Running
172
π’
Kacper GΕombiewski
Clausss
AI & ML interests
None yet
Recent Activity
liked
a Space
6 days ago
sambanovasystems/SambaNova-Qwen2.5-Coder-Artifacts
reacted
to
csabakecskemeti's
post
with π
6 days ago
The AMD Instinct MI50 (~$110) is surprisingly fast for inference Quantized models.
This runs a Llama 3.1 8B Q8 with Llama.cpp
https://huggingface.co/spaces/DevQuasar/Mi50
A little blogpost about the HW
http://devquasar.com/uncategorized/amd-radeon-instinct-mi50-cheap-inference/
liked
a Space
6 days ago
huggingface/open-source-ai-year-in-review-2024
Organizations
None yet
Clausss's activity
liked
a
Space
6 days ago
reacted to
csabakecskemeti's
post with π
6 days ago
Post
4443
The AMD Instinct MI50 (~$110) is surprisingly fast for inference Quantized models.
This runs a Llama 3.1 8B Q8 with Llama.cpp
DevQuasar/Mi50
A little blogpost about the HW
http://devquasar.com/uncategorized/amd-radeon-instinct-mi50-cheap-inference/
This runs a Llama 3.1 8B Q8 with Llama.cpp
DevQuasar/Mi50
A little blogpost about the HW
http://devquasar.com/uncategorized/amd-radeon-instinct-mi50-cheap-inference/
liked
a
Space
6 days ago
upvoted
an
article
10 days ago
Article
Introducing GGUF-my-LoRA
By
β’
β’
12
liked
a
Space
16 days ago
replied to
nroggendorff's
post
18 days ago
so what happen if storage beyond limit?
liked
2
Spaces
22 days ago
reacted to
mikelabs's
post with π₯
26 days ago
Post
1589
SAM: I can segment anything!
SAMURAI: Hold my sake while I track it through time and space without even training π―
Researchers really said what if we gave SAM object permanence and it worked π€―
https://www.aimodels.fyi/papers/arxiv/samurai-adapting-segment-anything-model-zero-shot
SAMURAI: Hold my sake while I track it through time and space without even training π―
Researchers really said what if we gave SAM object permanence and it worked π€―
https://www.aimodels.fyi/papers/arxiv/samurai-adapting-segment-anything-model-zero-shot
- 2 replies
Upload 81 files
1
#8 opened 29 days ago
by
Jorge-Ali
reacted to
fdaudens's
post with β€οΈ
27 days ago
Post
1883
π¦ Hug the butterfly! You can now add your Bluesky handle to your Hugging Face profile! β¨
Update app.py
1
#7 opened about 1 month ago
by
yesirecarolinasoto
reacted to
etemiz's
post with π₯
about 1 month ago
Post
969
if I host in hf spaces, can I interact with the app using an API?
- 1 reply
Error quantizing: b'/bin/sh: 1: ./llama.cpp/llama-quantize: not found\n'
6
#136 opened about 1 month ago
by
win10
reacted to
m-ric's
post with π
about 1 month ago
Post
2367
A non-Instruct LLM assistant is mostly useless. π§
Since it's mostly a model trained to complete text, when you ask it a question like "What to do during a stopover in Paris?", it can just go on and on adding more details to your question instead of answering, which would be valid to complete text from its training corpus, but not to answer questions.
β‘οΈ So the post-training stage includes an important Instruction tuning step where you teach your model how to be useful : answer questions, be concise, be polite... RLHF is a well known technique for this.
For people interested to understand how this step works, the folks at Adaptive ML have made a great guide!
Read it here π https://www.adaptive-ml.com/post/from-zero-to-ppo
Since it's mostly a model trained to complete text, when you ask it a question like "What to do during a stopover in Paris?", it can just go on and on adding more details to your question instead of answering, which would be valid to complete text from its training corpus, but not to answer questions.
β‘οΈ So the post-training stage includes an important Instruction tuning step where you teach your model how to be useful : answer questions, be concise, be polite... RLHF is a well known technique for this.
For people interested to understand how this step works, the folks at Adaptive ML have made a great guide!
Read it here π https://www.adaptive-ml.com/post/from-zero-to-ppo
reacted to
nroggendorff's
post with π
about 1 month ago
Post
2259
I still think whitespace in tokenizers are so dumb.
Congrats, you just doubled your vocab size for no reason.
Congrats, you just doubled your vocab size for no reason.
- 3 replies