| | ---
|
| | license: apache-2.0
|
| | pipeline_tag: text-generation
|
| | base_model: internlm/internlm3-8b-instruct
|
| | model_creator: InternLM
|
| | model_name: internlm3-8b-instruct
|
| | quantized_by: Second State Inc.
|
| | ---
|
| |
|
| | <!-- header start -->
|
| | <!-- 200823 -->
|
| | <div style="width: auto; margin-left: auto; margin-right: auto">
|
| | <img src="https://github.com/LlamaEdge/LlamaEdge/raw/dev/assets/logo.svg" style="width: 100%; min-width: 400px; display: block; margin: auto;">
|
| | </div>
|
| | <hr style="margin-top: 1.0em; margin-bottom: 1.0em;">
|
| | <!-- header end -->
|
| |
|
| | # internlm3-8b-instruct-GGUF
|
| |
|
| | ## Original Model
|
| |
|
| | [internlm/internlm3-8b-instruct](https://huggingface.co/internlm/internlm3-8b-instruct)
|
| |
|
| | ## Run with LlamaEdge
|
| |
|
| | - LlamaEdge version: [v0.16.1](https://github.com/LlamaEdge/LlamaEdge/releases/tag/0.16.1) and above
|
| |
|
| | - Prompt template
|
| |
|
| | - Prompt type: `chatml`
|
| |
|
| | - Prompt string
|
| |
|
| | ```text
|
| | <|im_start|>system
|
| | {system_message}<|im_end|>
|
| | <|im_start|>user
|
| | {prompt}<|im_end|>
|
| | <|im_start|>assistant
|
| | ```
|
| |
|
| | - Context size: `128000`
|
| |
|
| | - Run as LlamaEdge service
|
| |
|
| | - Chat
|
| |
|
| | ```bash
|
| | wasmedge --dir .:. --nn-preload default:GGML:AUTO:internlm3-8b-instruct-Q5_K_M.gguf \
|
| | llama-api-server.wasm \
|
| | --prompt-template chatml \
|
| | --ctx-size 128000 \
|
| | --model-name internlm3-8b-instruct
|
| | ```
|
| |
|
| | - Tool use
|
| |
|
| | ```bash
|
| | wasmedge --dir .:. --nn-preload default:GGML:AUTO:internlm3-8b-instruct-Q5_K_M.gguf \
|
| | llama-api-server.wasm \
|
| | --prompt-template internlm-2-tool \
|
| | --ctx-size 128000 \
|
| | --model-name internlm3-8b-instruct
|
| | ```
|
| |
|
| | - Run as LlamaEdge command app
|
| |
|
| | ```bash
|
| | wasmedge --dir .:. \
|
| | --nn-preload default:GGML:AUTO:internlm3-8b-instruct-Q5_K_M.gguf \
|
| | llama-chat.wasm \
|
| | --prompt-template chatml \
|
| | --ctx-size 32000
|
| | ```
|
| |
|
| | ## Quantized GGUF Models
|
| |
|
| | | Name | Quant method | Bits | Size | Use case |
|
| | | ---- | ---- | ---- | ---- | ----- |
|
| | | [internlm3-8b-instruct-Q2_K.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q2_K.gguf) | Q2_K | 2 | 3.45 GB| smallest, significant quality loss - not recommended for most purposes |
|
| | | [internlm3-8b-instruct-Q3_K_L.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q3_K_L.gguf) | Q3_K_L | 3 | 4.73 GB| small, substantial quality loss |
|
| | | [internlm3-8b-instruct-Q3_K_M.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q3_K_M.gguf) | Q3_K_M | 3 | 4.39 GB| very small, high quality loss |
|
| | | [internlm3-8b-instruct-Q3_K_S.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q3_K_S.gguf) | Q3_K_S | 3 | 3.99 GB| very small, high quality loss |
|
| | | [internlm3-8b-instruct-Q4_0.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q4_0.gguf) | Q4_0 | 4 | 5.09 GB| legacy; small, very high quality loss - prefer using Q3_K_M |
|
| | | [internlm3-8b-instruct-Q4_K_M.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q4_K_M.gguf) | Q4_K_M | 4 | 5.36 GB| medium, balanced quality - recommended |
|
| | | [internlm3-8b-instruct-Q4_K_S.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q4_K_S.gguf) | Q4_K_S | 4 | 5.12 GB| small, greater quality loss |
|
| | | [internlm3-8b-instruct-Q5_0.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q5_0.gguf) | Q5_0 | 5 | 6.13 GB| legacy; medium, balanced quality - prefer using Q4_K_M |
|
| | | [internlm3-8b-instruct-Q5_K_M.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q5_K_M.gguf) | Q5_K_M | 5 | 6.26 GB| large, very low quality loss - recommended |
|
| | | [internlm3-8b-instruct-Q5_K_S.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q5_K_S.gguf) | Q5_K_S | 5 | 6.13 GB| large, low quality loss - recommended |
|
| | | [internlm3-8b-instruct-Q6_K.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q6_K.gguf) | Q6_K | 6 | 7.23 GB| very large, extremely low quality loss |
|
| | | [internlm3-8b-instruct-Q8_0.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-Q8_0.gguf) | Q8_0 | 8 | 9.36 GB| very large, extremely low quality loss - not recommended |
|
| | | [internlm3-8b-instruct-f16.gguf](https://huggingface.co/second-state/internlm3-8b-instruct-GGUF/blob/main/internlm3-8b-instruct-f16.gguf) | f16 | 16 | 17.6 GB| |
|
| |
|
| | *Quantized with llama.cpp b4497* |
