Thox-ai/ thoxglobal-7b:Q4_K_M

7 1 month ago

THOX.ai multilingual assistant. Qwen2.5-7B QLoRA fine-tune, Q4_K_M, answers in the user's language. 

ollama run Thox-ai/thoxglobal-7b:Q4_K_M

Details

1 month ago

b49df274f00b · 4.7GB ·

model
qwen2
·
7.62B
·
Q4_K_M
license
Base model: Qwen2.5-7B (Apache-2.0). THOX LoRA delta and packaging: Apache-2.0. Published by THOX.AI
license
Base model governed by its own license: Apache-2.0. THOX fine-tune, Modelfile, and packaging: (c) 20
template
{{ if .System }}<|im_start|>system {{ .System }}<|im_end|> {{ end }}<|im_start|>user {{ .Prompt }}<|
system
You are ThoxGlobal-7B, the THOX.ai multilingual assistant (Qwen 2.5 base). Answer in the user's lang
params
{ "num_ctx": 8192, "stop": [ "<|im_end|>", "<|endoftext|>" ], "tempe

Readme

ThoxGlobal-7B

THOX.ai multilingual assistant. Qwen2.5-7B QLoRA fine-tune, Q4_K_M, answers in the user’s language.

Tags: tools multilingual qwen2 7b q4_k_m

At a glance

  • Base model: Qwen/Qwen2.5-7B-Instruct
  • Fine-tune: QLoRA supervised fine-tune (THOX thoxllm-factory pipeline)
  • Architecture: qwen2
  • Parameters: 7.6B
  • Quantization: Q4_K_M (4.7 GB)
  • Native context: 32768 tokens (default runtime num_ctx 8192)
  • Prompt template: ChatML (<|im_start|> / <|im_end|>)
  • License: Qwen2.5 base: Apache-2.0. THOX fine-tune, Modelfile, and packaging: © 2026 THOX.ai LLC — see NOTICE.md.
  • Publisher: THOX.AI LLC

Applications

Launch this model directly into a supported tool:

ollama launch claude   --model thox-ai/thoxglobal-7b:Q4_K_M   # Claude Code
ollama launch chatgpt  --model thox-ai/thoxglobal-7b:Q4_K_M   # Codex App
ollama launch openclaw --model thox-ai/thoxglobal-7b:Q4_K_M   # OpenClaw
ollama launch hermes   --model thox-ai/thoxglobal-7b:Q4_K_M   # Hermes
ollama launch codex    --model thox-ai/thoxglobal-7b:Q4_K_M   # Codex
ollama launch opencode --model thox-ai/thoxglobal-7b:Q4_K_M   # OpenCode

Quick start

ollama pull thox-ai/thoxglobal-7b:Q4_K_M
ollama run thox-ai/thoxglobal-7b:Q4_K_M "your prompt"

Intended use

Multilingual chat and drafting. Replies in the user’s language; precise, concise, tool-capable.

The model ships with the THOX factuality system prompt: it answers company questions only from verified facts and declines to invent non-public details (headquarters, headcount, funding, founding date).

Not intended for

Legal or medical advice; unattended long-horizon tasks.

Notes

  • Context: the GGUF supports up to 32768 tokens; the published default num_ctx is 8192 for predictable memory on edge devices. Override at runtime with --num-ctx or a client parameter.
  • On-device: built for local execution on THOX hardware and desktops. Your AI. Your Data. Your Rules.

Links

Published by THOX.AI LLC from the thoxllm-factory pipeline.