Update card links to COPA-AI

README.md

@@ -9,8 +9,8 @@ tags:
   - armenian
   - continued-pretraining
 datasets:
-  - C-O-P-A/armweb
-  - C-O-P-A/armstem
+  - COPA-AI/armweb
+  - COPA-AI/armstem
 ---
 
 # arm-gemma-e4b
@@ -18,8 +18,8 @@ datasets:
 **arm-gemma-e4b** is Gemma-4-E4B adapted to Armenian by continued pretraining.
 To our knowledge, it is the **first open Armenian LLM released together with
 its complete training corpus and recipe** — every training token is either
-public ([ArmWeb](https://huggingface.co/datasets/C-O-P-A/armweb),
-[ArmSTEM](https://huggingface.co/datasets/C-O-P-A/armstem), FineWeb-Edu,
+public ([ArmWeb](https://huggingface.co/datasets/COPA-AI/armweb),
+[ArmSTEM](https://huggingface.co/datasets/COPA-AI/armstem), FineWeb-Edu,
 Stack-smol) or reproducible from the released pipeline.
 
 This is a **base model**: no instruction tuning, no chat template. Use it for
@@ -102,7 +102,7 @@ side for translated data. The two remaining reported benchmarks, m-MMLU-hy
 and ARC-hy, were scanned post hoc: zero of 4,306,620 ArmWeb training
 documents and 4 of 372,907 ArmSTEM pairs share any 13-gram with their
 items. Existing public Armenian corpora overlap these benchmarks at
-7.9–17.4% (see the [ArmWeb card](https://huggingface.co/datasets/C-O-P-A/armweb)).
+7.9–17.4% (see the [ArmWeb card](https://huggingface.co/datasets/COPA-AI/armweb)).
 
 ## Limitations