Eyght

eyght-v-tx

README

License: apache-2.0

Quick start

python
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
base = "Qwen/Qwen2.5-7B-Instruct"
model = AutoModelForCausalLM.from_pretrained(base, torch_dtype="auto", device_map="auto")
model = PeftModel.from_pretrained(model, "Eyght/eyght-v-tx")
tok = AutoTokenizer.from_pretrained(base)
print(tok.decode(model.generate(**tok("Write a Python function to reverse a list.", return_tensors="pt").to(model.device), max_new_tokens=256)[0]))

Details

Table

Base model	`Qwen/Qwen2.5-7B-Instruct`
Adapter	PEFT LoRA (r=16, q/k/v/o/gate/up/down)
Format	`adapter_config.json` + `adapter_model.safetensors`
Trainable params	~40 M (~0.5% of 7.66 B)
Dataset	`eyght-v-tx` (~20,000 coding + reasoning examples)
Framework	Unsloth (4-bit + LoRA SFT)
Hardware	1x NVIDIA RTX 4070 Ti (12 GB)

Merge to a standalone model (optional)

python
merged = model.merge_and_unload()
merged.save_pretrained("eyght-v-tx-merged")

Then quantize to GGUF (Q4_K_M ~4.7 GB) for Ollama.

License & attribution

Adapter: Apache-2.0. Built, trained, and owned by Eyght.
Respect the base model license when redistributing.

Built by Eyght with the local Eyght Veta studio.

🧠 Project Janus — Dual-Loop Cognitive Architecture

This model is designed to operate within Project Janus, a dual-loop cognitive architecture that goes beyond simple prompt-response:

The Core Cognitive Loop

Perceive → Working Memory → Internal Critic → Action → Consolidation

Table with columns: Phase, Component, Function
Phase	Component	Function
1. Perceive	Input mapping	Raw data → dense vector space
2. Working Memory	Dynamic scratchpad	Active goal, hypotheses, constraints
3. Internal Critic	Value function	Evaluates outputs before execution (quality + safety)
4. Action	Response delivery	Approved output delivered to the user
5. Consolidation	Durable memory	Successful strategies stored for future retrieval

Three-Tier Memory

Episodic Store — append-only memory of past tasks and successes (retrieval-augmented)
Procedural Skill Library — reusable reasoning patterns compiled into adapters
Core Value Axioms — frozen safety rules preventing drift as the agent learns

Self-correction + Memory + Safety

The Internal Critic evaluates every draft response against core value axioms before delivery. This enables self-correction, memory-augmented reasoning (RAG), and continuous learning without catastrophic forgetting.

Built by Eyght. Project Janus — a cognitive architecture, not just a model.

Available on FriendliAI

Dedicated Endpoints

Run this model inference on single tenant GPU with unmatched speed and reliability at scale.

Learn more

Container

Run this model inference with full control and performance in your environment.

Learn more

Model Details

Model Provider

Eyght

Model Tree

Base

Qwen/Qwen2.5-7B-Instruct

Adapter

this model

Input Modalities

Text

Output Modalities