Sarah
Sarah
marian version：v1.11.0 using factor word embedding [2022-05-25 23:28:43] [marian] Marian v1.11.0 f00d0621 2022-02-08 08:39:24 -0800 [2022-05-25 23:28:43] [marian] Running on dnnl as process 7562 with command line: [2022-05-25 23:28:43] [marian]...
### Bug description the url of the paper : https://www.researchgate.net/publication/337074940_Fast_Transformer_Decoding_One_Write-Head_is_All_You_Need ### How to reproduce Describe steps or include command to reproduce the behavior. when i slice the key and value...
email like [email protected] nums like 1,234,000 can be written as 12,34,000 in Hindi, 1.234.000 in German, and 123.4万 in Chinese.
ie, how to tie all layers in the decoder (not the encoder)
The document describe the following para： --quantize-bits UINT=0 Number of bits to compress model to. Set to 0 to disable --quantize-optimization-steps UINT=0 Adjust quantization scaling factor for N steps --quantize-log-based...
### Bug description mairan version: Marian v1.10.0 6f6d484 when I set the configuration of quantize like this: --sync-sgd --quantize-bits 8 --quantize-optimization-steps 10 --quantize-biases true the file "marian-master\src\training\graph_group_sync.cpp" Line 379 will...
Doubt this line should be : scores.append((sentence_idx, max_cluster_score))
1、which version or branch of marian to complie? marian-master， marian-dev or https://github.com/afaji/Marian/tree/fixed-quant? 2、could the teacher model trained by teh marian-master?
**header_only** In file included from ./spdlog/fmt/fmt.h:27:0, from ./spdlog/common.h:50, from spdlog/spdlog.h:12, from server_main.cpp:28: ./spdlog/fmt/bundled/core.h:3293:59: error: macro "print" passed 2 arguments, but takes just 1 FMT_INLINE void print(format_string fmt, T&&... args) {...