i-Code

i-Code copied to clipboard

Trainig pipeline of CoDi in i-Code-V3

7

Hello, is the training code of CoDi available now? There is no discription in README.md.

zhanganny

I would try to finetune the model for a specific domain. Following the paper, i was able to isolate the prompt encoders for their contrastive training, but i don't know...

danielemolino

Hi, Thanks for sharing this awesome work. I have a question on COCO image captioning benchmark. The BLEU4 scores in Table 6 is around 40, but my implementation shows BLEU4...

HoinJung

Finetuning on InfographicVQA

12

I was unable to achieve the result shown in the [UDOP paper](https://arxiv.org/abs/2212.02623). I used the [udop-unimodel-large-224 checkpoint](https://huggingface.co/ZinengTang/Udop/resolve/main/udop-unimodel-large-224.zip). My ANLS score is **0.407903**. This is nowhere near 0.461 as shown in...

Caixin89

Generating bounding boxes with UDOP

7

Hi, by reading the UDOP paper, my understanding is that during pre-training the model is taught to predict the layout of a target (textual) sequence using special layout tokens. I...

AleRosae

Hi there! Firstly, thank you for your amazing work and research! I just wanted to ask, maybe you could provide an example of document editing/generation script using UDOP model? I...

boris-lok-pentadoc

Are there any plans to release the weights of the UDOP model trained on 1024x1024 resolution? Thanks

ageorgiev97

Finetuning on Due-Benchmark

10

Hi, I have been trying to finetune the model on due-benchmark using the provided script. However, the performance is quite low compared to the reported numbers. For example, DocVQA results...

swathikirans

Bumps [transformers](https://github.com/huggingface/transformers) from 4.30.0 to 4.36.0. Release notes Sourced from transformers's releases. v4.36: Mixtral, Llava/BakLlava, SeamlessM4T v2, AMD ROCm, F.sdpa wide-spread support New model additions Mixtral Mixtral is the new...

dependabot[bot]

In the directory i-Code-V3/core/models/encoders/clap_modules/training, I attempted to execute "python main.py," and I encountered the following information: Traceback (most recent call last): File "main.py", line 31, in from open_clip import create_model_and_transforms,...

iMoriton