Kshitij Thakkar's picture

Building on HF

Kshitij Thakkar PRO

kshitijthakkar

·

AI & ML interests

AI observability + MoE efficiency engineer. Building tools that make GenAI traceable, measurable, and production-ready.

Organizations

kshitijthakkar 's models 130

kshitijthakkar/moe-340m-107m-24x2-12L-24exp-specialized

kshitijthakkar/moe-350m-102m-16x1-12L-top1-routing

kshitijthakkar/moe-274m-132m-16x4-12L-top4-routing

kshitijthakkar/moe-240m-103m-12x2-16L-deep-narrow-16l

kshitijthakkar/moe-270m-132m-12x2-8L-shallow-wide-8l

kshitijthakkar/moe-229m-111m-12x2-10L-full-attention-no-gqa

kshitijthakkar/moe-284m-119m-12x2-14L-aggressive-gqa-1kv

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr5e-06

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr1e-05

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr3e-05

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr5e-05

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr1e-04

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr2e-04

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr3e-04

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr5e-04

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-lr1e-03

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-bs2-ctx512

kshitijthakkar/moe-255m-114m-12x2-12L-full-attention-no-gqa-bs2-ctx1024

kshitijthakkar/loggenix-moe-0.4B-0.2A-sft-s3.1

Text Generation • 0.4B • Updated Jan 28 • 145

kshitijthakkar/loggenix-moe-0.6B-base-pt-cpt-test

kshitijthakkar/loggenix-moe-1.3B-base-cpt-test

Updated Dec 24, 2025

kshitijthakkar/loggenix-moe-2.5B-base

Text Generation • 3B • Updated Dec 22, 2025 • 1

kshitijthakkar/loggenix-moe-0.12B-A0.08B-e5-lr5e4-b4-3060

Text Generation • 0.1B • Updated Dec 17, 2025 • 6 • 1

kshitijthakkar/loggenix-moe-1.3B-base

Text Generation • 1B • Updated Dec 15, 2025 • 3

kshitijthakkar/loggenix-moe-0.6B-base

Text Generation • 0.6B • Updated Dec 12, 2025 • 1

kshitijthakkar/loggenix-moe-0.3B-A0.1B-smolagents-v1-quantized

Text Generation • 0.3B • Updated Dec 12, 2025 • 4

kshitijthakkar/loggenix-moe-0.3B-A0.1B-smolagents-v0-grpo

Text Generation • 0.3B • Updated Dec 11, 2025 • 28

kshitijthakkar/loggenix-moe-0.3B-A0.1B-smolagents-v0

Text Generation • 0.3B • Updated Oct 9, 2025 • 38

kshitijthakkar/loggenix-moe-0.3B-A0.1B-e3-lr7e5-b16-4090-v7-sft-v3-dpo-v1-merged

Text Generation • 0.3B • Updated Oct 5, 2025 • 2 • 1

kshitijthakkar/loggenix-moe-0.3B-A0.1B-e3-lr7e5-b16-4090-v7-sft-v3-dpo-v1

Updated Oct 5, 2025