alexgusevski/Einstein-v6.1-Llama3-8B-mlx-3Bit
The Model alexgusevski/Einstein-v6.1-Llama3-8B-mlx-3Bit was converted to MLX format from Weyaxi/Einstein-v6.1-Llama3-8B using mlx-lm version 0.29.1.
Use with mlx
pip install mlx-lm
from mlx_lm import load, generate
model, tokenizer = load("alexgusevski/Einstein-v6.1-Llama3-8B-mlx-3Bit")
prompt="hello"
if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
messages = [{"role": "user", "content": prompt}]
prompt = tokenizer.apply_chat_template(
messages, tokenize=False, add_generation_prompt=True
)
response = generate(model, tokenizer, prompt=prompt, verbose=True)
- Downloads last month
- 7
Model size
1B params
Tensor type
BF16
·
U32 ·
Hardware compatibility
Log In to add your hardware
3-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for alexgusevski/Einstein-v6.1-Llama3-8B-mlx-3Bit
Base model
meta-llama/Meta-Llama-3-8B
Finetuned
Weyaxi/Einstein-v6.1-Llama3-8B Datasets used to train alexgusevski/Einstein-v6.1-Llama3-8B-mlx-3Bit
Evaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard62.460
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard82.410
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard66.190
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard55.100
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard79.320
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard66.110
- strict accuracy on IFEval (0-Shot)Open LLM Leaderboard45.680
- normalized accuracy on BBH (3-Shot)Open LLM Leaderboard29.380
- exact match on MATH Lvl 5 (4-Shot)Open LLM Leaderboard5.740
- acc_norm on GPQA (0-shot)Open LLM Leaderboard4.250
- acc_norm on MuSR (0-shot)Open LLM Leaderboard11.230
- accuracy on MMLU-PRO (5-shot)test set Open LLM Leaderboard23.680