OpenNMT/OpenNMT-py

Issues

• Error when translating with scores (--with_scores) option enabled

  #2583 opened 2 months ago by ncicio
  1

• UnicodeDecodeError: 'utf-8' codec can't decode byte 0x80 in position 55: invalid start byte

  #2579 opened a month ago by fkurushin
  1

• [Bug - Translation server] - Missing `tgt`param in `translator.translate` method (allows some multilingual/seq2seq models to work properly)

  #2586 opened a month ago by medfreeman
  0

• Supported languages

  #2584 opened a month ago by Mayyarkmp
  0

• Issues with Custom SentencePiece Models and Pretrained Embeddings in Training

  #2582 opened 2 months ago by HURIMOZ
  0

• Canʻt get past Sentencepiece subword tokenization with pretrained embeddings

  #2581 opened 2 months ago by HURIMOZ
  0

• AttributeError: 'Namespace' object has no attribute 'block_ngram_repeat'

  #2578 opened 2 months ago by HURIMOZ
  0

• Index out of range

  #2580 opened 2 months ago by MSKantulu
  0

• OpenNMT v3.5.0 training fails using Multi headed attention

  #2574 opened 2 months ago by RakshaPRao
  3

• Custom callbacks for metrics, saving checkpoints

  #2575 opened 2 months ago by Garfounkel
  3

• (Again, but different) AssertionError: assert model_dim % head_count == 0

  #2571 opened 3 months ago by James-Decatur
  2

• Support for torch 2.2

  #2560 opened 3 months ago by jakeBass
  5

• Device side assert triggered on AWQ Mistral converted model

  #2562 opened 4 months ago by kdcyberdude
  2

• NaN values when training big transformer model

  #2559 opened 4 months ago by PC91
  1

• Any plan to support "mps" backend?

  #2474 opened 4 months ago by everdark
  1

• NCCL timeout with 2B+ parameter model

  #2515 opened 4 months ago by Dagamies
  8

• Supported SentencePiece parameters

  #2546 opened 4 months ago by PC91
  1

• How to use Huawei‘s NPU Ascend310 to install OpenNMT-py？

  #2556 opened 4 months ago by michelleqyhqyh
  1

• Speech to Text Toy Data Could Not Be Downloaded

  #2554 opened 4 months ago by Keram-Yasin
  3

• Translation API Not Working

  #2555 opened 4 months ago by Keram-Yasin
  1

• List index out of range in onmt.utils.distributed.all_reduce_and_rescale_tensors:51

  #2549 opened 5 months ago by alexis-allemann
  0

• Error evaluating LM-prior checkpoint:

  #2541 opened 5 months ago by anthdr
  1

• Error message of `SequenceTooLongError`

  #2522 opened 5 months ago by PC91
  1

• Bug when training encoder-decoder models

  #2528 opened 6 months ago by JOHW85
  1

• Data generation when resuming from a checkpoint

  #2517 opened 6 months ago by PC91
  2

• Input size mismatch

  #2519 opened 6 months ago by pranjaliseth
  1

• set random seed for a multi-GPU model

  #2516 opened 7 months ago by Galaxy-Husky
  1

• transforms: filtertoolong failed in translating

  #2501 opened 7 months ago by ares89
  1

• Cannot load recurrent encoder-decoder model trained with copy attention

  #2503 opened 7 months ago by colincwilson
  7

• Training fails to start with rotary embedding (Latest OpenNMT-py)

  #2513 opened 7 months ago by Dagamies
  3

• Columns and DataType Not Explicitly Set on line 163 of run_mmlu_opennmt.py

  #2510 opened 7 months ago by CodeSmileBot
  0

• bash: scripts/onmt/train.sh: No such file or directory

  #2502 opened 7 months ago by Armilius
  4

• [help] onmt.inputters.text_dataset

  #2482 opened 7 months ago by timon49
  5

• Maybe exist some bug in attn_debug function in vesion 3,

  #2490 opened 8 months ago by zw-SIMM
  18

• Is it a requirement for the parallel training corpus to be 100% strictly correspondent paraghaph by paragraph?

  #2488 opened 8 months ago by fishfree
  2

• The result of opennmt-py translate different from CTranslate2.Translator (beamsearch10)

  #2483 opened 8 months ago by zw-SIMM
  1

• Bug in Encoder-Decoder Translation models

  #2468 opened 9 months ago by henyee
  3

• Translator fails with coverage penalty enabled

  #2471 opened 9 months ago by robertBrnnn
  3

• Please delete

  #2478 opened 9 months ago by BaGRoS
  0

• Translation fails with torch<2.0

  #2475 opened 9 months ago by robertBrnnn
  1

• number of features mismatch

  #2459 opened 9 months ago by totaltube
  2

• translate only one sentence input by user.

  #2461 opened 10 months ago by 13633491388
  3

• Can not finetune nllb-200-3.3B

  #2458 opened 10 months ago by ILG2021
  2

• cuda out of memeory

  #2456 opened 10 months ago by Mamooshe
  2

• Some questions about fine-tune nllb-200

  #2457 opened 10 months ago by ILG2021
  1

• Confusion about grad and loss computation in distributed multi-gpu training

  #2451 opened 10 months ago by Galaxy-Husky
  8

• ValueError on calling onmt_translate

  #2455 opened 10 months ago by Lguyogiro
  2

• Question about Falcon40b result on MMLU

  #2454 opened 10 months ago by MozerWang
  10

• A question about “world_size” and "gpu_rank"

  #2453 opened 10 months ago by fdggdfgg
  0

• some problem about "-world-size" and "gpu_ranks"

  #2450 opened 10 months ago by fdggdfgg
  9