Commit
•
1c11bda
1
Parent(s):
5779bd6
Push model using huggingface_hub.
Browse files- README.md +11 -0
- config.json +52 -0
- model.safetensors +3 -0
README.md
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- denseav
|
| 5 |
+
- pytorch_model_hub_mixin
|
| 6 |
+
- model_hub_mixin
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
| 10 |
+
- Library: https://github.com/mhamilton723/DenseAV
|
| 11 |
+
- Docs: [More Information Needed]
|
config.json
ADDED
|
@@ -0,0 +1,52 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"adaptive_clipping": true,
|
| 3 |
+
"audio_aligner_type": "audio_sa_3_3_pool_2",
|
| 4 |
+
"audio_lora": false,
|
| 5 |
+
"audio_lora_rank": 8,
|
| 6 |
+
"audio_model_type": "hubert",
|
| 7 |
+
"audio_pool_width": 1,
|
| 8 |
+
"cal_balance_weight": 0.1,
|
| 9 |
+
"cal_init": 1.0,
|
| 10 |
+
"channel_dropout": 0.0,
|
| 11 |
+
"code_dim": 384,
|
| 12 |
+
"disentangle_weight": 0.0,
|
| 13 |
+
"finetune_audio_model": true,
|
| 14 |
+
"finetune_image_model": false,
|
| 15 |
+
"gather_tensors": true,
|
| 16 |
+
"gradient_clipping": 10.0,
|
| 17 |
+
"head_agg": "max_elementwise",
|
| 18 |
+
"image_aligner_type": "image_linear",
|
| 19 |
+
"image_lora": true,
|
| 20 |
+
"image_lora_rank": 8,
|
| 21 |
+
"image_model_token_type": "token",
|
| 22 |
+
"image_model_type": "dino8",
|
| 23 |
+
"image_pool_width": 2,
|
| 24 |
+
"learn_audio_cls": true,
|
| 25 |
+
"loss_leak": 0.0,
|
| 26 |
+
"loss_margin": 0.0,
|
| 27 |
+
"loss_type": "nce",
|
| 28 |
+
"lr": 5e-05,
|
| 29 |
+
"lr_cycle_length": 50000,
|
| 30 |
+
"lr_schedule": null,
|
| 31 |
+
"lr_warmup": 1000,
|
| 32 |
+
"mask_silence": true,
|
| 33 |
+
"memory_buffer_size": 0,
|
| 34 |
+
"mixup_weight": 0.0,
|
| 35 |
+
"neg_audio": true,
|
| 36 |
+
"neg_audio_weight": 0.01,
|
| 37 |
+
"nonneg_pressure": 0.01,
|
| 38 |
+
"nonneg_sim": false,
|
| 39 |
+
"norm_vectors": false,
|
| 40 |
+
"optimizer": "adam",
|
| 41 |
+
"output_root": "/mnt/azureml/cr/j/5e46ccec7c6e408e8496ffdbf5769734/cap/data-capability/wd/INPUT_video_analysis_store_3_video_analysis_store_3",
|
| 42 |
+
"pretrain_lr": 5e-05,
|
| 43 |
+
"silence_l1": 0.01,
|
| 44 |
+
"silence_l2": 0.0,
|
| 45 |
+
"sim_agg_heads": 2,
|
| 46 |
+
"sim_agg_type": "misa",
|
| 47 |
+
"sim_use_cls": false,
|
| 48 |
+
"spatial_dropout": 0.0,
|
| 49 |
+
"specialization_weight": 0.05,
|
| 50 |
+
"tv_weight": 0.01,
|
| 51 |
+
"use_cached_embs": false
|
| 52 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c46217bb89ca8d6ab27b7c40cdee909aff29e0be38c28a4cd2201d33b3c3b72b
|
| 3 |
+
size 1365469596
|