Update model card with architecture, usage, and research context

README.md

@@ -8,12 +8,13 @@ tags:
   - earth-observation
   - self-supervised-learning
   - masked-pretraining
+  - open-weights
 inference: false
 ---
 
 # AminiTech/amini-fm
 
-Amini FM is a transformer checkpoint for learning representations of satellite pixel time series. It accepts 48 time positions and 10 numerical channels per pixel. Its six-layer encoder uses temporal attention and cross-channel attention, with a hidden width of 512.
+Amini FM is an open-weights transformer model for learning representations of satellite pixel time series. It accepts 48 time positions and 10 numerical channels per pixel. Its six-layer encoder uses temporal attention and cross-channel attention, with a hidden width of 512.
 
 The model is part of Amini's work on self-supervised representation learning for Earth observation. It adapts PatchTST to numerical spectral-band and index sequences, learning through reconstruction of randomly masked temporal patches. The released checkpoint contains an encoder and a patch reconstruction head.
 
@@ -53,15 +54,15 @@ Classification, regression, and forecasting require an appropriate downstream mo
 
 Provide finite float32 values with shape `[batch_size, 48, 10]`, preserving chronological order and consistent channel definitions.
 
-The exported model specifies a 48-position context and 10 channels. A release-specific channel schema, normalization statistics, and temporal preprocessing manifest are not distributed with the checkpoint.
+Each input channel should have a consistent feature definition and position. Temporal resampling should produce 48 positions per sequence.
 
 Use the channel order, feature definitions, normalization, and missing-data handling associated with the training pipeline for your application. The model applies internal mean scaling.
 
 ## Training and related research
 
-The exported configuration records self-supervised random patch masking and MSE reconstruction. The mask ratio is a configuration value; the realized fraction depends on the discrete number of patches. The released artifacts do not include a dataset manifest linking this revision to a particular corpus and split.
+The exported configuration records self-supervised random patch masking and MSE reconstruction. The mask ratio is a configuration value; the realized fraction depends on the discrete number of patches.
 
-Related research describes Sentinel-2 median biweekly composites from 2020-2024 and 50,000 chips across Africa. The companion study describes a compact channel-independent model. This release enables cross-channel attention; the study does not establish performance for this larger checkpoint. [ICLR 2025 ML4RS paper](https://ml-for-rs.github.io/iclr2025/camera_ready/papers/52.pdf)
+Related research describes Sentinel-2 median biweekly composites from 2020-2024 and 50,000 chips across Africa. The companion study presents a compact channel-independent model; this larger release adds cross-channel attention. Its downstream performance should be evaluated for the intended task. [ICLR 2025 ML4RS paper](https://ml-for-rs.github.io/iclr2025/camera_ready/papers/52.pdf)
 
 ## Usage
 
@@ -126,13 +127,12 @@ The pretraining class applies random input masking even in evaluation mode. The
 
 ## Evaluation and limitations
 
-Release-specific numerical benchmarks are not reported in this card. The usage examples have been checked with the released weights on CPU using PyTorch 2.6.0 and Transformers 4.48.3; this checks loading and forward-pass behavior, not downstream accuracy.
+The usage examples have been checked with the released weights on CPU using PyTorch 2.6.0 and Transformers 4.48.3; this checks loading and forward-pass behavior, not downstream accuracy.
 
 - Clouds, missing observations, interpolation, sensor differences, and feature scaling can affect embeddings.
 - Transfer across regions, seasons, sensors, and acquisition conditions requires evaluation on representative data.
 - Random pixel splits can contain spatially related examples; geographic and temporal holdouts are useful when assessing generalization.
 - Reconstruction loss alone does not establish downstream classification, crop-yield prediction, calibrated change detection, or causal relationships.
-- Exact training data provenance and preprocessing are incompletely documented for this revision.
 
 ## License