This is a 12B parameter model trained on 10T tokens.
It's also editorialized which is against HN.
Title is: "NVFP4 Trains with Precision of 16-Bit and Speed and Efficiency of 4-Bit"
patrickhogan1 · 1h ago
12B model
opcode84 · 4h ago
For narrow-precision formats to be practical in large-scale pretraining, they must ensure both model accuracy and stable convergence. To assess the viability of 4-bit precision in large-scale model training, experiments were conducted with FP8 and NVFP4 on a 12-billion parameter model based on a combined Mamba-Transformer architecture (12B Hybrid Mamba-Transformer model)—similar to NVIDIA Nemotron Nano 2. This model was trained on a massive dataset of 10 trillion tokens using a phased data-blending approach, switching to a different dataset mix in the second phase of training at 70%, and in the third phase of training at 90% during pretraining.
A version of the 12B Hybrid Mamba-Transformer model was initially trained with 8-bit precision—FP8, which has been shown in previous studies to closely match 16-bit precision, and hence served as our baseline for comparison. We then successfully trained this same 12B model from scratch using NVFP4, demonstrating that this new low-precision format can support full pretraining at trillion-token scale. The NVFP4 run exhibited stable convergence without the training instabilities or divergence issues that typically plague ultra-low precision training.
Figure 3 below shows that NVFP4’s validation loss curve closely matches the loss curves from the higher-precision baseline (i.e., FP8) throughout the entire duration of training. The quantization techniques outlined above ensure that even with aggressive bit-width reduction, the 4-bit pretraining dynamics closely resemble those of higher-precision runs.
This is a 12B parameter model trained on 10T tokens.
It's also editorialized which is against HN.
Title is: "NVFP4 Trains with Precision of 16-Bit and Speed and Efficiency of 4-Bit"
A version of the 12B Hybrid Mamba-Transformer model was initially trained with 8-bit precision—FP8, which has been shown in previous studies to closely match 16-bit precision, and hence served as our baseline for comparison. We then successfully trained this same 12B model from scratch using NVFP4, demonstrating that this new low-precision format can support full pretraining at trillion-token scale. The NVFP4 run exhibited stable convergence without the training instabilities or divergence issues that typically plague ultra-low precision training.
Figure 3 below shows that NVFP4’s validation loss curve closely matches the loss curves from the higher-precision baseline (i.e., FP8) throughout the entire duration of training. The quantization techniques outlined above ensure that even with aggressive bit-width reduction, the 4-bit pretraining dynamics closely resemble those of higher-precision runs.