This is an experimental merge of models RedPajama-INCITE-Chat-3B-V1 and RedPajama-INCITE-Instruct-3B-V1.
This model is adaptive to prompt templates, but this template is recommended:
HUMAN: {prompt}
ASSISTANT:
Feel free to change HUMAN or ASSISTANT. It will not change much.
GGML versions here (Note that this is only compatible with koboldcpp).
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 39.23 |
| ARC (25-shot) | 42.58 |
| HellaSwag (10-shot) | 67.48 |
| MMLU (5-shot) | 25.99 |
| TruthfulQA (0-shot) | 33.62 |
| Winogrande (5-shot) | 64.8 |
| GSM8K (5-shot) | 0.91 |
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 39.23 |
| AI2 Reasoning Challenge (25-Shot) | 42.58 |
| HellaSwag (10-Shot) | 67.48 |
| MMLU (5-Shot) | 25.99 |
| TruthfulQA (0-shot) | 33.62 |
| Winogrande (5-shot) | 64.80 |
| GSM8k (5-shot) | 0.91 |
- Downloads last month
- 942
This model does not have enough activity to be deployed to Inference API (serverless) yet.
Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Datasets used to train acrastt/RedPajama-INCITE-Chat-Instruct-3B-V1
Spaces using acrastt/RedPajama-INCITE-Chat-Instruct-3B-V1 23
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard42.580
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard67.480
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard25.990
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard33.620
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard64.800
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard0.910