Arcee Spark is a powerful 7B parameter language model that punches well above its weight class.

297 6 months ago

Readme

Arcee Spark

Arcee Spark

Arcee Spark is a powerful 7B parameter language model that punches well above its weight class. Initialized from Qwen2, this model underwent a sophisticated training process:

  1. Fine-tuned on 1.8 million samples
  2. Merged with Qwen2-7B-Instruct using Arcee’s mergekit
  3. Further refined using Direct Preference Optimization (DPO)

This meticulous process results in exceptional performance, with Arcee Spark achieving the highest score on MT-Bench for models of its size, outperforming even GPT-3.5 on many tasks.

Key Features

  • 7B parameters
  • State-of-the-art performance for its size
  • Initialized from Qwen2
  • Advanced training process including fine-tuning, merging, and DPO
  • Highest MT-Bench score in the 7B class
  • Outperforms GPT-3.5 on many tasks
  • Has a context length of 128k tokens, making it ideal for tasks requiring many conversation turns or working with large amounts of text.

Business Use Cases

Arcee Spark offers a compelling solution for businesses looking to leverage advanced AI capabilities without the hefty computational requirements of larger models. Its unique combination of small size and high performance makes it ideal for:

  1. Real-time applications: Deploy Arcee Spark for chatbots, customer service automation, and interactive systems where low latency is crucial.

  2. Edge computing: Run sophisticated AI tasks on edge devices or in resource-constrained environments.

  3. Cost-effective scaling: Implement advanced language AI across your organization without breaking the bank on infrastructure or API costs.

  4. Rapid prototyping: Quickly develop and iterate on AI-powered features and products.

  5. On-premise deployment: Easily host Arcee Spark on local infrastructure for enhanced data privacy and security.

Performance and Efficiency

Arcee Spark demonstrates that bigger isn’t always better in the world of language models. By leveraging advanced training techniques and architectural optimizations, it delivers:

  • Speed: Blazing fast inference times, often 10-100x faster than larger models.
  • Efficiency: Significantly lower computational requirements, reducing both costs and environmental impact.
  • Flexibility: Easy to fine-tune or adapt for specific domains or tasks.

Despite its compact size, Arcee Spark offers deep reasoning capabilities, making it suitable for a wide range of complex tasks including:

  • Advanced text generation
  • Detailed question answering
  • Nuanced sentiment analysis
  • Complex problem-solving
  • Code generation and analysis

Model Availability

  • Quants: Arcee Spark GGUF
  • FP32: For those looking to squeeze every bit of performance out of the model, we offer an FP32 version that scores slightly higher on all benchmarks.

Benchmarks and Evaluations

Benchmark Results
Additional Benchmark Results
Bigbenchhard Results

MT-Bench

########## First turn ##########
                     score
model       turn          
arcee-spark 1     8.777778
########## Second turn ##########
                     score
model       turn          
arcee-spark 2     8.164634
########## Average ##########
                score
model                
arcee-spark  8.469325

EQ-Bench

EQ-Bench: 71.4

TruthfulQA

Task Version Metric Value Stderr
truthfulqa_mc 1 mc1 0.4382 ± 0.0174
mc2 0.6150 ± 0.0155

AGI-Eval

Task Version Metric Value Stderr
agieval_aqua_rat 0 acc 0.3937 ± 0.0307
acc_norm 0.3937 ± 0.0307
agieval_logiqa_en 0 acc 0.4731 ± 0.0196
acc_norm 0.4854 ± 0.0196
agieval_lsat_ar 0 acc 0.2783 ± 0.0296
acc_norm 0.3000 ± 0.0303
agieval_lsat_lr 0 acc 0.5549 ± 0.0220
acc_norm 0.5451 ± 0.0221
agieval_lsat_rc 0 acc 0.6022 ± 0.0299
acc_norm 0.6208 ± 0.0296
agieval_sat_en 0 acc 0.8155 ± 0.0271
acc_norm 0.8107 ± 0.0274
agieval_sat_en_without_passage 0 acc 0.4806 ± 0.0349
acc_norm 0.4612 ± 0.0348
agieval_sat_math 0 acc 0.4909 ± 0.0338
acc_norm 0.4545 ± 0.0336

AGI-eval average: 51.11

GPT4All Evaluation

Task Version Metric Value Stderr
arc_challenge 0 acc 0.5333 ± 0.0146
acc_norm 0.5640 ± 0.0145
arc_easy 0 acc 0.8131 ± 0.0080
acc_norm 0.7668 ± 0.0087
boolq 1 acc 0.8471 ± 0.0063
hellaswag 0 acc 0.6206 ± 0.0048
acc_norm 0.8118 ± 0.0039
openbookqa 0 acc 0.3560 ± 0.0214
acc_norm 0.4600 ± 0.0223
piqa 0 acc 0.7987 ± 0.0094
acc_norm 0.8030 ± 0.0093
winogrande 0 acc 0.7690 ± 0.0130

Gpt4al Average: 69.37

Big Bench Hard

Task Version Metric Value Stderr
bigbench_causal_judgement 0 multiple_choice_grade 0.6053 ± 0.0356
bigbench_date_understanding 0 multiple_choice_grade 0.6450 ± 0.0249
bigbench_disambiguation_qa 0 multiple_choice_grade 0.5233 ± 0.0312
bigbench_geometric_shapes 0 multiple_choice_grade 0.2006 ± 0.0212
exact_str_match 0.0000 ± 0.0000
bigbench_logical_deduction_five_objects 0 multiple_choice_grade 0.2840 ± 0.0202
bigbench_logical_deduction_seven_objects 0 multiple_choice_grade 0.2429 ± 0.0162
bigbench_logical_deduction_three_objects 0 multiple_choice_grade 0.4367 ± 0.0287
bigbench_movie_recommendation 0 multiple_choice_grade 0.4720 ± 0.0223
bigbench_navigate 0 multiple_choice_grade 0.4980 ± 0.0158
bigbench_reasoning_about_colored_objects 0 multiple_choice_grade 0.5600 ± 0.0111
bigbench_ruin_names 0 multiple_choice_grade 0.4375 ± 0.0235
bigbench_salient_translation_error_detection 0 multiple_choice_grade 0.2685 ± 0.0140
bigbench_snarks 0 multiple_choice_grade 0.7348 ± 0.0329
bigbench_sports_understanding 0 multiple_choice_grade 0.6978 ± 0.0146
bigbench_temporal_sequences 0 multiple_choice_grade 0.4060 ± 0.0155
bigbench_tracking_shuffled_objects_five_objects 0 multiple_choice_grade 0.2072 ± 0.0115
bigbench_tracking_shuffled_objects_seven_objects 0 multiple_choice_grade 0.1406 ± 0.0083
bigbench_tracking_shuffled_objects_three_objects 0 multiple_choice_grade 0.4367 ± 0.0287

Big Bench average: 45.78

License

Arcee Spark is released under the Apache 2.0 license.

Acknowledgments

  • The Qwen2 team for their foundational work
  • The open-source AI community for their invaluable tools and datasets
  • Our dedicated team of researchers and engineers who push the boundaries of what’s possible with compact language models