Can you run telecomadm1145/mamba2_ja2zh_zh2ja_trans on RTX 3090 24GB?

Yes — telecomadm1145/mamba2_ja2zh_zh2ja_trans fits on RTX 3090 24GB: estimated 1.87 GB (weights + KV cache) against 24 GB.

Hugging Bay is intentionally crawlable for search engines, answer engines, user-requested AI fetches, frontier AI assistants, and open-source agents. Public pages describe verified open-source AI artifact metadata, provenance, licenses, community trust signals, and selective hosted mirrors with hashes. Hugging Bay does not bypass gated or restricted upstream access controls.

Machine-readable entrypoints: robots.txt, AI search guidance, well-known AI search JSON, AI citation packs, well-known AI citation JSON, llms.txt, llms-full.txt, AI crawler policy, AI bot allowlist, answer-engine manifest, agent manifest, OpenAPI, and indexing status.

Yes — telecomadm1145/mamba2_ja2zh_zh2ja_trans fits on RTX 3090 24GB: estimated 1.87 GB (weights + KV cache) against 24 GB.

Verdict across common GPUs

GPUVerdictVRAM
NVIDIA T4 16GB (free Colab)Yes16 GB
NVIDIA L4 24GBYes24 GB
RTX 3060 12GBYes12 GB
RTX 4080 16GBYes16 GB
RTX 3090 24GBYes24 GB
RTX 4090 24GBYes24 GB
A100 40GBYes40 GB
A100 80GBYes80 GB
H100 80GBYes80 GB
Apple M-series 16GB (unified)Yes16 GB
Apple M-series 32GB (unified)Yes32 GB
Apple M-series 64GB (unified)Yes64 GB
Apple M-series 128GB (unified)Yes128 GB
CPU / 32GB system RAMYes32 GB
CPU / 64GB system RAMYes64 GB

Quantizations of telecomadm1145/mamba2_ja2zh_zh2ja_trans

RepoQuantApprox weightsHosting
mradermacher/Arxh-2.29-12B-it-GGUFq3_k_m5.3 GBexternal
mradermacher/Nous-Hermes-2-Yi-34B-i1-GGUFunknown68 GBexternal
mradermacher/Kagura-Llama-3.1-8B-Instruct-GGUFq3_k_m3.5 GBexternal
Menlo/Lucy-ggufq3_k_m—external
mradermacher/Qwen3.5-27B-heretic-i1-GGUFunknown54 GBexternal
mradermacher/KansenSakura-Erosion-RP-12b-i1-GGUFunknown24 GBexternal
mradermacher/astrollama-3-8b-base_aic-GGUFq3_k_s3.2 GBexternal
unsloth/GLM-4.7-Flash-GGUFq4_k_s—hosted ✓

Next steps

  • Open telecomadm1145/mamba2_ja2zh_zh2ja_trans on Hugging Bay — trust verdict, hosted files, reviews.
  • Machine-readable deployment plan (JSON)
  • Raw fit report (JSON)

Methodology: weights + KV-cache estimate from parameter count, quant precision, and declared context. Not a benchmark; runtimes differ. Data recomputed hourly.