Vian Ai Shop Small
Developed by: Glavinnguyen Model type: Text Generation / Multimodal (Text/Vision/Audio Interface) Language(s) (NLP): Vietnamese (Primary), English (Primary) License: mit Finetuned from model: google/gemma-3-4b-it
Fuente del modelo
- Tipo
- Other
- Tamaño del archivo
- 73,5 MB–8,01 GB
- Modelo base
- google/gemma-3-4b-it
- Versión
- vRevision ed9f664
- Actualizado
- 22 oct 2026
- Publicador
- @GalvinNguyen
- Descargas
- 5
- Parámetros
- 4B
- Tareas
- Generación de texto
- Fuente
- Hugging Face
Descripción de la fuente
Model Card for Vian ai Fine-Tuned
Model Details
Model Description
- Developed by: Glavinnguyen
- Model type: Text Generation / Multimodal (Text/Vision/Audio Interface)
- Language(s) (NLP): Vietnamese (Primary), English (Primary)
- License: mit
- Finetuned from model: google/gemma-3-4b-it
Model Sources
- Repository: [More Information Needed]
Uses
Direct Use
This model is specifically fine-tuned on a small-scale dataset to standardize response formatting, style, and structure. It natively retains the advanced deep reasoning mechanisms (<|think|>) and inherent logical problem-solving capabilities of the base Gemma 3 architecture.
Out-of-Scope Use
The model should not be used for tasks requiring extensive broad-domain knowledge expansion outside the scope of the training dataset without human supervision. The fine-tuning process was focused on structuring behavioral output rather than massive knowledge injection.
Training Details
Training Data
The model was trained on a highly curated, high-quality alignment dataset.
Training Procedure
Training was conducted utilizing the QLoRA (Quantized Low-Rank Adaptation) method to minimize hardware resource consumption while aggressively preserving the base model's pre-trained weights.
Training Hyperparameters
The hyperparameters were carefully optimized for an ultra-small dataset to prevent catastrophic overfitting and achieve an ideal convergence point:
- Training regime: QLoRA (FP16/BF16 Mixed Precision)
- Learning Rate: 2e-4 to 1e-5 (Low Learning Rate)
- Per Device Train Batch Size: 1
- Gradient Accumulation Steps: 32 (Global Batch Size = 32)
- Number of Train Epochs: 1 to 2
- Max Length: 256 - 2048 tokens (Allocated to safeguard the generation of
<|think|>tokens) - Optimizer: AdamW
- LR Scheduler Type: Cosine / Constant
- LoRA Rank (r): 4 or 8
- LoRA Alpha ($\alpha$): 8 or 16
Speeds, Sizes, Times
- Final Training Loss:
0.69(The sweet spot convergence for low-sample fine-tuning—balancing structural alignment with base intelligence retention). - VRAM Consumption: ~2.2 GB (When running on a 4-bit Q4_0 execution profile).
Technical Specifications
Model Architecture and Objective
Built upon Google's next-generation Gemma 3 architecture, featuring integrated Quantization Aware Training (QAT) and an intrinsic multi-token prediction (MTP) engine. The model leverages an internal step-by-step reasoning loop before routing structural text outputs via designated generation tags.
Model Card Contact
- Contact: Glavinnguyen
Framework versions
- PEFT 0.19.1
Fuentes
1 fuente
Fuentes
1 fuenteVerificado 27 sept
Artefactos del modelo
2 artefactos
Artefactos del modelo
2 artefactosExtractos de fuentes
3 extractos
Extractos de fuentes
3 extractos--- base_model: google/gemma-3-4b-it library_name: peft pipeline_tag: text-generation tags: - base_model:adapter:google/gemma-3-4b-it - lora - transformers - marketing - support - ai - shop - sale - 4b - gemma license: mit language: - en - vi --- # Model Card for Vian ai Fine-Tuned ## Model Details ### Model Description - **Developed by:** Glavinnguyen - **Model type:** Text Generation / Multimodal (Text/Vision/Audio Interface) - **Language(s) (NLP):** Vietnamese (Primary), English (Primary) - **License:** mit - **Finetuned from model:** google/gemma-3-4b-it ### Model Sources - **Repository:** [More Information Needed] ## Uses ### Direct Use This model is specifically fine-tuned on a small-scale dataset to standardize response formatting, style, and structure. It natively retains the advanced deep reasoning mechanisms (`<|think|>`) and inherent logical problem-solving capabilities of the base Gemma 3 architecture. ### Out-of-Scope Use The model should not be used for tasks requiring extensive broad-domain knowledge expansion outside the scope of the training dataset without human supervision. The fine-tuning process was focused on structuring behavioral output rather than massive knowledge injection. ## Training Details ### Training Data The model was trained on a highly curated, high-quality alignment dataset. ### Training Procedure Training was conducted utilizing the **QLoRA (Quantized Low-Rank Adaptation)** method to minimize hardware resource consumption while aggressively preserving the base model's pre-trained weights. #### Training Hyperparameters The hyperparameters were carefully optimized for an ultra-small dataset to prevent catastrophic overfitting and achieve an ideal convergence point: - **Training regime:** QLoRA (FP16/BF16 Mixed Precision) - **Learning Rate:** 2e-4 to 1e-5 (Low Learning Rate) - **Per Device Train Batch Size:** 1 - **Gradient Accumulation Steps:** 32 (Global Batch Size = 32) - **Number of Train Epochs:** 1 to 2 - **Max Length:** 256 - 2048 tokens (Allocated to safeguard the generation of `<|think|>` tokens) - **Optimizer:** AdamW - **LR Scheduler Type:** Cosine / Constant - **LoRA Rank (r):** 4 or 8 - **LoRA Alpha ($\alpha$):** 8 or 16 #### Speeds, Sizes, Times - **Final Training Loss:** `0.69` (The sweet spot convergence for low-sample fine-tuning—balancing structural alignment with base intelligence...
Source context: 5 downloads · 0 likes · Pipeline text-generation · Library peft · Repo GalvinNguyen/vian_ai_shop_small
Source context: 762 downloads · 0 likes · Pipeline text-generation · Library peft · Repo GalvinNguyen/vian_ai_shop_small
