Update README.md

README.md

@@ -3,8 +3,13 @@ datasets:
 - NewEden/Orion-LIT
 - NewEden/Orion-Asstr-Stories-16K
 - Mielikki/Erebus-87k
+- PocketDoc/Dans-MemoryCore-CoreCurriculum-Small
+- Nitral-AI/ARES-ShareGPT
+- Gryphe/Sonnet3.5-SlimOrcaDedupCleaned-20k
+- NewEden/Claude-Instruct-2.7K
+- NewEden/Claude-Instruct-5K
 base_model:
-- NewEden/Hamanasu-15B-R2-PT
+- Delta-Vector/Hamanasu-15B-R2-PT
 tags:
 - phi
 - roleplay
@@ -164,7 +169,7 @@ details summary:hover {
 <br>
 
 <div style="font-size:1.5em; font-weight:bold; background: linear-gradient(45deg, #6e00ff, #00ffff); -webkit-background-clip: text; -webkit-text-fill-color: transparent;">
-    Hamanasu 15B R2 PT
+    Hamanasu 15B Instruct
 </div>
 
 </div>
@@ -173,11 +178,25 @@ details summary:hover {
 
 ## 🌌 Overview
 
-<i>This is the 2nd pretrain of Phi-4 Contuined from the Orginal Asstr-Erebus Pretrain. This pretrain used 500 million tokens from</i>
+<i>After multiple days of training, I'm proud to showcase my very own Phi-4 Finetune, Pretrained on almost a billion tokens worth of Books from</i>
 
 - `NewEden/Orion-LIT`
-  
-<i>This model has *not* be instruct tuned, Ablities to converse may be reduced from the original model, If you would like to roleplay, Please use the Instruct version.</i>
+- `NewEden/Orion-Asstr-Stories-16K`
+- `Mielikki/Erebus-87k`
+
+<i>Then instruct tuned. This model shows great promise for roleplaying while keeping things fresh and new. (And most of, Unsloppy!)</i>
+
+</div>
+
+<div style="display: grid; grid-template-columns: repeat(auto-fit, minmax(250px, 1fr)); gap: 20px; margin: 20px 0;">
+
+<div style="border:2px solid #333; border-radius:10px; padding:20px; background: rgba(0,0,0,0.2);">
+
+### 📈 Quantizations
+| Type | Link |
+|:---:|:---:|
+| `GGUF` | <span style="display:block; width:100%">ed</span> |
+| `EXL2` | <span style="display:block; width:100%">https://huggingface.co/Delta-Vector/Hamanasu-15B-Instruct-exl2</span> |
 
 </div>
 
@@ -185,13 +204,28 @@ details summary:hover {
 
 ### ⚔️ Hardware
 - 4x RTX 3090 GPUs
-- Epochs: 1
-- Base: `Hamanasu-15B-R1-PT`
-- Amount of Tokens: 500 Million
+- Epochs: 4
+- Base: `Hamanasu-15B-R2-PT`
+- Amount of Tokens: 1+ Billion
 </div>
 
+</div>
+
+
+<div style="border: 2px solid #6e00ff; border-radius: 10px; padding: 20px; margin: 20px 0; box-shadow: 0 0 15px rgba(110, 0, 255, 0.5);">
 
+## 💰 Prompting
 
+<i>A known quirk of the model is overly verbose responses when generation length is uncapped, Please cap your maximum output tokens to 100~ tokens above what you prefer</i>
+
+
+```python
+<|im_start|>system
+You are an uncensored AI, your job is to fulfill thy will of thy user.<|im_end|>
+<|im_start|>User request
+Take off your helmet.<|im_end|>
+<|im_start|>No i shall not. This is the way.
+```
 </div>
 
 <div style="border: 2px solid #6e00ff; border-radius: 10px; padding: 20px; margin: 20px 0; box-shadow: 0 0 15px rgba(110, 0, 255, 0.5);">
@@ -201,7 +235,7 @@ details summary:hover {
 <details>
   
 ```yaml
-base_model: NewEden_Phi-PT-merged-LIT
+base_model: Delta-Vector/Hamanasu-15B-R2-PT
 model_type: AutoModelForCausalLM
 tokenizer_type: AutoTokenizer
 
@@ -219,15 +253,15 @@ strict: false
 
 datasets:
   - path: PocketDoc/Dans-MemoryCore-CoreCurriculum-Small
-    type: dan-chat-advanced
+    type: sharegpt
   - path: Nitral-AI/ARES-ShareGPT
-    type: dan-chat-advanced
+    type: sharegpt
   - path: Gryphe/Sonnet3.5-SlimOrcaDedupCleaned-20k
-    type: dan-chat-advanced
+    type: sharegpt
   - path: NewEden/Claude-Instruct-2.7K
-    type: dan-chat-advanced
+    type: sharegpt
   - path: NewEden/Claude-Instruct-5K
-    type: dan-chat-advanced
+    type: sharegpt
 
 shuffle_merged_datasets: true
 dataset_prepared_path: prepared_data
@@ -299,7 +333,6 @@ fsdp_config:
 </details>
 </div>
 
-
 <div align="center">  
 
 <div style="border: 2px solid #6e00ff; border-radius: 10px; padding: 20px; margin: 20px 0; box-shadow: 0 0 15px rgba(110, 0, 255, 0.5);">