From ef41eb23b2cfb756241b9a560e7a48683e598e96 Mon Sep 17 00:00:00 2001 From: jxue16 <105090474+jxue16@users.noreply.github.com> Date: Tue, 4 Jun 2024 14:59:18 -0700 Subject: [PATCH] add assets --- assets/360.yaml | 22 +++++++++++ assets/adobe.yaml | 6 ++- assets/ai2.yaml | 3 +- assets/alibaba.yaml | 21 ++++++++--- assets/apple.yaml | 20 +++++++--- assets/cartesia.yaml | 22 +++++++++++ assets/cohere.yaml | 34 +++++++++++++++-- assets/deepmind.yaml | 63 ++++++++++++++++++++++++++++++++ assets/eleutherai.yaml | 12 ++++-- assets/fuse.yaml | 13 +++++-- assets/google.yaml | 17 ++++++--- assets/huggingface.yaml | 27 ++++++++++---- assets/konan.yaml | 6 ++- assets/ktai.yaml | 11 ++++-- assets/lg.yaml | 3 +- assets/llm360.yaml | 26 +++++++++++++ assets/meta.yaml | 43 +++++++++++++++++++--- assets/microsoft.yaml | 81 ++++++++++++++++++++++++++++++++++------- assets/mistral.yaml | 21 +++++++++++ assets/naver.yaml | 8 +++- assets/ncsoft.yaml | 9 +++-- assets/openai.yaml | 21 +++++++++++ assets/openbmb.yaml | 7 +++- assets/reka.yaml | 14 +++++-- assets/shanghai.yaml | 21 ++++++++--- assets/skt.yaml | 5 ++- assets/stability.yaml | 15 ++++++-- assets/tokyo.yaml | 14 ++++--- assets/xai.yaml | 14 +++++-- js/main.js | 2 + 30 files changed, 491 insertions(+), 90 deletions(-) create mode 100644 assets/360.yaml create mode 100644 assets/cartesia.yaml diff --git a/assets/360.yaml b/assets/360.yaml new file mode 100644 index 00000000..0926d33a --- /dev/null +++ b/assets/360.yaml @@ -0,0 +1,22 @@ +--- +- type: model + name: 360 Zhinao + organization: 360 Security + description: 360 Zhinao is a multilingual LLM in Chinese and English with chat capabilities. + created_date: 2024-05-23 + url: https://arxiv.org/pdf/2405.13386 + model_card: none + modality: text; text + analysis: Achieved competitive performance on relevant benchmarks against other 7B models in Chinese, English, and coding tasks. + size: 7B parameters + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknwon + quality_control: '' + access: open + license: unknown + intended_uses: '' + prohibited_uses: '' + monitoring: '' + feedback: none \ No newline at end of file diff --git a/assets/adobe.yaml b/assets/adobe.yaml index ec346907..081311f9 100644 --- a/assets/adobe.yaml +++ b/assets/adobe.yaml @@ -101,7 +101,8 @@ - type: dataset name: CulturaX organization: University of Oregon, Adobe - description: CulturaX is a substantial multilingual dataset with 6.3 trillion tokens in 167 languages, tailored for LLM development. + description: CulturaX is a substantial multilingual dataset with 6.3 trillion + tokens in 167 languages, tailored for LLM development. created_date: 2023-09-17 url: https://arxiv.org/pdf/2309.09400 datasheet: https://huggingface.co/datasets/uonlp/CulturaX @@ -116,6 +117,7 @@ access: open license: mC4, OSCAR intended_uses: '' - prohibited_uses: The data must not be utilized for malicious or harmful purposes towards humanity. + prohibited_uses: The data must not be utilized for malicious or harmful purposes + towards humanity. monitoring: unknown feedback: https://huggingface.co/datasets/uonlp/CulturaX/discussions diff --git a/assets/ai2.yaml b/assets/ai2.yaml index 4188c061..0c1f2189 100644 --- a/assets/ai2.yaml +++ b/assets/ai2.yaml @@ -259,7 +259,8 @@ - type: dataset name: MADLAD-400 organization: AI2 - description: MADLAD-400 is a document-level multilingual dataset based on Common Crawl, covering 419 languages in total. + description: MADLAD-400 is a document-level multilingual dataset based on Common + Crawl, covering 419 languages in total. created_date: 2023-09-09 url: https://arxiv.org/abs/2309.04662 datasheet: https://huggingface.co/datasets/allenai/MADLAD-400 diff --git a/assets/alibaba.yaml b/assets/alibaba.yaml index 1541b894..17d78f68 100644 --- a/assets/alibaba.yaml +++ b/assets/alibaba.yaml @@ -145,23 +145,34 @@ - type: model name: SeaLLM v2.5 organization: DAMO Academy, Alibaba - description: SeaLLM v2.5 is a multilingual large language model for Southeast Asian (SEA) languages. + description: SeaLLM v2.5 is a multilingual large language model for Southeast + Asian (SEA) languages. created_date: 2024-04-12 url: https://github.com/DAMO-NLP-SG/SeaLLMs model_card: https://huggingface.co/SeaLLMs/SeaLLM-7B-v2.5 modality: text; text - analysis: The model was evaluated on 3 benchmarks (MMLU for English, M3Exam (M3e) for English, Chinese, Vietnamese, Indonesian, and Thai, and VMLU for Vietnamese) and it outperformed GPT-3 and Vistral-7B-chat models across these benchmarks in the given languages. + analysis: The model was evaluated on 3 benchmarks (MMLU for English, M3Exam (M3e) + for English, Chinese, Vietnamese, Indonesian, and Thai, and VMLU for Vietnamese) + and it outperformed GPT-3 and Vistral-7B-chat models across these benchmarks + in the given languages. size: 7B parameters dependencies: [Gemma] training_emissions: unknown training_time: unknown training_hardware: unknown - quality_control: Despite efforts in red teaming and safety fine-tuning and enforcement, the creators suggest, developers and stakeholders should perform their own red teaming and provide related security measures before deployment, and they must abide by and comply with local governance and regulations. + quality_control: Despite efforts in red teaming and safety fine-tuning and enforcement, + the creators suggest, developers and stakeholders should perform their own red + teaming and provide related security measures before deployment, and they must + abide by and comply with local governance and regulations. access: open license: explanation: License can be found at https://huggingface.co/SeaLLMs/SeaLLM-13B-Chat/blob/main/LICENSE value: custom - intended_uses: The model is intended for multilingual tasks such as knowledge retrieval, math reasoning, and instruction following. Also, it could be used to provide multilingual assistance. - prohibited_uses: The model should not be used in a way that could lead to inaccurate, misleading or potentially harmful generation. Users should comply with local laws and regulations when deploying the model. + intended_uses: The model is intended for multilingual tasks such as knowledge + retrieval, math reasoning, and instruction following. Also, it could be used + to provide multilingual assistance. + prohibited_uses: The model should not be used in a way that could lead to inaccurate, + misleading or potentially harmful generation. Users should comply with local + laws and regulations when deploying the model. monitoring: unknown feedback: https://huggingface.co/SeaLLMs/SeaLLM-7B-v2.5/discussions diff --git a/assets/apple.yaml b/assets/apple.yaml index 4462a34d..30035dd2 100644 --- a/assets/apple.yaml +++ b/assets/apple.yaml @@ -25,21 +25,31 @@ - type: model name: OpenELM organization: Apple - description: OpenELM is a family of Open-source Efficient Language Models. It uses a layer-wise scaling strategy to efficiently allocate parameters within each layer of the transformer model, leading to enhanced accuracy. + description: OpenELM is a family of Open-source Efficient Language Models. It + uses a layer-wise scaling strategy to efficiently allocate parameters within + each layer of the transformer model, leading to enhanced accuracy. created_date: 2024-04-24 url: https://machinelearning.apple.com/research/openelm model_card: https://huggingface.co/apple/OpenELM-3B-Instruct modality: text; text - analysis: The models were evaluated in terms of zero-shot, LLM360, and OpenLLM leaderboard results. + analysis: The models were evaluated in terms of zero-shot, LLM360, and OpenLLM + leaderboard results. size: 3B parameters - dependencies: [RefinedWeb, The Pile, RedPajama-Data, Dolma, CoreNet library] + dependencies: + - RefinedWeb + - The Pile + - RedPajama-Data + - Dolma + - CoreNet library training_emissions: unknown training_time: unknown training_hardware: unknown quality_control: unknown access: open license: Apple - intended_uses: To empower and enrich the open research community by providing access to state-of-the-art language models. - prohibited_uses: No explicit prohibited uses stated, though it is noted that users should undertake thorough safety testing. + intended_uses: To empower and enrich the open research community by providing + access to state-of-the-art language models. + prohibited_uses: No explicit prohibited uses stated, though it is noted that users + should undertake thorough safety testing. monitoring: none feedback: https://huggingface.co/apple/OpenELM-3B-Instruct/discussions diff --git a/assets/cartesia.yaml b/assets/cartesia.yaml new file mode 100644 index 00000000..a3490c7d --- /dev/null +++ b/assets/cartesia.yaml @@ -0,0 +1,22 @@ +--- +- type: model + name: Sonic + organization: Cartesia + description: Sonic is a low-latency voice model that generates lifelike speech. Developed by Cartesia, it was designed to be an efficient real-time AI capable of processing any-sized contexts and running on any device. + created_date: 2024-05-29 + url: https://cartesia.ai/blog/sonic + model_card: none + modality: text; audio + analysis: Extensive testing on Multilingual Librispeech dataset resulted in 20% lower validation perplexity. In downstream evaluations, this leads to a 2x lower word error rate and a 1 point higher quality score. Sonic also displays impressive performance metrics at inference, achieving lower latency (1.5x lower time-to-first-audio), faster inference speed (2x lower real-time factor), and higher throughput (4x). + size: 2024-05-29 + dependencies: [Multilingual Librispeech dataset] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: '' + access: limited + license: unknown + intended_uses: Sonic has potential applications across customer support, entertainment, and content creation and is a part of Cartesias broader mission to bring real-time multimodal intelligence to every device. + prohibited_uses: unknown + monitoring: unknown + feedback: Contact through the provided form or via email at join@cartesia.ai. diff --git a/assets/cohere.yaml b/assets/cohere.yaml index 7b9074df..5ca4466a 100644 --- a/assets/cohere.yaml +++ b/assets/cohere.yaml @@ -546,12 +546,14 @@ - type: model name: Rerank 3 organization: Cohere - description: Rerank 3 is a new foundation model for efficient enterprise search and retrieval with 4k context length. + description: Rerank 3 is a new foundation model for efficient enterprise search + and retrieval with 4k context length. created_date: 2024-04-11 url: https://cohere.com/blog/rerank-3 model_card: none modality: text; text - analysis: Evaluated on code retrieval and data retrieval capabilities, with improvements compared to the standard in both. + analysis: Evaluated on code retrieval and data retrieval capabilities, with improvements + compared to the standard in both. size: unknown dependencies: [] training_emissions: unknown @@ -560,7 +562,33 @@ quality_control: '' access: limited license: unknown - intended_uses: Efficient enterprise search and retrieval. + intended_uses: Efficient enterprise search and retrieval. prohibited_uses: '' monitoring: unknown feedback: none +- type: model + name: Aya 23 + organization: Cohere + description: Aya 23 is an open weights research release of an instruction fine-tuned + model with multilingual capabilities. It focuses on pairing a highly performant + pre-trained Command family of models with the recently released Aya Collection. + This model supports 23 languages. + created_date: 2024-05-31 + url: https://arxiv.org/pdf/2405.15032 + model_card: https://huggingface.co/CohereForAI/aya-23-35B + modality: text; text + analysis: Evaluated across 23 languages with the highest results in all tasks + and languages compared to other multilingual language models. + size: 35B parameters + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: unknown + access: open + license: CC-BY-NC + intended_uses: This model is designed to be used for multilingual tasks covering + 23 languages. + prohibited_uses: unknown + monitoring: unknown + feedback: https://huggingface.co/CohereForAI/aya-23-35B/discussions diff --git a/assets/deepmind.yaml b/assets/deepmind.yaml index fad62991..f90458a8 100644 --- a/assets/deepmind.yaml +++ b/assets/deepmind.yaml @@ -684,3 +684,66 @@ prohibited_uses: '' monitoring: '' feedback: none +- type: model + name: Imagen 3 + organization: Google DeepMind + description: Imagen 3 is a high-quality text-to-image model, capable of generating images with better detail, richer lighting, and fewer distracting artifacts compared to previous models. Improved understanding of prompts allows for a wide range of visual styles and captures small details from longer prompts. It also understands prompts written in natural, everyday language, making it easier to use. Imagen 3 is available in multiple versions, optimized for different types of tasks, from generating quick sketches to high-resolution images. + created_date: 2024-05-14 + url: https://deepmind.google/technologies/imagen-3/ + model_card: none + modality: text; image + analysis: The model was tested and evaluated on various prompts to assess its understanding of natural language, its ability to generate high-quality images in various formats and styles and generate fine details and complex textures. Red teaming and evaluations were conducted on topics including fairness, bias, and content safety. + size: unknown + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: Extensive filtering and data labeling were used to minimize harmful content in datasets and reduce the likelihood of harmful outputs. Privacy, safety, and security technologies were leveraged in deploying the model, including watermarking tool SynthID. + access: limited + license: unknown + intended_uses: Generate high-quality images for various purposes, from photorealistic landscapes to textured oil paintings or whimsical claymation scenes. It is useful in situations where detailed visual representation is required based on the textual description. + prohibited_uses: unknown + monitoring: Through digital watermarking tool SynthID embedded in pixels for detection and identification. + feedback: unknown +- type: model + name: Veo + organization: Google DeepMind + description: Veo is Google DeepMind's most capable video generation model to date. It generates high-quality, 1080p resolution videos that can go beyond a minute, in a wide range of cinematic and visual styles. It accurately captures the nuance and tone of a prompt, and provides an unprecedented level of creative control. The model is also capable of maintaining visual consistency in video frames, and supports masked editing. + created_date: 2024-05-14 + url: https://deepmind.google/technologies/veo/ + model_card: none + modality: text; video + analysis: unknown + size: unknown + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: Videos created by Veo are watermarked using SynthID, DeepMinds tool for watermarking and identifying AI-generated content, and passed through safety filters and memorization checking processes to mitigate privacy, copyright and bias risks. + access: limited + license: unknown + intended_uses: Veo is intended to help create tools that make video production accessible to everyone. It can be used by filmmakers, creators, or educators for storytelling, education and more. Some of its features will be also brought to products like YouTube Shorts. + prohibited_uses: unknown + monitoring: unknown + feedback: Feedback from leading creators and filmmakers is incorporated to improve Veo's generative video technologies. +- type: model + name: Gemini 1.5 Flash + organization: Google DeepMind + description: Gemini Flash is a lightweight model, optimized for speed and efficiency. It features multimodal reasoning and a breakthrough long context window of up to one million tokens. It's designed to serve at scale and is efficient on cost, providing quality results at a fraction of the cost of larger models. + created_date: 2024-05-30 + url: https://deepmind.google/technologies/gemini/flash/ + model_card: none + modality: audio, image, text, video; text + analysis: The model was evaluated on various benchmarks like General MMLU, Code Natural2Code, MATH, GPQA, Big-Bench, WMT23, MMMU, and MathVista providing performance across various domains like multilingual translation, image processing, and code generation. + size: unknown + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: The research team is continually exploring new ideas at the frontier of AI and building innovative products for consistent progress. + access: limited + license: Googles Terms and Conditions + intended_uses: The model is intended for developer and enterprise use cases. It can process hours of video and audio, and hundreds of thousands of words or lines of code, making it beneficial for a wide range of tasks. + prohibited_uses: '' + monitoring: unknown + feedback: none diff --git a/assets/eleutherai.yaml b/assets/eleutherai.yaml index 2af747be..9996bfe3 100644 --- a/assets/eleutherai.yaml +++ b/assets/eleutherai.yaml @@ -299,12 +299,17 @@ - type: model name: Pile-T5 organization: EleutherAI - description: Pile-T5 is a version of the broadly used T5 model, but improved to eliminate weaknesses such as the omission of crucial code-related tokens. It utilizes LLaMA tokenizer and is trained on the Pile, offering enhancements for finetuning on downstream tasks, particularly those involving code. + description: Pile-T5 is a version of the broadly used T5 model, but improved to + eliminate weaknesses such as the omission of crucial code-related tokens. It + utilizes LLaMA tokenizer and is trained on the Pile, offering enhancements for + finetuning on downstream tasks, particularly those involving code. created_date: 2024-04-15 url: https://blog.eleuther.ai/pile-t5/ model_card: none modality: text; text - analysis: The models were evaluated on SuperGLUE, CodeXGLUE, as well as MMLU and Bigbench Hard. Comparisons were made with T5v1.1 and found that Pile-T5 models performed better in most conditions. + analysis: The models were evaluated on SuperGLUE, CodeXGLUE, as well as MMLU and + Bigbench Hard. Comparisons were made with T5v1.1 and found that Pile-T5 models + performed better in most conditions. size: unknown dependencies: [The Pile, T5x, LLaMA, umT5] training_emissions: unknown @@ -313,7 +318,8 @@ quality_control: '' access: open license: unknown - intended_uses: The model is aimed at downstream tasks that benefit from the encoder-decoder architecture. Particularly useful for tasks involving code. + intended_uses: The model is aimed at downstream tasks that benefit from the encoder-decoder + architecture. Particularly useful for tasks involving code. prohibited_uses: unknown monitoring: unknown feedback: unknown diff --git a/assets/fuse.yaml b/assets/fuse.yaml index 2607a5e2..4432f862 100644 --- a/assets/fuse.yaml +++ b/assets/fuse.yaml @@ -2,12 +2,17 @@ - type: model name: FuseChat organization: FuseAI - description: FuseChat is a powerful chat Language Learning Model (LLM) that integrates multiple structure and scale-varied chat LLMs using a fuse-then-merge strategy. The fusion is done using two stages + description: FuseChat is a powerful chat Language Learning Model (LLM) that integrates + multiple structure and scale-varied chat LLMs using a fuse-then-merge strategy. + The fusion is done using two stages created_date: 2024-02-26 url: https://arxiv.org/abs/2402.16107 model_card: https://huggingface.co/FuseAI/FuseChat-7B-VaRM modality: text; text - analysis: The FuseChat model was evaluated on MT-Bench which comprises 80 multi-turn dialogues spanning writing, roleplay, reasoning, math, coding, stem, and humanities domains. It yields an average performance of 66.52 with specific scores for individual domains available in the leaderboard results. + analysis: The FuseChat model was evaluated on MT-Bench which comprises 80 multi-turn + dialogues spanning writing, roleplay, reasoning, math, coding, stem, and humanities + domains. It yields an average performance of 66.52 with specific scores for + individual domains available in the leaderboard results. size: 7B parameters dependencies: [Nous Hermes 2, OpenChat 3.5] training_emissions: unknown @@ -16,7 +21,9 @@ quality_control: none access: open license: Apache 2.0 - intended_uses: FuseChat is intended to be used as a powerful chat bot that takes in text inputs and provides text-based responses. It can be utilized in a variety of domains including writing, roleplay, reasoning, math, coding, stem, and humanities. + intended_uses: FuseChat is intended to be used as a powerful chat bot that takes + in text inputs and provides text-based responses. It can be utilized in a variety + of domains including writing, roleplay, reasoning, math, coding, stem, and humanities. prohibited_uses: unknown monitoring: unknown feedback: https://huggingface.co/FuseAI/FuseChat-7B-VaRM/discussions diff --git a/assets/google.yaml b/assets/google.yaml index c6dee979..965e3289 100644 --- a/assets/google.yaml +++ b/assets/google.yaml @@ -1785,12 +1785,18 @@ - type: model name: Med-Gemini organization: Google - description: Med-Gemini is a family of highly capable multimodal models that are specialized in medicine with the ability to seamlessly integrate the use of web search, and that can be efficiently tailored to novel modalities using custom encoders. + description: Med-Gemini is a family of highly capable multimodal models that are + specialized in medicine with the ability to seamlessly integrate the use of + web search, and that can be efficiently tailored to novel modalities using custom + encoders. created_date: 2024-04-29 url: https://arxiv.org/pdf/2404.18416 model_card: none modality: image, text; text - analysis: Evaluated Med-Gemini on 14 medical benchmarks spanning text, multimodal and long-context applications, establishing new state-of-the-art (SoTA) performance on 10 of them, and surpassing the GPT-4 model family on every benchmark where a direct comparison is viable. + analysis: Evaluated Med-Gemini on 14 medical benchmarks spanning text, multimodal + and long-context applications, establishing new state-of-the-art (SoTA) performance + on 10 of them, and surpassing the GPT-4 model family on every benchmark where + a direct comparison is viable. size: unknown dependencies: [Gemini, MultiMedBench] training_emissions: unknown @@ -1799,8 +1805,9 @@ quality_control: '' access: closed license: unknown - intended_uses: To be used in areas of medical research including medical summarization, referral letter generation, and medical simplification tasks. - prohibited_uses: Unfit for real-world deployment in the safety-critical medical domain. + intended_uses: To be used in areas of medical research including medical summarization, + referral letter generation, and medical simplification tasks. + prohibited_uses: Unfit for real-world deployment in the safety-critical medical + domain. monitoring: '' feedback: none - diff --git a/assets/huggingface.yaml b/assets/huggingface.yaml index 75df5184..2b4053e9 100644 --- a/assets/huggingface.yaml +++ b/assets/huggingface.yaml @@ -174,28 +174,40 @@ - type: model name: Idefics2 organization: Hugging Face - description: Idefics2 is a general multimodal model that takes as input arbitrary sequences of text and images, generating text responses. It has the capability to describe visual content, answer questions about images, perform basic arithmetic operations, create stories grounded in multiple images, and extract information from documents. + description: Idefics2 is a general multimodal model that takes as input arbitrary + sequences of text and images, generating text responses. It has the capability + to describe visual content, answer questions about images, perform basic arithmetic + operations, create stories grounded in multiple images, and extract information + from documents. created_date: 2024-04-15 url: https://huggingface.co/blog/idefics2 model_card: https://huggingface.co/HuggingFaceM4/idefics2-8b modality: image, text; text - analysis: The performance of Idefics2 has been evaluated on numerous benchmarks. It is top of its class size and competes with much larger models such as LLava-Next-34B and MM1-30B-chat. + analysis: The performance of Idefics2 has been evaluated on numerous benchmarks. + It is top of its class size and competes with much larger models such as LLava-Next-34B + and MM1-30B-chat. size: 8B parameters dependencies: [The Cauldron] training_emissions: unknown training_time: unknown training_hardware: unknown - quality_control: The quality of the model has been ensured by training it on a mixture of openly available datasets and enhancing its OCR capabilities. Further improvements include manipulating images in their native resolutions and aspect ratios, better pre-trained backbones, and allowing for sub-image splitting. + quality_control: The quality of the model has been ensured by training it on a + mixture of openly available datasets and enhancing its OCR capabilities. Further + improvements include manipulating images in their native resolutions and aspect + ratios, better pre-trained backbones, and allowing for sub-image splitting. access: open license: Apache 2.0 - intended_uses: The model can be used for answering questions about images, describing visual content, creating stories grounded in multiple images, extracting information from documents, and performing basic arithmetic operations. + intended_uses: The model can be used for answering questions about images, describing + visual content, creating stories grounded in multiple images, extracting information + from documents, and performing basic arithmetic operations. prohibited_uses: unknown monitoring: unknown feedback: https://huggingface.co/HuggingFaceM4/idefics2-8b/discussions - type: dataset name: The Cauldron organization: Hugging Face - description: The Cauldron is an open compilation of 50 manually-curated datasets formatted for multi-turn conversations. + description: The Cauldron is an open compilation of 50 manually-curated datasets + formatted for multi-turn conversations. created_date: 2024-04-15 url: https://huggingface.co/blog/idefics2 datasheet: https://huggingface.co/datasets/HuggingFaceM4/the_cauldron @@ -203,8 +215,9 @@ size: 50 vision-language datasets sample: [] analysis: none - dependencies: - explanation: These are the datasets with the most tokens included; the full list of all 50 datasets can be found at https://huggingface.co/datasets/HuggingFaceM4/the_cauldron + dependencies: + explanation: These are the datasets with the most tokens included; the full + list of all 50 datasets can be found at https://huggingface.co/datasets/HuggingFaceM4/the_cauldron value: [LNarratives, Rendered Text, WebSight, DaTikz] included: '' excluded: '' diff --git a/assets/konan.yaml b/assets/konan.yaml index 55d90d11..144b051d 100644 --- a/assets/konan.yaml +++ b/assets/konan.yaml @@ -2,7 +2,9 @@ - type: model name: Konan LLM organization: Konan - description: Konan LLM is a Large Language Model developed in-house by Konan Technology. Optimized for super-large AI training, it leverages high-quality, large-scale data and over 20 years of expertise in natural language processing. + description: Konan LLM is a Large Language Model developed in-house by Konan Technology. + Optimized for super-large AI training, it leverages high-quality, large-scale + data and over 20 years of expertise in natural language processing. created_date: 2023-09-17 url: https://en.konantech.com/en/llm/konanllm model_card: none @@ -19,4 +21,4 @@ intended_uses: Document generation, document review, Q&A, customer response scenarios. prohibited_uses: '' monitoring: '' - feedback: none \ No newline at end of file + feedback: none diff --git a/assets/ktai.yaml b/assets/ktai.yaml index 56251384..5b5e0f1e 100644 --- a/assets/ktai.yaml +++ b/assets/ktai.yaml @@ -2,18 +2,23 @@ - type: model name: Midm organization: KT Corporation - description: Midm is a pre-trained Korean-English language model developed by KT. It takes text as input and creates text. The model is based on Transformer architecture for an auto-regressive language model. + description: Midm is a pre-trained Korean-English language model developed by + KT. It takes text as input and creates text. The model is based on Transformer + architecture for an auto-regressive language model. created_date: 2023-10-31 url: https://huggingface.co/KT-AI/midm-bitext-S-7B-inst-v1 model_card: https://huggingface.co/KT-AI/midm-bitext-S-7B-inst-v1 modality: text; text analysis: unknown size: 7B parameters - dependencies: [AI-HUB dataset, National Institute of Korean Language dataset] + dependencies: + - AI-HUB dataset + - National Institute of Korean Language dataset training_emissions: unknown training_time: unknown training_hardware: unknown - quality_control: KT tried to remove unethical expressions such as profanity, slang, prejudice, and discrimination from training data. + quality_control: KT tried to remove unethical expressions such as profanity, slang, + prejudice, and discrimination from training data. access: open license: CC-BY-NC 4.0 intended_uses: It is expected to be used for various research purposes. diff --git a/assets/lg.yaml b/assets/lg.yaml index 3ec7981c..78ea5a80 100644 --- a/assets/lg.yaml +++ b/assets/lg.yaml @@ -2,7 +2,8 @@ - type: model name: EXAONE 2.0 organization: LG AI Research - description: EXAONE 2.0 is a multimodal artificial intelligence that can be used to help develop new materials and medicines. + description: EXAONE 2.0 is a multimodal artificial intelligence that can be used + to help develop new materials and medicines. created_date: 2023-07-19 url: https://www.lgresearch.ai/exaone model_card: none diff --git a/assets/llm360.yaml b/assets/llm360.yaml index 31846a65..770838f6 100644 --- a/assets/llm360.yaml +++ b/assets/llm360.yaml @@ -58,3 +58,29 @@ prohibited_uses: '' monitoring: unknown feedback: https://huggingface.co/LLM360/CrystalCoder/discussions +- type: model + name: K2 + organization: LLM360 + description: K2 is a 65 billion parameter large language model built upon the + Llama 2 70B model. The model is also supported with a suite of research tools, + tutorials and step-by-step guides for learning pre-training and fine-tuning + techniques. + created_date: 2024-05-29 + url: https://www.llm360.ai/paper2.pdf + model_card: https://huggingface.co/LLM360/K2 + modality: text; text + analysis: Evaluated on the LLM360 Performance and Evaluation Collection that checks + standard best practice benchmarks, medical, math, and coding knowledge. + size: 65B parameters + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: unknown + access: open + license: Apache 2.0 + intended_uses: The model is intended for learning pre-training techniques or enhancing + research capabilities in large language models. + prohibited_uses: unknown + monitoring: unknown + feedback: https://huggingface.co/LLM360/K2/discussions diff --git a/assets/meta.yaml b/assets/meta.yaml index a2908725..9b831737 100644 --- a/assets/meta.yaml +++ b/assets/meta.yaml @@ -797,23 +797,54 @@ - type: model name: Llama 3 organization: Meta - description: Llama 3 is the third generation of Meta AI's open-source large language model. It comes with pretrained and instruction-fine-tuned language models with 8B and 70B parameters that can support a broad range of use cases. + description: Llama 3 is the third generation of Meta AI's open-source large language + model. It comes with pretrained and instruction-fine-tuned language models with + 8B and 70B parameters that can support a broad range of use cases. created_date: 2024-04-18 url: https://llama.meta.com/llama3/ model_card: https://github.com/meta-llama/llama3/blob/main/MODEL_CARD.md modality: text; text - analysis: The models were evaluated based on their performance on standard benchmarks and real-world scenarios. These evaluations were performed using a high-quality human evaluation set containing 1,800 prompts covering multiple use cases. The models also went through red-teaming for safety, where human experts and automated methods were used to generate adversarial prompts to test for problematic responses. + analysis: The models were evaluated based on their performance on standard benchmarks + and real-world scenarios. These evaluations were performed using a high-quality + human evaluation set containing 1,800 prompts covering multiple use cases. The + models also went through red-teaming for safety, where human experts and automated + methods were used to generate adversarial prompts to test for problematic responses. size: 70B parameters dependencies: [] training_emissions: unknown training_time: unknown training_hardware: 2 custom-built Meta 24K GPU clusters - quality_control: Extensive internal and external testing for safety, and design of new trust and safety tools. + quality_control: Extensive internal and external testing for safety, and design + of new trust and safety tools. access: open license: explanation: Can be found at https://github.com/meta-llama/llama3/blob/main/LICENSE value: Llama 3 - intended_uses: Llama 3 is intended for a broad range of use cases, including AI assistance, content creation, learning, and analysis. + intended_uses: Llama 3 is intended for a broad range of use cases, including AI + assistance, content creation, learning, and analysis. prohibited_uses: unknown - monitoring: Extensive internal and external performance evaluation and red-teaming approach for safety testing. - feedback: Feedback is encouraged from users to improve the model, but the feedback mechanism is not explicitly described. + monitoring: Extensive internal and external performance evaluation and red-teaming + approach for safety testing. + feedback: Feedback is encouraged from users to improve the model, but the feedback + mechanism is not explicitly described. +- type: model + name: Chameleon + organization: Meta FAIR + description: Chameleon is a family of early-fusion token-based mixed-modal models capable of understanding and generating images and text in any arbitrary sequence. + created_date: 2024-05-17 + url: https://arxiv.org/pdf/2405.09818 + model_card: none + modality: image, text; image, text + analysis: Evaluated on a comprehensive range of tasks, including visual question answering, image captioning, text generation, image generation, and long-form mixed modal generation. Chameleon demonstrates broad and general capabilities, including state-of-the-art performance in image captioning tasks, outperforms Llama-2 in text-only tasks while being competitive with models such as Mixtral 8x7B and Gemini-Pro. + size: 34B parameters + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: Meta's Research Super Cluster (powered by NVIDIA A100 80GB GPUs) + quality_control: '' + access: open + license: unknown + intended_uses: '' + prohibited_uses: '' + monitoring: '' + feedback: none diff --git a/assets/microsoft.yaml b/assets/microsoft.yaml index 646bc637..eb538758 100644 --- a/assets/microsoft.yaml +++ b/assets/microsoft.yaml @@ -913,23 +913,76 @@ monitoring: unknown feedback: https://huggingface.co/microsoft/Orca-2-13b/discussions - type: model - name: Phi-3 Mini + name: Phi-3 organization: Microsoft - description: Phi-3 Mini is a 3.8 billion-parameter, lightweight, state-of-the-art open model trained using the Phi-3 datasets. - created_date: 2024-04-23 - url: https://azure.microsoft.com/en-us/blog/introducing-phi-3-redefining-whats-possible-with-slms/ - model_card: https://huggingface.co/microsoft/Phi-3-mini-128k-instruct + description: Phi-3 is a 14 billion-parameter, lightweight, state-of-the-art open + model trained using the Phi-3 datasets. + created_date: 2024-05-21 + url: https://arxiv.org/abs/2404.14219 + model_card: https://huggingface.co/microsoft/Phi-3-medium-128k-instruct modality: text; text - analysis: The model has been evaluated against benchmarks that test common sense, language understanding, mathematics, coding, long-term context, and logical reasoning. The Phi-3 Mini-128K-Instruct demonstrated robust and state-of-the-art performance among models with fewer than 13 billion parameters. - size: 3.8B parameters + analysis: The model has been evaluated against benchmarks that test common sense, + language understanding, mathematics, coding, long-term context, and logical + reasoning. The Phi-3 Medium-128K-Instruct demonstrated robust and state-of-the-art + performance. + size: 14B parameters dependencies: [] training_emissions: unknown - training_time: 7 days - training_hardware: 512 H100-80G GPUs - quality_control: The model underwent post-training processes viz. supervised fine-tuning and direct preference optimization to increase its capability in following instructions and aligning to safety measures. + training_time: unknown + training_hardware: unknown + quality_control: The model underwent post-training processes viz. supervised fine-tuning + and direct preference optimization to increase its capability in following instructions + and aligning to safety measures. access: open license: MIT - intended_uses: The model's primary use cases are for commercial and research purposes that require capable reasoning in memory or compute constrained environments and latency-bound scenarios. It can also serve as a building block for generative AI-powered features. - prohibited_uses: The model should not be used for high-risk scenarios without adequate evaluation and mitigation techniques for accuracy, safety, and fairness. - monitoring: Issues like allocation, high-risk scenarios, misinformation, generation of harmful content and misuse should be monitored and addressed. - feedback: https://huggingface.co/microsoft/Phi-3-mini-128k-instruct/discussions + intended_uses: The model's primary use cases are for commercial and research purposes + that require capable reasoning in memory or compute constrained environments + and latency-bound scenarios. It can also serve as a building block for generative + AI-powered features. + prohibited_uses: The model should not be used for high-risk scenarios without + adequate evaluation and mitigation techniques for accuracy, safety, and fairness. + monitoring: Issues like allocation, high-risk scenarios, misinformation, generation + of harmful content and misuse should be monitored and addressed. + feedback: https://huggingface.co/microsoft/Phi-3-medium-128k-instruct/discussions +- type: model + name: Aurora + organization: Microsoft + description: Aurora is a large-scale foundation model of the atmosphere trained on over a million hours of diverse weather and climate data. + created_date: 2024-05-28 + url: https://arxiv.org/pdf/2405.13063 + model_card: none + modality: text; climate forecasts + analysis: Evaluated by comparing climate predictions to actual happened events. + size: 1.3B parameters + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: 32 A100 GPUs + quality_control: '' + access: closed + license: unknown + intended_uses: '' + prohibited_uses: '' + monitoring: '' + feedback: none +- type: model + name: Prov-GigaPath + organization: Microsoft + description: Prov-GigaPath is a whole-slide pathology foundation model pretrained on 1.3 billion 256 × 256 pathology image tiles. + created_date: 2024-05-22 + url: https://www.nature.com/articles/s41586-024-07441-w + model_card: none + modality: image; embeddings + analysis: Evaluated on a digital pathology benchmark comprising 9 cancer subtyping tasks and 17 pathomics tasks, with Prov-GigaPath demonstrating SoTA performance in 25 out of 26 tasks. + size: unknown + dependencies: [GigaPath] + training_emissions: unknown + training_time: 2 days + training_hardware: 4 80GB A100 GPUs + quality_control: '' + access: closed + license: unknown + intended_uses: '' + prohibited_uses: '' + monitoring: '' + feedback: none diff --git a/assets/mistral.yaml b/assets/mistral.yaml index cbe8263b..f8fd24ad 100644 --- a/assets/mistral.yaml +++ b/assets/mistral.yaml @@ -64,3 +64,24 @@ monthly_active_users: unknown user_distribution: unknown failures: unknown +- type: model + name: Codestral + organization: Mistral AI + description: Codestral is an open-weight generative AI model explicitly designed for code generation tasks. It helps developers write and interact with code through a shared instruction and completion API endpoint. Mastering code and English, it can be used to design advanced AI applications for software developers. It is fluent in 80+ programming languages. + created_date: 2024-05-29 + url: https://mistral.ai/news/codestral/ + model_card: none + modality: text; code + analysis: Performance of Codestral is evaluated in Python, SQL, and additional languages, C++, bash, Java, PHP, Typescript, and C#. Fill-in-the-middle performance is assessed using HumanEval pass@1 in Python, JavaScript, and Java. + size: 22B parameters + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: '' + access: open + license: Mistral AI Non-Production License + intended_uses: Helps developers write and interact with code, design advanced AI applications for software developers, integrated into LlamaIndex and LangChain for building applications, integrated in VSCode and JetBrains environments for code generation and interactive conversation. + prohibited_uses: unknown + monitoring: unknown + feedback: none diff --git a/assets/naver.yaml b/assets/naver.yaml index 37cb740f..a69a3f1e 100644 --- a/assets/naver.yaml +++ b/assets/naver.yaml @@ -28,12 +28,16 @@ - type: model name: HyperCLOVA X organization: NAVER - description: HyperCLOVA X is a family of large language models (LLMs) tailored to the Korean language and culture, along with competitive capabilities in English, math, and coding. + description: HyperCLOVA X is a family of large language models (LLMs) tailored + to the Korean language and culture, along with competitive capabilities in English, + math, and coding. created_date: 2024-04-13 url: https://arxiv.org/pdf/2404.01954 model_card: none modality: text; text - analysis: Evaluated on English and Korean benchmarks in comparison to open source English and multilingual LLMs, with HyperCLOVA X (closed) surpassing the models compared. + analysis: Evaluated on English and Korean benchmarks in comparison to open source + English and multilingual LLMs, with HyperCLOVA X (closed) surpassing the models + compared. size: unknown dependencies: [] training_emissions: unknown diff --git a/assets/ncsoft.yaml b/assets/ncsoft.yaml index 2bc86592..6cd7ef44 100644 --- a/assets/ncsoft.yaml +++ b/assets/ncsoft.yaml @@ -2,12 +2,14 @@ - type: model name: VARCO-LLM organization: NCSOFT - description: VARCO-LLM is NCSOFT’s large language model and is trained on English and Korean. + description: VARCO-LLM is NCSOFT’s large language model and is trained on English + and Korean. created_date: 2023-08-16 url: https://github.com/ncsoft/ncresearch model_card: none modality: text; text - analysis: Boasts the highest performance among the Korean LLMs of similar sizes that have been released to date, according to internal evaluations. + analysis: Boasts the highest performance among the Korean LLMs of similar sizes + that have been released to date, according to internal evaluations. size: 13B parameters dependencies: [] training_emissions: unknown @@ -18,7 +20,8 @@ license: explanation: Can be found at https://github.com/ncsoft/ncresearch/blob/main/LICENSE.txt value: custom - intended_uses: Developing various NLP-based AI services such as Q&A, chatbot, summarization, information extraction + intended_uses: Developing various NLP-based AI services such as Q&A, chatbot, + summarization, information extraction prohibited_uses: '' monitoring: '' feedback: none diff --git a/assets/openai.yaml b/assets/openai.yaml index 77331e3c..b32ee878 100644 --- a/assets/openai.yaml +++ b/assets/openai.yaml @@ -1421,3 +1421,24 @@ prohibited_uses: '' monitoring: unknown feedback: '' +- type: model + name: GPT-4o + organization: OpenAI + description: GPT-4o is OpenAI's new flagship model, as of release, that can reason across audio, vision, and text in real time. + created_date: 2024-05-13 + url: https://openai.com/index/hello-gpt-4o/ + model_card: none + modality: audio, text; image, text, video + analysis: When evaluated on standard performance benchmarks, achieves similar levels of performance to GPT-4 Turbo. + size: unknown + dependencies: [] + training_emissions: unknown + training_time: unknown + training_hardware: unknown + quality_control: Training data filtering and post-training refinement act as additional guardrails for preventing harmful outputs. + access: limited + license: unknown + intended_uses: '' + prohibited_uses: '' + monitoring: unknown + feedback: none diff --git a/assets/openbmb.yaml b/assets/openbmb.yaml index c961d875..4633c13b 100644 --- a/assets/openbmb.yaml +++ b/assets/openbmb.yaml @@ -79,7 +79,9 @@ url: https://arxiv.org/abs/2404.02078 model_card: https://huggingface.co/openbmb/Eurus-70b-nca modality: text; text - analysis: The model was comprehensively benchmarked across 12 tests covering five tasks. Eurus achieved the best overall performance among open-source models of similar sizes and even outperformed specialized models in many cases. + analysis: The model was comprehensively benchmarked across 12 tests covering five + tasks. Eurus achieved the best overall performance among open-source models + of similar sizes and even outperformed specialized models in many cases. size: 70B parameters dependencies: [Eurus SFT, UltraInteract, UltraFeedback] training_emissions: unknown @@ -88,7 +90,8 @@ quality_control: none access: open license: Apache 2.0 - intended_uses: The model can be used for reasoning tasks and is especially tailored for coding and math following specific prompts. + intended_uses: The model can be used for reasoning tasks and is especially tailored + for coding and math following specific prompts. prohibited_uses: none monitoring: unknown feedback: https://huggingface.co/openbmb/Eurus-70b-nca/discussions diff --git a/assets/reka.yaml b/assets/reka.yaml index 7dd3f20c..bee20908 100644 --- a/assets/reka.yaml +++ b/assets/reka.yaml @@ -24,12 +24,18 @@ - type: model name: Reka Core organization: Reka - description: Reka Core is a frontier-class multimodal language model comparable to industry leaders. It has powerful capabilities including multimodal understanding (including images, videos, and audio), superb reasoning abilities, code generation, and multilinguality with proficiency in 32 languages. + description: Reka Core is a frontier-class multimodal language model comparable + to industry leaders. It has powerful capabilities including multimodal understanding + (including images, videos, and audio), superb reasoning abilities, code generation, + and multilinguality with proficiency in 32 languages. created_date: 2024-04-15 url: https://www.reka.ai/news/reka-core-our-frontier-class-multimodal-language-model model_card: none modality: audio, image, text, video; text - analysis: Reka Core was evaluated against leading models such as OpenAIs GPT-4, Claude-3 Opus, and Gemini Ultra on a variety of tasks and metrics including multimodal and human evaluation conducted by a third party. It was found to be competitive or even surpassing these models. + analysis: Reka Core was evaluated against leading models such as OpenAIs GPT-4, + Claude-3 Opus, and Gemini Ultra on a variety of tasks and metrics including + multimodal and human evaluation conducted by a third party. It was found to + be competitive or even surpassing these models. size: unknown dependencies: [] training_emissions: unknown @@ -38,7 +44,9 @@ quality_control: '' access: limited license: unknown - intended_uses: Reka Core can be used in e-commerce, social media, digital content and video games, healthcare, robotics, and other industries for tasks that require multimodal understanding, coding, complex reasoning, and more. + intended_uses: Reka Core can be used in e-commerce, social media, digital content + and video games, healthcare, robotics, and other industries for tasks that require + multimodal understanding, coding, complex reasoning, and more. prohibited_uses: unknown monitoring: unknown feedback: unknown diff --git a/assets/shanghai.yaml b/assets/shanghai.yaml index c549fbfe..57ad187a 100644 --- a/assets/shanghai.yaml +++ b/assets/shanghai.yaml @@ -130,21 +130,28 @@ - type: model name: CosmicMan organization: Shanghai AI Laboratory - description: CosmicMan is a text-to-image foundation model specialized for generating high-fidelity human images with meticulous appearance, reasonable structure, and precise text-image alignment. + description: CosmicMan is a text-to-image foundation model specialized for generating + high-fidelity human images with meticulous appearance, reasonable structure, + and precise text-image alignment. created_date: 2024-04-01 url: https://cosmicman-cvpr2024.github.io/ model_card: none modality: text; image - analysis: The model was compared with SOTAs and has shown good performance in generating high-quality human images. + analysis: The model was compared with SOTAs and has shown good performance in + generating high-quality human images. size: unknown dependencies: [CosmicMan-HQ 1.0] training_emissions: unknown training_time: 1 week training_hardware: 32 80G NVIDIA A100 GPUs - quality_control: The quality control measures taken include modeling the relationship between dense text descriptions and image pixels in a decomposed manner and enforcing attention refocusing without adding extra modules. + quality_control: The quality control measures taken include modeling the relationship + between dense text descriptions and image pixels in a decomposed manner and + enforcing attention refocusing without adding extra modules. access: open license: unknown - intended_uses: The model is intended to generate high-quality, photorealistic human images from text descriptions. Applications include avatar generation and potentially virtual reality and video game character creation. + intended_uses: The model is intended to generate high-quality, photorealistic + human images from text descriptions. Applications include avatar generation + and potentially virtual reality and video game character creation. prohibited_uses: unknown monitoring: unknown feedback: unknown @@ -152,14 +159,16 @@ - type: dataset name: CosmicMan-HQ 1.0 organization: Shanghai AI Laboratory - description: CosmicMan-HQ 1.0 is a large-scale dataset with 6 million high-quality, real-world human images. + description: CosmicMan-HQ 1.0 is a large-scale dataset with 6 million high-quality, + real-world human images. created_date: 2024-04-28 url: https://arxiv.org/pdf/2404.01294 datasheet: none modality: image size: 6 million images sample: [] - analysis: Compared to other human image datasets on data quantity, image quality, and annotations. + analysis: Compared to other human image datasets on data quantity, image quality, + and annotations. dependencies: [] included: '' excluded: '' diff --git a/assets/skt.yaml b/assets/skt.yaml index b2c2d46a..8e942647 100644 --- a/assets/skt.yaml +++ b/assets/skt.yaml @@ -2,7 +2,8 @@ - type: model name: A.X organization: SK Telecom - description: A.X is SK Telecom's proprietary LLM, which has been trained on the Korean language. + description: A.X is SK Telecom's proprietary LLM, which has been trained on the + Korean language. created_date: 2023-09-26 url: https://www.sktelecom.com/en/press/press_detail.do?idx=1582 model_card: none @@ -19,4 +20,4 @@ intended_uses: '' prohibited_uses: '' monitoring: '' - feedback: none \ No newline at end of file + feedback: none diff --git a/assets/stability.yaml b/assets/stability.yaml index 128631bf..29f128b8 100644 --- a/assets/stability.yaml +++ b/assets/stability.yaml @@ -267,7 +267,9 @@ - type: model name: Stable Audio 2.0 organization: Stability AI - description: Stable Audio 2.0 sets a new standard in AI-generated audio, producing high-quality, full tracks with coherent musical structure up to three minutes in length at 44.1kHz stereo. + description: Stable Audio 2.0 sets a new standard in AI-generated audio, producing + high-quality, full tracks with coherent musical structure up to three minutes + in length at 44.1kHz stereo. created_date: 2024-04-03 url: https://stability-ai.squarespace.com/news/stable-audio-2-0 model_card: none @@ -278,10 +280,15 @@ training_emissions: unknown training_time: unknown training_hardware: unknown - quality_control: To protect creator copyrights, for audio uploads, Stability AI partners with Audible Magic to use their content recognition (ACR) technology to power real-time content matching and prevent copyright infringement. Opt-out requests were honored during the training phase. + quality_control: To protect creator copyrights, for audio uploads, Stability AI + partners with Audible Magic to use their content recognition (ACR) technology + to power real-time content matching and prevent copyright infringement. Opt-out + requests were honored during the training phase. access: open license: unknown - intended_uses: It can be used to generate melodies, backing tracks, stems, and sound effects. + intended_uses: It can be used to generate melodies, backing tracks, stems, and + sound effects. prohibited_uses: Uploading copyrighted material for transformation. - monitoring: Advanced content recognition is used to maintain compliance and prevent copyright infringement. + monitoring: Advanced content recognition is used to maintain compliance and prevent + copyright infringement. feedback: none diff --git a/assets/tokyo.yaml b/assets/tokyo.yaml index 7930e47e..6084473a 100644 --- a/assets/tokyo.yaml +++ b/assets/tokyo.yaml @@ -1,17 +1,21 @@ --- - type: model name: Aurora-M - organization: Tokyo Institute of Technology, MIT-IBM Watson Lab, Sapienza University of Rome - description: Aurora-M is a 15B parameter multilingual open-source model trained on English, Finnish, Hindi, Japanese, Vietnamese, and code. + organization: Tokyo Institute of Technology, MIT-IBM Watson Lab, Sapienza University + of Rome + description: Aurora-M is a 15B parameter multilingual open-source model trained + on English, Finnish, Hindi, Japanese, Vietnamese, and code. created_date: 2024-04-23 url: https://arxiv.org/pdf/2404.00399 model_card: none modality: text; text - analysis: Evaluated on all language datasets compared to similarly sized SOTA models, with Aurora-M achieving strong performance in most. + analysis: Evaluated on all language datasets compared to similarly sized SOTA + models, with Aurora-M achieving strong performance in most. size: 15B parameters dependencies: [StarCoderPlus] training_emissions: - explanation: The training process operated entirely on 100% hydro-powered energy and included waste heat recycling. + explanation: The training process operated entirely on 100% hydro-powered energy + and included waste heat recycling. value: unknown training_time: 48 days training_hardware: LUMI supercomputer, using 128 AMD MI250X GPUs @@ -21,4 +25,4 @@ intended_uses: '' prohibited_uses: '' monitoring: unknown - feedback: none \ No newline at end of file + feedback: none diff --git a/assets/xai.yaml b/assets/xai.yaml index 14dc7b20..b656754a 100644 --- a/assets/xai.yaml +++ b/assets/xai.yaml @@ -26,12 +26,18 @@ - type: model name: Grok-1.5V organization: xAI - description: Grok-1.5V is a first-generation multimodal model which can process a wide variety of visual information, including documents, diagrams, charts, screenshots, and photographs. + description: Grok-1.5V is a first-generation multimodal model which can process + a wide variety of visual information, including documents, diagrams, charts, + screenshots, and photographs. created_date: 2024-04-12 url: https://x.ai/blog/grok-1.5v model_card: none modality: image, text; text - analysis: The model is evaluated in a zero-shot setting without chain-of-thought prompting. The evaluation domains include multi-disciplinary reasoning, understanding documents, science diagrams, charts, screenshots, photographs and real-world spatial understanding. The model shows competitive performance with existing frontier multimodal models. + analysis: The model is evaluated in a zero-shot setting without chain-of-thought + prompting. The evaluation domains include multi-disciplinary reasoning, understanding + documents, science diagrams, charts, screenshots, photographs and real-world + spatial understanding. The model shows competitive performance with existing + frontier multimodal models. size: unknown dependencies: [] training_emissions: unknown @@ -40,7 +46,9 @@ quality_control: '' access: limited license: unknown - intended_uses: Grok-1.5V can be used for understanding documents, science diagrams, charts, screenshots, photographs. It can also translate diagrams into Python code. + intended_uses: Grok-1.5V can be used for understanding documents, science diagrams, + charts, screenshots, photographs. It can also translate diagrams into Python + code. prohibited_uses: unknown monitoring: unknown feedback: none diff --git a/js/main.js b/js/main.js index 03c72139..6b7611e5 100644 --- a/js/main.js +++ b/js/main.js @@ -677,6 +677,7 @@ function loadAssetsAndRenderPageContent() { 'assets/transformify.yaml', 'assets/paladin.yaml', 'assets/01ai.yaml', + 'assets/360.yaml', 'assets/ai2.yaml', 'assets/ai21.yaml', 'assets/aleph_alpha.yaml', @@ -694,6 +695,7 @@ function loadAssetsAndRenderPageContent() { 'assets/bloomberg.yaml', 'assets/brex.yaml', 'assets/cagliostro.yaml', + 'assets/cartesia.yaml', 'assets/causallm.yaml', 'assets/cerebras.yaml', 'assets/character.yaml',