Views
No views yet
gemma-31<bos><start_of_turn>user
2{user_message}<end_of_turn>
3<start_of_turn>model
4{model_message}<end_of_turn>model320001wasmedge --dir .:. --nn-preload default:GGML:AUTO:gemma-3n-E4B-it-Q5_K_M.gguf \
2 llama-api-server.wasm \
3 --prompt-template gemma-3 \
4 --ctx-size 32000 \
5 --model-name gemma-3n-E4B-it1wasmedge --dir .:. \
2 --nn-preload default:GGML:AUTO:gemma-3n-E4B-it-Q5_K_M.gguf \
3 llama-chat.wasm \
4 --prompt-template gemma-3 \
5 --ctx-size 32000| Name | Quant method | Bits | Size | Use case |
|---|---|---|---|---|
| gemma-3n-E4B-it-Q2_K.gguf | Q2_K | 2 | 2.76 GB | smallest, significant quality loss - not recommended for most purposes |
| gemma-3n-E4B-it-Q3_K_L.gguf | Q3_K_L | 3 | 3.60 GB | small, substantial quality loss |
| gemma-3n-E4B-it-Q3_K_M.gguf | Q3_K_M | 3 | 3.44 GB | very small, high quality loss |
| gemma-3n-E4B-it-Q3_K_S.gguf | Q3_K_S | 3 | 3.25 GB | very small, high quality loss |
| gemma-3n-E4B-it-Q4_0.gguf | Q4_0 | 4 | 4.09 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| gemma-3n-E4B-it-Q4_K_M.gguf | Q4_K_M | 4 | 4.24 GB | medium, balanced quality - recommended |
| gemma-3n-E4B-it-Q4_K_S.gguf | Q4_K_S | 4 | 4.10 GB | small, greater quality loss |
| gemma-3n-E4B-it-Q5_0.gguf | Q5_0 | 5 | 4.87 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| gemma-3n-E4B-it-Q5_K_M.gguf | Q5_K_M | 5 | 4.95 GB | large, very low quality loss - recommended |
| gemma-3n-E4B-it-Q5_K_S.gguf | Q5_K_S | 5 | 4.87 GB | large, low quality loss - recommended |
| gemma-3n-E4B-it-Q6_K.gguf | Q6_K | 6 | 5.70 GB | very large, extremely low quality loss |
| gemma-3n-E4B-it-Q8_0.gguf | Q8_0 | 8 | 7.35 GB | very large, extremely low quality loss - not recommended |
| gemma-3n-E4B-it-f16.gguf | f16 | 16 | 13.7 GB |