updates

config/preload-models-30b-superhot.yaml

@@ -0,0 +1,47 @@
+- id: huggingface@thebloke/gplatty-30b-superhot-8k-ggml/gplatty-30b-superhot-8k.ggmlv3.q2_k.bin
+  name: "gpt-4"
+  overrides: 
+    context_size: 8192
+    mmap: true
+    f16: true
+    mirostat: 2
+    mirostat_tau: 5.0
+    mirostat_eta: 0.1
+    parameters: 
+      temperature: 0.1
+      top_k: 40
+      top_p: 0.95
+      rope_freq_scale: 0.25
+- id: model-gallery@stablediffusion
+- id: model-gallery@voice-en-us-kathleen-low
+- url: github:go-skynet/model-gallery/base.yaml
+  name: all-MiniLM-L6-v2
+  overrides:
+    embeddings: true
+    backend: huggingface-embeddings
+    parameters:
+      model: all-MiniLM-L6-v2
+- id: huggingface@thebloke/gplatty-30b-superhot-8k-ggml/gplatty-30b-superhot-8k.ggmlv3.q2_k.bin
+  name: functions
+  overrides:
+    context_size: 8192
+    mirostat: 2
+    mirostat_tau: 5.0
+    mirostat_eta: 0.1
+    template:
+      chat: ""
+      completion: ""
+    roles:
+      assistant: "ASSISTANT:"
+      system: "SYSTEM:"
+      assistant_function_call: "FUNCTION_CALL:"
+      function: "FUNCTION CALL RESULT:"
+    parameters:
+      temperature: 0.1
+      top_k: 40
+      top_p: 0.95
+      rope_freq_scale: 0.25
+    function:
+      disable_no_action: true
+    mmap: true
+    f16: true

config/preload-models-30b.yaml

@@ -0,0 +1,45 @@
+- id: huggingface@thebloke/gplatty-30b-ggml/gplatty-30b.ggmlv3.q2_k.bin
+  name: "gpt-4"
+  overrides: 
+    context_size: 4096
+    mmap: true
+    f16: true
+    mirostat: 2
+    mirostat_tau: 5.0
+    mirostat_eta: 0.1
+    parameters: 
+      temperature: 0.1
+      top_k: 40
+      top_p: 0.95
+- id: model-gallery@stablediffusion
+- id: model-gallery@voice-en-us-kathleen-low
+- url: github:go-skynet/model-gallery/base.yaml
+  name: all-MiniLM-L6-v2
+  overrides:
+    embeddings: true
+    backend: huggingface-embeddings
+    parameters:
+      model: all-MiniLM-L6-v2
+- id: huggingface@thebloke/gplatty-30b-ggml/gplatty-30b.ggmlv3.q2_k.bin
+  name: functions
+  overrides:
+    context_size: 4096
+    mirostat: 2
+    mirostat_tau: 5.0
+    mirostat_eta: 0.1
+    template:
+      chat: ""
+      completion: ""
+    roles:
+      assistant: "ASSISTANT:"
+      system: "SYSTEM:"
+      assistant_function_call: "FUNCTION_CALL:"
+      function: "FUNCTION CALL RESULT:"
+    parameters:
+      temperature: 0.1
+      top_k: 40
+      top_p: 0.95
+    function:
+      disable_no_action: true
+    mmap: true
+    f16: true

config/preload-models.yaml

@@ -0,0 +1,45 @@
+- id: huggingface@thebloke/wizardlm-13b-v1.0-uncensored-ggml/wizardlm-13b-v1.0-uncensored.ggmlv3.q4_k_m.bin
+  name: "gpt-4"
+  overrides: 
+    context_size: 2048
+    mmap: true
+    f16: true
+    mirostat: 2
+    mirostat_tau: 5.0
+    mirostat_eta: 0.1
+    parameters: 
+      temperature: 0.1
+      top_k: 40
+      top_p: 0.95
+- id: model-gallery@stablediffusion
+- id: model-gallery@voice-en-us-kathleen-low
+- url: github:go-skynet/model-gallery/base.yaml
+  name: all-MiniLM-L6-v2
+  overrides:
+    embeddings: true
+    backend: huggingface-embeddings
+    parameters:
+      model: all-MiniLM-L6-v2
+- id: huggingface@thebloke/wizardlm-13b-v1.0-uncensored-ggml/wizardlm-13b-v1.0-uncensored.ggmlv3.q4_0.bin
+  name: functions
+  overrides:
+    context_size: 2048
+    mirostat: 2
+    mirostat_tau: 5.0
+    mirostat_eta: 0.1
+    template:
+      chat: ""
+      completion: ""
+    roles:
+      assistant: "ASSISTANT:"
+      system: "SYSTEM:"
+      assistant_function_call: "FUNCTION_CALL:"
+      function: "FUNCTION CALL RESULT:"
+    parameters:
+      temperature: 0.1
+      top_k: 40
+      top_p: 0.95
+    function:
+      disable_no_action: true
+    mmap: true
+    f16: true