Damysus 2.7B Chat is an open-source language model by prince-canuma. Features: 2.7b LLM, VRAM: 11.1GB, Context: 2K, License: mit, LLM Explorer Score: 0.13, Arc: 59.1, HellaSwag: 74.4, MMLU: 56.3, GSM8K: 50.2.
Damysus 2.7B Chat Parameters and Internals
| Model Type | | Transformer, Instruction-tuned |
|
| Use Cases |
| Areas: | | NLP, Instruction following |
|
| Applications: | | Text generation, Data extraction, Structured outputs, Q&A, Explanations |
|
| Primary Use Cases: | | Answer synthesizer, Summarizer, Query rewriter |
|
| Limitations: | | May generate inaccurate code, Limited code scope, particularly outside Python, Designed primarily for standard English |
|
| Considerations: | | Users are advised to verify outputs for accuracy, especially in technical domains. |
|
|
| Additional Notes | | Optimal performance derived from fine-tuning on smaller, curated datasets. Recommended for use in structured output tasks and conversational queries. |
|
| Supported Languages | |
| Training Details |
| Data Sources: | | Open-Orca/SlimOrca, prince-canuma/TinyOrca |
|
| Methodology: | | Instruction-tuned with additional training to better follow user instructions. Fine-tuned on a subset of Open-Orca data. |
|
| Context Length: | |
| Model Architecture: | |
|
Best Alternatives to Damysus 2.7B Chat
Note: green Score (e.g. "73.2") means that the model is better than prince-canuma/Damysus-2.7B-Chat.