You can deploy this model and start earning money today!

nvlz/
mistral-ed-only-v3

Name: mistral-ed-only-v3
Availability: InStock

mistralai/Mistral-7B-Instruct-v0.2

adapter

Views

No views yet

Mistral-7B — EmpatheticDialogues Fine-Tune v3

v2 → v3 Değişiklikleri

Problem	v2	v3
`_comma_` artefaktı	❌	✅ Temizlendi
min_out_words	20 (→ 3K örnek)	15 (→ 4664 örnek)
Anti-pattern filtresi	❌	✅
Kalite skoru filtresi	❌	✅
Sistem promptu	Genel	Anti-pattern yasaklı

Repo Yapısı

Klasör	İçerik
`final/`	Model ağırlıkları (LoRA adapter)
`checkpoints/epoch_N/`	Epoch checkpointları
`eval/`	Metrikler + sample outputs
`logs/`	Eğitim logu

Eğitim Sonuçları


Training Loss	0.7188
Süre	187.9 dk
Train samples	2785
Val samples	934
Platform	Google Colab (Tesla T4)

Yükleme

python

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

base = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2", torch_dtype=torch.float16, device_map="auto")
model = PeftModel.from_pretrained(base, "nvlz/mistral-ed-only-v3", subfolder="final")
tokenizer = AutoTokenizer.from_pretrained("nvlz/mistral-ed-only-v3", subfolder="final")