Upload README.md with huggingface_hub

README.md

@@ -1,43 +1,70 @@
-# AuraMind Auramind 270
+---
+license: mit
+base_model: google/gemma-2-270m
+tags:
+- conversational-ai
+- mental-health
+- productivity
+- smartphone
+- mobile-ai
+- therapy
+- assistant
+- gemma
+library_name: transformers
+pipeline_tag: text-generation
+model-index:
+- name: zail-ai/auramind-270m
+  results:
+  - task:
+      type: text-generation
+      name: Conversational AI
+    dataset:
+      type: zail-ai/auramind
+      name: AuraMind Dataset
+    metrics:
+    - type: inference_speed
+      value: 100-300ms on modern smartphones
+      name: Inference Speed
+    - type: memory_usage
+      value: ~680MB RAM
+      name: Memory Usage
+    - type: parameters
+      value: 270M
+      name: Model Parameters
+---
 
-A warm and caring AI companion model fine-tuned for emotional support and personal connection.
+# Auramind-270M - 270M Parameters
 
-## Model Details
-- **Base Model**: Gemma 3 270M (Released August 2025)
-- **Parameters**: 268,102,656
-- **Quantization**: INT4
-- **Context Length**: 2,048 tokens
-- **Vocabulary**: 256k tokens (large vocabulary for rare tokens)
+Full-featured smartphone deployment with balanced performance and capabilities
 
-## Architecture Details
-- **Total Parameters**: 270M (170M embedding + 100M transformer blocks)
-- **Training Data**: 6 trillion tokens
-- **Knowledge Cutoff**: August 2024
+## Specifications
 
-## Training Details
-- **Dataset**: AuraMind Companion Dataset
-- **Training Steps**: 1,000
-- **Final Loss**: 126.1678
-- **Device**: MPS
+- **Parameters**: 270M
+- **Base Model**: google/gemma-2-270m
+- **Memory Usage**: ~680MB RAM
+- **Quantization**: INT4 optimized
+- **Inference Speed**: 100-300ms on modern smartphones
+
+## Mobile Deployment
+
+This variant is specifically optimized for:
+- **Target Devices**: Premium smartphones
+- **Memory Requirements**: ~680MB RAM
+- **Performance**: 100-300ms on modern smartphones
 
 ## Usage
+
 ```python
 from transformers import AutoTokenizer, AutoModelForCausalLM
 
-tokenizer = AutoTokenizer.from_pretrained("models/auramind_270")
-model = AutoModelForCausalLM.from_pretrained("models/auramind_270")
-
-# Generate companion response
-prompt = "<|system|>\nYou are AuraMind, a warm and caring AI companion. You form genuine emotional bonds with users through empathy, understanding, and personal connection. You provide intimate emotional support while maintaining healthy boundaries. Focus on creating deep, meaningful relationships through authentic conversation.\n\n<|user|>\nI'm feeling lonely today.\n\n<|assistant|>\n"
-inputs = tokenizer(prompt, return_tensors="pt")
-outputs = model.generate(**inputs, max_length=200, temperature=0.7, do_sample=True)
-response = tokenizer.decode(outputs[0], skip_special_tokens=False)
+# Load this specific variant
+tokenizer = AutoTokenizer.from_pretrained("zail-ai/auramind-270m")
+model = AutoModelForCausalLM.from_pretrained(
+    "zail-ai/auramind-270m",
+    torch_dtype=torch.float16,
+    device_map="auto",
+    low_cpu_mem_usage=True
+)
 ```
 
-## Capabilities
-- Emotional support and companionship
-- Personal conversation and bonding
-- Empathetic responses
-- Intimate but appropriate interactions
-
-**Note**: This is a companion AI, not a replacement for professional therapy.
+Refer to the main [AuraMind repository](https://huggingface.co/zail-ai/Auramind) for complete documentation.