Commit message (Expand) | Author | Age | |
---|---|---|---|
* | Skip tensors with null buffers. | 2018-09-12 | |
* | Bidi lstm supports hybrid eval. | 2018-09-07 | |
* | Skip quantization of optional tensors (tensor_idx = -1) | 2018-09-05 | |
* | Make minimum num elements of quantizable weights tensor configurable. | 2018-09-04 | |
* | Improving logging of tensors being quantized. | 2018-08-31 | |
* | Hybrid operations need either all or none of their tensors quantized. | 2018-08-31 | |
* | Docs for quantize weights tool. | 2018-08-31 | |
* | Allocate QuantizationParam when the input model doesn't have it. | 2018-08-31 | |
* | Avoid using bracket initializer for vector initializer with correct types. | 2018-08-30 | |
* | Add support for hybrid evaluation in quantize weights transform. | 2018-08-29 | |
* | TFLite quantize_weights tool. | 2018-08-23 |