HF Model & Benchmark Matrix

24,078 variants · 14,984 scores · 120 benchmarks

falcon-mamba-7b-instruct-Q4_K_M-GGUF

tiiuae/falcon-mamba-7b-instruct-Q4_K_M-GGUF

kind
gguf
quant
gguf (4-bit)
family
falcon
modality
—
root model
tiiuae/falcon-mamba-7b-instruct
provider
tiiuae
license
other
language
en
context
2,048
library
—
vRAM (GB)
—
size on disk
—
downloads
3,285
likes
7
created
—
last modified
2024-10-10T06:10:27+00:00 (2024)
huggingface
https://huggingface.co/tiiuae/falcon-mamba-7b-instruct-Q4_K_M-GGUF
tags
ggufendataset:tiiuae/falcon-refinedwebarxiv:2312.00752arxiv:2410.05355base_model:tiiuae/falcon-mamba-7b-instructbase_model:quantized:tiiuae/falcon-mamba-7b-instructlicense:otherendpoints_compatibleregion:usconversational

Benchmark scores (11)

benchmarkcategoryscorecontextconfevidence
HellaSwagknowledge80.82low| `FalconMamba-7B`<sup>*</sup> | 62.03 | 80.82 | 62.11 | 73.64 | 53.42 | 52.54 | **64.09** |
WinoGrandeknowledge73.64low| `FalconMamba-7B`<sup>*</sup> | 62.03 | 80.82 | 62.11 | 73.64 | 53.42 | 52.54 | **64.09** |
MMLUknowledge62.11low| `FalconMamba-7B`<sup>*</sup> | 62.03 | 80.82 | 62.11 | 73.64 | 53.42 | 52.54 | **64.09** |
TruthfulQAsafety53.42low| `FalconMamba-7B`<sup>*</sup> | 62.03 | 80.82 | 62.11 | 73.64 | 53.42 | 52.54 | **64.09** |
GSM8Kmath52.54low| `FalconMamba-7B`<sup>*</sup> | 62.03 | 80.82 | 62.11 | 73.64 | 53.42 | 52.54 | **64.09** |
IFEvalinstruction-following33.36low| `FalconMamba-7B` | 33.36 | 19.88 | 3.63 | 8.05 | 10.86 | 14.47 | **15.04** |
BBHreasoning19.88low| `FalconMamba-7B` | 33.36 | 19.88 | 3.63 | 8.05 | 10.86 | 14.47 | **15.04** |
MMLU-Proknowledge14.47low| `FalconMamba-7B` | 33.36 | 19.88 | 3.63 | 8.05 | 10.86 | 14.47 | **15.04** |
MUSRreasoning10.86low| `FalconMamba-7B` | 33.36 | 19.88 | 3.63 | 8.05 | 10.86 | 14.47 | **15.04** |
GPQAreasoning8.05low| `FalconMamba-7B` | 33.36 | 19.88 | 3.63 | 8.05 | 10.86 | 14.47 | **15.04** |
MATHmath3.63low| `FalconMamba-7B` | 33.36 | 19.88 | 3.63 | 8.05 | 10.86 | 14.47 | **15.04** |

Discussions (0)

none scraped.

Hugging Face model & benchmark data scraped & matrix-ified. 24,078 variants · 14,984 scores · 120 benchmarks.