DeepCoder with tool calling(MCP) support

tools 14b

Updated 4 months ago

4 months ago

819b50856fb5 · 9.0GB

archqwen2

parameters14.8B

quantizationQ4_K_M

9.0GB

{{- if .Suffix }}<|fim_prefix|>{{ .Prompt }}<|fim_suffix|>{{ .Suffix }}<|fim_middle|> {{- else if .M

1.6kB

This model weights are licensed under the MIT License same as original DeepCoder Preview model. Deep

804B

{ "temperature": 0.6, "top_p": 0.95 }

33B

Readme

This model is a quantized version of the original DeepCoder Preview from Agentica Project.

Key changes from the original include quantization and chat template customization, as follows:

Quantization: It is quantized using imatrix with calibration data mixed from English and Japanese sources. For more details, see this Hugging Face project:
- https://huggingface.co/okamototk/DeepCoder-14B-Preview-imatrix-GGUF
Tool Support: Tool support has been added through chat template customization.