bigstral-12b-v0.2-32k
`ollama run ehartford/bigstral`
This is Mistral-7B-v0.2 self-interleaved into a larger 12B model using MergeKit. It is intended for further pretraining.
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the passthrough merge method.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
dtype: bfloat16
merge_method: passthrough
slices:
- sources:
- layer_range: [0, 8]
model: alpindale/Mistral-7B-v0.2-hf
- sources:
- layer_range: [4, 12]
model: alpindale/Mistral-7B-v0.2-hf
- sources:
- layer_range: [8, 16]
model: alpindale/Mistral-7B-v0.2-hf
- sources:
- layer_range: [12, 20]
model: alpindale/Mistral-7B-v0.2-hf
- sources:
- layer_range: [16, 24]
model: alpindale/Mistral-7B-v0.2-hf
- sources:
- layer_range: [20, 28]
model: alpindale/Mistral-7B-v0.2-hf
- sources:
- layer_range: [24, 32]
model: alpindale/Mistral-7B-v0.2-hf
- Downloads last month
- 5
Install from pip and serve model
# Install vLLM from pip: pip install vllm# Start the vLLM server: vllm serve "abacusai/bigstral-12b-v0.2-32k"# Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "abacusai/bigstral-12b-v0.2-32k", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'