Text Reconstruction on TokBench (val)
16.44T-ACCsInsightTok
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| InsightTok#Tokens=1,024, Codebook=16,384, BPP=0.0547, Resolution=512 x 5122026.05 | 16.44 | 53.05 | 42.31 | 71.4 | |
| Emu3.5-IBQ#Tokens=1,024, Codebook=131,072, BPP=0.0664, Resolution=512 x 5122026.05 | 12.99 | 41.52 | 39.92 | 65.39 | |
| VAR#Tokens=2,240, Codebook=4,096, BPP=0.1025, Resolution=512 x 5122026.05 | 3.71 | 29.31 | 18.01 | 50 | |
| O-MAGVIT2-262k#Tokens=1,024, Codebook=262,144, BPP=0.0703, Resolution=512 x 5122026.05 | 3.02 | 27.33 | 16.87 | 47.28 | |
| IBQ-16k#Tokens=1,024, Codebook=16,384, BPP=0.0547, Resolution=512 x 5122026.05 | 2.28 | 24.16 | 14.75 | 43.66 | |
| O-MAGVIT2-16k#Tokens=1,024, Codebook=16,384, BPP=0.0547, Resolution=512 x 5122026.05 | 1.62 | 20.62 | 12.71 | 39.96 | |
| OCR-VQGAN#Tokens=1,024, Codebook=16,384, BPP=0.0547, Resolution=512 x 5122026.05 | 1.13 | 12.76 | 10.24 | 28.58 | |
| LlamaGen-T2I#Tokens=1,024, Codebook=16,384, BPP=0.0547, Resolution=512 x 5122026.05 | 0.67 | 15.01 | 7.76 | 30.44 | |
| Chameleon#Tokens=1,024, Codebook=8,192, BPP=0.0508, Resolution=512 x 5122026.05 | 0.6 | 11.55 | 7.63 | 26.8 | |
| Cosmos-DI#Tokens=1,024, Codebook=64,000, BPP=0.0624, Resolution=512 x 5122026.05 | 0.52 | 12.26 | 7.61 | 27.95 | |
| VQGAN#Tokens=1,024, Codebook=16,384, BPP=0.0547, Resolution=512 x 5122026.05 | 0.15 | 6.12 | 5.2 | 17.32 |