GGUF models by keisuke-miyako

302 GGUF models quantized by keisuke-miyako on Hugging Face. Sizes run from 23M to 14.7B parameters. The longest trained context is 1024K tokens.

Most downloadedNewestLargestLongest contextSmallest
GE
springsea 0.1
Gemma · keisuke-miyako
Params4.65B
Context128K
Q4_K_M3.4 GB
CodingReasoningTool use
MU
multilingual e5 small gguf q8 0
Multilingual · keisuke-miyako
Params118M
Context512
Smallest0.1 GB
GGUF
GR
Granite Embedding 107m Multilingual
Granite · keisuke-miyako
Params107M
Context512
Smallest0.1 GB
Embedding
NL
nllb-200-3.3B
Nllb · keisuke-miyako
Params4.92B
Context—
Q4_K_M2.8 GB
GGUF
GE
Gemma-4-E2B-It
Gemma · keisuke-miyako
Params4.65B
Context128K
Q4_K_M3.1 GB
ReasoningTool useVision
MU
multilingual e5 small gguf q4 k m
Multilingual · keisuke-miyako
Params118M
Context512
Q4_K_M0.1 GB
GGUF
BG
bge m3 gguf q8 0
BGE · keisuke-miyako
Params567M
Context8K
Smallest0.6 GB
Embedding
BG
bge reranker v2 m3 gguf q4 k m
BGE · keisuke-miyako
Params568M
Context8K
Q4_K_M0.4 GB
Embedding
NL
nllb-200-distilled-600M
Nllb · keisuke-miyako
Params1.4B
Context—
Q4_K_M0.8 GB
GGUF
MU
multilingual e5 small gguf f16
Multilingual · keisuke-miyako
Params118M
Context512
Smallest0.2 GB
GGUF
GT
Gte Multilingual Reranker Base
Gte · keisuke-miyako
Params306M
Context8K
Q4_K_M0.2 GB
Embedding
ET
ettin reranker v1
Ettin · keisuke-miyako
Params149M
Context8K
Smallest0.0 GB
Embedding
MI
Ministral 3 8B Instruct 2512 gguf q4 k m
Mistral · keisuke-miyako
Params8.49B
Context256K
Q4_K_M5.2 GB
Tool useVision
NO
nomic-embed-text-v1.5
Nomic · keisuke-miyako
Params137M
Context2K
Smallest0.1 GB
Embedding
NL
nllb-200-3.3B
Nllb · keisuke-miyako
Params4.92B
Context—
Smallest9.8 GB
GGUF
NL
nllb-200-distilled-1.3B
Nllb · keisuke-miyako
Params2.16B
Context—
Q4_K_M1.2 GB
GGUF
RU
ruri v3 310m gguf q8 0
Ruri · keisuke-miyako
Params315M
Context8K
Smallest0.3 GB
GGUF
XL
xLAM 2 3b Fc R
Xlam · keisuke-miyako
Params3.09B
Context32K
Q4_K_M1.9 GB
Tool use
GE
gemma 3n E2B gguf q4 k m
Gemma · keisuke-miyako
Params4.46B
Context32K
Q4_K_M2.8 GB
GGUF
NO
Modernbert Embed
Nomic · keisuke-miyako
Params149M
Context8K
Smallest0.2 GB
Embedding
GT
Gte Reranker Modernbert Base
Gte · keisuke-miyako
Params150M
Context8K
Smallest0.2 GB
Embedding
GT
gte modernbert base
Gte · keisuke-miyako
Params149M
Context8K
Q4_K_M0.1 GB
GGUF
HA
harrier oss v1 0.6b gguf q8 0
Harrier · keisuke-miyako
Params596M
Context32K
Smallest0.6 GB
ReasoningTool use
GE
Gemma-4-E4B-It
Gemma · keisuke-miyako
Params7.52B
Context128K
Q4_K_M5.0 GB
ReasoningTool useVision
ZE
Zembed 1
Zembed · keisuke-miyako
Params4.02B
Context40K
Q4_K_M2.5 GB
EmbeddingReasoningTool use
MI
Ministral 3 8B Reasoning 2512 gguf q4 k m
Mistral · keisuke-miyako
Params8.49B
Context256K
Q4_K_M5.2 GB
ReasoningTool useVision
LL
Llama 3 ELYZA JP 8B
Llama · keisuke-miyako
Params8.03B
Context8K
Q4_K_M4.9 GB
GGUF
GE
gemma 3n E2B it gguf q4 k m
Gemma · keisuke-miyako
Params4.46B
Context32K
Q4_K_M2.8 GB
GGUF
NL
nllb-200-distilled-1.3B
Nllb · keisuke-miyako
Params2.16B
Context—
Smallest4.3 GB
GGUF
TR
translategemma 4b it gguf q4 k m
Translategemma · keisuke-miyako
Params3.88B
Context128K
Q4_K_M2.5 GB
Vision

About keisuke-miyako on Hugging Face

keisuke-miyako has published 302 GGUF models that ggufmodels.com tracks, across families such as BGE, Qwen and Granite. The most downloaded are springsea 0.1, multilingual e5 small gguf q8 0 and Granite Embedding 107m Multilingual.

The quants keisuke-miyako publishes most often are Q8_0 (144), Q4_K_M (140), F16 (20) and TQ2_0 (3). Their Q4_K_M files range from 0.0 to 9.1 GB. Parameter counts run from 23M to 14.7B, and the longest trained context window is 1024K tokens.

Frequently asked questions

What GGUF models has keisuke-miyako published?

302 models are listed here, mainly from the BGE, Qwen and Granite families. The most downloaded is springsea 0.1.

Which quants does keisuke-miyako publish?

The most common are Q8_0 (144 models), Q4_K_M (140 models), F16 (20 models) and TQ2_0 (3 models).

Keep browsing