Open-world classification — GRPO experiments
Collection
GRPO training datasets for open-world image classification (Oxford-IIIT Pet): direct and structured-reasoning variants. From my lmms-owc work. • 2 items • Updated
Datasets:
data_source stringclasses 1
value | prompt listlengths 2 2 | ability stringclasses 1
value | reward_model dict | images images listlengths 1 1 | extra_info dict |
|---|---|---|---|---|---|
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 0,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 1,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 2,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 3,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 4,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 5,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 6,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 7,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 8,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 9,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 10,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 11,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 12,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 13,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 14,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 15,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 16,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 17,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 18,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 19,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 20,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 21,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 22,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 23,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 24,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 25,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 26,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 27,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 28,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 29,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 30,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 31,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 32,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 33,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 34,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 35,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 36,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 37,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 38,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 39,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 40,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 41,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 42,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 43,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 44,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 45,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 46,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 47,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 48,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 49,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 50,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 51,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 52,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 53,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 54,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 55,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 56,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 57,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 58,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 59,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 60,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 61,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 62,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 63,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 64,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 65,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 66,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 67,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 68,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 69,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 70,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 71,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 72,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 73,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 74,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 75,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 76,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 77,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 78,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "abyssinian",
"style": "rule"
} | {
"gt_label": "abyssinian",
"index": 79,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 80,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 81,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 82,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 83,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 84,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 85,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 86,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 87,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 88,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 89,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 90,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 91,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 92,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 93,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 94,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 95,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 96,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 97,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 98,
"split": "train"
} | |
oxford_pets | [
{
"content": "You are a helpful assistant.",
"role": "system"
},
{
"content": "<image>\nClassify the image.\nOutput the thinking process in <think> </think> and the final answer in <answer> </answer> tags. The output answer format should be as follows: <think></think> <answer>a single label or the w... | classification | {
"ground_truth": "american bulldog",
"style": "rule"
} | {
"gt_label": "american bulldog",
"index": 99,
"split": "train"
} |
End of preview. Expand in Data Studio
Oxford-IIIT Pet — GRPO training data (structured reasoning)
GRPO (verl) training data for Oxford-IIIT Pet breed classification with a structured-reasoning prompt: the model emits a scratchpad tagging visible properties (HasProperty), parts (HasA), and setting (AtLocation) before the label. Reward: 0.30 for a well-formed think block, 0.70 for the label match.
Splits: train 2,944 rows, test 3,669 rows.
Schema
| column | type |
|---|---|
data_source |
string |
prompt |
list<element: struct<content: string, role: string>> |
ability |
string |
reward_model |
struct<ground_truth: string, style: string> |
images |
list<element: struct<bytes: binary, path: string>> |
extra_info |
struct<gt_label: string, index: int64, split: string> |
Format
Parquet in the verl RL schema, one row per image: a chat prompt (system + user), an images column (PNG-encoded), reward_model = {style: rule, ground_truth: <breed>}, ability: classification, and extra_info. Consumed directly by the verl GRPO trainer with data.image_key=images.
Provenance
Built with generate_grpo_dataset() for GRPO training experiments on open-world image classification with large multimodal models (Test-Time Warm-Up line). Derived from the Oxford-IIIT Pet Dataset (Parkhi et al., CVPR 2012).
- Downloads last month
- 210