YaTharThShaRma999/schnell_lora
Updated
Yatharth Sharma
YaTharThShaRma999
AI & ML interests
None yet
Recent Activity
updated
a model
1 day ago
YaTharThShaRma999/schnell_lora
published
a model
1 day ago
YaTharThShaRma999/schnell_lora
upvoted
a
paper
4 days ago
Compass Control: Multi Object Orientation Control for Text-to-Image
Generation
Organizations
None yet
YaTharThShaRma999's activity
updated
a
model
1 day ago
published
a
model
1 day ago
upvoted
a
paper
4 days ago
reacted to
AdinaY's
post with 🔥
5 days ago
Post
3086
Shanghai AI Lab - OpenGV team just released InternVL3 🔥
OpenGVLab/internvl3-67f7f690be79c2fe9d74fe9d
✨ 1/2/8/9/14/38/28B with MIT license
✨ Stronger perception & reasoning vs InternVL 2.5
✨ Native Multimodal Pre-Training for even better language performance
OpenGVLab/internvl3-67f7f690be79c2fe9d74fe9d
✨ 1/2/8/9/14/38/28B with MIT license
✨ Stronger perception & reasoning vs InternVL 2.5
✨ Native Multimodal Pre-Training for even better language performance
- 1 reply
upvoted
a
paper
5 days ago
reacted to
onekq's
post with 🚀
6 days ago
Post
2550
We desperately need GPU for model inference. CPU can't replace GPU.
I will start with the basics. GPU is designed to serve predictable workloads with many parallel units (pixels, tensors, tokens). So a GPU allocates as much transistor budget as possible to build thousands of compute units (Cuda cores in NVidia or execution units in Apple Silicon), each capable of running a thread.
But CPU is designed to handle all kinds of workloads. CPU cores are much larger (hence a lot fewer) with branch prediction and other complex things. In addition, more and more transistors are allocated to build larger cache (~50% now) to house the unpredictable, devouring the compute budget.
Generalists can't beat specialists.
I will start with the basics. GPU is designed to serve predictable workloads with many parallel units (pixels, tensors, tokens). So a GPU allocates as much transistor budget as possible to build thousands of compute units (Cuda cores in NVidia or execution units in Apple Silicon), each capable of running a thread.
But CPU is designed to handle all kinds of workloads. CPU cores are much larger (hence a lot fewer) with branch prediction and other complex things. In addition, more and more transistors are allocated to build larger cache (~50% now) to house the unpredictable, devouring the compute budget.
Generalists can't beat specialists.
- 4 replies
updated
a
model
6 days ago
liked
a
Space
7 days ago
liked
a
model
7 days ago
liked
a
model
8 days ago
liked
a
Space
8 days ago
liked
a
model
9 days ago
upvoted
a
collection
9 days ago
reacted to
sr-rai's
post with 🤯🤯🤗
9 days ago
Post
2593
ExLlamaV3 is out. And it introduces EXL3 - a new SOTA quantization format!
"The conversion process is designed to be simple and efficient and requires only an input model (in HF format) and a target bitrate. By computing Hessians on the fly and thanks to a fused Viterbi kernel, the quantizer can convert a model in a single step, taking a couple of minutes for smaller models, up to a few hours for larger ones (70B+) (on a single RTX 4090 or equivalent GPU.)"
Repo: https://github.com/turboderp-org/exllamav3
"The conversion process is designed to be simple and efficient and requires only an input model (in HF format) and a target bitrate. By computing Hessians on the fly and thanks to a fused Viterbi kernel, the quantizer can convert a model in a single step, taking a couple of minutes for smaller models, up to a few hours for larger ones (70B+) (on a single RTX 4090 or equivalent GPU.)"
Repo: https://github.com/turboderp-org/exllamav3
- 1 reply
reacted to
jsulz's
post with 🔥
10 days ago
Post
3586
Huge week for
xet-team
as Llama 4 is the first major model on Hugging Face uploaded with Xet providing the backing! Every byte downloaded comes through our infrastructure.
Using Xet on Hugging Face is the fastest way to download and iterate on open source models and we've proved it with Llama 4 giving a boost of ~25% across all models.
We expect builders on the Hub to see even more improvements, helping power innovation across the community.
With the models on our infrastructure, we can peer in and see how well our dedupe performs across the Llama 4 family. On average, we're seeing ~25% dedupe, providing huge savings to the community who iterate on these state-of-the-art models. The attached image shows a few selected models and how they perform on Xet.
Thanks to the
meta-llama
team for launching on Xet!
xet-team
as Llama 4 is the first major model on Hugging Face uploaded with Xet providing the backing! Every byte downloaded comes through our infrastructure.Using Xet on Hugging Face is the fastest way to download and iterate on open source models and we've proved it with Llama 4 giving a boost of ~25% across all models.
We expect builders on the Hub to see even more improvements, helping power innovation across the community.
With the models on our infrastructure, we can peer in and see how well our dedupe performs across the Llama 4 family. On average, we're seeing ~25% dedupe, providing huge savings to the community who iterate on these state-of-the-art models. The attached image shows a few selected models and how they perform on Xet.
Thanks to the
meta-llama
team for launching on Xet!
liked
a
model
11 days ago
liked
a
model
13 days ago
liked
a
model
14 days ago
