Text Generation on GSM8K
187.43Throughput (tokens/s)EVICT
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| EVICTTemperature=0, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 187.43 | — | — | — | — | — | — | |
| EVICTTemperature=1, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 184.44 | — | — | — | — | — | — | |
| EAGLE-3Temperature=0, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 146.38 | — | — | — | — | — | — | |
| EAGLE-3Temperature=1, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 143.43 | — | — | — | — | — | — | |
| DDDTemperature=0, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 137.71 | — | — | — | — | — | — | |
| DDDTemperature=1, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 135.65 | — | — | — | — | — | — | |
| VanillaTemperature=0, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 79.61 | — | — | — | — | — | — | |
| VanillaTemperature=1, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 79.01 | — | — | — | — | — | — | |
| LookaheadTemperature=0, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 65.5 | — | — | — | — | — | — | |
| LookaheadTemperature=1, Backbone=Qwen3-235B-A22B, Inference Engine=SGLang, Hardware=NVIDIA A100 SXM4 80GB GPUs, Batch Size=1, Tensor Parallelism=TP=82026.05 | 64.6 | — | — | — | — | — | — | |
| ARBackbone=Qwen3-4B, Temperature=1.02026.06 | — | — | 31.05 | — | — | — | — | |
| ARBackbone=Qwen3-8B, Temperature=1.02026.06 | — | — | 30.84 | — | — | — | — | |
| CADDTREEBackbone=Qwen3-4B, Temperature=1.02026.06 | — | — | 5.56 | 8.67 | — | 250 | — | |
| CADDTREEBackbone=Qwen3-8B, Temperature=1.02026.06 | — | — | 5.62 | 8.44 | — | 184 | — | |
| DDTree-oracleBackbone=Qwen3-4B, Temperature=1.02026.06 | — | — | 5.5 | 8.65 | 256 | — | — | |
| DDTree-oracleBackbone=Qwen3-8B, Temperature=1.02026.06 | — | — | 5.67 | 8.27 | 128 | — | — | |
| DFlashBackbone=Qwen3-4B, Temperature=1.02026.06 | — | — | 7.64 | 6 | — | — | — | |
| DFlashBackbone=Qwen3-8B, Temperature=1.02026.06 | — | — | 7.69 | 5.92 | — | — | — |