Code 1pct 9e18 Seed2 2026 07 12 00 08 42 554920 Pt is a machine learning model checkpoint hosted on Hugging Face. The model is trained with nanochat and represents a checkpoint at step 12,281. Its configuration specifies a sequence length of 2048, a vocabulary size of 65,536, 18 layers, 9 attention heads, 9 key-value heads, and an embedding size of 1,152. The license for this model is Apache 2.0, indicating it is available under an open-source license. Training details include the use of a matrix learning rate of 0.015 and an embedding learning rate, with a total training time of approximately 3,183 units (the unit is not specified in the evidence). The outcome metrics provided are smooth_train_loss and min_objective, along with computational statistics such as flops used and flops per token. The model is delivered as a downloadable checkpoint file from Hugging Face. No further information about its intended audience, specific applications, or integration capabilities is provided in the available evidence.
Code 1pct 9e18 Seed2 2026 07 12 00 08 42 554920 Pt sits in PulseGate's Other AI category. It focuses on providing an open-source code language model checkpoint for AI research and development. It is built as an open-source project for AI researchers and developers. Code 1pct 9e18 Seed2 2026 07 12 00 08 42 554920 Pt is open source under the MIT license. Code 1pct 9e18 Seed2 2026 07 12 00 08 42 554920 Pt is available on the web and API.
It is developed by alexkstern, and the product first shipped in 2025. Development happens publicly on GitHub with 56.2k stars and 34 commits in the last 90 days. It operates in a well-populated space: PulseGate tracks 19 similar tools. Key capabilities include code generation, model checkpoint, and open weights.
Latest indexed changes and source events
alexkstern/code_1pct_9e18_seed2_2026-07-12_00-08-42_554920-pt discovered by the PulseGate indexer
Other apps tracked under the same category.
See 35 alternatives to Code 1pct 9e18 Seed2 2026 07 12 00 08 42 554920 Pt
