license
stringclasses
1 value
messages
listlengths
12
230
tools
listlengths
2
6
uuid
stringlengths
36
36
used_in
listlengths
1
1
metadata
listlengths
0
0
CC BY 4.0
[ { "role": "user", "content": "Implement a vector element-wise multiplication kernel on the GPU.\nUse the provided GpuTimer struct to measure kernel execution time.\n\nYour task:\n1. Implement the elementWiseMultiply kernel in solution.cu\n2. Complete the applyMultiply function to launch the kernel and time ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
689d816b-c933-50af-881e-d12c6913bb70
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a kernel to compute the Intersection over Union (IoU)\nmatrix for pairs of bounding boxes. This is a fundamental operation\nin object detection for evaluating how well predicted boxes match\nground truth boxes.\n\nYour function must compute IoU for all pairs between t...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
5db80653-96f6-566a-98a7-5edc1dccf181
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function to convert a block sparse row (BSR) format matrix to compressed sparse row (CSR) format using cuSPARSE.\nThe function signature is in include/solution_header.h.\n\nBSR format stores non-zero values in blocks of dense elements (blockDim x blockDim).\nYour im...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
078a4c0c-2640-5244-9fbd-5c389b32887a
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that combines cuDNN transforms with row-wise normalization.\n\nYour task is to:\n1. Copy matrix A from leading dimension layout to dense layout\n2. Apply row-wise scaling (gamma) and shifting (beta) normalization\n3. Use cuDNN transform operations as needed...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
7cf1b1c6-3d23-5e31-b947-be23bce66829
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement an architecture-aware element-wise scaling operation that uses different\noptimization kernels based on the GPU's compute capability.\n\nThe function should compute output[i] = input[i] * scale for all elements, but\ndispatch to different kernel implementations depend...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
0b9b0b38-0113-5f35-ab0b-2cc80747365a
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a row-wise matrix reduction that computes the sum of each row.\n\nGiven an M x N matrix in row-major order, compute an M-length output array\nwhere output[i] = sum of all elements in row i.\n\nYour implementation should:\n1. Use efficient thread indexing for 2D matric...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
43458015-2e6d-5682-8aff-87cc5dfcc14c
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement eigenvalue decomposition of a symmetric matrix using cuSOLVER.\n\nYour function should compute the eigenvalues and eigenvectors of a real symmetric\nmatrix A such that A = V * Λ * V^T, where V contains the eigenvectors and Λ is a\ndiagonal matrix of eigenvalues.\n\nYo...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
e2e2ac66-2279-558a-9a15-fa7634060176
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a scaled matrix-vector multiplication using cuBLAS.\nThe function should compute y = alpha * A * x, where A is an m×n matrix in column-major format.\nYour implementation must use cublasSgemv with proper alpha and beta parameters.\nYour function signature is in include...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
ff9b4b89-f3b1-595c-b894-d2f400f29b3e
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that computes scaled matrix-vector multiplication: y = alpha * A * x + beta * y\nwhere A is a dense matrix in row-major format, x and y are vectors, and alpha/beta are scaling factors.\n\nYour function signature is in include/solution_header.h.\nUse CUBLAS ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
de98ecd8-eaa3-5667-a5f9-f4a0d1ae0e74
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a batched FFT roundtrip (forward + inverse) using cuFFT.\n\nYour implementation should:\n1. Use cufftXtMakePlanMany to create batched FFT plans for FP16 (half) data\n2. Execute forward FFT transforming real input to complex frequency domain\n3. Execute inverse FFT tra...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
aeb5f559-cb63-5706-a606-d38353a00c77
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement per-batch element scaling: output[i][j] = input[i][j] * scale_i for each batch i.\nUse a custom CUDA kernel with thread indexing and boundary checking.\n\nYour implementation should:\n1. Define a kernel that handles flattened indexing across batches and elements\n2. U...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
99f682f9-2f0c-5881-aaee-54f2c3f866f7
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a batch normalization apply kernel for matrices.\nThe operation applies per-column scaling and bias to each element:\n\n output[row, col] = input[row, col] * alpha_prime[col] + bias_prime[col]\n\nThis is a simplified form of batch norm where alpha_prime and bias_pr...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
4e5a2709-16fa-5fbb-bf7a-66ebeb9dba1b
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a parallel 3D point rotation kernel around the Y-axis.\nYour function signature is in include/solution_header.h.\n\nThe kernel must:\n- Use 1D thread grid with blockIdx.x and threadIdx.x\n- Handle boundary checking for arbitrary n\n- Ensure memory coalesced access (co...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
bd75852c-49b8-5101-b534-201fbeb8d9c2
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a triangular banded matrix-vector multiplication with post-scaling.\n\nYour function must compute x = scale_factor * op(A) * x where:\n- x is a complex double-precision vector\n- A is a triangular banded matrix with k super-or-sub-diagonals\n- op(A) is A, A^T, or A^H ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
1b0e5b17-9eea-56d1-8678-8f12e1a43ad5
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a sparse matrix-vector multiplication (SpMV) using cuSPARSE.\nThe matrix is stored in CSR (Compressed Sparse Row) format.\nYour function signature is in include/solution_header.h.\n\n## Environment Instructions\n- Write your implementation to `solution/solution.cu`.\n...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
94165885-ef50-505a-afff-81fe31261e8c
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a CUDA context manager singleton class that manages cuBLAS and cuRAND handles.\nThe class must provide thread-safe singleton access via a static Get() method and expose\nthe handles through inline getter functions.\n\n## Environment Instructions\n- Write your implemen...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
55f520a7-1de9-5adf-be66-fec2c8e29668
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a row-wise dot product of two matrices using cuBLAS.\n\nGiven two matrices A and B of shape (M x N), compute the dot product\nfor each row: result[i] = sum(A[i][j] * B[i][j]) for j = 0..N-1.\n\nUse cuBLAS cublasSdot for efficient computation. Create the cuBLAS\nhandle...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
1fdc97c3-4357-56b7-8fd5-08452be2c3f8
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement element-wise vector addition using Thrust.\nGiven two input vectors a and b, compute c = a + b element-wise.\nYour function signature is in include/solution_header.h.\n\n## Environment Instructions\n- Write your implementation to `solution/solution.cu`.\n- Tests may b...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
9417cf67-45d2-50a5-96ee-8a2e63886c47
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement an INT8 GEMM kernel with column and row scaling using raw CUDA.\nYour function signature is in include/solution_header.h.\n\nThe function should:\n- Perform matrix multiplication C = A * B where A is (m x k) and B is (k x n)\n- Apply column scaling: each output row i ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
c93182a6-341e-51e3-8076-e9b3c046fa91
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a staged element-wise operation using CUDA events for synchronization.\n\nThe operation applies `result[i] = a[i] * a[i] * k` (square elements, then multiply by constant).\nImplement this in TWO stages:\n1. First kernel: computes temp[i] = a[i] * a[i]\n2. Second kerne...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
1eea5418-1a18-52aa-8958-cf2d7631bd00
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function to set up a cuDNN 4D tensor descriptor with custom strides.\nThe function should configure the tensor descriptor to describe a 4D tensor with\ndimensions (n, c, h, w) where:\n- n: batch size\n- c: channels\n- h: height\n- w: width\n\nYour function signature...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
491f3ea7-7b53-5835-aa98-0c8504fd12a9
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a matrix addition kernel C = A + B using CUTLASS concepts.\n\nYour implementation should:\n1. Use 2D grid/block indexing for tiled processing\n2. Handle non-square matrices gracefully (rows and cols may differ)\n3. Include boundary checking for edges\n4. Use the funct...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
3331b6dd-a68a-5ba9-9ce4-28cd83ee0ff4
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a CUDA kernel that applies GELU activation with linear combination scaling.\n\nGiven input and optionally bias arrays, compute:\n```\nscaled = alpha * input + beta * bias\noutput = gelu(scaled)\n```\n\nThe GELU (Gaussian Error Linear Unit) activation is computed as:\n...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
5f6a6a1b-6041-538b-8719-258fc78ba445
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a high-precision reduction kernel that computes the sum of an array using\na split-precision technique inspired by the TF32x3 algorithm.\n\nThe kernel should:\n1. Split each floating-point value into high and low precision components\n2. Reduce using shared memory and...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
7eebcd51-b4e7-5924-87e8-98671cffdb92
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a matrix scaling function using cuBLAS: output = alpha * input\nThe matrix is stored in column-major format (like Fortran/Matlab).\nYour function signature is in include/solution_header.h.\n\nUse cublasScal to scale all elements of the matrix, treating it as a single ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
15aa9015-8e52-5da5-98dd-4b8617d5c53d
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a scale and quantize kernel that converts floating-point values to int8.\n\nYour implementation should:\n1. Apply a scaling factor to each input element: scaled = input[i] * scale\n2. Round to nearest integer using round-half-to-even (rn) mode\n3. Cast the result to i...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
ad67df40-b546-5d7a-ac4b-b8f9b0516c68
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a batched LU factorization solver using cuBLAS.\nYour function should perform LU decomposition on a batch of matrices.\n\nFunction signature:\nint solve_batched_lu(cublasHandle_t handle, \n double* A, int* ipiv_array, int* info_array, \n ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
30841772-d185-56f3-b58e-e62edd9d89d9
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a stable softmax kernel for FP16 matrices.\nYour function signature is in include/solution_header.h.\n\nThe kernel should:\n1. Apply a scale factor to each input element\n2. Compute softmax row-wise using the numerically stable approach (subtract row max before exp)\n...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
432ee545-b2f2-57a6-828c-832318548961
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a 2D convolution forward pass using cuDNN.\n\nYour implementation should:\n1. Create and configure cuDNN tensor descriptors for input and output\n2. Create and configure filter descriptor for the convolution kernel\n3. Create convolution descriptor with padding, strid...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
6b6f4486-39df-5085-af60-aefcc8d16b87
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a row-wise online softmax kernel for a 2D matrix.\n\nEach row is processed independently in parallel. The twist: implement the ONLINE\nsoftmax algorithm used in Flash Attention, which processes data incrementally\nwithout materializing intermediate exponentials. This ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
0825ffbc-62d4-53de-989d-3e110dd518ad
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a fused operation that scales a floating-point array by per-element and uniform scale factors, \nthen computes the sum of the result. This combines element-wise operations from two CUDA kernels: per-array scaling\nwith type conversion.\n\nThe operation for each elemen...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
ce26ba52-7183-5de2-8bcc-00efa6b38962
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function to convert a sparse matrix from CSR (Compressed Sparse Row) format\nto BSR (Block Sparse Row) format using cuSPARSE library.\n\nCSR format stores:\n- row_ptr[i] to row_ptr[i+1]-1: range of elements in row i\n- col_ind[j]: column index of element j\n- values...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
4bdeadef-b049-5d86-8c98-aee1bbc5918f
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a scaled matrix multiplication kernel (D = alpha * (A @ B) + beta * C).\n\nGiven matrices A (M x K) and B (K x N) of float32 values, compute:\n D = alpha * (A @ B) + beta * C\n\nWhere:\n- A, B, C are row-major float32 matrices\n- C may be null if beta = 0\n- alpha an...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
e19b16e2-fc1f-5dba-881a-154e7c453a61
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement an LSTM forward pass kernel using CUDA.\nYour function signature is in include/solution_header.h.\n\nThe function should:\n1. Parse the weight buffer into LSTM gate matrices (ii, if, ig, io, hi, hf, hg, ho, bi, bf, bg, bo)\n2. Apply each time step sequentially using t...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
6eed8996-7a67-5610-878b-28fb36de3ab0
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that solves triangular systems of linear equations op(A) * X = alpha * B using cuBLAS.\n\nThe function solves multiple right-hand side systems where:\n- A is a triangular matrix (n x n) \n- B is the right-hand side matrix (n x nrhs), which is overwritten wi...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
976f5e3e-b296-5d5d-8ed3-a79665f11540
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that fills a CUDA device buffer with a scalar value and downloads it back to host memory asynchronously.\n\nThe function should:\n1. Launch a kernel to fill the device buffer with the specified value\n2. Asynchronously copy the buffer from device to host us...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
a0f9d54f-9aaa-5646-9ef3-78a86ef7ca17
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function to compute eigenvalues within a specified range for a generalized\nHermitian eigenvalue problem A*x = lambda*B*x using cuSOLVER's cusolverDnZhegvdx and\ncusolverDnZhegvdx_bufferSize APIs.\n\nYour function signature is in include/solution_header.h.\n\n## Env...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
e081cbb2-05fb-5a4c-bea1-3b276d44461d
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a matrix scaling kernel using Thrust.\nGiven an array of 4x4 matrices and scale factors, apply scaling to each matrix.\nYour function signature is in include/solution_header.h.\n\n## Environment Instructions\n- Write your implementation to `solution/solution.cu`.\n- T...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
3fc6d631-398b-5c38-a5af-c188bb96306e
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that generates random float values in a specified range [min_val, max_val] using CUDA curand.\n\nYour function signature is in include/solution_header.h.\n\nRequirements:\n- Use curand_uniform() to generate random values in [0, 1]\n- Scale the output to the...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
e59c2d83-1ba5-5719-804d-256fc0044ba2
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement matrix addition using cuBLAS geam with async stream operations.\nYour function should compute C = alpha*A + beta*B elementwise.\nThe function signature is in include/solution_header.h.\n\n## Environment Instructions\n- Write your implementation to `solution/solution.c...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
d7b2022e-2f96-5a0d-bbe9-6db216ee8c85
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that performs triangular band matrix-vector multiplication using cuBLAS.\nThe operation computes: x = T * x (in-place), where T is a triangular band matrix.\n\nUse cublasZtbmv_v2 to perform the computation.\n\nYour function signature is in include/solution_...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
ead7a82a-f095-51d8-8728-b001e164a85e
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a sparse eigenvalue power method using cuSPARSE.\n\nGiven a symmetric sparse matrix in CSR format, use power iteration to approximate\nthe largest (dominant) eigenvalue and corresponding eigenvector.\n\nYour implementation should:\n1. Use cuSPARSE for the sparse matri...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
f2696606-2304-5183-9269-f1b38f86800c
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a Sobel gradient computation kernel that computes both gradient \nmagnitude and direction for an image. \nYour function signature is in include/solution_header.h.\n\nThe kernel should:\n1. Apply Sobel X and Sobel Y convolution kernels to compute horizontal (gx) and ve...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
0446290d-b6a5-59c2-80a1-e0389a158d40
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that computes the absolute maximum value of elements in a matrix with custom strided memory layouts.\n\nThe function signature is in include/solution_header.h.\n\nYour kernel should:\n1. Read matrix elements using the provided row_stride and col_stride\n2. ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
b82e292b-2646-5594-b703-467878f37fc7
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a tiled matrix multiplication kernel using shared memory and thread coalescing.\nThe kernel should compute C = alpha * A * B + beta * C where:\n- A is M x K matrix (row-major)\n- B is K x N matrix (row-major)\n- C is M x N matrix (row-major)\n- alpha and beta are scal...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
adaeaa53-a55a-5d99-89dc-e5f4074dbdce
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a cuDNN softmax forward inference function that computes softmax over\nthe channel dimension (C) for each batch element and spatial position.\n\nYour function should:\n1. Set up cuDNN tensor descriptors for input and output (4D NCHW format)\n2. Choose appropriate soft...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
23e6e0a8-800e-5828-bf0f-d7d87669fc8e
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement eigenvalue computation for real symmetric matrices using cuSOLVER Jacobi method.\n\nYour implementation should:\n1. Use cusolverDnSsyevj (Jacobi iterative method) for symmetric eigenvalue decomposition\n2. Compute only eigenvalues (CUSOLVER_EIG_MODE_NOVECTOR), not eig...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
29ddc74a-4084-5f6a-99ea-097d3503627a
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a BiasedLeakyReLU activation layer with forward and backward passes.\n\nThe BiasedLeakyReLU performs element-wise:\n output = leaky_relu(input + bias)\n\nwhere leaky_relu(x) = max(slope * x, x) with slope=0.01 (negative slope).\n\nYour implementation must:\n- Support...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
935b7f3f-0dbf-5ea7-ad3a-243f420a01f6
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a 2D max pooling operation using cuDNN.\n\nYour implementation should:\n1. Use cuDNN's pooling API to perform max pooling\n2. Configure pooling descriptor with window size, stride, and padding\n3. Create and configure tensor descriptors for input and output tensors\n4...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
c11cccbf-386c-59ea-8693-bf58ea4f06c6
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a CUDA function that prunes a CSR sparse matrix by removing values below a threshold and converts the result to COO format.\n\nYour function signature is in include/solution_header.h.\n\nThe function should:\n1. Remove entries from CSR matrix where |value| <= threshol...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
02e73239-28fb-5256-a31a-765fdd3892ed
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a multi-stream task executor that launches independent GPU kernels\nand uses CUDA events to synchronize between streams.\n\nYour function signature is in include/solution_header.h.\n\nThe implementation should:\n1. Create separate CUDA streams for each independent tas...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
9c1a9bab-ed78-501f-9900-4151e8227c51
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement element-wise tensor addition using cuDNN.\n\nYou need to use cuDNN tensor descriptors and the OpTensor API to perform \nelement-wise addition: output = input_a + input_b.\n\nYour function signature is in include/solution_header.h.\n\nThe function receives two 2D half-...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
6fd6fd90-9a21-583d-863c-b1034f04567b
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a vector operation sequence using cuBLAS.\n\nYour implementation should:\n1. Compute the dot product of two vectors using cublasDdot\n2. Scale vector A by the dot product result using cublasDscal\n3. Add scaled vector A to vector B using cublasDaxpy\n4. Store the fina...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
d06f0b91-cc00-5f7a-8e3a-7f4affa29684
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a matrix transpose kernel using shared memory.\nTransposing through shared memory avoids uncoalesced memory access patterns.\n\nYour function signature is in include/solution_header.h.\nUse TILE_SIZE=16 for the shared memory tile size.\n\n## Environment Instructions\n...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
49dcd4a0-c380-57b0-ada2-41ee054fae84
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a contrast and brightness adjustment kernel for grayscale images.\nYour function signature is in include/solution_header.h.\n\nThe transformation for each pixel is:\n output = clamp(contrast * (input - 128) + 128 + brightness, 0, 255)\n\nRequirements:\n- Use 1D grid/...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
8b6ea3a2-0281-5244-b0f2-6ac6571dbc30
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a FP32 to FP16 conversion and scaling kernel.\n\nYour implementation should:\n1. Convert FP32 input values to FP16 (half precision)\n2. Scale the FP16 values by a scale factor\n3. Convert back to FP32 and store results\n4. Handle arrays of arbitrary size with proper b...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
bcad976b-2032-5766-af15-4de006dffe98
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a fused element-wise scaling and type conversion kernel.\nThe function scales each element of a double array with a per-element alpha factor\nand a uniform beta factor (alpha[i] * x[i] + beta), then converts the result to float.\n\nYour function signature is in includ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
062c5a96-f681-5832-9607-5d8e08522e90
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a sparse matrix-vector multiplication (SpMV) using cuSPARSE in CSR format.\n\nYour implementation should:\n1. Use cusparseDcsrmv for double-precision SpMV\n2. Set up proper cuSPARSE matrix descriptor (general type, zero-based indexing)\n3. Compute y = A * x where A is...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
2897c4b6-2b96-588e-a30f-b00f49be87cf
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement batched singular value decomposition (SVD) for a set of matrices using cuSOLVER's gesvdjBatched function.\nGiven a batch of M×N matrices, compute their SVD decomposition A = U * S * V^T.\nYour function signature is in include/solution_header.h.\n\n## Environment Instr...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
2a816c66-4004-5ef6-b17e-c5032cf0132f
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a scaled sparse matrix-vector multiplication using cuSPARSE.\n\nGiven a CSR (Compressed Sparse Row) format sparse matrix A of size n x n,\ncompute y = alpha * A * x where alpha is a scalar multiplier.\n\nYour function signature is in include/solution_header.h.\n\nRequ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
9e62ac5f-585b-561d-b1d4-51d97e7dbc9a
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a matrix multiplication using cuBLAS with configurable alpha/beta scaling.\n\nYour implementation should:\n1. Use cublasSgemm for FP32 general matrix multiplication\n2. Compute C = alpha * A * B + beta * C\n3. Support matrix transposition for both A and B\n4. Handle m...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
086e8fda-1994-5a28-bc76-ce39856ea0df
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a transform-exclusive-scan operation on an array of floats.\nYour function signature is in include/solution_header.h.\n\nThe function should: \n1. Transform each input element by squaring it (x -> x*x)\n2. Compute an exclusive scan on the transformed values\n3. Use th...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
a7a5227d-05d3-5077-8952-65b5b5a09427
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function to perform Cholesky factorization of a symmetric POSITIVE DEFINITE matrix using cuSOLVER.\n\nThe function should:\n1. Query the workspace size using cusolverDnDpotrf_bufferSize\n2. Perform Cholesky factorization using cusolverDnDpotrf\n3. Check the status t...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
cce884fa-1882-579a-b199-65cd7060123a
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that orthonormalizes the columns of a matrix using QR decomposition.\nThe function signature is:\nint orthonormalize_columns(float* d_A, int64_t m, int64_t n, int64_t lda);\n\nThe input matrix A is of size m x n with m >= n (column-major storage).\nThe outp...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
f95ddc13-da79-53ff-bc57-19dd29ac80a2
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement workspace-aware matrix multiplication using cuBLAS with half-precision (FP16) data.\n\nYour task is to:\n1. Calculate total workspace size with 128-byte alignment for multiple buffers\n2. Perform FP16 matrix multiplication using cuBLAS with workspace management\n\nThe...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
f758f3e1-d0ae-580b-9017-04f603e4938d
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a sparse-dense matrix multiplication (SpMM) using cuSPARSE.\n\nThe function `cusparseSpMM` should compute C = alpha * A * B + beta * C where:\n- A is a sparse matrix in CSR (Compressed Sparse Row) format\n- B is a dense matrix (row-major order)\n- C is a dense output ...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
ac77edd0-6a92-5835-b5fa-3418ac569cd0
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a stochastic cellular automaton kernel using cuRAND.\n\nThe kernel performs Game of Life-style updates with probabilistic survival and birth:\n- A live cell survives if it has 2 or 3 live neighbors AND a random value <= survival_prob\n- A dead cell becomes alive if it...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
7c836695-a8a6-51e4-b740-e1617fa6600a
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement complex floating-point matrix multiplication using cuBLAS.\nYour function signature is in include/solution_header.h.\n\nCompute C = alpha * A * B + beta * C, where A is m x k, B is k x n, and C is m x n.\nAll matrices contain complex floating-point numbers (cuComplex)...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
1f6f31a7-fe3f-57ec-8452-39a1ea29badd
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that computes the 2D FFT of complex input data and then calculates \nthe magnitude (absolute value) of each complex element in the frequency domain.\n\nThe function should:\n1. Execute a 2D forward FFT on the input complex array using the provided cuFFT pla...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
741a2ac9-3351-5934-b7fa-980d6d0c1928
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a quantized matrix transposition kernel with per-group scaling.\nThe transposition should be done efficiently using shared memory tiling.\n\nYour function signature is in include/solution_header.h.\n\nInput: A quantized matrix storing int8_t values in row-major format...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
d3053aea-7153-5c78-9f2e-45c20828372c
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a fused linear combination operation that computes z = alpha * x + beta * y + z\nusing cuBLAS. Your function signature is in include/solution_header.h.\n\nThis operation combines two vector axpy operations into a single efficient computation.\nThe twist is that z is u...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
8c1bb451-08dd-5d59-9b3e-0809126e266b
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a kernel that converts a float32 matrix to packed int4 format.\n\nEach output byte packs two signed int4 values:\n- Lower 4 bits hold the first value (even index)\n- Upper 4 bits hold the second value (odd index)\n\nConversion formula: int4_out = clamp(float_in * scal...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
3f0fb281-288f-5d9f-a3ca-5a97e7a8c3e4
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that queries CUDA device attributes using the CUDA Driver API.\nYour function signature is in include/solution_header.h.\n\nThe function should:\n1. Load the cuDeviceGet and cuDeviceGetAttribute driver function pointers using cudaGetDriverEntryPoint\n2. Que...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
e85a98a5-e538-5369-b5c3-58364255f895
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement batched matrix multiplication with element-wise scaling using cuBLAS.\nCompute: C[i] = alpha * (A[i] * B[i]) for each matrix in the batch.\nYour function signature is in include/solution_header.h.\n\n## Environment Instructions\n- Write your implementation to `solutio...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
8211573f-4c07-5234-be51-7ed48344cdda
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function to convert a sparse matrix from CSR (Compressed Sparse Row) format to GEBSR (General Block Sparse Row) format using cuSPARSE.\n\nYour task:\n- Given a CSR matrix with values, row pointers, and column indices\n- Convert to GEBSR format with the specified blo...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
950e1181-dd6d-557e-8263-54a2ce49fd2b
[ "ultra_v3" ]
[]
CC BY 4.0
[ { "role": "user", "content": "Implement a function that prints GPU memory consumption information.\n\nGiven a CUDA device, retrieve information about free memory and total memory.\nOutput the percentage of free memory and the percentage of used memory.\n\nYour function signature is in include/solution_heade...
[ { "type": "function", "function": { "name": "bash", "description": "Executes a given bash command in a persistent shell session with optional timeout, ensuring proper handling and security measures.\n\nBefore executing the command, please follow these steps:\n\n1. Directory Verification:\n - I...
25cbee06-e979-59fd-9c11-86da1dcdde91
[ "ultra_v3" ]
[]