vikhyatk/moondream2 · Discussions

#74 opened about 1 month ago by

KeilahElla

Update model in ollama

#73 opened 2 months ago by

Yussef1234

ethan/q

#72 opened 2 months ago by

err805

Detailed Architecture information.

#71 opened 3 months ago by

Sharath07

❗️Inference Fails with Type Mismatch Error: self and mat2 must have the same dtype, but got Half and Char

#70 opened 4 months ago by

Manaf-paar

Fails to load local version: FileNotFoundError: ~/.cache/huggingface/modules/transformers_modules/rope.py

#69 opened 4 months ago by

treehugg3

Inference Fails with Type Mismatch Error: self and mat2 must have the same dtype, but got Half and Char

#68 opened 4 months ago by

Manaf-paar

ModuleNotFoundError: No module named 'transformers_modules.vikhyatk'

#67 opened 4 months ago by

anubhavshavaran

Detecting multiple classes and grounding

#66 opened 4 months ago by

harpreetsahota

Difficulty deploying to Sagemaker AI

#65 opened 4 months ago by

josiahschuller

How to get raw_logits from the Moondream model?

#64 opened 5 months ago by

dutta18

JSON Processing

#63 opened 5 months ago by

dalematthews9

cannot use the local code

#62 opened 6 months ago by

jz97

Mac MPS Example Code and Acceleration

#61 opened 6 months ago by

RamboRogers

Exploring the Potential for a MoonDream-R1 Model

❤️ 🔥 6

#60 opened 6 months ago by

PlayAI

Unable to use from transformers

➕ 👍 3

#59 opened 6 months ago by

sraj

Plans to make the processor compatible with transformers Trainer?

#58 opened 6 months ago by

RonanMcGovern

Unable to load finetuned model after saving

#57 opened 6 months ago by

Charlington

Error while deserializing header: InvalidHeaderDeserialization

#56 opened 6 months ago by

bbnns

Dope model - saw you talk about it in Vancouver!

#55 opened 6 months ago by

harpreetsahota

Very powerful, but the recognition of mathematical formulas is not ideal.

#54 opened 6 months ago by

tianyang1

cannot load the model with transformers

6

#53 opened 7 months ago by

rnyak

Multiple Images

#52 opened 7 months ago by

RonanMcGovern

Localization issues with bounding boxes

#51 opened 7 months ago by

Rainbowl

Recommended server endpoint

👍 1

#49 opened 7 months ago by

RonanMcGovern

Finetuning for pointing, object detection task

#48 opened 7 months ago by

yaneivan

Thundermoon is very impressive

#46 opened 7 months ago by

Hangover3832

No module named 'transformers_modules.moondream2.fourier_features'

5

#45 opened 7 months ago by

Hangover3832

How do we download the 0.5B model

👍 16

#44 opened 8 months ago by

sachin

Moondream 0.5B Ollama GGUF needed

#43 opened 8 months ago by

csepartha

ollama run hf.co/vikhyatk/moondream2 doesn't work

👍 5

#41 opened 9 months ago by

Davidasx

Create modelfile with "Projector"

#40 opened 9 months ago by

jphen

Vision Encoder does not scale well on batched images as input

#38 opened 9 months ago by

IanLi233

Update GGUF to latest and also 8bit / 4bit

👍 4

#36 opened 10 months ago by

ssmits

Fine tuning for a more diverse and accurate description of a person/an animal?

#35 opened 11 months ago by

Leo333

Fine tuning moondream for OCR

#33 opened 12 months ago by

sunilchandwani

How many objects can Moondriam2 recognize?

#31 opened 12 months ago by

canecho

How to fix the IndexError caused by latest transformers update

🔥 2

#28 opened about 1 year ago by

dylanhogg

Leaderboard

#26 opened about 1 year ago by

Stark2008

need new models versions (revisions) gguf files

#25 opened about 1 year ago by

mohan007

Getting logits and kv_cache of moondream

#24 opened about 1 year ago by

sachin

I try to create inference using a service and it fails every time, thoughts? (across all different services)

7

#23 opened about 1 year ago by

Sirianth

the model output：

#21 opened about 1 year ago by

Deepfor

Are the weights of MoonDream2 are initialized from SigLIP and Phi 1.5?

#20 opened about 1 year ago by

HoqueMahmudul

Why does the Moondream demo give better answers than my local one?

#19 opened about 1 year ago by

bermancheese

Proper Chat Format?

#17 opened about 1 year ago by

ElMater06

Flexibility to change max generated token.