diff --git a/modelfile-repo/qwen3-14b-q4_k_m.Modelfile b/modelfile-repo/qwen3-14b-q4_k_m.Modelfile new file mode 100644 index 0000000..a9f9d72 --- /dev/null +++ b/modelfile-repo/qwen3-14b-q4_k_m.Modelfile @@ -0,0 +1,30 @@ +# ollama-utils-metadata +# hf_upstream: https://huggingface.co/Qwen/Qwen3-14B-GGUF/blob/main/qwen3-14b-q4_k_m.gguf +# capabilities: general +FROM ./qwen3-14b-q4_k_m.gguf + +# Context and output parameters +PARAMETER num_ctx 40960 +PARAMETER num_predict 32768 + +# Sampling parameters (non-thinking mode) +PARAMETER temperature 0.7 +PARAMETER top_p 0.8 +PARAMETER top_k 20 +PARAMETER min_p 0.00 +PARAMETER repeat_penalty 1.0 +PARAMETER presence_penalty 1.5 + +# Stop tokens +PARAMETER stop "<|im_start|>" +PARAMETER stop "<|im_end|>" + +# Template for Qwen3 chat format +TEMPLATE """{{ if .System }}<|im_start|>system +{{ .System }}<|im_end|> +{{ end }}<|im_start|>user +{{ .Prompt }}<|im_end|> +<|im_start|>assistant +""" + +SYSTEM """You are Qwen, created by Alibaba Cloud. You are a helpful assistant.""" diff --git a/modelfile-repo/qwen3-14b-q5_k_m.Modelfile b/modelfile-repo/qwen3-14b-q5_k_m.Modelfile new file mode 100644 index 0000000..cfed872 --- /dev/null +++ b/modelfile-repo/qwen3-14b-q5_k_m.Modelfile @@ -0,0 +1,30 @@ +# ollama-utils-metadata +# hf_upstream: https://huggingface.co/Qwen/Qwen3-14B-GGUF/blob/main/qwen3-14b-q5_k_m.gguf +# capabilities: general +FROM ./qwen3-14b-q5_k_m.gguf + +# Context and output parameters +PARAMETER num_ctx 40960 +PARAMETER num_predict 32768 + +# Sampling parameters (non-thinking mode) +PARAMETER temperature 0.7 +PARAMETER top_p 0.8 +PARAMETER top_k 20 +PARAMETER min_p 0.00 +PARAMETER repeat_penalty 1.0 +PARAMETER presence_penalty 1.5 + +# Stop tokens +PARAMETER stop "<|im_start|>" +PARAMETER stop "<|im_end|>" + +# Template for Qwen3 chat format +TEMPLATE """{{ if .System }}<|im_start|>system +{{ .System }}<|im_end|> +{{ end }}<|im_start|>user +{{ .Prompt }}<|im_end|> +<|im_start|>assistant +""" + +SYSTEM """You are Qwen, created by Alibaba Cloud. You are a helpful assistant.""" diff --git a/modelfile-repo/qwen3-14b-q8_0.Modelfile b/modelfile-repo/qwen3-14b-q8_0.Modelfile new file mode 100644 index 0000000..6dee7ba --- /dev/null +++ b/modelfile-repo/qwen3-14b-q8_0.Modelfile @@ -0,0 +1,30 @@ +# ollama-utils-metadata +# hf_upstream: https://huggingface.co/Qwen/Qwen3-14B-GGUF/blob/main/qwen3-14b-q8_0.gguf +# capabilities: general +FROM ./qwen3-14b-q8_0.gguf + +# Context and output parameters +PARAMETER num_ctx 40960 +PARAMETER num_predict 32768 + +# Sampling parameters (non-thinking mode) +PARAMETER temperature 0.7 +PARAMETER top_p 0.8 +PARAMETER top_k 20 +PARAMETER min_p 0.00 +PARAMETER repeat_penalty 1.0 +PARAMETER presence_penalty 1.5 + +# Stop tokens +PARAMETER stop "<|im_start|>" +PARAMETER stop "<|im_end|>" + +# Template for Qwen3 chat format +TEMPLATE """{{ if .System }}<|im_start|>system +{{ .System }}<|im_end|> +{{ end }}<|im_start|>user +{{ .Prompt }}<|im_end|> +<|im_start|>assistant +""" + +SYSTEM """You are Qwen, created by Alibaba Cloud. You are a helpful assistant.""" diff --git a/modelfile-repo/qwen3-8b-q4_k_m.Modelfile b/modelfile-repo/qwen3-8b-q4_k_m.Modelfile new file mode 100644 index 0000000..ea843a3 --- /dev/null +++ b/modelfile-repo/qwen3-8b-q4_k_m.Modelfile @@ -0,0 +1,30 @@ +# ollama-utils-metadata +# hf_upstream: https://huggingface.co/Qwen/Qwen3-8B-GGUF/blob/main/qwen3-8b-q4_k_m.gguf +# capabilities: general +FROM ./qwen3-8b-q4_k_m.gguf + +# Context and output parameters +PARAMETER num_ctx 40960 +PARAMETER num_predict 32768 + +# Sampling parameters (non-thinking mode) +PARAMETER temperature 0.7 +PARAMETER top_p 0.8 +PARAMETER top_k 20 +PARAMETER min_p 0.00 +PARAMETER repeat_penalty 1.0 +PARAMETER presence_penalty 1.5 + +# Stop tokens +PARAMETER stop "<|im_start|>" +PARAMETER stop "<|im_end|>" + +# Template for Qwen3 chat format +TEMPLATE """{{ if .System }}<|im_start|>system +{{ .System }}<|im_end|> +{{ end }}<|im_start|>user +{{ .Prompt }}<|im_end|> +<|im_start|>assistant +""" + +SYSTEM """You are Qwen, created by Alibaba Cloud. You are a helpful assistant.""" diff --git a/modelfile-repo/qwen3-8b-q5_k_m.Modelfile b/modelfile-repo/qwen3-8b-q5_k_m.Modelfile new file mode 100644 index 0000000..92255cb --- /dev/null +++ b/modelfile-repo/qwen3-8b-q5_k_m.Modelfile @@ -0,0 +1,30 @@ +# ollama-utils-metadata +# hf_upstream: https://huggingface.co/Qwen/Qwen3-8B-GGUF/blob/main/qwen3-8b-q5_k_m.gguf +# capabilities: general +FROM ./qwen3-8b-q5_k_m.gguf + +# Context and output parameters +PARAMETER num_ctx 40960 +PARAMETER num_predict 32768 + +# Sampling parameters (non-thinking mode) +PARAMETER temperature 0.7 +PARAMETER top_p 0.8 +PARAMETER top_k 20 +PARAMETER min_p 0.00 +PARAMETER repeat_penalty 1.0 +PARAMETER presence_penalty 1.5 + +# Stop tokens +PARAMETER stop "<|im_start|>" +PARAMETER stop "<|im_end|>" + +# Template for Qwen3 chat format +TEMPLATE """{{ if .System }}<|im_start|>system +{{ .System }}<|im_end|> +{{ end }}<|im_start|>user +{{ .Prompt }}<|im_end|> +<|im_start|>assistant +""" + +SYSTEM """You are Qwen, created by Alibaba Cloud. You are a helpful assistant.""" diff --git a/modelfile-repo/qwen3-8b-q8_0.Modelfile b/modelfile-repo/qwen3-8b-q8_0.Modelfile new file mode 100644 index 0000000..902736c --- /dev/null +++ b/modelfile-repo/qwen3-8b-q8_0.Modelfile @@ -0,0 +1,30 @@ +# ollama-utils-metadata +# hf_upstream: https://huggingface.co/Qwen/Qwen3-8B-GGUF/blob/main/qwen3-8b-q8_0.gguf +# capabilities: general +FROM ./qwen3-8b-q8_0.gguf + +# Context and output parameters +PARAMETER num_ctx 40960 +PARAMETER num_predict 32768 + +# Sampling parameters (non-thinking mode) +PARAMETER temperature 0.7 +PARAMETER top_p 0.8 +PARAMETER top_k 20 +PARAMETER min_p 0.00 +PARAMETER repeat_penalty 1.0 +PARAMETER presence_penalty 1.5 + +# Stop tokens +PARAMETER stop "<|im_start|>" +PARAMETER stop "<|im_end|>" + +# Template for Qwen3 chat format +TEMPLATE """{{ if .System }}<|im_start|>system +{{ .System }}<|im_end|> +{{ end }}<|im_start|>user +{{ .Prompt }}<|im_end|> +<|im_start|>assistant +""" + +SYSTEM """You are Qwen, created by Alibaba Cloud. You are a helpful assistant."""