Upload folder using huggingface_hub

README.md

@@ -1,36 +1,25 @@
----
-tags:
-- transcriptomics
-- dimensionality-reduction
-- tabpfn
-- TRACERx
-license: mit
----
+# TabPFN Transcriptome 32D Model
 
-# TABPFN Model - transcriptome mode - 32D
+Pre-computed TabPFN embeddings for transcriptomic data.
 
-Pre-trained tabpfn model for transcriptomic data compression.
-
-## Details
-- **Mode**: transcriptome-centric compression
-- **Dimensions**: 32
-- **Training data**: TRACERx lung cancer transcriptomics
-- **Created**: 2026-01-10T10:17:30.385971
+## Contents
+- `model.joblib`: Model bundle with reference embeddings
 
 ## Usage
-
 ```python
 import joblib
-from huggingface_hub import snapshot_download
+from huggingface_hub import hf_hub_download
 
 # Download model
-local_dir = snapshot_download("jruffle/tabpfn_transcriptome_32d")
-model = joblib.load(f"{local_dir}/model.joblib")
+model_path = hf_hub_download(repo_id="jruffle/tabpfn_transcriptome_32d", filename="model.joblib")
+model = joblib.load(model_path)
 
-# For classical models (PCA/UMAP):
-# model contains: 'pca', 'umap', 'robust_scaler', 'gene_order'
-
-# For TabPFN models (with UMAP reduction):
-# model contains: 'tabpfn_embedding', 'umap_final', 'raw_embeddings', 'input_scaler', etc.
-# Uses UMAP instead of PCA for non-linear dimension reduction of TabPFN embeddings
+# Access embeddings
+embeddings = model['reference_embeddings']
 ```
+
+## Details
+- Mode: transcriptome
+- Dimensions: 32
+- Samples: 2405
+- Complete coverage: All samples have TabPFN embeddings

model.joblib

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c002f3f599806cdbe776800ef6cb0020972afb5555db0afc21df119485544f13
-size 132974786
+oid sha256:cbfbd98b958b8f05a180af43e708b0df8bfd49e84a87e086b2c322ac13af3ed2
+size 665798