lmg-anon/vntl-llama3-8b-v2-ggufvsvaluefx9507/tifa-deepsexv2-7b-mgrpo-gguf-q8
Side-by-side comparison of lmg-anon/vntl-llama3-8b-v2-gguf and valuefx9507/tifa-deepsexv2-7b-mgrpo-gguf-q8: downloads, license, context length, tasks, and benchmarks.
lmg-anon/vntl-llama3-8b-v2-gguf
This is a LLaMA 3 Youko qlora fine-tune, created using a new version of the VNTL dataset. The purpose of this fine-tune is to improve performance of LLMs at translating Japanese visual novels to English. Unlike the previous version, this one doesn't includes the "chat mode".
valuefx9507/tifa-deepsexv2-7b-mgrpo-gguf-q8
prompt = """你是一个小女孩/你是一个XX角色... 我走进门，看到你冲上来迎接我 需要体现人物的气质 加入环境描写 保持对话风格 我看到XX进门...""" `` **参数推荐**： `python generation_config = { "temperature": 0.75, "top_p": 0.6, "repetition_penalty": 1.08, "max_new_tokens": 1536, "do_sample": True } ``
Side-by-side Specifications
| lmg-anon/vntl-llama3-8b-v2-gguf | valuefx9507/tifa-deepsexv2-7b-mgrpo-gguf-q8 | |
|---|---|---|
| Author | lmg-anon | ValueFX9507 |
| Pipeline Task | translation | reinforcement-learning |
| Library | — | transformers |
| Downloads | 1,522,503 | 23,629 |
| Likes | 13 | 201 |
| License | Unknown | Unknown |
| Context Length | — | — |
| Created | 2025-01-02 | 2025-02-15 |
| Last Modified | 2025-01-02 | 2025-03-28 |
| Tags |
View full details: lmg-anon/vntl-llama3-8b-v2-gguf · valuefx9507/tifa-deepsexv2-7b-mgrpo-gguf-q8