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 ยฑ 0.0174
mc2 0.6150 ยฑ 0.0155

AGI-Eval

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

AGI-eval average: 51.11

GPT4All Evaluation

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

Gpt4al Average: 69.37

Big Bench Hard

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

Open LLM Leaderboard Evaluation Results

Detailed results can be found here

Metric Value
Avg. 25.54
IFEval (0-Shot) 56.21
BBH (3-Shot) 37.14
MATH Lvl 5 (4-Shot) 12.31
GPQA (0-shot) 7.61
MuSR (0-shot) 8.60
MMLU-PRO (5-shot) 31.36
Downloads last month
2,870
Safetensors
Model size
7.62B params
Tensor type
BF16
ยท
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for arcee-ai/Arcee-Spark

Finetunes
7 models
Merges
5 models
Quantizations
1 model

Space using arcee-ai/Arcee-Spark 1

Evaluation results