openmoss/anygpt issues and pull requests

#49 - CUDA Runtime Error: Device-Side Assert Triggered

Issue - State: open - Opened by Dj12367 about 1 month ago

#48 - About pretrain dataset

Issue - State: open - Opened by Yang-Liz about 1 month ago

#47 - BertLMHeadModel.from_pretrained("bert-base-uncased", config=encoder_config)

Issue - State: open - Opened by baipeng1 about 2 months ago

#46 - Where is output_models/mm_pretrain2/checkpoint-4000?

Issue - State: open - Opened by Psychomachia 3 months ago

#45 - Question about stage1_pretrain

Issue - State: closed - Opened by Psychomachia 4 months ago

#44 - High resolution images are not found

Issue - State: open - Opened by binisalegend 4 months ago - 3 comments

#43 - Does AnyGPT support multi images input?

Issue - State: closed - Opened by QAQdev 5 months ago

#42 - Details on TTS evaluation?

Issue - State: open - Opened by Btlmd 5 months ago - 3 comments

#41 - preprocessing audio to SpeechTokenizer codecs

Issue - State: open - Opened by ehosseiniasl 6 months ago

#40 - Regarding ASR testing

Issue - State: open - Opened by Simplesss 6 months ago - 4 comments

#39 - targets were set to all -100 in stage2_sft stage due to cur_len != expected_len

Issue - State: open - Opened by hchc007 6 months ago - 2 comments

#38 - Music Tokenizer Format

Issue - State: open - Opened by ayaan-together 7 months ago

#37 - 请问大佬有做过单独模态的消融实验吗

Issue - State: closed - Opened by silvercherry 7 months ago - 1 comment

#36 - fix:args parameters didn't pass to AnyGPTInference correctly

Pull Request - State: open - Opened by SoftEgLi 7 months ago

#35 - cannot import name 'split_torch_state_dict_into_shards' from 'huggingface_hub'

Issue - State: open - Opened by cubxx 7 months ago - 2 comments

#34 - data format in mmichat_*.jsonl

Issue - State: open - Opened by hchc007 7 months ago - 1 comment

#33 - 请教一下关于speech的问题

Issue - State: closed - Opened by silvercherry 7 months ago - 1 comment

#32 - Speech-to-Speech task prompt

Issue - State: open - Opened by ehosseiniasl 7 months ago - 6 comments

#31 - Train_loss = 0 and Eval_loss = NaN in stage2_sft

Issue - State: open - Opened by xuxiaoang 7 months ago - 4 comments

#30 - Question about training code

Issue - State: open - Opened by Gaffey 7 months ago - 4 comments

#29 - is there any way to reduce latency?

Issue - State: closed - Opened by kaen2891 7 months ago - 3 comments

#28 - 请教一下关于AnyGPT模型预训练的一些问题

Issue - State: closed - Opened by shihuai 8 months ago - 2 comments

#27 - please provide a TTS prompt example for chat.

Issue - State: closed - Opened by sipie800 8 months ago - 1 comment

#26 - How to train the speech tokneizer?

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

#25 - About input formats for training and inference

Issue - State: open - Opened by wen020 9 months ago - 2 comments

#24 - GPU resources for AnyGPT instruction turning

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

#23 - Consider next version to use LLM instead of UNet

Issue - State: closed - Opened by matbee-eth 9 months ago

#22 - 请教下music vocabulary size of 8192的实现

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

#21 - 关于论文中音乐tokenize和音乐生成示例的问题

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

#20 - Loss Masking

Issue - State: closed - Opened by gmltmd789 10 months ago - 2 comments

#19 - Some weights of BertLMHeadModel were not initialized from the model checkpoint at bert-base-uncased and are newly initialized

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

#18 - hi，when will the pre-train related codes&scripts be released?

Issue - State: closed - Opened by XL2248 10 months ago - 4 comments

#17 - text chat is not so good.

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

#16 - can i ask the model to choose which of the voices sound more natural?

Issue - State: closed - Opened by s-j-chung 11 months ago - 2 comments

#15 - RuntimeError: Error(s) in loading state_dict for SoundStorm

Issue - State: closed - Opened by empty2enrich 11 months ago - 1 comment

#14 - qformer_quantizer.py missing keys: 511 unexpected keys: 146

Issue - State: closed - Opened by empty2enrich 11 months ago - 1 comment

#13 - Question about training stage and dataset

Issue - State: closed - Opened by hyx100e 11 months ago - 3 comments

#12 - add Docker

Pull Request - State: open - Opened by Sunwood-ai-labs 11 months ago

#11 - Bug Fix (#9, #10) & README JP & Update README

Pull Request - State: open - Opened by Sunwood-ai-labs 11 months ago

#10 - huggingface_hub.utils._validators.HFValidationError: Repo id must be in the form 'repo_name' or 'namespace/repo_name': '/mnt/petrelfs/zhanjun.p/mllm/models/bert-base-uncased'. Use `repo_type` argument if needed.

Issue - State: closed - Opened by Sunwood-ai-labs 11 months ago - 1 comment

#9 - ModuleNotFoundError: No module named 'mmgpt.src'

Issue - State: closed - Opened by Sunwood-ai-labs 11 months ago - 1 comment

#8 - hi，when will the train code be released? && do you train image and text tokens all in autoregressive?

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

#7 - add JP readme

Pull Request - State: closed - Opened by Sunwood-ai-labs 11 months ago

#6 - Collaboration request

Issue - State: closed - Opened by DeepDream2045 11 months ago

#5 - When will code, dataset and checkpoints be released?

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

#4 - how many a100 it cost in training and if i want to train on v100 what is the number needed ?

Issue - State: closed - Opened by Yang-bug-star 12 months ago - 4 comments

#3 - code/dataset/model

Issue - State: closed - Opened by darkacorn almost 1 year ago - 1 comment

#2 - Update README.md

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

#1 - Is there any evaluation on VQA datasets?

Issue - State: closed - Opened by jzhang38 about 1 year ago - 1 comment

GitHub / openmoss/anygpt issues and pull requests