TinyLlama 1.1B Chat V0.6 is an open-source language model by TinyLlama. Features: 1.1b LLM, VRAM: 0.6GB, Context: 2K, License: apache-2.0, Quantized, LLM Explorer Score: 0.17, Arc: 31.7, HellaSwag: 55.8, MMLU: 26, GSM8K: 2.1.
TinyLlama 1.1B Chat V0.6 Parameters and Internals
| Model Type | |
| Use Cases |
| Areas: | | chatbot applications, text generation |
|
| Primary Use Cases: | | interactive chatbots, text completion |
|
|
| Additional Notes | | This model is compact and designed for efficient computation and memory footprint. |
|
| Supported Languages | |
| Training Details |
| Data Sources: | | cerebras/SlimPajama-627B, bigcode/starcoderdata, OpenAssistant/oasst_top1_2023-08-25 |
|
| Data Volume: | |
| Methodology: | | Pretraining with optimization, same architecture and tokenizer as Llama 2, further aligned with TRL's DPOTrainer on UltraFeedback dataset |
|
| Training Time: | |
| Hardware Used: | |
| Model Architecture: | |
|
| Input Output |
| Input Format: | | JSON-like structured input for messages |
|
| Accepted Modalities: | |
| Output Format: | |
| Performance Tips: | | Use proper transformers and accelerate installation for optimal performance |
|
|
Best Alternatives to TinyLlama 1.1B Chat V0.6