basically-ai/Pebble - GGUF adaptation
Collection
2 items โข Updated โข 1
ConReq.
ContextReq
AI & ML interests
DIY - mess around and findout, all and every aspect.
Recent Activity
upvoted a collection about 1 hour ago
basically-ai/Pebble - GGUF adaptation updated a collection about 1 hour ago
basically-ai/Pebble - GGUF adaptation updated a collection about 1 hour ago
basically-ai/Pebble - GGUF adaptationOrganizations
None yet
upvoted a collection about 1 hour ago
updated a
collection about 1 hour ago
upvoted a collection about 1 hour ago
updated 2
models about 18 hours ago
published 2
models about 18 hours ago
Holy sh, a small language model that is mamab!
๐ 1
10
#1 opened 1 day ago
by 
FlameF0X
reacted to Hoglet-33's post with ๐ฅ 2 days ago
Post
3079
Pebble 10M and Pebble 10M Chat are now released!
Both models use our Mamba/Transformer 3:1 hybrid architecture and were pretrained on 25 billion tokens.
Pebble 10M Chat was additionally fine-tuned on 250 million tokens of Smol-SmolTalk to improve its conversational capabilities.
You can find them here:
- basically-ai/Pebble-10M
- basically-ai/Pebble-10M-Chat
We hope you enjoy using them. The rest of the Pebble family will be released soon.
Follow for more:
@Hoglet-33
basically-ai
Both models use our Mamba/Transformer 3:1 hybrid architecture and were pretrained on 25 billion tokens.
Pebble 10M Chat was additionally fine-tuned on 250 million tokens of Smol-SmolTalk to improve its conversational capabilities.
You can find them here:
- basically-ai/Pebble-10M
- basically-ai/Pebble-10M-Chat
We hope you enjoy using them. The rest of the Pebble family will be released soon.
Follow for more:
@Hoglet-33
- 19 replies
liked 2
models 2 days ago
reacted to appvoid's post with ๐ฅ 4 days ago
Post
2612
Nobody knows what is doing, when you train a model, you are experimenting to advance the frontier, so keep failing ๐ซต
- 17 replies
"terrible engineering" - You've got to start somewhere to learn. As someone who has spent years as a hobby. If you care enough then you learn from. "This unscientific approach is why the ecosystem is flooded with overfitted, hollow checkpoints that break down outside of their strict test sets.You aren't advancing the frontier; you are just polluting the platform because you refuse to open a textbook." - Same goes to learn from. Partly projecting with this next part: Imagine gatekeeping ML out of elitism because people learn differently and have different goals. A public platform exists to share what we learn and discover so that others may advance with or separately.