| Name | Version | Summary | date |
| neurograd |
1.8.2 |
A Pure Python Deep Learning Framework with Automatic Differentiation |
2025-08-31 04:36:09 |
| fbgemm-gpu-genai-nightly |
2025.8.30 |
None |
2025-08-30 14:23:36 |
| fbgemm-gpu-nightly-cpu |
2025.8.30 |
None |
2025-08-30 13:26:46 |
| graphistry |
0.41.2 |
A visual graph analytics library for extracting, transforming, displaying, and sharing big graphs with end-to-end GPU acceleration |
2025-08-28 18:08:08 |
| b10-kernel |
0.1.3 |
Baseten Kernel Library |
2025-08-28 14:32:48 |
| aiconfigurator |
0.1.1 |
aiconfigurator: automatic disaggregated serving offline configuration |
2025-08-28 12:40:48 |
| qutip-cuquantum |
0.1.1 |
cuQuantum backend for QuTiP |
2025-08-27 16:30:33 |
| fbgemm-gpu-nightly |
2025.8.26 |
None |
2025-08-26 15:04:15 |
| pmcx |
0.5.1 |
Python bindings for Monte Carlo eXtreme photon transport simulator |
2025-08-26 05:33:34 |
| scalene |
1.5.54 |
Scalene: A high-resolution, low-overhead CPU, GPU, and memory profiler for Python with AI-powered optimization suggestions |
2025-08-25 19:13:06 |
| hf-vram-calc |
1.0.2 |
GPU memory calculator for Hugging Face models with different data types and parallelization strategies |
2025-08-25 15:50:02 |
| megatron-bridge |
0.2.0rc0 |
Megatron Bridge: Training Recipes for Megatron-based LLM and VLM models |
2025-08-25 06:57:40 |
| pmmc |
0.3.0 |
Python bindings for Mesh-based Monte Carlo (MMC) photon transport simulator |
2025-08-25 05:51:39 |
| pmcxcl |
0.3.2 |
Python bindings for Monte Carlo eXtreme (OpenCL) photon transport simulator |
2025-08-25 05:41:35 |
| openlit |
1.35.1 |
OpenTelemetry-native Auto instrumentation library for monitoring LLM Applications and GPUs, facilitating the integration of observability into your GenAI-driven projects |
2025-08-24 14:25:44 |
| prime |
0.3.20 |
Prime Intellect CLI + SDK |
2025-08-24 12:12:54 |
| magnetron |
0.1.0 |
High-performance, lightweight deep-learning library with a PyTorch like API and GPU support. |
2025-08-22 23:31:44 |
| megatron-fsdp |
0.1.0rc0 |
**Megatron-FSDP** is an NVIDIA-developed PyTorch extension that provides a high-performance implementation of Fully Sharded Data Parallelism (FSDP) |
2025-08-22 17:13:31 |
| modelbatch |
0.1.0 |
Train many independent PyTorch models simultaneously on a single GPU using vectorized operations |
2025-08-22 01:22:06 |
| fbgemm-gpu-genai |
1.3.0 |
None |
2025-08-21 22:23:51 |