Poro 34B GPTQ is an open-source language model by TheBloke. Features: 34b LLM, VRAM: 20.3GB, License: apache-2.0, Quantized, LLM Explorer Score: 0.13, Arc: 47, HellaSwag: 73.8, MMLU: 32.5, GSM8K: 5.1.
Poro 34B GPTQ Parameters and Internals
Model Type bloom, decoder-only transformer
Use Cases
Areas: Limitations: No meaningful proficiency in languages other than English, Finnish, and code
Considerations: Poro is a release of a partially trained model, and special care should be taken when using any output.
Additional Notes Poro is a research checkpoint, training has not been completed.
Supported Languages English (fluent), Finnish (fluent)
Training Details
Data Sources: cerebras/SlimPajama-627B, bigcode/starcoderdata, mc4, allenai/dolma
Data Volume: 1 trillion tokens (500 billion as of this release)
Methodology: BLOOM architecture with ALiBi embeddings for context length extrapolation
Context Length: Hardware Used: LUMI supercomputer using 512 AMD MI250X GPUs
Model Architecture: Generative pretrained transformer using a BLOOM architecture
Rank the Poro 34B GPTQ capabilities
Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.
Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation
Best Alternatives to Poro 34B GPTQ
Expand