Time Series Forecasting on CAF-7M HARD
-1.5CRPSTimeLLM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TimeLLMEvaluation Protocol=Trained, Training Dataset=CAF-7M, Context Condition=No context2026.03 | -1.5 | -34.5 | |
| DP GPT-5.2Evaluation Protocol=Zero-shot2026.03 | 0.212 | 0.75 | |
| DoubleCastEvaluation Protocol=Trained, Training Dataset=CAF-7M2026.03 | 0.251 | 0.79 | |
| DP Qwen3-14BEvaluation Protocol=Zero-shot2026.03 | 0.277 | 0.63 | |
| ChronosEvaluation Protocol=Zero-shot2026.03 | 0.304 | — | |
| ChatTimeEvaluation Protocol=Zero-shot2026.03 | 0.407 | 0.25 | |
| TimeLLMEvaluation Protocol=Trained, Training Dataset=CAF-7M2026.03 | 0.428 | 0.22 | |
| TimeLLMEvaluation Protocol=Zero-shot, Training Dataset=ETTm1, Input Sequence Length=1922026.03 | 0.549 | 0.12 | |
| TimeLLMEvaluation Protocol=Zero-shot, Training Dataset=ETTm2, Input Sequence Length=3362026.03 | 0.703 | 0.07 | |
| TimeLLMEvaluation Protocol=Zero-shot, Training Dataset=ETTm2, Input Sequence Length=962026.03 | 1.284 | 0.02 | |
| TimeLLMEvaluation Protocol=Trained, Training Dataset=CAF-7M, Context Condition=Swapped2026.03 | 2 | 7.3 | |
| TimeLLMEvaluation Protocol=Zero-shot, Training Dataset=ETTm1, Input Sequence Length=3362026.03 | 3.212 | 0 | |
| DoubleCastEvaluation Protocol=Trained, Training Dataset=CAF-7M, Context Condition=No context2026.03 | 22 | -41.9 | |
| DoubleCastEvaluation Protocol=Trained, Training Dataset=CAF-7M, Context Condition=Swapped2026.03 | 25.2 | -40.8 |