Text Generation on IFEval
74.49AccuracyLlama 3.1 8B Instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama 3.1 8B Instruct# Bits=BF16, Decoding Strategy=greedy2026.05 | 74.49 | |
| Llama 3.1 8B Instruct# Bits=BF16, Decoding strategy=greedy2026.05 | 74.49 | |
| FlexRound+LFQ+RILQ# Bits=W4, PTQ Method=FlexRound, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 72.46 | |
| FlexRound+LFQ# Bits=W4, PTQ Method=FlexRound, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 72.09 | |
| FlexRound+LFQ# Bits=W4, Decoding strategy=greedy2026.05 | 72.09 | |
| FlexRound+RILQ# Bits=W4, PTQ Method=FlexRound, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 71.9 | |
| FlexRound+RILQ# Bits=W4, Decoding strategy=greedy2026.05 | 71.9 | |
| OmniQuant+RILQ# Bits=W4, PTQ Method=OmniQuant, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 71.35 | |
| OmniQuant+LFQ# Bits=W4, PTQ Method=OmniQuant, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 71.35 | |
| OmniQuant+LFQ+RILQ# Bits=W4, PTQ Method=OmniQuant, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 71.35 | |
| OmniQuant+RILQ# Bits=W4, Decoding strategy=greedy2026.05 | 71.35 | |
| OmniQuant+LFQ# Bits=W4, Decoding strategy=greedy2026.05 | 71.35 | |
| OmniQuant# Bits=W4, PTQ Method=OmniQuant, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 70.61 | |
| OmniQuant# Bits=W4, Decoding strategy=greedy2026.05 | 70.61 | |
| FlexRound# Bits=W4, PTQ Method=FlexRound, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 70.24 | |
| FlexRound# Bits=W4, Decoding strategy=greedy2026.05 | 70.24 | |
| Block-AP+LFQ# Bits=W4, PTQ Method=Block-AP, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 68.76 | |
| Block-AP+LFQ# Bits=W4, Decoding strategy=greedy2026.05 | 68.76 | |
| Block-AP# Bits=W4, PTQ Method=Block-AP, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 68.58 | |
| Block-AP+RILQ# Bits=W4, PTQ Method=Block-AP, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 68.58 | |
| Block-AP+LFQ+RILQ# Bits=W4, PTQ Method=Block-AP, Decoding Strategy=greedy, Quantization Granularity=4-bit per-channel weight-only2026.05 | 68.58 | |
| Block-AP# Bits=W4, Decoding strategy=greedy2026.05 | 68.58 | |
| Block-AP+RILQ# Bits=W4, Decoding strategy=greedy2026.05 | 68.58 |