Model Type | |
Use Cases |
Areas: | |
Applications: | Legal document analysis, Case law review |
|
Primary Use Cases: | Text generation specific to the legal domain |
|
Considerations: | Model's accuracy may vary based on input complexity. |
|
|
Additional Notes | Supports various precision levels for different hardware efficiencies. |
|
Supported Languages | |
Training Details |
Data Sources: | Open-Orca/OpenOrca, GAIR/lima, WizardLM/WizardLM_evol_instruct_V2_196k, EleutherAI/pile |
|
Methodology: | Continued pre-training on domain-specific corpora using a reading comprehension approach to enhance prompting ability. |
|
Context Length: | |
Model Architecture: | Based on the LLaMA-1 architecture |
|
|
Input Output |
Input Format: | [INST] <> {system_message} <> {prompt} [/INST] |
|
Accepted Modalities: | |
Output Format: | |
|