Upload v1.0_-_Qwen_Image/README.md with huggingface_hub

v1.0_-_Qwen_Image/README.md

@@ -0,0 +1,46 @@
+---
+language:
+- en
+tags:
+- art
+---
+
+# Seizo Watase Style Lora
+
+**Creator**: [tarnished3029](https://civitai.com/user/tarnished3029)
+**Type**: LORA
+**Base Model**: Qwen
+**Version**: v1.0 - Qwen Image
+**Trigger Words**: `Seizo Watase style.`
+
+**Civitai Model ID**: 1601633
+**Civitai Version ID**: 2185088
+
+**Stats (at time of fetch for this version)**:
+*   Downloads: 52
+*   Rating: 0 (0 ratings)
+*   Favorites: N/A
+
+---
+
+## 📄 Description (Parent Model)
+
+Seizo Watase style lora. Pictures of Japanese aesthetic before the Bubble burst of the 1980's. Before the god killing cynicism of the 1990's. Before the plague of cute anime school girls of the 2000's. His artwork depicts an upbeat and optimistic era of Japan, back when the party never stopped and everyone wore the best clothes. Primarily focused on slice of life and romantic scenes. His art stands out from the bleak depressing era of Japan's lost decade. He is best known for his manga "Heart Cocktail" and "Chalk Colored People".
+Dataset features 171 images mostly from the "Heart Cocktail" series. LoRA Rank is 32. Batch size was 4, LR 2e-5, 28 epochs, 6000 steps. Took about 6 hours running on 2 rtx 5000 ada's.
+Qwen model was trained at LR 1e-4, at 4800 steps about 24 epochs.
+The HiDream model captures his early work from Heart Cocktail very well.  I think the Qwen model captures his later style better.
+
+## Civitai Links
+
+*   **[🔗 View This Version on Civitai →](https://civitai.com/models/1601633?modelVersionId=2185088)**
+*   [View Full Model Page →](https://civitai.com/models/1601633)
+*   [View Creator Profile →](https://civitai.com/user/tarnished3029)
+
+---
+
+## File Information
+
+*   **Filename**: `seizo_watase_qwen_lora.safetensors`
+*   **Size**: 562.72 MB
+*   **Hash (AutoV2)**: `B645684194`
+*   **Hash (SHA256)**: `B645684194319E61C3B4EF5E41A6043F09FF77140D0B59EA2D7F71635A04ABBB`