LiquidAI/LFM2-1.2B
Text Generation
•
1B
•
Updated
•
6.8k
•
189
Djuunaa
djuna
AI & ML interests
None yet
Recent Activity
liked
a model
6 days ago
LiquidAI/LFM2-1.2B
liked
a model
6 days ago
HaochenWang/TreeVGR-7B
liked
a model
8 days ago
Bifrost-AI/Qwen3-Bifrost-SOL-4B
Organizations
liked
2
models
6 days ago
liked
2
models
8 days ago
liked
3
models
10 days ago
reacted to
eaddario's
post with 👍
13 days ago
Post
3771
Layer-wise and Pruned versions of cognitivecomputations/Dolphin-Mistral-24B-Venice-Edition
* Tesor-wise: eaddario/Dolphin-Mistral-24B-Venice-Edition-GGUF
* Pruned: eaddario/Dolphin-Mistral-24B-Venice-Edition-pruned-GGUF
Summary in the model's card and test results in the ./scores directory. Questions/feedback is always welcomed.
* Tesor-wise: eaddario/Dolphin-Mistral-24B-Venice-Edition-GGUF
* Pruned: eaddario/Dolphin-Mistral-24B-Venice-Edition-pruned-GGUF
Summary in the model's card and test results in the ./scores directory. Questions/feedback is always welcomed.
liked
a
model
16 days ago
reacted to
eaddario's
post with 🚀
20 days ago
Post
3729
Layer-wise and Pruned versions of Qwen/Qwen3-30B-A3B
* Tesor-wise: eaddario/Qwen3-30B-A3B-GGUF
* Pruned: eaddario/Qwen3-30B-A3B-pruned-GGUF
Even though the Perplexity scores of the pruned version are 3 times higher, the ARC, HellaSwag, MMLU, Truthful QA and WinoGrande scores are holding remarkably well, considering two layers were removed (5 and 39). This seems to support Xin Men et al conclusions in
ShortGPT: Layers in Large Language Models are More Redundant Than You Expect (2403.03853)
Results summary in the model's card and test results in the ./scores directory. Questions/feedback is always welcomed.
* Tesor-wise: eaddario/Qwen3-30B-A3B-GGUF
* Pruned: eaddario/Qwen3-30B-A3B-pruned-GGUF
Even though the Perplexity scores of the pruned version are 3 times higher, the ARC, HellaSwag, MMLU, Truthful QA and WinoGrande scores are holding remarkably well, considering two layers were removed (5 and 39). This seems to support Xin Men et al conclusions in
ShortGPT: Layers in Large Language Models are More Redundant Than You Expect (2403.03853)
Results summary in the model's card and test results in the ./scores directory. Questions/feedback is always welcomed.
reacted to
FlameF0X's
post with 👍
21 days ago
Post
1931
SnowflakeCore-G1 development update: We're building a 24-layer transformer with 32K context and 1024 embedding dimensions - pretty ambitious! Even running at batch_size=1 with heavy gradient accumulation, we're hitting memory walls at 300GB RAM. Scaling up to ~1TB will take some time, but the architecture is looking promising. Thanks for following along with the journey! 😅
- 1 reply
liked
2
models
21 days ago
Any plans to release an updated version based on DeepSeek-V3-0526 + R1, or how to create the merge myself?
14
#4 opened about 2 months ago
by
Lissanro
liked
a
model
24 days ago
liked
4
models
25 days ago
liked
a
model
26 days ago
