Upload Full float16 model files

README.md

@@ -1,121 +1,48 @@
 ---
-license: apache-2.0
-language: en
+base_model:
+- openbmb/MiniCPM5-1B
+- GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking
+library_name: transformers
 tags:
-  - mergekit
-  - slerp
-  - minicpm
-  - zoder
-  - gguf
-  - ollama
-created_by: Komandan Nasa
----
-
-# 🤖 Zoder 1.0-1B
+- mergekit
+- merge
 
-Custom SLERP-merged AI assistant by **Komandan Nasa** (Bakso Bangi Pak Romdani, Kediri, Indonesia).
-
-## Features
+---
+# zoder-merged
 
-- ✅ Chat Template & System Prompt configured
-- ✅ GGUF Q4_K_M quantized version (~700MB)
-- ✅ Gradio Web UI included
-- ✅ Ollama compatible with Modelfile
-- ✅ Optimized for Termux deployment
-- ✅ 100% benchmark pass rate (8/8 tests)
+This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
 
 ## Merge Details
+### Merge Method
+
+This model was merged using the [SLERP](https://en.wikipedia.org/wiki/Slerp) merge method.
+
+### Models Merged
+
+The following models were included in the merge:
+* [openbmb/MiniCPM5-1B](https://huggingface.co/openbmb/MiniCPM5-1B)
+* [GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking](https://huggingface.co/GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking)
+
+### Configuration
+
+The following YAML configuration was used to produce this model:
+
+```yaml
+slices:
+  - sources:
+      - model: openbmb/MiniCPM5-1B
+        layer_range: [0, 24]
+      - model: GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking
+        layer_range: [0, 24]
+merge_method: slerp
+base_model: openbmb/MiniCPM5-1B
+parameters:
+  t:
+    - filter: self_attn
+      value: [0, 0.5, 0.3, 0.7, 1]
+    - filter: mlp
+      value: [1, 0.5, 0.7, 0.3, 0]
+    - value: 0.5
+dtype: float16
 
-| Parameter | Value |
-|-----------|-------|
-| Base Model A | openbmb/MiniCPM5-1B |
-| Base Model B | GnLOLot/MiniCPM5-1B-Claude-Opus-Fable5-V2-Thinking |
-| Method | SLERP |
-| T-value (self_attn) | [0, 0.5, 0.3, 0.7, 1] |
-| T-value (mlp) | [1, 0.5, 0.7, 0.3, 0] |
-| T-value (default) | 0.5 |
-| Dtype | float16 |
-| Layers | 24 |
-
-## Files
-
-| File | Size | Description |
-|------|------|-------------|
-| `model.safetensors` | ~2GB | Full float16 model |
-| `Zoder1.0-1B-Q4_K_M.gguf` | ~700MB | Quantized for mobile/Termux/Ollama |
-| `zoder_gradio.py` | ~3KB | Web UI script |
-| `Modelfile` | ~1KB | Ollama configuration |
-| `install_zoder_ollama.sh` | ~2KB | Termux installer script |
-
-## Quick Start with Ollama
-
-```bash
-# Pull from HuggingFace (after creating model locally)
-ollama create zoder -f Modelfile
-
-# Or pull directly (if published to Ollama library)
-ollama run zoder
-```
-
-## Usage
-
-### Transformers (Full Model)
-
-```python
-from transformers import AutoTokenizer, AutoModelForCausalLM
-
-tokenizer = AutoTokenizer.from_pretrained("Nasaawakening/Zoder1.0-1B", trust_remote_code=True)
-model = AutoModelForCausalLM.from_pretrained("Nasaawakening/Zoder1.0-1B", trust_remote_code=True, torch_dtype="float16", device_map="auto")
-
-messages = [
-    {"role": "system", "content": "You are Zoder, a helpful AI assistant."},
-    {"role": "user", "content": "Hello!"}
-]
-
-text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
-inputs = tokenizer(text, return_tensors="pt").to(model.device)
-outputs = model.generate(**inputs, max_new_tokens=256)
-print(tokenizer.decode(outputs[0], skip_special_tokens=True))
-```
-
-### GGUF (llama.cpp / Termux)
-
-```bash
-huggingface-cli download Nasaawakening/Zoder1.0-1B Zoder1.0-1B-Q4_K_M.gguf --local-dir ~/zoder
-llama-server -m ~/zoder/Zoder1.0-1B-Q4_K_M.gguf --port 8080
 ```
-
-### Gradio Web UI
-
-```bash
-pip install gradio transformers torch
-python zoder_gradio.py
-```
-
-### Termux Ollama Deployment
-
-```bash
-# Download and run installer
-bash install_zoder_ollama.sh
-
-# Run Zoder
-proot-distro login ubuntu -- ollama run zoder
-```
-
-## Benchmark Results
-
-| Test | Status |
-|------|--------|
-| Basic Conversation | ✅ PASS |
-| Logical Reasoning | ✅ PASS |
-| Code Generation (Python) | ✅ PASS |
-| Creative Writing | ✅ PASS |
-| Instruction Following | ✅ PASS |
-| Multi-turn Memory | ✅ PASS |
-| Math Problem | ✅ PASS |
-| Termux Knowledge | ✅ PASS |
-| **Pass Rate** | **100% (8/8)** |
-
-## Built With ❤️ in Kediri, Indonesia
-
-Created by Komandan Nasa from Bakso Bangi Pak Romdani.