Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -18,6 +18,7 @@ Checkpoints behind the [CELL-FM CondenSeq demo](https://huggingface.co/spaces/Bo
|
|
| 18 |
| `condenseq/cellfm_seq2img.bin` | CELL-FM CS sequence-to-image generator (includes the ESM-C 600M encoder) | `pretrain_condenseq/cellfm_seq2img/checkpoint-50000` |
|
| 19 |
| `condenseq/vae.bin` | Image VAE, 160x160, 3 down blocks, 4 latent channels | `pretrain_condenseq/vae/checkpoint-50000` |
|
| 20 |
| `condenseq/vit_cls.bin` | ViT condensed/diffuse classifier, 2-channel 160x160 input | `PT_CondenSeq_img_ViT_cls_R1/checkpoint-10000` |
|
|
|
|
| 21 |
| `hpa/cellfm_seq2img.bin` | CELL-FM virtual-staining generator for HPA, 256x256, 3-channel conditioning (includes the ESM-C 600M encoder) | `pretrain_hpa/cellfm_seq2img/checkpoint-50000` |
|
| 22 |
| `hpa/vae.bin` | Image VAE at 256x256, the one `hpa/cellfm_seq2img.bin` was trained against | `pretrain_hpa/vae/checkpoint-50000` |
|
| 23 |
| `hpa/vit.bin` | Image-embedding ViT for HPA: 12 layers, 512 hidden, 2048 MLP, 8 heads, patch 4, 256x256 input. Trained to identify the protein in an image over 13,908 classes; the embedding is the representation, not the prediction. Four input channels — protein, nucleus, ER, microtubules | `pretrain_hpa/vit/checkpoint-16000` |
|
|
|
|
| 18 |
| `condenseq/cellfm_seq2img.bin` | CELL-FM CS sequence-to-image generator (includes the ESM-C 600M encoder) | `pretrain_condenseq/cellfm_seq2img/checkpoint-50000` |
|
| 19 |
| `condenseq/vae.bin` | Image VAE, 160x160, 3 down blocks, 4 latent channels | `pretrain_condenseq/vae/checkpoint-50000` |
|
| 20 |
| `condenseq/vit_cls.bin` | ViT condensed/diffuse classifier, 2-channel 160x160 input | `PT_CondenSeq_img_ViT_cls_R1/checkpoint-10000` |
|
| 21 |
+
| `condenseq/reference_nucleus.npy` | The DAPI channel every CondenSeq generation is conditioned on: `(1, 160, 160)` float32 in [-1, 1]. CondenSeq protein index 12626, image 0 — the same one the offline seq2img runs used | built |
|
| 22 |
| `hpa/cellfm_seq2img.bin` | CELL-FM virtual-staining generator for HPA, 256x256, 3-channel conditioning (includes the ESM-C 600M encoder) | `pretrain_hpa/cellfm_seq2img/checkpoint-50000` |
|
| 23 |
| `hpa/vae.bin` | Image VAE at 256x256, the one `hpa/cellfm_seq2img.bin` was trained against | `pretrain_hpa/vae/checkpoint-50000` |
|
| 24 |
| `hpa/vit.bin` | Image-embedding ViT for HPA: 12 layers, 512 hidden, 2048 MLP, 8 heads, patch 4, 256x256 input. Trained to identify the protein in an image over 13,908 classes; the embedding is the representation, not the prediction. Four input channels — protein, nucleus, ER, microtubules | `pretrain_hpa/vit/checkpoint-16000` |
|