LocalFit
Submit a benchmark

Ternary-Bonsai-2-2B-gguf

prism-ml/Ternary-Bonsai-2-2B-gguf

ReleasedRelease date unknownSize2.2BRoleAssistantLicenseapache-2.0Contextunknown

Where it runs 8K context · ≥ 5 t/s floor

SetupBest planMemoryGenerate
8 GB GPU + 32 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
12 GB GPU + 32 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
16 GB GPU + 64 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
24 GB GPU + 64 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
32 GB GPU + 128 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
2× 24 GB GPUs + 128 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
96 GB workstation GPU + 256 GB RAMTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAFull GPUKV unknownUnknown
Mac, 24 GBTernary-Bonsai-2-2B-Q2_0llama.cpp MetalUnified memoryKV unknownUnknown
Mac, 48 GBTernary-Bonsai-2-2B-Q2_0llama.cpp MetalUnified memoryKV unknownUnknown
Mac, 128 GBTernary-Bonsai-2-2B-Q2_0llama.cpp MetalUnified memoryKV unknownUnknown
Mac, 512 GBTernary-Bonsai-2-2B-Q2_0llama.cpp MetalUnified memoryKV unknownUnknown
DGX Spark 128 GBTernary-Bonsai-2-2B-Q2_0llama.cpp CUDAUnified memoryKV unknownUnknown
Ryzen AI Max+ 395, 128 GBTernary-Bonsai-2-2B-Q2_0llama.cpp VulkanUnified memoryKV unknownUnknown
CPU only, 64 GB DDR5-5600 dual channelTernary-Bonsai-2-2B-PQ2_0llama.cpp CPUCPU onlyKV unknownUnknown

Same plan engine as the planner, with its default reserves and RAM. How plans are chosen.

Files exact Hugging Face bytes

ArtifactSizeFilesRevisionSource
Ternary-Bonsai-2-2B-PQ2_0 native GGUF option
prism-ml/Ternary-Bonsai-2-2B-gguf
0.48 GiB1683f7be262HF files
Ternary-Bonsai-2-2B-PTQ1_0 native GGUF option
prism-ml/Ternary-Bonsai-2-2B-gguf
0.40 GiB1683f7be262HF files
Ternary-Bonsai-2-2B-Q2_0 native GGUF option
prism-ml/Ternary-Bonsai-2-2B-gguf
0.51 GiB1683f7be262HF files

Not found in the representative quant search (not proven unavailable): GGUF FP16, GGUF Q8_0, GGUF Q5_K_M, GGUF Q4_K_M, MLX 8-bit, MLX 4-bit, EXL3, AWQ, GPTQ. Complete selected file sets; projectors, MTP/draft and lookup files are listed separately and budgeted only by plans that use them. File bytes are not allocated memory.

Quality

No Artificial Analysis evaluation mapped to this checkpoint.

Arena text rating

Unknown

Publisher-reported benchmarks

None extracted from the model card.

AA and Arena describe the hosted model in the stated mode, not each quantized file; scores are never inherited by base, pruned, abliterated or quantized variants and never averaged together.

How to run it

PrismML llama.cpp fork

source-documented, not executed source · prism-b10776-d00e252 orlater forMTP+quantizedKV onCUDA;branchprism

./bin/llama-server -m Ternary-Bonsai-2-2B-PQ2_0.gguf \
  --mmproj Ternary-Bonsai-2-2B-mmproj-Q8_0.gguf -ngl 99 -fa on -c 32768 --port 8080

Recipes are source examples (settings are the publisher’s or author’s, not measured optima) and were not executed by LocalFit. A downloadable file does not prove your backend build supports the architecture.

Reported speeds 0

No reported measurement yet. Submit one.