Granite 4.1: IBM's 8B Model Matching 32B MoE

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by EagnaIonat on Thursday April 30, 2026 @12:57PM (#47965281)

Can you clarify what you mean?

If you check HF you will see its Apache2 and the datasets were also permissive.

It's one of the few models on the market where the creator indemnifies it against copyright claims.

https://research.ibm.com/blog/granite-ethical-ai

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Thursday April 30, 2026 @01:24PM (#47965616)

Oh sorry. Do we have the sources like Nvidia's Nemotron?

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by EagnaIonat on Thursday April 30, 2026 @04:02PM (#47967514)

You could have found in 5 seconds. The weights are also open sourced as well.

https://github.com/ibm-granite

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Thursday April 30, 2026 @05:05PM (#47968219)

Maybe I suck but I didn’t find that in 5 seconds. Or with more time.

I meant the full training datasets and the complete recipes to make the models.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by EagnaIonat on Friday May 01, 2026 @02:28AM (#47971811)

The training datasets are listed there and are all open source.

> the complete recipes to make the models.

You mean the weights which most companies don't release. Again you can find from that link.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Friday May 01, 2026 @04:32AM (#47972463)

Where is the list?

No I didn't mean the weights, but the source code to make the weights.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by EagnaIonat on Saturday May 02, 2026 @05:03AM (#47984705)

I'd like to assume you are not trolling and just want everything handed to you.

The granite site covers everything you keep asking for. Granite is made using lm-engine and the details are there.

Without the weights you are not going to be able to build to the same level of accuracy without some serious work.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Saturday May 02, 2026 @11:56AM (#47987499)

I want open source to stay a label for actual open source software. That means handing everything to me yes.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by robotmaxtron on Saturday May 02, 2026 @01:06PM (#47988205)

No, it doesn't.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by jeraldbenny on Friday May 01, 2026 @02:36AM (#47971841)

if you are googling you can find so many open source dataset. Also use kaggle, they're also having training datasets which we can use.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Friday May 01, 2026 @04:32AM (#47972466)

I know, but that is not my point.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by jasonlotito on Friday May 01, 2026 @10:55AM (#47975507)

You never had one. You tried to be clever and failed.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Friday May 01, 2026 @12:45PM (#47976912)

I don’t know. Since you are perhaps clever, can you show me the training datasets and recipes so I can replicate this model locally? I have access to good HPCs.

I think it’s fair if you use a bit more than 5 seconds as someone stated above. I would gladly be proven stupid.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by EagnaIonat on Saturday May 02, 2026 @05:05AM (#47984723)

https://github.com/ibm-granite

https://huggingface.co/ibm-granite

I think if you were genuinely interested, you could have found this yourself.

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by speedgoose on Saturday May 02, 2026 @11:55AM (#47987491)

This is not what I asked?

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by robotmaxtron on Thursday April 30, 2026 @04:26PM (#47967768)

if I can't reproduce the artifact, is it really open source?

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by postalrat on Thursday April 30, 2026 @05:51PM (#47968700)

If IBM themselves can't reproduce the artifact do they have the source?

[ Parent | on HN ]

Re:Granite 4.1: IBM's 8B Model Matching 32B MoE by robotmaxtron on Thursday April 30, 2026 @07:29PM (#47969655)

I guess not

[ Parent | on HN ]