Deepspeed Chat Step2 Model Opt350m is an open-source language model by zen-E. Features: 350m LLM, VRAM: 0.7GB, Context: 2K, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.06.
Deepspeed Chat Step2 Model Opt350m Parameters and Internals
Model Type Reward Model, Fine-tuned Model
Use Cases
Areas: Research, Commercial Applications
Additional Notes Model is highly sensitive to effective batch size which impacts performance.
Supported Languages Training Details
Data Sources: Dahoas/rm-static, Dahoas/full-hh-rlhf, Dahoas/synthetic-instruct-gptj-pairwise, yitingxie/rlhf-reward-datasets, openai/webgpt_comparisons, stanfordnlp/SHP
Methodology: Fine-tuning with several datasets across different stages: SFT, reward modeling, and RLHF
Hardware Used:
Input Output
Input Format: Text input for prompts with specific padding requirements.
Accepted Modalities: Output Format: Reward values for input prompts
Performance Tips: Consider effective batch size adjustments for better sensitivity handling.
Rank the Deepspeed Chat Step2 Model Opt350m 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 Deepspeed Chat Step2 Model Opt350m
Note: green Score (e.g. "73.2 ") means that the model is better than zen-E/deepspeed-chat-step2-model-opt350m .
Expand