default_embedding_fake_quant

torch.ao.quantization.fake_quantize.default_embedding_fake_quant

Default fake_quant for embeddings. Observer is memoryless since averaging_constant is 1.