Use the NGC CLI to download:
Copied!
Model
SWIN Transformer based Re-Identification network to generate embeddings for identifying persons in different scenes.
Use the NGC CLI to download:
Copied!
5 Versions
09/04/2024 6:35 PM UTC267.04 MBBatch Size: 1GPU: V100Copied!
09/04/2024 6:35 PM UTC267.04 MBBatch Size: 1GPU: V100
Copied!
SHA256 digests
| Key | Value |
|---|---|
| experiment.yaml | 3207924387e6ce860268c63a6911f3098807aa0738b5fca179f53b4d67fe5499 |
| swin_tiny_market1501_aicity156_featuredim256.tlt | a1a1f5757b8d24ada96b6d4a57c922259b4bd8f2d3b093b1fb373d4563898cea |
TAO Toolkit API
| Key | Value |
|---|---|
| is_backbone | False |
| endpoints | ['re_identification'] |
| format_version | 0.0.1 |
| model_format | tlt |
| tao_version | ['>=5.5.0'] |
| trainable | True |
TAO Toolkit UI Metadata
| Key | Value |
|---|---|
| license | NVAIE EULA |
| task | Re Identification |
| format_version | 0.0.1 |
| backbone_type | transformer |
| domain | purpose-built |
| accuracy | 93.4% |
| num_parameters | 28.1M |
| backbone_class | swin |
09/04/2024 6:35 PM UTC832.98 MBBatch Size: 1GPU: V100Copied!
09/04/2024 6:35 PM UTC832.98 MBBatch Size: 1GPU: V100
Copied!
SHA256 digests
| Key | Value |
|---|---|
| experiment.yaml | 3207924387e6ce860268c63a6911f3098807aa0738b5fca179f53b4d67fe5499 |
| swin_base_market1501_aicity156_featuredim256.tlt | 6979806aca802f7ae4accc6e9c7f9e34121e2981959bc443424ba9a58587265e |
TAO Toolkit API
| Key | Value |
|---|---|
| is_backbone | False |
| endpoints | ['re_identification'] |
| format_version | 0.0.1 |
| model_format | tlt |
| tao_version | ['>=5.5.0'] |
| trainable | True |
TAO Toolkit UI Metadata
| Key | Value |
|---|---|
| license | NVAIE EULA |
| task | Re Identification |
| format_version | 0.0.1 |
| backbone_type | transformer |
| domain | purpose-built |
| accuracy | 93.3% |
| num_parameters | 87.5M |
| backbone_class | swin |
swin_base_1024Selected09/04/2024 6:35 PM UTC836.76 MBBatch Size: 1GPU: V100Copied!
swin_base_1024Selected
09/04/2024 6:35 PM UTC836.76 MBBatch Size: 1GPU: V100
Copied!
SHA256 digests
| Key | Value |
|---|---|
| swin_base_market1501_aicity156_featuredim1024.tlt | 1ebe761f15ced29d660095a9d2bf4974fcd24685691e14be759a85998d1797f9 |
| experiment.yaml | 2fa8c4fd5b7e8b014fdce1f76b9a7a380ea6bc24943bcc26d7b12b48dbd2354b |
TAO Toolkit API
| Key | Value |
|---|---|
| is_backbone | False |
| endpoints | ['re_identification'] |
| format_version | 0.0.1 |
| model_format | tlt |
| tao_version | ['>=5.5.0'] |
| trainable | True |
TAO Toolkit UI Metadata
| Key | Value |
|---|---|
| license | NVAIE EULA |
| task | Re Identification |
| format_version | 0.0.1 |
| backbone_type | transformer |
| domain | purpose-built |
| accuracy | 94.6% |
| num_parameters | 87.5M |
| backbone_class | swin |
01/26/2024 4:25 AM UTC1.89 GBAccuracy: 93.7120 EpochsBatch Size: 1GPU: V100Copied!
01/26/2024 4:25 AM UTC1.89 GBAccuracy: 93.7120 EpochsBatch Size: 1GPU: V100
Copied!
TAO Toolkit API
| Key | Value |
|---|---|
| sha256_digest | {'swin_tiny_market1501_aicity156_featuredim256.tlt': 'a1a1f5757b8d24ada96b6d4a57c922259b4bd8f2d3b093b1fb373d4563898cea'} |
| endpoints | ['re_identification'] |
| format_version | 0.0.1 |
| model_format | pth |
| tao_version | ['>=5.2.0'] |
| trainable | true |
01/26/2024 4:23 AM UTC784.5 MBAccuracy: 93.7120 EpochsBatch Size: 1GPU: V100Copied!
01/26/2024 4:23 AM UTC784.5 MBAccuracy: 93.7120 EpochsBatch Size: 1GPU: V100
Copied!
TAO Toolkit API
| Key | Value |
|---|---|
| sha256_digest | {'swin_tiny_market1501_aicity156_featuredim256.onnx': '37b008d887495d5fbad31c72aadbf381c5456cb619e0e2a9c246a6d48a2932ff'} |
| endpoints | ['re_identification'] |
| format_version | 0.0.1 |
| model_format | onnx |
| tao_version | ['>=5.2.0'] |
| trainable | false |