metadata

base_model: tiiuae/Falcon3-1B-Instruct
license: other
license_name: falcon-llm-license
license_link: https://falconllm.tii.ae/falcon-terms-and-conditions.html
model_creator: tiiuae
model_name: Falcon3-1B-Instruct
quantized_by: Second State Inc.
library_name: transformers
language:
  - en
  - fr
  - es
  - pt
tags:
  - falcon3

Falcon3-1B-Instruct-GGUF

Original Model

tiiuae/Falcon3-1B-Instruct

Run with LlamaEdge

LlamaEdge version: v0.16.0 and above

Prompt template

Prompt type: falcon3

Prompt string

<|system|>
You are a helpful friendly assistant Falcon3 from TII, try to follow instructions as much as possible.
<|user|>
{user_message}
<|assistant|>

Context size: 8000

Run as LlamaEdge service

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Falcon3-1B-Instruct-Q5_K_M.gguf \
    llama-api-server.wasm \
    --prompt-template falcon3 \
    --ctx-size 8000 \
    --model-name Falcon3-1B-Instruct

Run as LlamaEdge command app

wasmedge --dir .:. --nn-preload default:GGML:AUTO:Falcon3-1B-Instruct-Q5_K_M.gguf \
  llama-chat.wasm \
  --prompt-template falcon3 \
  --ctx-size 8000

Quantized GGUF Models

Name	Quant method	Bits	Size	Use case
Falcon3-1B-Instruct-Q2_K.gguf	Q2_K	2	727 MB	smallest, significant quality loss - not recommended for most purposes
Falcon3-1B-Instruct-Q3_K_L.gguf	Q3_K_L	3	934 MB	small, substantial quality loss
Falcon3-1B-Instruct-Q3_K_M.gguf	Q3_K_M	3	885 MB	very small, high quality loss
Falcon3-1B-Instruct-Q3_K_S.gguf	Q3_K_S	3	827 MB	very small, high quality loss
Falcon3-1B-Instruct-Q4_0.gguf	Q4_0	4	1.01 GB	legacy; small, very high quality loss - prefer using Q3_K_M
Falcon3-1B-Instruct-Q4_K_M.gguf	Q4_K_M	4	1.06 GB	medium, balanced quality - recommended
Falcon3-1B-Instruct-Q4_K_S.gguf	Q4_K_S	4	1.02 GB	small, greater quality loss
Falcon3-1B-Instruct-Q5_0.gguf	Q5_0	5	1.19 GB	legacy; medium, balanced quality - prefer using Q4_K_M
Falcon3-1B-Instruct-Q5_K_M.gguf	Q5_K_M	5	1.21 GB	large, very low quality loss - recommended
Falcon3-1B-Instruct-Q5_K_S.gguf	Q5_K_S	5	1.19 GB	large, low quality loss - recommended
Falcon3-1B-Instruct-Q6_K.gguf	Q6_K	6	1.37 GB	very large, extremely low quality loss
Falcon3-1B-Instruct-Q8_0.gguf	Q8_0	8	1.78 GB	very large, extremely low quality loss - not recommended
Falcon3-1B-Instruct-f16.gguf	f16	16	3.34 GB

Quantized with llama.cpp b4381