OpenNMT/CTranslate2 issues and pull requests

#1595 - Chat example and TinyLlama/TinyLlama-1.1B-Chat-v1.0

Issue - State: closed - Opened by AIWintermuteAI 9 months ago - 1 comment

#1594 - feature request: support 4d attention masks

Issue - State: open - Opened by poedator 9 months ago

#1593 - Mixtral-8x7B support

Issue - State: closed - Opened by tobrun 9 months ago - 3 comments

#1592 - Support of new option offset to ignore token score of special tokens

Pull Request - State: closed - Opened by ASenart 9 months ago - 2 comments

#1591 - sqlcoder (7b) from mistral model

Issue - State: closed - Opened by JunHyungKang 9 months ago - 1 comment

#1590 - update cuda 11.2 to cuda 12.2

Pull Request - State: closed - Opened by minhthuc2502 9 months ago - 21 comments

#1589 - model to translate fron latin to spanish

Issue - State: closed - Opened by halaneji 10 months ago - 1 comment

#1588 - faster_whisper batch operation time-consuming

Issue - State: open - Opened by dyyzhmm 10 months ago - 9 comments

#1587 - Yi-6B-Chat convert failed.

Issue - State: closed - Opened by ninehills 10 months ago - 1 comment

#1586 - Can we support CoreML of apple silicon?

Issue - State: open - Opened by ILG2021 10 months ago

#1585 - updated documentation

Issue - State: open - Opened by BBC-Esq 10 months ago
Labels: documentation

#1584 - Bad Performance after convert the OpenNMT-py model to ct2 model

Issue - State: closed - Opened by CreamyLong 10 months ago - 4 comments

#1583 - GEMM operator calculates the `c` output shape incorrectly when input `a` is transpose?

Issue - State: closed - Opened by kandrio 10 months ago - 4 comments

#1582 - Prefix tokens are ignored during translation with NLLB model

Issue - State: closed - Opened by winstxnhdw 10 months ago - 1 comment

#1581 - fix onmt converter

Pull Request - State: closed - Opened by vince62s 10 months ago

#1580 - [Feature request] providing C API

Issue - State: closed - Opened by hellodword 10 months ago - 1 comment

#1579 - All samples are the same if include_prompt_in_result=False

Issue - State: closed - Opened by darabos 10 months ago - 2 comments

#1578 - convert opennmt py on the fly and inference with model in memory

Pull Request - State: open - Opened by minhthuc2502 10 months ago

#1577 - Whisper - Correct way to get prediction probability of each token and timestamp alignment

Issue - State: open - Opened by huydang2106 10 months ago

#1576 - finetuned whisper model can not use “initial_prompt”

Issue - State: open - Opened by v-yunbin 10 months ago - 1 comment

#1575 - [Feature request] Allowing StorageView.to() method to support moving between devices

Issue - State: closed - Opened by AtomAnu 10 months ago - 5 comments
Labels: enhancement

#1574 - bump v3.23.0

Pull Request - State: closed - Opened by minhthuc2502 10 months ago

#1573 - Can someone update the latest version on pypi ?

Issue - State: closed - Opened by PL508 10 months ago

#1572 - fix forward batch

Pull Request - State: closed - Opened by minhthuc2502 10 months ago - 1 comment

#1571 - Support for core42/jais-13b-chat (Arabic LLM)

Issue - State: open - Opened by Apoorv7092 10 months ago - 4 comments

#1570 - Mistral errors when doing forward_batch

Issue - State: closed - Opened by NeonBohdan 10 months ago

#1569 - Can't load models in reproducing the benchmark numbers from the README

Issue - State: open - Opened by opticalor 10 months ago

#1568 - Adding PhiConfig

Pull Request - State: closed - Opened by michaelfeil 10 months ago - 8 comments

#1567 - Context Extension with Ctranslate2

Issue - State: open - Opened by pranavbhat12 10 months ago - 1 comment

#1566 - Inefficient tokens suppression during BeamSearch

Issue - State: open - Opened by rilshok 10 months ago

#1565 - check alignment_heads in gen config before using

Pull Request - State: closed - Opened by minhthuc2502 10 months ago - 1 comment

#1564 - 'GenerationConfig' object has no attribute 'alignment_heads'

Issue - State: closed - Opened by takenori-y 10 months ago - 3 comments

#1563 - CUDA 12.1 support coming and CUBLAS and CUDNN?

Issue - State: closed - Opened by BBC-Esq 10 months ago - 2 comments

#1562 - ValueError: unsupported device mps

Issue - State: open - Opened by jk2K 10 months ago - 1 comment

#1561 - bump v3.22.0

Pull Request - State: closed - Opened by nguyendc-systran 10 months ago

#1560 - Support MADLAD-400 - multilingual machine translation model based on the T5 architecture

Issue - State: closed - Opened by carolinaxxxxx 10 months ago - 14 comments
Labels: enhancement

#1559 - Phi-1.5 conversion not working

Issue - State: closed - Opened by NeonBohdan 11 months ago - 3 comments
Labels: help wanted

#1558 - LLaMa2 does not generate results for prompts with input tokens > 2100

Issue - State: closed - Opened by bhavnicksm 11 months ago - 2 comments
Labels: question

#1557 - Support for LongT5 Model

Issue - State: closed - Opened by Nick-infinity 11 months ago - 1 comment
Labels: enhancement, help wanted

#1556 - OpenCL backend

Issue - State: closed - Opened by Nick-infinity 11 months ago - 2 comments

#1555 - test pipeline mistral support

Pull Request - State: closed - Opened by minhthuc2502 11 months ago

#1554 - sliding window + chunking input

Pull Request - State: closed - Opened by minhthuc2502 11 months ago

#1553 - MIstral sliding window + chunking input

Pull Request - State: closed - Opened by minhthuc2502 11 months ago

#1552 - fixed decoder_start_token_id for T5

Pull Request - State: closed - Opened by Ehsan-Jahanbakhsh 11 months ago - 3 comments

#1551 - set python tests to transformers 4.35

Pull Request - State: closed - Opened by vince62s 11 months ago - 7 comments

#1550 - test failing investigation

Pull Request - State: closed - Opened by vince62s 11 months ago

#1549 - Fixing tests with changes to github

Pull Request - State: closed - Opened by vince62s 11 months ago - 5 comments

#1548 - Revert "various fixes"

Pull Request - State: closed - Opened by vince62s 11 months ago

#1547 - New library/code challenging ctranslate2 or just creative beam sizes/batching instead?

Issue - State: closed - Opened by BBC-Esq 11 months ago - 2 comments

#1546 - Better HF-to-CT2 conversion for Whisper model

Pull Request - State: closed - Opened by flyingleafe 11 months ago - 5 comments

#1545 - Fix iomp5 linking with latest Intel OpenAPI on Ubuntu

Pull Request - State: closed - Opened by flyingleafe 11 months ago - 2 comments

#1544 - various fixes

Pull Request - State: closed - Opened by vince62s 11 months ago

#1543 - Docker image generation issue for vmap generation

Issue - State: open - Opened by rkoystart 11 months ago - 2 comments

#1542 - Issue converting Mistral models

Issue - State: closed - Opened by hobodrifterdavid 11 months ago - 7 comments

#1541 - add mistral to OpenNMT-py converter

Pull Request - State: closed - Opened by vince62s 11 months ago - 2 comments

#1540 - convert whisper large v3 error

Issue - State: closed - Opened by lxygoodjob 11 months ago - 1 comment

#1539 - ctranslate2 conflicting with transformers library?

Issue - State: closed - Opened by BBC-Esq 11 months ago - 1 comment

#1538 - publishing on pypi?

Issue - State: closed - Opened by BBC-Esq 11 months ago

#1537 - Support for IBM Power10 and Matrix Math Extensions (MMA)

Issue - State: closed - Opened by Dagamies 11 months ago - 1 comment

#1536 - Build from code failed

Issue - State: closed - Opened by CheshireCC 11 months ago - 9 comments

#1535 - Whisper large-v3 model!

Issue - State: closed - Opened by BBC-Esq 11 months ago - 2 comments

#1534 - DistilRoBERTa config

Issue - State: open - Opened by grohit1810 11 months ago - 2 comments

#1533 - bump v3.21.0

Pull Request - State: closed - Opened by vince62s 11 months ago

#1532 - possible to cut a new release?

Issue - State: closed - Opened by stillmatic 11 months ago - 1 comment

#1531 - Fail build python in conda environments

Issue - State: closed - Opened by johnnynunez 11 months ago - 1 comment

#1530 - whisper-large-v3 compatibility

Pull Request - State: closed - Opened by funboarder13920 11 months ago - 4 comments

#1529 - Support conversion for distil-whisper model

Pull Request - State: closed - Opened by chiiyeh 11 months ago - 9 comments

#1528 - Minimal support for Mistral 7B model.

Pull Request - State: closed - Opened by vince62s 11 months ago - 2 comments

#1527 - Does amdgpu support it?

Issue - State: closed - Opened by taikai-zz 11 months ago - 1 comment

#1526 - Add setuptools dependency to setup.py

Pull Request - State: open - Opened by jncraton 11 months ago - 2 comments

#1525 - New Jina embedding model and ctranslate2 support

Issue - State: closed - Opened by BBC-Esq 11 months ago - 6 comments

#1524 - Sliding window + chunking input for mistral model

Pull Request - State: closed - Opened by minhthuc2502 11 months ago - 8 comments

#1523 - [Question] What is the value range for inputs for batch_generator and more

Issue - State: closed - Opened by rakesh-krishna 11 months ago - 2 comments

#1522 - install ctranslate2 faill

Issue - State: closed - Opened by lambda7xx 11 months ago - 3 comments

#1521 - Failing for the recent pull requests

Issue - State: open - Opened by homink 11 months ago - 1 comment

#1520 - Support Transformers in the Wav2Vec2 Encoder for the ASR Inference

Pull Request - State: closed - Opened by homink 11 months ago - 4 comments

#1519 - Inference issue after converting Bloom-7b1 to ct2

Issue - State: closed - Opened by thanhnew2001 11 months ago - 2 comments

#1518 - Whisper inference with 10-second window

Issue - State: closed - Opened by bil-ash 11 months ago - 4 comments

#1517 - Implementation with Llama2 and llama index via langchain generates responses which includes context.

Issue - State: closed - Opened by pranavbhat12 11 months ago - 1 comment

#1516 - Support for mistral

Pull Request - State: closed - Opened by jgcb00 12 months ago - 3 comments

#1515 - [Question] Can conv1d in CTranslate2 work with groups argument as same as used in PyTorch?

Issue - State: closed - Opened by homink 12 months ago

#1514 - Llava support

Issue - State: open - Opened by BBC-Esq 12 months ago - 2 comments

#1513 - Compiling for a particular combination of CUDA and GPU CC

Issue - State: closed - Opened by sgjohnson1981 12 months ago - 1 comment

#1512 - Build ctranslate2 wheel package

Issue - State: closed - Opened by prashantkh19 12 months ago - 4 comments

#1511 - sampling temperature causes zero division ?

Issue - State: closed - Opened by evalquote 12 months ago - 1 comment

#1510 - Add RoBERTaConfig to TransformerConverter

Pull Request - State: open - Opened by bhavnicksm 12 months ago - 2 comments

#1509 - Whisper using prefix shows no speed up

Issue - State: open - Opened by chiiyeh 12 months ago - 1 comment
Labels: question

#1508 - Support ARM for m1 or jetson?

Issue - State: closed - Opened by johnnynunez 12 months ago - 2 comments

#1507 - new repo faster whisper transcribe

Issue - State: closed - Opened by BBC-Esq 12 months ago

#1506 - Support Mistral-7B configuration

Issue - State: closed - Opened by wsxiaoys 12 months ago - 2 comments

#1505 - Quantization of T5encorderonly models like instructor-xl.

Issue - State: closed - Opened by BBC-Esq almost 1 year ago - 2 comments

#1504 - utilizing embedding model

Issue - State: closed - Opened by BBC-Esq almost 1 year ago - 5 comments

#1503 - NameError: name 'torch' is not defined

Issue - State: closed - Opened by sdugoten about 1 year ago - 10 comments

#1502 - Accept callback even on generate_tokens method

Pull Request - State: closed - Opened by jgcb00 about 1 year ago - 4 comments

#1501 - Support for "mistralai/Mistral-7B-Instruct-v0.1" model

Issue - State: closed - Opened by Matthieu-Tinycoaching about 1 year ago - 57 comments

GitHub / OpenNMT/CTranslate2 issues and pull requests