Qwen2.5-7B-Instruct
A 7-billion-parameter instruction-tuned language model from Alibaba's Qwen2.5 series, supporting multilingual reasoning and conversational tasks. It is part of the Qwen2.5 family of models released by Alibaba.
Timeline (last 90 days)
LocalAgentBrowser is an open-source project that enables a 7B-parameter model to run a browser agent on a consumer-grade CPU and autonomously complete research and writing tasks.
The LocalAgentBrowser agent runs the Qwen2.5-7B-Instruct model on an Intel i5-9500 CPU and can complete the full workflow from searching and reading to writing an answer.
On an Intel i5-9500, the Qwen2.5-7B-Instruct model reads at about 8 tokens per second, and reading the prompt accounts for about 90% of the time of each model call.
Requiring the 7B model to show its work process (breaking the goal into parts and labeling each as covered or not covered, and reporting the number of sources) is more effective than directly asking whether the goal is met.
Experiments were conducted on Qwen2.5-7B-Instruct using Math-Shepherd as the PRM, tested on the MATH500 dataset (n=500, three random seeds).
The study's limitations include being restricted to a single model, Qwen2.5-7B-Instruct, and a single dataset, LongBench-v2, with a small amount of training data (only 168 questions).
The research used all 503 multiple-choice questions from the LongBench-v2 dataset as the primary diagnostic set, with Qwen2.5-7B-Instruct as the answer model.
All Facts (7)
LocalAgentBrowser is an open-source project that enables a 7B-parameter model to run a browser agent on a consumer-grade CPU and autonomously complete research and writing tasks.
50%UnverifiedThe LocalAgentBrowser agent runs the Qwen2.5-7B-Instruct model on an Intel i5-9500 CPU and can complete the full workflow from searching and reading to writing an answer.
50%UnverifiedOn an Intel i5-9500, the Qwen2.5-7B-Instruct model reads at about 8 tokens per second, and reading the prompt accounts for about 90% of the time of each model call.
50%UnverifiedRequiring the 7B model to show its work process (breaking the goal into parts and labeling each as covered or not covered, and reporting the number of sources) is more effective than directly asking whether the goal is met.
50%UnverifiedExperiments were conducted on Qwen2.5-7B-Instruct using Math-Shepherd as the PRM, tested on the MATH500 dataset (n=500, three random seeds).
50%UnverifiedThe study's limitations include being restricted to a single model, Qwen2.5-7B-Instruct, and a single dataset, LongBench-v2, with a small amount of training data (only 168 questions).
50%UnverifiedThe research used all 503 multiple-choice questions from the LongBench-v2 dataset as the primary diagnostic set, with Qwen2.5-7B-Instruct as the answer model.
50%