mistralai/mistral-inference issues and pull requests

#130 - "evaluation pipeline" public?

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

#127 - "official documentation" link points to a missing page (quickstart)

Issue - State: open - Opened by dpkirchner 9 months ago - 1 comment

#124 - Installation Problem

Issue - State: open - Opened by jahbini 9 months ago - 4 comments

#112 - Gate is Linear Layer?!?!

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

#101 - Minor typos

Pull Request - State: closed - Opened by sethupavan12 11 months ago

#100 - Which is the actual way to store the Adapter after PEFT finetuning

Issue - State: open - Opened by pradeepdev-1995 11 months ago

#99 - vLLM Build Issue using the provided Dockerfile

Issue - State: closed - Opened by Good-Coffee 11 months ago - 4 comments

#98 - Create Issue templates

Issue - State: open - Opened by adityaraute 11 months ago

#97 - Docs: Add tutorials for using Python client to generating embeddings and chat completion

Issue - State: open - Opened by m-newhauser 11 months ago

#96 - Fixing typos in MD

Pull Request - State: open - Opened by Cassini-chris 11 months ago

#95 - Has any thought been given to using LoRA to increase the number of experts (100x) with minimal memory?

Issue - State: open - Opened by sixChar 11 months ago - 8 comments

#94 - Fix typo/spelling in README.md

Pull Request - State: open - Opened by GilesBathgate 11 months ago - 1 comment

#93 - Mixtral Feedbacks

Issue - State: open - Opened by titouandk 11 months ago

#92 - Incomplete Output even with max_new_tokens

Issue - State: open - Opened by pradeepdev-1995 11 months ago

#91 - Building Mistral docker container results in OOM kill of the entire system

Issue - State: open - Opened by codevbus 11 months ago

#90 - wrong link in documentation

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

#89 - Adds attention mask with `model.forward(..., cache=None)`.

Pull Request - State: open - Opened by andsteing 11 months ago

#88 - Why does `cache=None` produce different outputs?

Issue - State: open - Opened by andsteing 11 months ago

#87 - Is the code up to date? Is the code the same for different model versions？

Issue - State: open - Opened by zysNLP 11 months ago

#86 - Inquiry on Implementing Sliding Window Attention for Custom Sequence Lengths

Issue - State: open - Opened by yihong1120 11 months ago

#85 - fix minor typo in README.md

Pull Request - State: open - Opened by nheagy 11 months ago

#84 - Fix link to official documentation in README.md

Pull Request - State: open - Opened by webchick 11 months ago

#83 - Add MoE and pipelining support

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

#82 - Update classifier.ipynb

Pull Request - State: open - Opened by eltociear 11 months ago

#81 - Fix Dockerfile

Pull Request - State: open - Opened by nicholasjpaterno 11 months ago

#80 - on Jetson ORIN, Xformer, Memory-efficient attention, SwiGLU, sparse and more won't be available.

Issue - State: open - Opened by cj401 11 months ago

#79 - Is window attention technology also used during the training phase?

Issue - State: open - Opened by peiyingxin 11 months ago

#78 - How to process batch input in mistral-src/model.py ?

Issue - State: open - Opened by NLPwoods 11 months ago

#77 - repeated build failure

Issue - State: open - Opened by juanmf 11 months ago

#76 - The detected CUDA version (11.8) mismatches the version that was used to compile

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

#75 - Fix: no system prompt in request

Pull Request - State: open - Opened by michel-ds 11 months ago

#74 - No safetensors in HF model card?

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

#73 - What is the difference between the files you publish on GitHub and Hugging Face

Issue - State: open - Opened by zhzfight 12 months ago

#72 - Unabled to load to GPU with 24 GB vRAM with quantization

Issue - State: open - Opened by fangzhouli 12 months ago - 1 comment

#71 - How is The 131K Attention Span Achieved?

Issue - State: open - Opened by ThePerfectComputer 12 months ago

#70 - Update README

Pull Request - State: open - Opened by luv-bansal 12 months ago

#69 - How to train mistral?

Issue - State: open - Opened by mihalt about 1 year ago

#68 - Was Mistral Pretrained with Dropout Enabled?

Issue - State: open - Opened by zaptrem about 1 year ago

#67 - Question about finetune mistral 7B (data format)

Issue - State: open - Opened by xihajun about 1 year ago - 1 comment

#66 - model is giving answer in russian

Issue - State: open - Opened by Sanchit-404 about 1 year ago - 4 comments

#65 - how to explain Attention that input QKV tensor # xformers requires (B=1, S, H, D)

Issue - State: closed - Opened by dhcode-cpp about 1 year ago - 1 comment

#64 - Does mistral-instruct-7b support fast transformer deployment

Issue - State: open - Opened by lebronjamesking about 1 year ago

#63 - Update README.md

Pull Request - State: open - Opened by VinayKokate22 about 1 year ago - 1 comment

#62 - Embedding model and Engine??

Issue - State: open - Opened by muhtalhakhan about 1 year ago - 6 comments

#61 - More language support?

Issue - State: open - Opened by OnceJune about 1 year ago - 6 comments

#60 - sliding window size in prefill and decode stage

Issue - State: open - Opened by ofhwei about 1 year ago

#59 - Can't load xFormers because of PyTorch 2.1.0+cu121

Issue - State: open - Opened by russ22cox about 1 year ago - 2 comments

#58 - Feature: Adding contributors section to the README.md file.

Issue - State: open - Opened by Kalyanimhala about 1 year ago - 2 comments

#57 - Code complete?

Issue - State: open - Opened by zhoumengbo about 1 year ago - 3 comments

#56 - Update README.md

Pull Request - State: open - Opened by eltociear about 1 year ago

#55 - Batching, GQA and Flash Attnetion

Issue - State: open - Opened by maximzubkov about 1 year ago - 1 comment

#54 - Unable to build Docker image with cuda:11.8.0-devel-ubuntu20.04 - CUDA version (11.8) mismatches the version that was used to compile PyTorch (12.1)

Issue - State: closed - Opened by hammad26 about 1 year ago - 1 comment

#53 - What is the `max_seq_len` in Mistral?

Issue - State: open - Opened by ParadoxZW about 1 year ago - 1 comment

#52 - Add simple classification example

Pull Request - State: closed - Opened by timlacroix about 1 year ago

#51 - Ray qelr_async_event not implemented yet

Issue - State: open - Opened by Ryojikn about 1 year ago

#50 - Error on run main

Issue - State: open - Opened by lrx1213 about 1 year ago - 1 comment

#49 - [Model] Refactoring model.py into small modules

Pull Request - State: closed - Opened by sarveshwar-s about 1 year ago - 1 comment

#48 - it's fantastic! but can do 1.1b , 3b versions too?

Issue - State: open - Opened by hiqsociety about 1 year ago

#47 - Are `RotatingBufferCache` and `RollingBufferCache` the same thing?

Issue - State: closed - Opened by ParadoxZW about 1 year ago - 1 comment

#46 - Update and rename main.py to mainwithcomments.py

Pull Request - State: closed - Opened by nikcode9 about 1 year ago - 2 comments

#45 - Python 3.11.6 compatibility

Issue - State: open - Opened by MasterLivens about 1 year ago - 3 comments

#44 - Update README.md

Pull Request - State: closed - Opened by infwinston about 1 year ago - 2 comments

#43 - Update PyTorch to 2.2.0 to support NVIDIA H100 PCIe

Pull Request - State: closed - Opened by quantumsheep about 1 year ago - 3 comments

#42 - python process keeps getting killed

Issue - State: closed - Opened by 5hayanB about 1 year ago - 1 comment

#41 - How many tokens did Mistral-7B train on?

Issue - State: closed - Opened by ninjasaid2k about 1 year ago

#40 - Questions about layer-wise sliding window attention

Issue - State: closed - Opened by NormXU about 1 year ago - 13 comments

#39 - very good! thx! but...

Issue - State: closed - Opened by hiqsociety about 1 year ago

#38 - one_file_ref.py attention has an O(seqlen^2) matrix multiplication when prefilling

Issue - State: closed - Opened by Aniruddha-Deb about 1 year ago - 1 comment

#37 - 🦒 colab

Issue - State: closed - Opened by camenduru about 1 year ago - 1 comment

#36 - Can you provide lora tutorial for mistral 7b instruction model on custom dataset?

Issue - State: open - Opened by universewill about 1 year ago - 1 comment

#35 - System prompt handling in chat templates for Mistral-7b-instruct

Issue - State: closed - Opened by jamesr66a about 1 year ago - 5 comments

#34 - Mistral on CPU

Issue - State: open - Opened by pruthvi1990 about 1 year ago - 2 comments

#33 - .bin format?

Issue - State: open - Opened by StanislawKarnacky about 1 year ago

#32 - Tokenizer.model error on pycharm

Issue - State: open - Opened by dominique-AR about 1 year ago - 1 comment

#31 - Update Dockerfile

Pull Request - State: closed - Opened by lerela about 1 year ago

#30 - Mistral-7B-instruct-v0.1 compatibility with main.py

Issue - State: open - Opened by nvidal01 about 1 year ago - 4 comments

#29 - fix URL typo

Pull Request - State: closed - Opened by VictorNanka about 1 year ago

#28 - Dilation ?

Issue - State: open - Opened by edmondja about 1 year ago - 1 comment

#27 - Update README.md

Pull Request - State: closed - Opened by Emporea about 1 year ago

#26 - Add top_k text decoding

Pull Request - State: closed - Opened by aahouzi about 1 year ago

#25 - ValueError: No available memory for the cache blocks.

Issue - State: open - Opened by Stoobiedoo about 1 year ago - 1 comment

#24 - Update README.md

Pull Request - State: closed - Opened by numaroth about 1 year ago - 1 comment

#23 - test Mistral / llama2 with flowise and replicate

Issue - State: open - Opened by scenaristeur about 1 year ago

#22 - Passkey retrieval results

Issue - State: open - Opened by RonanKMcGovern about 1 year ago

#21 - Update README.md

Pull Request - State: closed - Opened by eltociear about 1 year ago

#20 - Out of Memory after training a few epochs

Issue - State: open - Opened by waylonli about 1 year ago

#19 - Add Dockerfile and build instructions

Pull Request - State: closed - Opened by lerela about 1 year ago - 2 comments

#18 - ONNX?

Issue - State: closed - Opened by DiTo97 about 1 year ago - 1 comment

#16 - python3: No module named main

Issue - State: open - Opened by happybeing about 1 year ago - 2 comments

#15 - Custom Training Pipeline ?

Issue - State: open - Opened by AMEERAZAM08 about 1 year ago

#14 - Error on interactive run

Issue - State: open - Opened by sreekarchigurupati about 1 year ago - 3 comments

#13 - Update README.md

Pull Request - State: closed - Opened by devendrachaplot about 1 year ago

#12 - best out of the box yet

Issue - State: closed - Opened by silvacarl2 about 1 year ago - 1 comment

#11 - Addition docs adhering PEP257 and PEP8

Pull Request - State: open - Opened by rajveer43 about 1 year ago - 4 comments

#10 - documentation is required

Issue - State: open - Opened by rajveer43 about 1 year ago

#9 - Missing model card / data sheet with info on pretraining and RLHF datasets

Issue - State: open - Opened by mdingemanse about 1 year ago - 4 comments

#8 - Are you using window attention for training?

Issue - State: open - Opened by logicwong about 1 year ago - 1 comment

#7 - Xformers cannot be installed on MAC M1 Pro

Issue - State: open - Opened by Naqqash about 1 year ago - 6 comments

#6 - Compatible with Intel Arc dGPUs?

Issue - State: open - Opened by prakal about 1 year ago - 1 comment

#5 - Prompt for RAG

Issue - State: open - Opened by Matthieu-Tinycoaching about 1 year ago

GitHub / mistralai/mistral-inference issues and pull requests