Delete video

video/classification/k710/instructions.json

@@ -1,12 +0,0 @@
-[
-    "Watch the video, notice the main parts of the action, and pick the best category from the options given.",
-    "Look at the key points of the action in the video and choose the most fitting option from the categories provided.",
-    "Focus on the important details of the action and select the right category among the similar choices.",
-    "Observe the video, find the unique aspects of the action, and decide which category suits it best from the list.",
-    "Identify what makes the action stand out in the video and pick the most accurate category from the available options.",
-    "While viewing the video, recognize the crucial elements of the action and select the best-matching category from the choices.",
-    "Pay attention to the core features of the action in the video and choose the category that represents them correctly from the options.",
-    "Examine the video's action carefully, find its key components, and determine the closest category from the choices given.",
-    "Spot the essential qualities of the action in the video, then decide on the right category from the provided options.",
-    "Study the important traits of the action while watching the video, and choose the best-fitting category from the list."
-]

video/classification/k710/train_new.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:76b629410e702dafd49debcb59b9b20e7df639cb83ddb402f0311f11b513c5d0
-size 19863352

video/classification/ssv2/instructions.json

@@ -1,12 +0,0 @@
-[
-    "Watch the video closely, focusing on how the person interacts with the object. Identify the most accurate action category from the provided options.",
-    "Examine the video and pay attention to the human-object interaction taking place. Determine the best-matching action category from the choices given.",
-    "Observe the video carefully, noting the unique aspects of the person's actions with the object. Select the fitting category that describes their interaction from the available options.",
-    "Study the video, concentrating on the way the individual manipulates or interacts with the object. Choose the appropriate action category from the listed choices.",
-    "Analyze the video, highlighting the key components of the person's actions while handling the object. Pick the most suitable category that represents the interaction from the provided options.",
-    "Review the video, focusing on the nature of the interaction between the person and the object. Select the category that best characterizes this interaction from the given choices.",
-    "Inspect the video, paying particular attention to the behavior of the individual as they interact with the object. Choose the corresponding category that captures this action from the offered options.",
-    "Evaluate the video, observing the distinctive features of the human-object interaction. Determine the action category that aligns best with these features from the selection provided.",
-    "Assess the video, considering the essential details of the person's actions in relation to the object. Pick the relevant category from the list of options.",
-    "Scrutinize the video, taking note of the specific manner in which the person engages with the object. Select the most accurate category from the available choices."
-]

video/classification/ssv2/train_new.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:095c8dc1daa0e0df80127596f0b8e727093846e938365e7a3125e6fd72b2a386
-size 31467204

video/conversation/videochatgpt/train_full_flat_fix.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:0ae1c2e7bdb277ea7f7f900db983a61fd3ca393ad22c218921867a4394ff99fe
-size 45730260

video/reasoning/clevrer_mc/instructions.json

@@ -1,12 +0,0 @@
-[
-    "Watch the video and understand what's happening. Choose the most accurate option that explains why it happened, what might happen next, or an alternate scenario.",
-    "Look at the video closely and think about the events. Pick the best answer that either tells why things happened, what could happen later, or a different possible outcome.",
-    "Observe the video carefully, thinking about the actions that took place. Choose the correct option that explains causes, makes predictions, or considers 'what if' situations.",
-    "Check out the video and pay attention to the sequence of events. Select the right option that gives reasons, guesses what happens next, or suggests other possibilities.",
-    "Analyze the video to figure out what's going on. Choose the correct answer that either explains why, forecasts what comes next, or imagines a different situation.",
-    "Study the video to catch all the details. Pick the best response that either says why things happened, predicts future events, or imagines a different outcome.",
-    "Review the video to understand the flow of events. Choose the right response that explains the reason, predicts ensuing actions, or considers an alternative scenario.",
-    "Go through the video attentively to get the gist of the events. Select the correct response that either clarifies the cause, foretells the forthcoming happenings, or proposes a hypothetical situation.",
-    "Examine the video to grasp the progression of events. Pick the accurate answer that either describes the cause, anticipates the next step, or contemplates a 'what if' scenario.",
-    "Explore the video, focusing on the actions occurring. Choose the most accurate answer that gives reasons, predicts the next occurrence, or hypothesizes a different possibility."
-]

video/reasoning/clevrer_mc/train_new.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:721b84830d6f7d28b4767d886d2fa6fb0e55e8a6e0edb1572dd0107d1d5af684
-size 23466499

video/reasoning/clevrer_qa/instructions.json

@@ -1,12 +0,0 @@
-[
-    "Examine the video presented, noting object details and movements to provide an accurate response to the question.",
-    "View the provided video, observing colors, shapes, and object counts to correctly answer the associated question.",
-    "Inspect the video carefully, focusing on the number of objects and their features for an accurate answer.",
-    "Study the video, count the objects, take note of their color and shape, and use this information for a precise response.",
-    "Observe the displayed video, identifying the objects, their properties, and count to formulate an accurate response.",
-    "Watch the video closely, looking out for details like the number of objects and their characteristics for a correct answer.",
-    "Review the shown video, detailing observations on objects, their color, shape, and motion status for an accurate answer.",
-    "Take a look at the video, using the observed objects, their features, and counts to respond accurately.",
-    "Analyze the video in detail, counting the objects, and noting their color, shape, and motion status for an exact response.",
-    "Check the video, tracking objects, their properties, and numbers for a precise response."
-]

video/reasoning/clevrer_qa/train_new.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:982b87e2b59bcceb19494608558f28693db30f5396a3d1cff38de9b1c6d32873
-size 15052437

video/reasoning/next_qa/instructions.json

@@ -1,12 +0,0 @@
-[
-    "Watch the video carefully, noticing the cause and sequence of events, and then choose the best option for the given question.",
-    "Observe the video, think about the relationships between actions and consequences, as well as their timing, before picking the most suitable choice.",
-    "While watching the video, focus on understanding the reasons and chronology of events, then decide on the right answer among the choices.",
-    "As you study the video, pay attention to how actions lead to results and their order, and select the most fitting choice.",
-    "Consider the cause-and-effect relationships and the order of events in the video, then pick the most appropriate choice as your answer.",
-    "View the video, identify patterns or connections related to causality and timing, and choose the best option.",
-    "Dive into the video, concentrate on the motives behind the actions and their timing, and find the correct choice.",
-    "Keep in mind the causal and temporal aspects of the actions while watching the video, then determine the most accurate choice.",
-    "Approach the video by thinking about the reasons behind the actions and their order in time, and choose the most relevant option.",
-    "During your review of the video, think about how actions influence outcomes and progress over time, then make your decision by selecting the most suitable choice."
-]

video/reasoning/next_qa/train_new.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2461df7dfb760e81a071efca75e7f6d4c0ad08df7f3adbdd6b64f25c68e56a98
-size 15399626

video/vqa/webvid_qa/instructions.json

@@ -1 +0,0 @@
-["Review the given video and answer the question associated with its visual elements.", "Watch the provided video and offer an accurate response to the related question.", "Scrutinize the video carefully, identifying relevant details such as objects, colors, and quantities in order to address the linked question.", "Take a close look at the presented visuals and deliver a precise answer to the corresponding question.", "Observe the video attentively and accurately respond to the associated question.", "View the video attentively and provide a suitable answer to the posed question.", "Examine the video and approach the connected question with an informed response.", "Assess the displayed video and answer the subsequent question with accuracy.", "Consider the video content and deliver a relevant answer to the corresponding question.", "Go through the video, taking into account key aspects, and respond to the question."]

video/vqa/webvid_qa/train.json

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8f2bde1819d61ae1f0c69a159d6defa034e2d158f90f1660db86ff0567174861
-size 28743495