HA

HarmonicHarlequin v5 20B i1

HarmonicharlequinGGUF by mradermacherUpdated 2024-10-02
Download on Hugging Face
Parameters33.3B
Context4K
Layers104
Architecturellama
Smallest quant7.3 GB
Quick answer

HarmonicHarlequin v5 20B i1 needs about 38.0 GB of memory at Q4_K_M (a 20.1 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 21 quants, from 7.3 GB (IQ1_S) to 27.3 GB (Q6_K).

HarmonicHarlequin v5 20B i1 – a 33.3B llama‑style model

HarmonicHarlequin v5 20B i1 is a 33.3B parameter llama architecture model with a 4096 token context window. It is a merged model that includes Undi95/PsyMedRP-v1-20B. The model is available in multiple GGUF quantizations.

Good at

  • Large context handling
  • Coherent text generation
  • Merge‑based versatility

Limits

  • No vision support
  • License not specified
  • Requires 8k context for memory estimates

Best for

  • Chat applications
  • Content drafting
  • Research summarization

Who it’s for. Developers and researchers who need a large llama‑style model for text generation and want to run it locally with GGUF support. The model is suitable for those who can manage the memory requirements of the larger quantizations.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S7.3 GB 16.5 GB
Too big Get file →
IQ1_M7.9 GB 17.1 GB
Too big Get file →
IQ2_XXS8.9 GB 18.2 GB
Too big Get file →
IQ2_XS9.8 GB 19.1 GB
Too big Get file →
IQ2_S10.6 GB 19.8 GB
Too big Get file →
IQ2_M11.5 GB 20.7 GB
Too big Get file →
Q2_K12.3 GB 21.5 GB
Too big Get file →
IQ3_XXS12.6 GB 21.8 GB
Too big Get file →
IQ3_XS13.6 GB 22.8 GB
Too big Get file →
IQ3_S14.4 GB 23.6 GB
Too big Get file →
Q3_K_S14.4 GB 23.6 GB
Too big Get file →
IQ3_M15.2 GB 24.5 GB
Too big Get file →
Q3_K_M16.1 GB 25.4 GB
Too big Get file →
Q3_K_L17.7 GB 26.9 GB
Too big Get file →
IQ4_XS17.8 GB 27.0 GB
Too big Get file →
Q4_018.8 GB 28.1 GB
Too big Get file →
Q4_K_S18.9 GB 28.1 GB
Too big Get file →
Q4_K_M20.1 GB 29.3 GB
Too big Get file →
Q5_K_S22.9 GB 32.2 GB
Too big Get file →
Q5_K_M23.6 GB 32.8 GB
Too big Get file →
Q6_K27.3 GB 36.6 GB
Too big Get file →

For 8 GB VRAM, use IQ1_S (7.27 GB file, 25.2 GB memory at 8k context). For 16 GB VRAM, IQ2_M (11.45 GB file, 29.4 GB memory) or Q3_K_M (16.15 GB file, 34.1 GB memory). For 24 GB VRAM, Q4_K_M (20.09 GB file, 38 GB memory) or Q5_K_M (23.6 GB file, 41.5 GB memory).

What GPU do you need for HarmonicHarlequin v5 20B i1?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: no quant fits at an 8K context.
  • 6 GB VRAM: no quant fits at an 8K context.
  • 8 GB VRAM: no quant fits at an 8K context.
  • 12 GB VRAM: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: no quant fits at an 8K context.
  • 32 GB VRAM: the best quant that fits is IQ3_XS (13.6 GB file, about 31.6 GB in use).
  • 48 GB VRAM: the best quant that fits is Q6_K (27.3 GB file, about 45.3 GB in use).

Which HarmonicHarlequin v5 20B i1 quant should you choose?

Q4_K_M is the usual starting point for HarmonicHarlequin v5 20B i1: it balances file size and quality well. Here it is a 20.1 GB file.

If memory is tight, the smallest file is IQ1_S at 7.3 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters33.3B
Layers104
Trained context4,096 tokens
KV cache (fp16)2080 KB per token

Source

Quantized bymradermacher
Downloads (HF)4,605
Last updated2024-10-02

Frequently asked questions about HarmonicHarlequin v5 20B i1

How much VRAM does HarmonicHarlequin v5 20B i1 need?

About 38.0 GB at Q4_K_M with an 8K context window. That is the 20.1 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run HarmonicHarlequin v5 20B i1 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 25.2 GB, so you would need a larger GPU or CPU offloading.

What is the file size of HarmonicHarlequin v5 20B i1 GGUF?

It depends on the quant: IQ1_S is 7.3 GB, IQ1_M is 7.9 GB, IQ2_XXS is 8.9 GB, IQ2_XS is 9.8 GB, IQ2_S is 10.6 GB, IQ2_M is 11.5 GB and 15 more.

What context length does HarmonicHarlequin v5 20B i1 support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 17.4 GB at 8K and 8.7 GB at 4K.

Where can I download HarmonicHarlequin v5 20B i1 GGUF?

The files are on Hugging Face in the mradermacher/HarmonicHarlequin_v5-20B-i1-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run HarmonicHarlequin v5 20B i1 on 8 GB VRAM?

Use the IQ1_S quantization. It requires 25.2 GB memory at 8k context, so it will need CPU fallback or a larger GPU.

Does this model support vision?

No, the model has no vision capabilities.

What context length does the model support?

The model supports up to 4096 tokens of context.

Other GGUFs of HarmonicHarlequin_v5-20B

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Harmonicharlequin models

All Harmonicharlequin →

Other medium models

All medium models →