Falcon3-7B-Instruct: Optimized for Qualcomm Devices

Falcon3 family of Open Foundation Models is a set of pretrained and instruct LLMs ranging from 1B to 10B.

This is based on the implementation of Falcon3-7B-Instruct found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up to run these models on a hosted Qualcomm® device.

Deploying Falcon3-7B-Instruct on-device

Follow the GenieX quickstart to install GenieX and deploy the model on a target device.

You'll need to export the model artifact using the steps below, then follow Run a Local Model with GenieX.

See the LLM-on-Genie tutorial to run with the Genie runtime. Note: Genie support will be deprecated soon.

Getting Started

Due to licensing restrictions, we cannot distribute pre-exported model assets for this model. Use the Qualcomm® AI Hub Models Python library to compile and export the model with your own:

• Custom weights (e.g., fine-tuned checkpoints)
• Custom input shapes
• Target device and runtime configurations

See our repository for Falcon3-7B-Instruct on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.text_generation

Model Stats:

• Supported languages: English, French, Spanish, Portuguese.
• TTFT: Time To First Token is the time it takes to generate the first response token. This is expressed as a range because it varies based on the length of the prompt. The lower bound is for a short prompt (up to 128 tokens, i.e., one iteration of the prompt processor) and the upper bound is for a prompt using the full context length (4096 tokens).
• Response Rate: Rate of response generation after the first response token.

Performance Summary

Model	Runtime	Precision	Chipset	Context Length	Response Rate (tokens per second)	Time To First Token (range, seconds)
Falcon3-7B-Instruct	GENIE	w4a16	Snapdragon® 8 Elite Gen 5 Mobile	4096	17.845033645629883	0.098628 - 3.156096
Falcon3-7B-Instruct	GENIE	w4a16	Snapdragon® 8 Elite Mobile	4096	15.377281188964844	0.121521 - 3.888672
Falcon3-7B-Instruct	GENIE	w4a16	Snapdragon® X2 Elite	4096	24.13884735107422	0.125103 - 4.003296
Falcon3-7B-Instruct	GENIE	w4a16	Snapdragon® X Elite	4096	5.356301307678223	0.194595 - 6.22704
Falcon3-7B-Instruct	GENIE	w4a16	Qualcomm® QCS9075	4096	10.775397300720215	0.162771 - 5.208672
Falcon3-7B-Instruct	GENIE	w4a16	Qualcomm® QCS8750	4096	15.504198265075683	0.1292663 - 4.1365216
Falcon3-7B-Instruct	GENIE	w4a16	Qualcomm® QCS7181	4096	8.573315954208374	0.21308370000000001 - 6.8186784000000005
Falcon3-7B-Instruct	GENIEX_QAIRT	w4a16	Snapdragon® 8 Elite Gen 5 Mobile	4096	17.911517	0.1257 - 4.0224
Falcon3-7B-Instruct	GENIEX_QAIRT	w4a16	Snapdragon® 8 Elite Mobile	4096	15.689129	0.1512 - 4.8384
Falcon3-7B-Instruct	GENIEX_QAIRT	w4a16	Snapdragon® X2 Elite	4096	24.471058	0.109 - 3.488
Falcon3-7B-Instruct	GENIEX_QAIRT	w4a16	Snapdragon® X Elite	4096	11.783296	0.1978 - 6.3296
Falcon3-7B-Instruct	GENIEX_QAIRT	w4a16	Qualcomm® QCS9075	4096	10.652838	0.197 - 6.304

License

• The license for the original implementation of Falcon3-7B-Instruct can be found here.

References

• Source Model Implementation

Community

• Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
• For questions or feedback please reach out to us.

Usage and Limitations

This model may not be used for or in connection with any of the following applications:

• Accessing essential private and public services and benefits;
• Administration of justice and democratic processes;
• Assessing or recognizing the emotional state of a person;
• Biometric and biometrics-based systems, including categorization of persons based on sensitive characteristics;
• Education and vocational training;
• Employment and workers management;
• Exploitation of the vulnerabilities of persons resulting in harmful behavior;
• General purpose social scoring;
• Law enforcement;
• Management and operation of critical infrastructure;
• Migration, asylum and border control management;
• Predictive policing;
• Real-time remote biometric identification in public spaces;
• Recommender systems of social media platforms;
• Scraping of facial images (from the internet or otherwise); and/or
• Subliminal manipulation