Add Olmo3 implementation

[2015aroras]

This PR adds the upcoming Olmo 3. The main architectural differences from Olmo 2 are:

• Sliding window attention is used for 3 out of 4 layers. RoPE scaling is not applied to sliding window attention layers.

Since the architecture is very similar to Olmo 2, this PR opts to merge Olmo 3 changes into the Olmo 2 implementation (similar to vllm-project/vllm#24534). I can create a separate Olmo 3 implementation instead if preferred.

[2015aroras]

I used the model conversion example for testing. I got the following results when using bf16 on shanearora/2025-sep-a-base-model, modified to have yarn rope scaling enabled.

📈 METRICS
==============================
MSE (Mean Squared Error):     1.592396e-02
Reference Variance:           6.831117e+00
NMSE:                         2.331092e-03
Max Absolute Error:           0.438750
Mean Absolute Error:          0.116665
NMSE (dB):                    -26.32 dB

🎯 INTERPRETATION
==============================
👍 Good match

📋 GUIDANCE
==============================
👍 GOOD: Your GGML conversion is working well.
   Small differences are likely due to precision/quantization.

📚 NMSE BENCHMARKS
==============================
✅ RESULT: PASS (NMSE = 2.33e-03)

Also, below is the allenai/OLMo-2-0425-1B with fp32.

📈 METRICS
==============================
MSE (Mean Squared Error):     1.594746e-03
Reference Variance:           9.219801e+00
NMSE:                         1.729697e-04
Max Absolute Error:           0.168732
Mean Absolute Error:          0.033951
NMSE (dB):                    -37.62 dB

🎯 INTERPRETATION
==============================
👍 Very good match

📋 GUIDANCE
==============================
✅ EXCELLENT: Your GGML conversion is working very well!
   The differences are negligible for practical use.

📚 NMSE BENCHMARKS
==============================
✅ RESULT: PASS (NMSE = 1.73e-04)

[pwilkin]

@2015aroras What tool are you using to compare the conversion?

[2015aroras]

@pwilkin I am using the model conversion tools inside this repo. These have been created to help make sure HF to llama.cpp conversion is accurate. The logs above are from the Model logits verfication step.

[pwilkin]

Ah, that's nice, I haven't used that specific one yet :)

[2015aroras]

All the check failures seem to be unrelated to this change. Before merging master again, an ios check was failing instead. So imo ready to merge.

[CISC]

All the check failures seem to be unrelated to this change. Before merging master again, an ios check was failing instead. So imo ready to merge.

Yes, sorry for the delay, just a minor cosmetic change and we'll merge. :)