mobilenetv3_large_100.miil_in21k_ft_in1k', pretrained=True, num_classes=0, # remove classifier nn.Linear ) model = model.eval() # get model specific transforms (normalization, resize) data_config = timm.data.resolve_model_data_config(model) transforms = timm.data.create_transform(**data_config, is_training=False) output = model(transforms(img).unsqueeze(0)) # output is (batch_size, num_features) shaped tensor # or equivalently (without needing to set num_classes=0) output = model.forward_features(transforms(img).unsqueeze(0)) # output is unpooled, a (1, 960, 7, 7) shaped tensor output = model.forward_head(output, pre_logits=True) # output is a (1, num_features) shaped tensor ``` ## Model Comparison Explore the dataset and runtime metrics of this model in timm [model results](https://github.com/huggingface/pytorch-image-models/tree/main/results). ## Citation ```bibtex @inproceedings{howard2019searching, title={Searching for mobilenetv3}, author={Howard, Andrew and Sandler, Mark and Chu, Grace and Chen, Liang-Chieh and Chen, Bo and Tan, Mingxing and Wang, Weijun and Zhu, Yukun and Pang, Ruoming and Vasudevan, Vijay and others}, booktitle={Proceedings of the IEEE/CVF international conference on computer vision}, pages={1314--1324}, year={2019} } ``` ```bibtex @misc{rw2019timm, author = {Ross Wightman}, title = {PyTorch Image Models}, year = {2019}, publisher = {GitHub}, journal = {GitHub repository}, doi = {10.5281/zenodo.4414861}, howpublished = {\url{https://github.com/huggingface/pytorch-image-models}} } ```

Helsinki-NLP/opus-mt-ine-en

Helsinki-NLP

transformers

translation

--- language: - ca - es - os - ro - fy - cy - sc - is - yi - lb - an - sq - fr - ht - rm - ps - af - uk - sl - lt - bg - be - gd - si - en - br - mk - or - mr - ru - fo - co - oc - pl - gl - nb - bn - id - hy - da - gv - nl - pt - hi - as - kw - ga - sv - gu - wa - lv - el - it - hr - ur - nn - de - cs - ine tags: - translation license: apache-2.0 --- ### ine-eng * source group: Indo-European languages * target group: English * OPUS readme: [ine-eng](https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ine-eng/README.md) * model: transformer * source language(s): afr aln ang_Latn arg asm ast awa bel bel_Latn ben bho bos_Latn bre bul bul_Latn cat ces cor cos csb_Latn cym dan deu dsb egl ell enm_Latn ext fao fra frm_Latn frr fry gcf_Latn gla gle glg glv gom gos got_Goth grc_Grek gsw guj hat hif_Latn hin hrv hsb hye ind isl ita jdt_Cyrl ksh kur_Arab kur_Latn lad lad_Latn lat_Latn lav lij lit lld_Latn lmo ltg ltz mai mar max_Latn mfe min mkd mwl nds nld nno nob nob_Hebr non_Latn npi oci ori orv_Cyrl oss pan_Guru pap pdc pes pes_Latn pes_Thaa pms pnb pol por prg_Latn pus roh rom ron rue rus san_Deva scn sco sgs sin slv snd_Arab spa sqi srp_Cyrl srp_Latn stq swe swg tgk_Cyrl tly_Latn tmw_Latn ukr urd vec wln yid zlm_Latn zsm_Latn zza * target language(s): eng * model: transformer * pre-processing: normalization + SentencePiece (spm32k,spm32k) * download original weights: [opus2m-2020-08-01.zip](https://object.pouta.csc.fi/Tatoeba-MT-models/ine-eng/opus2m-2020-08-01.zip) * test set translations: [opus2m-2020-08-01.test.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ine-eng/opus2m-2020-08-01.test.txt) * test set scores: [opus2m-2020-08-01.eval.txt](https://object.pouta.csc.fi/Tatoeba-MT-models/ine-eng/opus2m-2020-08-01.eval.txt) ## Benchmarks | testset | BLEU | chr-F | |-----------------------|-------|-------| | newsdev2014-hineng.hin.eng | 11.2 | 0.375 | | newsdev2016-enro-roneng.ron.eng | 35.5 | 0.614 | | newsdev2017-enlv-laveng.lav.eng | 25.1 | 0.542 | | newsdev2019-engu-gujeng.guj.eng | 16.0 | 0.420 | | newsdev2019-enlt-liteng.lit.eng | 24.0 | 0.522 | | newsdiscussdev2015-enfr-fraeng.fra.eng | 30.1 | 0.550 | | newsdiscusstest2015-enfr-fraeng.fra.eng | 33.4 | 0.572 | | newssyscomb2009-ceseng.ces.eng | 24.0 | 0.520 | | newssyscomb2009-deueng.deu.eng | 25.7 | 0.526 | | newssyscomb2009-fraeng.fra.eng | 27.9 | 0.550 | | newssyscomb2009-itaeng.ita.eng | 31.4 | 0.574 | | newssyscomb2009-spaeng.spa.eng | 28.3 | 0.555 | | news-test2008-deueng.deu.eng | 24.0 | 0.515 | | news-test2008-fraeng.fra.eng | 24.5 | 0.524 | | news-test2008-spaeng.spa.eng | 25.5 | 0.533 | | newstest2009-ceseng.ces.eng | 23.3 | 0.516 | | newstest2009-deueng.deu.eng | 23.2 | 0.512 | | newstest2009-fraeng.fra.eng | 27.3 | 0.545 | | newstest2009-itaeng.ita.eng | 30.3 | 0.567 | | newstest2009-spaeng.spa.eng | 27.9 | 0.549 | | newstest2010-ceseng.ces.eng | 23.8 | 0.523 | | newstest2010-deueng.deu.eng | 26.2 | 0.545 | | newstest2010-fraeng.fra.eng | 28.6 | 0.562 | | newstest2010-spaeng.spa.eng | 31.4 | 0.581 | | newstest2011-ceseng.ces.eng | 24.2 | 0.521 | | newstest2011-deueng.deu.eng | 23.9 | 0.522 | | newstest2011-fraeng.fra.eng | 29.5 | 0.570 | | newstest2011-spaeng.spa.eng | 30.3 | 0.570 | | newstest2012-ceseng.ces.eng | 23.5 | 0.516 | | newstest2012-deueng.deu.eng | 24.9 | 0.529 | | newstest2012-fraeng.fra.eng | 30.0 | 0.568 | | newstest2012-ruseng.rus.eng | 29.9 | 0.565 | | newstest2012-spaeng.spa.eng | 33.3 | 0.593 | | newstest2013-ceseng.ces.eng | 25.6 | 0.531 | | newstest2013-deueng.deu.eng | 27.7 | 0.545 | | newstest2013-fraeng.fra.eng | 30.0 | 0.561 | | newstest2013-ruseng.rus.eng | 24.4 | 0.514 | | newstest2013-spaeng.spa.eng | 30.8 | 0.577 | | newstest2014-csen-ceseng.ces.eng | 27.7 | 0.558 | | newstest2014-deen-deueng.deu.eng | 27.7 | 0.545 | | newstest2014-fren-fraeng.fra.eng | 32.2 | 0.592 | | newstest2014-hien-hineng.hin.eng | 16.7 | 0.450 | | newstest2014-ruen-ruseng.rus.eng | 27.2 | 0.552 | | newstest2015-encs-ceseng.ces.eng | 25.4 | 0.518 | | newstest2015-ende-deueng.deu.eng | 28.8 | 0.552 | | newstest2015-enru-ruseng.rus.eng | 25.6 | 0.527 | | newstest2016-encs-ceseng.ces.eng | 27.0 | 0.540 | | newstest2016-ende-deueng.deu.eng | 33.5 | 0.592 | | newstest2016-enro-roneng.ron.eng | 32.8 | 0.591 | | newstest2016-enru-ruseng.rus.eng | 24.8 | 0.523 | | newstest2017-encs-ceseng.ces.eng | 23.7 | 0.510 | | newstest2017-ende-deueng.deu.eng | 29.3 | 0.556 | | newstest2017-enlv-laveng.lav.eng | 18.9 | 0.486 | | newstest2017-enru-ruseng.rus.eng | 28.0 | 0.546 | | newstest2018-encs-ceseng.ces.eng | 24.9 | 0.521 | | newstest2018-ende-deueng.deu.eng | 36.0 | 0.604 | | newstest2018-enru-ruseng.rus.eng | 23.8 | 0.517 | | newstest2019-deen-deueng.deu.eng | 31.5 | 0.570 | | newstest2019-guen-gujeng.guj.eng | 12.1 | 0.377 | | newstest2019-lten-liteng.lit.eng | 26.6 | 0.555 | | newstest2019-ruen-ruseng.rus.eng | 27.5 | 0.541 | | Tatoeba-test.afr-eng.afr.eng | 59.0 | 0.724 | | Tatoeba-test.ang-eng.ang.eng | 9.9 | 0.254 | | Tatoeba-test.arg-eng.arg.eng | 41.6 | 0.487 | | Tatoeba-test.asm-eng.asm.eng | 22.8 | 0.392 | | Tatoeba-test.ast-eng.ast.eng | 36.1 | 0.521 | | Tatoeba-test.awa-eng.awa.eng | 11.6 | 0.280 | | Tatoeba-test.bel-eng.bel.eng | 42.2 | 0.597 | | Tatoeba-test.ben-eng.ben.eng | 45.8 | 0.598 | | Tatoeba-test.bho-eng.bho.eng | 34.4 | 0.518 | | Tatoeba-test.bre-eng.bre.eng | 24.4 | 0.405 | | Tatoeba-test.bul-eng.bul.eng | 50.8 | 0.660 | | Tatoeba-test.cat-eng.cat.eng | 51.2 | 0.677 | | Tatoeba-test.ces-eng.ces.eng | 47.6 | 0.641 | | Tatoeba-test.cor-eng.cor.eng | 5.4 | 0.214 | | Tatoeba-test.cos-eng.cos.eng | 61.0 | 0.675 | | Tatoeba-test.csb-eng.csb.eng | 22.5 | 0.394 | | Tatoeba-test.cym-eng.cym.eng | 34.7 | 0.522 | | Tatoeba-test.dan-eng.dan.eng | 56.2 | 0.708 | | Tatoeba-test.deu-eng.deu.eng | 44.9 | 0.625 | | Tatoeba-test.dsb-eng.dsb.eng | 21.0 | 0.383 | | Tatoeba-test.egl-eng.egl.eng | 6.9 | 0.221 | | Tatoeba-test.ell-eng.ell.eng | 62.1 | 0.741 | | Tatoeba-test.enm-eng.enm.eng | 22.6 | 0.466 | | Tatoeba-test.ext-eng.ext.eng | 33.2 | 0.496 | | Tatoeba-test.fao-eng.fao.eng | 28.1 | 0.460 | | Tatoeba-test.fas-eng.fas.eng | 9.6 | 0.306 | | Tatoeba-test.fra-eng.fra.eng | 50.3 | 0.661 | | Tatoeba-test.frm-eng.frm.eng | 30.0 | 0.457 | | Tatoeba-test.frr-eng.frr.eng | 15.2 | 0.301 | | Tatoeba-test.fry-eng.fry.eng | 34.4 | 0.525 | | Tatoeba-test.gcf-eng.gcf.eng | 18.4 | 0.317 | | Tatoeba-test.gla-eng.gla.eng | 24.1 | 0.400 | | Tatoeba-test.gle-eng.gle.eng | 52.2 | 0.671 | | Tatoeba-test.glg-eng.glg.eng | 50.5 | 0.669 | | Tatoeba-test.glv-eng.glv.eng | 5.7 | 0.189 | | Tatoeba-test.gos-eng.gos.eng | 19.2 | 0.378 | | Tatoeba-test.got-eng.got.eng | 0.1 | 0.022 | | Tatoeba-test.grc-eng.grc.eng | 0.9 | 0.095 | | Tatoeba-test.gsw-eng.gsw.eng | 23.9 | 0.390 | | Tatoeba-test.guj-eng.guj.eng | 28.0 | 0.428 | | Tatoeba-test.hat-eng.hat.eng | 44.2 | 0.567 | | Tatoeba-test.hbs-eng.hbs.eng | 51.6 | 0.666 | | Tatoeba-test.hif-eng.hif.eng | 22.3 | 0.451 | | Tatoeba-test.hin-eng.hin.eng | 41.7 | 0.585 | | Tatoeba-test.hsb-eng.hsb.eng | 46.4 | 0.590 | | Tatoeba-test.hye-eng.hye.eng | 40.4 | 0.564 | | Tatoeba-test.isl-eng.isl.eng | 43.8 | 0.605 | | Tatoeba-test.ita-eng.ita.eng | 60.7 | 0.735 | | Tatoeba-test.jdt-eng.jdt.eng | 5.5 | 0.091 | | Tatoeba-test.kok-eng.kok.eng | 7.8 | 0.205 | | Tatoeba-test.ksh-eng.ksh.eng | 15.8 | 0.284 | | Tatoeba-test.kur-eng.kur.eng | 11.6 | 0.232 | | Tatoeba-test.lad-eng.lad.eng | 30.7 | 0.484 | | Tatoeba-test.lah-eng.lah.eng | 11.0 | 0.286 | | Tatoeba-test.lat-eng.lat.eng | 24.4 | 0.432 | | Tatoeba-test.lav-eng.lav.eng | 47.2 | 0.646 | | Tatoeba-test.lij-eng.lij.eng | 9.0 | 0.287 | | Tatoeba-test.lit-eng.lit.eng | 51.7 | 0.670 | | Tatoeba-test.lld-eng.lld.eng | 22.4 | 0.369 | | Tatoeba-test.lmo-eng.lmo.eng | 26.1 | 0.381 | | Tatoeba-test.ltz-eng.ltz.eng | 39.8 | 0.536 | | Tatoeba-test.mai-eng.mai.eng | 72.3 | 0.758 | | Tatoeba-test.mar-eng.mar.eng | 32.0 | 0.554 | | Tatoeba-test.mfe-eng.mfe.eng | 63.1 | 0.822 | | Tatoeba-test.mkd-eng.mkd.eng | 49.5 | 0.638 | | Tatoeba-test.msa-eng.msa.eng | 38.6 | 0.566 | | Tatoeba-test.multi.eng | 45.6 | 0.615 | | Tatoeba-test.mwl-eng.mwl.eng | 40.4 | 0.767 | | Tatoeba-test.nds-eng.nds.eng | 35.5 | 0.538 | | Tatoeba-test.nep-eng.nep.eng | 4.9 | 0.209 | | Tatoeba-test.nld-eng.nld.eng | 54.2 | 0.694 | | Tatoeba-test.non-eng.non.eng | 39.3 | 0.573 | | Tatoeba-test.nor-eng.nor.eng | 50.9 | 0.663 | | Tatoeba-test.oci-eng.oci.eng | 19.6 | 0.386 | | Tatoeba-test.ori-eng.ori.eng | 16.2 | 0.364 | | Tatoeba-test.orv-eng.orv.eng | 13.6 | 0.288 | | Tatoeba-test.oss-eng.oss.eng | 9.4 | 0.301 | | Tatoeba-test.pan-eng.pan.eng | 17.1 | 0.389 | | Tatoeba-test.pap-eng.pap.eng | 57.0 | 0.680 | | Tatoeba-test.pdc-eng.pdc.eng | 41.6 | 0.526 | | Tatoeba-test.pms-eng.pms.eng | 13.7 | 0.333 | | Tatoeba-test.pol-eng.pol.eng | 46.5 | 0.632 | | Tatoeba-test.por-eng.por.eng | 56.4 | 0.710 | | Tatoeba-test.prg-eng.prg.eng | 2.3 | 0.193 | | Tatoeba-test.pus-eng.pus.eng | 3.2 | 0.194 | | Tatoeba-test.roh-eng.roh.eng | 17.5 | 0.420 | | Tatoeba-test.rom-eng.rom.eng | 5.0 | 0.237 | | Tatoeba-test.ron-eng.ron.eng | 51.4 | 0.670 | | Tatoeba-test.rue-eng.rue.eng | 26.0 | 0.447 | | Tatoeba-test.rus-eng.rus.eng | 47.8 | 0.634 | | Tatoeba-test.san-eng.san.eng | 4.0 | 0.195 | | Tatoeba-test.scn-eng.scn.eng | 45.1 | 0.440 | | Tatoeba-test.sco-eng.sco.eng | 41.9 | 0.582 | | Tatoeba-test.sgs-eng.sgs.eng | 38.7 | 0.498 | | Tatoeba-test.sin-eng.sin.eng | 29.7 | 0.499 | | Tatoeba-test.slv-eng.slv.eng | 38.2 | 0.564 | | Tatoeba-test.snd-eng.snd.eng | 12.7 | 0.342 | | Tatoeba-test.spa-eng.spa.eng | 53.2 | 0.687 | | Tatoeba-test.sqi-eng.sqi.eng | 51.9 | 0.679 | | Tatoeba-test.stq-eng.stq.eng | 9.0 | 0.391 | | Tatoeba-test.swe-eng.swe.eng | 57.4 | 0.705 | | Tatoeba-test.swg-eng.swg.eng | 18.0 | 0.338 | | Tatoeba-test.tgk-eng.tgk.eng | 24.3 | 0.413 | | Tatoeba-test.tly-eng.tly.eng | 1.1 | 0.094 | | Tatoeba-test.ukr-eng.ukr.eng | 48.0 | 0.639 | | Tatoeba-test.urd-eng.urd.eng | 27.2 | 0.471 | | Tatoeba-test.vec-eng.vec.eng | 28.0 | 0.398 | | Tatoeba-test.wln-eng.wln.eng | 17.5 | 0.320 | | Tatoeba-test.yid-eng.yid.eng | 26.9 | 0.457 | | Tatoeba-test.zza-eng.zza.eng | 1.7 | 0.131 | ### System Info: - hf_name: ine-eng - source_languages: ine - target_languages: eng - opus_readme_url: https://github.com/Helsinki-NLP/Tatoeba-Challenge/tree/master/models/ine-eng/README.md - original_repo: Tatoeba-Challenge - tags: ['translation'] - languages: ['ca', 'es', 'os', 'ro', 'fy', 'cy', 'sc', 'is', 'yi', 'lb', 'an', 'sq', 'fr', 'ht', 'rm', 'ps', 'af', 'uk', 'sl', 'lt', 'bg', 'be', 'gd', 'si', 'en', 'br', 'mk', 'or', 'mr', 'ru', 'fo', 'co', 'oc', 'pl', 'gl', 'nb', 'bn', 'id', 'hy', 'da', 'gv', 'nl', 'pt', 'hi', 'as', 'kw', 'ga', 'sv', 'gu', 'wa', 'lv', 'el', 'it', 'hr', 'ur', 'nn', 'de', 'cs', 'ine'] - src_constituents: {'cat', 'spa', 'pap', 'mwl', 'lij', 'bos_Latn', 'lad_Latn', 'lat_Latn', 'pcd', 'oss', 'ron', 'fry', 'cym', 'awa', 'swg', 'zsm_Latn', 'srd', 'gcf_Latn', 'isl', 'yid', 'bho', 'ltz', 'kur_Latn', 'arg', 'pes_Thaa', 'sqi', 'csb_Latn', 'fra', 'hat', 'non_Latn', 'sco', 'pnb', 'roh', 'bul_Latn', 'pus', 'afr', 'ukr', 'slv', 'lit', 'tmw_Latn', 'hsb', 'tly_Latn', 'bul', 'bel', 'got_Goth', 'lat_Grek', 'ext', 'gla', 'mai', 'sin', 'hif_Latn', 'eng', 'bre', 'nob_Hebr', 'prg_Latn', 'ang_Latn', 'aln', 'mkd', 'ori', 'mar', 'afr_Arab', 'san_Deva', 'gos', 'rus', 'fao', 'orv_Cyrl', 'bel_Latn', 'cos', 'zza', 'grc_Grek', 'oci', 'mfe', 'gom', 'bjn', 'sgs', 'tgk_Cyrl', 'hye_Latn', 'pdc', 'srp_Cyrl', 'pol', 'ast', 'glg', 'pms', 'nob', 'ben', 'min', 'srp_Latn', 'zlm_Latn', 'ind', 'rom', 'hye', 'scn', 'enm_Latn', 'lmo', 'npi', 'pes', 'dan', 'rus_Latn', 'jdt_Cyrl', 'gsw', 'glv', 'nld', 'snd_Arab', 'kur_Arab', 'por', 'hin', 'dsb', 'asm', 'lad', 'frm_Latn', 'ksh', 'pan_Guru', 'cor', 'gle', 'swe', 'guj', 'wln', 'lav', 'ell', 'frr', 'rue', 'ita', 'hrv', 'urd', 'stq', 'nno', 'deu', 'lld_Latn', 'ces', 'egl', 'vec', 'max_Latn', 'pes_Latn', 'ltg', 'nds'} - tgt_constituents: {'eng'} - src_multilingual: True - tgt_multilingual: False - prepro: normalization + SentencePiece (spm32k,spm32k) - url_model: https://object.pouta.csc.fi/Tatoeba-MT-models/ine-eng/opus2m-2020-08-01.zip - url_test_set: https://object.pouta.csc.fi/Tatoeba-MT-models/ine-eng/opus2m-2020-08-01.test.txt - src_alpha3: ine - tgt_alpha3: eng - short_pair: ine-en - chrF2_score: 0.615 - bleu: 45.6 - brevity_penalty: 0.997 - ref_len: 71872.0 - src_name: Indo-European languages - tgt_name: English - train_date: 2020-08-01 - src_alpha2: ine - tgt_alpha2: en - prefer_old: False - long_pair: ine-eng - helsinki_git_sha: 480fcbe0ee1bf4774bcbe6226ad9f58e63f6c535 - transformers_git_sha: 2207e5d8cb224e954a7cba69fa4ac2309e9ff30b - port_machine: brutasse - port_time: 2020-08-21-14:41

mosaicml/mpt-7b-chat

mosaicml

transformers

text-generation

--- license: cc-by-nc-sa-4.0 datasets: - jeffwan/sharegpt_vicuna - Hello-SimpleAI/HC3 - tatsu-lab/alpaca - Anthropic/hh-rlhf - victor123/evol_instruct_70k tags: - Composer - MosaicML - llm-foundry inference: false --- # MPT-7B-Chat MPT-7B-Chat is a chatbot-like model for dialogue generation. It was built by finetuning [MPT-7B](https://huggingface.co/mosaicml/mpt-7b) on the [ShareGPT-Vicuna](https://huggingface.co/datasets/jeffwan/sharegpt_vicuna), [HC3](https://huggingface.co/datasets/Hello-SimpleAI/HC3), [Alpaca](https://huggingface.co/datasets/tatsu-lab/alpaca), [HH-RLHF](https://huggingface.co/datasets/Anthropic/hh-rlhf), and [Evol-Instruct](https://huggingface.co/datasets/victor123/evol_instruct_70k) datasets. * License: _CC-By-NC-SA-4.0_ (non-commercial use only) * [Demo on Hugging Face Spaces](https://huggingface.co/spaces/mosaicml/mpt-7b-chat) This model was trained by [MosaicML](https://www.mosaicml.com) and follows a modified decoder-only transformer architecture. ## Model Date May 5, 2023 ## Model License _CC-By-NC-SA-4.0_ (non-commercial use only) ## Documentation * [Blog post: Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs](https://www.mosaicml.com/blog/mpt-7b) * [Codebase (mosaicml/llm-foundry repo)](https://github.com/mosaicml/llm-foundry/) * Questions: Feel free to contact us via the [MosaicML Community Slack](https://mosaicml.me/slack)! ### Example Dialogue **SamIAm85**: >I want you to come up with a tweet based on this summary of the article: >"Introducing MPT-7B, the latest entry in our MosaicML Foundation Series. >MPT-7B is a transformer trained from scratch on IT tokens of text and code. >It is open source and it matches the quality of LLaMA-7B. >MPT-7B was trained on the MosaicML platform in 9.5 days with zero human intervention at a cost of ~$200k. >Starting today, you can train, finetune, and deploy your own private MPT models, either starting from one of our checkpoints or training from scratch. >For inspiration, we are also releasing three finetuned models in addition to the base MPT-7B: MPT-7B-Instruct, MPT-7B-Chat, and MPT-7B-Story Writer-65k+, the last of which uses a context length of 65k tokens! **MPT-7B-Instruct**: >Here's my suggestion for a tweet: >"Exciting news! @MosaicML has launched #MPT_7B - a new standard for open-source LLMs. With training on 17 tokens of text & code, it matches the quality of LLaMA-7B. Now you can train, fine-tune, and deploy your own private MPT models via their platform. Check out the full blog post here: https://bit.ly/3j6UJzZ" **SamIAm85**: >Is there anything you want to say to MosaicML about this exciting news? **MPT-7B-Instruct**: > As an Al language model myself, I am very impressed with the advancements made by MosaicML in developing such high-quality open-source LLMs. > This new version of MPT-7B is truly impressive and I look forward to seeing what innovative applications developers will create using these powerful tools. > Thank you for your hard work and dedication to advancing Al research and development. ## How to Use This model is best used with the MosaicML [llm-foundry repository](https://github.com/mosaicml/llm-foundry) for training and finetuning. ```python import transformers model = transformers.AutoModelForCausalLM.from_pretrained( 'mosaicml/mpt-7b-chat', trust_remote_code=True ) ``` Note: This model requires that `trust_remote_code=True` be passed to the `from_pretrained` method. This is because we use a custom `MPT` model architecture that is not yet part of the Hugging Face `transformers` package. `MPT` includes options for many training efficiency features such as [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf), [ALiBi](https://arxiv.org/abs/2108.12409), [QK LayerNorm](https://arxiv.org/abs/2010.04245), and more. To use the optimized [triton implementation](https://github.com/openai/triton) of FlashAttention, you can load the model on GPU (`cuda:0`) with `attn_impl='triton'` and with `bfloat16` precision: ```python import torch import transformers name = 'mosaicml/mpt-7b-chat' config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True) config.attn_config['attn_impl'] = 'triton' config.init_device = 'cuda:0' # For fast initialization directly on GPU! model = transformers.AutoModelForCausalLM.from_pretrained( name, config=config, torch_dtype=torch.bfloat16, # Load model weights in bfloat16 trust_remote_code=True ) ``` Although the model was trained with a sequence length of 2048, ALiBi enables users to increase the maximum sequence length during finetuning and/or inference. For example: ```python import transformers name = 'mosaicml/mpt-7b-chat' config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True) config.max_seq_len = 4096 # (input + output) tokens can now be up to 4096 model = transformers.AutoModelForCausalLM.from_pretrained( name, config=config, trust_remote_code=True ) ``` This model was trained with the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer. ```python from transformers import AutoTokenizer tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neox-20b") ``` The model can then be used, for example, within a text-generation pipeline. Note: when running Torch modules in lower precision, it is best practice to use the [torch.autocast context manager](https://pytorch.org/docs/stable/amp.html). ```python from transformers import pipeline pipe = pipeline('text-generation', model=model, tokenizer=tokenizer, device='cuda:0') with torch.autocast('cuda', dtype=torch.bfloat16): print( pipe('Here is a recipe for vegan banana bread:\n', max_new_tokens=100, do_sample=True, use_cache=True)) ``` ## Model Description The architecture is a modification of a standard decoder-only transformer. The model has been modified from a standard transformer in the following ways: * It uses [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf) * It uses [ALiBi (Attention with Linear Biases)](https://arxiv.org/abs/2108.12409) and does not use positional embeddings * It does not use biases | Hyperparameter | Value | |----------------|-------| |n_parameters | 6.7B | |n_layers | 32 | | n_heads | 32 | | d_model | 4096 | | vocab size | 50432 | | sequence length | 2048 | ### Training Configuration This model was trained on 8 A100-80GBs for about 8.2 hours, followed by training for 6.7 hours on 32 A100-40GBs using the [MosaicML Platform](https://www.mosaicml.com/platform). The model was trained with sharded data parallelism using [FSDP](https://pytorch.org/docs/stable/fsdp.html) and used the AdamW optimizer. ## Limitations and Biases _The following language is modified from [EleutherAI's GPT-NeoX-20B](https://huggingface.co/EleutherAI/gpt-neox-20b)_ MPT-7B-Chat can produce factually incorrect output, and should not be relied on to produce factually accurate information. MPT-7B-Chat was trained on various public datasets. While great efforts have been taken to clean the pretraining data, it is possible that this model could generate lewd, biased or otherwise offensive outputs. ## Acknowledgements This model was finetuned by Sam Havens and the MosaicML NLP team ## Disclaimer The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please cosult an attorney before using this model for commercial purposes. ## MosaicML Platform If you're interested in [training](https://www.mosaicml.com/training) and [deploying](https://www.mosaicml.com/inference) your own MPT or LLMs on the MosaicML Platform, [sign up here](https://forms.mosaicml.com/demo?utm_source=huggingface&utm_medium=referral&utm_campaign=mpt-7b). ## Citation Please cite this model using the following format: ``` @online{MosaicML2023Introducing, author = {MosaicML NLP Team}, title = {Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs}, year = {2023}, url = {www.mosaicml.com/blog/mpt-7b}, note = {Accessed: 2023-03-28}, % change this date urldate = {2023-03-28} % change this date } ```

Yntec/fennPhoto

Yntec

diffusers

text-to-image

--- language: - en license: creativeml-openrail-m library_name: diffusers pipeline_tag: text-to-image tags: - Base Model - Film - Real - stable-diffusion - stable-diffusion-diffusers - diffusers - text-to-image inference: true --- # Fenn Photo Original page: https://civitai.com/models/153869/fenn-photo Samples and prompt:   macro studio photo of old antique Victorian marmor figurine of cute Rinoa, chibi Rinoa Heartilly, eyeliner, very slim, arm warmers, necklace, sleeveless coat, black shirt, blue skirt, arm ribbon, bike shorts, boots, long hair, black hair, cozy home in the background with candles and plants, Rinoa on flat round porcelain base, by Michelangelo

google/bigbird-pegasus-large-arxiv

google

transformers

summarization

--- language: en license: apache-2.0 datasets: - scientific_papers tags: - summarization model-index: - name: google/bigbird-pegasus-large-arxiv results: - task: type: summarization name: Summarization dataset: name: scientific_papers type: scientific_papers config: pubmed split: test metrics: - name: ROUGE-1 type: rouge value: 36.0276 verified: true - name: ROUGE-2 type: rouge value: 13.4166 verified: true - name: ROUGE-L type: rouge value: 21.9612 verified: true - name: ROUGE-LSUM type: rouge value: 29.648 verified: true - name: loss type: loss value: 2.774355173110962 verified: true - name: meteor type: meteor value: 0.2824 verified: true - name: gen_len type: gen_len value: 209.2537 verified: true - task: type: summarization name: Summarization dataset: name: cnn_dailymail type: cnn_dailymail config: 3.0.0 split: test metrics: - name: ROUGE-1 type: rouge value: 9.0885 verified: true - name: ROUGE-2 type: rouge value: 1.0325 verified: true - name: ROUGE-L type: rouge value: 7.3182 verified: true - name: ROUGE-LSUM type: rouge value: 8.1455 verified: true - name: loss type: loss value: .nan verified: true - name: gen_len type: gen_len value: 210.4762 verified: true - task: type: summarization name: Summarization dataset: name: xsum type: xsum config: default split: test metrics: - name: ROUGE-1 type: rouge value: 4.9787 verified: true - name: ROUGE-2 type: rouge value: 0.3527 verified: true - name: ROUGE-L type: rouge value: 4.3679 verified: true - name: ROUGE-LSUM type: rouge value: 4.1723 verified: true - name: loss type: loss value: .nan verified: true - name: gen_len type: gen_len value: 230.4886 verified: true - task: type: summarization name: Summarization dataset: name: scientific_papers type: scientific_papers config: arxiv split: test metrics: - name: ROUGE-1 type: rouge value: 43.4702 verified: true - name: ROUGE-2 type: rouge value: 17.4297 verified: true - name: ROUGE-L type: rouge value: 26.2587 verified: true - name: ROUGE-LSUM type: rouge value: 35.5587 verified: true - name: loss type: loss value: 2.1113228797912598 verified: true - name: gen_len type: gen_len value: 183.3702 verified: true - task: type: summarization name: Summarization dataset: name: samsum type: samsum config: samsum split: test metrics: - name: ROUGE-1 type: rouge value: 3.621 verified: true - name: ROUGE-2 type: rouge value: 0.1699 verified: true - name: ROUGE-L type: rouge value: 3.2016 verified: true - name: ROUGE-LSUM type: rouge value: 3.3269 verified: true - name: loss type: loss value: 7.664482116699219 verified: true - name: gen_len type: gen_len value: 233.8107 verified: true --- # BigBirdPegasus model (large) BigBird, is a sparse-attention based transformer which extends Transformer based models, such as BERT to much longer sequences. Moreover, BigBird comes along with a theoretical understanding of the capabilities of a complete transformer that the sparse model can handle. BigBird was introduced in this [paper](https://arxiv.org/abs/2007.14062) and first released in this [repository](https://github.com/google-research/bigbird). Disclaimer: The team releasing BigBird did not write a model card for this model so this model card has been written by the Hugging Face team. ## Model description BigBird relies on **block sparse attention** instead of normal attention (i.e. BERT's attention) and can handle sequences up to a length of 4096 at a much lower compute cost compared to BERT. It has achieved SOTA on various tasks involving very long sequences such as long documents summarization, question-answering with long contexts. ## How to use Here is how to use this model to get the features of a given text in PyTorch: ```python from transformers import BigBirdPegasusForConditionalGeneration, AutoTokenizer tokenizer = AutoTokenizer.from_pretrained("google/bigbird-pegasus-large-arxiv") # by default encoder-attention is `block_sparse` with num_random_blocks=3, block_size=64 model = BigBirdPegasusForConditionalGeneration.from_pretrained("google/bigbird-pegasus-large-arxiv") # decoder attention type can't be changed & will be "original_full" # you can change `attention_type` (encoder only) to full attention like this: model = BigBirdPegasusForConditionalGeneration.from_pretrained("google/bigbird-pegasus-large-arxiv", attention_type="original_full") # you can change `block_size` & `num_random_blocks` like this: model = BigBirdPegasusForConditionalGeneration.from_pretrained("google/bigbird-pegasus-large-arxiv", block_size=16, num_random_blocks=2) text = "Replace me by any text you'd like." inputs = tokenizer(text, return_tensors='pt') prediction = model.generate(**inputs) prediction = tokenizer.batch_decode(prediction) ``` ## Training Procedure This checkpoint is obtained after fine-tuning `BigBirdPegasusForConditionalGeneration` for **summarization** on **arxiv dataset** from [scientific_papers](https://huggingface.co/datasets/scientific_papers). ## BibTeX entry and citation info ```tex @misc{zaheer2021big, title={Big Bird: Transformers for Longer Sequences}, author={Manzil Zaheer and Guru Guruganesh and Avinava Dubey and Joshua Ainslie and Chris Alberti and Santiago Ontanon and Philip Pham and Anirudh Ravula and Qifan Wang and Li Yang and Amr Ahmed}, year={2021}, eprint={2007.14062}, archivePrefix={arXiv}, primaryClass={cs.LG} } ```

Yntec/photoMovieXFinal

Yntec

diffusers

text-to-image

--- license: creativeml-openrail-m library_name: diffusers pipeline_tag: text-to-image tags: - stable-diffusion - stable-diffusion-diffusers - diffusers - text-to-image - MagicArt35 --- # Photo Movie X Final Samples and prompts:  Portrait of beautiful pretty cute lady, wearing ice princess dress realistic, stunning realistic photograph, full lips, 3d render, octane render, intricately detailed, cinematic, trending on artstation | Isometric | Centered hyper realistic cover photo awesome full color, hand drawn, dark, gritty, realistic mucha, klimt, erte .12k, intricate. high definition , cinematic, Rough sketch, mix of bold dark lines and loose lines, bold lines, on paper ,  Pretty cute girl in a future where humanity has colonized the stars, a group of explorers embarks on a journey to a distant planet, hoping to discover new forms of life and unlock the secrets of the universe. But as they descend through the planet’s thick atmosphere, they discover that the world below is more dangerous and mysterious than they could have ever imagined. Original page: https://civitai.com/models/94687?modelVersionId=103445

huawei-noah/TinyBERT_General_4L_312D

huawei-noah

transformers

TinyBERT: Distilling BERT for Natural Language Understanding ======== TinyBERT is 7.5x smaller and 9.4x faster on inference than BERT-base and achieves competitive performances in the tasks of natural language understanding. It performs a novel transformer distillation at both the pre-training and task-specific learning stages. In general distillation, we use the original BERT-base without fine-tuning as the teacher and a large-scale text corpus as the learning data. By performing the Transformer distillation on the text from general domain, we obtain a general TinyBERT which provides a good initialization for the task-specific distillation. We here provide the general TinyBERT for your tasks at hand. For more details about the techniques of TinyBERT, refer to our paper: [TinyBERT: Distilling BERT for Natural Language Understanding](https://arxiv.org/abs/1909.10351) Citation ======== If you find TinyBERT useful in your research, please cite the following paper: ``` @article{jiao2019tinybert, title={Tinybert: Distilling bert for natural language understanding}, author={Jiao, Xiaoqi and Yin, Yichun and Shang, Lifeng and Jiang, Xin and Chen, Xiao and Li, Linlin and Wang, Fang and Liu, Qun}, journal={arXiv preprint arXiv:1909.10351}, year={2019} } ```

codellama/CodeLlama-13b-hf

codellama

transformers

text-generation

--- language: - code pipeline_tag: text-generation tags: - llama-2 license: llama2 --- # **Code Llama** Code Llama is a collection of pretrained and fine-tuned generative text models ranging in scale from 7 billion to 34 billion parameters. This is the repository for the base 13B version in the Hugging Face Transformers format. This model is designed for general code synthesis and understanding. Links to other models can be found in the index at the bottom. | | Base Model | Python | Instruct | | --- | ----------------------------------------------------------------------------- | ------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------- | | 7B | [codellama/CodeLlama-7b-hf](https://huggingface.co/codellama/CodeLlama-7b-hf) | [codellama/CodeLlama-7b-Python-hf](https://huggingface.co/codellama/CodeLlama-7b-Python-hf) | [codellama/CodeLlama-7b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-7b-Instruct-hf) | | 13B | [codellama/CodeLlama-13b-hf](https://huggingface.co/codellama/CodeLlama-13b-hf) | [codellama/CodeLlama-13b-Python-hf](https://huggingface.co/codellama/CodeLlama-13b-Python-hf) | [codellama/CodeLlama-13b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-13b-Instruct-hf) | | 34B | [codellama/CodeLlama-34b-hf](https://huggingface.co/codellama/CodeLlama-34b-hf) | [codellama/CodeLlama-34b-Python-hf](https://huggingface.co/codellama/CodeLlama-34b-Python-hf) | [codellama/CodeLlama-34b-Instruct-hf](https://huggingface.co/codellama/CodeLlama-34b-Instruct-hf) | ## Model Use To use this model, please make sure to install transformers from `main` until the next version is released: ```bash pip install git+https://github.com/huggingface/transformers.git@main accelerate ``` Model capabilities: - [x] Code completion. - [x] Infilling. - [ ] Instructions / chat. - [ ] Python specialist. ```python from transformers import AutoTokenizer import transformers import torch model = "codellama/CodeLlama-13b-hf" tokenizer = AutoTokenizer.from_pretrained(model) pipeline = transformers.pipeline( "text-generation", model=model, torch_dtype=torch.float16, device_map="auto", ) sequences = pipeline( 'import socket\n\ndef ping_exponential_backoff(host: str):', do_sample=True, top_k=10, temperature=0.1, top_p=0.95, num_return_sequences=1, eos_token_id=tokenizer.eos_token_id, max_length=200, ) for seq in sequences: print(f"Result: {seq['generated_text']}") ``` ## Model Details *Note: Use of this model is governed by the Meta license. Meta developed and publicly released the Code Llama family of large language models (LLMs). **Model Developers** Meta **Variations** Code Llama comes in three model sizes, and three variants: * Code Llama: base models designed for general code synthesis and understanding * Code Llama - Python: designed specifically for Python * Code Llama - Instruct: for instruction following and safer deployment All variants are available in sizes of 7B, 13B and 34B parameters. **This repository contains the base version of the 13B parameters model.** **Input** Models input text only. **Output** Models generate text only. **Model Architecture** Code Llama is an auto-regressive language model that uses an optimized transformer architecture. **Model Dates** Code Llama and its variants have been trained between January 2023 and July 2023. **Status** This is a static model trained on an offline dataset. Future versions of Code Llama - Instruct will be released as we improve model safety with community feedback. **License** A custom commercial license is available at: [https://ai.meta.com/resources/models-and-libraries/llama-downloads/](https://ai.meta.com/resources/models-and-libraries/llama-downloads/) **Research Paper** More information can be found in the paper "[Code Llama: Open Foundation Models for Code](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/)" or its [arXiv page](https://arxiv.org/abs/2308.12950). ## Intended Use **Intended Use Cases** Code Llama and its variants is intended for commercial and research use in English and relevant programming languages. The base model Code Llama can be adapted for a variety of code synthesis and understanding tasks, Code Llama - Python is designed specifically to handle the Python programming language, and Code Llama - Instruct is intended to be safer to use for code assistant and generation applications. **Out-of-Scope Uses** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in languages other than English. Use in any other way that is prohibited by the Acceptable Use Policy and Licensing Agreement for Code Llama and its variants. ## Hardware and Software **Training Factors** We used custom training libraries. The training and fine-tuning of the released models have been performed Meta’s Research Super Cluster. **Carbon Footprint** In aggregate, training all 9 Code Llama models required 400K GPU hours of computation on hardware of type A100-80GB (TDP of 350-400W). Estimated total emissions were 65.3 tCO2eq, 100% of which were offset by Meta’s sustainability program. ## Training Data All experiments reported here and the released models have been trained and fine-tuned using the same data as Llama 2 with different weights (see Section 2 and Table 1 in the [research paper](https://ai.meta.com/research/publications/code-llama-open-foundation-models-for-code/) for details). ## Evaluation Results See evaluations for the main models and detailed ablations in Section 3 and safety evaluations in Section 4 of the research paper. ## Ethical Considerations and Limitations Code Llama and its variants are a new technology that carries risks with use. Testing conducted to date has been in English, and has not covered, nor could it cover all scenarios. For these reasons, as with all LLMs, Code Llama’s potential outputs cannot be predicted in advance, and the model may in some instances produce inaccurate or objectionable responses to user prompts. Therefore, before deploying any applications of Code Llama, developers should perform safety testing and tuning tailored to their specific applications of the model. Please see the Responsible Use Guide available available at [https://ai.meta.com/llama/responsible-use-guide](https://ai.meta.com/llama/responsible-use-guide).

stabilityai/sd-x2-latent-upscaler

stabilityai

diffusers

--- license: openrail++ tags: - stable-diffusion inference: false --- # Stable Diffusion x2 latent upscaler model card This model card focuses on the latent diffusion-based upscaler developed by [Katherine Crowson](https://github.com/crowsonkb/k-diffusion) in collaboration with [Stability AI](https://stability.ai/). This model was trained on a high-resolution subset of the LAION-2B dataset. It is a diffusion model that operates in the same latent space as the Stable Diffusion model, which is decoded into a full-resolution image. To use it with Stable Diffusion, You can take the generated latent from Stable Diffusion and pass it into the upscaler before decoding with your standard VAE. Or you can take any image, encode it into the latent space, use the upscaler, and decode it. **Note**: This upscaling model is designed explicitely for **Stable Diffusion** as it can upscale Stable Diffusion's latent denoised image embeddings. This allows for very fast text-to-image + upscaling pipelines as all intermeditate states can be kept on GPU. More for information, see example below. This model works on all [Stable Diffusion checkpoints](https://huggingface.co/models?other=stable-diffusion) |  | |:--:| Image by Tanishq Abraham from [Stability AI](https://stability.ai/) originating from [this tweet](https://twitter.com/StabilityAI/status/1590531958815064065)| Original output image | 2x upscaled output image :-------------------------:|:-------------------------:  |  - Use it with 🧨 [`diffusers`](https://huggingface.co/stabilityai/sd-x2-latent-upscaler#examples) ## Model Details - **Developed by:** Katherine Crowson - **Model type:** Diffusion-based latent upscaler - **Language(s):** English - **License:** [CreativeML Open RAIL++-M License](https://huggingface.co/stabilityai/stable-diffusion-2/blob/main/LICENSE-MODEL) ## Examples Using the [🤗's Diffusers library](https://github.com/huggingface/diffusers) to run latent upscaler on top of any `StableDiffusionUpscalePipeline` checkpoint to enhance its output image resolution by a factor of 2. ```bash pip install git+https://github.com/huggingface/diffusers.git pip install transformers accelerate scipy safetensors ``` ```python from diffusers import StableDiffusionLatentUpscalePipeline, StableDiffusionPipeline import torch pipeline = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4", torch_dtype=torch.float16) pipeline.to("cuda") upscaler = StableDiffusionLatentUpscalePipeline.from_pretrained("stabilityai/sd-x2-latent-upscaler", torch_dtype=torch.float16) upscaler.to("cuda") prompt = "a photo of an astronaut high resolution, unreal engine, ultra realistic" generator = torch.manual_seed(33) # we stay in latent space! Let's make sure that Stable Diffusion returns the image # in latent space low_res_latents = pipeline(prompt, generator=generator, output_type="latent").images upscaled_image = upscaler( prompt=prompt, image=low_res_latents, num_inference_steps=20, guidance_scale=0, generator=generator, ).images[0] # Let's save the upscaled image under "upscaled_astronaut.png" upscaled_image.save("astronaut_1024.png") # as a comparison: Let's also save the low-res image with torch.no_grad(): image = pipeline.decode_latents(low_res_latents) image = pipeline.numpy_to_pil(image)[0] image.save("astronaut_512.png") ``` **Result**: *512-res Astronaut*  *1024-res Astronaut*  **Notes**: - Despite not being a dependency, we highly recommend you to install [xformers](https://github.com/facebookresearch/xformers) for memory efficient attention (better performance) - If you have low GPU RAM available, make sure to add a `pipe.enable_attention_slicing()` after sending it to `cuda` for less VRAM usage (to the cost of speed) # Uses ## Direct Use The model is intended for research purposes only. Possible research areas and tasks include - Safe deployment of models which have the potential to generate harmful content. - Probing and understanding the limitations and biases of generative models. - Generation of artworks and use in design and other artistic processes. - Applications in educational or creative tools. - Research on generative models. Excluded uses are described below. ### Misuse, Malicious Use, and Out-of-Scope Use _Note: This section is originally taken from the [DALLE-MINI model card](https://huggingface.co/dalle-mini/dalle-mini), was used for Stable Diffusion v1, but applies in the same way to Stable Diffusion v2_. The model should not be used to intentionally create or disseminate images that create hostile or alienating environments for people. This includes generating images that people would foreseeably find disturbing, distressing, or offensive; or content that propagates historical or current stereotypes. #### Out-of-Scope Use The model was not trained to be factual or true representations of people or events, and therefore using the model to generate such content is out-of-scope for the abilities of this model. #### Misuse and Malicious Use Using the model to generate content that is cruel to individuals is a misuse of this model. This includes, but is not limited to: - Generating demeaning, dehumanizing, or otherwise harmful representations of people or their environments, cultures, religions, etc. - Intentionally promoting or propagating discriminatory content or harmful stereotypes. - Impersonating individuals without their consent. - Sexual content without consent of the people who might see it. - Mis- and disinformation - Representations of egregious violence and gore - Sharing of copyrighted or licensed material in violation of its terms of use. - Sharing content that is an alteration of copyrighted or licensed material in violation of its terms of use. ## Limitations and Bias ### Limitations - The model does not achieve perfect photorealism - The model cannot render legible text - The model does not perform well on more difficult tasks which involve compositionality, such as rendering an image corresponding to “A red cube on top of a blue sphere” - Faces and people in general may not be generated properly. - The model was trained mainly with English captions and will not work as well in other languages. - The autoencoding part of the model is lossy - The model was trained on a subset of the large-scale dataset [LAION-5B](https://laion.ai/blog/laion-5b/), which contains adult, violent and sexual content. To partially mitigate this, we have filtered the dataset using LAION's NFSW detector (see Training section). ### Bias While the capabilities of image generation models are impressive, they can also reinforce or exacerbate social biases. Stable Diffusion vw was primarily trained on subsets of [LAION-2B(en)](https://laion.ai/blog/laion-5b/), which consists of images that are limited to English descriptions. Texts and images from communities and cultures that use other languages are likely to be insufficiently accounted for. This affects the overall output of the model, as white and western cultures are often set as the default. Further, the ability of the model to generate content with non-English prompts is significantly worse than with English-language prompts. Stable Diffusion v2 mirrors and exacerbates biases to such a degree that viewer discretion must be advised irrespective of the input or its intent.

classla/bcms-bertic-ner

classla

transformers

token-classification

--- language: - hr - bs - sr - cnr - hbs widget: - text: "Zovem se Marko i živim u Zagrebu. Studirao sam u Beogradu na Filozofskom fakultetu. Obožavam album Moanin." license: apache-2.0 --- # The [BERTić](https://huggingface.co/classla/bcms-bertic)&ast; [bert-ich] /bɜrtitʃ/ model fine-tuned for the task of named entity recognition in Bosnian, Croatian, Montenegrin and Serbian (BCMS) &ast; The name should resemble the facts (1) that the model was trained in Zagreb, Croatia, where diminutives ending in -ić (as in fotić, smajlić, hengić etc.) are very popular, and (2) that most surnames in the countries where these languages are spoken end in -ić (with diminutive etymology as well). This is a fine-tuned version of the [BERTić](https://huggingface.co/classla/bcms-bertic) model for the task of named entity recognition (PER, LOC, ORG, MISC). The fine-tuning was performed on the following datasets: - the [hr500k](http://hdl.handle.net/11356/1183) dataset, 500 thousand tokens in size, standard Croatian - the [SETimes.SR](http://hdl.handle.net/11356/1200) dataset, 87 thousand tokens in size, standard Serbian - the [ReLDI-hr](http://hdl.handle.net/11356/1241) dataset, 89 thousand tokens in size, Internet (Twitter) Croatian - the [ReLDI-sr](http://hdl.handle.net/11356/1240) dataset, 92 thousand tokens in size, Internet (Twitter) Serbian The data was augmented with missing diacritics and standard data was additionally over-represented. The F1 obtained on dev data (train and test was merged into train) is 91.38. For a more detailed per-dataset evaluation of the BERTić model on the NER task have a look at the [main model page](https://huggingface.co/classla/bcms-bertic). If you use this fine-tuned model, please cite the following paper: ``` @inproceedings{ljubesic-lauc-2021-bertic, title = "{BERT}i{\'c} - The Transformer Language Model for {B}osnian, {C}roatian, {M}ontenegrin and {S}erbian", author = "Ljube{\v{s}}i{\'c}, Nikola and Lauc, Davor", booktitle = "Proceedings of the 8th Workshop on Balto-Slavic Natural Language Processing", month = apr, year = "2021", address = "Kiyv, Ukraine", publisher = "Association for Computational Linguistics", url = "https://www.aclweb.org/anthology/2021.bsnlp-1.5", pages = "37--42", } ``` When running the model in `simpletransformers`, the order of labels has to be set as well. ``` from simpletransformers.ner import NERModel, NERArgs model_args = NERArgs() model_args.labels_list = ['B-LOC','B-MISC','B-ORG','B-PER','I-LOC','I-MISC','I-ORG','I-PER','O'] model = NERModel('electra', 'classla/bcms-bertic-ner', args=model_args) ```

lmsys/vicuna-13b-v1.3

lmsys

transformers

text-generation

--- inference: false --- **NOTE: New version available** Please check out a newer version of the weights [here](https://github.com/lm-sys/FastChat/blob/main/docs/vicuna_weights_version.md). <br> # Vicuna Model Card ## Model Details Vicuna is a chat assistant trained by fine-tuning LLaMA on user-shared conversations collected from ShareGPT. - **Developed by:** [LMSYS](https://lmsys.org/) - **Model type:** An auto-regressive language model based on the transformer architecture. - **License:** Non-commercial license - **Finetuned from model:** [LLaMA](https://arxiv.org/abs/2302.13971). ### Model Sources - **Repository:** https://github.com/lm-sys/FastChat - **Blog:** https://lmsys.org/blog/2023-03-30-vicuna/ - **Paper:** https://arxiv.org/abs/2306.05685 - **Demo:** https://chat.lmsys.org/ ## Uses The primary use of Vicuna is research on large language models and chatbots. The primary intended users of the model are researchers and hobbyists in natural language processing, machine learning, and artificial intelligence. ## How to Get Started with the Model - Command line interface: https://github.com/lm-sys/FastChat#vicuna-weights. - APIs (OpenAI API, Huggingface API): https://github.com/lm-sys/FastChat/tree/main#api. ## Training Details Vicuna v1.3 is fine-tuned from LLaMA with supervised instruction fine-tuning. The training data is around 125K conversations collected from ShareGPT.com. See more details in the "Training Details of Vicuna Models" section in the appendix of this [paper](https://arxiv.org/pdf/2306.05685.pdf). ## Evaluation Vicuna is evaluated with standard benchmarks, human preference, and LLM-as-a-judge. See more details in this [paper](https://arxiv.org/pdf/2306.05685.pdf) and [leaderboard](https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard). ## Difference between different versions of Vicuna See [vicuna_weights_version.md](https://github.com/lm-sys/FastChat/blob/main/docs/vicuna_weights_version.md)

Rostlab/prot_bert

Rostlab

transformers

fill-mask

--- tags: - protein language model - protein datasets: - Uniref100 --- # ProtBert model Pretrained model on protein sequences using a masked language modeling (MLM) objective. It was introduced in [this paper](https://doi.org/10.1101/2020.07.12.199554) and first released in [this repository](https://github.com/agemagician/ProtTrans). This model is trained on uppercase amino acids: it only works with capital letter amino acids. ## Model description ProtBert is based on Bert model which pretrained on a large corpus of protein sequences in a self-supervised fashion. This means it was pretrained on the raw protein sequences only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from those protein sequences. One important difference between our Bert model and the original Bert version is the way of dealing with sequences as separate documents. This means the Next sentence prediction is not used, as each sequence is treated as a complete document. The masking follows the original Bert training with randomly masks 15% of the amino acids in the input. At the end, the feature extracted from this model revealed that the LM-embeddings from unlabeled data (only protein sequences) captured important biophysical properties governing protein shape. This implied learning some of the grammar of the language of life realized in protein sequences. ## Intended uses & limitations The model could be used for protein feature extraction or to be fine-tuned on downstream tasks. We have noticed in some tasks you could gain more accuracy by fine-tuning the model rather than using it as a feature extractor. ### How to use You can use this model directly with a pipeline for masked language modeling: ```python >>> from transformers import BertForMaskedLM, BertTokenizer, pipeline >>> tokenizer = BertTokenizer.from_pretrained("Rostlab/prot_bert", do_lower_case=False ) >>> model = BertForMaskedLM.from_pretrained("Rostlab/prot_bert") >>> unmasker = pipeline('fill-mask', model=model, tokenizer=tokenizer) >>> unmasker('D L I P T S S K L V V [MASK] D T S L Q V K K A F F A L V T') [{'score': 0.11088453233242035, 'sequence': '[CLS] D L I P T S S K L V V L D T S L Q V K K A F F A L V T [SEP]', 'token': 5, 'token_str': 'L'}, {'score': 0.08402521163225174, 'sequence': '[CLS] D L I P T S S K L V V S D T S L Q V K K A F F A L V T [SEP]', 'token': 10, 'token_str': 'S'}, {'score': 0.07328339666128159, 'sequence': '[CLS] D L I P T S S K L V V V D T S L Q V K K A F F A L V T [SEP]', 'token': 8, 'token_str': 'V'}, {'score': 0.06921856850385666, 'sequence': '[CLS] D L I P T S S K L V V K D T S L Q V K K A F F A L V T [SEP]', 'token': 12, 'token_str': 'K'}, {'score': 0.06382402777671814, 'sequence': '[CLS] D L I P T S S K L V V I D T S L Q V K K A F F A L V T [SEP]', 'token': 11, 'token_str': 'I'}] ``` Here is how to use this model to get the features of a given protein sequence in PyTorch: ```python from transformers import BertModel, BertTokenizer import re tokenizer = BertTokenizer.from_pretrained("Rostlab/prot_bert", do_lower_case=False ) model = BertModel.from_pretrained("Rostlab/prot_bert") sequence_Example = "A E T C Z A O" sequence_Example = re.sub(r"[UZOB]", "X", sequence_Example) encoded_input = tokenizer(sequence_Example, return_tensors='pt') output = model(**encoded_input) ``` ## Training data The ProtBert model was pretrained on [Uniref100](https://www.uniprot.org/downloads), a dataset consisting of 217 million protein sequences. ## Training procedure ### Preprocessing The protein sequences are uppercased and tokenized using a single space and a vocabulary size of 21. The rare amino acids "U,Z,O,B" were mapped to "X". The inputs of the model are then of the form: ``` [CLS] Protein Sequence A [SEP] Protein Sequence B [SEP] ``` Furthermore, each protein sequence was treated as a separate document. The preprocessing step was performed twice, once for a combined length (2 sequences) of less than 512 amino acids, and another time using a combined length (2 sequences) of less than 2048 amino acids. The details of the masking procedure for each sequence followed the original Bert model as following: - 15% of the amino acids are masked. - In 80% of the cases, the masked amino acids are replaced by `[MASK]`. - In 10% of the cases, the masked amino acids are replaced by a random amino acid (different) from the one they replace. - In the 10% remaining cases, the masked amino acids are left as is. ### Pretraining The model was trained on a single TPU Pod V3-512 for 400k steps in total. 300K steps using sequence length 512 (batch size 15k), and 100K steps using sequence length 2048 (batch size 2.5k). The optimizer used is Lamb with a learning rate of 0.002, a weight decay of 0.01, learning rate warmup for 40k steps and linear decay of the learning rate after. ## Evaluation results When fine-tuned on downstream tasks, this model achieves the following results: Test results : | Task/Dataset | secondary structure (3-states) | secondary structure (8-states) | Localization | Membrane | |:-----:|:-----:|:-----:|:-----:|:-----:| | CASP12 | 75 | 63 | | | | TS115 | 83 | 72 | | | | CB513 | 81 | 66 | | | | DeepLoc | | | 79 | 91 | ### BibTeX entry and citation info ```bibtex @article {Elnaggar2020.07.12.199554, author = {Elnaggar, Ahmed and Heinzinger, Michael and Dallago, Christian and Rehawi, Ghalia and Wang, Yu and Jones, Llion and Gibbs, Tom and Feher, Tamas and Angerer, Christoph and Steinegger, Martin and BHOWMIK, DEBSINDHU and Rost, Burkhard}, title = {ProtTrans: Towards Cracking the Language of Life{\textquoteright}s Code Through Self-Supervised Deep Learning and High Performance Computing}, elocation-id = {2020.07.12.199554}, year = {2020}, doi = {10.1101/2020.07.12.199554}, publisher = {Cold Spring Harbor Laboratory}, abstract = {Computational biology and bioinformatics provide vast data gold-mines from protein sequences, ideal for Language Models (LMs) taken from Natural Language Processing (NLP). These LMs reach for new prediction frontiers at low inference costs. Here, we trained two auto-regressive language models (Transformer-XL, XLNet) and two auto-encoder models (Bert, Albert) on data from UniRef and BFD containing up to 393 billion amino acids (words) from 2.1 billion protein sequences (22- and 112 times the entire English Wikipedia). The LMs were trained on the Summit supercomputer at Oak Ridge National Laboratory (ORNL), using 936 nodes (total 5616 GPUs) and one TPU Pod (V3-512 or V3-1024). We validated the advantage of up-scaling LMs to larger models supported by bigger data by predicting secondary structure (3-states: Q3=76-84, 8 states: Q8=65-73), sub-cellular localization for 10 cellular compartments (Q10=74) and whether a protein is membrane-bound or water-soluble (Q2=89). Dimensionality reduction revealed that the LM-embeddings from unlabeled data (only protein sequences) captured important biophysical properties governing protein shape. This implied learning some of the grammar of the language of life realized in protein sequences. The successful up-scaling of protein LMs through HPC to larger data sets slightly reduced the gap between models trained on evolutionary information and LMs. Availability ProtTrans: \&lt;a href="https://github.com/agemagician/ProtTrans"\&gt;https://github.com/agemagician/ProtTrans\&lt;/a\&gt;Competing Interest StatementThe authors have declared no competing interest.}, URL = {https://www.biorxiv.org/content/early/2020/07/21/2020.07.12.199554}, eprint = {https://www.biorxiv.org/content/early/2020/07/21/2020.07.12.199554.full.pdf}, journal = {bioRxiv} } ``` > Created by [Ahmed Elnaggar/@Elnaggar_AI](https://twitter.com/Elnaggar_AI) | [LinkedIn](https://www.linkedin.com/in/prof-ahmed-elnaggar/)