C-RLFT mixed-quality tuning
According to the model card, the model is fine-tuned with C-RLFT on mixed-quality instruction data without preference labels.
Open Source Model Profile · openchat
openchat_3.5 is a Mistral text-generation fine-tune from openchat trained with C-RLFT on mixed-quality instruction data.
openchat_3.5 is published by openchat as a text-generation fine-tune. Captured configuration identifies MistralForCausalLM with a mistral model type under Apache-2.0. According to the model card, it is a 7B conversational model trained with C-RLFT on mixed-quality instruction data.
According to the model card, the model is fine-tuned with C-RLFT on mixed-quality instruction data without preference labels.
The model card documents an OpenAI-compatible API server using vLLM, listening on localhost:18888, with a 24GB consumer-GPU note and tensor-parallel option.
The model card provides separate chat and coding request examples, including a Code condition for code-oriented prompts and a conversation template.
Captured metadata and the model card both record Apache-2.0 licensing for the model and code.
Source: openchat/openchat_3.5
Captured: Unknown. Processed: 2026-09-07T19:34:54.672903+00:00.
OpenChat: Advancing Open-source Language Models with Mixed-Quality Data GitHub Repo • Online Demo • Discord • Twitter • Huggingface • Paper 🔥 The first 7B model Achieves Comparable Results with ChatGPT (March)! 🔥 🤖 #1 Open-source model on MT-bench scoring 7.81, outperforming 70B models 🤖 OpenChat is an innovative library of open-source language models, fine-tuned with C-RLFT - a strategy inspired by offline reinforcement learning. Our models learn from mixed-quality data without preference labels, delivering exceptional performance on par with ChatGPT, even with a 7B model. Despite our simple approach, we are committed to develo…
F001F002F003F004F005F006F009F010F011F012F014F015F017F018