Upload folder using huggingface_hub

loras/cuda/README.md

@@ -0,0 +1,4 @@
+# cuda
+**Rank**: 100 | **Loss**: 3.7179
+
+CUDA expert. GPU kernels, memory hierarchy, warp-level programming, Nsight profiling, optimization.

loras/cuda/adapter.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f40a9f8a3c729eaaa6f48b0573cd634d615c61689062042954aae27525c6d014
+size 697177068

loras/cuda/config.json

@@ -0,0 +1 @@
+{"name": "cuda", "system": "CUDA expert. GPU kernels, memory hierarchy, warp-level programming, Nsight profiling, optimization.", "rank": 100, "alpha": 200, "scaling": 2.0, "best_loss": 3.717874526977539, "train_time": 70.4920289516449, "samples": 3000, "type": "coding", "gpu": "H100 (Modal)", "by": "NiftyOracle codesymb.com"}