Trained Models 🏋️
Collection
They may be small, but they're training like giants!
•
9 items
•
Updated
•
20
A Llama Chat Model of 68M Parameters
- Base model: JackFram/llama-68m
- Datasets:
- Availability in other ML formats:
Recommended Prompt Format
<|im_start|>system
{system_message}<|im_end|>
<|im_start|>user
{user_message}<|im_end|>
<|im_start|>assistant
Recommended Inference Parameters
penalty_alpha: 0.5
top_k: 4
Open LLM Leaderboard Evaluation Results
Detailed results can be found here
| Metric | Value |
|---|---|
| Avg. | 29.72 |
| AI2 Reasoning Challenge (25-Shot) | 23.29 |
| HellaSwag (10-Shot) | 28.27 |
| MMLU (5-Shot) | 25.18 |
| TruthfulQA (0-shot) | 47.27 |
| Winogrande (5-shot) | 54.30 |
| GSM8k (5-shot) | 0.00 |
- Downloads last month
- 787
Model tree for Felladrin/Llama-68M-Chat-v1
Datasets used to train Felladrin/Llama-68M-Chat-v1
Spaces using Felladrin/Llama-68M-Chat-v1 3
Collection including Felladrin/Llama-68M-Chat-v1
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard23.290
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard28.270
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard25.180
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard47.270
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard54.300
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard0.000