GGUF models by transitionGap

45 GGUF models quantized by transitionGap on Hugging Face. Sizes run from 362M to 9.24B parameters. The longest trained context is 128K tokens.

Most downloadedNewestLargestLongest contextSmallest
LL
COURTPROCEDDING Llama3.2 3B smallset
Llama · transitionGap
Params362M
Context2K
Q4_K_M0.3 GB
GGUF
QW
PROPERTYMUTATION IN Qwen2.5 5B smallset
Qwen · transitionGap
Params1.54B
Context32K
Q4_K_M1.0 GB
GGUF
LL
DOMICILE IN Llama3.2 2B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
QW
EDUCATIONLOAN IN Qwen2.5 5B smallset
Qwen · transitionGap
Params1.54B
Context32K
Q4_K_M1.0 GB
GGUF
MI
DOMICILE IN mistral 7B smallset
Mistral · transitionGap
Params7.24B
Context32K
Q4_K_M4.4 GB
GGUF
LL
MARRAIGELAW IN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
LL
RATIONCARD IN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
SM
PENSION IN SmolLM 360M smallset
SmolLM · transitionGap
Params362M
Context2K
Q4_K_M0.3 GB
GGUF
GE
BIRTHCERT IN gemma2 9B smallset
Gemma · transitionGap
Params9.24B
Context8K
Q4_K_M5.8 GB
GGUF
LL
HEALTHINSURANCE IN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
GE
DEATHCERT IN gemma2 9B smallset
Gemma · transitionGap
Params9.24B
Context8K
Q4_K_M5.8 GB
GGUF
LL
PENSION IN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
LL
BUILDINGPLAN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
SM
BUILDINGPLAN SmolLM.360M smallset
SmolLM · transitionGap
Params362M
Context2K
Q4_K_M0.3 GB
GGUF
MI
HEALTHINSURANCE IN mistral 7B smallset
Mistral · transitionGap
Params7.24B
Context32K
Q4_K_M4.4 GB
GGUF
LL
EDUCATIONLOAN IN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
SM
RATIONCARD IN SmolLM 360M smallset
SmolLM · transitionGap
Params362M
Context2K
Q4_K_M0.3 GB
GGUF
QW
DRIVINGLICENE IN Qwen2.5 0.5B smallset
Qwen · transitionGap
Params494M
Context32K
Q4_K_M0.4 GB
GGUF
GE
FARMSUBSIDY IN gemma2 9B smallset
Gemma · transitionGap
Params9.24B
Context8K
Q4_K_M5.8 GB
GGUF
LL
DOMICILE IN Llama3.1 8B smallset
Llama · transitionGap
Params8.03B
Context128K
Q4_K_M4.9 GB
Tool use
LL
RERA IN Llama3.1 8B smallset
Llama · transitionGap
Params8.03B
Context128K
Q4_K_M4.9 GB
Tool use
GE
PROPERTYREG IN gemma2 9B smallset
Gemma · transitionGap
Params9.24B
Context8K
Q4_K_M5.8 GB
GGUF
LL
EDUCATIONLOAN IN Llama3.1 8B smallset
Llama · transitionGap
Params8.03B
Context128K
Q4_K_M4.9 GB
Tool use
GE
SUBSIDY IN gemma2 9B smallset
Gemma · transitionGap
Params9.24B
Context8K
Q4_K_M5.8 GB
GGUF
QW
DEATHCERT IN Qwen2.5 0.5B smallset
Qwen · transitionGap
Params494M
Context32K
Q4_K_M0.4 GB
GGUF
LL
BUILDINGPLAN IN Llama3.1 8B smallset
Llama · transitionGap
Params8.03B
Context128K
Q4_K_M4.9 GB
Tool use
QW
RTIACT IN Qwen2.5 0.5B smallset
Qwen · transitionGap
Params494M
Context32K
Q4_K_M0.4 GB
GGUF
LL
PROPERTYMUTATION IN Llama3.2 3B smallset
Llama · transitionGap
Params3.21B
Context128K
Q4_K_M2.0 GB
Tool use
QW
Qwen2.5 0.5B Bnb 4bit
Qwen · transitionGap
Params494M
Context32K
Q4_K_M0.4 GB
GGUF
QW
SUBSIDY IN Qwen2.5 0.5B smallset
Qwen · transitionGap
Params494M
Context32K
Q4_K_M0.4 GB
GGUF

About transitionGap on Hugging Face

transitionGap has published 45 GGUF models that ggufmodels.com tracks, across families such as Llama, Qwen and Gemma. The most downloaded are COURTPROCEDDING Llama3.2 3B smallset, PROPERTYMUTATION IN Qwen2.5 5B smallset and DOMICILE IN Llama3.2 2B smallset.

The quants transitionGap publishes most often are BF16 (45), Q8_0 (44), Q5_K_M (44) and Q4_K_M (44). Their Q4_K_M files range from 0.3 to 5.8 GB. Parameter counts run from 362M to 9.24B, and the longest trained context window is 128K tokens.

Frequently asked questions

What GGUF models has transitionGap published?

45 models are listed here, mainly from the Llama, Qwen and Gemma families. The most downloaded is COURTPROCEDDING Llama3.2 3B smallset.

Which quants does transitionGap publish?

The most common are BF16 (45 models), Q8_0 (44 models), Q5_K_M (44 models) and Q4_K_M (44 models).

Keep browsing