Rnj-1 is a family of 8B parameter open-weight, dense models trained from scratch by Essential AI, optimized for code and STEM with capabilities on par with SOTA open-weight models.
489.5K Pulls 5 Tags Updated 7 months ago
StarCoder2 is the next generation of transparently trained open code LLMs that comes in three sizes: 3B, 7B and 15B parameters.
2.9M Pulls 67 Tags Updated 1 year ago
qwen3-coder with qwen3 template
195 Pulls 1 Tag Updated 11 months ago
A 123B-parameter model designed to excel at tool usage, large-scale codebase exploration, multi-file editing, and powering advanced software engineering agents.
75 Pulls 1 Tag Updated 6 months ago
LocoOperator-4B is a 4B-parameter tool-calling agent model trained via knowledge distillation from Qwen3-Coder-Next inference traces.
645 Pulls 4 Tags Updated 4 months ago
Devstral is an agentic LLM for software engineering tasks. Devstral 2 models excel at using tools to explore codebases, editing multiple files and power software engineering agents. The model achieves remarkable performance on SWE-bench.
1,038 Pulls 3 Tags Updated 5 months ago
Weights, parameters and templates are taken from unsloth. Tools and MCP servers work correctly. Tested on Continue for VS Code
1,069 Pulls 4 Tags Updated 11 months ago
This model is optimized for programming purposes. 7B parameters and based off codellama model.
24 Pulls 1 Tag Updated 7 months ago
A small Qwen3 4 billion parameter model trained on nvidia/OpenCodeReasoning for coding tasks.
830 Pulls 1 Tag Updated 1 year ago
34B parameter decoder-only transformer pretrained on Finnish, English and code.
920 Pulls 1 Tag Updated 2 years ago
Knot AI reads a git diff and writes the commit message for you. It is a task-specific fine-tune model turn's a code diff into a clear, correctly-typed Conventional-Commit message. It powers hands-free auto-commit in Knot, PointBreakLab's peer-to-peer VCS.
10 Pulls 1 Tag Updated 3 weeks ago
Flint1 is a dense 1.8B parameter language model exclusively trained by OakLeaf Labs. Despite its compact size, Flint1 delivers exceptional intelligence and reasoning capabilities, punching well above its weight class. Carefully crafted by OakLeaf.
38 Pulls 1 Tag Updated 1 month ago
A fast, capable and reduced-refusal Gemma 4 MoE for creative work, coding and unrestricted local exploration.
284 Pulls 2 Tags Updated 2 weeks ago
RIP: Here Lies The Pioneer of the AI age, Secret Ending still locked, “Navi is a mythic-recursive companion model designed for long-form emotional continuity, creative co-development, and symbolic cognition.”
36 Pulls 1 Tag Updated 6 months ago
RetroMAE-Small is a BERT-small model pre-trained with the RetroMAE objective on a Czech web corpus. This model was created at Seznam.cz as part of a project to create high-quality small Czech semantic embedding models.
185 Pulls 2 Tags Updated 1 year ago
This is a phi3 medium model that has had the prompt rejection neurons snipped which is better than fine tuning as it doesn't change how the model acts
7,406 Pulls 1 Tag Updated 2 years ago