Feedback and support: TensorBlock's Twitter/X, Telegram Group and Discord server
Eurdem/megatron_1.1_MoE_2x7B - GGUF
This repo contains GGUF format model files for Eurdem/megatron_1.1_MoE_2x7B.
The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4242.
Prompt template
<s>GPT4 Correct System: {system_prompt}<|end_of_turn|>GPT4 Correct User: {prompt}<|end_of_turn|>GPT4 Correct Assistant:
Model file specification
Filename | Quant type | File Size | Description |
---|---|---|---|
megatron_1.1_MoE_2x7B-Q2_K.gguf | Q2_K | 4.761 GB | smallest, significant quality loss - not recommended for most purposes |
megatron_1.1_MoE_2x7B-Q3_K_S.gguf | Q3_K_S | 5.588 GB | very small, high quality loss |
megatron_1.1_MoE_2x7B-Q3_K_M.gguf | Q3_K_M | 6.207 GB | very small, high quality loss |
megatron_1.1_MoE_2x7B-Q3_K_L.gguf | Q3_K_L | 6.730 GB | small, substantial quality loss |
megatron_1.1_MoE_2x7B-Q4_0.gguf | Q4_0 | 7.281 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
megatron_1.1_MoE_2x7B-Q4_K_S.gguf | Q4_K_S | 7.342 GB | small, greater quality loss |
megatron_1.1_MoE_2x7B-Q4_K_M.gguf | Q4_K_M | 7.783 GB | medium, balanced quality - recommended |
megatron_1.1_MoE_2x7B-Q5_0.gguf | Q5_0 | 8.874 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
megatron_1.1_MoE_2x7B-Q5_K_S.gguf | Q5_K_S | 8.874 GB | large, low quality loss - recommended |
megatron_1.1_MoE_2x7B-Q5_K_M.gguf | Q5_K_M | 9.133 GB | large, very low quality loss - recommended |
megatron_1.1_MoE_2x7B-Q6_K.gguf | Q6_K | 10.567 GB | very large, extremely low quality loss |
megatron_1.1_MoE_2x7B-Q8_0.gguf | Q8_0 | 13.686 GB | very large, extremely low quality loss - not recommended |
Downloading instruction
Command line
Firstly, install Huggingface Client
pip install -U "huggingface_hub[cli]"
Then, downoad the individual model file the a local directory
huggingface-cli download tensorblock/megatron_1.1_MoE_2x7B-GGUF --include "megatron_1.1_MoE_2x7B-Q2_K.gguf" --local-dir MY_LOCAL_DIR
If you wanna download multiple model files with a pattern (e.g., *Q4_K*gguf
), you can try:
huggingface-cli download tensorblock/megatron_1.1_MoE_2x7B-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'
- Downloads last month
- 127
Model tree for tensorblock/megatron_1.1_MoE_2x7B-GGUF
Base model
Eurdem/megatron_1.1_MoE_2x7BEvaluation results
- normalized accuracy on AI2 Reasoning Challenge (25-Shot)test set Open LLM Leaderboard65.530
- normalized accuracy on HellaSwag (10-Shot)validation set Open LLM Leaderboard84.520
- accuracy on MMLU (5-Shot)test set Open LLM Leaderboard65.020
- mc2 on TruthfulQA (0-shot)validation set Open LLM Leaderboard51.580
- accuracy on Winogrande (5-shot)validation set Open LLM Leaderboard81.530
- accuracy on GSM8k (5-shot)test set Open LLM Leaderboard71.490