Joe lafrite

joelafrite's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in drowzeys/keys-DeepSeek-V4.1-Flash-1017-Abliterated 14 days ago

2x DGX Spark (Native Zig 1M TP=2) benchmark receipt: 0% refusal and zero coding regression vs stock LibertAI 2.9bpw

#7 opened 14 days ago by

joelafrite

New activity in unsloth/Qwen3.8-27B-NVFP4 14 days ago

This checkpoint ships calibrated fp8 KV-cache scales, but vLLM never loads them

👍 2

#28 opened 14 days ago by

joelafrite

liked a model 14 days ago

drowzeys/keys-Qwen3.8-Flash-Next-NVFP4-FP4-QAD-house-qsa-L3-47

Image-Text-to-Text • 120B • Updated 14 days ago • 311 • 26

liked a model 17 days ago

nvidia/GLM-5.3-Flash-NVFP4

Image-Text-to-Text • 321B • Updated 17 days ago • 66.5k • 191

New activity in LibertAIDAI/GLM-5.3-Flash-NVFP4 26 days ago

Datapoint: fp4 vs fp8 GLM-5.3-Flash endpoint parity on GSM8K + HumanEval

2

#8 opened 3 months ago by

joelafrite

liked a model 26 days ago

joelafrite/Qwen3.8-27B-fp8-KV-scale-calibration

Updated 28 days ago • 44 • 1

New activity in joelafrite/Qwen3.8-27B-fp8-KV-scale-calibration 27 days ago

How to generate the calibrated KV Scales?

1

#1 opened 27 days ago by

Goghor

New activity in LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw 27 days ago

Thank you for the quick upload

➕ 1

3

#1 opened 28 days ago by

joelafrite

New activity in dealignai/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw-UNCENSORED 27 days ago

All-EXL3 variant? `wo_b` / `shared_experts.w2` are BF16 in layers 10–36

❤️ 1

#1 opened 27 days ago by

joelafrite

liked a model 28 days ago

deepseek-ai/DeepSeek-V4.1-Flash-1017

Image-Text-to-Text • 763B • Updated 4 days ago • 398k • • 5.68k

updated a model 28 days ago

joelafrite/Qwen3.8-27B-fp8-KV-scale-calibration

Updated 28 days ago • 44 • 1

liked a model 28 days ago

LibertAIDAI/DeepSeek-V4.1-Flash-1017-EXL3-2.9bpw

Text Generation • 105B • Updated 26 days ago • 7.23k • 60

New activity in RadixArk/Qwen3.8-Flash-Next-NVFP4 about 1 month ago

Degeneration Loop when calling tool (PDF/Image)

🔥 1

4

#3 opened 3 months ago by

dangnsh

liked 4 models about 2 months ago

flukethoughts/Behemoth-128B-v3-NVFP4

72B • Updated Sep 8 • 36 • 3

Mia-AiLab/DeepSeek-V4.1-Flash-EXL3-2.9bpw

Image-Text-to-Text • 105B • Updated Sep 13 • 2.11k • 23

dealignai/DeepSeek-V4.1-Flash-UNCENSORED-EXL3-2.9bpw

Image-Text-to-Text • 105B • Updated Sep 17 • 895 • 47

RedHatAI/GLM-5.3-Flash-NVFP4

Image-Text-to-Text • 169B • Updated 5 days ago • 10.6k • 50

liked 2 models 2 months ago

deepseek-ai/DeepSeek-V4.1-Flash

Image-Text-to-Text • 763B • Updated 29 days ago • 1.77M • • 4.97k

deepseek-ai/DeepSeek-V4-Flash-Vision-Exp

Image-Text-to-Text • 305B • Updated Sep 1 • 138k • • 970

New activity in gratex/Qwen3.8-27B-DFlash2-W4A16-g128-sym-GPTQ 2 months ago

apply_dense_kv_fix.py: silent corruption risk with NVFP4 drafters (+ working variant)

3

#1 opened 3 months ago by

joelafrite