Files
llama.cpp/conversion
Daniel Bevenius 0ef27309b7 conversion : use original model tensor names in _flush_nvfp4_experts (#31427)
This commit updates the _flush_nvfp4_experts to take a new parameter
which is a dictionary mapping the key to the original model tensor name.

The motivation for this is that I noticed that the tensor names being
looked up were not the original model tensor names, for example:
```console
  model.layers.1.mlp.experts.up_proj.weight
```
But _nvfp4_layer_algo uses the original per-expert names:
```console
(Pdb) p self._nvfp4_layer_algo
{
'backbone.layers.1.mixer.experts.0.up_proj': 'W4A16_NVFP4',
...
```
This means that the lookup will never match for experts and fall back to
_nvfp4_global_algo. If the global algo is W4A16_NVFP4 then this would
go unnoticed, but for a model like
nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-NVFP4 which is
MIXED_PRECISION, the expert tensors will not get prec_a4=false so they
would use the native W4A4 path on Blackwell even though the original
model tensors declare them as W4A16_NVFP4.
2026-11-03 11:40:51 +01:00
..
2026-11-02 18:37:27 +01:00
2026-10-17 00:07:55 +02:00
2026-11-02 18:37:27 +01:00
2026-10-20 18:33:32 +02:00
2026-10-19 23:01:40 +02:00
2026-09-07 21:10:06 +02:00
2026-10-29 15:43:33 +01:00
2026-10-27 11:49:12 +01:00
2026-10-29 18:14:37 +01:00
2026-08-17 10:15:11 +02:00