Scheduled Commit

data/retrieval_battle-8e53f2d0-435d-4bdf-8b8a-f8c64f4908ac.jsonl

@@ -2,3 +2,4 @@
 {"tstamp": 1731367360.6292, "task_type": "retrieval", "type": "tievote", "models": ["", ""], "ip": "", "0_conv_id": "0a0c5041f2014d03bcac56e7ff012679", "0_model_name": "mixedbread-ai/mxbai-embed-large-v1", "0_prompt": "who does marge's voice on the simpsons", "0_output": [["who does marge's voice on the simpsons", "Bart Simpson\n\nVoice Bart's voice is provided by Nancy Cartwright, who voices several other child characters on The Simpsons, including Nelson Muntz, Ralph Wiggum, Todd Flanders, and Kearney. While the roles of Homer and Marge were given to Dan Castellaneta and Julie Kavner because they were already a part of The Tracey Ullman Show cast, the producers decided to hold casting for the roles of Bart and Lisa. Yeardley Smith had initially been asked to audition for the role of Bart, but casting director Bonita Pietila believed her voice was too high. Smith later recalled, \"I always sounded too much like a girl. I read two lines as Bart and they said, 'Thanks for coming! Smith was given the role of Lisa instead. On March 13, 1987, Nancy Cartwright went in to audition for the role of Lisa. After arriving at the audition, she found that Lisa was simply described as the \"middle child\" and at the time did not have much personality. Cartwright became more interested in the role of Bart, who was described as \"devious, underachieving, school-hating, irreverent, [and] clever\". Matt Groening let her try out for the part instead, and upon hearing her read, gave her the job on the spot. Cartwright is the only one of the six main Simpsons cast members who had been professionally trained in voice acting prior to working on the show."]], "0_corpus": "wikipedia", "1_conv_id": "c1a500138c13415da4eb79a32b54060b", "1_model_name": "nomic-ai/nomic-embed-text-v1.5", "1_prompt": "who does marge's voice on the simpsons", "1_output": [["who does marge's voice on the simpsons", "Bart Simpson\n\nVoice Bart's voice is provided by Nancy Cartwright, who voices several other child characters on The Simpsons, including Nelson Muntz, Ralph Wiggum, Todd Flanders, and Kearney. While the roles of Homer and Marge were given to Dan Castellaneta and Julie Kavner because they were already a part of The Tracey Ullman Show cast, the producers decided to hold casting for the roles of Bart and Lisa. Yeardley Smith had initially been asked to audition for the role of Bart, but casting director Bonita Pietila believed her voice was too high. Smith later recalled, \"I always sounded too much like a girl. I read two lines as Bart and they said, 'Thanks for coming! Smith was given the role of Lisa instead. On March 13, 1987, Nancy Cartwright went in to audition for the role of Lisa. After arriving at the audition, she found that Lisa was simply described as the \"middle child\" and at the time did not have much personality. Cartwright became more interested in the role of Bart, who was described as \"devious, underachieving, school-hating, irreverent, [and] clever\". Matt Groening let her try out for the part instead, and upon hearing her read, gave her the job on the spot. Cartwright is the only one of the six main Simpsons cast members who had been professionally trained in voice acting prior to working on the show."]], "1_corpus": "wikipedia"}
 {"tstamp": 1731367631.8702, "task_type": "retrieval", "type": "leftvote", "models": ["", ""], "ip": "", "0_conv_id": "09aca70a5d704c30bd418b7da4b48441", "0_model_name": "Salesforce/SFR-Embedding-2_R", "0_prompt": "how to play music from your iphone on your laptop?", "0_output": [["how to play music from your iphone on your laptop?", "Apple Stackexchange\n\nQ: Can I play audio from my iPhone on my Mac? Can I attach my iPhone 3GS to my Mac Pro so that the iPhone's audio plays through the Mac Pro's speakers?\nI can easily attach my Bose headphones so that the music from the iPhone plays through them… that's cool… but I want others to hear the music through my Mac Pro's speaker system.\n\nA: You can play your iPhone's music out your mac via iTunes. You don't even have to be in sync with that iTunes library, just plug it in, navigate to your music, and press play.\nIf you're talking about like game audio coming out your mac's speakers... I'd take your speakers and plug them straight into the phone. There's no way I know of to use the mac as an audio route from the phone.\n"]], "0_corpus": "stackexchange", "1_conv_id": "8741ba6643fc4a229d510014230bc7bf", "1_model_name": "BM25", "1_prompt": "how to play music from your iphone on your laptop?", "1_output": [["how to play music from your iphone on your laptop?", "Apple Stackexchange\n\nQ: How can I remote control music on my computer from my iPhone? I want to change, play or pause my music using my iPhone while it is playing on my laptop.\n\nA: Get Remote from the App Store which lets you control iTunes playback on your computer.\n"]], "1_corpus": "stackexchange"}
 {"tstamp": 1731367935.9788, "task_type": "retrieval", "type": "leftvote", "models": ["", ""], "ip": "", "0_conv_id": "75510c169b644b06bdc3aff96c7d3a2d", "0_model_name": "GritLM/GritLM-7B", "0_prompt": "I'm searching for a paper that first introduced semantic textual similarity task.", "0_output": [["I'm searching for a paper that first introduced semantic textual similarity task.", "Title: SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluation\n\nAbstract: Semantic Textual Similarity (STS) measures the meaning similarity of sentences. Applications include machine translation (MT), summarization, generation, question answering (QA), short answer grading, semantic search, dialog and conversational systems. The STS shared task is a venue for assessing the current state-of-the-art. The 2017 task focuses on multilingual and cross-lingual pairs with one sub-track exploring MT quality estimation (MTQE) data. The task obtained strong participation from 31 teams, with 17 participating in all language tracks. We summarize performance and review a selection of well performing methods. Analysis highlights common errors, providing insight into the limitations of existing models. To support ongoing work on semantic representations, the STS Benchmark is introduced as a new shared training and evaluation set carefully selected from the corpus of English STS shared task data (2012-2017)."]], "0_corpus": "arxiv", "1_conv_id": "02ff720f6d904d67a30f5b9395c0791b", "1_model_name": "mixedbread-ai/mxbai-embed-large-v1", "1_prompt": "I'm searching for a paper that first introduced semantic textual similarity task.", "1_output": [["I'm searching for a paper that first introduced semantic textual similarity task.", "Title: A Massive Scale Semantic Similarity Dataset of Historical English\n\nAbstract: A diversity of tasks use language models trained on semantic similarity data. While there are a variety of datasets that capture semantic similarity, they are either constructed from modern web data or are relatively small datasets created in the past decade by human annotators. This study utilizes a novel source, newly digitized articles from off-copyright, local U.S. newspapers, to assemble a massive-scale semantic similarity dataset spanning 70 years from 1920 to 1989 and containing nearly 400M positive semantic similarity pairs. Historically, around half of articles in U.S. local newspapers came from newswires like the Associated Press. While local papers reproduced articles from the newswire, they wrote their own headlines, which form abstractive summaries of the associated articles. We associate articles and their headlines by exploiting document layouts and language understanding. We then use deep neural methods to detect which articles are from the same underlying source, in the presence of substantial noise and abridgement. The headlines of reproduced articles form positive semantic similarity pairs. The resulting publicly available HEADLINES dataset is significantly larger than most existing semantic similarity datasets and covers a much longer span of time. It will facilitate the application of contrastively trained semantic similarity models to a variety of tasks, including the study of semantic change across space and time."]], "1_corpus": "arxiv"}
+{"tstamp": 1731368282.7933, "task_type": "retrieval", "type": "rightvote", "models": ["", ""], "ip": "", "0_conv_id": "4fb31c293678458ba9ce6ca5219e1ba9", "0_model_name": "sentence-transformers/all-MiniLM-L6-v2", "0_prompt": "What does battleground states in U.S. politics mean?", "0_output": [["What does battleground states in U.S. politics mean?", "Partisan (military)\n\nA partisan is a member of a domestic irregular military force formed to oppose control of an area by a foreign power or by an army of occupation by some kind of insurgent activity.\nThe term can apply to the field element of resistance movements. The most common use in present parlance in several languages refers to occupation resistance fighters during World War II, especially under the Yugoslav partisan leader Josip Broz Tito.\nHistory before 1939\nThe initial concept of partisan warfare involved the use of troops raised from the local population in a war zone (or in some cases regular forces) who would operate behind enemy lines to disrupt communications, seize posts or villages as forward-operating bases, ambush convoys, impose war taxes or contributions, raid logistical stockpiles, and compel enemy forces to disperse and protect their base of operations.\nGeorge Satterfield has analyzed the \"partisan warfare\" () in the Netherlands campaigns of 1673-1678 during the Franco-Dutch War of 1672-1678. Some of the practices of ninjas in feudal Japan resembled irregular partisan warfare."]], "0_corpus": "wikipedia", "1_conv_id": "f2d5a946c09d4727bc049967a684f38b", "1_model_name": "intfloat/e5-mistral-7b-instruct", "1_prompt": "What does battleground states in U.S. politics mean?", "1_output": [["What does battleground states in U.S. politics mean?", "Swing state\n\nIn American politics, a swing state (also known as battleground state, toss-up state, or purple state) is any state that could reasonably be won by either the Democratic or Republican candidate in a statewide election, most often referring to presidential elections, by a swing in votes. These states are usually targeted by both major-party campaigns, especially in competitive elections. Meanwhile, the states that regularly lean to a single party are known as \"safe states\" (or more specifically as \"red states\" and \"blue states\" depending on the partisan leaning), as it is generally assumed that one candidate has a base of support from which a sufficient share of the electorate can be drawn without significant investment or effort by the campaign.\nDue to the winner-take-all method most states use to determine their presidential electors, candidates often campaign only in competitive states, which is why a select group of states frequently receives a majority of the advertisements and candidate visits. The battlegrounds may change in certain election cycles and may be reflected in overall polling, demographics, and the ideological appeal of the nominees."]], "1_corpus": "wikipedia"}

data/retrieval_individual-8e53f2d0-435d-4bdf-8b8a-f8c64f4908ac.jsonl

@@ -6,3 +6,5 @@
 {"tstamp": 1731367583.5833, "task_type": "retrieval", "type": "chat", "model": "BM25", "gen_params": {}, "start": 1731367581.7993, "finish": 1731367583.5833, "ip": "", "conv_id": "8741ba6643fc4a229d510014230bc7bf", "model_name": "BM25", "prompt": "how to play music from your iphone on your laptop?", "output": [["how to play music from your iphone on your laptop?", "Apple Stackexchange\n\nQ: How can I remote control music on my computer from my iPhone? I want to change, play or pause my music using my iPhone while it is playing on my laptop.\n\nA: Get Remote from the App Store which lets you control iTunes playback on your computer.\n"]], "corpus": "stackexchange"}
 {"tstamp": 1731367874.9055, "task_type": "retrieval", "type": "chat", "model": "GritLM/GritLM-7B", "gen_params": {}, "start": 1731367874.6341, "finish": 1731367874.9055, "ip": "", "conv_id": "75510c169b644b06bdc3aff96c7d3a2d", "model_name": "GritLM/GritLM-7B", "prompt": "I'm searching for a paper that first introduced semantic textual similarity task.", "output": [["I'm searching for a paper that first introduced semantic textual similarity task.", "Title: SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluation\n\nAbstract: Semantic Textual Similarity (STS) measures the meaning similarity of sentences. Applications include machine translation (MT), summarization, generation, question answering (QA), short answer grading, semantic search, dialog and conversational systems. The STS shared task is a venue for assessing the current state-of-the-art. The 2017 task focuses on multilingual and cross-lingual pairs with one sub-track exploring MT quality estimation (MTQE) data. The task obtained strong participation from 31 teams, with 17 participating in all language tracks. We summarize performance and review a selection of well performing methods. Analysis highlights common errors, providing insight into the limitations of existing models. To support ongoing work on semantic representations, the STS Benchmark is introduced as a new shared training and evaluation set carefully selected from the corpus of English STS shared task data (2012-2017)."]], "corpus": "arxiv"}
 {"tstamp": 1731367874.9055, "task_type": "retrieval", "type": "chat", "model": "mixedbread-ai/mxbai-embed-large-v1", "gen_params": {}, "start": 1731367874.6341, "finish": 1731367874.9055, "ip": "", "conv_id": "02ff720f6d904d67a30f5b9395c0791b", "model_name": "mixedbread-ai/mxbai-embed-large-v1", "prompt": "I'm searching for a paper that first introduced semantic textual similarity task.", "output": [["I'm searching for a paper that first introduced semantic textual similarity task.", "Title: A Massive Scale Semantic Similarity Dataset of Historical English\n\nAbstract: A diversity of tasks use language models trained on semantic similarity data. While there are a variety of datasets that capture semantic similarity, they are either constructed from modern web data or are relatively small datasets created in the past decade by human annotators. This study utilizes a novel source, newly digitized articles from off-copyright, local U.S. newspapers, to assemble a massive-scale semantic similarity dataset spanning 70 years from 1920 to 1989 and containing nearly 400M positive semantic similarity pairs. Historically, around half of articles in U.S. local newspapers came from newswires like the Associated Press. While local papers reproduced articles from the newswire, they wrote their own headlines, which form abstractive summaries of the associated articles. We associate articles and their headlines by exploiting document layouts and language understanding. We then use deep neural methods to detect which articles are from the same underlying source, in the presence of substantial noise and abridgement. The headlines of reproduced articles form positive semantic similarity pairs. The resulting publicly available HEADLINES dataset is significantly larger than most existing semantic similarity datasets and covers a much longer span of time. It will facilitate the application of contrastively trained semantic similarity models to a variety of tasks, including the study of semantic change across space and time."]], "corpus": "arxiv"}
+{"tstamp": 1731368271.4428, "task_type": "retrieval", "type": "chat", "model": "sentence-transformers/all-MiniLM-L6-v2", "gen_params": {}, "start": 1731368270.9168, "finish": 1731368271.4428, "ip": "", "conv_id": "4fb31c293678458ba9ce6ca5219e1ba9", "model_name": "sentence-transformers/all-MiniLM-L6-v2", "prompt": "What does battleground states in U.S. politics mean?", "output": [["What does battleground states in U.S. politics mean?", "Partisan (military)\n\nA partisan is a member of a domestic irregular military force formed to oppose control of an area by a foreign power or by an army of occupation by some kind of insurgent activity.\nThe term can apply to the field element of resistance movements. The most common use in present parlance in several languages refers to occupation resistance fighters during World War II, especially under the Yugoslav partisan leader Josip Broz Tito.\nHistory before 1939\nThe initial concept of partisan warfare involved the use of troops raised from the local population in a war zone (or in some cases regular forces) who would operate behind enemy lines to disrupt communications, seize posts or villages as forward-operating bases, ambush convoys, impose war taxes or contributions, raid logistical stockpiles, and compel enemy forces to disperse and protect their base of operations.\nGeorge Satterfield has analyzed the \"partisan warfare\" () in the Netherlands campaigns of 1673-1678 during the Franco-Dutch War of 1672-1678. Some of the practices of ninjas in feudal Japan resembled irregular partisan warfare."]], "corpus": "wikipedia"}
+{"tstamp": 1731368271.4428, "task_type": "retrieval", "type": "chat", "model": "intfloat/e5-mistral-7b-instruct", "gen_params": {}, "start": 1731368270.9168, "finish": 1731368271.4428, "ip": "", "conv_id": "f2d5a946c09d4727bc049967a684f38b", "model_name": "intfloat/e5-mistral-7b-instruct", "prompt": "What does battleground states in U.S. politics mean?", "output": [["What does battleground states in U.S. politics mean?", "Swing state\n\nIn American politics, a swing state (also known as battleground state, toss-up state, or purple state) is any state that could reasonably be won by either the Democratic or Republican candidate in a statewide election, most often referring to presidential elections, by a swing in votes. These states are usually targeted by both major-party campaigns, especially in competitive elections. Meanwhile, the states that regularly lean to a single party are known as \"safe states\" (or more specifically as \"red states\" and \"blue states\" depending on the partisan leaning), as it is generally assumed that one candidate has a base of support from which a sufficient share of the electorate can be drawn without significant investment or effort by the campaign.\nDue to the winner-take-all method most states use to determine their presidential electors, candidates often campaign only in competitive states, which is why a select group of states frequently receives a majority of the advertisements and candidate visits. The battlegrounds may change in certain election cycles and may be reflected in overall polling, demographics, and the ideological appeal of the nominees."]], "corpus": "wikipedia"}