Repository Issue Activity (beta)

#15 Question: Larger images + Text modalities

#47 Working with missing bands and reconstruction from embeddings

#46 How to build TerraMind encdec model pretrained

#44 Stability of `tiled_inference`

#45 Continual pre-training for TerraMind

#38 Bad segmentation results in Sen1Floods11 notebook

#30 [Bug] Output modality "Caption" not recognized (KeyError: 'caption')

#43 Deterministic execution

#42 `terramind_v1_small_sen1floods11` example doesn't work (for me)

#35 Potential checkerboard artifacts in terramind_v1_tokenizer_s1grd reconstruction

#40 S1GRD to S2L2A Generation

#39 Metadata encoding for fine-tuning?

#37 [Discussion/Question] Guidance on Latent-Space Generative Modeling with the VAE Tokenizer

#36 [Guidance Request] Best Practices for Reconstructing Partial-Band (10-Band) Sentinel-2 Data

#34 Optimal number of timesteps for tokenizer decoders

#32 Use different seeds to get different outputs

#31 Multimodal temporal wrapper

#29 Need help with TerraMind/TerraTorch: crop output not available, NDVI/LULC working, next steps

#28 Terramind Huggingface

#26 Clarification on dual scale early fusion in terramind_v1_base

#23 Loading pretrained backbone with Sentinel2 RBG+IR and Copernicus DEM bands

#25 Inconsistent Metric Results in terramind_v1_base_sen1floods11.ipynb Testing

#27 Define output modalities in forward for `_generate` models

#24 Support TiM models based on subset of input modalities

#21 Object Detection as a downstream task