---

base_model: liuhaotian/llava-v1.5-7b inference: false library_name: transformers license: llama2 model_creator: liuhaotian model_name: Llava v1.5 7B quantized_by: Second State Inc.

---

---

Llava-v1.5-7B-GGUF

Original Model

liuhaotian/llava-v1.5-7b

Run with LlamaEdge

• LlamaEdge version: v0.16.2
• Prompt template

- Prompt type: vicuna-llava

- Prompt string

<systemprompt>\nUSER:<imageembeddings>\n<textual_prompt>\nASSISTANT:

• Context size: 4096
• Run as LlamaEdge service

wasmedge --dir .:. \
    --nn-preload default:GGML:AUTO:llava-v1.5-7b-Q5KM.gguf \
    llama-api-server.wasm \
    --prompt-template vicuna-llava \
    --ctx-size 4096 \
    --llava-mmproj llava-v1.5-7b-mmproj-model-f16.gguf \
    --model-name llava-v1.5

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
llava-v1.5-7b-Q2K.gguf	Q2K	2	2.53 GB	smallest, significant quality loss - not recommended for most purposes
llava-v1.5-7b-Q3KL.gguf	Q3K_L	3	3.6 GB	small, substantial quality loss
llava-v1.5-7b-Q3KM.gguf	Q3K_M	3	3.3 GB	very small, high quality loss
llava-v1.5-7b-Q3KS.gguf	Q3K_S	3	2.95 GB	very small, high quality loss
llava-v1.5-7b-Q40.gguf	Q40	4	3.83 GB	legacy; small, very high quality loss - prefer using Q3KM
llava-v1.5-7b-Q4KM.gguf	Q4K_M	4	4.08 GB	medium, balanced quality - recommended
llava-v1.5-7b-Q4KS.gguf	Q4K_S	4	3.86 GB	small, greater quality loss
llava-v1.5-7b-Q50.gguf	Q50	5	4.65 GB	legacy; medium, balanced quality - prefer using Q4KM
llava-v1.5-7b-Q5KM.gguf	Q5K_M	5	4.78 GB	large, very low quality loss - recommended
llava-v1.5-7b-Q5KS.gguf	Q5K_S	5	4.65 GB	large, low quality loss - recommended
llava-v1.5-7b-Q6K.gguf	Q6K	6	5.53 GB	very large, extremely low quality loss
llava-v1.5-7b-Q80.gguf	Q80	8	7.16 GB	very large, extremely low quality loss - not recommended
llava-v1.5-7b-mmproj-model-f16.gguf	f16	8	624 MB

Quantized with llama.cpp b2230